2015 ARS Supplement
Transcription
2015 ARS Supplement
A Supplement to: PERSPECTIVES ON THE TREATMENT OF HEMATOLOGIC AND SOLID MALIGNANCIES April 2015 | Volume 29 • Supplement No. 1 For ONCOLOGY online, go to ONCOLOGY • APRIL 2015 • VOLUME 29 • SUPPLEMENT NUMBER 1 Integrating High-Tech With High-Touch in Radiation Oncology Program Abstracts From the 97th Annual Meeting of the American Radium Society Kenneth E. Rosenzweig, MD, President Mount Sinai School of Medicine, Department of Radiation Oncology, New York, New York Benjamin Movsas, MD, Chair, Scientific Program Committee Henry Ford Hospital, Department of Radiation Oncology, Detroit, Michigan cancernetwork.com SUPPORTED BY THE AMERICAN RADIUM SOCIETY GRAND HYATT KAUAI KAUAI, HAWAII MAY 2–5, 2015 Committees of the ARS EXECUTIVE COMMIT TEE Kenneth E. Rosenzweig, MD, President John “Drew” Ridge, MD, PhD, President-Elect Ben J. Slotman, MD, PhD, Secretary Quynh-Thu Le, MD, Treasurer Kaled M. Alektiar, MD, Industry Relations Chair Kenneth B. Roberts, MD, Member-at-Large Beth M. Beadle, MD, Member-at-Large Benjamin Movsas, MD, 2015 SPC Chair Elin R. Sigurdson, MD, PhD, Immediate Past President 2014-2015 2014-2015 2014-2015 2014-2015 2014-2015 2013-2015 2014-2016 2014-2015 2014-2015 CONSTITUTION AND BY-LAWS COMMIT TEE Lynn D. Wilson, MD, MPH (Chair)2014-2017 Elin Sigurdson, MD, PhD (Ex officio) Kenneth Russell, MD 2012-2015 2014-2016 Erich Sturgis, MD, MPH EDUCATIOn and website RESOURCES COMMIT TEE Jonathan J. Beitler, MD (Co-Chair) Elin R. Sigurdson, MD, PhD (Ex officio) Roy Decker, MD Jim Wallace, MD Mike Kupferman, MD David Wazer, MD Peter A.S. Johnstone, MD Cynthia Ballenger, MD 2014-2017 2014-2017 2012-2015 2014-2017 2012-2015 2014-2016 2014-2017 INDUSTRY RELATIONS/DEVELOPMENT COMMIT TEE Kaled M. Alektiar, MD (Chair)2014-2015 Elin R. Sigurdson, MD, PhD (Ex officio) Alan Pollack, MD, PhD 2012-2016 Sue S. Yom, MD, PhD 2014-2018 2015-2019 Meena S. Moran, MD JANEWAY LECTURE COMMIT TEE WEBSITE AND PUBLIC RELATIONS COMMIT TEE Jonathan J. Beitler, MD (Chair)2011-2014 Thomas A. Buchholz, MD (Ex officio) David Wazer, MD 2012-2015 2010-2013 John Ward, MD Cynthia Ballenger, MD 2011-2014 SCIENTIFIC PROGRAM COMMIT TEE Benjamin Movsas, MD (Program Chair) Kenneth E. Rosenzweig, MD (President) Matthew C. Abramowitz, MD Beth M. Beadle, MD, PhD Thomas A. Buchholz, MD Todd J. Carpenter, MD Indrin Chetty, PhD Ted DeWeese, MD Mohamed Elshaikh, MD Mary Feng, MD Thomas Galloway, MD Mary K. Hayes, MD Mark A. Henderson, MD Joe Herman, MD Peter A.S. Johnstone, MD Andre Konski, MD, MBA, MA, FACR W. Robert Lee, MD Stella Ling, MD Rahul R. Parikh, MD, PhD Alan Pollack, MD, PhD John "Drew" Ridge, MD, PhD Kenneth Roberts, MD Elin R. Sigurdson, MD, PhD Charles B. Simone, II, MD Jean Wright, MD Sue S. Yom, MD, PhD rePRESENTATIVE TO THE BOARD OF CHANCELLORS OF THE AMERICAN COLLEGE OF RADIOLOGY J. Frank Wilson, MD (Chair)2012-2015 John “Drew” Ridge, MD, PhD 2014-2015 Peter A.S. Johnstone, MD 2014-2017 Elin R. Sigurdson, MD, PhD (Ex officio) Peter A.S. Johnstone, MD MEMBERSHIP AND CREDENTIALS COMMIT TEE Dr. James D. Cox, MD Ben J. Slotman, MD (Chair) Elin R. Sigurdson, MD, PhD (Ex officio) Mack Roach, III, MD Kelly K. Hunt, MD Erich Sturgis, MD, MPH Quynh-Thu Le, MD Meena S. Moran, MD Matthew C. Abromowitz, MD Arnold Paulino, MD NOMINATING COMMIT TEE Jay S. Cooper, MD Sue S. Yom, MD, PhD Kenneth Russell, MD 2014-2016 2014-2017 The UICC Global Task Force on Radiotherapy for Cancer Control (GTFRCC) 2013-2015 AMERICAN COLLEGE OF RADIOLOGY COUNCILOR 2012-2015 2014-2016 2012-2015 2014-2016 2014-2017 2014-2017 2012-2015 Andre Konski, MD 2012-2015 2014-2016 2014-2017 REPRESENTATIVE TO THE NATIONAL COUNCIL ON RADIATION PROTECTION & MEASUREMENTS RESIDENT AND AT TENDING EDUCATIONAL COMMIT TEE Jeffrey Michalski, MD Elin R. Sigurdson, MD, PhD (Ex officio) Ashesh B. Jani, MD Joe Herman, MD 2014-2015 2012-2015 2014-2017 2014-2017 2012-2015 AMERICAN COLLEGE OF RADIOLOGY ASSISTANT COUNCILOR Martin Colman, MD 2014-2015 REPRESENTATIVE TO THE COMMISSION ON CANCER John “Drew” Ridge, MD, PhD Ritsuko Komaki, MD 2012-2015 2014-2017 TRUSTEES OF THE AMERICAN BOARD OF RADIOLOGY Lynn Wilson, MD Kaled Alektiar, MD 2011-2015 2013-2017 New Membership Application AMERICAN RADIUM SOCIETY ARS Membership Services 11300 W Olympic Blvd #600 Los Angeles CA 90064 Phone: 310-437-0581 ext. 126 Fax: 310-437-0585 Email: [email protected] Twitter: @RadiumSociety Web Site: http://www.americanradiumsociety.org/ APPLICATION FOR MEMBERSHIP PLEASE TYPE OR PRINT CLEARLY PLEASE ACCEPT MY APPLICATION FOR THE FOLLOWING MEMBERSHIP CATEGORY (CHECK ONE): � ACTIVE -- Active membership may be held by physicians and allied scientists: Physicians Graduates of recognized medical colleges; adequate formal training and ongoing practice in those branches of medicine and surgery that are closely allied in the management of cancer; must be certified by a medical or surgical board that governs their area of specialty or have equal qualifications acceptable to the Membership and Credentials Committee. If there is an oncology board in their specialty, they must be certified by that subspecialty board or provide evidence of oncologic qualifications; must have published in peer-reviewed journals or have made a single contribution to science, in keeping with the criteria of the Membership and Credentials Committee. Allied scientists shall be graduates of recognized graduate schools and have qualifications acceptable to the Executive Committee of this Society. They shall have adequate formal training, ongoing professional experience, and clinical or academic contributions in the oncologic sciences. They must have published in peer-reviewed journals or have made a single contribution to science, in keeping with the criteria of the Membership and Credentials Committee. Membership Dues: $190 paid at time of application; upon acceptance will be applied to first year’s dues. Thereafter $275 per year. � CANDIDATE -- Candidate membership may be held by physicians and allied scientists. These members will be non-voting members: Physicians shall be graduates of a recognized medical college and be enrolled in an oncology training program (surgical, radiation, medical, gynecologic.) Allied scientists shall have corresponding qualifications; eligible for Active membership upon completion of training and board certification as described for active membership. Their membership status shall exist until the candidate is selected for Active membership or until no more than two years have passed since board certification. Candidate membership shall be terminated three years from the completion of their training program, unless they have paid the dues as an Active member. No annual dues. � MEDICAL STUDENT— These members will be non-voting members. Medical Student membership may be held by those interested in pursuing a career in an oncology specialty training program (surgical, radiation, medical or gynecologic) and currently enrolled in an Accredited medical school acceptable to the ARS. No annual dues. APPLICANT’S FULL NAME: (LAST/FAMILY NAME) � MD � DO � PhD (FIRST/GIVEN NAME) (MIDDLE NAME OR INITIAL) � Other Degrees: ______________________________________________________________________ Date of Birth (month/day/year): __________________ Gender: M F Country of Birth __________________________ CURRENT ADDRESS – Please indicate if professional-mailing address or home-mailing address: P H Institution Name (If applicable) Department/Division Name (If applicable) Street Address (including: suite#; floor#; mail-stop#; building name; room# -- if applicable) (City) (State/Province) (Indicate Cell or Office Telephone Number / Fax Number) (Zip/Postal Code) (Country) (E-mail Address #1) (E-Mail Address #2) Please list all degrees earned below, including honorary degrees and fellowships: ___________________________________________________________________________________________________________ Conferring Institution Degree Date Started Date Completed _________________________________________________________________________________________ Conferring Institution Degree Date Started Date Completed ARS PROCEEDINGS 2015 i New Membership Application Please list professional training below: Internship: Institution Type Date Started Date Completed ______________________________________________________________________________________________________________________________________ Residency: Institution Type Date Started Date Completed ______________________________________________________________________________________________________________________________________ Specialty Training/Fellowship: Institution Type Date Started Date Completed ______________________________________________________________________________________________________________________________________ Specialty of Medical Practice: Type Years in Practice % of Total Practice in Oncology ______________________________________________________________________________________________________________________________________ Specialty Board (or Equivalent): Qualification Year of Certification/Certificate # _________________________________________________________________________________________________________ Other EXPERIENCE: Hospital (or Professional) affiliations, past and present, with dates: ____________________________________________________________________________________________________________________________________ Current Academic Title (if any): Hospital Title: ____________________________________________________________________________________________________________________________________ Medical and Professional Societies: ____________________________________________________________________________________________________________________________________ Teaching Positions: _____________________________________________________________________________________________________________________________________ State Licensure: Number of State License(s): _____________________________________________________________________________________________________________________________________ What part of your activity is devoted to the treatment of patients with cancer, cancer-related research and/or teaching? _____________________________________________________________________________________________________________________________________ Have you attended any meetings of the ARS? What year(s)? Signature of Applicant: Date: Required ARS Application Documentation: • ACTIVE Applicant: Must be Board Certified • Sponsorship form completed by an Active or Senior ARS member • Current CV • $190 application fee • CANDIDATE Applicant: Sponsorship form completed by an Active or Senior ARS member • A letter of good standing completed by the Program Director of your Residency or Fellowship program • Current CV • MEDICAL STUDENT Applicant: Sponsorship form completed by an Active or Senior ARS member • A letter of good standing • Current CV ACTIVE APPLICANTS ONLY: PLEASE ENCLOSE $190 APPLICATION FEE: � A check (USD only) is enclosed with this application. Please make checks payable to ARS. � I authorize you to charge my: � VISA � MasterCard � American Express CC Number: ___________________________________ Expiration Date: ________________ Amount: _____________ Cardholder Name: ________________________________ Signature: ________________________________________ Date: ____________________________________________ ii 2016 CAll foR AbstRACts The American Radium Society invites you and your colleagues to participate in the ARS 98th Annual Meeting April 16–19, 2016 Hilton Philadelphia at Penn’s Landing Philadelphia, PA President: John “Drew” Ridge, MD, PhD Program Chair: W. Robert Lee, MD CALL FOR ABSTRACTS All authors who wish to present papers for the ARS 98th Annual Meeting may submit an abstract online starting in summer 2015. Abstracts may be selected for either oral or poster presentations. Deadline for submission is November 2015. We encourage everyone to submit an abstract, whether or not you are a member of the ARS. Visit www.americanradiumsociety.org/abstracts for more information and complete submission instructions. Original contributions should be submitted for consideration; any work that has already been accepted for publication or previously presented is not eligible. For more information, please visit www.americanradiumsociety.org or contact the ARS office: 11300 West Olympic Blvd, suite 600 | Los Angeles, CA 90064 Travel Grants and Essay Awards: Trainees who submit a completed manuscript, dealing with subjects related to clinical, translational or basic research, by November 2015, are eligible for the Young Oncologist Essay Awards. These awards provide an honorarium of $500 plus reimbursement up to $2,000 for travel. Young Oncologist Travel Grants of $1,000 will also be available for outstanding abstracts that are reviewed favorably for presentation at the Annual Meeting. Phone: (310) 437-0581 | Fax: (310) 437-0585 | Email: [email protected] | Twitter: @RadiumSociety | Facebook: www.facebook.com/AmericanRadiumSociety American Radium Society Scientific Papers and Posters 2015 SCIENTIFIC PAPERS (S001) Tumor Control and Toxicity Outcomes for Head and Neck Cancer Patients Re-Treated With IntensityModulated Radiation Therapy (IMRT)—A Fifteen-Year Experience (S002) Prognostic Value of Intraradiation Treatment FDG-PET Parameters in Locally Advanced Oropharyngeal Cancer Erqi L. Pollom, MD, Jie Song, PhD, Madhu Sudhan, Benjamin Y. Durkee, Sonya Aggarwal, BA, Rie von Eyben, Timothy T. Bui, BS, Ruijiang Li, PhD, Billy Loo, Quynh-Thu X. Le, Wendy Y. Hara, MD; Stanford University Vinita Takiar, MD, PhD, Dominic Ma, BS, Adam S. Garden, MD, Beth M. Beadle, MD, PhD, Clifton D. Fuller, MD, PhD, Gary B. Gunn, MD, William H. Morrison, MD, David I. Rosenthal, MD, Jack Phan, MD, PhD; UT MD Anderson Cancer Center BACKGROUND: To determine whether fluorodeoxyglucose (FDG) positron emission tomography (PET) parameters measured at an early time point during radiation for locally advanced oropharyngeal cancer (OPC) correlate with outcomes. PURPOSE AND OBJECTIVES: The probability of locoregional failure after definitive treatment for cancers in the head and neck (H&N) area approaches 50%, with 80% of such failures occurring within previously high-dose–treated radiation volumes, within 2 years of treatment. Historically, H&N cancers arising in previously irradiated volumes were rarely re-treated with radiotherapy due to toxicity concerns. With improved precision in planning and delivery, re-irradiation is now used with greater frequency. Here, we review and analyze our 15-year institutional experience using only intensity-modulated radiation therapy (IMRT) to treat previously irradiated H&N carcinoma. METHODS: Patients with stage III–IVB, intact OPC who were treated with definitive radiation with curative intent were included in this study if they underwent both pre- and midtreatment PETcomputed tomography (CT) planning scans in our department. The treatment-planning CT was registered with the PET from the same session, and the metabolic tumor volume (MTV) was extracted from within the primary and nodal tumor volumes contoured by the treating physician. MTV was defined as the volume with standardized uptake value (SUV) > 2.5. MTV velocity was defined as the difference between pre- and midtreatment nodal MTV divided by time elapsed between these two scans. Extraction of imaging features was performed using MATLAB. MATERIALS AND METHODS: We retrospectively reviewed the records of 227 patients who were re-irradiated to the H&N using IMRT between 1999 and 2014. Radiation-related acute and late toxicity, including events requiring hospitalization or urgent intervention and death were recorded. Outcome variables included gender, age, surgery, chemotherapy, radiotherapy dose, radiotherapy volume, and time between initial irradiation and re-irradiation. RESULTS: A total of 206 patients (91%) were treated with definitive intent. Of them, 104 patients (50%) underwent salvage resection and 136 patients (66%) received chemotherapy. Median follow-up after re-irradiation for definitely treated patients was 24.7 months. The 5-year rates of locoregional control, progression-free survival, and overall survival for definitively treated patients were 54%, 25%, and 39%, respectively. Actuarial rates of grade ≥ 3 toxicity were 32% at 2 years and 48% at 5 years, with dysphagia or odynophagia requiring feeding tube placement representing the most common toxicity of grade ≥ 3. On multivariate analysis, concurrent chemotherapy and retreatment site influenced tumor control, whereas response to induction chemotherapy and initial disease site influenced survival. High-dose clinical tumor volume (CTV1) > 50 cc and concurrent chemotherapy were significantly associated with increased grade ≥ 3 toxicity. Notably, patients who were treated to a CTV1 < 25 cc experienced no grade ≥ 3 toxicity. CONCLUSIONS: Re-irradiation for H&N cancers with IMRT and concurrent chemotherapy results in promising local control and survival outcomes in selected patients. Treatment-related toxicity continues to be significant despite improvements in systemic therapy and radiation dose conformality, warranting careful patient selection and target volume delineation. RESULTS: In total, 60 patients who fulfilled the inclusion criteria were treated from February 2009 to January 2014 at Stanford. Median age was 59 years (range: 27–83 yr). The p16 status was positive in 51 patients, negative in 8 patients, and unknown in 1 patient. Nine patients received induction chemotherapy, and 59 patients received concurrent chemotherapy (cisplatin: n = 26; cetuximab: n = 25; carboplatin: n = 8). A total of 25 patients had a > 10-year smoking history. Patients were treated to a median dose of 70 Gy (range: 63.6–70 Gy, in 30–35 fractions). Patients underwent a planning PET at a median of 11 days (range: 2–26 d) prior to radiation start and an intratreatment planning PET after receipt of a median of 16 fractions (range: 10–22 fractions). Median pretreatment MTVs for the entire cohort at the primary, nodal, and combined primary and nodal sites were 16.7 cc (range: 0.9–143.0 cc), 11.5 cc (range: 0–195.1 cc), and 32.7 cc (range: 2.0– 225.2 cc), respectively. Median intratreatment MTVs for the entire cohort at the primary, nodal, and combined primary and nodal sites were 7.6 cc (range: 0.4–150.2 cc), 3.8 cc (range: 0–95.6 cc), and 14.7 cc (range: 0.3–150.2 cc), respectively. Median follow-up was 17 months (range: 2–63 mo). One-year overall survival was 98%. Age, smoking status, chemotherapy, and stage did not predict for survival. For the entire cohort, there was a trend for worse survival with less metabolic response, as measured by MTV velocity (P = .09; hazard ratio [HR] = 1.6). Within the p16+ patients, less metabolic response at the combined primary and nodal sites was a significant predictor for worse survival (P = .03; HR = 2.2). CONCLUSION: Metabolic response during radiotherapy predicts for survival in p16+ OPC patients and may help in risk stratification of these patients for potential treatment de-intensification. ARS PROCEEDINGS 2015 1 American Radium Society Scientific Papers and Posters 2015 (S003) Weekly IGRT Volumetric Response Analysis as a Predictive Tool for Locoregional Control in Head and Neck Cancer Radiotherapy (S004) Combination of Radiotherapy and Cetuximab for Aggressive, High-Risk Cutaneous Squamous Cell Cancer of the Head and Neck: A Propensity Score Analysis Raj Davuluri, MD, Jeff Krase, PharmD, Michael K. Cheung, MD, Brian N. Allen, MS, Sun K. Yi, MD; Department of Radiation Oncology, University of Arizona Cancer Center Joshua D. Palmer, MD, Jon Strasser, MD, Michael Dzeda, MD, Neil Hockstein, MD, Charles Schneider, MD, Adam Raben, MD; Sidney Kimmel Medical College of Thomas Jefferson University; Helen F. Graham Cancer Center of Christiana Care Health System INTRODUCTION: Organ-preserving chemoradiation (CRT) has become standard of care for head and neck squamous cell carcinoma (HNSCC). Volumetric image-guided radiotherapy (IGRT), though primarily utilized for daily setup precision, may also allow for treating physicians to evaluate disease response throughout therapy. In other disease sites, such as lung, IGRT-measured tumor response has been found to be predictive of oncologic outcome. To our knowledge, no such study has been performed in patients treated for HNSCC. METHODS: After institutional review board (IRB) approval, patients who were treated definitively with CRT for stage III–IV HNSCC on TomoTherapy (Madison, WI) with daily megavoltage cone-beam computed tomography (MVCT) IGRT were retrospectively reviewed for volumetric response. A resident physician (RD), alongside an expert HN radiation oncologist (SY), contoured identifiable disease on representative weekly imaging while blinded to patient outcome. Response was measured against formal posttreatment clinical, radiological, and/or histological assessment. Data were stratified according to primary site or largest nodal disease. Student t-test was utilized to compare percent volume reduction in those with recurrence vs those without recurrence. RESULTS: Thirty-five patients were identified with a median followup of 15.8 months. Fifty-six percent of patients were HPV-positive, and 67% was oropharyngeal, 12% was laryngeal, 9% was hypopharyngeal, and 6% was nasopharyngeal in origin. Twenty-six percent of patients were T4, 26% was T3, 21% was T2, and 17% was T1. Twelve percent of patients were N3, 65% was N2, 6% was N1, and 11% was N0. Eighty-five percent of patients had stage IV disease, and 15% had stage III disease. The median dose to gross disease was 70 Gy in 35 fractions. A total of 6 recurrences at a median follow-up of 8 months [range: ] were identified. Mean percent volume of disease at the end of treatment for those with recurrence vs those without recurrence was significantly different (31% vs 6%; P = .001). For primary tumors, the difference in mean residual disease was 20% vs 5% (P = .005) for recurrences vs nonrecurrences, respectively. For nodal disease, the difference in mean residual disease was 47% vs 6% (P = .0001) for recurrences vs nonrecurrences, respectively. When percent residual disease was less than 30% of original volume, the positive predictive value, negative predictive value, sensitivity, and specificity were 100%, 93%, 66%, and 100%, respectively. CONCLUSIONS: Assessment of volumetric IGRT disease response for HNSCC that is treated with definitive CRT is valuable for predicting disease control. Patients with less than 70% volume reduction should be considered for early surgical intervention. 2 BACKGROUND: Locally advanced, high-risk cutaneous squamous cell carcinoma (CSCC) of the head and neck is typically aggressive and treated with combined modality therapy. These patients tend to be older and frail, with multiple comorbidities, which makes chemotherapy difficult to tolerate. Cetuximab is a monoclonal antibody against the epidermal growth factor (EGF) receptor and has demonstrated activity in CSCC. We investigate the safety and preliminary efficacy of combined therapy in advanced, high-risk CSCC with the addition of cetuximab. METHODS: Patients who were identified with locally advanced CSCC with high-risk or very-high-risk features were treated with cetuximab and radiotherapy between 2006 and 2013. A matched cohort over the same time period was identified that was treated with radiation. Propensity score analysis was performed with weighted factors, including Charlson comorbidity index score (age-adjusted), age, Karnofsky performance status (KPS), primary location, T and N stage, recurrent status, margin status, lymphovascular space invasion (LVSI), perineural invasion (PNI), and grade. Overall survival (OS), progression-free survival, and freedom from local or distant recurrence were evaluated with the Kaplan-Meier method for both the unadjusted and propensity score-adjusted groups. Multivariate analysis was performed using Cox proportional hazard models. RESULTS: A total of 29 patients were in the cetuximab group, and 39 were in the control group. Median follow-up for living patients was 30 months. Patients in the cetuximab group were more likely to have advanced N stage, positive margins, and recurrent disease. After matching of propensity scores, the groups were well balanced. OS was not statistically significantly different between the two groups, but there were approximately 20% more long-term survivors in the cetuximab group after matching, with 80% vs 61% surviving at 4 years. Local control rate was 76% and 79% in the cetuximab and control groups, respectively. The rate of distant metastases was lower in the cetuximab group (6.8% vs 10%). The incidence of grade 2–3 toxicity was 41% in the cetuximab group. There was one grade 3 toxicity (cetuximab-induced acneiform rash), one grade 4 toxicity (dysphagia), and no grade 5 toxicity. CONCLUSIONS: Although limited by small numbers, we found that there were more long-term survivors and less distant metastasis in the cetuximab group. This is the largest report of CSCC patients treated with cetuximab. In the absence of prospective data, we believe that these data reveal that the addition of cetuximab is well tolerated and reveal signs of efficacy in this typically poorly performing group of patients and should be pursued in clinical trials. American Radium Society Scientific Papers and Posters 2015 (S005) Radiotherapy for Carcinoma of the Hypopharynx Over Five Decades: Experience at a Single Institution Julian Johnson, MD, Stephen Shiao, MD, PhD, Vivian Weinberg, PhD, Jeanne Quivey, MD, Sue Yom, MD, PhD; University of California, San Francisco; Cedars Sinai BACKGROUND AND PURPOSE: To determine the effect of treat- ment decade and utilization of intensity-modulated radiation therapy (IMRT) on locoregional control and overall survival in patients with squamous cell carcinoma of the hypopharynx. METHODS: Between 1962 and 2008, 116 patients with squamous cell carcinoma of the hypopharynx underwent definitive radiotherapy. We retrospectively reviewed our experience treating these patients with radiotherapy (IMRT, 3-dimensional conformal radiotherapy [3DCRT], intraoperative radiotherapy [IORT], and external beam radiation therapy [EBRT]). This report focuses on the pattern of locoregional control, overall survival, and toxicity rates over the study period. RESULTS: Median follow-up duration from diagnosis was 17 months (range: < 2–441 mo). The 2-year estimates of overall survival and locoregional control were 41% (95% confidence interval [CI], 32%– 50%) and 55% (95% CI, 44%–65%), respectively. The 5-year estimates of overall survival and locoregional control for the entire patient population were 25% (95% CI, 17%–33%) and 49% (95% CI, 37%–60%), respectively. The median overall survival for all patients was 18.3 months (range: 2–441 mo). With respect to treatment type, the median overall survival with EBRT was 13.8 months, 20.1 months for 3DCRT, and 37.8 months for IMRT (log-rank test: P = .04). Median overall survival estimates by decade were not statistically significantly different: 10.5 months for 1960–1969, 11.3 months for 1970–1979, 18.2 months for 1980–1989, 17.0 months for 1990–1999, and 37.8 months for 2000–2010 (P = .22). There was a trend for improved locoregional control comparing post-1980 treatment to pre-1980 treatment (P = .09) but not for overall survival (P = .52). Treatment with chemotherapy was increasingly more common over the 5 decades studied (P < .001) but did not impact locoregional or survival control (P = .69). Having surgery did not impact overall survival (P = .28) but improved locoregional control, resulting in 2-year and 5-year estimates of 66% vs 48% and 60% vs 41%, respectively (P = .045). The frequency of grade ≥ 3 mucositis decreased over time by decade (P = .01), whereas dysphagia did not. Data on xerostomia were not consistently recorded. CONCLUSIONS: With similar 10-year follow-up, there is a trend for improved locoregional control if treated as of 1980. Our data suggest that overall survival is longer for patients treated with IMRT. Our data also suggest that toxicity from mucositis has declined over time. The current study lends further support to the body of evidence suggesting that in contrast to squamous cell carcinoma of the larynx, overall survival is improving for patients with squamous cell carcinoma of the hypopharynx. (S006) Impact of Delays to Adjuvant Radiation Therapy on Survival in Squamous Cell Carcinoma of the Oral Cavity and Oropharynx Erik Liederbach, BS, Carol M. Lewis, MD, MPH, Chi-Hsiung Wang, PhD, Arif Shaikh, MD, Mihir K. Bhayani, MD; Department of Surgery, Center for Biomedical Research Informatics, Department of Radiation Oncology, Department of Head and Neck Surgery, NorthShore University HealthSystem; Department of Head and Neck Surgery, UT MD Anderson Cancer Center INTRODUCTION: Few studies have analyzed how treatment delays to postoperative radiation therapy (RT) for oropharynx and oral cavity squamous cell carcinoma (OSCC) affect overall survival (OS). This study investigates time intervals from surgery to RT using the National Cancer Data Base (NCDB). METHODS: Utilizing the NCDB, we selected 14,058 stage I–IV OSCC patients (4,786 oropharynx, 9,272 oral cavity) treated with surgery and RT from 1998 to 2011. Patients who received neoadjuvant radiation or chemotherapy, adjuvant chemotherapy, and biopsy alone were excluded. Analysis of Variance (ANOVA) and Cox proportional hazard models were utilized. Patients were categorized into four delay groups based on their time from surgery to RT initiation (group 1: ≤ 30 d, group 2: 31–60 d, group 3: 61–90 d, and group 4: > 90 d). RESULTS: The median age was 59 years (range: 18–90 yr), and 9,570 (68.1%) of patients were male. Half of the patients (51.2%) were treated at academic/research facilities. The overall median time from surgery to initiation of adjuvant RT increased from 42 days in 1998 to 52 days in 2011 (P < .001). In 2010–2011, stage I patients on average received RT 11 days earlier than stage IV patients (41 d vs 52 d; P = .007), and patients with no insurance waited 6 days longer compared to patients with private insurance (55 d vs 49 d; P < .001). Patients treated at academic/research hospitals experienced longer delays compared to patients treated at community hospitals (51 d vs 44 d; P = .028), and those treated in the Middle Atlantic region waited the longest for RT (57 d) compared to the West South Central region (42 d; P < .001). There were no significant delays identified among age groups, races, and genders. Between 1998 and 2006, there were 9,677 (68.8%) patients available for survival analysis, who had a median follow-up of 3.7 years (range: 1–13 yr). The unadjusted 5-year OS rates for each group were 53.0% for group 1, 48.4% for group 2, 44.3% for group 3, and 36.0% for group 4 (P < .001). After adjusting for patient, facility, and tumor factors, survival was significantly reduced in group 3 (hazard ratio [HR] = 1.17; 95% confidence interval [CI], 1.06–1.29; P < .001) and group 4 (HR = 1.35; 95% CI; 1.19–1.54; P < .001) compared to group 1. Patients who started RT in group 2 had equivalent survival to group 1. CONCLUSIONS: Delays to adjuvant RT > 60 days increased mortality by 17% to 35% compared to patients treated earlier. Current efforts should be focused on timely delivery of care, and further investigation into factors associated with delays is necessary. (S007) Intensity of Follow-Up After Radiotherapy for HPV-Positive Oropharyngeal Cancer Jessica M. Frakes, MD, Stephanie Demetriou, BS, Tobin Strom, MD, Jeffrey Russell, MD, PhD, Julie A. Kish, MD, Judith McCaffrey, MD, Kristen Otto, Tapan Padhya, MD, Andy Trotti, MD, Jimmy J. Caudell, MD, PhD; Moffitt Cancer Center; Florida Atlantic University ARS PROCEEDINGS 2015 3 American Radium Society Scientific Papers and Posters 2015 PURPOSE AND OBJECTIVES: According to the American Cancer Society, human papilloma virus-positive (HPV+) oropharynx cancer is an epidemic. Fortunately, outcomes for these patients with radiotherapy +/− chemotherapy are excellent. We reviewed our institutional experience with HPV+ oropharynx cancer to determine the time to recurrence or to grade ≥ 3 toxicities to better define an optimal follow-up schedule. MATERIALS AND METHODS: An institutional database of patients with head and neck cancer seen between 2006–2014 was queried, and 232 patients with a biopsy-proven diagnosis of nonmetastatic HPV+ oropharynx squamous cell carcinoma were identified with at least 6 months of follow-up. Charts were reviewed to capture patients’ tumor, treatment, toxicity, and outcome information. Recommended follow-up was every 3 months in the first year, every 4 months in Year 2, and every 6 months in Years 3–5. Locoregional control (LRC), distant control (DC), and overall survival (OS) were calculated according to the Kaplan-Meier (KM) method from the end of treatment. RESULTS: Median follow-up of all patients was 33 months (range: 6–99 mo). Based on Radiation Therapy Oncology Group [RTOG] 0129 risk stratification, 162 patients (70%) were low-risk and 70 (30%) were intermediate-risk. Concurrent systemic therapy was utilized in 85% of patients (n = 196). Three-year LRC, DC, and OS rates were 94%, 91%, and 91%, respectively. Late grade ≥ 3 toxicity occurred in 9% (n = 21) of patients. There were a total of 19 grade 3 toxicities (most commonly feeding tube) and 2 grade 4 toxicities (tracheostomy), with resolution in 15 patients and 1 patient, respectively, at time of last follow-up. Overall, local, regional, or distant relapse or grade ≥ 3 toxicity occurred in 27 patients (68% of all events) within the first 6 months. Subsequently, there were very few events at each time point over 48 months (< 2% at each time point). As expected, recurrence or toxicity events were more common in the intermediate-risk group, while the time to an event was most likely to occur within the first 6 months after therapy. CONCLUSIONS: Following radiotherapy +/− chemotherapy for HPV+ oropharynx cancer, there is a low risk of disease recurrence or late grade ≥ 3 toxicity. As most events occur within 6 months of treatment completion, it may be reasonable to reduce the intensity of follow-up appointments to an every-6-month basis beyond this window. (S008) The Impact of HPV, HIV, and Smoking on Oncologic and Functional Outcomes in Patients With Head and Neck Cancer Waleed F. Mourad, Kenneth S. Hu, Catherine Concert, Daniel Shasha, Louis B. Harrison; Beth Israel Medical Center OBJECTIVES: To report the clinical outcomes and the impact of HPV, HIV, and smoking on patients with head and neck cancer (HNC). MATERIALS AND METHODS: This is a single-institution retrospective study of 105 HIV+ pts with HNC treated from 1998–2013. The median age at radiation therapy (RT) and HIV diagnosis was 51 years (range: 32–72 yr) and 34 (range: 25–50 yr), respectively. HIV 4 duration was 11 years (range: 6–20 yr). A total of 22%, 27%, and 51% had stage I-II, III, and IV disease, respectively. A total of 37% were treated with RT alone, while 63% received concurrent chemoradiation (CRT), and 50% of patients were on highly active antiretroviral therapy (HAART) during treatment. A total of 34 patients had oropharyngeal squamous cell carcinoma (SCC) and metastases of unknown primary origin, 50% of whom were HPV+. Median doses of 70, 63, and 54 Gy were delivered at 1.8–2-Gy/fraction to gross disease and high- and low-risk neck, respectively. Twelve patients underwent neck dissection for N3 disease. RESULTS: Acute skin desquamation and mucositis grade ≤ 2 and 3 rates were 70% and 30%, respectively. Rates of treatment breaks ≥ 10 and 5 days were 10% and 20%, respectively. One patient died from induction chemotherapy (CT), 1 died several weeks post transoral robotic surgery (TORS) for T2N1 SCC of the tonsil, 1 developed grade 4 mucositis, and 1 developed osteoradionecrosis during CRT. The median weight loss was 25 lbs (range: 6–40lbs). With a median follow-up of 60 months (range: 12–140 mo), rates of late dysphagia grades ≤ 2, 3, and 4 were 74%, 15%, and 11%, respectively. Rates of late xerostomia grades ≤ 2 and 3 were 77% and 23%, respectively. The median CD4 counts and viral loads before, during, and after treatment were 370, 135, and 100 and 0, 160, and 260 cells/μL, respectively. Seven patients developed second primary malignancy. The 4-year locoregional control (LRC) and overall survival (OS) rates were 65% and 50%, respectively. Chi-square test showed a significant relationship between LRC and both RT duration and CT, as well as a relationship between lower CD4 counts and higher viral load) (P = .001). Positive trends were observed between weight loss ≤ 10% and LRC and between absence of second malignancy and OS. There was no significant relationship between HPV positivity, smoking, or CT with either LRC or OS. CONCLUSIONS: HIV+ patients with HNC have inferior oncologic and functional outcomes compared to HIV− patients. HPV positivity and smoking did not have a statistically significant impact on clinical outcomes. Innovative treatment modalities and approaches with better efficacy and less morbidity need to be developed for this growing patient population. (S009) Radiation Therapy Improves Outcomes With Desmoplastic Melanoma of the Head and Neck Tobin J. Strom, MD, Jimmy J. Caudell, MD, PhD, Jonathan S. Zager, MD, C. Wayne Cruse, MD, Jane L. Messina, MD, Vernon K. Sondak, MD, Louis B. Harrison, MD, Andy M. Trotti, MD; H. Lee Moffitt Cancer Center and Research Institute BACKGROUND: Desmoplastic melanomas are considered to have a high risk of local recurrence after resection alone, especially in the head and neck. We hypothesized that adjuvant radiotherapy might reduce the risk of local recurrence. METHODS: A single-institution institutional review board (IRB)approved study was performed including 140 patients with desmoplastic melanoma without distant metastatic disease treated from 1990–2010 with wide excision ± sentinel lymph node dissection ± regional lymph node dissection. Patient, tumor, and treatment characteristics were compared between the groups based on American Radium Society Scientific Papers and Posters 2015 receipt of adjuvant radiotherapy. Adjuvant radiotherapy was delivered to the primary tumor bed in all cases with a 2–4-cm margin as feasible and to the draining lymphatics in a minority of cases (n = 5). Patients were treated to a total dose of either 30 Gy in 5-Gy fractions dosed twice per week (n = 37) or 50–68 Gy in 25–34 daily fractions (n = 32). Adjuvant systemic therapy was delivered in 18 cases (interferon in 16 of 18 cases). The primary study outcome was local control, and the secondary outcome was locoregional control. Kaplan-Meier (KM) analysis and the log-rank test were used to compare outcomes. A Cox hazards multivariate (MV) model was created for the primary outcome. RESULTS: Median follow-up was 47 months. Receipt of radiotherapy was associated with deeper tumors (median 5.4 mm vs 2.8 mm; P < .001) and positive margins (28% vs 13%; P = .03), compared with no radiotherapy. Nevertheless, adjuvant radiotherapy was associated with improved local control compared with patients who did not receive it (4-yr KM estimate: 94% vs 74%; P = .02) and locoregional control (4-yr KM estimate: 86% vs 69%; P = .03). On Cox MV analysis, radiotherapy was independently associated with improved local control (hazard ratio [HR] = 0.17; 95% confidence interval [CI], 0.06–0.51]; P = .002) and locoregional control (HR = 0.26; 95% CI, 0.11–0.63; P = .003). Variables associated with local recurrence on MV analysis included age > 70 years (HR = 4.1; 95% CI, 1.6–10.6; P = .004) and positive margins (HR = 5.6; 95% CI, 2.1–15.1; P = .001). Among patients with positive margins (n = 28), those treated with radiation therapy had improved local control (4-yr KM estimate: 88% vs 18%; P = .01) compared with those not treated with adjuvant radiation. Similarly, patients who had negative margins also benefited from adjuvant radiation therapy (4-yr KM local control estimate: 96% vs 80%; P = .048). CONCLUSIONS: Radiotherapy improves both local and locoregional control in patients with desmoplastic melanoma of the head and neck, regardless of margin status. (S010) A Phase III Randomized Trial of MRI-Mapped Dose-Escalated Salvage Radiotherapy Post-Prostatectomy: The MAPS Trial—Feasibility and Acute Toxicity Amber Orman, MD, Alan Pollack, MD, PhD, Kelin Wang, PhD, Radka Stoyanova, PhD, Elizabeth Bossart, PhD, Deukwoo Kwon, PhD, Matthew Abramowitz, MD; University of Miami PURPOSE AND OBJECTIVES: MAPS is the first phase III randomized trial of magnetic resonance imaging (MRI)-mapped dose-escalated salvage radiotherapy. In this planned feasibility analysis, we relate dosimetry to acute toxicity in the setting of dose escalation using a simultaneous incorporated hypofractionated boost (SIHB) to MRI-identified lesions in the prostate bed. MATERIALS AND METHODS: Two intensity-modulated radiothera- py (IMRT) plans were generated for each patient treated on the MAPS protocol, regardless of actual randomization arm: one for the standard fraction radiotherapy (SFRT) arm and one for the SIHB arm. In the SFRT arm, 68 Gy in 34 fractions was prescribed to ≥ 95% of the planning target volume (PTV). In the SIHB arm, an additional 2.25 Gy daily SIHB was prescribed to the gross tumor volume (GTV) (2.25 Gy daily, total dose of 76.5 Gy in 34 fractions). The trial stipulates that no more than 35% and 55% of the rectum should receive ≥ 65 Gy and ≥ 40 Gy, respectively, and that no more than 50% and 70% of the bladder minus the clinical target volume (B-CTV) should receive ≥ 65 Gy and ≥ 40 Gy, respectively. Acute toxicities were recorded for patients at designated times per protocol according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0. RESULTS: Data from the first 14 enrolled patients were reviewed to ensure dosimetric adequacy of the protocol requirements. In all plans, at least 95% of the PTV received 68 Gy, and at least 95% of the GTV received 76.5 Gy. Dosimetric constraints were achieved for all organs at risk (OARs) except B-CTV. Five cases had > 70% of the bladder receiving ≥ 40 Gy in both plans, and one case had > 50% of the bladder receiving ≥ 65 Gy in the SIHB plan only. The only difference between the SFRT and SIHB plans, per patient or overall, was a higher percent volume of the PTV receiving 68 Gy in the SIHB plans. In 13 patients for whom full data were available, the highest toxicity recorded was grade 2 gastrointestinal toxicity: one episode in each arm. CONCLUSIONS: This study demonstrates that even though most MRI-identified GTVs are located in close proximity to critical structures, dose escalation is achievable without exceeding rectal constraints in all cases, and bladder constraints in the majority of cases. These variations are in cases with small bladders encompassed in the CTV and are not associated with increased acute toxicity. (S011) Hypofractionated vs Standard Fractionated Proton Beam Therapy for Early-Stage Prostate Cancer: Interim Results of a Randomized Prospective Trial William F. Hartsell, MD, Megan Dunn, PhD, Gary Larson, MD, Carlos Vargas, MD; CDH Proton Center; Proton Collaborative Group; ProCure Proton Center; Mayo Clinic PURPOSE: After prostate cancer treatment, most adverse event (AE) and quality of life (QoL) changes can be initially identified within the first 2 years. The purpose of this interim analysis is to determine if there are differences in terms of QoL, International Prostate Symptom Score (IPSS), or AEs among prostate cancer patients treated on a randomized prospective trial with either standard fractionation or hypofractionation. MATERIALS AND METHODS: Eighty-two patients were randomized to 38 Gy(relative biologic effectiveness [RBE]) in 5 treatments (n = 49) vs 79.2 Gy(RBE) in 44 treatments (n = 33). All patients had stage I prostate cancer and were treated with proton therapy using fiducial markers and daily image guidance. RESULTS: Median follow-up for both groups was 18 months, with 33 patients reaching follow-up of 2 years or more. Patient characteristics for both groups were similar, with most patients being T1c (84%) and all having a Gleason score of 6 and a prostate-specific antigen (PSA) level < 10 ng/mL (median, 5.6 ng/mL). Baseline median IPSS was 5 for the 5-fraction arm (range: 0–15), and median IPSS was also 5 for the 44-fraction arm (range: 0–14). There was no difference between the two groups with regard to Expanded Prostate Index Composite (EPIC) urinary, bowel, or sexual function scores at 3, 6, 9, 12, 18, 24, ARS PROCEEDINGS 2015 5 American Radium Society Scientific Papers and Posters 2015 or 36 months. The only significant difference was the IPSS score at 12 months: 5 for the 44-fraction arm vs 8 for the 5-fraction arm (P = .03), but there was no difference in the IPSS scores at the other time points. No grade ≥ 3 AEs were seen in either arm. CONCLUSIONS: Patients tolerated proton therapy in this randomized trial well, with excellent QoL scores, persistent low IPSS, and no grade ≥ 3 AEs in either arm. Thus far, there is no apparent clinical difference in outcomes with hypofractionated proton beam therapy compared to standard fractionation. (S012) Trends in the Selection of Definitive Treatment for Newly Diagnosed Prostate Cancer in Men < 60 Years Old Joseph Safdieh, MD, J. Rineer, MD, M. Shao, MD, E. Navo, MD, D. Schreiber, MD; SUNY Downstate Medical Center; University of Florida Health Cancer Center at Orlando; Department of Veteran Affairs, NY Harbor Campus INTRODUCTION: According to the National Comprehensive Cancer Network (NCCN) Guidelines, alternatives to definitive treatment include active surveillance for men with low-risk (LR) prostate cancer and a life expectancy ≥ 10 years, as well as androgen deprivation alone for men with high-risk (HR) disease who are not candidates for definitive therapy. In this study, we utilized the Surveillance, Epidemiology, and End Results (SEER) Database to analyze the trends for selection of definitive treatment in a young patient population (age < 60 years). MATERIALS AND METHODS: We identified all men < 60 years old who were diagnosed with nonmetastatic prostate adenocarcinoma from 2010–2011 and had a follow-up time > 6 months. Detailed clinical and pathologic factors were collected, including the T-stage, Gleason scores, and prostate-specific antigen, and these data were used to group men into NCCN risk groupings. Receipt of treatment was categorized as no definitive treatment, local treatment (such as cryotherapy or transurethral resection of the prostate [TURP]), radical prostatectomy (RP), and radiation therapy (RT). Since androgen deprivation is not recorded in the SEER database, these men were included as having no definitive treatment. Descriptive analyses were used to determine rates of treatment receipt. Multivariate logistic regression was used to determine predictors for selecting treatment other than definitive therapy. RESULTS: There were a total of 16,630 men identified. Of these subjects, 7,281 (43.8%) had LR disease, 5,660 (34%) had intermediaterisk (IR) disease, and 3,689 (22.2%) had HR disease. The median age at diagnosis was 55 years (interquartile range [IQR]: 52–58 yr). For those with LR disease, 20.5% pursued no definitive treatment, 0.3% pursued local treatment, 57.8% pursued RP, and 21.3% underwent RT. For IR disease, 8.0% pursued no definitive treatment, 0.7% underwent local treatment, 69.2% underwent RP, and 22.0% received RT. For HR disease, 25.5% underwent no definitive treatment, 1.3% underwent local treatment, 51.5% received RP, and 21.8% received RT. On multivariate logistic regression, AfricanAmerican race (odds ratio [OR] = 0.76; 95% confidence interval [CI], 0.68–0.85; P < .001), and the presence of HR disease (OR = 0.74; 95% CI, 0.67–0.81; P < .001) were associated with a decreased likelihood of selecting definitive treatment. 6 CONCLUSION: In this young patient population, no definitive treatment was selected in 22.5% of men with LR disease. For those with HR disease, 25.5% did not undergo active treatment, indicating that these men appear to be receiving less aggressive therapy than recommended by NCCN guidelines. In addition, African-American race is also associated with a decreased likelihood for receipt of definitive therapy. (S013) Heterogeneity Within the Prostate and Risk-Adapting Dose-Volume Analysis With SBRT for Prostate Cancer Zachary A. Seymour, MD, Li Zhang, PhD, Albert J. Chang, MD, I-Chow J. Hsu, MD, Alexander R. Gottschalk, MD, PhD; Department of Radiation Oncology, Cancer Center, University of California, San Francisco PURPOSE AND OBJECTIVES: To evaluate dose-volume relationships of genitourinary (GU) toxicity after stereotactic body radiotherapy (SBRT) for prostate cancer to determine optimal cut points for dosimetric parameters based on prostatic volume. MATERIALS AND METHODS: Fifty-one of 56 patients treated with SBRT for prostate cancer had evaluable dose-volume histograms. All patients were treated in a uniform manner, receiving a total dose of 38 Gy in 4 fractions. Acute, late, and overall GU toxicities were documented according to the Common Terminology Criteria for Adverse Events version 4 (CTCAE v4). Dose volumes were assessed via receiver operating characteristic (ROC) curves to determine optimal cut points at 0.5-Gy dose-volume intervals, and probabilities of toxicity for each cut point were produced. Only dose volumes with a sensitivity and specificity > 0.65 were considered for analysis of GU grade 2+ toxicity. RESULTS: The median age at treatment was 68 years, and median prostate volume was 45.4 mL. The median prescription isodose line was 68%. The median clinical follow-up was 35.49 months. Acute and late grade 2+ GU toxicities occurred in 35.7% and 23.2%, respectively, with only 2 grade 3 GU toxicities. Overall toxicity was associated with baseline prostate volume (continuous, P = .006; hazard ratio [HR] = 1.06; 95% confidence interval [CI], 1.02–1.10) with an overall risk of 39% for grade 2+ toxicity in patients with prostate volumes > 45 mL (AUC 0.722, sensitivity 67%, specificity 71%). The probability of late grade 2+ GU toxicity was associated with absolute and normalized prostate volumes across doses from 46–50 Gy, suggestive of increasing late toxicity with intraprostatic heterogeneity. Normalized V50 Gy of 9% was the strongest normalized prostate volume associated with a 19% late grade 2+ GU toxicity rate (AUC 0.763, sensitivity 82%, specificity 68%). Urethra 42 Gy volume was also associated with grade 2+ toxicity with an optimal cutoff of 2.1 mL (AUC 0.62, sensitivity 82%, specificity 70%, probability 19%). No bladder volume cut points were strongly associated with late or overall GU grade 2+ toxicity. CONCLUSIONS: Greater intraprostatic heterogeneity was associated with late grade 2+ GU toxicity. Given the high correlation of prostate volume with toxicity, SBRT dose parameters should be individualized and risk-adapted based on normalized prostate volumes, including a V50 not to exceed 9% of the prostate. The urethra is an important organ at risk, and the 42-Gy dose-volume should be lim- American Radium Society Scientific Papers and Posters 2015 ited to 2 mL, while bladder dose-volumes appear to be poor predictors of GU grade 2+ toxicity. (S014) pN+ Prostate Cancer (CaP) Does Not Imply Incurable Disease Jan Poleszcuk, PhD, Heiko Enderling, PhD, Peter Johnstone, MD, FACR; Moffitt Cancer Center OBJECTIVES: Mathematical modeling of cancer provides mechanisms not simply to predict tumor behavior based on cellular signatures but also to predict patient options based on population dynamics. We have previously reported that populations with cancer die according to an inverse Gompertzian model. Further, we have shown that “incurability” is associated with an inflection point in the Gompertz curve and a nadir in the derivatives of these curves over time. Before reaching this inflection, potential curative interventions may be applied; after this point, reproducible cure is unlikely. Randomized data supporting optimal therapy for pN+ prostate cancer (CaP) patients are frankly nonexistent beyond recommending hormonal therapy. We sought to stratify curability of this cadre of patients by assessing survival by degree of lymph node (LN) positivity (%LN+). METHODS: Overall or observed survival data in the CaP population are of less utility than in other cancers because of frequent comorbid diagnoses. Thus, national registry-based data were specifically used: relative survival data were retrieved by year postdiagnosis and stratified by the number of positive LNs and by the number of LNs evaluated postoperatively. Data were obtained from the public access Surveillance, Epidemiology, and End Results (SEER) between 1988 and 2011. Observed survival was corrected for an age-, race-, and gender-matched national sample. These data, obtained from a national sample, may be presumed to include hormonal therapy in the disease trajectory. Raw data were then converted to %LN+ and modeled using inverse Gompertzian kinetics. Inflection points were determined for %LN+ values less than 10%, between 10% and 25%, between 25% and 50%, and over 50%. RESULTS: Data were retrieved from 8,018 pN+ patients. The inflection point varies inversely with increasing %LN+. For patients with less than 10% LN+, the inflection was almost 14 years postoperatively. Patients with 10% to 25% LN+ had an inflection point at almost 10 years; for those with more than 50% LN+, the inflection point was just over 4 years. The difference in time to inflection between cadres was significant. CONCLUSIONS: When corrected for comorbid conditions in this patient cohort, CaP patients with < 10% %LN+ have almost 14 years before their disease becomes incurable. Progressively larger %LN+ yields smaller windows of such time. Above 50% LN+, the inflection point is only about 4 years. While better therapies for pN+ CaP must be defined, this patient cadre is not homogenous and should be stratified by %LN+ in future clinical trials. (S015) Impact of Health Insurance Status on Prostate Cancer Treatment Modality Selection in the United States Trevor Bledsoe, MD, Henry Park, MD, MPH, Charles Rutter, MD, Sanjay Aneja, MD, James Yu, MD, MHS; Department of Therapeutic Radiology, Yale University School of Medicine INTRODUCTION: A variety of treatment modalities are available for the management of men with clinically localized prostate cancer in the United States. In addition to clinical factors, treatment choice may be influenced by a patient’s insurance status. We investigated the influence of health insurance on prostate cancer treatment modality selection in the United States. METHODS: Men aged 18–65 years treated for localized prostate cancer from 2010–2011 were identified in the National Cancer Data Base. Patients with no insurance or private insurance were included. Treatment modalities included minimally invasive surgery alone (MIS), open surgery alone, external beam radiotherapy alone (EBRT), proton therapy alone, brachytherapy alone, hormone therapy alone, active surveillance, and combinations of treatments. Demographic and clinical covariates included age, race, income, education level, year of diagnosis, treatment facility type, D’Amico risk classification, and Charlson/Deyo score. Chi-square and multivariable logistic regression analyses were used to evaluate the association of insurance status and other covariates with treatment modality selection. RESULTS: We identified 67,370 patients with either no insurance (3.2%) or private insurance (96.8%). The greatest disparities in treatment modality by insurance status were observed among men receiving MIS and EBRT. For patients with no insurance, 35.1% received MIS and 25.6% received EBRT. For patients with private insurance, 60.1% received MIS and 9.7% received EBRT. Insurance status was the strongest predictor of receipt of both MIS and EBRT on multivariable analysis. Lack of insurance was associated with decreased utilization of MIS (odds ratio [OR] = 0.39; 95% confidence interval [CI], 0.36–0.43; P < .001) and increased utilization of EBRT (OR = 2.56; 95% CI, 2.27–2.85; P < .001). CONCLUSIONS: Patients without insurance were less than half as likely to receive MIS and more than twice as likely to receive EBRT compared with patients with private insurance in our national cohort. Our findings suggest that with expanding access to private insurance under the Affordable Care Act, there may be significant shifts in the selection of treatment modality for men with prostate cancer in the United States. (S016) Intermediate-Risk Prostate Cancer: A MedicareBased Cost Comparison of Five Radiotherapy Regimens Romaine C. Nichols, MD, Kathy McIntyre, Juana Gifford, Steve Ritz, Stuart Klein, Curtis M. Bryan, MD, MPH, Randal H. Henderson, MD, MBA, William M. Mendenhall, MD, Nancy P. Mendenhall, MD, Bradford S. Hoppe, MD, MPH; Proton Therapy Institute, Department of Radiation Oncology, University of Florida BACKGROUND: Patients with intermediate-risk prostate cancer choosing radiotherapy may be treated with a number of regimens. The cur- ARS PROCEEDINGS 2015 7 American Radium Society Scientific Papers and Posters 2015 rent study compares the direct treatment cost for five therapeutic options based on fiscal year 2014 Medicare-allowable reimbursements. METHODS: Hypothetical charge sheets were generated along with the expected Medicare-allowable reimbursements (based on global billing where applicable) for the following regimens: (1) image-guided intensity-modulated radiotherapy (IGIMRT) to a dose of 78 Gy in 39 fractions with one field reduction (IGIMRT); (2) dose-escalated IGIMRT to 84.60 Gy in 47 fractions with one field reduction (MSKCC-IGIMRT); (3) IGIMRT to a dose of 45 Gy in 25 fractions followed by a 90-seed I125 prostate implant (IGIMRT-BTX); (4) image-guided proton therapy to a dose of 78 Gy(relative biologic effectiveness [RBE]) in 39 fractions with one field reduction (SFPT); and (5) image-guided hypofractionated proton therapy to a dose of 72.50 Gy(RBE) in 29 fractions with one field reduction (HFPT). RESULTS: Based on fiscal year 2014 Medicare-allowable reimbursements, the direct cost, including professional fees, technical fees, isotope costs, and facility fees, for each intervention is as follows: IGIMRT, $25,204.12; MSKCC-IGIMRT, $29,130.76; IGIMRT-BTX, $31,104.39; SFPT, $46,652.66; and HFPT, $34,977.54. CONCLUSIONS: These data present a framework for evaluating the cost-effectiveness of proton therapy as compared to competing therapeutic options. Under current Medicare-allowable reimbursements, the cost of proton therapy relative to the cost of other therapeutic options is highly dependent on the number of radiotherapy fractions delivered. The feasibility of delivering hypofractionated proton therapy for patients with localized prostate cancer is being investigated at a number of institutions, as well as within the framework of a multicenter protocol. Ultimately, the cost of proton therapy will need to be weighed against tumor control probabilities, as well as the economic and quality of life benefits associated with reduced normal tissue exposure. motherapy, and 9.2% received adjuvant chemotherapy. Median patient age was 64 years (range: 23–93 yr), and median follow-up was 3.7 years. A total of 23 men experienced RR following lymphadenectomy. Among men who failed to respond to treatment, median time to RR was 4.5 months (interquartile range: 3.6–9.0 mo). Three- and 5-year cumulative incidence rates of RR were 15.5% and 16.8%, respectively. On univariate analysis, pathologic T-stage at penile surgery (P = .013), clinical nodal stage before lymphadenectomy (P < .001), the presence of extranodal extension (ENE) at lymphadenectomy (P < .001), and involvement of > 3 nodes at lymphadenectomy (P < .001) were associated with RR. The 3-year RR rate was 44.3% in men with ENE and 4.0% in men without ENE. For men with > 3 involved nodes, the 3-year RR rate was 49.0% vs 8.7% in men with < 3 involved nodes. The 3-year RR rate for men with cN0, cN1, cN2, and cN3 disease was 6.4%, 15.0%, 10.5%, and 38.6%, respectively. On multivariate analysis, the presence of cN3 disease before lymphadenectomy (vs cN0; adjusted hazard ratio [AHR] = 6.91; 95% confidence interval [CI], 1.23–38.8; P = .028), > 3 pathologically involved nodes (AHR = 10.92; 95% CI, 2.51–47.5; P = .001), ENE (AHR = 77.89; 95% CI, 12.59–482.0; P < .001), and pT4 disease at penile surgery (vs pT1; AHR = 50.15; 95% CI, 5.04– 499.6; P < .001) continued to be associated with RR. Conversely, adjuvant chemotherapy was associated with reduced RR (AHR = 0.09; 95% CI, 0.01–0.59; P = .012). Median survival for men who experienced RR was 11.4 months compared with 15 years for men who did not develop RR. CONCLUSION: The presence of cN3 disease, ENE, and > 3 lymph nodes at lymphadenectomy and the presence of pT4 disease at penile surgery were associated with increased risk of early RR, while adjuvant chemotherapy decreased RR. Since RR portends a dismal prognosis with few salvage options, men with these adverse factors should be considered for adjuvant therapy, including radiation therapy, to reduce RR. (S017) Factors Associated With Regional Recurrence Following Lymphadenectomy for Penile Squamous Cell Carcinoma (S018) Intensity-Modulated Radiation Therapy or Conformal Radiation Therapy and Cardiopulmonary Mortality Risk in the Elderly With Esophageal Cancer Jay Reddy, MD, PhD, Curtis A. Pettaway, MD, Lawrence Levy, MS, Lance Pagliaro, MD, Pherose Tamboli, MD, Priya Rao, MD, Isuru Jayaratna, MD, Karen Hoffman, MD, MPH; UT MD Anderson Cancer Center Steven H. Lin, MD, PhD, Ning Zhang, PhD, Joy Godby, BS, Jingya Wang, MD, Gary D. Marsh, BS, Zhongxing Liao, MD, Ritsuko Komaki, MD, Linus Ho, MD, Wayne L. Hofstetter, MD, Stephen G. Swisher, MD, Thomas A. Buchholz, MD, Linda S. Elting, PhD, Sharon H. Giordano, MD; UT MD Anderson Cancer Center PURPOSE: Factors associated with regional recurrence (RR) following lymphadenectomy for penile cancer were assessed to determine which patients might benefit from adjuvant therapy. METHODS: Men who underwent lymphadenectomy for penile squamous cell carcinoma between 1990–2014 were identified from an institutional database. Kaplan-Meier curves estimated time to RR calculated from the date of lymphadenectomy. Cox proportional hazards models evaluated the association between RR and patient and tumor characteristics. Backward selection with a P value cutoff of .05 selected covariates into the multivariate model. RESULTS: A total of 153 men underwent lymphadenectomy and did not receive adjuvant radiation therapy: 54.9% of patients underwent inguinal lymphadenectomy, and 44.4% underwent inguinal and pelvic lymphadenectomy; 28.8% of patients received neoadjuvant che- 8 PURPOSE: We performed a population-based assessment for the all-cause and cardiopulmonary mortality risk in esophageal cancer (EC) patients treated with chemotherapy and radiation, comparing conventional radiotherapy (CRT) or advanced radiation modality defined by the use of intensity-modulated radiation therapy (IMRT). PATIENTS AND METHODS: We identified 2,578 patients aged over 65 years from the Surveillance, Epidemiology, and End Results (SEER)/Texas Cancer Registry-Medicare databases who had nonmetastatic EC diagnosed between 2002 and 2009 (CRT = 2,265; IMRT = 313). We defined radiation modality by delivery claims, whether it was by conventional radiation therapy, which could either have been two-dimensional (2D) or three-dimensional (3D) (CRT: Healthcare Common Procedure Coding System [HCPCS]: American Radium Society Scientific Papers and Posters 2015 77401–77416) or by IMRT delivery (HCPCS: 77418, G0174) within 6 months of diagnosis. Patients in both cohorts were compared using propensity score–based adjustment. Cause-specific and overall mortality rates were evaluated using Kaplan-Meier LIFETEST and a multivariate (MVA) Cox proportional hazards model. RESULTS: Except for marital status and SEER region, both radiation cohorts were well balanced for various patient, tumor, and treatment characteristics, including the distribution of the use of IMRT vs CRT in urban/metropolitan or rural areas. CRT was done primarily by 3D delivery (98.9%). IMRT use increased from 2.6% in 2002 to 30% in 2009, while 3D use decreased from 97.4% in 2002 to 70% in 2009. In the unadjusted analysis, all-cause mortality, EC-specific mortality, and cardiac mortality were significantly reduced in the IMRT group (all: 52.4% vs 74.5%, P < .0001; EC: 40.3% vs 55.6%, P < .0001; and cardiac: 1.6% vs 5.3%, P = .0043). However, no difference was seen in deaths from pulmonary (0.96% vs 1.55%; P = .419) or other causes (9.6% vs 12.1%; P = .204). On propensity score– adjusted MVA, IMRT was not associated with EC-specific mortality (hazard ratio [HR] = 0.87; 95% confidence interval [CI], 0.69–1.06), pulmonary (HR = 1.04; 95% CI, 0.29–3.79), or other-cause mortality (HR = 0.81; 95% CI, 0.54–1.22) but was significantly associated with higher overall survival (HR = 0.83; 95% CI, 0.70–0.98) and lower cardiac mortality (HR = 0.35; 95% CI, 0.14–0.88). Similar associations were seen even after adjusting for physician experience, the type of chemotherapy used, and sensitivity analysis removing hybrid radiation claims. CONCLUSION: In this population-based analysis, IMRT use was significantly associated with improved overall survival and reduced cardiac mortality in patients with esophageal cancer. (S019) Radiation-Induced Liver Disease Following Liver SBRT for Primary Hepatic Malignancies: Analysis of a Prospective Institutional Study Ashley Weiner, MD, PhD, Jeffrey Olsen, MD, Daniel Ma, MD, Pawel Dyk, MD, Todd Dewees, PhD, Parag Parikh, MD; Washington University; Mayo Clinic PURPOSE AND OBJECTIVES: Although primary liver tumors are commonly treated by surgery, transplant, radiofrequency ablation, or transarterial therapies, there is an increasing patient population that is not suited to these treatments. Stereotactic body radiotherapy (SBRT) allows precise delivery of hypofractionated radiation—a strategy that is thought to minimize the risk of radiation-induced liver disease (RILD) over conventionally fractionated radiation. The objective of this study was to evaluate the feasibility, safety, and efficacy of SBRT in the treatment of inoperable hepatocellular cancer (HCC) and intrahepatic cholangiocarcinoma (IHC). MATERIALS AND METHODS: A total of 28 patients with inoperable HCC or IHC and Child-Pugh score ≤ 8 were enrolled on a singleinstitution prospective protocol. A total of 32 lesions in 26 patients were treated with a planned dose of 55 Gy in five fractions, which was reduced to maintain the mean liver dose < 20 Gy or spare 700 cc of liver with doses < 20 Gy. Treatment was delivered via linear accelerator with immobilization, daily imaging, and end-expiratory gating on fiducial markers. Kaplan-Meier survival analysis was per formed. Toxicities were graded by Common Terminology Criteria for Adverse Events version 4 (CTCAE v4) criteria, and radiographic response to treatment was scored by European Association for the Study of the Liver (EASL) guidelines. RESULTS: The study cohort included 12 patients with HCC, 13 patients with IHC, and 1 patient with a biphenotypic tumor. Median prescribed dose was 55 Gy (range: 40–55 Gy) in five fractions. Mean tumor diameter was 5.3 cm (range: 1.6–12.3 cm), and mean PTV volume was 342 cc (range: 54–964 cc). Median follow-up was 7.7 months with a median overall survival (OS) of 9.8 months. The 6and 12-month OS rates were 80% and 44%, respectively. On imaging 8 weeks posttreatment, complete radiographic response was noted in four patients, while six patients had a partial treatment response. Two grade 5 toxicities occurred during study follow-up. One patient with cirrhosis and a renal transplant history who was being treated with tacrolimus, a potentially hepatotoxic medication, developed RILD, which progressed to hepatic failure and death. Another patient with longstanding cirrhosis succumbed to a gastrointestinal bleed in the setting of RILD. CONCLUSIONS: Primary hepatic malignancies that are not amenable to surgical resection portend a poor prognosis, despite available treatment options. Though RILD is rare following SBRT, this study demonstrates a risk, despite close adherence to standard dose and volume constraints. Further analysis of this prospective study will seek to elicit patient parameters that may increase susceptibility to toxicities, such as RILD. (S020) Predictors of Local-Regional Failure and the Impact on Overall Survival in Patients With Resected Exocrine Pancreatic Cancer Kenneth Merrell, MD, MS, Christopher Hallemeier, MD, William Harmsen, MS, Fernando Quevedo, MD, Michael Kendrick, MD, Michael Haddock, MD, Robert Miller, MD; Department of Radiation Oncology, Division of Biomedical Statistics and Informatics, Department of Medical Oncology, Department of General Surgery, Mayo Clinic PURPOSE: Resection of exocrine pancreatic cancer is necessary for cure, but most patients subsequently develop progressive disease. In prospective trials, rates of local-regional failure (LRF) range from 20% to 40%. We evaluated our institutional experience to better understand the risk factors for LRF and the impact on overall survival (OS). METHODS: We reviewed 665 consecutive patients with nonmetastatic exocrine pancreatic carcinoma who underwent resection (R0 or R1) at our institution between March 1985 and January 2011 and received adjuvant therapy. A total of 458 patients had adequate follow-up and imaging to determine LR control (LRC) rates. All patients received adjuvant chemotherapy (n = 80, 17.5%) or chemoradiotherapy (CRT) (n = 378, 82.5%). Chemotherapy and CRT most frequently consisted of six cycles of gemcitabine and 50.4 Gy in 28 fractions with concurrent 5-fluorouracil (5-FU), respectively. In addition to CRT, 207 (45%) patients received additional chemotherapy. LRF was defined as tumor bed or regional node progression with or without distant metastatic progression. When available, CA19.9 values were used to validate radiographic findings. LRC and ARS PROCEEDINGS 2015 9 American Radium Society Scientific Papers and Posters 2015 OS were estimated using the Kaplan-Meier method. Univariate and multivariate (MVA) analyses were performed using Cox proportional hazards regression models. RESULTS: Median patient age was 64.5 years (range: 37–88 yr). Median follow-up for living patients was 84 months (range: 6–300 mo). Tumor location was pancreas head (80%), body (5%), and tail (8%) or unspecified (7%). Tumor grade was I (0.5%), II (11.5%), III (76%), or IV (12%). T-stage was T1 (7%), T2 (20%), T3 (70.5%), or T4 (2.5%). N stage was N0 (38%) or N1 (62%). Extent of resection was R0 (84%) or R1 (16%). The overall crude incidence of LRF was 17% (n = 79). Three-year rates of LRC for patients with and without RT were 81.6% vs 68.3%, respectively (P = .003; hazard ratio [HR] = 0.45; 95% confidence interval [CI], 0.28–0.76). On MVA, no radiation therapy (P = .0097; HR = 2.2; 95% CI, 1.22–3.79), elevated preoperative CA19.9 (P = .046; HR = 1.92 per 20 units; 95% CI, 1.02– 3.59), and positive nodes/total nodes ratio ≥ 0.2 (P = .059; HR = 1.61; 95% CI, 0.98–2.65) were associated with LRF. When used as a time-dependent covariate for OS, LRF was associated with worse OS (P < .0001; HR = 4.07; 95% CI, 3.2–5.2) compared with no LRF. CONCLUSIONS: In our analysis of 458 patients with resected pancreatic cancer and adjuvant therapy, elevated preoperative CA 19.9 and no adjuvant RT were associated with increased risk of LRF. LRF was associated with poor OS. As such, RT should be considered as an adjunctive LR treatment modality for patients undergoing pancreatic cancer resection. (S021) The Role of Neoadjuvant Stereotactic Body Radiation Therapy in Pancreatic Cancer tively. Fluorouracil, leucovorin, irinotecan, and oxaliplatin (FOLFIRINOX)-based chemotherapy was administered to 61.5% and 45.9% of SBRT and chemotherapy-alone patients, respectively. The majority (61.5%) of SBRT patients were deemed unresectable, while only 29.7% in the chemotherapy-alone group had LAPC. Pancreaticoduodenectomy was performed in 66.7% of SBRT patients compared with 75.7% of chemotherapy-alone patients. Median time to surgery was 2.0 months (range: 0.1–10.5 mo) from the end of SBRT. The overall rate of margin-negative resection in patients who received SBRT was 87.2%, with 86.7% in BRPC and 87.5% in LAPC. In comparison, the overall margin-negative resection rate in chemotherapy-alone patients was 48.6% (34.3% in BRPC, 54.5% in LAPC). Node-negative resections were achieved in 71.7% of patients who received SBRT (60.0% in BRPC, 79.2% in LAPC) and in 45.9% of patients who received chemotherapy alone (50.0% in BRPC, 36.4% in LAPC). The pCR rate was 10.3% in the SBRT group (6.7% in BRPC, 12.5% in LAPC) and 2.7% in the chemotherapy-alone group (0% in BRPC, 9.1% in LAPC). The npCR rate was 23.1% in the SBRT group (20.0% in BRPC, 25.0% in LAPC) and 5.4% in the chemotherapy-alone group (7.7% in BRPC, 0% in LAPC). CONCLUSIONS: Selected patients who are initially deemed unresectable may now undergo resection after receiving neoadjuvant induction chemotherapy and SBRT. Furthermore, improved surgical outcomes are observed with neoadjuvant SBRT in comparison with neoadjuvant chemotherapy alone. Longer follow-up is needed to validate its impact on survival. Lauren M. Rosati, BS, Jin He, MD, Shalini Moningi, BA, Amy HackerPrietz, PAC, Thomas A. Huebner, MD, Daniel A. Laheru, MD, John L. Cameron, MD, Timothy M. Pawlik, MD, MPH, PhD, Matthew J. Weiss, MD, Christopher L. Wolfgang, MD, Joseph M. Herman, MD, MSc; Johns Hopkins University School of Medicine (S022) Experimental Insight Into the Preferential Cytotoxicity of Cancerous vs Noncancerous Cells of Metformin BACKGROUND: Recent prospective data demonstrate that stereotactic body radiation therapy (SBRT) is safe and effective in locally advanced pancreatic cancer (LAPC); however, little is known regarding the role of SBRT in the neoadjuvant setting. This study compared the role of neoadjuvant chemotherapy with and without SBRT in patients with borderline resectable pancreatic cancer (BRPC) or LAPC. PURPOSE AND OBJECTIVE: Metformin is known to produce direct cancer cell cytotoxicity, inhibit tumor growth, radioprotect normal tissues, and radiosensitize cancer cells. However, the underlying mechanism behind the anticancer activity is not yet understood. This study was designed to investigate the biochemical effect of metformin on cancerous and noncancerous cells under varying conditions from the perspective of reactive oxygen species (ROS). METHODS: All patients who underwent surgical resection following chemotherapy alone or induction chemotherapy followed by SBRT (SBRT group) were retrospectively reviewed. Disease stage was determined by multidisciplinary review. Chemotherapy regimens were heterogeneous; the cumulative SBRT dose range was 25–33 Gy in five fractions. Pathologic complete response (pCR) was defined as no residual tumor, and near pCR (npCR) was defined as microscopic foci of single cells of adenocarcinoma scattered among an area of dense fibrosis. MATERIALS AND METHODS: A549 (lung cancer), MCF-7 (breast cancer), and MCF-10A (normal breast) cell lines were maintained in cell culture and exposed to 5 mM metformin for 4–6 hours in environments in which oxygen and glucose concentrations varied. The level of ROS production was measured using a standard dihydroethidium (DHE) fluorescence reagent and flow cytometric spectroscopy. Oxygen levels were maintained with atmospheric oxygen or a mixture of carbon dioxide/nitrogen gas. Clonogenic and MTT assays were used to assess cellular response to radiation therapy. RESULTS: Among 76 resected patients with BRPC or LAPC, 37 received chemotherapy alone and 39 received SBRT. Median age was 60.4 years (range: 44.2–83.6 yr) and 64.4 years (range: 39.2–83.2 yr) in the SBRT group and chemotherapy-alone group, respec- RESULTS: Metformin consistently and significantly increased ROS levels at least 2-fold in A549 and 3-fold in MCF-7 cell lines. ROS levels in MCF-10A cells were only moderately affected. Hypoxic A549 cells treated with metformin showed a 137% increase in ROS 10 Derek Isrow, MD, PhD, Karen Lapanowski, Andrew Kolozsvary, Stephen Brown, PhD, Jae Ho Kim, MD, PhD; Henry Ford Hospital American Radium Society Scientific Papers and Posters 2015 levels over baseline hypoxic cells without metformin (165% at 6 hr). Oxygenated A549 cells treated with metformin showed a 117% ROS increase compared with control cells without metformin (139% at 6 hr). When A549 cells were placed in glucose-free medium (GFM) and exposed to metformin, ROS levels were 203% greater than A549 cells in GFM without metformin. Hypoxic A549 cells in GFM with metformin produced a ROS level 120% greater than cells in hypoxic GFM without metformin. Cellular response studies confirmed the ROS results. Increased cytotoxicity to radiation in the presence of metformin was observed with A549 cells by clonogenic assay and with MCF-7 cells by MTT assay. In agreement with the ROS studies, no increased radiosensitization or cytotoxicity was observed in MCF-10A cells using metformin. CONCLUSIONS: Metformin was demonstrated to increase ROS levels, cytotoxicity, and radiosensitization in A549 and MCF-7 cancer cells under oxic, hypoxic, glucose-full, and glucose-free conditions. Normal MCF-10A cells did not show increased toxicity with metformin. The preferential metformin-induced increase in ROS levels found in cancer cells, particularly hypoxic cells, may provide some explanation for the therapeutic benefit seen in diabetic patients taking metformin while undergoing cancer treatment. Further studies are underway in the hope of translating these findings into clinical practice. (S023) Ultrasound Tissue Characterization of Breast Fibrosis Following Hypofractionated Breast Radiotherapy Xiaofeng Yang, Mylin Torres, MD, Simone Henry, Donna Mister, Tatiana Han, Walter Curran, Tian Liu; Emory University PURPOSE: Hypofractionation whole-breast radiotherapy (RT) following conservation surgery has been used in many institutions for several decades. However, the hypofractionation regimen has not been widely accepted in the United States. Criticism has focused on concerns about late normal-tissue effects. The purpose of this study is to evaluate breast fibrosis in patients receiving hypofractionated RT using quantitative ultrasound tissue characterization. MATERIALS AND METHODS: Twenty postlumpectomy women were enrolled in a prospective, longitudinal ultrasound imaging study to examine the development of RT-induced normal-tissue toxicity. Each received simultaneous integrated boost (SIB) RT (40 Gy plus simultaneous cavity boost to 48 Gy in 15 fractions). A baseline ultrasound scan was performed 1 week prior to RT, plus one scan during RT, followed by evaluation at the 1-year follow-up. Patients were imaged with 10-MHz ultrasonography at the 12, 3, 6, and 9 o’clock positions of the breast. To compensate for individual variation, scans taken on the untreated contralateral breast served as a control. Ultrasonography parameters, including skin thickness, Pearson coefficient, and midband fit, were calculated to quantify cutaneous and glandular tissue toxicity. Quantitative ultrasonography parameters were generated by normalizing to the corresponding contralateral breast. Ultrasonography findings were compared with clinical assessments based on the Radiation Therapy Oncology Group (RTOG) toxicity scheme. RESULTS: Based on the ultrasonography measurements, 50% to 60% of patients developed mild to moderate late toxicity (fibrosis) 1-year post-RT. For the 20 patients, the average skin thickness ratios were 1.23 prior to RT, 1.33 during RT, and 1.37 at the 1-year followup. Prior to RT, the diseased breast experienced 23% greater skin thickening secondary to postsurgical changes. The midband fit ratios were 1.12 prior to RT, 1.70 during RT, and 2.45 1-year postRT. The Pearson ratios were 0.92 prior to RT, 0.78 during RT, and 0.63 1-year post-RT. Increased skin thickness and midband fit values correlated with increased cutaneous toxicity assessed using RTOG scores (P < .05). The ultrasonography findings were consistent with the clinical assessments: 60% developed RTOG grade 1 toxicity, and 10% developed grade 2 toxicity at 1 year. CONCLUSIONS: This is the first prospective imaging study to objectively document normal-tissue toxicity in patients treated with hypofractionation breast RT using ultrasonography tissue characterization. Contrary to the criticism, patients receiving hypofractionation RT recovered better and experienced less late toxicity at the 1-year follow-up. With this noninvasive ultrasonography technology, physicians can objectively compare the late effects of hypofractionation with the standard regimen and gain a better understanding of treatment responses in individual patients. (S024) Impact of Pelvic Radiotherapy on Sexuality Reported by Women Surviving Cancer James M. Metz, MD, Erin Davis, Carolyn Vachani, Margaret K. Hampshire, Christine E. Hill-Kayser, MD; University of Pennsylvania INTRODUCTION: Many women are cured of pelvic malignancies through multimodality treatment. Some treatments, including radiation, are recognized to impact female sexual function in the long term; however, patient perspectives on these outcomes are not well understood, nor are contributions of other treatment forms. Here, we describe patient-reported outcomes after pelvic radiation, as well as radiation elsewhere and chemotherapy alone. METHODS: Patient-reported data were gathered via convenience sample frame from cancer survivors voluntarily using an internet tool for the creation of survivorship care plans. The tool requires entry of data regarding diagnosis, demographics, and treatments and provides customized guidelines for future care. It is publically available and free of charge, with > 300,000 total users since 2007. During use of the tool, survivors are queried regarding their experience with late effects associated with specific treatments. RESULTS: The tool was used by 7,258 female cancer survivors from November 2012–October 2014. Overall, the group was 82% Caucasian, with a median age of 51 years (range: 18–84 yr) and a median time of 2.5 yr from diagnosis (range: < 1–38 yr). Of these patients, 503 had received pelvic radiation as part of treatment for gastrointestinal or gynecologic malignancies (37% vs 63%, respectively). Of the pelvic radiation group, 78% reported receiving external beam radiation (EBRT) with x-rays, 35% received brachytherapy (55% in conjunction with EBRT and 45% alone), and 3% received proton radiation. When the entire group was queried, “Have you experienced sexual changes since completing treatment?” 42% responded “yes,” 23% answered “no,” and 17% responded “I don’t know.” The brachytherapy group was more likely to respond “yes” (56%) than the EBRT-alone group (40%) (P < .05). ARS PROCEEDINGS 2015 11 American Radium Society Scientific Papers and Posters 2015 The pelvic radiation group was compared with patients receiving radiation outside of the pelvis. Of this group, 34% responded “yes,” 32% responded “no,” and 34% responded “I don’t know” (P < .05 when compared with the pelvic radiation group). The pelvic radiation group was also compared with patients receiving chemotherapy alone, with 38% responding “yes,” 29% responding “no,” and 33% responding “I don’t know” (P < .05). CONCLUSIONS: Reports of sexual changes in women after cancer treatment are highest in patients receiving pelvic radiation, particularly brachytherapy. Rates of sexual changes are higher than expected in patients receiving radiation elsewhere and/or chemotherapy alone. These findings support the need for both adequate patient counseling prior to treatment and support for late-effect management afterwards. (S025) Discovery and Validation of Predictive Factors for Safety Incident Reports in Patients Receiving Radiation Therapy: Comparative Results From a Large Safety Variance Database Shereef M. Elnahal, MD, MBA, Amanda Blackford, SCM, Eric C. Ford, PhD, Annette N. Souranis, Valerie Briner, Koren Smith, PhD, Todd R. McNutt, PhD, Theodore L. Deweese, MD, Jean Wright, MD, Stephanie A. Terezakis, MD; Johns Hopkins University School of Medicine; Department of Radiation Oncology, University of Washington; Department of Radiation Oncology, Johns Hopkins Hospital OBJECTIVES: Patient safety is a vital concern in radiotherapy, but little is known about the factors that predict patients at risk for safety incidents. We aimed to: a) describe patients who have experienced safety incident reports; and b) compare them with a cohort without events to identify predictive factors and build a nomogram tool to identify patients at risk. MATERIALS AND METHODS: We used our institution’s radiation treatment safety incident reporting system to build a database of patients who experienced events from January 2011–October 2014. Patient and treatment-specific data were reviewed from our institution’s radiation management software program for a random selection of patients with events in our reporting system. A control group of sequential patients in 2014 was generated for comparison. Summary statistics, likelihood ratios, and mixed-effect logistic regression models were used for group comparisons. RESULTS: The safety incident group comprised 1050 events. Errors in treatment planning (eg, incorrect planning target volume [PTV] margins submitted) (35.2%) and documentation (eg, wrong patient name denoted) (34.9%) were the most common incident types, followed by communication (17.5%) and treatment delivery (12.4%). Several clinical and treatment-specific factors correlated with safety reports. Incidents were much more frequently reported in minors (age < 18 yr) than in adults (25.3% vs 5.4%; P < .001). Patients with head and neck (16% vs 8%; P < .001) and breast (20% vs 15%; P = .03) primaries were more frequently associated with an incident. Larger tumors (19% vs 11% had T4 lesions; P = .02), a greater mean number of fractions (24 vs 20; P = .02), and cases on protocol (9% vs 5%; P = .005) or with intensity-modulated radiation therapy (IMRT)/image-guided IMRT (IGMRT) (52% vs 43%; P = .001) were 12 associated with incident events. Of note, IMRT was used in 70.5% of minors treated. Other independently associated factors were found in patients with T4 tumors (19.5% protocol, 50.7% head and neck, mean of 24 fractions). On multivariate analysis, dose per fraction, head and neck cases, pediatric patients, protocol enrollment, and average pain score independently predicted for safety events. CONCLUSIONS: We determined several patient and treatment-specific characteristics that predicted for treatment incident events. Children were more susceptible to incident alerts, possibly related to increased staff vigilance or more frequent use of complex modalities. Patients with head and neck tumors, greater number of fractions, treatment on protocol, and use of IMRT also predicted for incidents. Independently predictive variables have been built into a nomogram that will be validated in a prospective dataset, the results of which will be available at the time of the meeting. (S026) A Novel Method for Detecting Serious Cardiac Device Errors in Patients Receiving Radiotherapy Using Daily Pulse Checks Jonathan D. Grant, MD, Anne Dougherty, MD, Karimzad Kaveh, MD, Daniel R. Gomez, MD, Marc A. Rozner, PhD, MD; UT MD Anderson Cancer Center INTRODUCTION: Little consensus exists for the proper monitoring of cardiac implantable electronic devices (CIEDs) in patients undergoing radiotherapy. We present data supporting a novel, easy, safe, and cost-effective method for detecting serious CIED problems arising during radiotherapy using daily heart rate (HR) monitoring. MATERIALS AND METHODS: Since all CIEDs default to pacing at 60–72 beats per minute (bpm) without rate-response function (RRF) upon detection of a serious error, we instituted the following daily posttreatment pulse check protocol in 2010. Any patient who can be safely paced at a lower set rate of 75 bpm with RRF undergoes reprogramming at his preradiation CIED interrogation. Following each fraction of radiotherapy, these patients rest for 5 minutes and then undergo a pulse check. An HR of ≥ 74 bpm verifies the absence of a serious error. Because paroxysmal ventricular contractions may permit perfused pulses that register < 74 bpm, “tapping” on the CIED ensues to simulate patient exercise. An HR increase verifies the appropriate programming and the absence of a serious error. In a patient with an HR ≤ 73 and nonresponse to the tapping maneuver, a serious error is suspected, and a CIED interrogation is performed. At the completion of radiotherapy, device interrogation is performed for all patients, with return of the CIED to the appropriate settings. RESULTS: Between February 2010 and October 2013, a total of 36 patients with CIEDs received treatment at the proton therapy center, 29 (81%) of whom were eligible for the pulse protocol. We found four patients (14%) with five serious CIED errors, manifested as a device parameter reset. A total of 845 fractions of radiotherapy were delivered in this cohort, giving an error rate per fraction of 0.6%. All affected patients were detected by a decrease in the measured HR of ≤ 73. No additional resets were discovered on any other device at routine posttreatment interrogation. No adverse events were experienced by any patient as a result of the increased lower pace rate. In one patient American Radium Society Scientific Papers and Posters 2015 with HR < 74 bpm, the “tapping maneuver” confirmed the absence of a reset and prevented an unnecessary device interrogation. CONCLUSIONS: A novel method for monitoring CIEDs during radiotherapy is presented, which we report as effective, easy, and lacking adverse side effects. This simple technique is a cost-effective alternative to frequent device interrogations during the course of radiotherapy and allows for consistent daily monitoring. (S027) Novel Mechanisms of Adaptive Resistance in Head and Neck Cancer Vinita Takiar, MD, PhD, Jennifer B. Dennison, PhD, Shreya Mitra, PhD, Dominic Ma, BS, Jack Phan, MD, PhD, Gordon B. Mills, MD, PhD; UT MD Anderson Cancer Center OBJECTIVE: Locoregional recurrence after definitive radiation therapy (XRT) for head and neck (H&N) cancer is a significant clinical challenge, with many patients failing after combined doses > 100 Gy. Although multimodality treatment is prescribed for many cancers, such treatment algorithms are generally empiric, with no identified molecular basis. The efficacy of XRT is limited by intrinsic and acquired radioresistance. We hypothesize that resistance to XRT is, at least in part, mediated by adaptive signaling events induced by the XRT itself. Using signal transduction to better understand these adaptive responses is therefore crucial to developing rational treatment combinations that increase the therapeutic efficacy of XRT. MATERIALS AND METHODS: Two representative H&N cell lines (FaDu and HN5) were grown in both two dimensions (2D) and three dimensions (3D) (spheroids suspended in Matrigel). Based on cell survival assays conducted in 2D culture, a dose of 4 Gy (~LD50) was chosen for subsequent experiments. Cells in both 2D and 3D were exposed to either 0 Gy or 4 Gy with a cobalt-60 source and then subjected to high-throughput reverse phase protein analysis (RPPA) to screen for differential protein expression in cells able to survive radiation. As the expression of multiple energy-related metabolites was increased, oxygen consumption rate (OCR), which is a readout of aerobic respiration, was subsequently analyzed using an extracellular flux analyzer. RESULTS: Both HN5 and FaDu cells formed intact monolayers in 2D and spheroids in 3D. Radiation of 4 Gy in one fraction resulted in differential expression of multiple signaling pathways in both cell lines, with significant upregulation of multiple components of the mammalian target of rapamycin (mTOR) pathway (P < .05), including pACC, pAMPK, pAkt, and p70S6K—collectively suggesting altered energy metabolism. Glutamine metabolism, regulated by the mTOR pathway, serves as an anaplerotic and limiting reagent for citric acid cycle progression and thus serves as a proof-of-concept pathway. Using a glutaminase-specific inhibitor that prevents the conversion of glutamine to glutamate, we report that the cellular OCR was significantly reduced (P < .001) in the presence of glutamine. However, the OCR was unchanged in the presence of glucose alone, suggesting that glutamine metabolism is indeed the driver of aerobic respiration in these cells. CONCLUSIONS: Targeted combinatorial therapy with XRT is neces- sary to overcome adaptive radioresistance. RPPA is a powerful pro teomic platform, suggesting alterations in energy metabolism following XRT that are targetable by inhibition of the enzyme glutaminase. Further in vivo experiments with glutaminase inhibition and xenograft models to assess combinatorial efficacy with radiation are warranted. (S028) Patient-Level DNA Damage and Repair Pathway Profiles Are Prognostic After Prostatectomy for High-Risk Prostate Cancer Joseph R. Evans, MD, PhD, Shuang Zhao, MSE, Scott A. Tomlins, MD, PhD, Karen E. Knudsen, PhD, Eric A. Klein, MD, Robert B. Den, MD, R. Jeffrey Karnes, MD, Elai Davivioni, PhD, Felix Y. Feng, MD; University of Michigan; Thomas Jefferson University; Cleveland Clinic; Mayo Clinic; GenomeDx Biosciences Inc. PURPOSE/OBJECTIVES: A substantial number of patients who are diagnosed with high-risk prostate cancer are at risk for metastatic progression after primary treatment. Better biomarkers are needed to identify patients who are at the highest risk so that therapy can be intensified. MATERIALS AND METHODS: We developed a novel patient-level DNA damage and repair (DDR) pathway profiling approach in a cohort of 1,090 patients undergoing prostatectomy for high-risk prostate cancer who had long follow-up (mean, 7–13 yr). Highdensity microarray gene expression data from prostatectomy samples were analyzed for DDR pathway enrichment with the gene set enrichment analysis algorithm and clustered hierarchically. The prognostic performance of pathway-based biomarkers for metastatic progression and overall survival (OS) was analyzed with univariate and multivariate logistic regression models. RESULTS: We found that there were distinct clusters of DDR pathway patterns within the cohort. DDR pathway enrichment correlated only weakly with the clinical variables age, Gleason score, and PSA (Spearman’s rho ≤ 0.20) and correlated much more strongly with AR pathway enrichment (Spearman’s rho up to 0.82). Many of the individual pathways were significantly associated with increased metastatic progression risk (odds ratio [OR] = 1.28–1.78) and worse OS (OR = 1.51–2.23) on univariate and multivariate analysis, and a composite DDR pathway biomarker showed improved prognostic performance for metastatic progression risk (OR = 2.09 [1.33, 3.28]). Interestingly, this prognostic significance was restricted to patients aged < 70 years. CONCLUSIONS: Patient-level DDR pathway profiling revealed distinct clusters. Individual DDR pathways and a composite biomarker showed strong prognostic performance with the long-term outcomes of metastatic progression and OS, which may be useful for risk stratification of high-risk prostate cancer patients aged < 70 years. (S029) Evaluating the Role of a 21-Gene Expression Assay in Directing Adjuvant Radiotherapy Decisions for Elderly Women With Early-Stage Breast Cancer Charles E. Rutter, MD, Nataniel H. Lester-Coll, MD, Suzanne B. Evans, MD, MPH; Department of Therapeutic Radiology, Yale University School of Medicine ARS PROCEEDINGS 2015 13 American Radium Society Scientific Papers and Posters 2015 Objective: Adjuvant breast radiotherapy (RT) after breast-conserving surgery (BCS) in elderly women with early-stage breast cancer (ESBC) confers a local control benefit but no improvement in overall survival. Hence, the role of RT in this population is uncertain. In the separate context of adjuvant chemotherapy recommendations for node-negative ESBC, gene expression testing has allowed risk-adapted decision making and has had a marked impact on clinical practice. We sought to determine the potential role and costeffectiveness of gene expression testing in directing adjuvant RT decisions for elderly ESBC patients, given a published relationship of Oncotype DX (ODX) results with local-regional recurrence risk. METHODS: The distribution of ODX recurrence scores (RSs) among elderly women (aged ≥ 70 yr) with estrogen receptor–positive ESBC (pT1N0M0) post-BCS was defined within the National Cancer Data Base (NCDB), using standard definitions of low (< 18), intermediate (18–30), and high risk (> 30). Next, a decision tree was constructed to determine the overall cost-effectiveness of ODX in directing RT recommendations. Local-regional recurrence estimates at 10 years were extrapolated from the literature and assumed to be 1%, 3%, and 5% in patients with low, intermediate, and high RSs, respectively, following BCS, RT, and 5 years of tamoxifen. Rates of adjuvant radiotherapy utilization for each RS stratum were based upon published patterns of RS-guided adjuvant chemotherapy utilization. RESULTS: We identified 2,564 elderly ESBC patients treated between 2007 and 2011 within the NCDB with ODX RS results. Median age and tumor size were 73 years (range: 70–90 yr) and 1.2 cm (range: 0.1–2 cm), respectively. Median RS was 15 (standard deviation [SD] = 8.3). RS was low, intermediate, and high in 69.2%, 28.3%, and 2.5% of patients, respectively. Neither age nor tumor size was associated with RS. ODX-directed adjuvant radiotherapy recommendations resulted in a net savings of $1,163.05 per patient and a utility decrease of 0.024 quality-adjusted life years (QALYs) relative to current clinical practice, yielding an incremental cost-effectiveness ratio for ODX-directed adjuvant radiotherapy of $47,920.78/QALY. CONCLUSIONS: We observed a wide range of RSs among elderly ESBC patients in the NCDB. Further research is needed to determine if gene expression assays are capable of risk-stratifying within this population, as a means for directing adjuvant RT recommendations after BCS. Should this be proven, our findings would suggest that ODX-directed decision-making is theoretically cost-effective at a conservative willingness-to-pay threshold of $50,000/QALY. varying results, with some showing equivalence and others favoring M. Outcomes for BCT have improved significantly over time. The most appropriate comparison for BCT and M should reflect treatment techniques from the modern era. We hypothesize that in the modern treatment era, BCT and M have equivalent outcomes. METHODS: Breast cancers diagnosed between 1975 and 2012 in women aged ≤ 40 years were collected from our institutional tumor registry. Four cohorts were then created based on year of diagnosis and treatment approach: BCT before 2000, BCT during or after 2000, M before 2000, and M during or after 2000. Vital status and cause of death were obtained from the government cancer registry. Kaplan-Meier survival estimates were used to analyze the primary outcome (freedom from locoregional recurrence [FFLR]) and secondary outcomes (overall survival [OS] and relapse-free survival [RFS]). RESULTS: A total of 427 BCT candidates were identified from our institutional tumor registry. Among those who received BCT, OS and RFS were similar for those diagnosed before and after 2000. FFLR, however, was significantly improved after 2000. FFLR at 10 years was 78.8% vs 95.4% for BCT before and after 2000, respectively (P = .03). Among those who received M, OS (P = .03) and RFS (P < .0001) were significantly improved with diagnosis after 2000. FFLR at 10 years was 85.6% vs 93.1% for M before and after 2000, respectively (P = .08). When comparing BCT and M after 2000, FFLR, RFS, and OS were all similar. FFLR at 10 years was 95.4% vs 93.1% for BCT and M, respectively (P = .75). When comparing BCT and M before 2000, OS and RFS rates were similar. CONCLUSIONS: Outcomes for women with breast cancer aged ≤ 40 years undergoing BCT and M have improved significantly over time, likely reflecting the advent of improved local and systemic therapies. In addition, for women treated after 2000, BCT appears to be safe and equivalent to M at 10 years in terms of FFLR, OS, and RFS. (S031) Dosimetric Analysis of Left-Sided Breast Cancer Treatment With Tomotherapy IMRT, IMRT, VMAT, and 3D-CRT Twisha Chakravarty, MD, Eugene C. Endres, Brent Parker, PhD, DABR, Sandra Hatch, MD; University of Texas Medical Branch Jonathan Frandsen, MD, David Ly, MD, MPA, Gita Suneja, MD, MPH, Cindy Matsen, MD, Matthew Poppe, MD; Department of Radiation Oncology, Department of General Surgery, Huntsman Cancer Hospital, University of Utah School of Medicine OBJECTIVES: Previous studies have demonstrated that the risk of ischemic heart disease is increased as a result of exposure to ionizing radiation in women treated for breast cancer. Alternative radiation techniques, such as intensity-modulated radiation therapy (IMRT), volumetric-modulated arc therapy (VMAT), and TomoTherapy IMRT, have been shown to improve dosimetric parameters of the heart and substructures. However, these techniques have not been compared with each other to potentially guide treatment decisions. INTRODUCTION: Women aged < 40 years with breast cancer have worse outcomes compared with older women. Mastectomy (M) is often recommended as primary therapy, even for women with earlystage disease, because young women have been underrepresented in the landmark trials showing equivalence of breast conservation therapy (BCT) and M. Retrospective series comparing BCT and M show METHODS: Treatment plans from 10 patients treated to the whole breast for left-sided breast cancer in 2014 were collected. Treatment plans using TomoTherapy IMRT, IMRT, VMAT, and three-dimensional (3D) conformal radiation therapy with TomoDirect and opposed tangents were generated for each patient. Dosimetric parameters for the heart, left ventricle, and left anterior descending (S030) Breast Conservation in Young Women in the Modern Era 14 American Radium Society Scientific Papers and Posters 2015 artery (LAD), including V2, V5, V10, Dmax, and mean doses, were collected and analyzed using paired t-test and analysis of variance (ANOVA). RESULTS: For V2 of the LAD, no statistically significant difference was found between TomoTherapy IMRT and 3D conformal plans (67.2% vs 70.5% vs 75.0%; P = .49); however, TomoTherapy IMRT plans had significantly reduced V5 of the LAD when compared with 3D conformal plans, as well as IMRT and VMAT (2.0% vs range: 12.0–49.0%; P < .01). With regard to the heart and left ventricle, the V2 was significantly lower for 3D conformal plans vs TomoTherapy IMRT, VMAT, and IMRT (P < .01). This difference was maintained for V5 for the heart but not for the left ventricle. Across all three cardiac parameters, the average Dmax was significantly lower with TomoTherapy IMRT plans when compared with all other planning techniques (P < .01). CONCLUSIONS: Our analysis of this single-institution population of women with left-sided breast cancer treated to the whole breast demonstrates that differing radiation treatment techniques have statistically significant impacts on dosimetric parameters of the heart, left ventricle, and LAD. TomoTherapy IMRT was shown to be superior for reducing low-dose radiation to the LAD and maximum dose to all cardiac structures, particularly when compared with VMAT and IMRT. No significant difference was demonstrated in low-dose radiation exposure from treatment with TomoDirect vs opposed tangents. These results provide insight into treatment differences that may help guide clinical practice in the future, with an emphasis on reducing long-term patient toxicity. (S032) Complications of Contralateral Prophylactic Mastectomy With Tissue Expander Reconstruction and Potential Impact on Adjuvant Oncologic Therapy Rahul D. Tendulkar, MD, Neil M. Woody, MD, Mihir Naik, DO, Chandana A. Reddy, MS, Paul Durand, MD, Adekunle Elegbede, Eliana Duraes, Stephen R. Grobmyer, MD, Joseph P. Crowe, MD, Risal Djohan, MD; Cleveland Clinic PURPOSE: One concern of contralateral prophylactic mastectomy (CPM) is that a complication may delay adjuvant chemotherapy or radiotherapy for breast cancer. We report acute (< 6 mo), delayed (> 6 mo), and total complication rates in patients undergoing synchronous bilateral mastectomies and tissue expander reconstruction (TER), comparing complications between the ipsilateral (therapeutic) side and the contralateral (prophylactic) side within a perfectly matched cohort. METHODS: We conducted a retrospective review of breast cancer patients treated at Cleveland Clinic from 2000–2007, and 88 women undergoing synchronous bilateral mastectomies with TER for unilateral invasive breast cancer were identified. Complications were defined as hematoma or seroma requiring reoperation, blood transfusion, capsular contracture (Baker grade III–IV), wound infection requiring intravenous (IV) antibiotics or implant removal, wound dehiscence, implant leak, and extrusion. The timing and laterality of complications were determined to compare the therapeutic and prophylactic sides. RESULTS: The median age was 43 years (range: 23–63 yr), and the median follow-up was 6.1 years. Overall, 64 patients (73%) were premenopausal, 39 (44%) were lymph node-positive, 18 (20%) received preoperative chemotherapy, 57 (65%) received postoperative chemotherapy, and 24 (28%) received postoperative radiation therapy to the ipsilateral side. Twelve patients (14%) had received prior breast radiotherapy and underwent mastectomy for recurrence. Total complications on the therapeutic side were more common than on the prophylactic side, including wound dehiscence (4% vs 1%; P = .17), capsular contracture (13% vs 5%; P = .06), implant extrusion (10% vs 2%; P = .03), wound infection (20% vs 9%; P = .04), any complication (38% vs 25%; P = .08), and reoperation for a complication (34% vs 21%; P = .048). On the therapeutic side, 23 of 36 (64%) radiated TER patients experienced a complication compared with 10 of 52 (19%) nonirradiated patients (P < .0001). Delayed complications were more common on the therapeutic side than on the prophylactic side (24% vs 13%; P = .04), but acute complication rates were similar (14% vs 13%, respectively; P = .8). Among 11 patients with an acute complication of a CPM, receipt of postoperative chemotherapy or radiotherapy was actually delayed by a complication in 2 of 63 (3%) patients who received adjuvant treatment. CONCLUSIONS: CPM with TER resulted in a complication in 25% of patients, half of which occurred within 6 months of surgery. However, adjuvant chemotherapy or radiotherapy was delayed by acute complications in only 3% after CPM. Delayed complication rates were higher on the therapeutic side, possibly due to effects of radiotherapy. These data may better inform patients considering bilateral mastectomies with TER. (S033) Predictive Capacity of Three Comorbidity Indices in Estimating Survival Endpoints in Women With Early-Stage Endometrial Carcinoma Sean Vance, MD, Meredith Mahan, Mohamed Elshaikh, MD; Henry Ford Hospital PURPOSE/OBJECTIVES: The negative impact of medical comorbidity on survival endpoints in women with endometrial carcinoma (EC) is well known. Few validated comorbidity indices are available for clinical use—eg, the Charlson Comorbidity Index (CCI), ageadjusted Charlson Comorbidity Index (AACCI), and Adult Comorbidity Evaluation-27 (ACE-27). The study goal is to evaluate which index correlates best with survival endpoints in women with early-stage EC. MATERIALS AND METHODS: For this institutional review board (IRB)-approved study, our prospectively maintained database of 1,920 women with endometrial cancer was reviewed. We identified 1,132 women with endometrioid carcinoma International Federation of Gynecology and Obstetrics (FIGO) stages I–II who underwent hysterectomy from 1987–2011. The three comorbidity indices at the time of hysterectomy were retrospectively calculated by physician chart review. Univariate and multivariate modeling with Cox regression analysis was used to determine the significant predictors of survival endpoints. Kaplan-Meier and log-rank test methods were used to evaluate survival outcomes. ARS PROCEEDINGS 2015 15 American Radium Society Scientific Papers and Posters 2015 RESULTS: After a median follow-up of 60 months, 262 deaths were recorded (42 from EC [16%] and 220 [84%] from other causes). For each of the studied comorbidity indices, the highest scores correlated significantly with poorer overall survival (OS). The hazard ratio of death from any cause was 3.92 (95% confidence interval [CI], 2.95–5.20) for AACCI, 2.25 (95% CI, 1.73–2.94) for CCI, and 1.57 (95% CI, 1.23–2.01) for ACE-27. Lymphovascular space involvement, tumor grade, lower uterine segment involvement, and AACCI were independent predictors of OS. None of the three comorbidity indices was significantly predictive of disease-specific or recurrence-free survival. ropathy, 4/105 (3.8%) metabolic, and 1/105 (0.95%) venous thromboembolism. Dose reduction occurred in 6/115 cycles (5.2%), and dose delay occurred in 12/115 cycles (10.4%). Since the study onset, 1 of 21 pts had a recurrence (lung) and died of disease. CONCLUSIONS: While all three comorbidity indices correlated significantly with OS in women with early-stage EC, AACCI was the only independent predictor of OS and should be considered for evaluating comorbidity in future endometrial cancer patients. (S035) Cervical Cancer Outcome Prediction to HighDose-Rate Brachytherapy Using Quantitative Magnetic Resonance Imaging Analysis of Tumor Response to External Beam Radiotherapy (S034) Pilot Phase II Trial of “Sandwich” Radiation and Combination Carboplatin and Paclitaxel Chemotherapy in Patients With High-Risk Endometrial Cancer Lori Spoozak, MD, MHS, Divya Gupta, MD, Laura Reimers, MPH, Jennifer Jorgensen, MD, Rafi Kabarriti, MD, Keyur Mehta, MD, Gary Goldberg, MD, Mark Einstein, MD, MS, Dennis Yi-Shin Kuo, MD; Albert Einstein College of Medicine; Weill Cornell Medical College OBJECTIVES: We sought to determine the toxicity, tolerability, and outcome of radiation therapy (RT) sandwiched between combination chemotherapy in patients (pts) with high-risk endometrial cancer. CONCLUSIONS: Sandwich chemotherapy and radiation therapy is a tolerable treatment modality for patients with high-risk endometrial cancer. Rates of hematologic toxicities are acceptable, and nonhematologic toxicities are uncommon. Further enrollment of patients is underway to determine efficacy. Beant S. Gill, MD, David Minkoff, BA, Hayeon Kim, MS, DABR, Christopher Houser, MS, Sushil Beriwal, MD; Magee-Womens Hospital, University of Pittsburgh Medical Center PURPOSE: Image-guided brachytherapy (IGBT) has been shown to improve outcomes for cervical cancer. Integration of magnetic resonance imaging (MRI) allows visualization of residual disease. In order to assess tumor regression and outcomes, a volumetric analysis was conducted among patients treated with MRI-based IGBT. METHODS: High-risk endometrial cancer was defined as surgically staged endometrioid adenocarcinoma, IA with grade (G) 3 tumor and with lymphovascular space involvement, IB with G2 or 3 tumor, and stage II–IV disease, any grade. Treatment involved paclitaxel (T) (175 mg/m2) and carboplatin (C) (area under the concentration-time curve [AUC] = 6.0 or 6.5) every 21 days × 3 doses, followed by pelvic external beam radiation therapy (45 Gy) and brachytherapy (15 Gy). Per protocol, brachytherapy only (25 Gy) was administered to pts with stage IA G3 and IB G2 disease. Three additional cycles of T/C (AUC = 5) were administered after RT. Toxicity was graded by Common Terminology Criteria for Adverse Events version 4.0 (CTCAE v4.0). Analysis of toxicities was performed on all evaluable pts. METHODS: Consecutive patients with International Federation of Gynecology and Obstetrics (FIGO) stage IB1–IVA cervical cancer receiving definitive chemoradiation from 2007–2013 were identified. Patients were excluded if they were undergoing perineal template-based interstitial brachytherapy or if MRI was not completed prior to therapy and at first brachytherapy application. All patients were treated using a ring and tandem with or without interstitial needles. High-dose-rate brachytherapy was delivered in five onceor twice-weekly fractions of 5–6 Gy/fraction. Optimization was completed to meet the following cumulative equivalent 2-Gy doses (EQD2): high-risk clinical target volume (HRCTV) 75–85 Gy and 2 cc of rectum ≤ 70 Gy, sigmoid ≤ 70 Gy, and bladder ≤ 85 Gy. T2-weighted imaging using 1.5-T MRI was completed following brachytherapy applicator insertion. Gross tumor volumes (GTVs) were retrospectively contoured and defined: GTV prior to therapy (GTVPre-EBRT), GTV at first application (GTVIGBT), and percentage residual GTV at first application (GTV%Residual). RESULTS: A total of 21 pts were enrolled between 2008–2014. The median age was 62 years (range: 48–74 yr): 66.7% (14/21) pts had stage I/II disease, while the remaining 33.3% (7/21) had stage IIIC disease; 42.9% (9/21) had International Federation of Gynecology and Obstetrics (FIGO) G1 disease, and 23.8% (5/21) and 33.3% (7/21) had G2 and 3 disease, respectively; 66.7% (14/21) pts completed the treatment per protocol; and 76.2% (16/21) pts completed six cycles of chemotherapy. One pt refused chemotherapy after enrollment, and four pts could not complete chemotherapy due to hematologic toxicities. Of 115 evaluable chemotherapy cycles, the G3 or 4 combined hematologic toxicity events were as follows: 33/115 cycles (28.7%) neutropenia, 9/115 cycles (7.8%) thrombocytopenia, and 10/115 cycles (8.7%) anemia. Of 105 evaluable chemotherapy cycles, the G3 or 4 combined nonhematologic events were as follows: 2/105 cycles (1.9%) infection, 1/105 cycles (0.95%) neu- RESULTS: Eighty-four patients were identified. The majority had FIGO stage IIB disease (57.1%) and squamous histology (82.1%) and received median external beam and brachytherapy doses of 45.0 Gy and 27.5 Gy, respectively. With a 20.8-month (range: 3–74 mo) median follow-up, the 2-year Kaplan-Meier estimates of local control (LC), disease-free survival (DFS), and overall survival (OS) were 91.3%, 79.8%, and 85.0%, respectively. Median GTVPre-EBRT, GTVIGBT, and GTV%Residual values were 31.9 cc (range: 2.6–171.3 cc), 3.5 cc (range: 0.0–36.6 cc), and 9.7% (range: 0.0%–67.3%). Multivariate Cox regression revealed adenocarcinoma (hazard ratio [HR] = 5.76; P = .03) and GTVIGBT (HR = 1.17; P < .01) as predictors for local failure. Additionally, GTVIGBT was associated with any disease recurrence (HR = 1.17; P < .01) and overall mortality (HR = 1.20; P < .01). GTVIGBT > 7.5 cc was associated with inferior 2-year LC (75.0% vs 96.6%; P < .01), DFS (42.6% vs 91.6%; P < .01), and OS (65.2% vs 16 American Radium Society Scientific Papers and Posters 2015 91.5%; P < .01). No difference in mean HRCTV D90 EQD2 was seen between these groups (83.2±2.2 Gy vs 83.5±2.7 Gy; P = .61). CONCLUSION: Aside from the known benefits of IGBT, MRI-based planning allows for assessment of tumor regression and prognosticates patients, as shown in the present study. If these findings are replicated in prospective trials, alternative methods, such as dose escalation and surgical salvage, should be considered to offset poor prognoses. (S036) Circulating Tumor DNA and Implications for Clinical Decision-Making in Stage I NSCLC Julie Koenig, Ben Durkee, Erqi Pollom, Iris Gibbs, Max Diehn; Stanford University School of Medicine; Department of Radiation Oncology, Stanford University Medical Center; Institute for Stem Cell Biology and Regenerative Medicine, Stanford University BACKGROUND: A nontrivial fraction of stage I non–small-cell lung cancer (NSCLC) is occult metastatic disease [Rusch, JCO 2011]. These patients have worse disease-free survival and overall survival (OS) [Rusch, JCO 2011]. The phase III Cancer and Leukemia Group B (CALGB) 9633 trial found a small benefit for adjuvant chemotherapy in stage I patients with high-risk disease based on size criterion [Strauss, JCO 2008]. However, the authors could not reproducibly identify these high-risk patients [Strauss, JCO 2008]. Circulating tumor DNA (ctDNA) is highly predictive of residual disease after definitive therapy [Newman, Nat Med 2014]. We hypothesize that ctDNA could identify high-risk patients with stage I NSCLC who could benefit from early systemic therapy. METHODS: We created a Bayesian model to simulate posterior probabilities after screening for occult metastatic disease by ctDNA. We used prevalence data from the American College of Surgeons Oncology Group (ACOSOG) Z0040 trial, and recently published receiver operating characteristics for ctDNA [Newman, Nat Med 2014]. Next, we built a two-state Markov model, with the assumption that occult micrometastases conferred worse survival (hazard ratio [HR] = 1.82), based on ACOSOG. Patients receiving chemotherapy were assigned a survival benefit (HR = 0.69, high risk) and detriment (HR = 1.12, low risk) by extrapolation from Cancer and Leukemia Group B (CALGB) trial 9633. The model was run with no patient receiving chemotherapy and rerun with patients who had detectable ctDNA receiving adjuvant chemotherapy. high-risk patients could benefit from systemic therapy. There are limitations to this study: a model does not substitute for a clinical trial. We did not do a sensitivity analysis. Our time horizon (3 yr) is short. We assumed that our high-risk patients (defined by ctDNA) would derive the same benefit as high-risk patients in the CALGB trial (defined by size). To support this, the data show that ctDNA levels correlate with tumor volume [Newman, Nat Med 2014]. We assumed that occult micrometastases and ctDNA had clinical significance, which is supported in published literature. (S037) Phenotypic Diversity Measured in PET/CT Scans Predicts Overall Survival in Early-Stage Lung Adenocarcinoma Jennifer S. Chang, Viola Walther, Natalie Lui, Aleah F. Caulin, David Jablons, Carlo Maley, Trevor Graham, Sue S. Yom; University of California, San Francisco; Barts Cancer Institute, Queen Mary University of London BACKGROUND: Although most patients with early-stage lung adenocarcinoma have favorable outcomes, stratification of a high-risk subset may facilitate clinical decision making on therapy and surveillance. The degree of within-tumor heterogeneity may be a prognostic biomarker, as more diverse tumors are more likely to contain a subpopulation that is adapted to a new selective pressure (eg, radiotherapy), whereas homogeneous tumors are less likely to harbor a resistant subclone. We recently found that within-tumor genetic heterogeneity predicts overall survival (OS) in patients with early-stage lung adenocarcinoma. Here, we investigated whether quantification of withintumor phenotypic diversity, measured by 18-fluorodeoxyglucosepositron emission tomography (FDG-PET) and computed tomography (CT) imaging, was also a predictor of OS. METHODS: Genetic analysis had previously been performed on 58 patients initially diagnosed with stage I/II lung adenocarcinoma, treated with surgery and no neoadjuvant therapy. We obtained the presurgery PET and CT scans. The primary lung lesion was contoured on PET and CT scans using the software program MIM 5.6. Phenotypic diversity was then quantified using the Moran I statistic, which describes spatial autocorrelation. The prognostic value of the Moran I coefficient was assessed using Cox proportional hazards (CPH) models and Kaplan-Meier curves. Statistical analyses were performed in R. RESULTS: The 3-year OS rate was 71% and 73% for high- and lowrisk patients, respectively, with a total payoff of 2.56 life-years. These modeled estimates correlate well with CALGB data. The positive likelihood ratio for ctDNA was 10.0. After screening for ctDNA, the posterior probability of occult disease for a positive test improved to 67.0%, from a baseline prevalence of 16.9%. Treating high-risk patients with adjuvant chemotherapy improved survival to 79%. The number needed to treat was 12.5. RESULTS: Tumors were contoured on 46 CT and 32 PET scans. Differences in phenotypic diversity were identified across the cohort for both PET and CT images. Patients in the upper quartile of Moran I values (phenotypic diversity) had significantly shorter survival using PET scans (log-ranked: P = .02; CPH: P = .08; hazard ratio [HR] = 2.72; 95% confidence interval [CI], 0.88–8.39), and the upper quartile was borderline significant for CT scans (log-ranked: P = .051; CPH: P = .061; HR = 2.83; 95% CI, 0.95–8.45). There was a weak rank correlation between Moran’s I statistic and genetic diversity for both PET (P = .85) and CT (P = .21). Moran I and genetic diversity were significant predictors for overall survival using PET scans (Moran I: P = .045; genetic diversity: P = .032). CONCLUSION: Our model suggests that ctDNA could improve clinical decision-making for stage I NSCLC. Prospectively identified CONCLUSIONS: These results suggest that within-tumor phenotypic diversity, as quantified in PET-CT scans, can predict OS in ARS PROCEEDINGS 2015 17 American Radium Society Scientific Papers and Posters 2015 patients with early-stage lung adenocarcinoma. Quantification of within-tumor heterogeneity in this routine clinical imaging may provide a noninvasive method for identifying a high-risk subset of patients with early-stage non–small-cell lung cancer. (S038) Pulmonary Dose-Volume Predictors of Radiation Pneumonitis Following Stereotactic Body Radiation Therapy Eileen M. Harder, BS, Henry S. Park, MD, MPH, Zhe Chen, PhD, FAAPM, Roy H. Decker, MD, PhD; Department of Therapeutic Radiology, Yale University School of Medicine PURPOSE: Radiation pneumonitis (RP) can be a significant risk after stereotactic body radiation therapy (SBRT). The purpose of this study was to identify clinical and dosimetric predictors of symptomatic (grade ≥ 2) RP following pulmonary SBRT. MATERIALS AND METHODS: Patients with ≥ 3 months of follow-up who received SBRT for primary lung cancer were selected from an institutional database. RP was determined retrospectively from all available records, including those from appointments with radiation oncology, pulmonology, and medical oncology and hospitalizations. RP was scored per Common Terminology Criteria for Adverse Events version 4.0 (CTCAE v4.0). Normal lung volume was defined as total lung volume minus gross tumor volume (GTV) on the planning computed tomography (CT). Pulmonary Dmax (maximum point dose), mean lung dose (MLD), and Vx (volume of lung receiving ≥ x Gy) in 5-Gy increments were collected. Univariate analyses were performed with the chi-square or Student’s t-test. Dosimetric predictors of RP were identified using multivariate logistic regression with a manual forward stepwise selection technique. RESULTS: A total of 264 patients were included (median follow-up 29.4 mo). Median prescription dose was 54 Gy (range: 40–60 Gy). Patient characteristics were as follows: 27 (10.2%) had multifocal disease, 58 (22.0%) had T-stage ≥ 2, 26 (9.8%) had prior lung radiotherapy (4.2% with SBRT), and 72 (27.3%) had prior lung resection. Grade ≥ 2 RP occurred in 61 patients (23.1%) with a median onset time of 1.8 months (range: 0.1–16.2 mo). Grade ≥ 3 RP occurred in 23 patients (8.7%). Lung V5, V10, V15, V20, V25, V30, V35, V40, V45, and MLD were significantly associated with grade ≥ 2 RP on univariate analysis (P < .05), but no correlation was seen with lung V50, lung Dmax, age, gender, Eastern Cooperative Oncology Group (ECOG) performance status, cigarette use, T stage, location (central/peripheral), lobe (upper/lower) synchronicity, prescribed BED, GTV, ITV, PTV, pretreatment pulmonary comorbidity, or any heart dosimetric volume. Among these significant factors, lung V10 was the strongest predictor of RP on multivariate analysis (P = .006). The median lung V10 was 10.9 Gy. Symptomatic RP was present in 18.9% of patients receiving lung V10 < 10.9 Gy, compared with 27.3% with lung V10 ≥ 10.9 Gy. CONCLUSIONS: Symptomatic RP occurred in 23.1% of our patients treated with SBRT. Lung V10 was the strongest predictor of grade ≥ 2 RP on multivariate logistic regression, associated with a 30% decrease in risk for patients with V10 < 10.9 Gy compared with ≥ 10.9 Gy. Further research is needed to validate these findings and the importance of lung V10 in predicting symptomatic RP following SBRT. 18 (S039) High-Resolution 4D Ventilation and Perfusion PET/CT Facilitates Functionally Adapted IntensityModulated Radiation Therapy (IMRT) in Lung Cancer David Ball, MD; Peter MacCallum Cancer Centre PURPOSE: To assess the utility of functional lung avoidance using intensity-modulated radiation therapy (IMRT) informed by fourdimensional ventilation/perfusion positron emission tomography/ computed tompgraphy (4D-V/Q PET/CT) in patients with non– small-cell lung cancer (NSCLC). MATERIALS AND METHODS: In a prospective clinical trial, patients underwent 4D-V/Q PET/CT scanning before 60 Gy of definitive chemoradiation. Both “highly perfused” (HPLung) and “highly ventilated” (HVLung) lung volumes were delineated using a visually adapted 70th percentile standardized uptake value (SUV) threshold method. The HVLung was universally smaller than HPLung; so, a “ventilated lung volume” (VLung) was created to approximate the HPLung using a 50th percentile SUV threshold. For each patient, four IMRT plans were created, optimized to the anatomical lung, HPLung, VLung, and HPLung volumes. IMRT that was optimized to the anatomical lung was compared with functionally adapted IMRT using functional lung volumetrics, including mean lung dose (MLD), V5, V10, V20, V30, V40, V50, and V60 parameters. Plan quality was assessed by dose to 95% and 5% of planning target volume (PTV) (D95 and D5), conformity index (CI), and heterogeneity index (HI). RESULTS: The study cohort consisted of 20 patients with 80 IMRT plans. The mean (+/– SD) lung volume for the HPLung and VLung was similar at 1,876 cc (+/– 677 cc) and 1,904 cc (+/– 505 cc), respectively, while HVLung was smaller at 932 cc (+/– 258 cc). Plans that were optimized to HPLung resulted in a significant reduction of functional MLD by a mean of 13.0% (1.7 Gy; P = .02). Functional V5, V10, and V20 were improved by 13.2%, 7.3%, and 3.8%, respectively (P < .04). There was no sparing of dose to functional lung when adapting to VLung or HVLung. Plan quality was highly consistent with a mean PTV D95 and D5, ranging from 60.8–61.0 Gy and 63.4–64.5 Gy, respectively, and with mean CI and HI ranging from 1.11–1.17 and 0.94–0.95, respectively. CONCLUSIONS: IMRT plans that were adapted to perfused but not ventilated lung on 4D-V/Q PET/CT allowed for reduced dose to the functional lung while maintaining consistent plan quality. (S040) Has Severe Acute Esophagitis Been Reduced by 21st Century Treatment Modalities Among Patients With Limited-Stage Small-Cell Lung Cancer? Ritsuko Komaki, MD, Pamela K. Allen, PhD, Xiong Wei, Emma B. Holliday, MD, Brett Carter, MD, Stephen D. Bilton, CMD, James D. Cox, MD; UT MD Anderson Cancer Center PURPOSE: Severe acute esophagitis (grade 3–5) is a limiting factor when twice-daily radiation treatment (RT) with concurrent chemotherapy (CChRT) is applied. The previous Intergroup study (IG 0096) has shown a significant 27 % rate of severe acute esophagitis with twice-daily RT compared to an 11 % esophagitis rate with daily RT and concurrent etoposide and cisplatin (EP) among patients American Radium Society Scientific Papers and Posters 2015 with limited-stage small-cell lung cancer (SCLC) (P < .01), although twice-daily treatment significantly improved 5-year overall survival (OS). Neither intensity-modulated radiation therapy (IMRT) nor three-dimensional conformal radiotherapy (3D-CRT) was used for the IG 0069 study. We investigated if IMRT has reduced severe acute esophagitis with concurrent EP in our institution. METHODS: We identified 577 patients with limited-stage SCLC treated at a single institution from 1986–2009 with definitive CChRT to a total dose of at least 45 Gy. Candidate variables included tumor size, year of diagnosis, gender, age, Karnofsky performance status (KPS), ethnicity, radiation dose, cycles of induction chemotherapy, use of IMRT, and once- vs twice-daily fractionation. Tumor measurement used the biggest dimension of the primary or regional node. The chi-square test was used for between-group comparisons for categorical variables, and the median test was used for between-group comparisons for continuous variables. KaplanMeier estimates were constructed for OS, disease-free survival (DFS), local recurrence–free survival (LRFS), and distant metastasis–free survival (DMFS). Analysis was performed using logistic regression analysis with severe acute esophagitis as the primary endpoint. RESULTS: Of the 577 patients included in this analysis, 520 patients had tumor size data available. Patients with severe esophagitis were treated with immediate CChRT (n = 99 [95%] vs n = 5 [5%]; P < .001), more likely to have a KPS score < 90 (n = 50 [48%] vs n = 45 [43%]; P = .008), treated with radiation therapy twice daily (n = 66 [63%] vs n = 34 [33%]; P = .01), had been treated since 2000 (n = 47 [55%] vs n = 47 [45%]; P = .02), and had tumors ≥ 5 cm (n = 66 [63%] vs n = 34 [33%]; P = .01) All other patient, tumor, and treatment characteristics were not significantly different between the groups. For the entire cohort, Kaplan-Meier estimates showed severe esophagitis to be associated with reduced local-regional failure–free survival (5-year survival 39.3% vs 53.2%; P = .027). Logistic regression analysis found immediate CChRT (odds ratio [OR] = 2.39; P = .001) and tumor size ≥ 5 cm (OR = 1.66; P = .024) to be associated with severe esophagitis. CONCLUSIONS: Patients with tumors ≥ 5 cm, receiving immediate CChRT or twice-daily RT, and having local-regional failure experienced significantly higher rates of severe acute esophagitis. IMRT did not reduce acute severe esophagitis. (S041) The Role of Chemoradiation in Elderly Limited-Stage Small-Cell Lung Cancer Patients Christopher D. Corso, MD, PhD, Charles E. Rutter, MD, Henry S. Park, MD, MPH, Nataniel Lester-Coll, MD, Roy H. Decker, MD, PhD; Department of Therapeutic Radiology, Yale University School of Medicine INTRODUCTION: Small-cell lung cancer (SCLC) in elderly patients is a frequent problem, with prior studies suggesting that 40% to 55% of patients with limited stage SCLC (LS-SCLC) are aged 70 years or older. A meta-analysis examining the role of thoracic radiotherapy (RT) in LS-SCLC suggested that there is no clear benefit to combined modality therapy over chemotherapy alone in patients aged over 70 years. In this retrospective study, we sought to investigate outcomes for elderly patients treated with chemotherapy alone vs chemotherapy (chemo)-RT in the modern era using a national database. METHODS: We identified elderly patients (age ≥ 70 yr) diagnosed with LS-SCLC (cT1-T4, cN0-N3, cM0) in the National Cancer Data Base from 1998–2006. Patients treated with surgery or RT initiated > 30 days before or > 270 days after chemotherapy were excluded. Patients were stratified as receiving chemotherapy alone vus chemoRT. The group receiving chemo-RT was further subdivided into concurrent vs sequential treatment. Concurrent therapy was defined as those starting RT 30 days before to 90 days after chemotherapy began. Overall survival (OS) from the start of treatment was compared between groups using log-rank analysis. Conditional survival analysis was performed for patients who were alive 6 months after the initiation of treatment. RESULTS: We identified 20,836 patients who met the inclusion criteria; 47.9% received chemo alone, and 52.1% received chemo-RT. The median RT dose was 55.8 Gy. OS was significantly higher in the chemo-RT group when compared with patients who received chemo alone (median OS 20.0 mo vs 10.7 mo; log-rank P < .001). This survival benefit was observed in elderly patients with no comorbid illness and in those with multiple comorbidities. Analysis of the treatment sequence revealed a significant difference between concurrent vs sequential therapy by log-rank analysis (P < .037). The Kaplan-Meier curves crossed at approximately 18 months, suggesting that there is initially improved OS with sequential therapy, but long-term survival is improved with chemo-RT. Conditional survival analysis showed that patients who were alive 6 months after diagnosis had a significant survival benefit with concurrent chemoRT over sequential therapy (2-yr OS 35% vs 27%; log-rank P < .001). CONCLUSION: With modern treatment, there appears to be a significant benefit to the use of combined modality therapy in elderly patients with LS-SCLC, even in those with medical comorbidities. For patients who would be expected to tolerate the acute effects of concurrent chemo-RT, there appears to be a significant long-term survival benefit over sequential therapy. (S042) Factors Influencing Brain Recurrence After PCI Among Patients With Limited Small-Cell Lung Cancer Emma B. Holliday, Pamela K. Allen, Brett Carter, Xiong Wei, Stephen D. Bilton, Ritsuko Komaki; UT MD Anderson Cancer Center PURPOSE: The brain is the most common site of distant failure for patients with small-cell lung cancer (SCLC) after complete response (CR) to treatment. Prophylactic cranial irradiation (PCI) was established as the standard of care for limited-stage disease after definitive chemoradiation (CRT). PCI confers an overall survival (OS) advantage but is not without toxicity. Therefore, we sought to identify a low-risk population for whom PCI may not be necessary. We hypothesized that patients with small tumor size may be at lower risk for brain failure, even in the absence of PCI. METHODS: We identified 577 patients with limited-stage SCLC who were treated with definitive CRT at a single institution from 1986– 2009 to a dose ≥ 45 Gy. The chi-square test was used for betweengroup comparisons for categorical variables, and the median test was used for between-group comparisons for continuous variables. ARS PROCEEDINGS 2015 19 American Radium Society Scientific Papers and Posters 2015 Kaplan-Meier estimates were constructed for brain metastasis–free survival (BMFS) and OS. Analysis was performed using competing risk regression with BMFS as the primary endpoint. Factors identified as significantly associated with improved BMFS were entered into the multivariate model, and log-rank test was performed. Tumor size was recorded as the largest dimension of the primary parenchymal tumor or the largest regional node, if the primary tumor could not be visualized. RESULTS: Of the 577 patients identified, 307 (53.2%) received PCI. Patients who did not receive PCI were older (median age 64 years [range: 27–95 yr] vs 61 years [range: 31–79 yr]; P = .021) and had a higher rate of those with Karnofsky performance status (KPS) score < 80 (n = 49 [18.1%] vs n = 24 [7.8%]; P < .001). Fifty-four (17.6%) patients who received PCI and 71 (26.3%) patients who did not receive PCI ultimately developed brain metastases. Tumor size was available for 520 patients. Among patients who received PCI, those with tumor size ≥ 5 cm trended towards increased brain failure (n = 21 [13.5%] vs n = 25 [22.3%]; P = .058). On multivariate competing risk analysis with other distant metastases DM as the competing risk, those with a tumor ≥ 5 cm were 50% more likely to develop brain failure (P = .035). For patients who received PCI, tumor size was still significantly associated with increased brain failure (subdistribution hazard ratio [SHR] = 1.79; 95% confidence interval [CI], 1.01–3.18; P = .048). For patients who did not receive PCI, tumor size was not associated with increased brain failure. There was no difference in OS based on tumor size. CONCLUSIONS: Tumor size appears to be a significant prognostic factor for brain recurrence in patients with limited-stage SCLC after PCI. Further investigation is warranted to in order to best individualize treatment for limited-stage SCLC patients with large tumors. (S043) Changes in Quality of Life After Radiation Therapy for Localized Prostate Cancer After Dissemination of Intensity-Modulated Radiation Therapy Elyn H. Wang, Shiyi Wang, Pamela Soulos, Ronald Chen, Cary P. Gross, James B. Yu; Yale University School of Medicine; University of North Carolina School of Medicine PURPOSE: Although intensity-modulated radiation therapy (IMRT) has replaced three-dimensional conformal radiotherapy (3D-CRT) to become the dominant form of external beam radiation therapy (EBRT) for prostate cancer, it is unclear whether this change in practice has been associated with improved patient outcomes. To estimate the impact of IMRT on health-related quality of life (HRQOL), we examined the HRQOL of elderly patients who had undergone EBRT during two eras: 1998–2001, when 3D-CRT was the standard of care for EBRT, and 2006–2009, when IMRT became the dominant form of EBRT. METHODS: The Surveillance, Epidemiology, and End Results (SEER)-Medicare Health Outcomes Survey was used to obtain HRQOL measurements of Medicare beneficiaries who had received external beam radiotherapy for nonmetastatic prostate cancer and control subjects without cancer. Control subjects during each era were matched 10:1 by use of pro20 pensity scores estimated from demographics and comorbid medical conditions. A total of 90 patients during the early era and 75 patients during the late era had eligible surveys both before and after treatment. Descriptive statistics and random effects linear regression were used to compare HRQOL between cancer patients and matched controls during the two eras. Area under the curve (AUC) was calculated to determine quality-adjusted life-years (QALYs) gained. RESULTS: During the first 2 years after treatment, patients who were treated in the late era had 0.19 more QALYs compared with patients treated in the early era, relative to matched noncancer controls. Additionally, health utility and mental component scores at followup among cases during the early era declined significantly compared with controls, whereas they were similar between cases treated in the late era and their controls (P = .02 for difference in change in health utility and mental component scores between cases and controls during early vs late era). CONCLUSIONS: External beam radiotherapy during an era after IMRT has become widespread is associated with improved HRQOL measures among prostate cancer patients compared with the era prior to IMRT dissemination. (S044) Patient-Reported Quality of Life After Stereotactic Body Radiotherapy (SBRT), IntensityModulated Radiotherapy (IMRT), and Brachytherapy Joseph R. Evans, MD, PhD, Shuang Zhao, MSE, Stephanie Daignault, MS, Martin G. Sanda, MD, Jeff Michalski, MD, Howard M. Sandler, MD, Deborah A. Kuban, MD, Jay Ciezki, MD, Irving D. Kaplan, MD, Anthony L. Zietman, MD, Larry Hembroff, PhD, Felix Y. Feng, MD, Simeng Suy, PhD, Ted A. Skolarus, MD, P. William McLaughlin, MD, John T. Wei, MD, Rodney L. Dunn, MS, Steven E. Finkelstein, MD, Constantine A. Mantz, MD, Sean P. Collins, MD, PhD, Daniel A. Hamstra, MD, PhD; University of Michigan; Emory University; Washington University; Cedars-Sinai Medical Center; UT MD Anderson Cancer Center; Cleveland Clinic; Beth Israel Deaconess Medical Center; Massachusetts General Hospital; Michigan State University; Georgetown University; University of Michigan; VA Ann Arbor Healthcare System; 21st Century Oncology background: Stereotactic body radiotherapy (SBRT) is gaining popularity as a radiotherapy option for localized prostate cancer that is less invasive than brachytherapy and less costly and time-intensive than intensity-modulated radiotherapy (IMRT). However, there are ongoing concerns about excess toxicity after SBRT compared with other radiotherapy options. METHODS: We conducted a multi-institutional pooled cohort analysis of patient-reported quality of life (QoL) using the Expanded Prostate Cancer Index (EPIC-26) survey collected serially from baseline to 2 years. Men were treated with IMRT, brachytherapy, or SBRT at multiple institutions. Mean EPIC symptom domain scores were evaluated, as was the proportion of patients with a minimal clinically detectable difference (MCD) in each domain. Multivariate analyses were used to determine the independence and association of standard clinical variables and treatment type with domain scores at 2 years. Sensitivity analysis was performed for 12-Item Short Form Health Survey (SF-12) data, as a minor proportion of the SBRT patients were missing SF-12 data, and for year of IMRT American Radium Society Scientific Papers and Posters 2015 treatment, as we suspected that IMRT technique may have evolved over time, which would be expected to decrease toxicity and reduce the impact on QoL. RESULTS: Data were analyzed from 803 patients at baseline and from 645 patients at 2 years. Mean declines at 2 years across all patients were −1.9, −4.8, −4.9, and −13.3 points for the urinary obstructive, urinary incontinence, bowel, and sexual symptom domains, respectively. Based upon MCD, 29%, 20%, and 28% of all patients had detectable differences in urinary, bowel, and sexual domains, respectively, at 2 years. On multivariate analysis, brachytherapy had worse urinary irritation (−6.8 [−9.9, −3.6] points vs IMRT, P < .0001; −5.8 points [−9, −2.5] vs SBRT, P < .0001), but there were no differences in the bowel (P = .48), urinary incontinence (P = .21), or sexual domains (P = .18). QoL after SBRT was similar to IMRT for the urinary (P = .55, obstruction; P = .74, incontinence) and sexual domains (P = .57) but was associated with better bowel score (+6.7 [3.2, 10] vs IMRT, P < .0002; +5.5 [2.2, 8.8] vs brachytherapy, P = .001). We observed a positive association between SF-12 Mental and Physical component scores and 2-year domain scores for nearly all domains. Sensitivity analyses demonstrated a minimal effect of missing SF-12 data and no changes in 2-year domain scores with year of IMRT treatment. CONCLUSIONS: There were slightly smaller declines in bowel QoL after SBRT and slightly larger declines in urinary QoL after brachytherapy. While QoL after SBRT, IMRT, or brachytherapy is largely similar and supports SBRT as a reasonable radiotherapy option for localized prostate cancer, a randomized comparison is needed for the strongest evidence. (S045) Equivalent Survival With Breast-Conserving Therapy and Mastectomy in Young Women Under the Age of 40 With Early-Stage Breast Cancer: A National Registry-Based Stage-by-Stage Comparison Jason C. Ye, MD, Weisi Yan, MD, PhD, Paul Christos, DrPH, MS, Dattatreyudu Nori, MD, MBBS, Akkamma Ravi, MD, MBBS; Department of Radiation Oncology, New York Presbyterian Hospital/Weill Cornell Medical College; Department of Radiation Oncology, New York Hospital Queens/Weill Cornell Medical College; Division of Biostatistics & Epidemiology, Department of Public Health, Weill Cornell Medical College PURPOSE: Studies have shown that young patients with early-stage breast cancer (BC) are increasingly getting mastectomy instead of breast-conserving therapy (BCT) despite the lack of clear survival evidence in this approach. We examined the difference between outcomes in young women (aged < 40 yr) treated with BCT vs mastectomy. METHODS: The Surveillance, Epidemiology, and End Results (SEER) database was queried for women aged < 40 years and diagnosed with stage I or II invasive BC treated with surgery from 1998– 2003. Breast cancer–specific survival (BCSS) and overall survival (OS) were evaluated by Kaplan-Meier survival analysis, and the logrank test was used to compare survival between treatment categories of interest. Multivariable Cox regression model analysis was performed to estimate the predictors of BCSS. RESULTS: Of the 7,665 women identified, 3,249 (42%) received BCT, while 2,627 (34%) patients had mastectomy and no radiation. There were also 994 (14%) patients who had lumpectomy but no radiation, and 795 (10%) were treated with mastectomy plus radiation. On multivariate analysis, higher stage, estrogen receptor–negative status, and mastectomy plus radiation were found to have a higher risk of BC mortality. When separated by stage, with a median follow-up of 111 months (based on living patients), the BCT and mastectomy-only groups showed no statistically significant difference in BCSS or OS. There was also no difference in non-BC mortality in any-stage disease between the two treatment groups. The group aged 35–39 years (66% of total) was associated with better 10-year BCSS (88%) and OS (86.1%) compared with patients aged 20–34 years (34% of total), who had 10-year BCSS and OS rates of 84.1% and 82.3%, respectively (P < .001 for both BCSS and OS). When patients of each age group were further subdivided into stages I, IIA, and IIB, there was no statistically significant difference in BCSS or OS between the BCT and mastectomy groups for any stage for those aged 35–39 years or for stages I and IIA in the 20–34-yearold age group. However, in patients aged 20–34 years with stage IIB disease only, the mastectomy-only group (n = 183) had significantly inferior 10-year BCSS (64% vs 79%; P = .004) and OS (61% vs 77%; P = .002) compared with the BCT group (n = 183). CONCLUSION: Our study suggests that while young age may be a poor prognostic factor for breast cancer, there is no evidence that these patients have better outcomes with mastectomy over BCT, supporting the continued use of BCT. (S046) Prognosis for Patients With Metastatic Breast Cancer Who Achieve ‘No-Evidence-of-Disease’ Status After Systemic or Local Therapy Andrew J. Bishop, Joe Ensor, Stacy L. Moulder, Simona F. Shaitelman, Mark A. Edson, Gary J. Whitman, Sandra Bishnoi, Karen E. Hoffman, Michael C. Stauder, Vicente Valero, Thomas A. Buchholz, Naoto T. Ueno, Gildy Babiera, Wendy A. Woodward; UT MD Anderson Cancer Center; Rice University PURPOSE/OBJECTIVES: Newer systemic therapy regimens are more commonly associated with complete or near-complete radiographic response in patients with metastatic breast cancer (MBC). We investigated outcomes for patients with MBC with no evidence of disease (NED) after treatment and identified factors that were predictive of outcome once NED status was achieved. MATERIALS AND METHODS: We reviewed 570 patients with MBC consecutively treated between January 2003 and December 2005. NED was defined as a complete metabolic response by positron emission tomography (PET) and/or sclerotic healing of bone metastases on computed tomography (CT) or magnetic resonance imaging (MRI). Median follow-up was 27 months (range: 0–134 mo) for all patients and 100 months (range: 14–134 mo) for NED patients. The Kaplan-Meier method estimated overall survival (OS) and progression-free survival (PFS); log-rank tests assessed for equality among groups. Cox proportional hazard models were used. Estimated hazard ratios (HRs) are reported. RESULTS: The 3- and 5-year OS rates for the entire group were 44% ARS PROCEEDINGS 2015 21 American Radium Society Scientific Papers and Posters 2015 and 24% compared with 96% and 78%, respectively, for the NED subset (n = 90, 16%). Patients who attained NED status had a longer median survival (112 vs 24 months; P < .001), which correlated strongly with survival on multivariate analysis (P < .001; HR = 0.15). Several other factors were also associated with more favorable survival on multivariate analysis, including presenting with de novo MBC (P < .001; HR = 0.60) and distant metastases involving only bone (P < .001; HR = 0.52) or only lung (P < .001; HR = 0.47). Patients who developed NED were more likely to have a normal body mass index (BMI) (P < .001), have either estrogen receptor (ER)+ or human epidermal growth factor receptor 2 (HER2)+ tumors (P < .001), have single sites of metastasis (P < .001), and present with de novo MBC (P = .04) when compared with the nonNED group. For patients with NED, the 3- and 5-year PFS rates (after NED) were 54% and 38%, respectively, with a median NED duration of 41 months (range: 0–115 mo). The only characteristics associated with PFS after NED were receptor status (HER2+ 58 mo vs ER+ 35 mo vs TNBC 13 mo; P = .02), inflammatory breast cancer at presentation (28 mo vs 45 mo; P = .02), and trastuzumab use (73 mo vs 31 mo; P = .02). Of the 63 patients who had disease progression, 20 achieved a second NED status. Ultimately, 31 patients (34%) with NED remained in remission at last follow-up. CONCLUSIONS: Achieving NED status correlates strongly with outcome in MBC, making this a potentially valuable short-term clinical trial endpoint akin to pathologic complete response. MBC patients who attain NED have prolonged survival, with perhaps up to onethird achieving complete remission. (S047) Radiotherapy vs Chemotherapy Effects on Neuronal Architecture and Spine Density in the Hippocampus Nevine Hanna, MD, MPH, Munjal Acharya, PhD, Charles Limoli, PhD; University of California, Irvine PURPOSE: Studies have shown that cognitive function is compromised by both radiation treatment (RT) and the use of chemotherapy (ie, the “chemobrain” phenomenon). Given that advancements in diagnosis, beam delivery, and drug treatments have extended long-term cancer survivorship, it has become increasingly critical to address the extent and persistence of, and neurobiological mechanisms underlying, cognitive dysfunction associated with cancer treatments. Through immunofluorescence studies in mice, our lab previously established that irradiation elicits significant reductions in neuronal morphology. In the present study, we examine the consistency of the effects of radiation or chronic cyclophosphamide (CYP) treatment, a commonly prescribed chemotherapeutic agent, on animal behavior and neuronal morphology via different staining techniques. METHODS: Adult athymic nude rats were treated with 9 Gy of x-rays every other day to a total of 27 Gy (biologically effective dose [BED] = 108 Gy) or with CYP (100 mg/kg) once weekly for 4 weeks and compared with sham-irradiated or saline controls, respectively. At 1 month posttreatment, animals were administered hippocampus- and cortex-dependent cognitive tasks, including novel place recognition (NPR) and a temporal order (TO) task. Following cognitive testing, immunohistochemical staining was used to trace immature (doublecortin, DCX+) and mature (Golgi- 22 Cox–impregnated) neurons in the brain for an assessment of neuronal morphology in the CA1 region. Spine density was also counted to delineate the number of long/thin, mushroom, and stubby spines in the same region. RESULTS: Both irradiated and CYP-treated rats showed significant decrements in learning and memory when assessed on both the NPR and TO tasks. CYP-treated animals were impaired in hippocampus-dependent place recognition memory and cortex-dependent recency memory compared with controls. Quantification of ultrastructural parameters of neurons in the hippocampus using Neurolucida software indicated compromised dendritic morphology in the immature (DCX+) and mature (Golgi-Cox) neurons in the CA1 hippocampal region. Chronic CYP treatment (n = 20) led to significant reductions in the apical dendritic volume (P = .0033), basal total dendritic length (P = .0161), endings (P = .0352), volume (P = .0004), and complexity (including branching and three-dimensional morphology; P = .01) compared with sham. Early analysis (n = 5) is also showing similar significance in the irradiated cohort. With regard to spine density, overall group effects were found for altered numbers of long/thin (P = .0079), mushroom (P = .0048), and stubby (P = .006) spine types in the CA1 region. CONCLUSIONS: Exposure to chemotherapy or RT disrupts the formation and establishment of proper synaptic connections by causing significant alterations in neuronal structure that compromise neurotransmission and cognition. (S048) Identification of Excellent and Poor Prognostic Groups After Stereotactic Radiosurgery for Spinal Metastasis: Secondary Analyses of Mature Prospective Trials Chad Tang, Kenneth Hess, Andrew Bishop, Hubert Pan, Eva Christiansen, Nizar Tannir, Behrang Amini, Claudio Tatsui, Lawrence Rhines, Paul Brown, Amol Ghia; UT MD Anderson Cancer Center BACKGROUND /PURPOSE: There is uncertainty in the prognosis of patients following treatment of spinal metastases. To stratify patients between poor and excellent predicted survival after stereotactic spine radiosurgery (SSRS), we created a Cox proportional hazards regression model. PATIENTS AND METHODS: Patients who were enrolled (between 2002 and 2011) in two prospective trials investigating SSRS for spinal metastasis were analyzed. To ensure mature survival data, living patients with < 3 years of follow-up were excluded. A multivariate Cox regression model was utilized to create a survival model via backward selection at P < .05. Pretreatment variables included race, sex, age, performance status, tumor histology, extent of vertebrae involvement, prior therapy at SSRS site, disease burden, and timing of diagnosis and treatment. Four prognostic groups were generated based on the model-derived prognostic index (PI). To assess whether pretreatment symptoms were associated with survival and survival model predictions, patients were prospectively queried for their pretreatment symptoms via the MD Anderson Symptom Inventory (MDASI) and Brief Pain Index (BPI). American Radium Society Scientific Papers and Posters 2015 RESULTS: A total of 206 patients were included in this analysis. Median follow-up time was 70 months (range: 37–133 mo) among all living patients (n = 40). Seven variables were selected in the prediction model: female sex (hazard ratio [HR] = 0.7; P = .04), Karnofsky performance status (KPS) (HR = 0.7 per 10% increase; P = .005), prior surgery at the SSRS site (HR = 0.6; P = .005), prior radiation at the SSRS site (HR = 1.7; P = .003), SSRS site as the only site of disease (HR = 0.5; P < .001), number of organ systems involved (HR = 1.4 per involved system; P < .001), and time between initial diagnosis and spine metastasis after variable normalization (HR = 0.6 per log10[time in months+5]; HR = 0.01). The c-index of the stratified and unstratified model was 0.68 and 0.70, respectively. The median survival time among all patients included in the analysis was 25.5 months and was significantly different among prognostic groups (Group 1 [excellent prognosis]: not reached, Group 2: 32.6 mo, Group 3: 19.7 mo, Group 4 [poor prognosis]: 8.2 mo; P < .001). Patients within the excellent prognosis group exhibited a Kaplan-Meier estimated 10-year survival rate of 71%. Furthermore, all pretreatment symptom metrics were predictive of overall survival and correlated with the model-derived PI (all P ≤ .01). 0.33–65.2 mo), with 12- and 18-month OS rates of 56.6% and 44.3%, respectively. On univariate analysis, inferior survival was associated with lower Karnofsky performance status (KPS) score (P = .038), presence of extracranial metastases at iSRS (P = .003), new metastases at the time of rSRS (P = .027), having rSRS < 9 months from iSRS (P = .019), and melanoma histology (P = .015). For all metastases from rSRS, the 12- and 18-month LC rates were 77.5% and 71.6%, respectively. Univariate analysis for local failure showed no significant association. Regarding DBF, the 12- and 18-month estimates were 52.8% and 62.9%, respectively. Univariate analysis for DBF was significant for melanoma (P < .01) and persistent systemic disease at rSRS (P = .012). Multivariate analysis showed a significant association for DBF with melanoma (hazard ratio [HR] = 22.34; P value < .00), presence of extracranial disease at rSRS (HR = 2.89; P = .07), and having at least one new brain metastasis at rSRS (HR = 3.32; P = .03), with an overall model P value < .00. Twelve patients (11.1%) had grade 3 radiation toxicities following rSRS at a median time of 4.0 months (range: 1.2–10.6 mo). The grade 3 toxicities consisted of radiation necrosis (10), seizure (3), headache (3), and an optic nerve disorder. No grade 4 or 5 toxicities were seen. Neither a dose nor volume relationship with toxicity was observed. CONCLUSIONS: We present a survival prediction model that has identified patient subgroups with poor (Group 4) to excellent (Group 1) prognoses. In addition, pretreatment symptoms were predictive of survival and correlated with the prediction of the model. If validated, we believe that this model, possibly in conjunction with patient symptoms, may aid in determining optimal treatment strategies. CONCLUSIONS: Repeat SRS represents a potential salvage therapy for patients with locally recurrent brain metastases, providing additional tumor control with acceptable toxicity, even in the setting of prior SRS, surgical resection, and/or WBRT. Repeat SRS may also be reasonable to use to either avoid or delay the treatment of WBRT. (S049) Salvage Stereotactic Radiosurgery for Locally Recurrent Brain Metastases Treated Previously With Stereotactic Radiosurgery Ben Y. Durkee, MD, PhD, Joseph Sanford, MD, Anna Oh, PhD, Daniel Slate, BS, Brandon Turner, BS, Erqi L. Pollom, MD, Iris C. Gibbs, MD, Scott G. Soltys, MD; Department of Radiation, Department of Radiation Oncology, Stanford University; Department of Nursing, University of California, San Francisco; Stanford University Graduate School of Business; Stanford University School of Medicine Douglas E. Holt, BS, Beant S. Gill, MD, David A. Clump, MD, PhD, Steven A. Burton, MD, John C. Flickinger, MD, Jonathan A. Engh, MD, Nduka Amankulor, MD, PhD, Dwight E. Heron, MD; University of Pittsburgh Medical Center PURPOSE AND OBJECTIVES: Patients with local recurrence of brain metastases following prior stereotactic radiosurgery (SRS) can be challenging to manage. Given the concerns of neurotoxicity with whole-brain radiotherapy (WBRT), we evaluated the efficacy of repeat SRS (rSRS) for patients with locally recurrent brain metastases after initial SRS (iSRS). MATERIALS AND METHODS: A retrospective review from 2004 to 2014 identified 108 patients (133 lesions) who received rSRS due to locally recurrent brain metastases after iSRS. Among these patients, 19.4% had WBRT prior to rSRS, with 40.6% of the lesions previously treated with surgical resection. Kaplan-Meier estimates were calculated from rSRS for overall survival (OS), local control (LC), and distant brain failure (DBF), as well as radiation-related toxicity. Cox proportional hazards modeling was conducted to establish predictive factors for OS, LC, DBF, and toxicity (P < .05) from the time of rSRS. RESULTS: With a median follow-up time of 12.0 months (range: 0.03–65.7 mo) from rSRS, the median OS was 14.2 months (range: (S050) Cost-Effectiveness of Neurocognitive Preservation for Whole-Brain Radiotherapy BACKGROUND: Whole-brain radiotherapy (WBRT) is the standard of care for nonsurgical intracranial metastatic disease. Patients receiving WBRT are at risk for neurocognitive degeneration, which is weighed against the neurocognitive detriment of tumor progression. Recent results from RTOG 0614 and RTOG 0933 have shown cognitive benefit with memantine (WBRT+M) or hippocampal avoidance (HA-WBRT) [Li, JCO 2007; Brown, Neuro Onc 2013; Gondi, JCO 2014]. Neurocognitive compromise is associated with lower quality of life (QoL). Utility scores for these health states are well described and reproducible. The cost of these cognitive preservation strategies must be weighed against the gains in QoL during the patients’ final months. METHODS: We created a decision tree with an integrated three-state Markov model. Treatment strategies included best supportive care (BSC), WBRT, WBRT+M, and HA-WBRT. Health states were cognitively intact, cognitively impaired, and dead. Cycle-specific probabilities for cognitive states were derived from PCI-P120-9801, RTOG 0614, and RTOG 0933. Survival data were derived from ARS PROCEEDINGS 2015 23 American Radium Society Scientific Papers and Posters 2015 recursive partitioning analysis (RPA) of three RTOG trials [Gaspar, IJROBP 1997]. Utility scores for the base case of a patient with metastatic cancer were derived from the available literature. The effect of cognitive detriment was extrapolated from patients with mild dementia, who scored similarly on a Hopkins Verbal Learning Test. The model was run using each RPA class (I/II/III) as the base case scenario. The analysis was done from a societal perspective, with a single payer system. Threshold for cost-effectiveness was set at a willingness to pay (WTP) of $100,000 per quality-adjusted life-year (QALY). Costs were derived from the Medicare Physician Fee Schedule. We performed multiway sensitivity analyses to address uncertainties in cost, utility scores, efficacy of intervention, life expectancy, and distribution of baseline cognitive states. RESULTS: BSC is the dominant strategy for the base case scenario. No strategy for neurocognitive preservation is cost-effective at a WTP of $100,000/QALY. Neurocognitive-preservation strategies become cost-effective in the theoretical cohort of patients with perfect baseline cognition and long expected survival (> 14 mo for base case distribution; > 10 mo for perfect baseline cognition). The model was sensitive to assumptions about the initial distribution of patients’ cognitive states and cost. It was minimally sensitive to utility scores and efficacy of the intervention. CONCLUSION: Neurocognitive-preservation strategies for WBRT are not cost-effective from a societal perspective, though they may be effective for patients who are cognitively intact at baseline and have a long expected survival. (S051) The Role of Temozolomide in the Treatment of Low-Grade Glioma Emi J. Yoshida, MD, Alicia Ortega, BA, Chirag G. Patil, MD, MS, Stephen L. Shiao, MD, PhD; Cedars-Sinai Medical Center BACKGROUND/OBJECTIVE: World Health Organization (WHO) grade II gliomas are relatively slow-growing brain tumors, with 5-year progression-free survival rates ranging widely, from 13% to 70%. A subset of these tumors will transform into high-grade aggressive cancers with dismal prognoses. A number of phase II trials have demonstrated a role for well-tolerated temozolomide in the prevention of tumor progression and malignant transformation, despite a lack of impact on survival. The institutional paradigm at Cedars-Sinai Medical Center is to manage low-grade gliomas with single-agent temozolomide and to reserve radiotherapy for progression. The aim of this study is to report the preliminary results of our experience with temozolomide in the management of low-grade glioma. MATERIALS AND METHODS: The records of 234 patients diagnosed with WHO grade II glioma at Cedars-Sinai Medical Center from January 1991 to April 2014 were reviewed. Median age was 47.5 years (range: 39.0–60.0 yr). Ninety-four (40.2%) patients were treated with temozolomide +/− radiotherapy after surgery. Surgery included biopsy (75 patients, 33.8%), subtotal resection (56 patients, 25.2%), and gross total resection (91 patients, 40.9%). Median follow-up time was 51.0 months (range: 25.0–84.4 mo). Statistical 24 analysis controlled for tumor histology, tumor size, extent of resection, and patient age. P values < .05 were considered statistically significant. RESULTS: Of the 234 patients analyzed, 211 patients recurred or progressed (90.2%). Average time to progression was 42.7 months among patients treated with temozolomide +/− radiotherapy compared with 46.5 months among patients who did not receive temozolomide (P = .35). Five-year progression-free survival (PFS) was 20.7% in those who received temozolomide +/− radiotherapy compared with 29.5% among those who were observed +/− radiotherapy. CONCLUSION: Temozolomide has become widely used as a primary intervention for low-grade glioma in the past decade. Our preliminary results suggest that the use of temozolomide in the management of low-grade glioma is neither deleterious nor beneficial in terms of PFS. In comparison with the PFS reported by the European Organisation for Research and Treatment of Cancer (EORTC) 22844/22845 and Radiation Therapy Oncology Group (RTOG) 9802 trials, our PFS is markedly worse. This finding is likely attributable to the high median age of our patient population, which is noticeably older and thus assumed to have a poorer prognosis. Ongoing studies are investigating the impact of temozolomide on time to radiotherapy and examining survival outcomes after concurrent temozolomide and radiotherapy. (S052) Inverse Optimization for Correlating 4DCT Ventilation Imaging and Radiation Dose Edward Castillo, PhD, Richard Castillo, PhD, Thomas Guerrero, MD, PhD; Beaumont Health System; UT Medical Branch PURPOSE: Four-dimensional computed tomography (4DCT) ventilation imaging is an emerging modality with utility in thoracic radiotherapy treatment planning. Though recent studies have demonstrated its potential for quantifying the functional response of lung tissue to irradiation, ventilation image analysis is challenging and difficult to reproduce because of issues such as spatial inaccuracies in the required deformable image registration (DIR), cine mode phase-binning artifacts, and variations in the patient’s breathing. In order to address these issues, we have developed a numerical method for computing 4DCT ventilation images that correlate perfectly with a given radiation dose distribution or dose-response model while maintaining high-spatial accuracy in the corresponding DIR solution. METHODS: Ventilation images are defined by a DIR spatial transformation that maps the position of inhale lung voxels to their corresponding position at exhale (or vice versa). A voxel’s volume change under the transformation, described mathematically by the Jacobian matrix, is the intensity value of the ventilation image and quantifies the air exchanged. Given an initial DIR estimate and a target ventilation image, we define an optimization problem describing the spatial transformation closest to our initial estimate (according to the L2-norm), with Jacobian values equal to those in the target ventilation image. RESULTS: The inhale/exhale phases of a treatment-planning 4DCT American Radium Society Scientific Papers and Posters 2015 for a patient with non–small-cell lung cancer were registered using a previously reported DIR method. The spatial accuracy of the DIR solution, given as the average (standard deviation) millimeter error with respect to 417 expert-determined landmark points, was 1.03 (1.01) mm. The radiation dose distribution image was used to generate a target ventilation image using a linear dose-response model applied to the original ventilation image. Our numerical method was then applied to the initial DIR solution to produce a spatial transformation and corresponding ventilation image that matched the target ventilation image within 1e-2.The average millimeter error for the new transformation was 1.11 (1.10). CONCLUSION: A numerical method for computing a DIR transformation according to a target ventilation image was used to generate a ventilation image that correlates precisely with the dose distribution while maintaining high DIR spatial accuracy. Thus, by employing this approach, the focus of future CT ventilation studies that are designed to assess radiation dose response is reduced to assessing the physical feasibility of the DIR transformation that generates the ventilation image predicted by the dose-response model. (S053) Nodal Surveillance With Diffusion-Weighted MRI After Definitive (Chemo) Radiotherapy for HPV-Predominant Squamous Cell Cancers of the Oropharynx and Unknown Primary Yao Yu, MD, Marc Mabray, MD, William Silveira, MD, PhD, Peter Y. Shen, MD, PhD, William Ryan, MD, Alina Uzelac, DO, Sue S. Yom; University of California, San Francisco BACKGROUND: Diffusion-weighted MRI has been proposed as a method to differentiate treatment effect from persistent or recurrent nodal disease after definitive treatment with radiotherapy. METHODS: Records and imaging were reviewed for 70 patients treated with definitive radiotherapy with or without chemotherapy for squamous cell carcinomas of the oropharynx or p16-positive unknown primary of the head and neck. A total of 40 patients were available for analysis. Surveillance imaging with magnetic resonance imaging (MRI) was obtained 6–8 weeks after treatment, followed by positron emission tomography/computed tomography (PET/CT) at 12 weeks after treatment. Apparent diffusion coefficient (ADC) values were calculated for each node and for each hemineck. PET/CT results and ADC values were correlated with regional control at 6 months based on histopathology and clinical follow-up. RESULTS: The mean ADC was significantly lower for lymph nodes corresponding with recurrent disease compared with control at 6 months (1,301 μm2/s vs 2,049 μm2/s; P = .04). Using receiver operating characteristic (ROC) analysis, an optimal threshold of 1,600 μm2/s was identified; lymph nodes with ADC values below this threshold were found to be at higher risk for recurrent disease. Sensitivity and specificity were 100% and 84%, respectively, with positive and negative predictive values of 41% and 100%, respectively. When analyzed by hemineck, the sensitivity and specificity were 100% and 85%, respectively. PET/CT at 12 weeks yielded a sensitivity and specificity of 100% and 85%, respectively. On Kaplan-Meier analysis, ADC was predictive of nodal progression-free survival (P = .00023). CONCLUSION: Early surveillance imaging with diffusion-weighted MRI at 6–8 weeks following definitive treatment with radiotherapy is feasible and may be predictive of early nodal failure. Further validation in a prospective cohort is warranted. (S054) Long-Term Survival and Racial Differences in Pediatric Hodgkin Lymphoma Patients From the State of Florida: Three Decades of Experience Hanmanth Neboori, MD, William Grubb, BS, Hua Li, PhD, MD, Joseph Panoff; University of Miami/Jackson Health System PURPOSE/METHODS: Hodgkin lymphoma (HL) represents 9% of all pediatric malignancies in the United States. Data are conflicting with regard to racial/ethnic survival differences. We sought to investigate overall survival (OS) differences in a large cohort of racially and ethnically diverse patients with 30 years of follow-up in the state of Florida. METHODS: The Florida Cancer Database System was used to retrospectively assess the long-term outcomes of 1,778 pediatric patients (age range: 1 mo–21 yr) diagnosed with HL between 1981 and 2010. Log-rank test and Cox univariate and multivariate regression analysis were used to identify predictors of OS. RESULTS: Median age at diagnosis was 1.7 years (range: 0.1–21 mo). Males and females were equally represented (50.5% vs 49.5%). The database consisted of 68% white, 13% African American (AA), 18% Hispanic, and 2% unknown. The breakdown of diagnosis by decade was as follows: 27% from 1981–1990; 31% from 1991–2000; and 42% from 2000–2010. There were 16% of patients diagnosed as stage I, 42% diagnosed as stage II, 31% diagnosed as stages III and IV, and 11% unknown. Nodular sclerosing was diagnosed in 69% of patients, mixed cellularity was diagnosed in 13%, lymphocyte-rich disease was diagnosed in 4%, lymphocyte-depleted (LD) disease was diagnosed in 2%, and nodular lymphocyte-predominant HL (NLPHL) was diagnosed in 2%. Radiation was administered in 43% of patients, and chemotherapy was given to 73% of patients. Median survival of the entire cohort at the time of last follow-up was 23.8 years. The 5-year OS rate was 84.2%. Men had worse OS than women at 25 years (36% vs 58%; P < .0001). AA patients had worse OS than white and Hispanic patients at 25 years (33% vs 49.2% vs 44.7%, respectively; P = .0005), and this finding persisted after controlling for decade of treatment (P < .001). There was no difference in OS between whites and Hispanics. There were no OS differences regarding decade of treatment, chemotherapy, or age at diagnosis. Patients who had radiation therapy had better OS (hazard ratio [HR] = 1.49; P = .0025). Additionally, patients with the LD subtype had worse OS (HR = 3.85; P = .01). CONCLUSION: This is the largest retrospective review with the longest outcome to specifically evaluate pediatric Hodgkin lymphoma patients. Furthermore, this is the first analysis to find that AA patients have inferior OS when compared with whites and Hispanics. These differences remained significant over the course of 30 years, indicating that modern treatment modalities have not improved this racial disparity. ARS PROCEEDINGS 2015 25 American Radium Society Scientific Papers and Posters 2015 SCIENTIFIC POSTERS* (P001) Disparities in the Local Management of Breast Cancer in the United States According to Health Insurance Status Thomas M. Churilla, MD, Brian Egleston, PhD, Joshua Meyer, MD, Yanqun Dong, MD, PhD, Penny Anderson, MD; Fox Chase Cancer Center INTRODUCTION: Various studies have suggested disparities in breast cancer care due to inadequate health insurance coverage. The purpose of our study was to test for associations between patient insurance status and the presentation and local therapy patterns among a large, nationally representative cohort of patients with localized breast cancer. METHODS: We queried the National Cancer Institute (NCI) Survival, Epidemiology, and End Results (SEER) database for breast cancer cases diagnosed from 2007–2011 in women aged 18–64 years, limited to ductal or lobular histology, nonmetastatic stage, and treated with mastectomy or lumpectomy. Patients ≥ 65 years were excluded based on their eligibility for Medicare. We characterized clinical and demographic variables according to insurance status (insured vs Medicaid vs uninsured). We tested for associations between patient insurance status and choice of definitive surgical procedure (mastectomy vs breast-conserving surgery), omission of radiation therapy (RT) following breast-conserving surgery, and administration of postmastectomy RT (PMRT). We calculated odds ratios (ORs), performed Pearson’s chi-square test for univariable analysis, and used multiple logistic regression analysis to adjust for clinical and demographic covariates. RESULTS: A total of 129,565 patients with localized breast cancer were analyzed. The health insurance statuses included insured (84.5%), Medicaid (11.5%), uninsured (2.1%), and unknown (1.9%). Patients with Medicaid or uninsured status were more likely to present with large, node-positive tumors and be black or unmarried and reside in low-income counties. The breast-conserving surgery rate was 51.3% among all patients and varied according to insurance status: insured (52.2%), uninsured (47.7%), and Medicaid (45.2%) (P < .001). Medicaid insurance status remained significantly associated with receipt of mastectomy in the multivariable analysis (OR =1.07; 95% confidence interval [CI], 1.03– 1.11; P < .001). Radiation therapy was more frequently omitted after breast-conserving surgery in both Medicaid (OR = 1.14; 95% CI, 1.07–1.21) and uninsured (OR = 1.29; 95% CI, 1.14–1.47) patients in the multivariable analysis (P < .001 for both). PMRT was more frequently administered among Medicaid and uninsured patients; however, only Medicaid status remained significantly associated with PMRT in the multivariable analysis (OR = 1.10; 95% CI, 1.04–1.18; P = .002). CONCLUSION: Patients with inadequate health insurance were more likely to receive mastectomy, omit RT following breast-conserving surgery, and receive PMRT. Differences in clinical presen*Scientific posters P001 through P014 have been designated by the American Radium Society as Posters of Distinction. 26 tation and demographics according to insurance status incompletely explain the variation in therapy. Further study is needed to validate and address these disparities and to evaluate the impact of health insurance legislative efforts in localized breast cancer. (P002) Predictors of CNS Disease in Metastatic Melanoma: Desmoplastic Subtype Associated With Higher Risk Gary B. Deutsch, MD, MPH, Samuel Yost, BA, Mariel B. Deutsch, MD, Ji Hey Lee, PhD, Leland Foshag, MD, Garni Barkhoudarian, MD, Daniel F. Kelly, MD, Mark B. Faries, MD; John Wayne Cancer Institute at Providence Saint John’s Health Center; David Geffen School of Medicine at UCLA and West Los Angeles VA Healthcare System INTRODUCTION: Brain metastases are a major cause of mortality in metastatic melanoma. Histologic subtype—specifically, the desmoplastic subtype and its propensity for neurotropism—has not been well studied as a predictor of central nervous system (CNS) disease. We report the largest series of brain metastases in order to better define the clinicopathologic risk factors and help guide management and surveillance. METHODS: A prospective institutional melanoma database was used to identify patients diagnosed with metastatic melanoma between 1971 and 2013. Patient age and sex; primary tumor location, thickness, ulceration, and histology; and types of recurrence and treatment were analyzed. Primary endpoints were development of brain metastases and 2-year brain metastases–free survival (BMFS). The secondary endpoint was overall survival (OS) from the date of stage IV diagnosis. RESULTS: Among 3,756 patients with metastatic melanoma, 711 (18.9%) developed brain metastases. Histology was available for 2,132 patients, 397 (18.6%) of whom developed brain metastases (32 [8.1%] brain only, 38 [9.6%] brain as the first site of stage IV disease). Head and neck (H&N) location (P = .01), presence of ulceration (P = .04), and desmoplastic variant (P = .04) were associated with a higher risk of CNS disease. Multivariable analysis identified presence of ulceration (hazard ratio [HR] = 1.49; P < .01), primary location (upper extremity vs H&N: HR = 0.55, P < .01; lower extremity vs H&N: HR = 0.63, P = .01; mucosal vs H&N: HR = 0.42, P = .04; ocular vs H&N: HR = 0.29, P = .04), and histologic subtype (desmoplastic vs superficial spreading; HR = 2.38; P = .01) as independent predictors of 2-year BMFS. Improved 2-year OS was seen with female sex, younger age, upper extremity location, lack of ulceration, and ipilimumab therapy. Of the 25 patients diagnosed with desmoplastic melanoma, 9 (36%) were found to have brain metastases, all within the first year after the diagnosis of another systemic disease. Thick lesions (> 4 mm) of the H&N region were at greatest risk. CONCLUSIONS: Desmoplastic histologic subtype is a strong predictor of brain metastasis development and decreased 2-year BMFS in patients with metastatic melanoma. Patients with desmoplastic melanoma, particularly thick lesions involving the H&N, should be imaged frequently during the first year after the diagnosis of stage IV disease. American Radium Society Scientific Papers and Posters 2015 (P003) Identification of Somatic Mutations Using Fine Needle Aspiration: Correlation With Clinical Outcomes in Patients With Locally Advanced Pancreatic Cancer Arti Parekh, Lauren M. Rosati, Vicente Valero III, Matthew J. Weiss, Ryan K. Assadi, Daniel A. Laheru, Christopher L. Wolfgang, Christine A. Iacobuzio-Donahue, Joseph M. Herman; Johns Hopkins Hospital PURPOSE/OBJECTIVE: Historically used to biopsy tumors that raise suspicion for pancreatic cancer, fine needle aspirates (FNAs) can also be obtained at the time of fiducial placement to perform molecular studies. We present an exploratory analysis of genomic sequencing data from patients who received stereotactic body radiation (SBRT) for locally advanced pancreatic cancer (LAPC). The purpose of this study was to identify somatic mutations associated with clinical outcomes and survival. MATERIALS AND METHODS: FNAs from 15 patients undergoing SBRT for LAPC were sequenced for 409 known cancer genes. All FNAs were obtained prior to the delivery of 25–33 Gy SBRT in five fractions. Induction gemcitabine was administered to all patients, and the majority (93%) of patients received maintenance gemcitabine following a 1-week break from SBRT. Kaplan-Meier survival analysis was used to evaluate the association between gene variants and treatment response to SBRT. RESULTS: Median follow-up time was 13.6 months (range: 2.4–26.2 mo). Median age was 66.3 years, and 66.7% of patients had pancreatic head tumors. Fourteen (93%) patients had disease progression, with 14.3% initially experiencing local progression alone, 71.4% experiencing distant progression alone, and 14.3% experiencing synchronous local and distant progression. Median overall survival was 16.7 months (95% confidence interval [CI], 11.6–21.7 mo), and the 1-year and 2-year survival rates were 63.4% and 22.6%, respectively. Advanced age (≥ 65 yr) was associated with inferior survival (10.8 vs 21.1 mo; P = .048). Patients who first experienced local failure alone had inferior survival compared with those who experienced distant failure only (6.9 vs 16.7 mo; P = .01). At least two patients in this cohort expressed mutations in the following genes: KRAS, CDKN2A, TP53, SMAD4, TRIM33, ARID1A, GRM8, and NOTCH. No specific correlations were observed between most driver mutations known in pancreatic cancer (CDKN2A, TP53, SMAD4) and overall survival. Interestingly, improved survival was observed in patients harboring KRAS mutations compared with those with the wild-type variant (17.9 vs 8.5 mo; P = .015). One patient (6.7%) had an ATM mutation and survived 19.2 months. CONCLUSIONS: This is the first study to demonstrate the feasibility of genomic sequencing of FNAs from pancreatic tumors. We have been able to successfully identify unique genetic signatures in patients with LAPC; however, the small sample size limits our ability to identify generalizable patterns. With this exploratory analysis, we propose that routine FNA sequencing, as well as additional molecular studies, such as immunohistochemistry, of larger cohorts may allow for identification of unique patterns that guide individualized selection of patients for SBRT. (P004) A Retrospective Study to Assess Disparities in the Utilization of Intensity-Modulated Radiotherapy (IMRT) and Proton Therapy (PT) in the Treatment of Prostate Cancer (PCa) Kristina L. Demas, MD, Neha Vapiwala, MD, Stefan Both, PhD, Curtiland Deville, MD; University of Pennsylvania BACKGROUND: Despite its increase in use, proton therapy (PT) is a relatively limited resource. The purpose of this study was to examine clinical and demographic differences in intensity-modulated radiotherapy (IMRT) and PT utilization for prostate cancer (PCa). METHODS: All patients with low- and intermediate-risk PCa (n = 350) undergoing definitive RT (2.5 Gy × 28 fractions or 1.8 Gy ×44 fractions) between 2010–2012 at a single institution were divided into IMRT (n = 58) and PT (n = 292) comparison groups. Pretreatment characteristics, including age, race, socioeconomic status (SES) (low vs high, defined as geocoded census tract 20% below or above poverty level, respectively), prostate-specific antigen (PSA), clinical tumor stage, Gleason score, risk group, prostate volume, and patient-reported outcomes, were retrospectively collected. Chi-square and independent sample t-tests were used for analyses. RESULTS: Of PT patients, 228 (78%), 51 (18%), 4 (1%), and 9 (3%) were white, black, Asian, or other, respectively; 256 patients (88%) had high SES, and 36 (12%) had low SES. Mean age, distance from center, PSA level, prostate volume, International Prostate Symptom Score (IPSS), and International Index of Erectile Function (IIEF) in the PT group were 65 ± 7.1 years, 86 ± 190 miles, 5.6 ± 2.9 ng/mL, 41 ± 18 cc, 8 ± 6, and 19 ± 6, respectively; 142 (49%) patients were lowrisk, and 150 (51%) were intermediate-risk. A total of 236 (81%), 46 (16%), and 10 (3%) PT patients were T1c, T2a, and T2b, respectively; 154 (53%) and 138 (47%) patients were Gleason 6 and 7. In the IMRT group (n = 58), 28 (48%), 24 (42%), 3 (5%), and 3 (5%) patients were white, black, Asian, or other, respectively; 40 (69%) patients had higher SES, and 18 (31%) had low SES. Mean age, distance, PSA, prostate volume, IPSS, and IIEF were 69 ± 8.6, 16 ± 18 miles, 7.4 ± 4.5 ng/mL, 54 ± 40 cc, 8 ± 7, and 14 ± 8, respectively; 142 (49%) were low-risk, and 150 (51%) were intermediaterisk patients. A total of 236 (81%), 46 (16%), and 10 (3%) IMRT patients were T1c, T2a, and T2b, respectively; 154 (53%) and 138 (47%) patients were Gleason 6 and 7. The cohorts varied in average age (P = .0005), race (P < .0001), SES status (P = .0007), and average miles traveled to the facility (P = .0054)—ie, IMRT patients were older, resided closer, and consisted of more black and low-SES patients. Baseline PSA (P = .0001), Gleason score (P = .0244), prostate volume (P = .0040), and IIEF (P < .001) were significantly increased for IMRT, while risk group, T stage, and IPSS were not (P > .05 for all). Therapeutically, IMRT patients were less likely to receive hypofractionated therapy (P < .0001) and more likely to receive androgen deprivation therapy (P = .0006). CONCLUSION: Disparities exist in PT utilization compared with IMRT by age, race, and SES and merit further investigation. ARS PROCEEDINGS 2015 27 American Radium Society Scientific Papers and Posters 2015 (P005) Ultrasensitive PSA Identifies Patients With Organ-Confined Prostate Cancer Requiring Postop Radiotherapy Jung Julie Kang, MD, PhD, Robert Reiter, MD, Patrick Kupelian, MD, Michael Steinberg, MD, Christopher R. King, PhD, MD; Department of Radiation Oncology, Department of Urology, UCLA PURPOSE/OBJECTIVES: Randomized controlled trials have shown that adjuvant radiotherapy (RT) after radical prostatectomy (RP) improves biochemical relapse-free and overall survival in patients with extracapsular disease. However, even patients with organconfined disease are at risk for failure. This study analyzes postoperative (postop) ultrasensitive prostate-specific antigen (uPSA) in the surveillance of these patients. MATERIALS AND METHODS: From 1991–2013, 146 patients with pT2N0 disease who were referred for a rise in PSA after RP were identified: 85 were margin-positive (m+), and 61 were margin-negative (m−). Surgical approach (open vs robotic), initial PSA (iPSA), Gleason grade, margin status, and postop uPSA were covariates for analysis. Median first postop PSA and follow-up were 3 and 38 months, respectively. The uPSA threshold was 0.01 ng/mL. Benign uPSA patterns occurred from 0.01–0.02 ng/mL; thus, ≥ 0.03 ng/mL was defined as uPSA failure. True (conventional) biochemical relapse (tBCR) was defined as PSA ≥ 0.2 ng/mL. Patients were censored at last follow-up or adjuvant therapy. Kaplan-Meier and Cox multivariate analyses were used to compare tBCR rates. RESULTS: Median time to tBCR for the entire cohort was 50 months. The m+ patients revealed a more indolent course than m− patients (median time to relapse: 86 mo for m+ vs 33 mo for m−; P = .0003). No differences in Gleason grade or iPSA between m+ and m− patients were seen. On multivariate analysis (MVA), only first postop uPSA ≥ 0.03 ng/ mL (hazard ratio [HR] = 4.1; P < .001) and margin status (m−: HR = 5.6; P = .0315) independently predicted for time to tBCR. First postop uPSA ≥ 0.03 ng/mL discerned tBCR with much greater sensitivity than undetectable first conventional PSA < 0.2 ng/mL (44% vs 19%). First postop uPSA < 0.03 ng/mL vs. ≥ 0.03 ng/mL predicted median tBCR at 61 vs 10 months (P = .0003). Any postop uPSA ≥ 0.03 ng/mL captured all failures missed by analyzing only the first postop value (100% sensitivity). Using uPSA ≥ 0.03 ng/mL to identify relapse yields a substantial (33 mo) lead time advantage over waiting until conventional relapse at PSA ≥ 0.2 ng/mL (median 17 vs 50 mo in favor of uPSA ≥ 0.03 ng/mL). CONCLUSIONS: Only first postop uPSA ≥ 0.03 ng/mL and margin status independently predicted biochemical relapse for organ-confined prostate cancer. The m− patients exhibited much earlier failures, suggesting greater biologic aggressiveness. Biochemical failure can be called at uPSA ≥ 0.03 ng/mL with excellent sensitivity, and adopting it offers an impressive lead time advantage over the conventional failure definition (PSA ≥ 0.2 ng/mL). 28 Integrating uPSA into the immediate and continued frequent surveillance of RP patients with organ-confined cancer will improve postop RT outcomes by identifying failures sooner and promoting an early salvage strategy. (P006) The Role of Sequential Imaging in Cervical Cancer Management A. MacDonald, C. Tung, M. Bonnen, M.Y. Williams-Brown, C.R. Diaz-Arrastia, M. Ludwig, M.L. Anderson; UT Health Science Center at Houston; Baylor College of Medicine OBJECTIVES: In the absence of International Federation of Gynecology and Obstetrics (FIGO) guidelines, optimal imaging of women diagnosed with cervical cancer in settings where multiple modalities are available for treatment planning remains unclear. The purpose of this study was to determine whether sequential imaging by contrast-enhanced computed tomography (CT), 18-fluoro-deoxyglucose positron emission tomography (FDGPET)-CT, and/or magnetic resonance imaging (MRI) enhances radiation treatment planning for women with cervical cancer. METHODS: After obtaining institutional review board (IRB) approval, all women diagnosed with cervical cancer who were eligible for definitive chemoradiation (FIGO stages IB1–IVA) in a regional safety-net health system between July 2012 and August 2014 were identified. Clinical demographics and treatment plans were reviewed and abstracted so that the impact of imaging modalities could be compared. RESULTS: A total of 93 women (mean age: 51.0 ± 13.2 yr) with IB1 (n = 3, 3.2%), IB2 (n = 16, 17.2%), IIA (n = 11, 11.8%), IIB (n = 33; 35.4%), IIIA (n = 5, 5.3%), IIIB (n = 22; 23.7%), and IVA (n = 3; 3.2%) disease were identified. Histologies included squamous cell (n = 78, 83.8%), adenosquamous (n = 2, 2.1%), poorly differentiated (n = 3, 3.2%), small cell (n = 1, 1.1%), and adenocarcinomas (n = 9, 9.6%). Pretreatment, 48 women underwent contrastenhanced CT (abdomen/pelvis and/or chest) alone, 28 received CT followed by PET, 6 received PET and MRI, 4 received CT and MRI, 1 received PET only, and 6 underwent all three tests. For the 34 women who had CT followed by PET, no anatomic findings were identified by CT that were not also detected by PET. In contrast, PET identified lesions that were not seen in anatomic fields evaluated by CT in eight women. PET resulted in radiation treatment modifications for 22 (65%) of these women. Of the 12 patients who received PET and MRI, treatment was modified in response to PET but not MRI for 8 (67%). Treatment modifications due to PET included nodal boost (n = 20, 67%), extension of irradiated field (n = 8, 26%), or both (n = 2, 7%). CONCLUSIONS: Despite uncertainty regarding the anatomic resolution of PET, sequential use of contrast-enhanced CT, PETCT, and/or MRI had no impact on treatment planning that was not accomplished by the use of PET alone. Future work should focus on determining the optimal pretreatment imaging for women with cervical cancer and developing guidelines to optimize outcomes while minimizing cost and radiation exposure. American Radium Society Scientific Papers and Posters 2015 (P007) Pooled Analysis of Locoregional Relapse After Minimally Invasive Surgery Alone for Intermediateor High-Risk HPV+ Oropharyngeal Squamous Cell Carcinoma Ryan K. Funk, MD, David G. Stoddard, MD, Kanograt Tangsriwong, MD, Michael G. Keeney, MD, Qihui Zhai, MD, Eneida F. Vencio, DDS, Alexander Lin, MD, Joaquin J. Garcia, MD, Eric J. Moore, MD, Robert L. Foote, MD, Katharine A. Price, MD, Daniel J. Ma, MD; Department of Radiation Oncology, Department of Otolaryngology, Department of Laboratory Medicine and Pathology, Division of Medical Oncology, Mayo Clinic; Department of Radiation Oncology, University of Pennsylvania; Department of Laboratory Medicine and Pathology, Mayo Clinic, Jacksonville, FL; Department of Oral Pathology, School of Dentistry, Federal University Goiás PURPOSE/OBJECTIVES: Traditional risk factors predicting locoregional relapse (LRR) following surgery for head and neck squamous cell carcinomas were identified in the pre–human papillomavirus (HPV) era. Patients with HPV-positive (HPV+) oropharyngeal squamous cell carcinoma (OPSCC) present at younger age and have an approximately 50% reduction in death when compared with their HPV− counterparts. Minimally invasive surgery (MIS) (eg, transoral robotic surgery [TORS] and transoral laser microsurgery [TLM]) allows for improved intraoperative assessment of the tumor. It is unknown if traditional risk factors fully apply to the HPV+ population that undergoes MIS. We examined whether traditional indications for adjuvant radiation therapy (RT) (perineural invasion [PNI], angiolymphatic invasion [ALI], T3–4, or N2a–b) or adjuvant chemoradiation therapy (CRT) (extracapsular extension [ECE] or positive margins) also predict for LRR in patients with HPV+ OPSCC who undergo MIS alone. MATERIALS AND METHODS: After institutional review board (IRB) approval, the medical records of patients treated with TORS (two institutions) or TLM (one institution) for OPSCC were reviewed to identify patients with HPV+ tumors who had indications for adjuvant therapy but did not receive adjuvant therapy. HPV status was confirmed by p16 immunohistochemistry. Individual data from all three institutions were pooled into a single database for analysis. RESULTS: A total of 43 patients with HPV+ base of tongue (n = 16) or tonsil (n = 27) OPSCC with a median age of 60 years (range: 37–81 yr) who did not receive adjuvant therapy despite the presence of intermediate- or high-risk features were identified. Seventeen of 43 patients had a smoking history > 10 pack-years. Fourteen patients had ECE as an indication for CRT. Margins were close in all patients due to the nature of MIS, but the final margins were all negative. Intermediate-risk features were ALI (n = 4), PNI (n = 3), T3 (n = 6), N2a (n = 13), and N2b (n = 6). Three patients had two intermediate risk factors. Median follow-up after surgery was 30.1 months (range: 3.1–161.7 mo). Eleven of 47 (26%) patients developed LRR at a median of 5.7 months (range: 2.9–39.2 mo). Kaplan-Meier estimate of 2-year relapse-free survival was lower for patients with ECE vs patients with intermediate-risk factors only (43% vs 88%, respectively; P = .02). LRR was detected locally (n = 3, base of tongue), locally and in ipsilateral regional lymph nodes (ILNs) (n = 1), in ILNs alone (n = 4), in contralateral regional LNs (n = 2), or in bilateral neck nodes (n = 1). CONCLUSION: In the HPV+ population, ECE remains a significant risk factor for LRR without adjuvant therapy after MIS. Patients with intermediate-risk disease after MIS have moderate rates of LRR. (P008) Pretreatment FDG Uptake of Nontarget Lung Tissue Correlates With Symptomatic Pneumonitis Following Stereotactic Ablative Radiotherapy (SABR) Aadel A. Chaudhuri, MD, PhD, Michael S. Binkley, BA, Justin N. Carter, BA, Maximilian Diehn, MD, PhD, Billy W. Loo, MD, PhD; Stanford University School of Medicine INTRODUCTION: Radiation pneumonitis (RP) is the most common toxicity associated with radiation therapy to the lung. It can be debilitating and even life-threatening, especially since many patients with lung malignancies have other pulmonary comorbidities. Here, we report our institution’s experience in using pretreatment positron emission tomography-computed tomography (PETCT) to calculate baseline nontarget lung fluorodeoxyglucose (FDG) avidity, which we have found to be significantly associated with symptomatic RP following stereotactic ablative radiotherapy (SABR). METHODS: We retrospectively reviewed outcomes in 73 patients with lung tumors treated with SABR. All patients underwent whole body PET-CT and four-dimensional (4D) chest CT simulation scan within 2 weeks prior to treatment. Gross tumor volume (GTV) was contoured on axial CT slices using lung windows. Respiration-induced motion was assessed by 4D CT, used to design the internal target volume (ITV), and a 0.5-cm margin was added to form the planning target volume (PTV). Total lung minus PTV and ipsilateral lung minus PTV were contoured using lung windows. All contouring was done using Varian Eclipse and MIM software. We used the simulation PET-CT scans to calculate the standard uptake value (SUV) at the 85th (SUV85), 90th (SUV90), and 95th (SUV95) percentiles and mean SUV for total lung minus PTV and ipsilateral lung minus PTV using MIM and compared these values between patients who experienced symptomatic RP and those who did not. All strata were compared using Student’s t-test and Cox regression for univariate and multivariate analyses. RESULTS: Median follow-up time was 15 months (range: 4–45 mo). A total of 70 patients (95.9%) were treated for T1 to T3 non–smallcell lung cancer, while 3 (4.1%) were treated for oligometastatic disease from non-lung primary cancers. Sixty-six patients (90.4%) were treated for one lung tumor, and seven patients (9.6%) were treated for two lung tumors. Sixteen patients (21.9%) experienced grade ≥ 2 RP, 3 of whom experienced grade ≥ 3 RP. Median time to pneumonitis was 4.95 months (range: 2.5–24 mo). Analysis of simulation PET-CT scans from all patients revealed that those who experienced grade ≥ 2 RP had significantly higher total lung minus PTV SUV85 (P = .0005), SUV90 (P = .0014), and SUV95 (P = .0078) than those with no symptomatic pneumonitis. We saw similar results when we analyzed ipsilateral lung minus PTV, with significantly higher baseline SUV85 (P = .0021), SUV90 (P = .0036), SUV95 (P = .0108), and mean SUV (P = .0012) in patients who experienced grade ≥ 2 RP. A significant association was observed between grade ≥ 2 RP and total lung minus PTV SUV85 on uniARS PROCEEDINGS 2015 29 American Radium Society Scientific Papers and Posters 2015 variate (hazard ratio [HR] = 28.9; 95% confidence interval [CI], 3.66–216.61; P =.001) and multivariate (HR = 25.22; 95% CI, 3.3– 192.94; P =.002) analysis. CONCLUSION: Patients with higher nontarget lung FDG avidity appear to be at greater risk for radiation pneumonitis following SABR. (P009) Monte Carlo Dosimetry Evaluation of Lung Stereotactic Body Radiosurgery Colbert A. Parker, BS, Roger Ove, MD, PhD, Madhu B. Chilikuri, PhD, Suzanne M. Russo, MD; University of South Alabama School of Medicine; Mitchell Cancer Institute BACKGROUND: Promising results have been obtained using stereotactic body radiosurgery (SBRT) for early-stage lung cancer. The calculation of dose in pulmonary parenchyma can be inaccurate. MATERIALS AND METHODS: We retrospectively analyzed 47 cases treated over a 2-year period with CyberKnife SBRT, planned with the standard pencil beam (PB) algorithm. Cases were a mixture of early-stage lung cancer and oligometastatic cases. The median prescribed dose was 50 Gy in four or five fractions. We compared the planned dose with the dose actually delivered, as estimated with Monte Carlo (MC) dosimetry to the 1% level. We correlated the dosimetric deficiencies with recurrences, using deformable registration to determine the dose delivered to the site of recurrence. RESULTS: With a median follow-up of 2 years, the local control at 1 year was 90%, declining to 70% at 2 years. The total number of local recurrences was 10, and 8 of these patients died with progressive disease. Two recurrences occurred synchronously with metastases, and two recurrences were in palliative cases treated to lower doses with tight margins, and disease was never cleared locally. MC calculations showed that the mean dose delivered to the planning target volume (PTV), averaged over all cases, was 7% lower than planned. Most cases were planned with an expansion on the PTV (PTVmicro = GTV + 8-mm expansion in lung + 3 mm), representing a region at risk for microscopic extension, and were intended to receive a minimum dose of 80% of the prescription dose. MC calculations showed that the minimum dose to this structure, averaged over all cases, was 47% lower than the intended dose. For cases that recurred, the mean dose to the PTVmicro was 8% lower than intended, while it was only 2% lower for those whose disease was controlled. There were no other significant differences in target coverage between patients with local control and local recurrence. The PB algorithm and MC estimates for pulmonary exposure were assessed, recording the V5, V10, and V20 for the ipsilateral and total lung volumes. These estimates roughly agreed for the two algorithms, with the MC results almost universally lower than PB by an absolute difference of 1% to 3% on average. CONCLUSIONS: Without the use of MC planning, target structures were substantially underdosed. Local failures were associated with PTVmicro undercoverage, which suggests that delivering a therapeutic dose to this expanded microscopic disease target volume is beneficial. MC dosimetry is preferable for lung SBRT, while the PB algorithm is adequate for predicting pulmonary toxicity. 30 (P010) Stereotactic Body Radiotherapy for Treatment of Adrenal Gland Metastasis: Toxicity, Outcomes, and Patterns of Failure William W. Chance, MD, Quynh-Nhu Nguyen, MD, Reza J. Mehran, MD, James W. Welsh, MD, Daniel R. Gomez, MD, Peter A. Balter, PhD, Ritsuko Komaki, MD, Zhongxing Liao, MD, Joe Y. Chang, MD, PhD; UT MD Anderson Cancer Center PURPOSE: A single-institution, retrospective review of toxicity, patterns of failure, and outcomes in patients undergoing stereotactic body radiotherapy (SBRT) for metastasis to the adrenal gland. MATERIALS AND METHODS: From 2009–2014, a total of 36 patients with 40 adrenal metastases were treated with SBRT. The median age of the patient population was 63.7 years (range: 50–77 yr). Primary sites included lung (n = 31), ovary (n = 2), bladder (n = 1), esophagus (n = 1), and melanoma (n = 1). A total of 32 patients received treatment to a single adrenal gland, while 4 patients received treatment to the bilateral adrenal glands. The prescription dose to the target was 60 Gy in 10 fractions (n = 25), 50 Gy in 10 fractions (n = 10), 50 Gy in 4 fractions (n = 4), or other (n = 4). Failures within the prescribed high-dose irradiated region were considered local failures. New or progressive distant metastases outside of the treated adrenal gland were considered distant failures. After review of the radiation treatment plan, local failures were characterized as in-field (epicenter within the 100% isodose line) or marginal (between the 50% and 100% isodose lines). RESULTS: Median follow-up was 12.6 months after radiation treatment (range: 1.4–37.1 mo). Six patients developed grade 1 gastrointestinal toxicity. Two patients who were treated to the bilateral adrenal glands developed grade 2 adrenal insufficiency. There was no grade 3 or 4 toxicity observed. Median overall survival after treatment was 19.5 months. The 1- and 2-year overall survival rates were 67% and 55%, respectively. The 1-year freedom from local failure was 87%. The median time to local failure was not reached. The 1-year freedom from disease progression was 25% (median time to progression, 5.9 mo). The 1-year freedom from distant failure was 33% (median time to distant failure, 6.0 mo). There were six in-field local failures and one marginal failure observed. CONCLUSION: SBRT is an effective treatment modality, achieving excellent local control with minimal toxicity for patients with adrenal metastases. The development of progressive distant metastasis is the predominant pattern of failure affecting patients’ survival outcomes. (P011) Stereotactic Radiosurgery and BRAF Inhibitor Therapy for Melanoma Brain Metastases Is Associated With Increased Risk for Radiation Necrosis. Kirtesh Patel, MD, Roshan Prabhu, MD, Mohammad K. Khan, MD; Winship Cancer Institute, Emory University; Levine Cancer Institute, Carolinas Medical Center INTRODUCTION: Melanoma is an aggressive malignancy with a deplorable penchant for spreading to the brain. BRAF inhibitors, as single agents, have demonstrated intracranial efficacy. The 2015 National Comprehensive Cancer Network (NCCN) guidelines American Radium Society Scientific Papers and Posters 2015 have raised concerns about increased toxicity when BRAF inhibitors are combined with radiotherapy, based on limited case reports. Thus, we investigate the safety and efficacy of stereotactic radiosurgery (SRS) and BRAF inhibitors for melanoma brain metastases patients. METHODS: We reviewed melanoma patients with newly diagnosed brain metastases from 2005–2012. Radiation necrosis was compared by Fisher’s exact test; local control, intracranial control, and overall survival were estimated by the Kaplan-Meier method. RESULTS: A total of 72 patients received SRS, with 12 (16.7%) also receiving BRAF inhibitor therapy. BRAF inhibitor patients were similar to SRS-alone patients, except for having a higher percentage of RPA class 3 patients (25.0% vs 0.0%; P = .0030) and lower rates of solitary metastases (25.0% vs 55.0%; P = .034). No significant differences between radiation therapy total dose and dose per fraction, gross tumor volume (GTV), prescription isodose line, or conformality index were identified. Rates of all grades of radiation necrosis were statistically higher in the BRAF cohorts (58.3 vs 26.7%; P = .044); furthermore, symptomatic radiation necrosis was also more frequent in the BRAF inhibitor group (41.7% vs 15.0%; P = .048). One-year local control (83.5% vs 83.5%, P = .835) and intracranial control were similar between cohorts (30.3% vs 26.3%; P = .395). One-year overall survival was higher in the BRAF inhibitor group but not statistically significant (72.7% vs 38.1%; P = .172). CONCLUSION: Rates of symptomatic radiation necrosis appear to be higher for the BRAF inhibitor therapy group. Prospective studies investigating BRAF inhibitor therapy and SRS for melanoma brain metastases should consider incorporating methods to decrease potential radiation necrosis, including fractionating radiosurgery. (P012) Characteristics of Mentorship During Radiation Oncology Residency Gurleen Dhami, MD, Wendy Gao, MD, Michael Gensheimer, MD, Andrew D. Trister, MD, PhD, Gabrielle M. Kane, MD, Jing Zeng, MD; University of Washington Medical Center PURPOSE: The mentor-mentee relationship is crucial during residency training to foster professional and personal growth. Multiple surveys in other residency specialties have identified the mentorship’s role in professional development and career satisfaction. This national survey of radiation oncology residents was conducted to identify key characteristics of resident mentorship to identify areas for improvement. MATERIALS AND METHODS: An anonymous questionnaire was sent to all radiation oncology residents/recent graduates at Accreditation Council for Graduate Medical Education (ACGME)accredited residency programs, identified per the Association of Residents in Radiation Oncology (ARRO) member directory. Questions assessed demographics, prevalence of formal mentorship program, opinions about the value of mentorship, and details of the relationship. Responses were scored on a 5-point ordinal Likert scale. RESULTS: Out of 596 survey invitations, 157 residents responded (25%). Just over half of respondents had a current mentor (53%). The majority found their mentors early in training, with 49% identified during postgraduate year (PGY)-2, 28% establishing relationships during medical school, and only 14% during PGYs 3,4, and 5. Mentors were most commonly selected from faculty members whom the residents knew (64%). They were also identified by contacting faculty members with similar interests (33%) and by recommendations from others (29%). Most mentors were chosen from the junior faculty (29%) but also included senior faculty members (19%), radiation oncologists at another academic center (18%), other faculty members at the same institution (10%), program directors (11%), department chairs (6%), and others (7%). Residents often had more than one active mentor, with most having two to three active mentors (68%), while 26% had one mentor, and 6% had more than active mentors. Frequent and regular meetings were a common feature of the mentor-mentee relationship, with 28% meeting more than once per week, 30% meeting once per week, and 30% meeting once per month. Only 13% of residents met twice per year or less with their mentors. The most valued trait that was desired from a mentor was approachability (90%). Other desired traits included availability (82%), ability to provide opportunities (77%), and a clinical role model (68%). The majority of residents found mentorship helpful for career development (83%), research (75%), lifestyle/personal advice (71%), networking (46%), and coping with residency (46%). CONCLUSION: Almost half of current radiation oncology residents do not have a mentor. Of those with mentors, most established relationships early in their training, during PGY-2 or prior. Therefore, it is imperative to intervene early in the training process to produce successful mentorship experiences. Many residents require more than one active mentor, which enables multiple goals to be met, such as career development, increasing one’s research portfolio, networking, and coping with residency. (P013) Perioperative Mortality in Nonelderly Adult Patients With Cancer: A Population-Based Study Evaluating Healthcare Disparities in the United States According to Insurance Status Arya Amini, Norman Yeh, Bernard Jones, Yevgeniy Vinogradskiy, Edward Bedrick, Chad G. Rusthoven, Ava Amini, William T. Purcell, Brian D. Kavanagh, Sana D. Karam, Christine M. Fisher; University of Colorado; Northwestern University PURPOSE: Cancer survival is known to vary based on socioeconomic factors, including insurance status. The purpose of this study was to evaluate predictors for perioperative mortality (death within 30 d of cancer-directed surgery) for the 15 most common surgically treated cancers in the United States. PATIENT AND METHODS: The Surveillance, Epidemiology, and End Results (SEER) database was examined for the 15 most common surgically resected cancers. The database was queried from 2007 to 2011, with a total of 506,722 patients included in the analysis. Binomial logistic regression was used to assess the effect of ARS PROCEEDINGS 2015 31 American Radium Society Scientific Papers and Posters 2015 patient and tumor characteristics on perioperative mortality under multivariable analysis. RESULTS: The insurance status for all patients was as follows: non- Medicaid insurance (83%), any Medicaid (10%), uninsured (4%), and unknown (3%). Under univariable analysis, predictors for perioperative mortality included Medicaid or uninsured status (P < .001), older age (≥ 60 yr) (P = .015), nonwhite race (P < .001), being unmarried (P < .001), urban and rural residence (vs metropolitan) (P = .002), higher percent of county below the federal poverty level (P < .001), and lower median household income (P < .001). Perioperative mortality was also associated with more advanced disease, including higher tumor stage (P < .001) and metastasis (P < .001). After adjusting for age, race, sex, marital status, residence (urban or rural), extent of disease (in situ, local, regional, distant), and percentage of county below federal poverty level, patients with either Medicaid insurance (odds ratio [OR] = 1.22; 95% confidence interval [CI], 1.15–1.29; P < .001) or uninsured status (OR = 1.75; 95% CI, 1.61–1.87; P < .001) were more likely to die within 30 days of surgery compared with patients with non-Medicaid insurance. Additional statistically significant predictors for perioperative mortality under multivariable analysis included rural residence (OR = 1.07), race—including being African-American (OR = 1.07), Hispanic, (OR = 1.27), and Asian or Pacific Islander (OR = 1.19)— and being unmarried (OR = 1.08). CONCLUSION: In the largest reported analysis of perioperative mortality evaluating the 15 most common surgically treated malignancies, those with Medicaid coverage or without insurance were more likely to die within 30 days of surgery. (P014) Absence of Infection From Injection of a Rectal Spacer Into the Anterior Perirectal Fat Space Jekwon Yeh, MD, Justin Ren, Kenneth Tokita, MD, John Ravera, MD; Cancer Center of Irvine OBJECTIVE/PURPOSE: There is now increasing literature to support the use of a rectal spacer to decrease rectal side effects during radiation for prostate cancer. A rectal spacer (polyethylene glycol hydrogel) is usually placed via transperineal injection behind Denonvillier’s fascia to enhance the separation between the prostate and rectum. This causes a decrease in radiation dose to the rectum. Occasionally, the needle can penetrate the rectal wall, and the spacer material is accidently injected into the rectal lumen. This study aims to evaluate the rate of infection from rectal spacer placement and in patients who experienced rectal wall penetration with the rectal spacer applicator needle. MATERIALS AND METHODS: From January 2010 to May 2014, a total of 370 patients had a rectal spacer placed via transperineal injection. Patients were instructed to perform an enema the night before and immediately prior to the surgery. The perineum was also sterilized with Betadine prior to the procedure. Patients also took ciprofloxacin 500 mg bid for 10 days, starting the day prior to the procedure. Patients also received gentamicin 80 mg and cefazolin 1 gram intraoperatively. Afterwards, a computed tomography (CT) scan and magnetic resonance imaging (MRI) were performed on all patients to confirm placement of the rectal spacer and for 32 external beam radiation planning. RESULTS: Out of 370 patients who received the hydrogel spacer, no patient developed a rectal infection. A total of 18 (5%) patients had known rectal penetration seen on imaging or clinically. None of these patients experienced any rectal infections. CONCLUSIONS: With proper preparation and antibiotics, placement of the rectal spacer into the anterior perirectal fat is extremely safe, without any risk of infection, in our series. Also, infections were not found in patients who had rectal wall penetration with the rectal spacer applicator needle. (P015) Radiation Biological Responses of MRI-Linac vs Linac in Human Head and Neck and Lung Cancer Cells Li Wang, Marco van Vulpen, Zhifei Wen, Stan Hoogcarspel, David P. Molkentine, Jan Kok, Steven Hsesheng Lin, R. Broekhuizen, Kie-kian Ang, A.N. Bovenschen, Bas Raaymakers, Steven Jay Frank; UT MD Anderson Cancer Center; University of Utrecht Medical Center OBJECTIVE: The MRI-Linac (MRL) is a novel radiotherapy technology designed to enable high cure rates with low toxicity and treatment response monitoring. However, the biological influence on radioresponses of human solid tumors within the 1.5-T magnetic field (MF) generated by MRL is unknown. Our hypothesis is that the 1.5-T MF has no influence on cell viability or radiation effects. METHODS: Two non–small-cell lung carcinoma (NSCLC) lines (H460 and H1299) and two head and neck squamous cell carcinoma (HNSCC) lines (HN-5 and UMSCC-47) were used. The influences of a 1.5-T MF on in vitro cell viability and on cellular radioresponses were determined by cell plate efficiency (PE) and by clonogenic cell survival assay (CSA) after exposing the cells to graded single doses or fractionated 6 MV radiation using Linac (100 MU/min) or MRL (134 MU/min), respectively. The experiments were performed with cell flasks placed inside a 37°C water bath with an optimized radiation dose distribution. The radiation output of Linac and the MRL was measured with a corrected chamber. The physical dose received by the cells was verified using radiation detectors. The results were analyzed by t-test. RESULTS: Our results supported a very strong trend that MF, as a single factor, had no influence on cell viability. No significant PE change occurred after single exposure or multiple exposures to MF compared with the Linac environment for the four cell lines (12 of 16 experiments; all P > .05) except for one experiment with the H460 cells (borderline; P = .044) and three experiments with UMSCC-47 cells. Similarly, single exposure or multiple exposures to MF had no influence on cell radioresponse (all P > .05). When the cells were exposed to MRL or Linac treatment for up to four times, no significant changes were observed for the D0s for with MF vs without MF (all P > .05). CONCLUSIONS: The 1.5-T MF generated by MRL had no effect on the viability or radioresponses of NSCLC or HNSCC cell lines in vitro. These results suggest that MRL, as a novel cancer treatment technology, has the potential not to influence the radiotherapy out- American Radium Society Scientific Papers and Posters 2015 come of patients. Considering the complicated in vivo microenvironment, further in vivo study is warranted. (P016) The Cell Cycle Inhibitor P21 Regulates Langerhans Cell Radiation Resistance and Promotes T Regulatory Cell Induction Upon Exposure to Radiotherapy Jeremy G. Price, Juliana Idoyaga, Brandon Hogstad, Helene Salmon, Marylene Leboeuf, Miriam Merad; Icahn School of Medicine at Mount Sinai; Stanford University Several studies have revealed that exposure to ionizing irradiation (IR) may lead to increased accumulation of tumor-infiltrating T regulatory cells (Tregs), which in turn promotes tumor resistance to radiation therapy. Here, we questioned the contribution of tissue-resident antigen-presenting cells (APCs) to the induction of Tregs upon exposure to IR. Specifically, we focused on Langerhans cells (LCs), the resident APCs of the epidermis, because of their unique ability to resist depletion by high-dose IR. Therefore, LC IR resistance may inform us of the underlying IR resistance mechanisms utilized by other progenitor cells. However, a comprehensive study of the molecular and cellular mechanisms conferring LC IR resistance has never been undertaken. We found that LCs do not undergo apoptosis following IR as do other dendritic cell (DC) subsets but instead persist and migrate to skin-draining lymph nodes. Subsequent analysis in migrationdeficient CCR7−/− mice revealed a constant number of epidermal LCs, indicating that changes in LC number following IR are due solely to migration and not to cell death. Moreover, we show for the first time that LCs are resistant to the formation of DNA damage, as measured by induction of H2AX foci and COMET assay. In contrast, other members of the myeloid lineage, notably lymphoid tissue-resident DCs, are exquisitely sensitive to IR-induced DNA damage. In addition, both steady-state and postlethal IR LCs express a unique repertoire of prosurvival and stress-related proteins and diminished levels of proapoptotic molecules by microarray analysis. Additionally, we found that the cell cycle regulator p21 is overexpressed in LCs at rest and that in contrast to WT LCs, p21−/− LCs undergo apoptosis, accumulate significant DNA damage, and fail to experience cell cycle arrest following IR. Strikingly, upon skin exposure to IR, WT LCs lead to significant expansion of Treg, whereas p21−/− LCs fail to do so. In a cutaneous tumor model, we show that p21−/− LCs cannot promote tumor-infiltrating Tregs and correlate with smaller tumor volumes. Moving forward, these data suggest a novel means by which targeting LC IR resistance can be used to increase the response to radiotherapy of cutaneous tumors. LCs uniquely express an IR resistance module of genes that permits them to persist and repopulate the epidermal niche following IR. Future analyses will seek to elaborate the functional implications of LCs and IR-sensitive DC IR-exposure as it relates to the priming and development of graft-versus-host disease (GVHD) and local antitumor immunity following radiation therapy. (P017) Dosimetric Evaluation of Respiratory-Gated Radiotherapy for Left-Sided Breast Cancer Veronica Finnegan, MD, Varun Chowdhry, MD, Weidong Li, PhD, Katrina Stellingworth, CMD, Jeffrey Bogart, MD, Anna Shapiro, MD; SUNY Upstate; Massachusetts General Hospital PURPOSE/OBJECTIVES: Adjuvant radiotherapy is associated with improvements in local control and survival in patients with breast cancer. As oncological outcomes have improved, there is a greater importance in preventing long-term toxicity from treatment. Technological advancements in radiotherapy delivery have the potential to allow better coverage of complex targets while reducing doses to normal structures. The use of four-dimensional computed tomography (4DCT) and respiratory gating has increased in clinical practice, but there is limited information regarding the utility of respiratory gating in breast cancer. The purpose of the study is to assess the dosimetric benefits of respiratory gating for patients receiving radiotherapy for left-sided breast cancer. PATIENTS/METHODS: A total of 38 women with left-sided breast cancer were treated with respiratory gating between 2009 and 2013. All patients underwent 4DCT simulation using the Varian RealTime Position Management respiratory gating system. The determination to use respiratory gating was at the discretion of the treating physician. For this study, maximum intensity projection (MIP) image sets containing images from all phases were also created. The original plans were copied to both the gated MIP and the all-phase MIP sets for dosimetric evaluations. Dose-volume histograms were calculated and compared. Doses to predefined heart and lung parameters were compared on two plans for each patient. Cardiac and pulmonary doses were compared for each patient using a twosided paired difference test (t-test). RESULTS: The use of respiratory gating resulted in statistically significant dose reductions to the heart on nearly all parameters evaluated. Mean whole heart dose was 368 cGy with gating and 389 cGy without gating (P < .001). V5 heart was 77% with gating and 84% without gating (P < .001). Max left anterior descending (LAD) dose was 3,990 cGy with gating and 4,264 cGy without gating (P = .009), and mean left ventricular dose was 511 cGy with gating and 549 cGy without gating (P < .001). There were trends toward a reduction of mean LAD dose (2,298 cGy with gating and 2,569 cGy without gating; P = .059) and max left ventricular dose (4,021 cGy with gating and 4,183 cGy without gating; P = .067). Statistically significant differences were not noted in any of the lung parameters tested (V5, V20, or V40). CONCLUSION: The use of respiratory gating has resulted in small but statistically significant reductions in heart dose. Further studies are needed to understand the clinical implications of these differences. (P018) Accelerated Partial Breast Radiotherapy Using VMAT—A Preliminary Dosimetric Comparison to Single-Entry Brachytherapy Yucel Saglam, Yasemin Bolukbasi, Steve Kirsner, Vildan Alpan, Duygu Sezen, Ugur Selek; American Hospital, MD Anderson Radiation Treatment Center, Istanbul; Department of Radiation Oncology, UT MD Anderson Cancer Center; Department of Radiation Oncology, Koc University ARS PROCEEDINGS 2015 33 American Radium Society Scientific Papers and Posters 2015 PURPOSE: To compare the dosimetric aspects of accelerated partial breast irradiation (APBI) with a single-entry brachytherapy device, strut-adjusted volumetric (SAVI), and external APBI using volumetric arc therapy (VMAT), with the intent of investigating the feasibility of using VMAT for the delivery of APBI treatments. MATERIALS AND METHODS: Five left breast cancer patients who were planned and treated with SAVI at MD Anderson Cancer Center in Istanbul at American Hospital were randomly and retrospectively chosen from a database for this preliminary study. All patients were staged with T1N0M0 invasive breast cancer after partial mastectomy and sentinel lymph node dissection. The age of the patients ranged from 50 to 82 years. None of the patients received chemotherapy. The average cavity size was 12.6 mm (range: 2.2–18 mm), and all patients had invasive ductal carcinoma. The plans and their contours were sent from the Oncentra Treatment Planning System to the Philips Pinnacle treatment planning system for replanning. VMAT APBI treatment plans were run on these same patients using the same contours delineated in the Oncentra planning system. The target volumes and planning parameters were set according to the National Surgical Adjuvant Breast and Bowel Project (NSABP)-39 protocol for external beam APBI, with a prescription dose of 38.5 Gy in 10 fractions. Dosimetric criteria on the heart and ipsilateral lung were used to compare the two techniques. Paired two-tailed Student’s t-test was performed to assess the differences in lung mean dose, heart mean dose, V20, V10, and V5 (%). RESULTS: Even VMAT resulted in a decreased low-dose volume of the lung (V5: 0.1% vs 12.2%; P = .03); mean dose (204.1 cGy vs 287.2 cGy; P = .08), and V10 (0.02% vs 2.83%; P = .132) and V20 (0.001% vs 4.540%; P = .063) percentages of the lung were found to be similar. The difference in heart mean dose (243.6 cGy vs 342.8 cGy; P = .022) was influenced by the significant difference of heart V5 (0.083% vs 21.4%; P = .004), but the V20 (0.0001% vs 0.126%; P = .374) and V10 (0.008% vs 3.448%; P = .104) values of the heart were found to be comparable between the two techniques. CONCLUSION: VMAT planning for the delivery of APBI in this preliminary dosimetric evaluation has been shown to be a viable option for APBI. VMAT planning gives acceptable lung and heart doses and appears to deliver lower doses to the heart and ipsilateral lung than APBI delivered with the single-entry SAVI applicator. (P019) Predictors of Outcomes in Breast Cancer Patients With Oligometastases with oligometastases tend to fare better than patients with oligometastases from other primary sites. However, no studies have examined predictors of outcomes in patients with oligometastatic breast cancer. MATERIALS AND METHODS: We identified 380 patients treated at our institution with SBRT or SRS from 2010–2014. We retrospectively reviewed the records of 62 patients with metastatic breast cancer. We defined oligometastases as metastases at ≤ 5 distinct clinical sites. Oligometastases were determined by imaging and clinical documentation. Kaplan-Meier curves were used to determine survival after metastasis. A Cox proportional hazards model was used to assess the effect of patient, tumor, and treatment characteristics as predictors of survival after metastases. RESULTS: There were 30 patients with oligometastases and 32 patients with non-oligometastases. A total of 60% of patients had initial stage I, II, or III disease, while 40% had stage IV. The tumor was estrogen receptor (ER)-positive, progesterone receptor (PR)positive, and human epidermal growth factor type 2 (HER2)positive in 63%, 46%, and 20% of oligometastatic patients and in 66.7%, 57%, and 28% of nonoligometastatic patients, respectively. Triple-negative disease was noted in 27% of patients in both arms. Further, 95% received chemotherapy and about 60% received hormonal therapy in both groups. The 5-year survival rate after metastases was 79.6% for oligometastatic vs 45.7% for nonoligometastatic patients. Univariate models showed that the predictors of inferior survival after metastases included: absence of oligometastases (hazard ratio [HR] = 3.39; 95% confidence interval [CI], 1.28–9.19; P = .016), HER2-negative status (HR = 12.42; 95% CI, 1.63–94.56; P = .015), and triple-negative status (HR = 3.61; 95% CI, 1.43–9.07; P = .006). On bivariate analyses, absence of oligometastases remained a significant predictor of worse survival after metastasis: absence of oligometastases (HR = 6.07; 95% CI, 1.56– 23.710; P = .009) adjusted for HER2-positive status (HR = 17.55; 95% CI, 2.13–144.84; P = .008) in model 1 and absence of oligometastases (HR = 5.31; 95% CI, 1.54–18.31; P = .008) adjusted for triple-negative status (HR =3.65; 95% CI, 1.43–9.29; P = .007) in model 2. CONCLUSION: Our study showed that oligometastatic breast cancer patients have improved 5-year survival after metastases compared with non-oligometastatic patients. In patients with oligometastases and HER2-positive disease or without triple-negative disease, survival after metastases was superior. Further studies are needed to identify a favorable subset of patients with oligometastases who would benefit from aggressive therapy. Shefali Gajjar, MD, Arthy Yoga, MD, Isildinha M. Reis, PhD, Youssef Zeidan, MD, PhD, Cristiane Takita, MD; Department of Radiation Oncology, Department of Epidemiology & Public Health, University of Miami/Sylvester Comprehensive Cancer Center; Department of Surgery, University of Miami (P020) 3D Conformal External Beam Radiation Therapy May Result in Lower Heart Dose and Risk of RadiationInduced Major Coronary Events Compared With Multicatheter Balloon High-Dose-Rate Brachytherapy for Left-Sided Breast Cancer Patients BACKGROUND: Oligometastases are hypothesized to represent a potentially curable disease. With the emergence of stereotactic body radiotherapy (SBRT) and stereotactic radiosurgery (SRS), additional treatment options have been presented to patients with limited metastases. Studies have shown that breast cancer patients Jason C. Ye, MD, Brittney Wilson, BS, CMD, Igor Shuryak, MD, Jenghwa Chang, PhD, Samuel Trichter, MSc, David Brenner, MD, DSc, Dattatreyudu Nori, MD, John Ng, MD; Department of Radiation Oncology, Weill Cornell Medical College; Center for Radiologic Research, Columbia University Medical Center 34 American Radium Society Scientific Papers and Posters 2015 PURPOSE: To determine and compare the heart dose and potential increased risks in long-term cardiac toxicity from three-dimensional conformal external beam radiation therapy (3DCRT) and multicatheter balloon high-dose-rate (HDR) brachytherapy treatments for left-sided breast cancer patients. METHODS: Fifteen consecutive patients who were treated with balloon HDR (3,400 cGy in 10 fractions, twice daily) after lumpectomy for left-sided breast cancer between 2011 and 2014 were included in this study. Target volumes drawn for HDR brachytherapy. Left breast, heart, and left anterior descending artery (LAD) were contoured by a physician in the brachytherapy treatment planning system and independently confirmed by another physician. 3DCRT plans (5,040 cGy in 28 fractions using opposed tangential beams) were developed using the same computed tomography (CT) scans and contours by the same team by transferring the image and structure sets to the external beam planning system. Appropriate target coverage, dose distribution, and dose homogeneity (± 7%) were confirmed independently. Cardiac blocks were allowed, as long as they did not significantly affect the breast and lumpectomy cavity coverage. The radiation doses to the heart and LAD using 3DCRT and HDR techniques were recorded. The risk of long-term additional cardiac toxicity was estimated by calculating the 10-year radiation-induced estimated actual risk (EAR) for major coronary events using a previously published cardiac risk model. RESULTS: The 15 women (median age: 65 yr, range: 59–86 yr) all had early-stage localized disease that fit the American Society for Radiation Oncology (ASTRO) guidelines for accelerated partial breast irradiation. Average of the mean heat doses (MHDs) delivered using HDR was 251.6 cGy (range: 137–427 cGy), while the average MHD that they would have received using 3DCRT was 128.99 cGy (range: 90–236.4 cGy) (Student’s t-test: P < .001). The mean dose to the LAD was not statistically significant (551 cGy for HDR vs 814 cGy for 3DCRT; P = .11 by Student’s t-test). After factoring in MHD, age, smoking status, history of hyperlipidemia, hypertension, and diabetic history, the mean 10-year radiationinduced EAR for 3DCRT and HDR for the 15 women was 0.51% (range: 0.17%–1.51%) and 1.01% (range: 0.31%–3.5%), respectively (Student’s t-test: P = .0064). CONCLUSIONS: Both 3DCRT and balloon HDR can achieve relatively low MHDs and result in minimal increases in the risk of additional major coronary events. In certain cases, compared with HDR brachytherapy, 3DCRT may result in lower MHDs and a lower risk of long-term cardiac toxicity. (P021) Improving Clinical Documentation and Prospectively Populating a Research Database Through an Electronic Data Capture System for Routine Clinical Care Hubert Y. Pan, MD, Timothy J. Edwards, BS, David P. Giragosian, PsyD, Emma B. Holliday, MD, Cameron W. Swanick, MD, Geoffrey V. Martin, MD, Karen E. Hoffman, MD, Simona F. Shaitelman, MD, Wendy A. Woodward, MD, PhD, Benjamin D. Smith, MD; UT MD Anderson Cancer Center INTRODUCTION: With the current emphasis on value-based health care, there is increased importance of quantifying value and reducing cost. Electronic health records (EHRs) are being adopted at an increasing rate but often capture information as unstructured text, requiring time-consuming retrospective reviews to extract outcomes data. Electronic data capture (EDC) using case report forms is commonly used to collect structured data for clinical trials. We implemented a web-based EDC system for routine clinical care and describe our experience piloting the system for our breast radiation therapy (RT) service. METHODS: Our institution uses dictation and transcription for clinical documentation stored in an in-house–developed EHR. A separate RT-specific system contains RT prescriptions, schedules, and treatment records. The implemented EDC serves as an intermediary between these two systems. Providers specify patient, tumor, and treatment characteristics through the EDC using structured data fields. These fields are merged with relevant data extracted from the RT system to generate template-based notes in the EHR for simulation, treatment planning, quality assurance (QA), weekly on-treatment visits (OTVs), and treatment summaries. Dictation and EDC times are reported as means and compared using t-tests. RESULTS: The EDC had been used by 21 providers and generated 850 notes for 127 patients during the most recent month. A consult form collecting tumor and treatment details was completed in the EDC as an initial additional step, requiring 2.5 minutes. This form generated a patient identifier used in all subsequent notes. Simulation notes, treatment planning notes, and treatment summaries were all completed more quickly using the EDC as compared with dictation (P < .001). QA notes recording the results of weekly treatment plan reviews were completed in real time during the conference (2.1 min with dictation). OTV notes were often completed during the patient encounter using a tablet (1.8 min with dictation). The total documentation time for a typical course of breast RT was 5.4 minutes with EDC and 22.1 minutes with dictation. CONCLUSION: We implemented an EDC system for routine clinical use in our breast RT service that resulted in significant time savings for clinical documentation and a prospective population of a database for future outcomes research. Additional follow-up is needed to determine how easily this system can be generalized to other RT disease sites and practices. (P022) Proton Therapy on an Incline Beam Line: Acute Toxicity Outcomes in Locally Advanced Breast Cancer Patients Lisa A. McGee, MD, Molly McGue, Megan Dunn, PhD, Stacey Schmidt, Darren Kaplan, Mark Pankuch, William Hartsell, MD; CDH Proton Center; Proton Collaborative Group background: Adjuvant radiotherapy for locally advanced breast cancer (LABC) is known to reduce local recurrence. Proton therapy can be used to improve the therapeutic ratio by sparing the dose to adjacent nontargeted tissues, including the heart and lung, in breast cancer patients requiring treatment of comprehensive regional lymphatics, including the internal mammary lymph nodes (IMNs). This retrospective series evaluates the acute toxicity outARS PROCEEDINGS 2015 35 American Radium Society Scientific Papers and Posters 2015 comes for LABC patients treated with uniform scanning (US) proton therapy (PT) on an incline beam line (IBL). METHODS: From September 2011 to May 2014, a total of 35 patients with LABC received adjuvant US PT targeting either the chest wall (n = 25) or intact breast (n = 10) plus comprehensive regional lymphatics, including the IMNs. Patients were treated on the IBL with one of the following techniques: superior and inferior anterior superior oblique (ASO) fields alternated every other day with a superior and inferior en face anterior oblique (AO) field (n = 12), all 4 fields daily (n = 22), or a superior and inferior en face AO field only (n = 1). Patients were simulated in a supine position, immobilized in an alpha cradle with the thorax rotated 30 degrees to approximate an en face angle achieved with the AO fields. Toxicity was prospectively assessed using Common Terminology Criteria for Adverse Events version 4.0 (CTCAE v4.0) weekly during treatment and at 4 weeks following treatment completion. Outcomes: Median follow-up was 3 months. A total of 34 women and 1 man were treated with adjuvant US PT for stage I (n = 1), II (n = 7), III (n = 27) LABC; 20 were left-sided, 12 were right-sided, and 3 were bilateral. Of the 25 postmastectomy patients, 14 were reconstructed. Initial fields were planned to 50.4 cobalt gray equivalent (CGE) (n = 33) or 45 CGE (n = 2), with 13 patients having a planned boost of 10 CGE (n = 9), 14 CGE (n = 1), 16 CGE (n = 1), and 20 CGE (n = 1). Median total dose received was 50.51 CGE (range: 45.11–70.36 CGE) All patients experienced acute radiation dermatitis: grade 1 (n = 8), grade 2 (n = 26), and grade 3 (n = 3). Five patients required a break in treatment, and two patients did not complete their full course of prescribed PT due to skin toxicity. Also, 16 patients experienced grade 1 esophagitis, and 13 patients experienced grade 1 chest wall pain. Two patients experienced skin infection following treatment, requiring antibiotic treatment, which occurred the second week following treatment completion. CONCLUSIONs: These early results demonstrate the feasibility of adjuvant breast cancer treatment with US PT on an IBL. Acute toxicity results appear acceptable. Longer follow-up is needed. (P023) Dosimetric Evaluation of Accelerated Partial Breast Irradiation Utilizing the ViewRay Magnetic Resonance Image-Guided Radiation Therapy System Benjamin W. Fischer-Valuck, MD, Karl Sona, Sahaja Acharya, MD, Min Yang, Rojano Kashani, PhD, Imran Zoberi, MD, Maria Thomas, MD, PhD; Washington University PURPOSE: External beam radiation therapy (EBRT) is a noninvasive alternative to deliver accelerated partial breast irradiation (APBI). However, due to the larger target volume, cosmetic outcomes after EBRT APBI may be inferior to more invasive techniques, such as brachytherapy. Utilizing the ViewRay magnetic resonance imaging (MRI)-guided radiation therapy system, MRI can be used for planning and daily positioning, as well as for precise targeting during treatment. This prefraction imaging allows for a reduction in the planning target volume (PTV) margin, which may improve acute skin toxicity and cosmetic outcomes compared with 36 traditional EBRT methods. MATERIALS AND METHODS: Ten patients with ductal carcinoma in situ (DCIS) or early-stage invasive breast cancer and negative axillary lymph nodes were treated with APBI using the ViewRay system. Prescription dose to the PTV was 38.5 Gy, delivered in 10 fractions, twice daily. The PTV for ViewRay treatment planning included a 1-cm margin around the lumpectomy cavity. Comparison three-dimensional conformal EBRT (3DCRT) plans for each patient were generated using the Pinnacle planning system and a 2-cm margin around the lumpectomy cavity. Dosimetric parameters for the PTV, ipsilateral breast, and other critical structures for each treatment plan were compared. RESULTS: Using the ViewRay system to obtain MR imaging for setup prior to each fraction, we observed that the lumpectomy cavity could be precisely localized and aligned prior to each fraction. A decrease in the PTV margin using the ViewRay system resulted in a mean PTV volume of 85 cc compared with 177 cc for 3DCRT. The median PTV receiving 95% of the prescribed dose for both the ViewRay and 3DCRT plans was 100%. Mean volume of the ipsilateral breast receiving the prescription dose was 11.72% for ViewRay treatment plans compared with 21.56% for 3DCRT treatment plans (P < .04). Mean volume of the ipsilateral breast receiving 50% of the prescribed dose using the ViewRay treatment plan was 31.34% compared with 52.71% for the 3DCRT treatment plans (P < .02). Reductions in the mean volumes of the ipsilateral lung, heart, and contralateral breast with the ViewRay plans were also observed, although they were not statistically significant. CONCLUSION: MRI-guided EBRT using the ViewRay system is a novel approach to deliver APBI. ViewRay APBI is noninvasive yet maintains a high degree of precision by using prefraction MR imaging, thus allowing a reduction in the PTV margin. The resultant decrease in the ipsilateral breast dose may reduce acute skin toxicity and improve cosmetic outcomes. Thus, the ViewRay system is an attractive alternative to existing APBI techniques. (P024) Predictors of Radiation-Induced Skin Toxicity: Data From a Prospective Cohort Receiving Breast Radiation Jean Wright, MD, Cristiane Takita, MD, Isildinha M. Reis, Wei Zhao, MD, BS, Eunkyung Lee, MS, Jennifer Hu, PhD; Johns Hopkins University; University of Miami PURPOSE/OBJECTIVES: Breast radiation (RT) is generally well tolerated, but acute skin toxicity is a common side effect that can impact receipt of treatment and quality of life. We sought to identify predictors of radiation-induced skin toxicity in a racially and ethnically diverse cohort of women receiving RT to the intact breast. MATERIALS AND METHODS: We evaluated the first 392 patients in an ongoing prospective study assessing radiation-induced skin toxicity in patients receiving breast RT. We recorded patient demographics, body mass index (BMI), and disease and treatment characteristics. Skin toxicity grade using Common Terminology Criteria for Adverse Events (CTCAE) and a modified scale captur- American Radium Society Scientific Papers and Posters 2015 ing moist desquamation were captured at Week 3 and at RT completion. Logistic regression analyses were conducted to evaluate the effect of potential predictors on the risk of skin toxicity. RESULTS: A total of 20.2% self-identified as African American, 15% self-identified as non-Hispanic white, 61.5% self-identified as Hispanic white, and 3.3% self-identified as other. Mean age was 56.2 years, and 74.2% had a body mass index (BMI) ≥ 25. Disease was stage 0 in 20.2% (in situ), I in 49.2%, and II or III in 30.6%. Further, 17.2% of patients were treated with hypofractionated regimens to the breast +/− regional nodes with a whole-breast dose of < 45 Gy, and 82.8% of patients were treated with conventionally fractionated regimens with a dose of ≥ 45 Gy. On multivariate analysis, both scales identified higher BMI, higher disease stage, invasive ductal histology, progesterone receptor (PR)-negative status, and conventionally fractionated regimens with total radiation dose ≥ 45 Gy as predictors for higher skin toxicity grade. The modified scale also identified larger breast volume as a predictor for moist desquamation, and the CTCAE scale identified chemotherapy use as a predictor for grade 2/3 skin toxicity. CONCLUSION: In this prospectively followed cohort of breast cancer patients receiving RT to the intact breast, novel predictors for more severe skin toxicity were identified, including PR-negative status, invasive ductal histology, and receipt of chemotherapy, as well as known risk factors, including BMI and fractionation scheme. (P025) T3 Tumors and Breast Conservation Richard J. Bleicher, MD, Karen Ruth, MS, Elin R. Sigurdson, MD, PhD, Marcia Boraas, MD, Penny Anderson, MD, John M. Daly, MD, Brian L. Egleston, PhD; Fox Chase Cancer Center INTRODUCTION: Although breast conservation remains one stan- dard for treatment of breast cancer, tumors > 5 cm have been consistently excluded from breast conservation clinical trials. While many surgeons perform breast conservation in select patients whose tumors are T3 primaries, very few small series have compared this with mastectomy for tumors > 5 cm. This study was performed to determine if survival remains equivalent between the two surgical options, using a large national dataset. METHODS: Surveillance, Epidemiology, and End Results (SEER)Medicare–linked cases were identified for patients aged ≥ 66 years undergoing breast conservation who had invasive noninflammatory, nonmetastatic breast cancer between 1992 and 2009. Propensity score-based adjustment was used to account for age, gender, race, year of diagnosis, Charlson comorbidity score, Elixhauser comorbidity index, income, education, marital status, SEER registry, urban/rural setting, United States (US) location, tumor size, histology, tumor grade, number of lymph nodes examined, number of positive lymph nodes, American Joint Committee on Cancer (AJCC) stage, administration of chemotherapy, and administration of radiotherapy. RESULTS: There were 5,890 patients with tumors > 5.0 cm who underwent breast surgery, with 16.1% having breast conservation and 83.9% undergoing mastectomy. Mean age of the breast conservation and mastectomy patients was 77.6 and 77.3 years, and the mean tumor size was 7.97 vs 7.51 cm, respectively. While 28.4% of stage II patients with T3 tumors had breast conservation, only 8.1% of stage III patients did so. Over 64% of breast conservation patients and 68% of mastectomy patients had ≥ 5 years of follow-up. Use of breast conservation was associated with younger age at diagnosis, race, a history of other cancers, higher income, a more rural setting, region of the country, lower Charlson comorbidity index, histology, fewer nodes examined, fewer positive nodes, lower-grade tumors, lower AJCC stage, nonuse of chemotherapy, and the use of radiotherapy. Overall survival and breast cancer–specific survival, adjusted for demographic, tumor, and treatment factors, were not different between those having breast conservation and those having mastectomy (breast cancer–specific survival: subdistribution hazard ratio [SHR] = 1.07; 95% confidence interval [CI], 0.75–1.51; P = .71; overall survival: SHR = 0.92; 95% CI, 0.75–1.13; P = .43). CONCLUSIONS: For patients with tumors > 5 cm, survival in the Medicare population remains similar between breast conservation and mastectomy, as it does for smaller primaries. Despite the exclusion from prospective randomized trials, breast conservation should remain a standard option for women with larger tumors when deemed clinically and cosmetically amenable to resection. (P026) Is Cause-Specific Survival Similar for Estrogen Receptor- and Progesterone Receptor-Negative EarlyStage Invasive Lobular and Invasive Ductal Cancers? A National Registry SEER Database Study Justin M. Mann, Weisi Yan, Guojiao Wu, Dattatreyudu Nori, Akkamma Ravi; New York Presbyterian-Weill Cornell Medical Center BACKGROUND: Invasive ductal carcinoma (IDC) and invasive lobular carcinoma (ILC) are the two most common breast cancer histologies. IDC is more common and confers a worse prognosis than ILC. Biomarkers, including estrogen receptor (ER) and progesterone receptor (PR) status, improve prognostic accuracy and will likely be added to a future edition of the American Joint Committee on Cancer (AJCC) TNM classification. As a whole, patients with an ER- and PR-negative phenotype have a worse prognosis when compared with receptor-positive phenotypes. Using Surveillance, Epidemiology, and End Results (SEER) data, we performed statistical analysis to determine if receptor-matched statuses in IDC compared with ILC differ in overall outcome. METHODS: A total of 50,658 cases of early-stage (I–IIB) breast cancer (groups = IDC and ILC) diagnosed from 1998–2002 who underwent breast conservation surgery were queried from the SEER 18 database. Patients without known ER or PR receptor status were excluded. Kaplan-Meier survival analysis and log-rank test were used to compare breast cancer cause-specific survival (BCSS). RESULTS: There were 2,852 (5.63%) cases of ILC and 47,806 (94.37%) cases of IDC in this analysis. A total of 40,099 (79.2%) cases were ER+, and 34,901 (68.9%) cases were PR+. For ER+ cases, 60-month BCSS was 98% for ILC and 97.7% for IDC (P = .79). For ER− cases, 60-month BCSS was 95.5% for ILC and 89.7% for IDC (hazard ratio [HR] = 0.45; P = .02). For PR+ cases, 60-month BCSS was 98.3% for ILC and 97.8% for IDC (P = .35). For PR− cases, ARS PROCEEDINGS 2015 37 American Radium Society Scientific Papers and Posters 2015 60-month BCSS was 96.1% for ILC and 92.0% for IDC (HR = 0.77; P = .03). CONCLUSION: For early-stage breast cancer cases with ER+ and PR+ status, histology was not associated with a difference in BCSS. Alternatively, ILC cases that were ER− or PR− had an increased BCSS compared with receptor-matched IDC cases. These findings add to the growing evidence supporting ILC as a more favorable histology, which is important for guiding treatment and prognostication. (P027) Increasing Use of Postlumpectomy Radiotherapy for Ductal Carcinoma In Situ of the Breast in the United States Yi An, MD, Henry S. Park, MD, MPH, John M. Stahl, MD, Sue B. Evans, MD, MPH, Charles E. Rutter, MD; Yale University School of Medicine INTRODUCTION: Breast ductal carcinoma in situ (DCIS) is a preinvasive neoplasm historically treated with mastectomy prior to the adoption of the breast conservation paradigm. Although postlumpectomy radiation therapy (RT) for DCIS reduces local recurrence risk, its role in patients with favorable-risk disease is controversial. As such, RT utilization rates for DCIS may vary. Here, we are the first to study the change in usage of postlumpectomy RT for DCIS and the factors associated with these changes on a national level. METHODS: We identified patients diagnosed with DCIS and treated with lumpectomy between 1998 and 2011 within the National Cancer Database. Chi-square tests and multivariate logistical regression analyses were used to identify demographic, tumor, and treatment facility factors associated with increased likelihood of receiving postlumpectomy RT. RESULTS: We identified 144,861 DCIS patients who underwent lumpectomy. Median age was 59 years, 84% of patients were white, and 39% had high-grade disease. A total of 99,639 (69%) women received postlumpectomy RT. The proportion receiving postlumpectomy RT increased from 61% in 1999 to 72% in 2008 (P < .01). Statistically significant predictors of postlumpectomy RT included younger age, higher grade, fewer comorbidities, rural residence, and increased distance from treatment center (P < .01 for all). Interestingly, medium-sized tumors and negative surgical margins were also significant predictors of postlumpectomy RT relative to patients with larger tumors and positive margins, suggesting that these higher-risk patients may undergo complete mastectomy rather than receive adjuvant radiotherapy. CONCLUSION: The percentage of DCIS patients receiving postlumpectomy RT increased between 1999 and 2008. Younger patients with higher grade, medium-sized tumors, and fewer comorbidities were more likely to receive postlumpectomy RT. (P028) Adjuvant Radiation Therapy After Lumpectomy According to Insurance Status Norman Yeh, Arya Amini, MD, Christine Fisher, MD; University of Colorado Denver 38 PURPOSE: After a lumpectomy in early-stage breast cancer, adjuvant radiation therapy as a component of breast conservation is standard of care. The purpose of this study was to evaluate patterns of care for postlumpectomy radiation according to insurance status. PATIENT AND METHODS: The Surveillance, Epidemiology, and End Results (SEER) database was examined for patients with breast cancer aged under 65 years and with a stage of T1N0. The database was queried from 2007 to 2011, with a total of 72,257 patients included in the analysis. Multinomial logistic regression was used to assess patient and tumor characteristics under multivariable analysis. RESULTS: All patients had the following insurance status: nonMedicaid insurance (90%), any Medicaid (9%), and uninsured (2%). Medicaid (odds ratio [OR] = 0.77; 95% confidence interval [CI], 0.66–0.89; P < .001) and uninsured status (OR = 0.53; 95% CI, 0.40–0.71; P < .001) were more likely to not receive postlumpectomy radiation on multivariate analysis after adjusting for age, race, sex, residence (urban vs rural), marital status, percentage of county below federal poverty limit, percentage of county below ninth grade education, and estrogen receptor (ER)/progesterone receptor (PR)/human epidermal growth factor receptor type 2 (HER2) status. Additional statistically significant predictors for patients not receiving postlumpectomy radiation included African-American race (OR = 0.86), being unmarried (OR = 0.88), living in a county with a higher percentage of residents with less than a ninth grade education (OR = 0.60), and higher percent of county below the federal poverty level (OR = 0.57). Among uninsured patients, there was also a significant association with African-American race and residence in a county with a higher percentage of people below the federal poverty level. CONCLUSION: Patients with Medicaid coverage or without insurance are more likely to not receive radiation after lumpectomy. (P029) A Comparison of CT- and MRI-Defined Lumpectomy Cavity for Radiotherapy Planning of Breast Cancer Wei Huang, MD, PhD, Adam Currey, MD, Victor Chen, PhD, Frank Wilson, MD, Allen Li, PhD; Medical College of Wisconsin PURPOSE: To compare lumpectomy (LC) and planning target volume (PTV) that is delineated using magnetic resonance imaging (MRI) and computed tomography (CT) and examine the possibility of replacing CT with MRI for radiotherapy (RT) planning for breast cancer. MATERIALS AND METHODS: MRI and CT data were acquired for 14 patients with early-stage breast cancer who had undergone lumpectomy at their radiation treatment positions (prone) using a large-bore CT scanner (HighSpeed, GE) and a 3-T large-bore MRI scanner (Vero, Siemens) during RT simulation. All patients had dense breast tissues. The LCs were delineated manually on both CT (LC-CT) and MRI acquired with three sequences (T1, T2, and short TI inversion recovery [STIR]) (LC-T1, LC-T2, and LC-STIR, respectively) by a radiation oncologist and verified by another radiation oncologist. The PTV (PTV-MRI) was created by expand- American Radium Society Scientific Papers and Posters 2015 ing a uniform 15-mm margin from the union of the LC-T1, LC-T2, and LC-STIR and was compared with those from CTs (PTV-CT). Differences were measured in terms of cavity visualization score (CVS), volume, Dice coefficient (DC), and distance between centers of mass (COMs). RESULTS: The mean CVSs for T1-, T2-, STIR-, and CT-defined LCs were 3.36, 3.36, 3.79, and 2.50, respectively, implying that LC is mostly visible with the MR STIR sequence. For most cases (12/14), the LC volumes or PTVs from T1, T2, and STIR were smaller than those from CT. The mean reductions of LCs for T1, T2, and STIR from those for CT were 20%, 44%, and 36%, respectively. However, the differences between the volumes of PTV-MRI (the union of three sequences) and PTV-CT were smaller. The DCs between CT- and MRI (union of T1, T2, and STIR)-defined volumes were 0.60 ± 0.15 for LCs and 0.85 ± 0.08 for PTVs. The COM shifts from PTV-MRI to PTV-CT were 0.31 ± 0.25, 0.35 ± 0.39, and 0.36 ± 0.33 cm in the x-, y- and z-axis, respectively. The average PTV-MRI:PTV-CT volume ratio was 1.11±0.22, ranging from 1.00 to 1.25 for most cases (10/14). In 12 of 14 cases, MRIdefined LC included extra regions that would not be visible from CT. CONCLUSIONS: MRI substantially improves the visibility and accu- racy of lumpectomy cavity definition as compared with CT. Although the LC and PTV volumes that are delineated from an individual MRI sequence are generally smaller as compared with those from CT, the volumes, shapes, and locations for the PTVMRI, defined by the union of T1, T2, and STIR, were comparable with PTV-CT for most of the cases studied. It is feasible to use MRI to replace CT in RT simulations for breast-conserving RT. (P030) Metaplastic Breast Carcinoma at a Single Institution: Clinical-Pathologic Characteristics and Outcome Sharang Tenjarla, Sheema Chawla, MD, Jiqing Ye, MD, Brittany Heatherington, Lori Medeiros, MD, Peter Bushunow, MD; Rochester Regional Health System OBJECTIVES: Metaplastic breast carcinomas (MBCs) are rare primary breast malignancies, constituting < 1% of all breast cancers. They are characterized by differentiation of the neoplastic epithelium into squamous cells and/or mesenchymal-like elements. There is a dearth of information on clinically relevant pathologic features and clinical outcomes for these rare tumors. METHODS: A search of MBCs in the tumor registry and pathology archives of our center was performed. A review of the demographic and clinical-pathologic features and outcomes of all cases of MBC was performed between 2005 and 2014. Cases were reviewed retrospectively after institutional review board (IRB) approval. RESULTS: A total of 32 cases were retrieved. Median age was 62 years (range: 42–79 yr). The median tumor size was 3.5 cm (range: 0.8–6.7 cm), and 72% (n = 23) of patients had grade 3 tumors. Histology outcomes showed: squamous in 13 (41%), adenosquamous in 4 (12%), osseous and chondroid in 6 (19%), and spindle cell/sarcomatoid in 9 patients (28%); 28 patients (88%) were triple negative. Human epidermal growth factor receptor type 2 (HER2)/ neu overexpression was not seen in any of the patients. A total of 19 patients underwent mastectomy (59%), and the rest underwent breast conservation surgery. The median follow-up was 17.5 months (range: 3–108 mo). Stage distribution was: stage IA in 11 (34%), IIA in 12 (38%), IIB in 7 (22%), and IIIA in 2 (6%) patients. At a median follow-up of 22 months, five patients had recurrence: one with local and four with distant (two brain, one bone, and one with multiple sites), two of whom are alive with disease. Overall, 28 patients (87%) were alive at last follow-up. CONCLUSION: Similar to other reported series, the majority of our patients had triple-negative disease, and our patient population did not express the HER2/neu oncoprotein. Our predominant histology was squamous differentiation, whereas heterogeneity in histology is described in the literature. Despite high-grade disease, the outcomes in our study are favorable in comparison with previously reported series, although the follow-up is short. (P031) Effects of Oncoplastic Surgery on Delivery of Standard Adjuvant Radiotherapy Elisabeth Arrojo, MD, Alvaro Martinez, MD, FACR, Michael Ghilezan, MD, James Fontanesi, MD, Frank Vicini, MD, FACR; 21st Century Oncology; Botsford Hospital BACKGROUND: Oncoplastic breast surgery (OBS) is a surgical effort to remove additional breast tissue and improve cosmesis. However, local control with routine lumpectomy and cosmesis after adjuvant radiotherapy (A-RT) are both very good, questioning the need for OBS. We wanted to explore if A-RT practices have changed due to OBS. METHODS: A search in the PubMed and Ovid MEDLINE databases was carried out from 2010 to 2014, using the keywords “therapeutic mammoplasty” and “OBS.” Use of boost and tumor bed marking in OBS were analyzed and compared with the results of the 2014 survey (Thomas et al, Pract Radiat Oncol.) reported as “Radiation practice patterns among United States (US) radiation oncologists (ROs) for postmastectomy breast reconstruction and oncoplastic breast reduction (RT.OBS).” RESULTS: We found six studies, totaling 1,180 patients. Four studies did not reported clip-marking. Three of these four studies did not give a boost to any of the OBS patients, and the other one gave a boost only to OBS patients with +/close margins. The two studies that reported clips gave a boost to the patients. The overall analysis showed that 71% of patients did not receive a boost treatment. A total of 271 ROs from the US completed the RT.OBS survey, showing that 65.4% of them did not give a boost to any of the patients systematically, 8.3% indicated never having utilized a boost, and 38.7% only gave a boost to patients with clips marking the tumor bed. Only 33.1% of respondents indicated that they routinely collaborated with surgeons for clip placement at the time of breast reduction or complex tissue rearrangement. CONCLUSIONS: The standard patterns of care for breast conservation therapy include A-RT with a boost to the surgical bed. For the ARS PROCEEDINGS 2015 39 American Radium Society Scientific Papers and Posters 2015 PubMed and Medline articles, 71% of the 1,180 patients without clips and clear margins did not receive a boost. These results correspond with the ROs’ survey, which showed that 65.4% of them did not give a boost to the patients systematically. While OBS was perceived by the surgeons as a technique to remove more tissue and improve cosmesis, our review demonstrated that it negatively impacts radiotherapy techniques that are proven to achieve adequate local control. OBS is clearly changing patterns of delivery of adjuvant RT, without long-term outcomes supporting its safety. (P032) Current Management of Low-Grade Central Nervous System Glioma Ron R. Allison, MD, Dioval Reymond, Sharon Salenius, MPH, Andrej Hnatov, MD, Cynthia Ballenger, MD, Constantine Mantz, MD, Eduardo Fernandez, MD, PhD, Daniel Dosoretz, MD, Vershalee Shukla, MD, Timothy Shafman, MD, Steven Finkelstein, MD; 21st Century Oncology; The Brody School of Medicine, East Carolina University BACKGROUND: Low-grade central nervous system (CNS) glioma is a rare diagnosis but is becoming more common as neuroimaging is now often undertaken for headache and other nonspecific neurologic signs and symptoms. Current management may include resection alone, radiation alone, or their combination, with chemotherapy as an adjunct. We reviewed the current management for this diagnosis based on a large cohort of freestanding and hospitalbased cancer centers. MATERIALS AND METHODS: An institutional review board (IRB)approved chart review for all patients with low-grade primary CNS glioma who were treated at our facilities between 1989 and 2012 was undertaken. This search returned 25 patients (10 males, 15 females; mean age: 50.5 yr). Twenty patients were Caucasian. Presenting signs and symptoms were nonspecific CNS complaints: mainly generalized headache, vision changes, numbness, and weakness. Workup included magnetic resonance imaging (MRI) and/or computed tomography (CT). Biopsy reported low-grade glioma, though imaging alone was used for diagnosis in some patients. Only two patients underwent gross total resection, and one patient underwent surgical debulking. All patients underwent external beam radiation therapy (RT), usually intensity-modulated RT (IMRT) or three-dimensional (3D) treatment with a mean dose of 48.6 Gy. Image guidance was employed in 24% of patients. Temodar was delivered to six patients (24%). RESULTS: All patients have been followed for a mean of 9.3 months (range: 0.3–153.3 mo). Most patients were considered unresectable. RT was well tolerated, with most complications being grade I/ II. One patient developed seizures as a late complication. All failures were local, occurring in 28% of patients. With regard to survival, eight patients are currently alive (32%). Statistical analysis showed no survival advantage for gender, age, performance status, biopsy, IMRT, radiation dose, chemotherapy, or surgery. CONCLUSION: Currently, low-grade glioma is commonly treated in the community setting by RT alone. A relatively high rate of local failure is noted (28%), and long-term survival appears to be shorter than might be expected. A number of patients are treated based on imaging alone without a detriment to survival. Low-grade glio- 40 ma appears to be a more aggressive disease than usually considered, and efforts to improve the outcome would be served through clinical trials. (P033) MRI Resection Cavity Dynamics Following Brain Metastasis Resection and Permanent Iodine-125 Brachytherapy David R. Raleigh, MD, PhD, Zachary A. Seymour, MD, Bryan Tomlin, PhD, Michael W. McDermott, MD, Manish K. Aghi, MD, Philip V. Theodosopolous, MD, Mitchel S. Berger, MD, Penny K. Sneed, MD; University of California, San Francisco; California State University Channel Islands PURPOSE/OBJECTIVES: Surgical resection and permanent iodine-125 (125I) brachytherapy provide good local control for brain metastases. However, resection cavity remodeling has been postulated to alter treatment efficacy and toxicity. The purpose of this study was to investigate cavity volumetrics following surgery and 125I brachytherapy for brain metastases. MATERIALS AND METHODS: A total of 96 patients with 106 brain metastases, treated from September 1997 to July 2013, were retrospectively identified. The efficacy and toxicity of 125I brachytherapy in this cohort have previously been reported. In brief, the median age at surgery was 59 years, with a median overall survival of 12 months and overall local control of 92%. The overall risk of necrosis was 15% and trended lower without prior stereotactic radiosurgery (SRS) to the surgical site or with activity ≤ 0.73 mCi per source. All patients underwent magnetic resonance imaging (MRI) prior to surgery and were followed with surveillance MRIs beginning a median of 1 day after implantation (MRI1). Volumetric data were calculated from three-dimensional contours on T1-weighted postcontrast images by a single radiation oncologist (DRR). Cavities were censured from volumetric analyses at the time of tumor progression or necrosis. RESULTS: A total of 476 brain MRIs were analyzed (median 3 per patient; range: 0–22) with a median imaging follow-up of 4 months (range: 0 mo–13.6 yr). Median metastasis volume was 13.5 cm3 (range: 0.21–76.2 cm3), and median cavity volume on MRI1 was 5.2 cm3 (n = 101; range: 0.3–23.2 cm3). At a median of 1.7 (n = 32), 3.6 (n = 46), 5.9 (n = 38), 11.7 (n = 30), and 20.5 (n = 22) months after surgery, cavity volumes decreased by 25%, 35%, 42%, 47%, and 60% relative to MRI1, respectively. Metastasis size was the strongest predictor of cavity volume and shrinkage on both univariate and multivariate linear regression modeling (P < .0001). Factors that were associated with an increase in cavity volume included prior SRS to the surgical site, periventricular location, and lobar tip location (P < .05). The cavity-to-metastasis volume ratio decreased with increasing metastasis size, and the rate of cavity shrinkage was greater for metastases > 13.5 cm3 (P < .0001). Despite these findings, multinomial logistic regression modeling, with or without adjustment for source activity and prior SRS, failed to predict the likelihood of local failure or necrosis using either metastasis or cavity volume. CONCLUSIONS: Surgical resection with 125I brachytherapy is an effective strategy for local control of brain metastases. Although American Radium Society Scientific Papers and Posters 2015 metastasis volume significantly influences resection cavity size and remodeling, volumetric parameters do not appear to influence local control or necrosis. larly in the setting of bevacizumab-refractory tumors. (P034) Proton Therapy (PT) Large-Volume Re-Irradiation for Recurrent Glioma: Overall Survival (OS) and Toxicity Outcomes Yuan J. Rao, MD, Stephanie Perkins, MD, Jiayi Huang, MD; Washington University (P035) Prognostic Factors of Early Deaths in Patients With Craniopharyngioma From the SEER Registry Brijal M. Desai, MD, Russell C. Rockne, PhD, Irene B. Helenowski, MS, PhD, Jeffrey J. Raizer, MD, Nina Paleologos, MD, Ryan Merrell, MD, Sean Grimm, MD, Syed Azeem, MD, William F. Hartsell, MD, Patrick Sweeney, MD, Kristin R. Swanson, PhD, Vinai Gondi, MD; Northwestern University Feinberg School of Medicine; Rush University Medical Center; Northshore University Health Systems; Cadence Brain Tumor Center; CDH Proton Center PURPOSE: Craniopharyngioma is a locally invasive tumor that is challenging to resect, and the roles of aggressive surgery or adjuvant radiation therapy (RT) are unclear. The current study analyzed the Surveillance, Epidemiology, and End Results (SEER) registry to identify prognostic factors of early deaths after the diagnosis of craniopharyngioma, as well as to examine the impact of gross total resection (GTR) or adjuvant RT on overall survival (OS). BACKGROUND: The therapeutic benefit of targeting T2/fluid-attenuated inversion recovery (FLAIR) in addition to contrast-enhancing (CE) tumor during re-irradiation for recurrent glioma is attenuated by augmented toxicity. Given its steep dose falloff and narrow penumbrae, proton therapy (PT) minimizes the volume of brain parenchyma outside of the target volume, potentially permitting safer delivery of large-volume re-irradiation. METHODS: The SEER database was queried for craniopharyngioma patients from 2004 to 2011. Inclusion criteria included histologically confirmed craniopharyngioma with adequate information on age, surgical extent, and adjuvant RT. Patients who were aged < 3 years or > 69 years were excluded. Follow-up time was determined from the time of diagnosis. Cox proportional hazards models were used for univariate analysis (UVA) and multivariate analysis (MVA). OS rates were calculated using the Kaplan-Meier method and compared between groups using log-rank statistics. METHODS: From February 2011 to April 2014, a total of 21 consecutive adult patients with recurrent glioma who were treated with PT re-irradiation at a single institution were retrospectively analyzed. Planning target volume (PTV) included T2/FLAIR and CE abnormalities. The covariates that were assessed were age, gender, Karnofsky performance status (KPS) at time of PT, number of prePT recurrences, grade at initial diagnosis, interval between PT and prior radiotherapy (PRT), PT dose, PT PTV, bevacizumab failure, concurrent use of temozolomide and/or bevacizumab, and post-PT radiation necrosis. Overall survival (OS) time from PT start was estimated with Kaplan-Meier analysis; comparisons used the logrank statistic. Multivariate analysis used the Cox proportional hazards model. RESULTS: Median age was 43 years, and median KPS was 90. The median number of salvage treatments was 2 (range: 1–9). Median interval between PRT and PT was 32.8 months (range: 6.9–162.9 mo). A total of 13 patients (62%) were bevacizumab-refractory. Median PT dose was 50.51 Cobalt Gray Equivalent (CGE), and median PTV was 224.2 cc. Five patients (24%) remain alive. Median OS was 10.5 months overall, 6.3 months among bevacizumab-refractory patients, and 12.4 months among bevacizumabnaive patients. Prior bevacizumab failure (hazard ratio [HR] = 3.77; P = .02), lower KPS score (HR = 1.03; P = .047), and decreased interval between PRT and PT (≤ 40 mo; HR = 3.46; P = .04) were prognostic of inferior OS. On multivariate analysis, decreased interval between PRT and PT (P = .02) remained prognostic of inferior OS, but lower KPS and prior bevacizumab failure trended to significance (P = .07). One patient had grade 3 necrosis in the setting of PT re-irradiation for progressive brainstem glioma. One patient had grade 2 necrosis, and another had grade 2 stroke. No other grade ≥ 3 toxicities were observed. RESULTS: A total of 788 patients met the inclusion criteria, and the median follow-up was 38 months (range: 1–95 mo). Nine pediatric patients died within 5 years of diagnosis, with 78% attributed to craniopharyngioma. Further, 81 adult patients died within 5 years of diagnosis, and the two most common causes of death were craniopharyngioma (36%) and cardiac/diabetic causes (31%). On both UVA and MVA, older age (hazard ratio [HR] = 1.03; 95% confidence interval [CI], 1.02–1.04), black race (HR = 3.52; 95% CI, 2.29–5.40), adamantinomatous histology (HR = 1.84; 95% CI, 1.19–2.84), and papillary histology (HR = 2.07; 95% CI, 1.12–3.82) were significant predictors of worse OS. In contrast, gender, GTR, adjuvant RT, and tumor size were not significantly correlated with OS. Adult patients (aged ≥ 18 yr) had significantly worse 5-year OS than pediatric patients (80% vs 95%, respectively; P < .01). Patients of black race had significantly worse 5-year OS than patients of other races (67% vs 87%, respectively; P < .01). Adamantinomatous histology and papillary histology were associated with worse 5-year OS than classic craniopharyngioma histology (80% and 77% vs 88%, respectively; P = .05). Patients with GTR had a 5-year OS of 83%, while those with biopsy or subtotal resection had a 5-year OS of 86% (P = .26). Patients with adjuvant RT had a 5-year OS of 89%, while those who did not receive RT had a 5-year OS of 83% (P = 0.15). CONCLUSIONS: Craniopharyngioma-related death represents the most common cause of early deaths in both pediatric and adult patients. GTR and adjuvant RT do not appear to have a significant impact on OS within 5 years of diagnosis. Older age, black race, and adamantinomatous/papillary histologies are significant prognostic factors for early deaths after the diagnosis. CONCLUSION: Large-volume PT re-irradiation for recurrent glioma is safe and associated with promising OS outcomes, particu ARS PROCEEDINGS 2015 41 American Radium Society Scientific Papers and Posters 2015 (P036) Patterns of Care and Outcomes of Adjuvant Radiotherapy for Meningiomas: A Surveillance, Epidemiology, and End Results and Medicare-Linked Analysis Mark J. Amsbaugh, MD, Beatrice Ugiliweneza, PhD, MSPH, Eric Burton, MD, Maxwell Boakye, MD, MPH, MBA, Shiao Woo, MD; University of Louisville PURPOSE: To identify patterns of care and outcomes of adjuvant radiotherapy for meningiomas in the linked Surveillance, Epidemiology, and End Results (SEER) Medicare data. MATERIALS AND METHODS: Patients over 66 years of age, diagnosed with meningioma, and treated with a craniotomy in the SEER-Medicare data were included. Patients were grouped according to adjuvant treatment with fractionated radiotherapy (XRT), stereotactic radiosurgery (SRS), or none. Demographic, tumor, treatment, and outcome variables were collected. The MannWhitney U-test and chi-square test were used to analyze continuous and categorical variables. Time to event was analyzed with the Kaplan-Meier methods and log-rank test. Multivariate comparisons were conducted with logistic regression and proportional hazard models. RESULTS: A total of 1,964 patients were included for analysis (1,701 with no adjuvant treatment, 175 with XRT, and 88 with SRS). Patients were less likely to receive adjuvant therapy if they were older than 75 years (odds ratio [OR] = 0.730; 95% confidence interval [CI], 0.548–0.973), female (OR = 0.731; 95% CI, 0.547–0.978), or unmarried (OR = 0.692; 95% CI, 0.515–0.929). Patients were more likely to receive adjuvant treatment for grade II/III tumors (OR = 5.586; 95% CI, 2.135–13.589), tumors over 5 cm (OR = 1.850; 95% CI, 1.332–2.567) or partial resection (OR = 3.230; 95% CI, 2.327–4.484). For those receiving adjuvant therapy, SRS was less likely than XRT in patients diagnosed with grade II/III tumors (OR = 0.061; 95% CI, 0.006–0.655) or a 1-unit increase in Gagne comorbidity score (OR = 0.761; 95% CI, 0.599–0.968). Yearly between 2000 and 2010, 10.65% to 19.77% of patients received adjuvant therapy (7.143%–42.105% of those who received SRS). Although no survival benefit was seen with the addition of adjuvant therapy (P = .1236), the subgroup of patients receiving SRS had better survival compared with those receiving surgery alone (adjusted hazard ratio [aHR] = 0.544; 95% CI, 0.318–0.929). Furthermore, when stratifying patients by degree of resection, those who underwent partial or local resection and did not receive SRS had an increased risk of death compared with those who did (aHR = 1.934; 95% CI, 0.992–3.771). CONCLUSIONS: Utilization of adjuvant XRT and SRS remained stable between 2000 and 2010. Male sex, young age, marriage, partial resection, grade II/III tumors, and large tumors predicted for use of adjuvant therapy. For all patients, SRS improved survival compared with craniotomy alone. For patients with incomplete resection, SRS improved survival compared with craniotomy alone and adjuvant XRT. Randomized, prospective clinical trials are needed to better define the role of adjuvant XRT or SRS. 42 (P037) Analysis of Survival Outcomes in Patients With Multifocal Glioblastoma Omar H. Gayar, MD, Anant Gopal, PhD, Lisa Scarpace, MS, Steven Kalkanis, MD, Tom Mikkelson, MD, Farzan Siddiqui, MD, PhD; Department of Radiation Oncology, Hermelin Brain Tumor Center, Henry Ford Health System INTRODUCTION: Glioblastoma multiforme (GBM) is associated with extremely poor prognosis and survival. A small subset of these patients present with more than one focus of disease (multifocal or multicentric). We analyzed survival outcomes in patients with multifocal or multicentric GBM treated at our institution. METHODS: An institutional review board (IRB)-approved retrospective analysis was performed to study patients with GBM who were noted to have multifocal lesions at initial diagnosis and were treated and followed up at our institution between 2005 and 2014. We reviewed patient gender, age at diagnosis, tumor location(s), extent of surgery, pathologic details, and treatment delivered: radiation therapy (RT) ± chemotherapy (CT). Median survival (MS) was calculated for patients who had at least 6 months of follow-up after completion of RT. RESULTS: A total of 30 patients with multifocal GBM were treated and had adequate follow-up for analysis; 21 patients (70%) were male, and 9 (30%) were female. Median age at diagnosis was 58 years. Regarding surgery, only 6 patients (20%) had subtotal tumor resection (STR), while 23 (77%) had biopsy. No patients had gross total resection (GTR), and one patient did not have resection or biopsy and was treated with RT. The O(6)-methylguanine-DNA methyltransferase (MGMT) gene was found to be methylated in 9 (30%) and unmethylated in 11 (37%) patients. MGMT gene methylation status was unknown for the remaining 10 (33%) patients. A total of patients (90%) had adjuvant RT; 14 patients (52%) were treated to a dose of 60 Gray (Gy), and 5 (19%) received 40–45 Gy. Three patients did not complete RT due to enrollment in hospice or death. Two patients had stereotactic RT as part of their primary RT. A total of 25 patients (83%) had CT: 23 had concurrent CT with RT, and 13 had adjuvant CT. Also, 26 patients had at least 6 months of follow-up after RT completion. Their overall MS was 10.1 months. MS was 16.6 months for patients who had STR and 5.5 months for patients who had biopsy. CONCLUSIONS: Median survival of multifocal GBM is incredibly short, even compared with the already short median survival of single-lesion GBM. The majority of our patients had biopsy alone, likely due to the nature of multifocal GBM. This most likely contributes to a worse MS. (P038) Leukoencephalopathy Following Stereotactic Radiosurgery for Brain Metastases Daniel M. Trifiletti, MD, Cheng-Chia Lee, MD, David Schlesinger, PhD, Jason P. Sheehan, MD, PhD, James M. Larner, MD; University of Virginia; National Yang-Ming University PURPOSE/OBJECTIVE: The use of stereotactic radiosurgery (SRS) in the treatment of brain metastases has increased dramatically over the last decade in order to avoid the neurocognitive dysfunc- American Radium Society Scientific Papers and Posters 2015 tion that is induced by whole-brain radiotherapy (WBRT). Technical improvements in SRS delivery have greatly enhanced the SRS workflow and allowed for the treatment of numerous intracranial metastases in a single session. It is now common practice to treat up to 5–10 lesions with SRS in a single session, followed by subsequent SRS sessions based on surveillance imaging; however, the cumulative neurocognitive effect of numerous SRS sessions remains unknown. As leukoencephalopathy is a sensitive marker for radiation-induced central nervous system (CNS) damage, we investigated the dosimetric thresholds for SRS-induced leukoencephalopathy in patients treated with SRS alone, as well as SRS in combination with WBRT. MATERIALS AND METHODS: All patients treated at our institution with at least two sessions of SRS for brain metastases between 2007 and 2013 were reviewed. Pre- and post-SRS fluid-attenuated inversion recovery and T1- and T2-weighted MRI sequences were reviewed and graded for white matter changes associated with radiation leukoencephalopathy using a previously validated scale. Patient characteristics and SRS dosimetric parameters were reviewed for factors that contributed to radiographic leukoencephalopathy using Cox proportional hazards modeling. RESULTS: A total of 103 patients meeting the inclusion criteria were identified. The overall incidence of leukoencephalopathy was 53% at 3 years, and four factors predicted for radiation-induced leukoencephalopathy: the use of WBRT (P = .001), SRS integral cranial dose of > 3 J (P = .026), total number of tumors (P = .001), and total tumor volume (P = .009). The volumes receiving 8 Gy (V8), V10, V12, and V15 were not predictive. CONCLUSIONS: Our results establish that WBRT + SRS produces leukoencephalopathy at a much higher rate than SRS alone. Surprisingly, an SRS integral dose of over only 3 J predicts for leukoencephalopathy in patients treated with SRS alone. Our data define a dosimetric threshold at which radiation-induced leukoencephalopathy is likely to occur following SRS. As the survival of patients with CNS metastases increases and as the neurotoxicity of chemotherapeutic and targeted agents becomes established, the threshold of 3 J may influence the therapeutic management of patients with multiple brain metastases. (P039) Outcomes of Patients With Glioblastoma Receiving Concurrent Radiation Treatment and Antiepileptic Agents With Histone Deacetylase Inhibitor Activity Nicholas S. Boehling, MD, Erik P. Sulman, MD, PhD; UT MD Anderson Cancer Center BACKGROUND: Preclinical studies have demonstrated increased radiosensitivity of gliomas using histone deacetylase inhibitors (HDACis), such as valproic acid (VA). We hypothesized that patients with glioblastoma (GBM) who are treated with antiepileptics with HDACi properties during radiotherapy (RT) would exhibit improved outcomes. METHODS: A retrospective analysis was performed of patients treated with RT for GBM. Outcomes included overall survival (OS), progression-free survival (PFS), and RT response. Patients were noted for use of VA and other antiepileptic mediations, as well as temozolomide (TMZ) during RT. RESULTS: Data for 267 patients were available for analysis, with a median age of 57 years. Median OS and PFS for the cohort were 13.9 months and 4.1 months, respectively. A total of 19 (11%) patients received VA, and 42 (24%) patients received other antiepileptics with HDACi activity during the course of RT. Positive RT response was significantly associated with improved median OS (21.6 mo vs 10.1 mo; P < .01) and PFS (10.2 mo vs 2.7 mo; P < .01). No significant benefit in OS (14.9 mo vs 13.8 mo; P = NS), PFS (5.0 mo vs 4.1 mo; P = NS), or RT response (odds ratio [OR] = 1.19; P = NS) was seen for patients receiving VA. For patients receiving any HDACi, there was a significant improvement in median OS (15.6 mo vs 13.5 mo; P < .05) but no significant change in median PFS (5.2 mo vs 3.7 mo; P = NS). A total of 32 (55%) patients receiving HDACis showed a positive response to RT compared with 98 (47%) patients not receiving HDACi s(OR = 1.39; P = NS). When stratified by TMZ use, HDACi use did not result in any significant change in median OS, PFS, or RT response. TMZ use was associated with a significant improvement in median OS (19.7 mo vs 11.4 mo; P < .01) and RT response (OR = 1.91; P = .01). On multivariate analysis, a Cox proportional hazards model showed similar results as compared with univariate analysis. HDACi use resulted in a nonsignificant improvement in outcomes (HR = 0.92; P = .64). CONCLUSIONS: The use of antiepileptic HDACi drugs during RT for GBM did not significantly improve outcomes or RT response in this retrospective analysis. Further study of the concomitant use of HDACis and TMZ is necessary to elucidate clinically relevant results. (P040) Radiosurgery for Primary Central Nervous System Lymphoma Zachary A. Seymour, MD, Sarah Westcott, James L. Rubenstein, Penny K. Sneed, MD; Department of Radiation Oncology, Department of Medicine, University of California, San Francisco BACKGROUND AND PURPOSE: The role of stereotactic radiosurgery (SRS) in primary central nervous system lymphoma (PCNSL) is unknown. This represents a case series of five patients treated with SRS for PCNSL. METHODS: All patients who were treated with SRS for PCSNL were retrospectively reviewed. All clinical and treatment parameters were evaluated to assess treatment outcomes, symptomatic response, complications, and disease control. Near-complete response was defined as ≥ 75% volumetric reduction from the time of treatment. RESULTS: A total of five patients with PCNSL underwent SRS to seven lesions following 5–10 cycles of chemotherapy, which consisted of a regimen of high-dose methotrexate, rituximab, and temozolomide, with the last administration at least 14 days before SRS. Three patients were treated with SRS in lieu of whole-brain radiotherapy (WBRT), one patient was treated for salvage SRS after focal failure after WBRT, and one patient received no adjuvant radiation following chemotherARS PROCEEDINGS 2015 43 American Radium Society Scientific Papers and Posters 2015 apy and was treated only at the time of focal recurrence. The median age at the time of SRS was 60 years (range: 44–75 yr). The median imaging follow-up post-SRS was 16.8 months (range: 1.2–46.9 mo). The median dose was 15 Gy (range: 14–17 Gy), with a median prescription isodose line of 50% for a median target volume of 2.76 mL (range: 0.69–14.6 mL). All lesions had at least a near-complete response, with 82% response being the smallest volumetric reduction and four lesions obtaining a complete response. No patients experienced a local failure. Prior to SRS treatment, four patients had focal neurologic deficits, all of which improved with SRS. No neurotoxicity or adverse effects were observed. Whole-brain control was not achieved in any patient, with a median time to distant brain failure of 2.3 months (range: 0.7–21.6 mo) after first SRS treatment, and two patients had salvage SRS for a focal distant recurrence 3 and 33 months after initial SRS, respectively. CONCLUSIONS: SRS appears to be a reasonable treatment option for focally persistent or recurrent PCNSL in select patients, especially in the setting of focal neurologic deficits. As a radiosensitive entity, all PCNSL lesions had a substantial volumetric reduction with at least 14 Gy, and all patients with a focal neurologic deficit at treatment experienced symptomatic improvement with SRS. Further investigation should be completed regarding the benefits of SRS for focally appearing PCNSL as a potential way to avoid neurotoxicity and improve symptoms in selected patients. (P041) Clinical Outcomes of Gamma Knife Stereotactic Radiosurgery (GK-SRS) for Painful Trigeminal Neuropathy (TNP) David Zaenger, MD, M.N. Woodall, MD, Bryan M. Rabatic, PhD, MD, John R. Vender, MD, Waleed F. Mourad, MD, PhD, Joseph Kaminski, MD; Georgia Regents University RESULTS: With a median follow-up of 20 months (range: 3–70 mo), the overall response was 60%. Specifically, CR, near-CR, and PR rates were 27%, 20%, and 13%, respectively. No response (NR) was seen in 40% of patients, who had persistent, unchanged pain. For the 60% of patients who responded to GK-SRS, the mean time latency from SRS to response was 1 month (range: 0–2 mo). The pattern of pretreatment pain in the responders group was described as being dull rather than sharp compared with nonresponders. Additionally, when compared with the NR group, the CR group was significantly less likely to be exacerbated by daily activities pretreatment (0/4 [0%] vs 5/6 [83%]). No GK-SRS-induced grade ≥ 2 toxicities were reported. CONCLUSIONS: SRS is both safe and efficacious for PTN. Additionally, prior to treatment, PTN patients who characterize their facial pain as dull and not exacerbated by daily activities are more likely to receive therapeutic benefit, with a mean response time of 1 month. (P042) Toxicity and Treatment Outcomes in Single vs Multifractionated Radiotherapy for Acoustic Neuromas Ajaykumar B. Patel, MD, Jennifer L. Peterson, MD, Colleen S. Thomas, MS, Michael G. Heckman, MS, Stephen J. Ko, MD, Katherine S. Tzou, MD, Robert C. Miller, MD, Laura A. Vallow, MD, Steven J. Buskirk, MD; Department of Radiation Oncology, Division of Medical Statistics and Informatics, Mayo Clinic Florida PURPOSE: To review our institution’s experience in treating acoustic neuromas by comparing toxicity and treatment outcomes using stereotactic radiosurgery (SRS) vs fractionated stereotactic radiation therapy (FSRT). trigeminal neuralgia (TN) that necessitates prior trigeminal nerve injury, in addition to facial pain. While the evidence supporting the safety and efficacy of Gamma Knife stereotactic radiosurgery (GK-SRS) for medically refractory TN is well established, there is a limited body of literature concerning SRS for medically refractory PTN. We report our long-term clinical outcomes using GK-SRS for medically refractory PTN. MATERIALS AND METHODS: A total of 57 consecutive patients were treated with either single-fraction SRS (n = 26) or FSRT (n = 31) for acoustic neuroma at our institution between March 2000 and November 2013. Median dose was 1,200 cGy (range: 1,160– 1,600 cGy) for SRS and 2,000 cGy (range: 2,000–2,500 cGy) for FSRT. Data were collected on treatment toxicities and progression in 22 SRS and 29 FSRT patients with sufficient follow-up data. Toxicities were graded by the Common Terminology Criteria for Adverse Events version 4.0 (CTCAE v4.0). METHODS: This is a single-institution retrospective study of 320 patients treated with GK-SRS for TN between 2000 and 2013. From this cohort, 20 XRT-naive patients with PTN were identified. All patients failed initial treatment with medications, and seven (35%) received other invasive treatment (eg, microvascular decompression, rhizotomy) prior to SRS. All patients underwent frame-based single-fraction GK-SRS. Mean age was 61 years (range: 38–80 yr), and 75% of patients were female. Fifteen patients (80%) were Caucasian, four (15%) were African American, and one (5%) was Asian. Mean dose was 80 Gy (range: 80–90 Gy) to the 100% isodose line in a single fraction to the root entry zone of the involved trigeminal nerve (2–3 mm from the anterolateral surface of the pons). Complete response (CR), near-CR, and partial response (PR) were defined as being pain-free without medication, being pain-free with medication, and having reduced pain with medication, respectively. RESULTS: Median follow-up length was 39 months in SRS patients and 27 months in FSRT patients. Median maximum tumor dimension for SRS and FSRT was 1.4 cm and 1.6 cm, respectively (P = .097). The most common complication was grade 3 ipsilateral hearing loss, which was experienced in 23% of SRS patients and 35% of FSRT patients. Grade 2 vestibular disorder was experienced in 27% of SRS patients and 21% of FSRT patients. No patient experienced grade ≥ 3 vestibular disorder in either group. Grade 1 and 2 facial nerve toxicity occurred in 14% of SRS patients and 10% of FSRT patients. Grade 1 or 2 trigeminal nerve toxicity occurred in 23% of SRS patients and 10% of FSRT patients. No patient experienced grade ≥ 3 facial nerve or trigeminal nerve toxicity. Only three patients (one in FSRT, two in SRS) had disease progression. At 1, 3, and 5 years after the start of treatment, the cumulative incidence of progression was 0%, 8%, and 8% for SRS patients and 0%, 0%, and 11% for FSRT patients, respectively (P = .47). There was no statisti- PURPOSE: Painful trigeminal neuropathy (PTN) is a rare variant of 44 American Radium Society Scientific Papers and Posters 2015 cally significant difference in likelihood of any individual type of complication or disease progression between patients treated with SRS or FSRT (all P >.12). There was no evidence of an association with occurrence of different types of complications for maximum tumor dimension, SRS dose, or FSRT dose (all P ≥ .19). CONCLUSION: FSRT and SRS in treatment for acoustic neuromas had similar outcomes and toxicity at our institution. Both modalities appear to be successful at providing high tumor control with acceptable toxicities in the noninvasive treatment of acoustic neuromas. (P043) Central Neurocytoma: Impact of Resection Extent and Adjuvant Radiotherapy on Survival Outcomes Yi An, MD, Yu B. James, MD, Jiang Wen, MD, PhD, Henry S. Park, MD, MPH; Yale University School of Medicine; UT MD Anderson Cancer Center INTRODUCTION: Central neurocytoma (CNC) is an uncommon benign/borderline malignant tumor of the central nervous system that often causes symptoms due to extension into the ventricular system. Maximal surgical resection is the standard of care, but indications for adjuvant radiotherapy (RT) are unclear. Our goal was to analyze the impact of resection extent and adjuvant radiotherapy on survival outcomes using a population-based CNC dataset. METHODS: We identified all patients in the Surveillance, Epidemiology, and End Results (SEER) database diagnosed with CNC (ICD-O-3 histology code 9506-1) in 2004–2011. Only those who received gross total resection (GTR) and subtotal resection (STR) were included. Chi-square analysis assessed the associations between demographic/clinical characteristics and the utilization of GTR and RT. Overall survival (OS) outcomes were analyzed using Kaplan-Meier analysis and the log-rank test. RESULTS: We included 203 patients, with a median age of 31 years; 46% of patients were male, and 80% of patients were white. GTR was performed in 47% of patients, with the remainder receiving STR. Adjuvant RT was delivered to 15% of patients, including 9% after GTR and 20% after STR (P = .018). Age, sex, race, diagnosis year, World Health Organization (WHO) grade, and tumor size were not significantly associated with GTR or RT utilization. OS was 87% at 5 years for all patients, with a median follow-up of 39 months. There was no significant OS benefit with GTR compared with STR (89% vs 86%; P = .82). There was also no significant OS benefit with RT after either GTR (86% with RT vs 90% without RT at 5 years; P = .76) or STR (79% with RT vs 89% without RT at 5 years; P = .53). CONCLUSION: Adjuvant RT was delivered to a minority of CNC patients after either GTR or STR in this national database, though patients were more likely to receive RT after STR. Long-term OS was excellent for all subgroups, and there was no clear evidence of resection extent or adjuvant RT influencing survival outcomes. Since our database is subject to selection bias and limited by a lack of information regarding local recurrence, salvage therapies, exact extent of STR, and RT technique, further research is needed to validate our findings. (P044) Treatment Outcomes of WHO Grade III Malignant Meningioma With and Without Postoperative Radiation Therapy Hanako Yamauchi Farol, DO, Michael R. Girvigian, MD, Michael J. Miller, MD, Joseph C. Chen, MD, Javad Rahimian, PhD, Christine Chang-Halpenny, MD, Brandon N. Glousman, Najeeb Alshak, MD, Kenneth Lodin, MD; UCI; Kaiser Permanente; USC OBJECTIVE: Malignant meningioma is a rare disease, the optimal management of which is unclear. Our goal was to review our institution’s treatment and outcomes of World Health Organization (WHO) grade III malignant meningioma. METHODS: From January 2000 to December 2011, through a retrospective chart review, we identified 16 patients with a pathologic diagnosis of WHO grade III meningioma; 11 of these patients had presented with primary malignant meningioma, and 5 presented after transformation into a malignant meningioma from earliergrade disease. RESULTS: Median follow-up was 20.5 months (range: 0.4–140 mo). All patients underwent surgical resection with or without radiation therapy (RT). Doses given ranged from 5,040 to 6,000 cGy. Of the 11 patients with primary malignant meningioma, 6 had gross total resection (GTR), 4 had subtotal resection (STR), and 1 had unknown status. RT was given to 6 of the 11 patients. Median survival was 88.2 months with RT and 8.7 months without RT (P = .022). Median time to progression was 48.1 months with RT and 5.1 months without RT (P = .045). Of the five patients with transformed meningioma, two received GTR and three had STR. RT was given to four of the five patients. Median survival for these five patients was 16.1 months, with median time to progression of 8.3 months. For all patients, overall survival (OS) and progression-free survival (PFS) rates were 68.8% and 56.3% at 1 year and 39.4% and 21.4% at 5 years, respectively. CONCLUSIONS: Our study showed that patients with primary malignant meningioma had better outcomes after maximal resection followed by postoperative radiation. In contrast, transformed meningiomas demonstrated more aggressive behavior, with lower median survival despite RT. Further multi-institutional or randomized studies are required to evaluate the effectiveness of postoperative RT to determine the best approach to managing these tumors. (P045) Multimodality Therapy With Intensity-Modulated Radiotherapy for Locally Advanced Esophageal Cancer Nitesh N. Paryani, MD, Stephen J. Ko, MD, Corey Hobbs, MD, Kristin Kowalchik, MD, Elizabeth Johnson, MD, Laura Vallow, MD, Jennifer Peterson, MD, Katherine Tzou, MD, Steven J. Buskirk, MD; Mayo Clinic PURPOSE: The current standard of care for locally advanced esophageal cancer includes chemoradiotherapy with or without surgery. Radiation is usually delivered via a three-dimensional (3D) technique. Intensity-modulated radiation therapy (IMRT) has been utilized in the treatment of multiple tumors and has demonstrated similar efficacy while offering the possibility of decreased toxicity. ARS PROCEEDINGS 2015 45 American Radium Society Scientific Papers and Posters 2015 MATERIALS AND METHODS: A total of 36 patients were treated with IMRT and chemotherapy; 21 patients underwent surgical resection—11 underwent open surgery, and the remainder underwent minimally invasive surgery. Chemotherapy consisted primarily of 5-fluorouracil (5-FU) with oxaliplatin or cisplatin. All but two patients received 50.4 Gy; one patient received 41.4 Gy without surgery, and one patient discontinued treatment after 25.2 Gy. Eleven patients required a treatment break during radiotherapy. The median age was 69 years (range: 46–87 yr). Approximately two-thirds of tumors were adenocarcinomas located in the lower thorax. Two-thirds of patients were stage T3 and had positive lymph nodes. The median tumor size was 5 cm (range: 2–13 cm). RESULTS: With a median follow-up of 21.3 months for all patients (range: 2.4–44.8 mo) and 33.9 months for survivors (range: 3.744.8 mo), overall survival (OS) at 24 months was 55%. The 24-month OS was 75% vs 24% for surgical and nonsurgical patients, respectively. Seven patients had a complete pathologic response. A total of 24 patients experienced grade ≥ 3 acute toxicity, and there was one grade 5 toxicity. Acute toxicity was similar between surgery and nonsurgery patients. Also, 14 patients experienced grade ≥ 3 late toxicity (9 surgery and 5 nonsurgery patients). The most frequent late toxicity was grade 3 stricture (21%). On multivariate analysis, advanced age (relative risk [RR] for 10-year increase in age = 2.01; P = .032) and heart maximum dose > 55 Gy (RR = 3.73; P = .011) were associated with decreased survival. CONCLUSION: Patients who undergo surgery after chemoradiotherapy demonstrate improved survival; however, this may be related to underlying comorbidities that preclude surgery. IMRT appears to be a reasonable treatment option that may reduce complications from radiotherapy. Careful attention should be given to heart dose during treatment planning. (P046) Prediction of Pathologic Complete Response After Neoadjuvant Chemoradiation Therapy for Rectal Cancer Using Radiographic Texture Analysis Martin T. King, MD, PhD, William F. Sensakovic, PhD, Albert Koong, MD, PhD, Daniel T. Chang, MD; Stanford Cancer Institute; Florida Hospital PURPOSE: To determine if certain radiographic texture features from pre–radiation therapy (pre-RT) computed tomography (CT) scans can predict pathologic complete response (pCR) after neoadjuvant chemoradiation therapy for rectal cancer. METHODS: We conducted an institutional review board (IRB)approved retrospective analysis of 30 patients with rectal cancer who received neoadjuvant chemoradiation therapy, followed by total mesorectal excision, between 2009 and 2012. We collected relevant demographic, staging, and treatment-related information from the electronic medical records. We also downloaded treatment planning CT scans and structure sets onto a specialized computerized workstation. Gross tumor volumes (GTVs) of the primary rectal lesion were adjusted, if necessary, based on careful review of the pre-RT non–contrast-enhanced CT and positron emission tomography (PET) data. GTV voxels corresponding to air, defined as less than −50 Hounsfield units (HU), were excluded. Multiresolution texture analysis was performed by passing CT 46 scans through a three-dimensional spatial bandpass (Laplacian of Gaussian) filter using sigma values of 1.0 mm, 2.0 mm, and 3.0 mm in order to enhance textural features at fine, medium, and coarse scales, respectively. Mean texture feature values within the GTV for entropy, uniformity, kurtosis, skewness, and standard deviation were then calculated. The performance of each feature in predicting pCR was evaluated using receiver operating characteristic (ROC) analysis. Then, 95% confidence intervals (CIs) for area under the curve (AUC) values were estimated using a bootstrapping method with 2,000 iterations. Features with an AUC above a 0.75 threshold were identified as potential predictors of pCR. RESULTS: Treatment planning PET scans were available for 29 of 30 patients. Six patients achieved a pCR at the time of surgery. Kurtosis demonstrated AUC values above the 0.75 threshold for fine (0.78; 95% CI, 0.46–0.89), medium (0.77; 95% CI, 0.48–0.91), and coarse (0.74; 95% CI, 0.38–0.87) scales. Skewness also exhibited AUC values above the 0.75 threshold for fine (0.77; 95% CI, 0.49–0.89), medium (0.79; 95% CI, 0.51–0.90), and coarse (0.78; 95% CI, 0.51– 0.90) scales. Skewness on the medium scale allowed for a sensitivity of 66.3% at a specificity of 80.1% for predicting pCR. CONCLUSION: Several radiographic texture features from the preRT CT scan were identified as potential predictors of pCR in rectal cancer patients after neoadjuvant chemoradiation therapy. Future work will focus on validating these features in a larger dataset. (P047) The Role of Radiation Therapy Following Adjuvant Chemotherapy in Pancreatic Adenocarcinoma Craig J. Baden, MD, MPH, Andrew McDonald, MD, Rojymon Jacob, MD; University of Alabama, Birmingham OBJECTIVES: Early studies of adjuvant treatment for resected pancreatic adenocarcinoma demonstrated significant improvements in disease-related outcomes with the administration of radiation therapy. However, with the advent of gemcitabine and other newer chemotherapy regimens, the value and proper role of adjuvant radiation have come into question. In this study, we examine disease-related outcomes in patients treated with surgery and modern chemotherapy with or without subsequent radiation therapy in order to clarify the role of adjuvant radiation in this particular setting and to identify subsets of patients most likely to benefit from radiation. METHODS: We conducted an institutional retrospective review of all patients with pancreatic adenocarcinoma treated with curativeintent surgery and adjuvant chemotherapy between December 2001 and January 2013. Actuarial estimates of local control (LC), local failure–free survival (LFFS), overall survival (OS), and median survival (MS) were determined by the Kaplan-Meier method, with examination of the benefit of adjuvant radiation completed with the log-rank test. Patients were also stratified by node and margin status in order to determine the effect of radiation within these subsets. RESULTS: A total of 71 patients (median age: 64.6 years) with pancreatic adenocarcinoma underwent treatment with curative-intent surgery and adjuvant gemcitabine-based (n = 66) or (fluorouracil [5-FU], leucovorin, irinotecan, oxaliplatin) (FOLFIRINOX) (n = 5) American Radium Society Scientific Papers and Posters 2015 chemotherapy. Surgical margins were negative in 44 patients, close in 8 patients, and positive in 19 patients. Lymph nodes were pathologically involved in 46 cases. Following completion of adjuvant chemotherapy, 21 patients went on to receive radiation therapy, with nearly all receiving 5,040 cGy delivered in 28 fractions. Sensitizing fluoropyrimidine-based chemotherapy was administered with radiation in 20 of the 21 patients. Median follow-up for the cohort was 21.9 months. Patients receiving radiation (vs those with no radiation) did not have statistically significant improvements in 18-month LC (64.1% vs 53.7%; P = .42), LFFS (51.6% vs 51.4%; P = .63), OS (69.6% vs 80.6%; P = .49), or MS (28.5 mo vs 32.6 mo; P = .49). In the subset of patients with positive nodes and negative margins, radiation was associated with a trend toward improved local control (18-mo LC: 62.2% vs 34.2%; P = .08). CONCLUSIONS: We discerned no significant improvements with the addition of radiation therapy in these patients with resected pancreatic adenocarcinoma treated with modern adjuvant chemotherapy, but statistical power was limited. The eventual results of the ongoing Radiation Therapy Oncology Group (RTOG) study RTOG 0848 will provide definitive data regarding the appropriate role for radiation in the era of modern adjuvant chemotherapy. (P048) Mutational Analysis by Next-Generation Sequencing in Patients With Pancreatic Adenocarcinoma Andrea L. Arnett, MD, PhD, Kenneth Chang, BS, Terence T. Sio, MD, MS, Robert C. Miller, MD, MS; Department of Radiation Oncology, Mayo Clinic, Rochester; Department of Radiation Oncology, Mayo Clinic, Jacksonville BACKGROUND: Pancreatic adenocarcinoma carries a poor prognosis, and currently available systemic therapies have demonstrated only modest efficacy in advanced disease. In this retrospective study, next-generation exomic sequencing (NGS) was utilized in pancreatic adenocarcinoma samples to identify potential novel therapeutic targets that are not routinely assayed in the clinical setting. MATERIALS AND METHODS: Eight patients with confirmed pan- creatic adenocarcinoma were selected based on availability of tissues. These patients were treated at our institution from 2001 to 2013. A total of 236 somatic genes were surveyed in this review, including 3,230 exons and 47 introns at > 900x mapping coverage. NGS reports were generated from 2011 to 2013. Statistical analysis was performed using Kaplan-Meier survival analyses. RESULTS: The most frequent genomic alterations were found within KRAS (88%) and TP53 (75%). Three patients were found to have mutations within the SMAD family of genes. All patients with SMAD alterations were also found to have concurrent KRAS mutations, which is consistent with the reported literature. KRAS mutations most commonly involved codon 12, while the locations of SMAD family mutations were heterogeneous. In addition, concurrent mutations were found within genes that have been shown to potentially modulate or interact with KRAS-mediated signaling pathways, including CCND3, CDKN2A/B, and RB1. Furthermore, 75% of patients had multiple, novel mutations that have not tradi tionally been associated with pancreatic adenocarcinoma. The average number of mutations was 8.1 (range: 0–17 mutations). The majority of patients had greater than six mutations identified, but there was substantial heterogeneity in the location and type of genomic alterations. Median survival and 5-year overall survival (OS) were 30.1 months and 41%, respectively. There was no significant correlation between the number of mutations and OS. Median age at diagnosis was 54 years (range: 35–82 yr). Overall, 88% of patients were found to have mutations associated with targeted therapies. One-third of patients possessed multiple concurrent molecular targets for which US Food and Drug Administration (FDA)-approved chemotherapeutic agents are currently available. CONCLUSION: Novel mutations were identified in the majority of patients, including mutations within a number of genes that have the potential to influence KRAS-mediated signaling, as well as other prominent signaling pathways. These results could potentially serve to identify targets for novel chemotherapeutic agents and guide personalized, combinatorial therapy in appropriately selected patients. (P049) Increased Portal Venous Enhancement of Hepatocellular Carcinoma Following Transcatheter Arterial Chemoembolization Alexander Lam, MD, Megha Nayyar, BS, Chandana Lall, MD; UCI Medical Center Hepatocellular carcinoma (HCC) is a growing cause of mortality throughout the United States, despite advancements in diagnosis and treatment. As advanced HCC progresses, it becomes more dependent on arterial blood. This translates to the typical imaging characteristics of sharp contrast enhancement during arterial phase, followed by brisk portal venous washout on multiphasic computed tomography (CT). Liver-directed therapies, such as transarterial chemoembolization (TACE), have utilized this preferential arterial flow to administer targeted chemotherapy agents and embolize parasitized arterial branches. By compromising arterial flow, these embolic therapies alter the imaging characteristics that are typically dependent on hemodynamics, such as multiphasic CT. We hypothesize that the portal venous supply to the tumor experiences a sustained, compensatory increase following chemoembolization, resulting in increased enhancement on portal venous phase when compared with the pretreatment tumor bed. To evaluate this hypothesis, the images of 102 patients with advanced HCC were retrospectively reviewed by two fellowship-trained radiologists. Appropriate preprocedural and posttreatment CT images were performed at noncontrast, arterial, and portal venous phases. Digital subtraction angiographic images were reviewed to confirm the target lesion for evaluation and ensure adequate obliteration of the arterial vessels feeding the tumor. Degree of enhancement was quantified by drawing a circular region of interest (ROI) within the lesion margin of at least 1 cm2. Differences in attenuation of the tumor bed were standardized against normal liver parenchyma and underlying background on the same slice. Enhancement was defined as a difference greater than 15 Hounsfield units (HU) from baseline attenuation on the noncontrast study. Enhancement patterns among the phases were ARS PROCEEDINGS 2015 47 American Radium Society Scientific Papers and Posters 2015 compared between the pre- and posttreatment scans. A total of 50% (51 out of 102) of HCC cases following TACE demonstrated definite increased enhancement on venous phase compared with the preprocedural images, of which the majority consisted of increased peripheral enhancement. Arterial enhancement was decreased or unchanged for these patients. Also, 24% of cases (25 patients) showed unchanged enhancement on arterial phase images following TACE therapy, with no change in the appropriate venous washout. A total of 11% of cases (11 patients) showed unchanged enhancement on both arterial and portal venous phase after therapy, and 15% of cases (15 patients) demonstrated increased enhancement on arterial phase and not on venous phase. The majority of cases exhibited increased enhancement on portal venous phase following treatment with TACE, demonstrating the importance of surveillance with proper multiphase imaging to assess for recurrence, especially in light of a decrease in arterial phase enhancement following treatment. (P050) Yttrium-90 Radioembolization for Unresectable, Chemorefractory Colorectal Cancer Liver Metastases in KRAS Wild-type and Mutant Patients Einsley Janowski, MD, PhD, Olga Timofeeva, PhD, Sergey Chasovskikh, PhD, Max Goldberg, Alexander Kim, MD, Filip Banovac, MD, Dalong Pang, PhD, Anatoly Dritschilo, MD, Keith Unger, MD; Georgetown University INTRODUCTION: We report our institutional experience on the efficacy of resin-based yttrium-90 (90Y) radioembolization for the treatment of unresectable, chemorefractory colorectal cancer liver metastases (CRCLMs). METHODS: From 2011 to 2014, a total of 51 patients underwent 90Y treatment for CRCLMs at our institution. A retrospective review was conducted for clinical outcomes, demographic information, and tumor mutation status. In 38 patients, interval imaging was available for tumor response assessment using Response Evaluation Criteria in Solid Tumors (RECIST) criteria. Serum (plasma) was prospectively collected in patients both pretreatment (n = 9) and posttreatment (n = 7), and circulating cell-free DNA (ccfDNA) concentration and integrity index were measured using quantitative PCR. DNA integrity results were validated in two patients using atomic force microscopy (AFM). RESULTS: Our patient population consisted of 51 patients diagnosed with colon cancer at a median age of 56 years (range: 31–85 yr). Tumor mutation information was available for 41 patients: 24 (58%) patients were KRAS mutant, 15 (37%) were KRAS wild-type, and 2 (5%) had microsatellite instability (MSI). The average survival after 90Y was 5.7 months ± 4.1 (range: 0–21 mo) in the entire cohort, with a 12-month survival of 10%. Average survival stratified by mutation status in KRAS wild-type, KRAS mutant, and MSI patients was 6 months ± 3.5, 5.29 months ± 4.86, and 5 months ± 2.83, respectively. Imaging assessment showed a partial response in 8 patients (21%), stable disease in 16 (42%), and progressive disease in 14 (37%) at a median follow-up of 2 months after treatment. 48 Tumor local control after 90Y treatment averaged 2.17 months ± 2.97 for the entire cohort. Local control response, assessed according to tumor mutation, averaged 2.62 months ± 4.25 for KRAS wild-type, 1.16 months ± 1.43 for KRAS mutant, and 4.5 months ± 3.54 for MSI patients. ccfDNA was detected in 100% of the analyzed samples. Median pretreatment plasma ccfDNA levels of 4.6 ng/mL decreased to 1.8 ng/mL after single-lobe treatment. DNA integrity index was reduced from a median of 0.62 to 0.23 after treatment. Analysis by AFM of paired pre- and posttreatment samples of a KRAS-mutant patient and an MSI patient revealed minimal change in the KRAS sample but a 35% average fragment size drop in the MSI sample. CONCLUSIONS: 90Y radioembolization is an effective treatment for CRCLMs in extending local control for liver-dominant metastatic disease. However, KRAS-mutant tumors may be more radioresistant to treatment. (P051) Intraductal Papillary Neoplasm of the Bile Duct: Prognostic Factors in SEER Outcomes of Benign and Malignant Cases Sean Szeja, MD, Todd Swanson, MD, PhD; UT Medical Branch INTRODUCTION: Intraductal papillary neoplasm of the bile duct (IPNB) may be invasive or noninvasive. Given its rarity, there are limited data on the management and clinical outcomes. The purpose of this study is to use the Surveillance, Epidemiology, and End Results (SEER) database to evaluate prognostic factors: stage, anatomical location, extent of surgery, and the use of radiation therapy (RT). METHODS: Cases diagnosed from 1978–2011 were downloaded from the SEER database. Inclusion criteria were first primary, known status of surgery, and RT history. Analysis of malignant diagnoses from 2004–2011 incorporated TNM staging. KaplanMeier curves calculated overall survival (OS) and disease-specific survival (DSS) in months. Log-rank tests were performed to compare survival. RESULTS: There were 31 benign cases, with an OS of 92 months; surgery was used in 26 cases, definitive RT was used in 1 case, and adjuvant RT was used in another. Ampulla of Vater (AoV) and other extrahepatic ductal (EHD) locations had statistically similar OS (120 vs 79 mo, respectively; P = .93). For EHD, trends suggested that subtotal resection had the best OS (P = .157), and for AoV locations, radical resection trended toward worse OS (P = .128), with statistical power limited by having eight patients with defined surgical extent at each location. There were 1,309 malignant cases; 542 of these patients did not have surgery, and of this group 77 received RT alone that extended median OS from 3 to 7 months (P = .026) and DSS from 4 to 8 months (P = .074). There were 323 malignant cases diagnosed from 2004–2011: 54% with N0M0 and 20% being T1N0M0. Analysis of all stages combined by location showed, in decreasing order, significantly different median survival times (P < .01): AoV: OS 48 mo, DSS 57 mo; EHD: OS 12 mo, DSS 15 mo; and intrahepatic ductal (IHD): OS 5 mo, DSS 5 mo. Analysis by treatment modality showed that, with regard to OS and DSS, surgery alone was better than surgery and radiation (P < .01), American Radium Society Scientific Papers and Posters 2015 which was better than radiation alone (P < .01), which was similar to results with no treatment. Analysis of T1N0M0 cases showed that smaller extent of resection of primary location correlated with better OS and DSS (P < .05, P < .02, respectively) at EHD but not AoV locations. CONCLUSIONS: Both benign and malignant cases had outcomes dependent upon the location and extent of surgical resection. In malignant cases that are not amenable to surgery, radiation offers a survival benefit. Given the cohort in this analysis, selection bias likely plays a significant role. Further study is required to define the optimal management of IPNB. (P052) Human Papillomavirus in Esophageal Cancer: An Institutional Retrospective Analysis Sarah E. James, MD, PhD, Terence T. Sio, MD, Rondell P. Graham, MBBS, Michael Keeney, MD, David I. Smith, PhD, Robert C. Miller, MD; Department of Radiation Oncology, Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester; Department of Radiation Oncology, Mayo Clinic, Jacksonville OBJECTIVES: Human papillomavirus (HPV) is a well-established oncogenic and prognostic factor in a number of cancers. This retrospective study aims to estimate the prevalence of HPV in esophageal cancer and investigate the potential effect of infection on treatment response. MATERIALS AND METHODS: Esophagectomy tissue samples were obtained for 94 stage II/III esophageal cancer patients treated with trimodality therapy at our institution between 1998 and 2003. Patients with both adenocarcinoma and squamous cell histology were included. Most patients were treated with 50.4 Gy in 28 fractions with concurrent chemotherapy. Chemotherapy consisted of a platinum doublet of either cisplatin with 5-fluorouracil (5-FU) or carboplatin with paclitaxel. Subsequent to the neoadjuvant therapy course, patients underwent Ivor-Lewis esophagectomy. Freshfrozen paraffin-embedded specimens were evaluated using quantitative polymerase chain reaction (qPCR) to measure expression of the HPV oncogenes E6 and E7 and human p16. Patient characteristics were obtained from a chart review of electronic medical records. Hematoxylin and eosin (H&E)-stained slides of freshfrozen paraffin-embedded esophageal tumor specimens that were taken at the time of surgery were carefully reviewed and graded for treatment response. Chi-square testing was used for multivariate analysis. RESULTS: None of the 94 samples was positive for HPV oncogene expression. Median age at time of diagnosis was 61 years. The majority of patients was male (87%). Patients had advanced disease, with most having T3N0 (13%) or T3N1 (57%) disease. A total of 89% of patients had adenocarcinoma, and 11% had squamous cell carcinoma. At the time of esophagectomy, 28 (30%) patients were found to have a complete pathologic response (pCR), 40 (43%) had a partial response, and 26 (28%) had no significant treatment response. There was no significant correlation found between pCR and age, T stage, N stage, histology, tumor location, grade, or gender. However, patients aged > 70 years were more likely to have a pCR when compared with younger counterparts (P = .07). CONCLUSIONS: There was no HPV oncogene expression in our patient cohort, which corresponds with a low-to-no prevalence of esophageal HPV infection in a population of patients in the United States. However, further studies including a larger patient cohort with pretreatment tissue analysis would still be helpful in determining the true prevalence of HPV in esophageal cancer. Patients who are treated with trimodality therapy experienced a high rate of pathologic response. (P053) Esophageal Cancer Pathologic Complete Response Rate After Neoadjuvant Chemoradiation: Is There a Difference Between Academic Centers vs Community Centers? Wendy Gao, MD, Gurleen Dhami, MD, Brant K. Oelschlager, MD, Veena Shankaran, MD, Shilpen Patel, MD, Smith Apisarnthanarax, MD, Jing Zeng, MD; University of Washington INTRODUCTION: Trimodality treatment, consisting of preoperative chemoradiation followed by surgical resection, has been established as the standard of care for locally advanced esophageal cancer. Rates of pathologic complete response (pCR) range from 29% to 40% in phase III trials. We reviewed our institution’s pCR rate and assessed whether there is a difference for patients treated at our academic institution vs in the community. METHODS: Consecutive patients with esophageal cancer who underwent esophagectomies at our institutionafter chemoradiati on from January 2012 to October 2014 were included in this retrospective analysis. Patient characteristics, staging, histology, and pathologic response data were collected. Chi-square and t-tests were used to compare patient groups. RESULTS: A total of 51 patients were found to have undergone resection after chemoradiation for esophageal cancer between January 2012 to October 2014; 28 patients (55%) were treated at our academic center, and 23 (45%) were treated in the community. Patients treated in the community were older (median age: 65 vs 61 yr; P = .047). Staging distribution was similar for the two patient groups: community stage: II = 34%, IIIA = 43%, and IIIB = 23%; academic center stage: II = 39%, IIIA = 53%, and IIIB = 7%. Most patients had adenocarcinoma (88.2%) vs 9.8% squamous cell and 2% adenosquamous. Location of the tumors was distal esophagus in 47 patients (92.1%) and midesophagus in 4 (7.9%). Median radiation dose was 50.4 Gy (range: 37.8–50.4 Gy) for all patients. All patients treated at our center received carboplatin and paclitaxel (carbo/taxol) vs 10% of patients in the community receiving regimens other than carbo/taxol (one docetaxel, cisplatin, and 5-fluorouracil [5-FU] [DCF]; one 5-FU/oxaliplatin; and one alternating carbo/taxol with 5-FU/oxaliplatin). Pathologic complete tumor response occurred in 21.5% of patients. By histology, the pCR rate was 50% for squamous cell (2/4 patients), and 19% for adenocarcinoma (9/47 patients). For the primary tumor, there was downstaging in 41.2% of tumors, no change in 35.3%, and upstaging in 2%. Pathologic complete nodal response occurred in 41.2%, downstaging occurred in 3.9%, there was no change in 35.3%, and upstaging occurred in 19.6% of patients. There was no statistically significant difference in pCR rate between ARS PROCEEDINGS 2015 49 American Radium Society Scientific Papers and Posters 2015 patients who received neoadjuvant chemoradiation at University of Washington Medical Center (21%) vs at an outside institution (17%) (P = 1.0). CONCLUSIONS: Pathologic outcomes after neoadjuvant chemoradiation for esophageal cancer were similar between patients treated at an academic center and community setting, although patients treated in the community tended to be older than patients treated at our academic center. These results will need to be validated with a larger dataset. The pCR rate after neoadjuvant chemoradiation at our institution was 21%, consistent with published data. (P054) Carbon Ion Therapy for Chinese Patients With Prostate Cancer: Primary Reports Shen Fu, Qing Zhang, Xiaomeng Zhang, Xin Cai, Jin Meng, Jingfang Zhao, Yinxiangzi Sheng, Kambiz Shahnazi, Michael F. Moyer; Shanghai Proton and Heavy Ion Center/Fudan University Shanghai Cancer Center; Shanghai Proton and Heavy Ion Center; Shanghai Jiao Tong University Affiliated Sixth People’s Hospital PURPOSE: To evaluate the toxicities and efficacy of carbon ion therapy for patients with prostate cancer. MATERIALS AND METHODS: A total of 19 patients with pathologically confirmed prostate cancer were treated with carbon ion therapy from June 2014 to September 2014; 2 had low-risk, 10 had intermediate-risk, and 7 had high-risk prostate cancer. The patients with intermediate-/high-risk disease received neoadjuvant hormone therapy for 2–3 months, followed by concurrent hormone and carbon ion irradiation (63–66 GyE/23–24 Fx with one fraction per day, 5 days per week). The target area included the prostate and partial seminal vesicle, depending on the prognostic risk. Conedown strategy was adapted in order to avoid the toxicities of the bladder and rectum. Peripheral blood was withdrawn before, during, and after carbon ion therapy. RESULTS: Immediate outcome: Prostate-specific antigen (PSA) was used to evaluate the response to carbon ion therapy: 7 of 19 (37%) patients had biochemical control (PSA < 0.1 ng/mL) just after the conclusion of carbon ion therapy, and PSA values will be used to determine efficacy of the treatment 3 months after treatment. Also, 7 of 19 patients were diagnosed with abnormal 11 choline–positron emission tomography-computed tomography (PET-CT) before carbon ion therapy. Therefore, 11 choline–PET-CT will be conducted for those patients 3 months after treatment, and the data will be available at the time of the ARS meeting. A total of 7 of 19 patients had statistically higher apparent diffusion coefficient (ADC) values detected by diffusion-weighted imaging (DWI) before radiotherapy than after treatment (P = .0049). Circulating tumor cells (CTCs) in peripheral blood samples could be detected in 7 of 19 patients before carbon ion therapy, but there were no CTCs in 6 of these 7 patients after treatment. A total of 6 of 19 (32%) patients had variable acute toxicities per the Common Terminology Criteria for Adverse Events version 4.03 (CTCAE v4.03): hematologic system (5), genitourinary (GU) (4), and gastrointestinal (GI) (1). Most toxicities were grade 1; two patients had grade 2 hematologic toxicities. There were no signifi50 cant differences in specific immunity (T cells: CD3, CD4, CD8 for cellular immunity; B cells: CD19 for humoral immunity) or nonspecific immunity (NK cells: CD16, CD56) before and after carbon ion therapy. CONCLUSION: This is the first report about a Chinese population with prostate cancer treated with carbon ion therapy. Our primary data showed that carbon ion therapy was well tolerated, and the immediate effect was encouraging. Long-term follow-up is needed for the analysis of final treatment responses and toxicities. (P055) Can High-Grade Prostate Cancer (Gleason 8–10) Be Cured With Definitive Local Therapy Without Testosterone Suppression? Five-Year Outcomes Employing Up-Front Prostatectomy in Patients With Clinically Localized, Nonmetastatic Disease Darrion L. Mitchell, MD, PhD, Kyle Russo, MD, Mark C. Smith, MD, Sarah L. Mott, MS, John M. Watkins, MD; Department of Radiation Oncology, Holden Comprehensive Cancer Center, University of Iowa; Bismarck Cancer Center PURPOSE: High-grade prostate cancer (HGPC) is associated with an aggressive clinical course and poor outcomes; thus, a common approach involves the combination of long-term testosterone suppression with definitive local therapy. Small single-institution case series report that long-term disease control can be obtained in selected patients who undergo definitive local therapy alone; however, prognostic factors for this approach remain to be identified. The current investigation seeks to describe disease control and survival outcomes for patients with clinically localized HGPC at biopsy who were managed with primary radical prostatectomy (RP) without systemic therapy, with analyses performed to identify prognostic factors associated with disease control. MATERIALS AND METHODS: Patients were retrospectively identified for inclusion by biopsy-proven Gleason 8–10 adenocarcinoma managed with primary RP, without preoperative evidence of nodal or distant metastasis. Patient who received any preoperative intervention or adjuvant hormone therapy were excluded, as were patients with insufficient prostate-specific antigen (PSA) follow-up (< 12 mo). Patient-, tumor-, and treatment-related factors were analyzed for association with freedom from failure (FFF, defined as PSA > 0.2 ng/mL and rising or upon initiation of salvage therapy), employing Cox proportional hazards regression. The Kaplan-Meier method was employed for estimation of FFF and survival. RESULTS: From 2003–2010, a total of 69 eligible patients were identified. Median age was 63 years (range: 48–75 yr) and median PSA was 11.7 ng/mL (range: 3.5–64.9 ng/mL). Gleason score (GS) at RP was < 7, 8, and > 9 for 22, 17, and 29 patients, respectively. Extraprostatic extension, involved surgical margin, seminal vesicle invasion, and lymph node involvement were identified in 32, 33, 18, and 6 patients, respectively, with adjuvant radiotherapy delivered to 5 patients. At a median follow-up of 67.3 months (range: 13.3–141.2 mo), 40 patients had disease recurrence, and 8 patients died (6 cancer-specific). The 5-year FFF and overall survival (OS) rates were 39% (95% confidence interval [CI], 21%–58%) and 87% (95% CI, 72%–94%), respectively. Primary and overall Gleason American Radium Society Scientific Papers and Posters 2015 score at RP, involved surgical margin, seminal vesicle involvement, nodal involvement, and elevated initial postprostatectomy PSA were significantly associated with FFF in the univariate analysis, with primary GS at RP (hazard ratio [HR] = 1.80; P < .01) and post-RP PSA (HR = 4.64; P < .01) significant in the multivariate analysis. CONCLUSIONS: Patients with HGPC at diagnosis have high rates of early disease recurrence, though mortality at 5 years remains low. Following RP without systemic therapy, high primary GS and initial post-RP PSA were independently associated with worse FFF outcomes. (P056) Ten-Year Outcomes of Patients With Gleason Score 9/10 Prostate Cancer Treated With Trimodality Therapy Andrew T. Wong, MD, A. Sivathayalan, R. Salant, MD, L. Harrison, MD, R. Stewart, MD, T. Carpenter, MD, D. Shasha, MD; Mount Sinai Beth Israel; Carleton University OBJECTIVE: Multiple publications have reported poor biochemical control in patients with localized Gleason score (GS) 9/10 prostate cancer treated with either surgery or external beam radiation therapy (EBRT). Few studies have specifically addressed brachytherapy outcomes in this cohort or reported outcomes beyond 5 years. The purpose of this study is to report long-term clinical outcomes in patients with GS 9/10 prostate cancer treated with combination EBRT, low-dose-rate brachytherapy (LDRBT), and androgen deprivation therapy (ADT). MATERIALS AND METHODS: We retrospectively reviewed 73 patients with localized GS 9/10 prostate cancer treated with combination therapy from 1998 to 2012 at a single institution; 27 patients had one high-risk feature (stage T3a/b or PSA ≥ 20 ng/ mL), 33 had two features, and 13 had three features. Patients received pelvic EBRT to 50.4 Gy followed by an LDRBT boost (108 Gy 125I, 90 Gy 103Pd). All implants and contours were performed by a single physician (DS). Luteinizing hormone releasing hormone (LHRH) agonist was given for a median duration of 25 months. Biochemical failure was defined using the Phoenix criteria (PSA nadir + 2). Biochemical progression–free survival (BPFS), prostate cancer–specific survival (PCSS), and overall survival (OS) were calculated using Cox regression analysis. Univariate and multivariate Cox regression was performed to identify factors that may impact BPFS: 1 high-risk feature vs > 1 feature (P = .0358), stage < T3c vs T3c (P = .240), prostate-specific antigen (PSA) < 50 ng/mL vs ≥ 50 ng/mL (P = .753), and isotope 125I vs 103Pd (P = .471). Toxicity was graded using the Common Toxicity Criteria for Adverse Effects version 3.0 (CTCAE v3.0). RESULTS: Minimum and median follow-ups were 20 and 55 months, respectively. Five-year actuarial BPFS, PCSS, and OS were 76.4%, 95.1%, and 88.2%, respectively. Ten-year BPFS, PCSS, and OS were 63.4%, 80.5%, and 59.1%. Median time to biochemical failure was 51 months. Multivariate Cox regression analysis showed that a higher number of high-risk features was significant (P = .057), but T3c (P = .378) and PSA ≥ 50 ng/mL (P = .600) were not significant predictors of BPFS. No patients developed acute urinary retention requiring urinary catheterization. One patient developed grade 3 toxicity, and there was no other cases of grade 3/4 genitourinary or gastrointestinal toxicity. CONCLUSIONS: In this large, retrospective series, we report very good long-term oncologic outcomes and minimal toxicity in patients with GS 9/10 prostate cancer treated with combination EBRT, LDRBT, and ADT. Trimodality therapy is a well-tolerated and effective treatment for these very-high-risk patients. (P057) A Population-Based Study of Men With Low-Volume, Low-Risk Prostate Cancer: Does African-American Race Predict for More Aggressive Disease? Arpit Chhabra, MD, Justin Rineer, MD, Jeremy Weedon, PhD, David Schreiber, MD; State University of New York Downstate Medical Center; University of Florida Orlando Health; Department of Veterans Affairs, New York Harbor Healthcare System INTRODUCTION: Recent studies have suggested that AfricanAmerican (AA) patients with clinical low-risk prostate cancer may not be ideal candidates for active surveillance due to a higher likelihood of harboring aggressive disease. In this study, we utilized the Surveillance, Epidemiology, and End Results (SEER) database to analyze whether race plays a role in pathological upstaging after radical prostatectomy (RP). MATERIALS AND METHODS: This study consisted of patients in the SEER database for whom race was identified either as white or AA. Eligible men were diagnosed between 2010 and 2011 with prostate-specific antigen (PSA) < 10 ng/mL as well as cT1cN0M0, Gleason score 6 disease with adenocarcinoma in 2 or fewer cores of a ≥ 12-core biopsy, treated by RP. Detailed pathologic information regarding pathologic T stage, margin status, and primary and secondary pathologic Gleason scores were collected. Adverse pathology was characterized individually, as well as in a composite metric, which was defined as pT2 and Gleason ≥ 4 + 3; pT3a and Gleason 3 + 3 with positive margins; pT3a and Gleason ≥ 3 + 4; or pT3b–pT4 with any Gleason score. We also analyzed the Cancer of the Prostate Risk Assessment score (CAPRA-S) for each patient. Chi-square was used to characterize differences in pathological extent of disease by race. Univariate and multivariate logistic regression was used to look for potential predictors for adverse pathology. Statistical significance was defined as a P value of < .05. RESULTS: There were a total of 1,794 patients who met the selection criteria, of whom 1,565 (87.2%) were identified as white and 229 (12.8%) were identified as AA. No statistical difference was observed between white and AA men with regard to pathological Gleason score (P = .99), pathological extent of disease (P = .34), margin status (P = .43), CAPRA-S score (P = .56), or adverse features (P = .45). On multivariate analysis, only increasing PSA and increasing age were significant predictors of adverse pathology. AA race was not predictive for adverse pathology on univariate (odds ratio [OR] = 1.19; 95% confidence interval [CI], 0.75–1.88; P = .45) or multivariate (OR = 1.43; 95% CI, 0.87–2.33; P = .16) analysis. ARS PROCEEDINGS 2015 51 American Radium Society Scientific Papers and Posters 2015 CONCLUSION: In this large population-based cohort of 1,794 men with low-risk, low-volume prostate cancer, AA race was not associated with more aggressive pathology compared with Caucasians. (P058) The Long-Term Economic Value of Hypofractionated Prostate Radiation: A Cost Minimization Analysis of a Randomized Trial Khinh Ranh Voong, MD, MPH, Lincy S. Lal, PhD, Deborah A. Kuban, MD, Thomas J. Pugh, MD, J. Michael Swint, PhD, Robert R. Trujillo, Joy Godby, Seungtaek Choi, MD, Andrew K. Lee, MD, MPH, Pamela J. Schlembach, MD, Usama Mahmood, MD, Steven J. Frank, MD, Sean E. McGuire, MD, PhD, Karen E. Hoffman, MD, MPH; UT MD Anderson Cancer Center; UT School of Public Health PURPOSE: Hypofractionated prostate radiation (HIMRT) shortens the treatment course while providing outcomes comparable with conventionally fractionated radiation (CIMRT). To determine the long-term economic value of HIMRT, including the costs of managing radiation toxicities that develop after treatment, a cost minimization analysis compared CIMRT with dose-escalated HIMRT using data from a randomized trial. PATIENTS AND METHODS: Men with localized prostate cancer were randomized to CIMRT (75.6 Gy in 42 fractions over 8.4 weeks) or HIMRT (72 Gy in 30 fractions over 6 wk). A decision tree modeled trial probabilities of maximum late bowel and urinary toxicities using patient-level data with a median follow-up of 6 years. Costs were estimated from the healthcare perspective, using 2014 national reimbursement rates for services received. Patientlevel institutional costs, adjusted to 2014 dollars, verified reimbursements. Sensitivity analysis assessed model uncertainty. RESULTS: The cost for HIMRT and toxicity management was $22,957, saving $7,000 compared with CIMRT ($30,241). CIMRT was the common factor among the five most influential scenarios that contributed to total costs. Toxicity represented a small part (< 10%) of the average total cost for patients with either grade 2/3 bowel or urinary toxicity. However, toxicity management reached up to 26% of total costs for patients with both high-grade bowel and urinary toxicities. There was no threshold at which CIMRT became the less costly regimen. Institutional costs confirmed the economic value of HIMRT ($6,000 savings). CONCLUSION: HIMRT is more cost-efficient than CIMRT for treating prostate cancer, taking into account additional costs owing to late radiation toxicities. (P059) Trends in the Utilization of Radiotherapy in the Management of Renal Cell Carcinoma Talha Shaikh, MD, Elizabeth A. Handorf, PhD, Colin Murphy, MD, Alexander Kutikov, MD, Robert G. Uzzo, MD, Mark Hallman, MD, PhD, Eric M. Horwitz, Marc C. Smaldone, MD, MSHP; Fox Chase Cancer Center PURPOSE: The role of radiotherapy in the management of renal cell carcinoma (RCC) has been limited due to the belief that these tumors are relatively radioresistant. More recently, retrospective 52 series have demonstrated good local control in patients undergoing radiation. We examined the temporal trends and patterns of use of radiotherapy in patients with localized RCC. MATERIALS AND METHODS: Patients diagnosed with RCC were identified using the National Cancer Data Base. Our primary objective was to describe the temporal trends in the utilization of radiotherapy. Our second objective was to identify patient and treatment factors associated with receipt of radiation. Data were analyzed using the chi-square and Cochran-Armitage tests for trend. RESULTS: A total of 330,426 were diagnosed with RCC between 1998 and 2010, with 18,522 (5.6%) patients receiving radiotherapy. After excluding patients with metastatic disease, 280,208 patients were diagnosed with localized RCC, with 3,552 (1.3%) patients receiving radiation therapy to the primary site. Factors associated with receipt of radiation included age > 71 years, no surgery (P < .0001), chemotherapy use (P < .0001), higher stage (P < .0001), higher grade (P < .0001), positive nodes (P < .0001), and sarcomatoid histology (P < .0001). A total of 257,304 patients underwent surgical resection in this cohort. Of these patients, 2,265 (0.9%) received adjuvant radiation. Patients receiving adjuvant radiotherapy were more likely to receive chemotherapy (P < .0001), have a higher stage (P < .0001), have a higher grade (P < .0001), have positive nodes (P < .0001), and have sarcomatoid histology (P < .0001). CONCLUSIONS: There has been a decrease in the use of radiotherapy for patients with localized RCC, although patients with more aggressive disease were more likely to receive radiation. Modern trials are needed to better identify the role of radiation in the management of these patients. (P060) Impact of Major Psychiatric Disorders on Tolerance and Outcomes for Men With Prostate Cancer Undergoing Dose-Escalated Radiation Therapy Joseph Safdieh, MD, J. Rineer, MD, A. Wong, MD, D. Schwartz, MD, D. Schreiber; SUNY Downstate Medical Center; University of Florida Health Cancer Center at Orlando; Department of Veteran Affairs, NY Harbor Campus PURPOSE/OBJECTIVES: Prior studies have revealed that patients with psychiatric disorders and prostate cancer have worse survival outcomes compared with the general population. However, many of these studies have associated these outcomes with delayed diagnosis and/or reduced access to definitive treatment. The purpose of this study was to investigate the toxicity and outcomes of patients with or without psychiatric disorders who were a primarily prostate-specific antigen (PSA)-screened population and who accepted definitive treatment with external beam radiation. MATERIALS AND METHODS: The charts of 469 patients diagnosed with prostate cancer from 2003–2010 who were treated with external beam radiation (minimum dose 7,560 cGy) were reviewed. Patients were identified as having no psychiatric disorder (−Psy) or having a psychiatric disorder (+Psy) consisting of a Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV diagnosis of American Radium Society Scientific Papers and Posters 2015 posttraumatic stress disorder, depression, schizophrenia, bipolar disorder, or generalized anxiety disorder. Kaplan-Meier analysis was used to analyze biochemical control, distant control, prostate cancer–specific survival, and overall survival (OS). Multivariate Cox regression was used to look for covariates associated with toxicity or biochemical control. RESULTS: The charts of 469 patients were reviewed, and 100 patients (21.3%) were found to be +Psy. At a median follow-up of 73 months, there was no difference between the two groups regarding 6-year biochemical failure–free survival (79.8% −Psy vs 80.4% +Psy; P = .50) or 6-year distant metastatic-free survival (96.4% − Psy vs 98.0% +Psy; P = .36). There were also no differences regarding 6-year prostate cancer–specific survival (98.4% −Psy vs 99.0% +Psy; P = .45) or 6-year OS (80.2% −Psy vs 82.2% +Psy; P = .35). Likewise, short- and long-term genitourinary and gastrointestinal toxicities were similar between the groups. On multivariate analyses, the presence of +Psy was not a significant predictor for toxicity or biochemical recurrence. CONCLUSIONS: This study did not find any significant differences in treatment tolerance or any outcome endpoints between men with and without psychiatric disorders undergoing definitive radiation treatment for prostate cancer. This finding suggests that early diagnosis and reduced barriers to definitive treatment will alleviate the decreased cancer-specific mortality in this patient population. (P061) Anticoagulation Use and Improved Outcomes in a Predominantly African-American Population With High-Risk Prostate Cancer Elliot B. Navo, MD, Shan-Chin Chen, Justin Rineer, MD, David Schreiber, MD; Department of Veterans Affairs, NY Harbor Healthcare System, SUNY Downstate Medical Center; MD Anderson Cancer Center Orlando INTRODUCTION: Several studies have reported improved biochemical and prostate cancer–specific mortality (PSM) outcomes, particularly for men with high-risk (HR) disease, associated with the use of anticoagulation (AC), in addition to radiation therapy. Numerous other studies have suggested that prostate cancer in African-American (AA) men tends to be more aggressive than in Caucasians and/or other races. Therefore, we analyzed our cohort of predominantly AA men with HR disease to determine whether anticoagulation use is associated with any improvements in biochemical or clinical outcomes. MATERIALS AND METHODS: There were 469 consecutive men treated at the New York Harbor Veterans Hospital with dose-escalated radiation therapy (minimum dose 7,560 cGy) for nonmetastatic prostate cancer between 2003 and 2010. Of these patients, 143 had HR disease and were included in this study. Men were categorized by use of AC at the time of consultation and/or follow-up examinations. Chi-square or Fisher’s exact test was used to compare patient characteristics. Kaplan-Meier curves were generated to compare biochemical-free survival (bFS), distant metastasis–free survival (DMFS), and PSM; outcomes were compared using the log-rank test. Multivariate Cox regression was also performed to identify covariates associated with increased risk for all clinical endpoints. RESULTS: The median follow-up was 65 months. A total of 55.9% of patients were identified as AA. There were no significant differences in patient characteristics between +AC and −AC. There was significantly improved bFS for AC patients at 5 years (84.6% vs 65.1%; P = .048). The 5-year DMFS was 96.4% for +AC vs 92% for −AC (P = .069). The 5-year PSM was 96.1% for +AC vs 95.9% for −AC (P = .21). On multivariate analysis for biochemical control, only use of AC was associated with improved bFS (hazard ratio [HR] = 0.47; 95% confidence interval [CI], 0.23–0.95; P = .036). AA race was not a significant predictor for biochemical failure (HR = 1.50; 95% CI, 0.50–4.50; P = .47). Similarly, in regard to multivariate analysis for distant metastases, AC was associated with improved DMFS (HR = 0.24; 95% CI, 0.06–0.88; P = .03). However, in regard to PSM, there was no benefit associated with the use of AC (HR = 0.26; 95% CI, 0.05–1.32; P = .10). CONCLUSIONS: In our population of predominantly minority men with HR disease, the use of anticoagulation is associated with improved bFS and DMFS on multivariate analysis. However, AA race was not associated with any differences in outcome. (P062) Comparison of Stereotactic Body Radiotherapy (SBRT) and Conventional External Beam Radiotherapy (EBRT) in Renal Cell Carcinoma (RCC) Christopher L. Tinkle, MD, PhD, Stephen L. Shiao, MD, PhD, Vivian K. Weinberg, PhD, Amy M. Lin, MD, Alexander R. Gottschalk, MD, PhD; University of California, San Francisco; Cedars-Sinai Medical Center BACKGROUND: Renal cell carcinoma (RCC) is considered a radiation-resistant histology, often with poor response to conventionally fractionated external beam radiotherapy (EBRT). We compared outcomes for patients treated with EBRT vs stereotactic body radiotherapy (SBRT) for RCC. METHODS: From 2004 and 2012, a total of 89 patients were treated with either EBRT or SBRT and retrospectively reviewed. Patients with locally recurrent RCC, bone or soft tissue RCC metastases, or primary RCC in a solitary kidney were included; 51 patients received EBRT, while 38 patients received SBRT. The median biologically effective dose (BED), assuming an α/β ratio of 10, was 32.6 Gy10 for the EBRT group and 48.0 Gy10 for the SBRT group. Local failure (LC) was defined pathologically or by imaging according to Response Evaluation Criteria in Solid Tumors (RECIST) 1.1, and toxicity was reported according to Common Terminology Criteria for Adverse Events version 4.0 (CTCAE v4.0) guidelines. Univariable and multivariable analyses using Cox’s regression model was performed to determine predictors of local control. RESULTS: Median follow-up from radiotherapy was 9.8 months (range: < 1–73 mo) with EBRT and 19.7 months (range: < 1–61 mo) with SBRT (P = .26). EBRT patients were younger (P = .02), and more of them were M1 (P = .04); yet, other baseline features did not differ significantly. Total RT dose, dose/fraction, and BED10 were significantly higher in the SBRT group (P ≤ .002 for each), while number of fractions was significantly lower (P < .001). The 1-year LC estimate was 88% (95% confidence interval [CI], 72%–96%) with SBRT and 50% (95% CI, 32%–65%) with EBRT (P ARS PROCEEDINGS 2015 53 American Radium Society Scientific Papers and Posters 2015 = .001), with no significant difference in rate of distant recurrences (P = .37). The 1-year progression-free survival (PFS) and overall survival (OS) rates between the EBRT and SBRT groups were 17% (95% CI, 8%–29%) vs 39% (95% CI, 24%–54%) (P = .06) and 39% (95% CI, 25%–52%) vs 82% (95% CI, 65%–91%) (P = .002), respectively. The use of SBRT was the most important independent factor that was significantly predictive of local control on multivariable analysis (P = .001, log-likelihood ratio [LLR] test; hazard ratio [HR] = 0.29; 95% CI, 0.13–0.61), while neither age nor metastasis at diagnosis was predictive. No grade 3/4 toxicity was observed in either RT group. CONCLUSIONS: The data support that SBRT improves local control over standard fractionation schemes. Higher dose per fraction, with a BED in the range of 48 Gy10, is a safe and effective local treatment modality for RCC. (P063) Can Some High-Risk Prostate Cancer Patients Be Treated With a Shorter Course of Androgen Deprivation Therapy (ADT)? Seungtaek Choi, MD, Quynh-Nhu Nguyen, MD, Sean E. McGuire, MD, PhD, Karen E. Hoffman, MD, MPH, Thomas J. Pugh, MD, Steven J. Frank, MD, Usama Mahmood, MD, Deborah A. Kuban, MD, Andrew K. Lee, MD, MPH; UT MD Anderson Cancer Center PURPOSE: To determine the efficacy of a shorter course of androgen deprivation therapy (ADT) in patients with high-risk prostate cancer. MATERIALS AND METHODS: Charts of 472 high-risk prostate cancer (stage ≥ T3a, Gleason ≥ 8, or prostate-specific antigen [PSA] > 20.0 ng/mL) treated between 2000 and 2012 were reviewed. Eighty patients who were treated with a shorter course of ADT (≤ 12 mo of a gonadotropin-releasing hormone [GnRH] agonist) were analyzed for this study. All patients were treated with image-guided radiation therapy with either intensity-modulated radiation therapy or proton therapy. The median total dose given was 76 Gy in 2-Gy fractions (range: 74–86.4 Gy in 1.8–2.0-Gy fractions). Patients were followed with PSA measurements taken at 3–6-month intervals after the radiation therapy. Biochemical failure was defined as nadir + 2 as per the Phoenix definition. Biochemical relapse–free survival and prostate cancer–specific survival were calculated using the Kaplan-Meier method. RESULTS: Median age was 73 years (range: 52–89 yr). The ADT duration was ≤ 6 months in 18 patients (22.5%) and 9–12 months in 62 patients (77.5%). The distribution of T stage was as follows: 1 patient with TX (1.3%), 31 with T1c (38.8%), 23 with T2a (27.5%), 10 with T2b (12.5%), 5 with T2c (6.2%), 6 with T3a (7.5%), and 5 with T3b (6.2%). The distribution of Gleason scores was as follows: 6 patients (7.5%) with Gleason 7 (3 + 4), 9 (11.2%) with Gleason 7 (4 + 3), 53 (66.2%) with Gleason 8 (4 + 4), 11 (13.8%) with Gleason 9 (4 + 5), and 1 (1.3%) with Gleason 9 (5 + 4). Four patients had PSA < 4 ng/mL (5.0%), 48 patients had PSA 4.1–10 ng/mL (60.0%), 19 patients had PSA 10.1–20 ng/mL (23.8%), and 9 patients had PSA > 20 ng/mL (11.2%). The majority of patients had disease at the lower end of the high-risk spectrum (stage ≤ T2 [85.0%], Gleason ≤ 8 [84.9%], and PSA ≤ 20 ng/ mL [88.8%]). Median follow-up was 38 months (range: 3–158 54 mo). Actuarial 4-year biochemical relapse–free survival was 91%. There were four biochemical failures seen, with a median time to failure of 33.5 months (range: 27–48 mo). There was no significant correlation between T stage, Gleason score, PSA, or duration of hormone ablation therapy and recurrence. Although there were 10 deaths in this patient population, none of the deaths was related to prostate cancer. CONCLUSIONS: This study suggests that certain patients with highrisk prostate cancer may be effectively treated with a shorter course of hormone ablation therapy. Longer follow-up and more patients will be needed to verify the efficacy of a shorter course of hormone ablation therapy in this patient population. (P064) Outcomes After Adjuvant Radiation Therapy for Prostate Cancer at a Comprehensive Cancer Center Emma B. Holliday, Deborah A. Kuban, Lawrence Levy, Priya Master, Seungtaek Choi, Steven J. Frank, Andrew K. Lee, Sean E. McGuire, Usama Mahmood, Thomas J. Pugh, Karen E. Hoffman; UT MD Anderson Cancer Center BACKGROUND: Randomized trials have shown that adjuvant radiation therapy (RT) improves prostate cancer control for men with extracapsular extension (ECE), seminal vesicle (SV) involvement, and/or positive surgical margins (SMs) at the time of prostatectomy. We report cancer control outcomes for men who received adjuvant RT at a comprehensive cancer center. METHODS: Men who received adjuvant RT within 12 months of prostatectomy from 1987 through 2010 were identified in an institutional database. All men had PSA < 0.2 ng/mL at the time of RT. Failure was defined as a rising postradiation PSA of 0.2 ng/mL; local, nodal, or distant recurrence; or the initiation of salvage treatment. Time to failure was calculated from the end of radiation treatment. Cox proportional hazards models evaluated the association between patient characteristics, tumor characteristics, radiation treatment, and cancer control. Kaplan-Meier product-limit estimator was used to estimate 5- and 10-year failure. RESULTS: A total of 137 men received adjuvant RT. Median time from prostatectomy was 5.1 months (interquartile range [IQR]: 1.4–6.4 mo). Most men had positive SM (n = 127, 92.7%) and ECE (n = 98, 71.5%). A total of 38% had Gleason 8/9 disease (n = 52) at prostatectomy. Median radiation dose was 60 Gy (IQR: 60–66 Gy). Few men (n = 24, 17.5%) received concurrent hormone therapy (HT) with RT. Patients were more likely to receive HT if their prostatectomy was performed outside of the institution (P = .021) or if they had involved SV (P < .001). With a median follow-up of 11.4 years (IQR: 5.6–17.5 yr), 22 men failed. The 5-year failure for the entire cohort was 11.9%, and the 10-year failure was 16%. Gleason score was the only factor significantly associated with failure. The 5- and 10-year failure rates for men with Gleason 8/9 disease were 18.4% and 20.8%, respectively, while the 5- and 10-year failure rates for men with Gleason < 8 disease were 6.7% and 11.9%, respectively (P = .013). Radiation dose (< 66 Gy vs ≥ 66 Gy), surgical margin status, and receiving HT were not associated with failure. American Radium Society Scientific Papers and Posters 2015 CONCLUSIONS: At a comprehensive cancer center, men who are referred for adjuvant RT often have margin involvement and/or ECE. Patients did well overall, with high 5- and 10-year freedom from failure. Only Gleason 8/9 disease was associated with increased failure after adjuvant RT. (P065) Lack of Variation in Pathologic Upgrading and Upstaging by Race Among Patients With Low-Risk Prostate Cancer Twisha Chakravarty, MD, Karen Hoffman, MD, Lawrence Levy, MS, Tj Pugh, MD, Sean McGuire, MD, Seungtaek Choi, MD, Steven Frank, MD, Andrew Lee, MD, Deborah Kuban, MD, Usama Mahmood, MD; Department of Radiation Oncology, UT Medical Branch; Department of Radiation Oncology, UT MD Anderson Cancer Center PURPOSE/OBJECTIVES: Previous studies have demonstrated that the clinical presentation of prostate cancer varies according to race. Nonetheless, it is unclear whether rates of pathologic upgrading and/or upstaging vary according to race after accounting for factors at clinical presentation. Such variation could potentially impact management decisions. MATERIALS AND METHODS: Using the Surveillance, Epidemiology, and End Results (SEER) database, information was obtained for all men diagnosed with low-risk or very-low-risk (per the National Comprehensive Cancer Network [NCCN] definition) adenocarcinoma of the prostate who underwent prostatectomy in 2010 or 2011. Multivariable analyses were performed to determine predictors of pathologic upgrading (increase in total Gleason score) and pathologic upstaging (presence of extraprostatic extension or seminal vesicle invasion) at the time of prostatectomy. Both patient demographic (age, race) and clinical factors (T stage, Gleason score, prostate-specific antigen [PSA], number of positive cores, and number of examined cores) were included in the analyses. RESULTS: A total of 10,620 patients were identified, of whom 3,650 (34%) had very-low-risk disease. Median age for the entire cohort was 60 years; 7,717 (73%) patients were white, 1,251 (12%) were black, 1,023 (10%) were Hispanic, 441 (4%) were Asian, and 188 (2%) were Native American/unknown race. In total, 4,511 (42%) patients had pathologic upgrading, and 992 (9%) had pathologic upstaging at the time of prostatectomy. Among the entire cohort, increasing age at diagnosis, increasing PSA, increasing number of positive cores, and decreasing number of examined cores were all significant predictors of the presence of both pathologic upgrading and pathologic upstaging on multivariable analyses (all P < .001). On the other hand, race was not found to be a significant predictor of pathologic upgrading (P = .089) or pathologic upstaging (P = .522). Similar findings were noted among patients with very-lowrisk disease, with race once again not found to be a significant predictor of pathologic upgrading (P = .068) or pathologic upstaging (P = .410). CONCLUSIONS: Our analysis of this large, population-based database demonstrates that after accounting for other demographics and clinical factors at diagnosis, race does not predict for pathologic upgrading or pathologic upstaging at the time of prostatectomy among patients with low-risk prostate cancer. As such, race by itself should not be used to select potential candidates for active surveillance or treatment. (P066) Dosimetric Analysis of Proton Therapy for Prostate Cancer: A Single Institutional Experience Jing Zeng, Kent McCune, Matthias Cook, Grayden MacLennan, Malin Mao, George E. Laramore, Kenneth Russell, Jay Liao; University of Washington BACKGROUND: With the increasing number of proton therapy centers in the United States, more patients with prostate cancer have access to proton radiation as a treatment option. While dose constraints for organs at risk (OARs) are well defined for photon treatment, less is known for proton radiation, and photon constraints are often used after correcting for relative biological effectiveness (RBE = 1.1). However, it is often possible to achieve lower doses to OARs with proton therapy. Therefore, we investigated dose to OARs in a series of patients treated at our institution to consider lowering dose constraints for the proton population. METHODS: Consecutive patients with prostate cancer treated at our center from April–September 2014 were analyzed. All patients were treated with fiducial markers, daily kV image-guided radiation therapy (IGRT), full bladder, and rectal balloon. Clinical characteristics and OAR doses were extracted. Low-risk patients were treated to prostate plus 3–5-mm planning target volume (PTV) margin. Intermediate- and high-risk patients were treated to prostate plus proximal seminal vesicles, with optional pelvic nodal irradiation for high-risk patients. Treatment groups were compared with t-tests. RESULTS: A total of 40 patients with prostate cancer were treated, with a median age of 68 years (range: 51–79 yr); 10 patients had low-risk disease, 21 had intermediate-risk disease, and 9 had highrisk disease (5 received whole-pelvis radiation). A total of 31 patients were treated with pencil beam scanning, and 9 patients were treated with uniform scanning. Median prescription was 79.2 Cobalt Gray Equivalent (CGE). For the bladder: median V80 CGE = 4.9 cc (range: 0–13.8 cc) with constraint < 8 cc; median V70 CGE = 6.6% (range: 3.6%–22.1%) with constraint < 10%; and median V50 CGE = 11.5% (range: 6.1%– 32.8%) with constraint < 35%. For the bladder wall (3-mm internal rind): median V80 CGE = 3.5% (range: 0%–14.5%) with constraint < 15%; median V70 CGE = 13% (range: 8%–31%) with constraint < 35%, and median V55 CGE = 16.5% (range: 10.8%–36.5%) with constraint < 50%. For the rectum: median V81 CGE = 0 cc (range: 0–6.61cc) with constraint < 1 cc; median V70 CGE = 10.8% (range: 3.0%–19.3%) with constraint < 25%; and median V50 CGE = 19.6% (range: 10.9%–33.8%) with constraint < 35%. For the rectal wall (3–mm internal rind): median V81 CGE = 0 cc (range: 0–5.36 cc) with constraint < 1 cc; median V70 CGE = 22.6% (range: 5.7%–34.3%) with constraint < 25%; and median V50 CGE = 32.5% (range: 15.8%– 43%) with constraint < 50%. Comparing the low-, intermediate-, and high-risk disease groups, the only significant dose difference was in rectal wall V50 CGE (low ARS PROCEEDINGS 2015 55 American Radium Society Scientific Papers and Posters 2015 = 28.4%, intermediate = 32.1%, high = 35.0%). CONCLUSIONS: Proton therapy for prostate cancer typically generates treatment plans that are below dose constraints defined in the photon literature, with the exception of a small subset of patients. In order to promote the best possible proton treatment plans, lower dose constraints for OARs should be considered. (P067) The Role of Offline PET-CT Imaging in Evaluating the Particle Beam Range and Beam Stop for Prostate Cancer Treated With Heavy Ion Therapy Qing Zhang, Jingyi Cheng, Jiandong Zhao, Xiaomeng Zhang, Xin Cai, Jin Meng, Shen Fu; Shanghai Proton and Heavy Ion Center; Fudan University Shanghai Cancer Center; Shanghai Jiao Tong University, Shanghai No. 6 People’s Hospital; Shanghai Proton and Heavy Ion Center/Fudan University Shanghai Cancer Center PURPOSE: To evaluate the possibility of offline positron emission tomography-computed tomography (PET-CT) in monitoring both particle beam range and beam stop after carbon ion therapy. MATERIALS AND METHODS: From June 2014 to September 2014, a total of 19 patients with pathology-confirmed primary prostate adenocarcinoma underwent carbon ion therapy in Shanghai Proton and Heavy Ion Center. All patients received a PET-CT scan shortly after carbon ion therapy. We divided these patients into two groups; 3 of 19 patients were treated with a single lateral beam in the first 4 days of carbon ion therapy, followed by a bilateral beam from day (D) 5, and PET-CT images were collected on D1, D4, D11, and the last day of the treatment. Another 16 patients were treated with bilateral beam only, and PET-CT imaging was collected on D1, D4/5, and the last day of the treatment. After irradiation, patients were transferred immediately to the PET-CT room nearby the irradiation center. A PET-CT scan was done on a Siemens Biograph mCT 5–7 minutes after carbon ion therapy, with no additional radioactive tracer injection. PET-CT images were collected for approximately 30 minutes for each patient, while the patient’s position during the PET-CT scan was the same as during treatment using the same immobilization device. Finally, a total of 60 PET-CT images were investigated. PET-CT and planning CT were further registered by matching the pelvic bones. An additional comparison with the expected β+ activity distribution was performed to identify either beam range or beam stop with accurate anatomical coregistration. RESULTS: Based on the different tissue densities, we found different activation effects shown on PET imaging, such as skin, bone, fat, and prostate. Compared with the diagnostic CT, the measured activity with different anatomic structures was correlated, which indicated an accurate beam range. For the three patients treated with single-beam irradiation, the PET imaging also indicated the beam stop well. Our data further showed that the PET-CT–defined beam range covered the clinical target volume well, especially for the isocenter, by matching the PET-CT imaging with treatment planning. CONCLUSION: Posttreatment offline PET-CT imaging has the potential to evaluate the particle beam range and beam stop for 56 prostate cancer patients treated with heavy ion therapy, which might overcome the pitfalls of simple bone structure fusions with two-dimensional imaging, and provides another approach to monitor the treatment accuracy of prostate cancer treated with carbon ion therapy. Further investigation needs to be performed. (P068) VMAT vs Eight-Field IMRT: Dosimetric Comparison of Pelvic Radiotherapy for High-Risk Prostate Cancer Patients in Terms of Bone Marrow Sparing Yasemin Bolukbasi, Vildan Alpan, Yucel Saglam, Ugur Selek; American Hospital, MD Anderson Radiation Treatment Center, Istanbul PURPOSE: Although there is no complete consensus on elective pelvic nodal irradiation for high-risk prostate cancer patients, pelvic radiotherapy with androgen ablation have been more commonly used in many centers. An important part of bone marrow (BM) reserve remains in the pelvic radiation treatment field. We intend to evaluate and compare the intensity-modulated radiotherapy (step and shoot IMRT [ssIMRT]) and volumetric arc radiotherapy (VMAT) techniques for pelvic radiotherapy in terms of pelvic bone marrow doses. MATERIALS AND METHODS: This study was based on the simulation scan data of 10 prostate cancer patients as 3-mm slice thickness using a full bladder and rectal balloon. The first phase of the treatment planning was prescribed to the pelvic lymphatics, prostate, and seminal vesicles (46 Gy, 2 Gy/fraction); then, the second phase consisted of the seminal vesicles and prostate (32 Gy, 2 Gy/ fraction). Planning target volume (PTV) margin was 0.4 cm posteriorly due to the rectum and 0.6 cm in all other directions. Using the same target volumes, ssIMRT with eight angles (225°, 260°, 295°, 330°, 30°, 65°, 100°, and 135°) and double-arc (182°–178° arc angle) VMAT were planned for each patient dataset. The planning objective was to cover the PTV by at least 95% of the prescribed isodose and clinical target volume (CTV) by 98% of the prescribed isodose line. No special dose constraint was given for bone marrow sparing. Each technique was compared by using dose-volume histograms (DVHs) of V5, V10, V20, V30, and V40 of the sacrum BM, iliac BM, ischium, pubis, proximal femora (lower pelvis), and femoral BM; V20, V30, V40, and V70 for the bladder; and V30, V40, V76, and V80 for the rectum, homogeneity index ??and the monitor units (MU). Two-sided Wilcoxon’s test was used for statistical analysis (P < .05). RESULTS: For the same PTV coverage, VMAT and ssIMRT plans had similar dose distributions for femur, iliac, sacrum, and total BM, as well as the other critical structures. However, VMAT plans in comparison with IMRT ensured significantly lower high-dose volumes on the rectum, such as bringing V80 from 1.6% to 0.9% (P = .01), and provided similar homogeneity index with lowered MUs (1,048 vs 1,591; P = .018). CONCLUSION: In this cohort, VMAT plans without a specific constraint for BM were not found to be superior to ssIMRT in terms of BM reserve irradiation, while VMAT could be encouraged for patients with higher rectum doses, such as V80. American Radium Society Scientific Papers and Posters 2015 (P069) Gleason 7 Prostate Adenocarcinoma Treated With High- or Low-Dose-Rate Brachytherapy: Impact of External Beam Radiotherapy and/or Androgen Deprivation Therapy Julian Johnson, MD, Charles Hsu, I-Chow Hsu, MD, Vivian K. Weinberg, PhD, Alexander Gottschalk, MD, PhD, Barby Pickett, MS, Mack Roach; University of California, San Francisco; Texas Oncology PURPOSE: To determine the impact of hormone therapy (HT) and/ or external beam radiotherapy (EBRT) on prostate-specific antigen (PSA) nadir (nPSA) in patients with Gleason score (GS) 3 + 4 or 4 + 3 prostate cancer treated with low-dose-rate (LDR) or high-doserate (HDR) brachytherapy (BT) at a single institution. MATERIALS AND METHODS: A total of 148 men were retrospectively identified with GS 7 (21% GS4 + 3, 79% GS3 + 4) cT1–T2cN0 prostate cancer receiving LDR (76%) or HDR (24%) BT as a component of treatment. LDR or HDR monotherapy was administered to 29% (EBRT was administered in 58%; HT was administered to 51%). Median follow-up from BT until last PSA or death was 72 months (range: 1–141 mo). nPSA was defined as current postimplant PSA nadir as of last visit. Disease failure was defined as PSA failure, local recurrence, metastasis, or salvage therapy. RESULTS: Median time to nPSA was 43.4 months vs 29.8 months for patients treated with LDR vs HDR (P = .03). Patients treated with HDR were more likely to have T2 disease (P = .01) and had higher median baseline PSA (8.6 vs 6.2 ng/mL; P = .004). Patients treated with HDR were more likely to receive HT (42% vs 69%; P = .01) but not EBRT (54% vs 61%; P = .56). There was no statistically significant difference between nPSA after LDR vs HDR (median: 0.1 vs 0.1 ng/mL; P = .27). Median nPSA after LDR for GS 3 + 4 vs 4 + 3 was 0.1 ng/mL and 0.05 ng/mL, respectively (P = .32). Median nPSA after HDR for GS 3 + 4 vs GS 4 + 3 was 0.1 vs 0.06 ng/mL, respectively (P = .62). Treatment with HT resulted in a median nPSA of 0.045 vs 0.1 ng/mL (P < .0001). EBRT resulted in median nPSA of 0.06 ng/mL vs 0.1 ng/mL (P = .05). Patients treated with LDR brachytherapy had a lower nPSA if treated with HT (0.05 ng/ mL vs 0.1 ng/mL; P = .0002). HT did not result in a lower nPSA in patients treated with HDR (P = .18). There was a statistically significant lower nPSA among LDR patients when combined with EBRT vs no EBRT (0.05 ng/mL vs 0.1 ng/mL; P = .003) but not among HDR patients (P = .52). Freedom from disease failure rate at 5 years was 92% vs 94% for LDR vs HDR, respectively (P = .00). There was no statistically significant difference with ADT (95% vs 89%; P = .33) or with EBRT (93% vs 91%; P = .37). CONCLUSIONS: nPSA has been related to biochemical progression–free survival, freedom from metastasis, and death from prostate cancer. BT with either EBRT or HT achieves a lower PSA nadir. There was no difference in disease failure. Longer follow-up may be necessary to see differences in disease failure in this population. (P070) High-Risk Prostate Adenocarcinoma Treated With Whole-Pelvis Radiotherapy HDR Brachytherapy Boost Results in Very High Disease-Specific Survival Julian Johnson, MD, Mack Roach, MD, Alexander Gottschalk, MD, PhD, Adam Cunha, PhD, Zachary Seymour, David Raleigh, MD, PhD, Katsuto Shinohara, MD, I-Chow Hsu, MD, Albert Chang, MD, PhD; University of California, San Francisco PURPOSE/OBJECTIVES: To report the long-term clinical outcomes of patients with high-risk adenocarcinoma of the prostate treated with pelvic radiotherapy followed by inverse planned, templatefree high-dose-rate (HDR) brachytherapy boost treated at a single institution. MATERIALS AND METHODS: A total of 115 patients with high-risk prostate cancer (≥ cT3, Gleason score [GS] 8–10, or PSA ≥ 20 ng/ mL) treated between July 1997 and November 2005 were included in this study. All patients underwent whole-pelvis external beam radiotherapy to 45 Gy followed by HDR brachytherapy boost. HDR brachytherapy boost consisted of 6 Gy × 3 (38 patients) or 9.5 Gy × 2 (77 patients) to the prostate and seminal vesicles. A total of 47%, 48%, and 5% of patients received long-term (> 18 mo), short-term (4–6 mo), or no androgen deprivation therapy, respectively. RESULTS: The median age at diagnosis was 64.8 years; 50 (43%), 90 (78.2%), and 24 (20.8%) patients were diagnosed with GS 8–10, cT3 disease, and PSA ≥ 20 ng/mL, respectively. Mean PSA was 14.94 ng/mL (range: 0.21–99.3 ng/mL). Further, 21 patients (18.1%) had seminal vesicle invasion, and 42 patients (37%) had at least two high-risk features. With a median follow-up time of 94 months, 5- and 10- year biochemical-free survival, as defined by the Phoenix definition, was 90% and 73%, respectively. At 10 years, four patients failed locally, as determined by biopsy, for a local control rate of 94%. The 5- and 10-year distant metastasis-free survival rates were 97% and 87%, respectively. The 5- and 10-year disease-specific survival rates were 100% and 92.5%, respectively. Six secondary malignancies (bladder carcinoma 1 year posttreatment, colon, lung, melanoma, and hepatocellular carcinoma) developed. One patient developed grade 3 ureteral stricture that caused hydronephrosis and required stent placement. There were no other grade ≥ 3 genitourinary or gastrointestinal toxicities. CONCLUSION: Pelvic radiotherapy followed by HDR brachytherapy boost is an effective treatment for high-risk prostate cancer. It provides excellent long-term disease control and very low rates of severe (grade ≥ 3) toxicity when delivered in two or three fractions. (P071) Socioeconomic Disparities in Baseline Magnetic Resonance Imaging (MRI) Utilization and Imaging Characteristics for Prostate Cancer (PCa) Patients Undergoing Radiotherapy Ayobami Ajayi, BA, Atu Agawu, BS, Neha Vapiwala, MD, Stefan Both, PhD, Zelig Tochner, MD, Curtiland Deville, MD; Department of Radiation Oncology, Perelman School of Medicine PURPOSE: To determine whether socioeconomic disparities exist in magnetic resonance imaging (MRI) utilization and imaging characteristics for prostate cancer (PCa) patients undergoing radiotherapy. METHODS: An institutional database identified 598 nonmetastatic PCa patients treated with proton and/or intensity-modulated radiotherapy from 2005–2013 with pretreatment pelvic MRI ARS PROCEEDINGS 2015 57 American Radium Society Scientific Papers and Posters 2015 (T2-weighted, diffusion-weighted, and/or dynamic contrastenhanced sequences; 91% with endorectal coil). Dominant intraprostatic lesion (IPL) presence, size, extracapsular extension (ECE), and seminal vesicle invasion (SVI) were reviewed. Patients of higher socioeconomic status (SES) vs low SES (defined as geocoded census tract > 20% below the poverty level) overall and stratified by risk group were compared using Fisher’s exact test for categorical variables and Student’s t-test for continuous variables (P < .05 significant). RESULTS: In this cohort, 437 were classified as higher-SES (73%) and 161 were classified as low-SES (27%). Mean age (67 ± 8 yr), Gleason score, and clinical stage did not differ significantly for higher-SES vs low-SES patients. Low-SES patients were more likely to be black (71%) than higher-SES patients (16%) (P < .001). LowSES patients also had significantly higher prostate-specific antigen (PSA) levels (higher-SES: 9.5 ± 21 ng/mL, low-SES: 15.6 ± 22 ng/ mL; P < .001) and were more likely to be high-risk (higher-SES: 22% vs low-SES: 34%; P = .006). Higher-SES patients had a greater proportion of low-risk patients (38%) than did low-SES subjects (25%; P = .005). Overall, 457 (76%) patients underwent baseline MRI. Low-SES patients (69%) were less likely than higher-SES patients (79%; P = .01) to undergo MRI. This disparity was greatest for low-risk patients (higher-SES: 84%, low-SES: 56%; P < .001) and not significant for intermediate-risk (77% vs 71%; P = .389) or high-risk patients (77% vs 81%; P = .675). There were no significant differences between SES groups in dominant IPL presence, diameter, area, ECE, or SVI—overall or within intermediate- and highrisk groups. Low-risk, low-SES patients demonstrated larger dominant IPL area (133.5 ± 141 mm2 vs 66.7 ± 58 mm2; P = .017), with no differences in dominant IPL diameter or presence, additional IPLs, ECE, or SVI. CONCLUSIONS: In this urban, academic center cohort, PCa patients of lower SES were significantly less likely to undergo staging MRI, particularly in the low-risk group. No differences were found in dominant IPL presence, area, ECE, or SVI, except in the low-risk group. Further investigation is required to better understand trends in pretreatment MRI utilization and dominant IPL characteristics differing by SES. (P072) Analysis of Composite EQD2 Dose Distribution in Radiotherapy for Cervical Cancer Using Central Shielding Technique Tomoaki Tamaki, MD, PhD, Shin-ei Noda, MD, PhD, Tatsuya Ohno, MD, PhD, Shingo Kato, MD, PhD, Takashi Nakano, MD, PhD; Saitama Medical University International Medical Center; Gunma University Graduate School of Medicine PURPOSE: To analyze the three-dimensional equivalent dose in 2-Gy fractions (EQD2) dose distribution of external beam radiotherapy (EBRT) plus intracavitary brachytherapy (ICBT) for cervical cancer using the central shielding technique. MATERIALS AND METHODS: In a phantom study, a whole-pelvis irradiation (WP) plan was created using the four-field box technique and pelvis irradiation with the central shielding technique (CS) using AP-PA fields with a central block having a 3-cm or 4-cm 58 width. Three patterns of EBRT were created for WP and CS: 30 Gy/15 fractions plus 20 Gy/10 fractions (Plan 30+20); 40 Gy/20 fractions plus 10 Gy/5 fractions (Plan 40+10); and 45 Gy/25 fractions and 0 Gy (Plan 45+0). For ICBT, two plans were created using Point-A prescription: 24 Gy/4 fractions (BTPlan 24/4) for Plan 30+20 and Plan 40+10, and 28 Gy/4 fractions (BTPlan 28/4) for Plan 45+0. The physical doses were converted to EQD2, and the composite EQD2 dose distributions were analyzed. RESULTS: In Plan 30+20 plus BTPlan 24/4 and Plan 40+10 plus BTPlan 24/4, the area covered with 60 Gy (EQD2) in the lateral direction was larger than that in Plan 45+0 plus BTPlan 28/4. There were no “cold” spots within the lateral axis, which indicates that even with the use of CS, the treatment could provide adequate dose coverage for the central tumor and the parametrial tissue. The Point-A doses of Plan 30+20 plus BTPlan 24/4, Plan 40+10 plus BTPlan 24/4, and Plan 45+0 plus BTPlan 28/4 were 78.0 Gy (CS: 3 cm) or 71.8 Gy (CS: 4 cm), 80.1 Gy (CS: 3 cm) or 77.0 Gy (CS: 4 cm), and 84.1 Gy, respectively. These values were higher than the total Point-A doses reported in previous clinical studies, which omitted the doses provided by CS. On the other hand, the coverage in the anterior-posterior direction was smaller in Plan 30+20 plus BTPlan 24/4 and Plan 40+10 plus BTPlan 24/4 as a result of CS, indicating that the high dose to the bladder and the rectum can be avoided, while the adequate tumor coverage in this direction needs to be assessed carefully. CONCLUSIONS: Three-dimensional composite dose distribution analysis plays a significant role in the correct understanding of the dose distribution of the combination of EBRT and ICBT for cervical cancer. The use of CS in radiotherapy for cervical cancer provided tumor doses higher than those referred by the Point-A dose in previous reports, with no irregularly “cold” regions around the central target. (P073) Optimal Epidural Analgesia During Interstitial Brachytherapy for Treatment of Gynecological Cancer Ashley K. Amsbaugh, MD, Mark J. Amsbaugh, MD, Moataz N. El Ghamry, MD, Brian M. Derhake, MD, MS; Department of Anesthesiology, Department of Radiation Oncology, University of Louisville PURPOSE: To determine optimal epidural analgesia for patients receiving interstitial brachytherapy (ISBT) for gynecologic cancer. MATERIALS AND METHODS: Records of all patients who underwent interstitial brachytherapy (ISBT) at our institution between January 2009 and July 2014 were reviewed. ISBT was delivered over the course of 2 to 3 days, and maximum pain scores (measured on a scale from 1 to 10 points) were recorded every 8 hours. The primary analgesia method was epidural catheter. In addition to epidural anesthetic, patients received “as-needed” medications (intravenous narcotics, oral narcotics, and acetaminophen) from a standard order set. Antiemetics and diphenhydramine were available for nausea and pruritus, respectively. Pain scores and administered medications were collected, and all narcotic medications were converted to intravenous morphine equivalent (IVME). Statistical analysis was performed with SAS (Statistical Analysis System) software (SAS Institute, Cary NC). American Radium Society Scientific Papers and Posters 2015 RESULTS: Epidural catheters were successfully placed in 71 of 73 patients. Twelve patients received ropivacaine alone, 14 patients received ropivacaine with fentanyl, and 45 patients received ropivacaine with hydromorphone. Patients receiving ropivacaine alone had higher pain scores than patients receiving ropivacaine with fentanyl or ropivacaine with hydromorphone on the morning of day 2 (4.2 vs 1.71 vs 0.6; P = .001), the afternoon of day 2 (4.9 vs 2.5 vs 1.7; P = .005), and the night of day 2 (2.4 vs 2.0 vs 0.6; P < .001). Patients receiving narcotics in their epidural had lower pain scores on the night of placement (P = .050), the morning of day 2 (P < .001), the afternoon of day 2 (P = .002), and the night of day 2 (P < .001). Patients receiving ropivacaine alone used more oral narcotics than those receiving ropivacaine with fentanyl or ropivacaine with hydromorphone on day 3 (5.9 mg vs 3.8 mg vs 2.8 mg IVME) and received more intravenous narcotics on day 1 (5.8 mg vs 0.0 mg vs 0.7 mg IVME; P = .004) and day 2 (20.6 mg vs 4.8 mg vs 1.0 mg IVME; P = .042). There were no differences in antiemetic use on days 1 (P = .146), 2 (P = .266), or 3 (P = .360). There were no differences in diphenhydramine usage on days 1 (P = .829), 2 (P = .678), or 3 (P = .413). No epidural complications occurred. Obstetrics (FIGO) stage was I (n = 1), II (n = 2), and IV (n = 1); the remaining three patients had vaginal cuff recurrence. All patients were simulated and treated after insertion of a Foley urinary catheter with manual bladder filling to a predetermined volume. Patients also underwent bowel preparation at the time of simulation and before each treatment fraction, and daily cone-beam image-guided radiotherapy (IGRT) was utilized. The dose-volume histograms for planning target volumes, as well as organs at risk, were all analyzed. CONCLUSIONS: Epidural analgesia provides safe and effective pain control in patients undergoing ISBT. Epidural delivery of narcotics with ropivacaine improves pain control and lowers oral and intravenous narcotic requirements without increased risk of adverse effects. CONCLUSION: SBRT boost appears to be an effective and welltolerated alternative treatment method for patients with gynecologic malignancies who cannot undergo brachytherapy. Target coverage and dose to organs at risk with SBRT appear to be comparable with those of brachytherapy. Acute toxicity is minimal. These initial clinical results substantiate the need for further evaluation of this treatment approach and its long-term efficacy and safety. (P074) The Use of Stereotactic Body Radiotherapy in Lieu of Brachytherapy in Patients With Cervical or Endometrial Cancer Nabila L. Waheed, DO, Michelle Ludwig, MD, PhD, Celestine Tung, MD, Marian Williams-Brown, MD, Creighton L. Edwards, MD, Concepcion Diaz-Arrastia, MD, Matthew L. Anderson, MD, PhD, Lois Ramondetta, MD, Timothy Wagner, MD, Joshua Asper, PA, Danny Tran, CMD, Umang Patel, MD, Mark Bonnen, MD; Department of Radiation Oncology, Department of Obstetrics and Gynecology, Baylor College of Medicine; Department of Gynecologic Oncology, UT MD Anderson Cancer Center PURPOSE: To evaluate the use of stereotactic body radiation therapy (SBRT) as an alternative treatment method to brachytherapy in patients with cervical or endometrial carcinoma with anatomically difficult-to-implant tumor volumes or other contraindications to brachytherapy. MATERIALS AND METHODS: A total of seven patients undergoing radiation therapy with curative intent for cervical or uterine cancer were identified after they had been deemed ineligible for brachytherapy due to unfavorable anatomy, including large tumor size, location, and/or prior hysterectomy. Patients diagnosed with either cervical (n = 5) or endometrial (n = 2) cancer were treated using 45–48.6 Gy external beam radiation to the pelvis +/− para-aortic lymph nodes. Either three-dimensional (3D) radiation or intensitymodulated radiotherapy (IMRT) was followed by Linac-based SBRT to the cervix or vaginal cuff at a dose of 5–6 Gy in five daily fractions to a total of 25–30 Gy. Median age was 52 years (range: 47–64 yr). The International Federation of Gynecology and RESULTS: SBRT was well tolerated. No acute severe urinary or gastrointestinal toxicity > grade 2 was observed during treatment or up to 5 months posttreatment. Target volume of 95% gross tumor volume (GTV) coverage goal was > 95% (range: 95%–99%). D0.2cc to the bladder was kept at or below 90 Gy (range: 87.3–90 Gy). Total bladder D0.2cc contribution from SBRT for all five fractions was 20.5–40 Gy. D0.2cc to the rectum was kept at or below 75 Gy (range: 69–73.8 Gy). Total rectal D0.2cc contribution from SBRT for all five fractions was 13.1–23.8 Gy. No patient has had any sign of progression or local failure at follow-up thus far, with a median follow-up of 3 months. (P075) Management of Nodal Recurrences of Endometrial Cancer With IMRT Jennifer Ho, Anuja Jhingran, MD, Shannon Westin, MD, Karen Lu, MD, Patricia Eifel, Ann Klopp, MD, PhD; UT MD Anderson Cancer Center PURPOSE: Pelvic and para-aortic lymph node regions are frequent sites of relapse in women with endometrial cancer who have not undergone adjuvant pelvic radiation. We investigated outcomes following radiation therapy with intensity-modulated radiation therapy (IMRT) for definitive treatment of nodal relapses of endometrial cancer at our institution. MATERIALS AND METHODS: Between 2002 and 2012, a total of 42 patients with endometrial cancer who had no prior pelvic external beam radiation were treated definitively using IMRT for pelvic and/or para-aortic nodal recurrences. A total of 12 patients (29%) had pelvic nodal recurrences only, 8 (19%) had para-aortic recurrences only, 10 (24%) had simultaneous pelvic and para-aortic recurrences, and 12 (28%) had simultaneous pelvic and other regionally confined recurrences. Also, 15 patients (35%) had chemotherapy before radiation, and 21 (50%) had concurrent chemotherapy with radiation. The median size of the largest nodal recurrence site was 2.9 cm (range: 1.3–9.1 cm). The nodal basins at risk were typically treated to 45–50 Gy, with a boost to the gross tumor, for a mean total dose of 64.8 Gy (range: 59–73 Gy). Survival rates were calculated using the Kaplan-Meier method. RESULTS: The median overall survival (OS) from date of recurARS PROCEEDINGS 2015 59 American Radium Society Scientific Papers and Posters 2015 rence was 45.1 months (95% confidence interval [CI], 28.3–61.8 mo), and the 2-year survival was 71%. A total of 16 (38%) patients developed local failures within the salvage radiation fields at a median time of 7.2 months (range: 2.4–28.6 mo), of which 11 were failures located within the high-dose regions. Further, 20 (48%) patients developed distant recurrences at a median time of 6.8 months (range: 1.2–31.9 mo). Patients who received concurrent chemotherapy had longer median survival than patients treated without concurrent chemotherapy (61.9 mo vs 28.9 mo; P = .029). Patients who received chemotherapy prior to radiation had shorter median survival compared with those who did not (28.3 mo vs 61.9 mo; P = .001) and a lower rate of survival free of local recurrence (28% vs 73% at 2 yr; P = .012). No significant survival difference was detected in survival or local recurrence based on histology, tumor size, or site of recurrence. A total of 11 patients (26%) experienced grade ≥ 3 gastrointestinal toxicity. CONCLUSION: Long-term survival can be achieved following salvage radiation for nodal recurrence of endometrial cancer. However, central and distant recurrences remain a challenge. Chemotherapy prior to radiation was associated with an increased rate of central recurrences and reduced survival, while the use of concurrent chemotherapy was associated with higher rates of survival. (P076) Postoperative Treatment Recommendations for Stage I Endometrial Cancer: A Survey of Society of Gynecologic Oncology Members Brian S. De, BA, Elena Pereira, MD, Valentin Kolev, MD, Konstantin Zakashansky, MD, Peter R. Dottino, MD, Sheryl Green, MD, Vishal Gupta, MD; Icahn School of Medicine at Mount Sinai OBJECTIVES: To assess postoperative adjuvant therapy (AT) recommendations for International Federation of Gynecology and Obstetrics (FIGO) stage I endometrioid endometrial cancer (EC) among members of the Society of Gynecologic Oncology (SGO). METHODS: A 19-question survey was developed, approved by our institutional review board, and emailed to SGO members. Data were collected anonymously using Internet-based survey software. Demographic questions included specialty, years in practice, practice setting, and EC patient volume. Respondents were asked questions regarding preoperative workup, surgical approach, lymph node dissection (LND), and AT based on various clinicopathologic scenarios. AT options included no further treatment or any combination of brachytherapy (BT), external beam radiotherapy (EBRT), and/or chemotherapy (CT). Here, we report the results of AT recommendations. Statistical analysis was performed using Statistical Package for the Social Sciences version 22.0 (SPSS v22.0). RESULTS: Of the 1,399 SGO members, 320 (23%) completed the survey: 97% of respondents were gynecologic oncologists or fellows, and 2% were radiation oncologists; 49% of respondents had > 10 years of experience, 81% practiced at a university hospital or a community hospital with an academic affiliation, and 87% treated > 30 EC patients yearly. As expected, AT was chosen more frequently with greater myometrial invasion, higher tumor grade (G), and lymphovascular invasion (LVI). Respondents typically did not select any AT for stage IA, G1–2 without LVI. For stage IA, 60 G3, +LVI disease, respondents chose BT only (55%), BT + CT (16%), BT + EBRT (7%), EBRT only (5%), CT only (5%), or no further therapy (8%); used alone or in combination, respondents most frequently selected BT (82%), followed by CT (24%) and EBRT (15%). For stage IB, G3, +LVI disease, respondents chose BT only (29%), BT + CT (27%), EBRT only (13%), BT + EBRT (10%), CT only (8%), BT + EBRT + CT (6%), EBRT + CT (5%), or no further therapy (2%); used alone or in combination, respondents most frequently selected BT (73%), followed by CT (48%) and EBRT (34%). A total of 70% of respondents considered age when planning treatment. Older patients were recommended to have AT, particularly BT, in earlier stages of disease, with similar use of CT and EBRT. CONCLUSIONS: BT is the most common AT modality recommended by SGO members in the postoperative management of stage I EC. CT was recommended in a substantial number of scenarios and even exceeded the use of EBRT in G3 disease. Although there was generally agreement about the management of low-risk EC, there was much more variability in high-intermediate–risk patients. Further studies are needed to compare these SGO members’ recommendations with those of radiation oncologists and to determine optimal management. (P077) Defining the Pendulum Swing From WholePelvic Radiation Therapy (WPRT) Alone to Vaginal Brachytherapy (VB) in the Adjuvant Setting for Endometrial Cancer: A SEER-Based Analysis From Years 2000–2011 Sagar C. Patel, MD, Sudershan Bhatia, MD, PhD, William Rockey, MD, PhD, Mindi Tennapel, BSMS, MBA, Wenqing Sun, MD, PhD; University of Iowa Hospitals and Clinics PURPOSE: To examine the change in the utilization trends of radiation therapy in the adjuvant setting for endometrial cancer in a Surveillance, Epidemiology, and End Results (SEER) analysis from 2000–2011. PATIENTS AND METHODS: We analyzed the change in utilization of whole-pelvic radiation therapy (WPRT) alone, vaginal brachytherapy (VB) alone, and WPRT + VB in all patients with endometrial cancer in the adjuvant setting, stratified by age, race, marital status, specific US registry, histology, and grade. In addition, a subset analysis of endometrial cancer patients with localized-stage uterine disease, endometrioid histology, and varying age and grade was completed to model intermediate- and high-risk groups. Percentage change (PC) and annual percentage change (APC) in treatment utilization were calculated for each radiation therapy modality. APC characterizes utilization trends over time and is comparable across scales in SEER*Stat. Utilization rates were analyzed by using SEER*Stat with a log-linear regression model. RESULTS: In the years 2000–2011, there was a 220.1% (PC) and 11.7% (APC) increase in the utilization of VB, while there was a 33.1% (PC) and 4.1% (APC) decrease in the utilization of WPRT alone (P < .05). For each demographic and clinical variable, the most pronounced increase in the utilization of VB over WPRT alone was in individuals aged ≥ 70 years (248.4% [PC] and 12.2% American Radium Society Scientific Papers and Posters 2015 [APC]), in those of Asian race (584% [PC] and 13.4% [APC]), in the SEER registry Louisiana (1,216.5% [PC] and 12.5% [APC]), in grade 3 tumors (303.3% [PC] and 12.9% [APC]), and in those with carcinosarcoma histology (1,251.5% [PC] and 32.3% [APC]). On the contrary, the most pronounced decrease in the utilization of WPRT alone was in individuals aged ≥ 70 years (41.4% [PC] and 5.2% [APC]), in those of white race (37.1% [PC] and 4.4% [APC]), in the SEER registry Louisiana (60.1% [PC] and 7.0% [APC]), in grade 2 tumors (52.2% [PC] and 6.8% [APC]), and in those with carcinosarcoma histology (56.8% [PC] and 3.3% [APC]). CONCLUSION: This SEER study demonstrates the pronounced increase in the utilization of VB over WPRT alone in the adjuvant setting for endometrial cancer. This observation may spark critical evaluation of US practice patterns seen from 2000 to 2011 and its potential impact on cost, new policies, and the patient’s quality of life. (P078) Radiation Therapy in the Management of Carcinosarcoma of the Uterus Jillian R. Gunther, MD, PhD, Eva N. Christensen, MD, PhD, Pamela K. Allen, PhD, Lois M. Ramondetta, MD, Anuja Jhingran, MD, Nicole D. Fleming, MD, Elizabeth Euscher, MD, Karen H. Lu, Patricia J. Eifel, MD, Ann H. Klopp, MD, PhD; UT MD Anderson Cancer Center; Radiation Oncology Associates PURPOSE: Uterine mixed Müllerian tumor (MMMT) is often treated with multimodality therapy, including surgery, chemotherapy, vaginal cuff brachytherapy (VCB), and/or pelvic radiation. However, the indications for pelvic and vaginal cuff radiation are unclear. To investigate this, we reviewed our institutional experience with treatment of uterine MMMT. METHODS: We performed a retrospective review of 155 women with stage I–III uterine carcinosarcoma who underwent total abdominal hysterectomy-bilateral salpingo-oophorectomy (TAHBSO) at our institution between 1990 and 2011. Overall survival (OS), disease-specific survival (DSS), and pelvic relapse–free survival (PRFS) were calculated using the Kaplan-Meier method, with differences assessed using a log-rank test. Multivariate Cox proportional hazards regression analyses were performed to evaluate the influence of different factors on DSS and PRFS. RESULTS: The study included patients with stage I (n = 98), II (n = 11), and III (n = 46) uterine MMMT. Median follow-up was 51 months (range: 2–278 mo). A total of 70 patients (45%) received chemotherapy (21 concurrent, 58 adjuvant, 9 both). Also, 108 patients (70%) received radiation therapy (RT); 35% of these patients received pelvic radiation, and 65% received pelvic radiation with or without additional brachytherapy. The 5-year OS for all patients was 48.6% (stage I, 53.8%; II, 30.0%; and III, 42.5%). The 5-year DSS was 57.17% (stage I, 60.88%; II, 44.44%; and III, 51.82%). The first site of recurrence was in the pelvis and/or paraaortic (PA) lymph nodes for 28 patients (44%), abdomen ± pelvis for 13 patients (21%), and distant for 22 patients (35%). On univariate analysis, lower rates of DSS were seen in patients aged ≥ 65 years (P = .001) and with hypertension (P = .03), higher tumor grade (P = .02), cervical involvement (P = .001), and lymphovascu lar space invasion (LVSI) (P = .004). On multivariate analysis, age ≥ 65 years (P = .001), cervical involvement (P = .03), and LVSI (P = .01) remained independently associated with lower DSS. PRFS was higher in patients receiving pelvic radiation as compared with VCB or no radiation (88.3% at 5 yr for pelvic radiation vs 67.4% for VCB and 71.2% for no RT; P = .04). Among patients who did not receive pelvic radiation, cervical involvement (hazard ratio [HR] = 3.2; P = .03) and heterologous elements (HR = 2.9; P = .04) were associated with pelvic relapse. In stage III patients, pelvic radiation was associated with higher rates of 5-year PRFS (90.0% vs 55.5%; P = .046), DSS (64.6% vs 46.4%; P = .13) and OS (64.6% vs 34.0%; P = .04). CONCLUSIONS: Pelvic radiation reduces the risk of pelvic relapse in patients with MMMT and may be indicated with patients at high risk for pelvic recurrence, including patients with cervical involvement, heterologous elements, and stage III disease. (P079) Retrospective Review of Chemoradiation in the Preoperative or Definitive Management of Locally Advanced Vulvar Cancer Melissa Joyner, MD, Gwyn Richardson, MD, Lyuba Levine, Sandra Hatch; UT Medical Branch, Galveston OBJECTIVES: To review historical use of preoperative chemoradiotherapy at a single institution in patients with locally advanced vulvar cancer who were not surgical candidates due to extent of disease. Patients were treated with intent to improve local control with organ preservation. METHODS: Historical chart review of 12 patients with an average age of 52 years (range: 40–72 yr) treated between 1997 and 2014, all with locally advanced clinical stage T3 or T4 squamous cell carcinomas of the vulva not amenable to surgical resection. A total of 2 of 12 patients presented with locally advanced recurrent disease and nodal relapse. All patients were treated with external beam to 4,760 cGy (1.7 Gy per fraction × 28 fractions) using an accelerated fractionation schema consistent with the Gynecologic Oncology Group (GOG) 101 protocol with a planned treatment break in conjunction with concurrent cisplatin and 5-fluorouracil (5-FU). A single patient treated with a modified fractionation schema after the first cycle of radiation demonstrated superior treatment response, which facilitated surgical resection; this patient was then followed with additional radiation to treat residual microscopic disease. Further, 7 out of 12 patients (58.3%) also received a boost ranging from 7.2 Gy to 17.2 Gy, with a single patient receiving 12 Gy standard deviation [SD] via vaginal cylinder. RESULTS: The study patients had an average of 75 months of follow-up. Sustained local control was achieved in 8 of 12 patients (66.7%). A complete clinical response (cCR) was seen in 100% of patients following treatment. Despite extensive disease at presentation, only 2 of 12 patients (16.7%) failed in the vulva. A significant number of patients remained disease-free with no evidence of distant metastases (58%)—5 of 12 patients were alive without any disease, and an additional 2 of 12 expired without evidence of disease. While 2 out of 12 patients exhibited evidence of active disease, they ARS PROCEEDINGS 2015 61 American Radium Society Scientific Papers and Posters 2015 are alive and undergoing additional therapy. Only 2 of 12 patients to date have expired with evidence of disease, and 1 patient expired with disease status unknown. Biopsies only were done of the primary and/or node in 4 of 12 patients (33%), which were negative for disease. Organ preservation was achieved in 100% of patients. CONCLUSIONS: This treatment schema provided excellent tolerance with sustained local control. All patients obtained cCR and were able to avoid pelvic exenteration surgery and maintain preservation of bladder and rectal function. (P080) Favorable Outcomes Using Radiation Therapy After Chemotherapy in the Management of Primary, Recurrent, and Metastatic Ovarian Cancer Carolina E. Fasola, MD, MPH, Daniel S. Kapp, MD, Elizabeth Kidd, MD; Department of Radiation Oncology, Stanford University School of Medicine PURPOSE: This study aimed to evaluate treatment outcomes and toxicity of radiation therapy (RT) in patients with primary, recurrent, or metastatic ovarian cancer following treatment with chemotherapy. MATERIALS AND METHODS: We reviewed the medical records of 133 patients with stage I–IV ovarian cancer treated at our institution between 1997 and 2014. Patients underwent the following treatment regimens after initial tumor debulking surgery and chemotherapy: whole-abdominopelvic RT (n = 15), RT at the time of localized recurrence (n = 57), or palliative RT in the metastatic setting (n = 61). Freedom from progression (FFP) and overall survival (OS) were estimated using Kaplan-Meier analysis. RESULTS: The median follow-up time among all patients was 15.6 months (range: 1–200 mo) and 22.5 months among patients treated with curative intent for primary or recurrent disease. For patients treated with curative intent, the majority had International Federation of Gynecology and Obstetrics (FIGO) stage III disease at initial presentation (65%) with papillary serous histology (47%). A total of 55 (76%) patients had platinum-sensitive disease, and 17 (24%) were considered platinum-resistant. The median prescription RT dose in this cohort of patients was 45.9 Gy (range: 30–57.6 Gy) delivered using three-dimensional (3D) conformal RT (n = 60) or intensity-modulated radiotherapy (IMRT)/ stereotactic body radiotherapy (SBRT) (n = 12). The 2-year FFP and OS rates among this cohort were 84.4% and 63.6%, respectively. The 2-year OS rates for platinum-sensitive disease vs platinum-resistant disease were 65.2% and 35.7%, respectively (P = .02). Predictors of decreased FFP included a diagnosis of primary peritoneal carcinoma (P = .02) and time to RT < 12 months (P = .03). RT was well tolerated, with the majority of patients reporting no acute effects (56%) or developing mild symptoms, such as grade 1 gastrointestinal or urinary toxicity (19%). Late toxicity consisting of small bowel obstruction occurred in seven patients with other risk factors, including previous surgical resections; five (71%) of these patients had received whole-abdominopelvic radiation. For patients with metastatic ovarian cancer treated with palliative intent, the median dose of RT was 30 Gy (range: 10–50.4 Gy), delivered using 3D conformal RT (n = 44) and IMRT/stereotactic radiosurgery (SRS) (n = 17). 62 Symptom palliation was achieved in 96% of cases. Reports of toxicity were low for this cohort of patients, with no grade ≥ 3 toxicity. The 2-year OS for patients treated with palliative intent was 53.8%. CONCLUSIONS: The use of RT for primary or recurrent ovarian cancer is well tolerated, with durable rates of FFP. RT with palliative intent achieved high rates of symptom palliation without much additional toxicity. RT should be considered for patients with advanced ovarian cancer. (P081) SBRT Boost as a Substitute for Brachytherapy in the Definitive Treatment of Gynecologic Malignancies With Radiotherapy Neilayan Sen, MD, Jessica Zhou, MD, Ryan Jozwiak, Yixiang Liao, PhD, Krystyna Kiel, MD; Rush University Medical Center PURPOSE: Delivery of high doses of radiation within a prescribed period of time is associated with local control when treating primary or recurrent gynecologic cancers. Occasionally, patients can not undergo a brachytherapy boost. We report our experience with alternative stereotactic body radiation treatment (SBRT) boost. MATERIALS AND METHODS: From 2012 to 2014, a total of 8 patients with locally advanced squamous cell carcinoma of the cervix (2 patients) and recurrent endometrial cancer in the vaginal cuff (6 patients) received an SBRT boost after pelvic external beam (EB) radiotherapy (range: 45–50 Gy). One patient received SBRT after EB and 2 high-dose-rate (HDR) brachytherapy fractions. Patients either refused brachytherapy or were high-risk (by medical comorbidities) for brachytherapy. Patients were immobilized using a CIVCO body frame (CIVCO Medical Solutions, Coralville, IA) with abdominal compression. Vaginal and fiducial markers were used to localize tumor at simulation and treatment. Doses typically used for brachytherapy were prescribed to D90 of the planning target volume (PTV) (0–5-mm expansion on clinical target volume [CTV] excluding the rectum when not involved by the tumor). Dose was limited by organ-at-risk tolerances. The Eclipse planning system (Varian, Palo Alto, CA) was used to generate RapidArc plans with 6-MV photons. Treatment was delivered using a True Beam STx linear accelerator (Varian, Palo Alto, CA). Daily conebeam computed tomographies (CTs) were performed using the rectum, bladder, visible tumor, and markers for image guidance, and ExacTrac (BrainLab, Germany) was used to ensure precision of delivery. Tumor status and toxicities were recorded at regular follow-up intervals; toxicity was graded according to Common Terminology Criteria for Adverse Events version 4.0 (CTCAE v4.0). RESULTS: Dose/fractionation schemes were 7 Gy × 2 for one patient (after two HDR brachytherapy fractions), 6 Gy × 5 for six patients, and 5.8 Gy × 5 for one patient. Cumulative equalized total dose in 2 Gy/fraction (EQD2Gy) to D90 of the target volume ranged from 74.6 Gy to 84.3 Gy (mean 81.3 Gy). Mean D2cc rectum and bladder doses were 66.3 Gy (range: 59.4–75.8 Gy) and 77.3 Gy (range: 69.6– 83.6 Gy), respectively. Mean overlap between the rectum as contoured on daily cone-beam CTs and the PTV was 0.29 cc (range: 0.00–1.42 cc). There were no local recurrences at a mean follow-up of 14.5 months. One patient developed distant metastases at 7 American Radium Society Scientific Papers and Posters 2015 months. One patient developed grade 3 vaginal fibrosis. No grade 4 toxicities were observed. CONCLUSIONS: With respect to local control and toxicity, SBRT may be a reasonable method of boosting patients who cannot safely undergo brachytherapy boosts for locally advanced gynecologic malignancies. (P082) Outcomes of Definitive Radiotherapy for T2N0 Squamous Cell Carcinoma of the Glottis: A SingleInstitution Retrospective Study Bassem Y. Youssef, MD, Abdallah S. Mohamed, MD, MSc, Blaine D. Smith, G. Brandon Gunn, MD, Jack Phan, MD, PhD, William H. Morrison, MD, Adam S. Garden, MD, David I. Rosenthal, MD, Clifton D. Fuller, MD, PhD; UT MD Anderson Cancer Center BACKGROUND: The aim of this study is to report the oncologic and functional outcomes of patients with T2N0M0 squamous cell carcinoma (SCC) of the glottis treated with radiation therapy (RT). METHODS: Sequential patients treated with definitive RT for T2 glottic SCC at our facility between 2000 and 2013 were retrospectively reviewed under an approved institutional review board (IRB) protocol. Demographics, disease stage, and treatment characteristics were extracted. Local control and survival outcomes at 2 and 5 years of follow-up were calculated. Both univariate analysis and multivariate Cox proportional hazards assessment were performed to investigate the correlation of patient- and treatment-related factors with disease control and survival endpoints. RESULTS: A total of 68 patients were included in the analysis. The median follow-up was 55 months (range: 5–173 mo); 58 (85%) of the patients were male, and the median age was 63 years (range: 18–88 yr). Of the 68 patients, 20 (29%) were treated using intensitymodulated radiotherapy (IMRT), and the remainder was treated using conventional three-dimensional conformal radiotherapy (3DCRT). Further, 18 patients (26%) were treated using altered fractions schemes, while the rest were treated conventionally. The local control (LC) and locoregional control (LRC) rates at 2 and 5 years were 89% and 84%, and 87% and 82%, respectively. The rate of ultimate LRC was 100%, with no single second recurrence after successful surgical salvage. The disease-specific survival (DSS) was 95% at both 2 and 5 years, with a corresponding OS of 92% and 88%, respectively. In both univariate and multivariate analyses, none of the examined variables (age, pathologic grade, radiation technique, fractionation scheme, total dose, and chemotherapy) was associated with local recurrence or mortality, except for older age, which was associated with worse OS only in the univariate analysis (P < .0001).Regarding functional outcome, only one patient was feeding tube–dependent at the 1 year follow-up due to persistence of grade 3 dysphagia; 60% of patients had no hoarseness of voice at the last follow-up assessment. CONCLUSIONS: Excellent 5-year oncologic and functional outcomes were achieved for patients presented in the study. The use of altered fractionation, concurrent chemotherapy, and different radiation techniques did not show any significant differences in outcomes. However, the reduced radiation dose delivered to the carotid arteries using IMRT suggests that it is potentially advantageous for reduction of long-term vascular toxicity and is therefore recommended as the treatment of choice. (P083) Concurrent Chemotherapy + IMRT in Locally Advanced Squamous Cell Carcinoma of Head and Neck: What Is the Appropriate Chemotherapy? Jeanann L. Suggs, MD, PhD, Shankar P. Giri, MD, Madhava Kanakamedala, MD; University of Mississippi Medical Center PURPOSE: To evaluate patterns of failure and survival among patients with locally advanced squamous cell carcinoma of the head and neck treated definitively with cetuximab, low-dose cisplatin, or high-dose cisplatin together with intensity-modulated radiation therapy (IMRT). MATERIALS AND METHODS: A total of 158 patients treated between 2005 and 2010 were reviewed retrospectively. All had biopsy-proven squamous cell carcinoma of the head and neck and were treated with IMRT with concurrent chemotherapy. Low-dose weekly cisplatin, high-dose cisplatin every 3 weeks, or weekly cetuximab was utilized for each patient. Local treatment failure, distant failure, and median survival were analyzed using Fisher’s exact test. RESULTS: Among 158 evaluated patients, 66 (41.8%) were treated with cetuximab, 57 (36.1%) were treated with low-dose cisplatin, and 32 (20.3%) were treated with high-dose cisplatin. Median age at diagnosis was 53.5 years. Primary tumor locations included the oropharynx (32%), larynx (32%), oral cavity (8%), paranasal spine (5%), hypopharynx (4%), nasopharynx (4%), and other (32%). Median radiation dose was 70 Gy (range: 60–70 Gy). Locoregional failure was not statistically different between the cetuximab, low-dose cisplatin, and high-dose cisplatin groups (15.15%, 7.01%, and 12.5%, respectively; P = .377). Distant failure was not statistically different between the cetuximab, low-dose cisplatin, and high-dose cisplatin groups (10.6%, 17.54%, and 15.63%, respectively; P = .525). Median survival for cetuximab, low-dose cisplatin, and high-dose cisplatin was 18 months, 19 months, and 22 months, respectively (P = .384). CONCLUSIONS: Cetuximab may be considered as an alternative to cisplatin with concurrent RT, particularly for patients with locally advanced head and neck squamous cell carcinoma who are not candidates for platinum therapy. These results indicate no difference in patterns of local or distant failure between cetuximab, lowdose weekly cisplatin, or q3 weekly high-dose cisplatin in this patient population. (P084) Long-Term Functional and Oncologic Outcomes of Esthesioneuroblastoma David Zaenger, MD, Bryan M. Rabatic, PhD, MD, Joseph M. Kaminski, MD, Waleed F. Mourad, MD, PhD; Georgia Regents University ARS PROCEEDINGS 2015 63 American Radium Society Scientific Papers and Posters 2015 PURPOSE: We report our multimodality functional and oncologic outcomes in the management of esthesioneuroblastoma (ENB). MATERIALS: This is a single-institution retrospective study of 22 patients treated between 1998 and 2014 for ENB. The median age was 47 years (range: 29–68 yr); 13 patients (59%) were male. The percentages of patients in Kadish stages A, B, and C were 45%, 32%, and 23%, respectively. Two patients (9%) received definitive induction chemotherapy, followed by chemoradiation therapy (CRT) (70 Gy), and 20 patients (91%) received postoperative RT (PORT) to a median dose of 59.4 Gy (range: 54–63 Gy). Intensity-modulated radiation therapy (IMRT) was utilized in 64% of the whole cohort. PORT fields included the preoperative tumor bed and elective bilateral upper neck lymph node (LN) levels IB and II. RESULTS: The median follow-up for the whole cohort was 40 months (range: 12–150 mo). The actuarial 4-year disease-free survival (DFS) was 68%. The 4-year actuarial overall survival (OS) was 86%. The 4-year actuarial locoregional control (LRC) was 68%, and distant control (DC) was 95%. Of the seven failures, two developed infield local failure and underwent successful salvage surgery and SRS boost, three patients failed locoregionally (primary site and neck), one failed regionally (neck alone), and one developed simultaneous local and distant failure. Recurrences manifested as late as 112 months. No adverse events related to vision were reported. A total of 10% developed short-term grade 3 dysphagia, without long-term percutaneous endoscopic gastrostomy (PEG) tube dependency. CONCLUSIONS: ENB is a rare malignancy, with optimal management remaining uncertain. Our experience suggests that aggressive management seems successful in providing sustained LRC with acceptable long-term toxicity. The role of elective nodal irradiation to the upper neck remains unclear. (P085) Oncologic and Functional Outcomes of Salivary Gland Tumors (SGTs) With Pathologically Proven Perineural Invasion (PNI) 13% of cases were parotid, minor salivary, and submandibular gland primaries, respectively. Pathologically adenoid cystic carcinoma and adenocarcinoma were the most common histologies (39% and 24%, respectively). RESULTS: The median age was 53 years (range: 20–78 yr). Males and Caucasians made up 50% and 50% of the population, respectively. With a median follow-up of 5 years (range: 1–10 yr), the 5-year actuarial disease-free survival (DFS), locoregional control (LRC), and distant control (DC) rates were 85%, 94%, and 91%, respectively. The median time to LR failure (LRF) and distant metastases (DMs) was 8 months (range: 6–16 mo) and 30 months (range: 18–60 mo), respectively. Late RT toxicity was grade ≤ 2 xerostomia (20%), altered taste (15%), trismus (6%), dysphagia (4%), and neck stiffness (2%). CONCLUSIONS: A multimodality approach to SGTs with PNI provides excellent LRC. Further studies are warranted to identify the patients at higher risk of LRF and DMs to optimize their management. (P086) Proton vs Photon/Electron-Based Therapy in the Treatment of Pediatric Salivary Gland Tumors: A Comparison of Dosimetric Data and Acute Toxicities Stephen R. Grant, BS, David R. Grosshans, MD, PhD, Anita Mahajan, MD; Baylor College of Medicine; Department of Radiation Oncology, UT MD Anderson Cancer PURPOSE: Adjuvant radiotherapy (RT) is frequently indicated for high-risk salivary gland tumors. For pediatric patients, minimizing radiation to surrounding normal tissues is of particular importance. We retrospectively compared clinical outcomes and toxicity profiles for pediatric patients with parotid and submandibular tumors treated with either adjuvant proton- or photon/electronbased RT. PURPOSE: To report the clinical outcomes and patterns of failure of salivary gland tumors (SGTs) with perineural invasion (PNI) status postmultimodality approach (ie, surgery, postoperative radiation therapy [RT], and chemotherapy). MATERIALS AND METHODS: We retrospectively identified all patients aged ≤ 18 years treated with radiation for salivary gland tumors between 1996 and 2014 at our institution. Demographic, disease-control, and survival data were recorded, and dosimetric data were compiled. Toxicities were scored according to National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events version 4.0 (CTCAE 4.0) criteria. Statistical analyses were performed using GraphPad Prism, with two-sided t-tests determined to be significant when P < .05. MATERIALS: This is a single-institution retrospective study that was fully approved by our institutional review board (IRB). From March 1997–2010, a total of 62 patients with SGTs underwent the multimodality approach due to the presence of PNI ± other highrisk features (recurrent tumor, R1 or R2 resection, positive margins, multiple positive lymph nodes, and extracapsular extension). Intensity-modulated RT (IMRT) was utilized in 50% of the patients. RT fields included the tumor bed, ipsilateral neck, and occasionally the contralateral neck. The most common pathologically involved nerves (facial, V2, and V3) were included in the RT fields to the skull base. The median RT dose delivered was 63 Gy (range: 50–70 Gy at 1.8–2 Gy/fraction). A total of 53%, 34%, and RESULTS: A total of 24 pediatric patients treated with adjuvant RT were identified (20 parotid, 4 submandibular). Histologies included mucoepidermoid carcinoma (12), adenoid cystic carcinoma (5), adenocarcinoma (2), acinic cell carcinoma (2), myoepithelioma (1), undifferentiated carcinoma (1), and pleomorphic adenoma (1). A total of 13 patients received proton therapy, and 11 received photon/electron-based therapy. The mean prescribed dose was 60 Gy in each cohort (range: 54–66 Gy). No acute grade 3 dermatitis or mucositis was seen in the proton cohort vs 27% and 18% in the photon/electron cohort, respectively. Significantly greater weight loss was reported in those receiving photon/electron radiotherapy (median 5.2% vs 0%), with one patient requiring Bryan M. Rabatic, PhD, MD, David Zaenger, MD, Joseph M. Kaminski, MD, Waleed F. Mourad, MD, PhD; Georgia Regents University 64 American Radium Society Scientific Papers and Posters 2015 placement of a feeding tube (compared with none in the proton cohort). With a mean follow-up of 49 months (range: 2 mo–18 yr), no disease recurrence or deaths were observed in either cohort. Compared with photon/electron-based therapy, proton therapy was associated with significantly lower mean doses to the spinal cord (0.2 Gy vs 19 Gy), pituitary gland (0.0 Gy vs 5.7 Gy), optic nerves (0.0 Gy vs 3 Gy), oral cavity (4.6 Gy vs 19 Gy), thyroid gland (1.5 Gy vs 24 Gy), larynx (11 Gy vs 39 Gy), and contralateral structures, including the eye (0.0 Gy vs 1.4 Gy), mandible (0.0 Gy vs 7.7 Gy), and parotid (0.0 Gy vs 5.2 Gy) and submandibular glands (0.0 Gy vs 8.1 Gy) (all P < .05). to G1. There were no other G3 events. G0, 1, 2, and 3 dysphagia rates were 87%, 4%, 9%, and 0% at baseline and 48%, 26%, 26%, and 0% at the end of RT, respectively. Among patients without G2 dysphagia at baseline, there was no difference between the rates of G0/1 and G2+ dysphagia at the end of RT compared with baseline (P = .104). Similar results were noted for xerostomia, dysgeusia, and weight loss. G2 mucositis occurred in 35% of patients—a significant change from baseline (P = .046)—but 43% of patients had no mucositis upon completing RT. The median weight at the end of RT was 97.4% of baseline. Only six patients (23%) lost > 5% of their baseline weight, and one patient (4%) lost > 10%. CONCLUSION: This report of adjuvant RT for pediatric salivary gland tumors is one of the largest to date and the only one to document outcomes following proton therapy. Compared with conventional photon/electron-based therapy, proton therapy significantly reduced doses to multiple normal tissues. Moreover, clinically, no grade 3 toxicities were observed in the proton group vs 45% in the photon/electron cohort. Continued follow-up is required to determine long-term outcomes. CONCLUSIONS: Proton therapy for parotid cancers results in very low rates of GI toxicity. Nutritional status and weight were preserved throughout therapy. These results should be validated with patient-reported outcomes. (P087) Proton Therapy Results in Low Rates of Acute GI Toxicity for Parotid Cancers Roi Dagan, MD, MS, Curtis Bryant, MD, MPH, Christopher G. Morris, MS, Daniel J. Indelicato, Julie A. Bradley, MD, William M. Mendenhall, MD; University of Florida BACKGROUND: Acute gastrointestinal (GI) toxicity from headand-neck radiotherapy (RT) results in decreased quality of life. Increasingly, conformal RT has improved toxicity outcomes. The target for parotid tumors is ideal for proton therapy, because distal to the Bragg peak, there is no exit of radiation to the oral cavity, swallowing musculature, or contralateral major salivary glands. PURPOSE: To evaluate acute GI toxicity from proton therapy for parotid cancers. METHODS: A total of 23 patients were treated with primary (n =7) or adjuvant (n = 16) proton therapy for cancers of the parotid region, including salivary gland carcinoma (n = 16), skin carcinoma with perineural invasion (n = 5), and Ewing sarcoma (n = 1). The most common indication for proton therapy was skull base invasion. Double-scattered conformal proton therapy, typically involving three fields, was delivered to a median dose of 70 Gy (relative biological effectiveness [RBE]) (range: 55,8–74.4 Gy[RBE]) using various fractionations (range: 1.8–2 Gy[RBE]) once daily, 5 fractions/week (n = 15); 1.2 Gy(RBE) twice daily, 10 fractions/week (n = 7); and 2 Gy(RBE) once daily, 6 fractions/week (n = 1). Seven patients received concurrent chemotherapy. Acute toxicities were prospectively evaluated weekly during radiotherapy using Common Terminology Criteria for Adverse Events version 3 (CTCAE v3). RESULTS: Grade (G) 3 dysphagia occurred during 1 week of therapy in one patient with baseline G2 dysphagia before RT after undergoing parotidectomy, mandible resection, and full-mouth extractions. A percutaneous gastrostomy tube was placed before RT and removed upon completing RT when dysphagia improved (P088) Nonadherence to NCCN Guidelines Negatively Impacts Survival in Early-Stage Squamous Cell Carcinoma of the Larynx Chi-Hsiung Wang, PhD, Erik Liederbach, BS, Arif Shaikh, MD, Mihir K. Bhayani, MD; Center for Biomedical Research Informatics, Department of Surgery, Department of Radiation Oncology, Department of Head and Neck Surgery, NorthShore University HealthSystem INTRODUCTION: Primary treatment for early-stage squamous cell carcinoma of the larynx (LSCC) is radiation therapy (RT) alone or larynx-preserving surgery per National Comprehensive Cancer Network (NCCN) guidelines. The guidelines do not recommend the addition of chemotherapy in this setting. We investigated the National Cancer Data Base (NCDB) to determine adherence to these guidelines and its effects on overall survival (OS). METHODS: Utilizing the NCDB, we selected 34,461 stage I–IV LSCC patients treated with RT from 1998 to 2011. Patients who received primary surgery were excluded. Chi-square tests, logistic regression models, and Cox proportional hazards models were utilized for analyses. RESULTS: In the cohort, there were 19,610 (56.9%) early-stage (stage I/II) LSCCs and 14,851 (43.1%) late-stage (stage III/IV) LSCCs; 1,634 (8.5%) early-stage LSCCs received chemotherapy (CRT), as did 11,468 (77.9%) of the late-stage patients, over the entire time period studied. For early-stage cancers, the proportion of patients receiving CRT increased from 3.3% in 1998 to 9.4% in 2011 (P < .001). For late-stage cancers, the proportion of patients receiving CRT increased from 55.2% in 1998 to 87.8% in 2011 (P < .001). In a multivariate model, patients with early-stage LSCC were significantly more likely to receive chemotherapy if patients were younger (aged < 50 yr), female, African American or Hispanic, treated at a community-based hospital, and living in the Middle Atlantic (NJ, NY, and PA) or East North Central (IL, IN MI, OH, and WI) regions of the United States. Survival analyses were restricted to 1998–2006, and patients had a median follow-up of 4.4 years (range: 1–13 yr). Early-stage LSCC patients receiving RT alone had superior 5-year unadjusted OS rates compared with CRT patients (64.3% vs 52.7%; P < .001). These results persisted after adjusting for patient, facility, and tumor characteristics; and ARS PROCEEDINGS 2015 65 American Radium Society Scientific Papers and Posters 2015 patients with CRT had worse OS (hazard ratio [HR] = 1.18; 95% confidence interval [CI], 1.07–1.31; P < .001). Late-stage LSCC patients receiving CRT had superior 5-year unadjusted OS rates compared with patients treated with RT alone (42.6% vs 25.9%; P < .001), and after similar risk adjustment, patients with CRT had superior OS (HR = 0.69; 95% CI, 0.65–0.73; P < .001) compared with those who received RT alone. CONCLUSION: NCCN guidelines were not followed in 9.4% of patients with early-stage LSCC in 2011, because chemotherapy was added to their RT-based regimen. In these patients, the addition of chemotherapy reduced patient survival by 18%. These results indicate the importance of adherence to NCCN guidelines when prescribing RT-based regimens in the early-stage LSCC setting. (P089) p16 as a Complementary Prognostic Marker for EBV-Positive Nasopharyngeal Carcinoma Wen Jiang, MD, PhD, Paul Chamberlain, Erich Sturgis, MD, Adam S. Garden, MD, Jack Phan, MD, PhD; UT MD Anderson Cancer Center BACKGROUND: p16, also known as cyclin-dependent kinase inhibitor 2A, is a tumor suppressor protein that hinders cell cycle progression by inhibiting the formation of cyclin-dependent kinase complexes. Overexpression of p16 is associated with improved outcomes in oropharyngeal carcinoma patients. However, its role in nasopharyngeal cancer pathogenesis remains incompletely understood. Here, we set out to determine whether p16 overexpression status acts as a prognostic marker among nasopharynx cancer patients. METHODS: We retrospectively identified 86 patients with nasopharyngeal carcinoma who were treated at MD Anderson from 2000 to 2014. Epstein-Barr virus (EBV) status was determined by in situ hybridization, while p16 expression was analyzed by immunohistochemistry. All statistical analyses were performed using the IBM Statistical Package for the Social Sciences (SPSS) v22.0 software package. RESULTS: A total of 44 patient samples were EBV-positive, and 40 were p16-positive. EBV positivity was associated with improved survival (44.9 vs 95.0 mo; P < .004), progression-free survival (PFS) (28.1 vs 80.4 mo; P < .013) and time to locoregional failure (expressed as duration of locoregional control [LRC]) (65.5 vs 104.4 mo; P < .043). Although p16 overexpression showed trends toward improved PFS and LRC, they failed to reach statistical significance. Overexpression of p16, however, was found to be a significant predictor for improved PFS (27.1 vs 106.3 mo; P < .02) and LRC (64.5 vs 93.6 mo; P < .02) in EBV-positive nasopharyngeal cancer patients using both univariate and multivariate analysis. CONCLUSION: Our results suggest that p16 overexpression can act as a marker for PFS and LRC in EBV-positive nasopharyngeal carcinoma patients. This interesting finding raises the possibility of further stratifying more aggressive phenotypes within EBVpositive tumors. Therefore, molecular testing of p16 expression may complement EBV status to provide more detailed and comprehensive guidance in determining the prognosis and predicting 66 treatment outcomes for patients diagnosed with nasopharyngeal carcinoma. (P090) Pediatric Metastatic Odontogenic Ghost Cell Carcinoma: A Multimodal Treatment Approach Safia K. Ahmed, Masayo Watanabe, MD, Daphne E. deMello, MD, Thomas B. Daniels; Department of Radiation Oncology, Mayo Clinic; Phoenix Children’s Hospital for Cancer and Blood Disorders; Department of Pathology and Laboratory Medicine, Phoenix Children’s Hospital INTRODUCTION: Odontogenic ghost cell carcinoma (OGCC) is an extremely rare tumor, wherein the optimal management remains uncertain. We report the first pediatric metastatic OGCC case treated with multimodal therapy: surgery, adjuvant chemoradiation, and adjuvant immunotherapy. CASE DESCRIPTION: A 10-year-old Hispanic male presented with facial swelling. A right maxillary mass was appreciated on exam. Imaging demonstrated a 3.3-cm soft tissue mass in the right maxilla with destruction of adjacent bone and displacement of molars. Pathology confirmed OGCC. Presurgical imaging obtained 5 weeks later noted growth of the mass to 6.7 cm, with right submandibular and posterior cervical adenopathy. Surgery entailed right-sided modified radical maxillectomy, palatectomy, and neck dissection. Multiple re-excisions were performed to obtain negative margins. Seven level 1 and 2 lymph nodes were positive for tumor. Immunohistochemistry was positive for epidermal growth factor receptor (EGFR), indicating cellular expression of EGFR protein. Adjuvant chemoradiation was recommended, the given rapid presurgery growth, adenopathy, and concern for microscopic residual disease. A dose of 44 Gy in 22 fractions was administered to the postoperative bed and bilateral neck via intensitymodulated radiation therapy (IMRT) prior to treatment break for toxicity. Reimaging then demonstrated new 1.7-cm left gingivolabial disease. Radiation plans were redesigned, with the postoperative bed treated to total doses of 60 Gy/30 fractions and the left gingivolabial disease treated to 39 Gy/13 fractions. Treatmentrelated toxicity included dehydration, dermatitis, and malnutrition. Monthly cetuximab was initiated, and the patient remains disease-free at 14 months. DISCUSSION: Adjuvant treatment was recommended, given the metastatic disease, rapid growth, and concern for microscopic residual disease. As adjuvant chemoradiation and immunotherapy are associated with improved outcomes in primary head and neck cancer, a similar approach was adopted. Multimodal therapy is thought to be unsuitable in OGCC due to the poor outcomes documented in seven patients treated with radiation and/or chemotherapy in the current literature. These conclusions are difficult to draw, given the small number of cases. Moreover, these modalities were utilized over 15 years ago. As significant advances have been made with multimodal therapy since then, further investigation is warranted, especially since this approach was tolerated in our patient and since he remains disease-free. CONCLUSION: OGCC can exhibit aggressive progression, warranting investigation into multimodal therapy. Given that adjuvant chemoradiation and immunotherapy are associated with improved American Radium Society Scientific Papers and Posters 2015 outcomes in primary head and neck cancer, a similar application in OGCC may help guide optimal treatment. This approach was well tolerated in our pediatric patient, and he remains disease-free at 14 months. (P091) Sarcopenia/Cachexia Is Associated With Reduced Survival and Locoregional Control in Head and Neck Cancer Patients Receiving Radiotherapy: Results From Quantitative Imaging Analysis of Lean Body Mass Sasikarn Chamchod, MD, Clifton D. Fuller, MD, PhD, Aaron J. Grossberg, MD, PhD, Abdallah S. Mohamed, MD, MSc, Jolien Heukelom, MD, Hillary Eichelberger, BA, BS, Michael E. Kantor, BS, Gary B. Gunn, MD, Adam S. Garden, MD, Steven J. Frank, MD, Jack Phan, MD, PhD, Beth M. Beadle, MD, PhD, Heath D. Skinner, MD, PhD, William H. Morrison, MD, Debra A. Ruzensky, RD, David I. Rosenthal, MD; Radiation Oncology Unit, Chulabhorn Hospital; Department of Radiation Oncology, UT MD Anderson Cancer Center; Department of Radiation Oncology, Netherlands Cancer Institute; UT Medical School, Houston BACKGROUND: Major weight loss before or during head and neck squamous cell cancer (HNSCC) treatment is common. We investigated the impact of weight loss, cachexia, and sarcopenia—the isolated loss of lean body mass—as determined by a novel method using routine staging positron emission tomography-computed tomography (PET-CT) scans on cancer treatment outcomes. METHODS: Biometric data were collected on consecutive patients with American Joint Committee on Cancer (AJCC) stage IVA–IVB HNSCC treated with radiation therapy (RT) with or without concurrent chemotherapy between 2003 and 2013 and who had paired pre- and posttreatment PET-CTs. Cachexia was defined as > 5% weight loss for < 6 months or body mass index (BMI) < 20 kg/m2 with 2% weight loss. Sarcopenia was defined by CT-measured L3 skeletal muscle index of < 52.4 cm2/m2 for men and < 38.5 cm2/ m2 for women, as described by Prado et al (2013). We evaluated the effect of pre- and post-RT sarcopenia on outcomes. Survival curves were constructed using the Kaplan-Meier technique. Log-rank test was used to compare outcomes. Univariate and multivariate overall survival (OS) modeling was performed using parametric survival fitting to allow intramodel comparison using corrected Bayesian information criteria (BIC). RESULTS: A total of 175 patients were identified, and the median follow-up was 67.9 months. We detected sarcopenia in 65 patients (37.1%) prior to RT. Sarcopenia was identified in an additional 47 of 110 patients (42.7%) on the post-RT scan. All patients who developed sarcopenia on the post-RT study had decreased locoregional control (LRC), OS, and disease-specific survival (DSS) (P < .05 for all). Sarcopenia that was identified on the pre-RT PET-CT showed the same patterns but only for nonoperated patients, where the 5-year OS was decreased from 66.7 ± 1.0% to 17.8 ± 10.2% (P < .001). Posttreatment sarcopenia was more substantive in competing multivariate models of mortality risk than simple weight and BMI-based cachexia metrics (ΔBIC ≥ 12.9). DISCUSSION: We confirm that analysis of routine CT scans for sarcopenia can predict outcomes for HNSCC patients. Pre-RT sarcopenia is associated with more poor outcomes for nonoper ated patients and post-RT sarcopenia for all HNSCC patients. Post-RT sarcopenia, as measured by routine CT, outperformed simple weight loss and BMI-derived cachexia metrics, because loss of lean muscle mass can occur independently of BMI status. These findings suggest the potential benefit for investigating intervention with aggressive nutritional and physiatric methods to prevent sarcopenia during RT and to study how these inter ventions might affect outcomes in nonoperated, pre-RT sarcopenia patients. (P092) Definitive Chemoradiotherapy or Radiotherapy for Unresectable, Very Locally Advanced, or Medically Inoperable Paranasal Sinus and Nasal Cavity Cancer Lindsay M. Burt, Ying Hitchcock; Huntsman Cancer Institute, University of Utah PURPOSE: To review radiotherapy (RT) technique and outcomes for definitive RT or chemoradiotherapy (CRT) for unresectable, very locally advanced, or medically inoperable paranasal sinus (PNS) or nasal cavity (NC) cancer at a single institution. METHODS: Between 1998 and 2010, there were 11 patients with unresectable, very locally advanced, or medically inoperable PNS/ NC cancers treated with definitive CRT (7) or RT alone (4) at the University of Utah. There were 10 males and 1 female, with a mean age of 57.3 years (range: 28–75 yr). CRT was given to stage IVA (5) and stage IVB (2) patients, and RT alone was given to stage II (1), stage III (1), and stage IVA (2) patients. One patient was treated with a three-dimensional (3D) conformal technique, and 10 were treated with intensity-modulated RT (IMRT). The median dose was 70.2 Gy (range: 70–72.4 Gy). In order to adapt to tumor shrinkage and prevent critical structures from receiving a high dose due to tumor regression and weight loss, most patients underwent two treatment planning phases, with a resimulation at a dose of 45–50 Gy. One patient received a boost of 11 Gy using stereotactic radiosurgery (SRS) following 70.2 Gy external beam RT (EBRT). The most common chemotherapy agent was cisplatin at 40 mg/m2, given on a weekly basis for 6–7 cycles. A flexible nasal endoscopy with a biopsy was performed 3 months posttreatment to evaluate tumor response. RESULTS: There was a median follow-up of 39 months (range: 1–70 mo). Six patients underwent a debulking surgery. Overall, six (55%) patients remained disease-free, two (18%) developed local recurrences, one (9%) developed regional recurrence, and 2 (18%) developed distant metastasis. Local control was seen in 9 of 11 (81.8%) patients. The treatment was well tolerated, with only one patient experiencing a grade 3 late toxicity (trismus), two patients experiencing grade 2 late toxicities (cataract, retinal detachment, and trismus), and five patients experiencing grade 1 late toxicities (dysgeusia, dry mouth, fibrosis, skin telangiectasia, and nasal congestion). Three patients had no late toxicities. CONCLUSION: Definitive CRT or RT is feasible for unresectable, very locally advanced, or medically inoperable PNS/NC cancer with minimal late toxicity. Local disease control is encouraging, with acceptable treatment-related complications, when treating with the described two-phase IMRT treatment technique. ARS PROCEEDINGS 2015 67 American Radium Society Scientific Papers and Posters 2015 (P093) Cesium-131 Brachytherapy in High-Risk and Recurrent Head and Neck (HN) Cancers: Long-Term Results of a Pilot Study patients with sinonasal neuroendocrine tumors who were treated with a combination of surgery, radiation therapy, and/or chemotherapy at the Mayo Clinic in Arizona. Bhupesh Parashar, Anthony Pham, Dattatreyudu Nori, A. Gabriella Wernicke; Weill Cornell Medical Center MATERIALS AND METHODS: An institutional review board (IRB)approved retrospective study of 27 patients treated at the Mayo Clinic from January 25, 1996 through November 6, 2013. Of these patients, 9 (33%) had olfactory neuroblastoma (ONB), 11 (41%) had sinonasal undifferentiated carcinoma (SNUC), 5 (19%) had small-cell carcinoma, and 2 (7%) had neuroendocrine carcinoma. BACKGROUND: The feasibility and efficacy of re-irradiation using contemporary radiation techniques to treat recurrent head and neck cancer have been demonstrated, but the role of brachytherapy is unclear. Here, we describe the use of cesium (Cs)-131 brachytherapy with concurrent salvage surgery in 18 patients. MATERIALS AND METHODS: Eligible patients underwent maximal gross resection of the tumor with implantation of Cs-131 brachytherapy seeds, delivering a minimum dose of 80 Gy to the tumor bed. Rates of overall survival (OS), locoregional progression-free survival (LRPFS), disease-free survival (DFS), and radiationinduced toxicity were analyzed. RESULTS: Between 2010 and 2013, a total of 18 patients with 20 implants were enrolled and treated with surgical resection and brachytherapy for the management of locoregional recurrences of head and neck cancer. The majority of histology was squamous cell carcinoma (10 of 18). All but one patient had a history of prior radiation in the area of tumor recurrence. One patient had gross residual disease following surgical resection. Two patients underwent an additional surgical resection and brachytherapy implantation 2 months and 5 months later for local recurrence that developed outside of the treatment volume. A total of 13 patients had previous locoregional recurrence treated with surgical salvage therapy. The total dose following initial definite external beam radiation therapy (RT) ranged from 5,000 cGy to 7,000 cGy. Two patients developed grade 3 toxicity; no grade 4 or 5 complications were observed. With a median follow-up of 38 months (range: 1–44 mo), 11 patients developed another recurrence or progression of head and neck cancer. In 6 of these 10 cases, the failure was locoregional, and in 4 patients, it was isolated distant failure. One patient was found to have simultaneous locoregional and distant progression of disease. The median OS was 15 months, and median DFS was 11 months. The 6-, 12-, and 18-month OS rates in this study were 77%, 71%, and 45%, respectively. The 6-, 12-, and 18-month LRPFS rates in this study in patients were 69%, 62%, and 52%, respectively. The 6-, 12-, and 18-month DFS rates in this study were 57%, 45%, and 37%, respectively. CONCLUSION: Compared with prior literature, our study shows comparable rates of survival with a decreased rate of radiationinduced toxicity. (P094) Treatment Outcomes of Sinonasal Neuroendocrine Cancer: A Retrospective Review Priya Parikh, BA, Samir H. Patel, MD, Mauricio Gamez, MD, William Wong, MD, Michele Halyard, MD, Kelly Curtis, MD, Devyani Lal, MD; Department of Radiation Oncology, Department of Medical Oncology, Department of Otorhinolaryngology, Mayo Clinic Arizona PURPOSE/OBJECTIVES: To report the treatment outcomes of 68 Of the 27 patients, 22 patients received a combination of surgery, radiation therapy, and/or chemotherapy: 9 (41%) underwent surgery plus adjuvant radiation, 8 (36%) underwent surgery plus adjuvant chemoradiation, 4 (18%) were treated with chemoradiation, and 1 (5%) had neoadjuvant chemotherapy followed by surgery and adjuvant radiation. Two patients were treated with surgery alone, one was treated with radiation therapy alone, and two were treated with chemotherapy alone. The median dose of radiation was 60 Gy (range: 42–66 Gy). RESULTS: Median follow-up was 16.3 months (range: 3–315 mo). The 5-year actuarial overall survival (OS), distant metastasis–free survival (DMFS), and local control (LC) rates were 44%, 62%, and 51%, respectively. When comparing outcomes of ONB vs nonONB histologies, 5-year OS was 56% vs 38% (P = .4), and 5-year LC was 67% vs 37%, respectively (P = .6). Specifically, SNUC patients had the worst 5-year OS of 25% and 5-year LC of 39%. Moreover, 50% of SNUC patients developed distant failure within 1 year. There were five regional recurrences in the neck—all in ONB patients. One of these five patients had a regional-only recurrence and did not receive elective radiation to the neck. CONCLUSIONS: Sinonasal neuroendocrine tumors are a challenging group of diseases to manage, due to their rarity and heterogeneous natural histories. New multimodality strategies need to be explored to potentially enhance outcomes, especially in non-ONB histologies. (P096) Late Radiation-Associated Dysphagia (Late-RAD) With Lower Cranial Neuropathy After Oropharyngeal IMRT Katherine A. Hutcheson, PhD, Denise A. Barringer, MS, CCCSLP, G. Brandon Gunn, MD, Stephen Y. Lai, MD, Merrill S. Kies, MD, David L. Schwartz, MD, Jan S. Lewin, PhD; UT MD Anderson Cancer Center; UT Southwestern Medical Center PURPOSE: Late radiation-associated dysphagia (late-RAD) is a debilitating, delayed toxicity of nonsurgical organ preservation for head and neck cancers. Herein, we examine late-RAD in long-term oropharyngeal cancer survivors after intensity-modulated radiation therapy (IMRT). METHODS: A pooled dataset was analyzed from two institutional single-arm clinical trials of nonsurgical organ preservation with radiation. Prospective functional analysis included radiographic swallow studies (videofluoroscopy) and questionnaires pretreatment and 6, 12, and 24 months after treatment. Functional recovery American Radium Society Scientific Papers and Posters 2015 at 1 year was assessed by index measure: solid food diet, feeding tube–free, functional airway protection, and no lower cranial neuropathy or stricture. A diagnostic algorithm was developed to classify late-RAD events based on significant dysphagia on late (≥ 2 yr) videofluoroscopic swallow referrals among patients with early functional recovery from acute toxicities of IMRT by 1 year. RESULTS: A total of 57 oropharyngeal cancer survivors with a minimum 2-year disease-free clinical follow-up after IMRT (range: 66–72 Gy) with systemic therapy were included. T-stage distribution was: TX (4), T1 (13), T2 (21), T3 (15), and T4 (4). Also, 52 patients (91%) achieved functional recovery by 1 year. At a median follow-up of 5 years, the cumulative incidence of late-RAD was 9% (5 of 57), 3 of whom had delayed lower cranial neuropathy (XII or X). An additional five patients were considered at risk for late-RAD per self-report of progressive dysphagia symptoms ≥ 2 years after treatment without videofluoroscopic confirmation. CONCLUSION: Although the majority of oropharyngeal cancer survivors enjoy functional recovery in early survivorship, almost 10% suffer severe, late deterioration of swallowing abilities. Delayed lower cranial neuropathies often precipitate late-RAD. (P097) Tumor Density, Size, and Histology in the Outcome of Stereotactic Body Radiation Therapy for Early-Stage Non–Small-Cell Lung Cancer: A Single-Institution Experience Jason C. Ye, MD, Jenghwa Chang, PhD, ZhiQiu Li, PhD, A. Gabriella Wernicke, MD, MSc, Dattatreyudu Nori, MD, Bhupesh Parashar, MD; Stich Radiation Oncology, Weill Cornell Medical College; Vantage Oncology PURPOSE: Stereotactic body radiation therapy (SBRT) for early- stage non–small-cell lung cancer (NSCLC) has been shown to have excellent local control (LC). NSCLC has diverse sizes and histologies, which can affect treatment response. It also has variable densities, which may affect radiation dose accumulation within the target to therapeutic dose. This study represents a single-institution outcome of lung SBRT with respect to various factors. METHODS: A retrospective chart review was conducted of all localized NSCLCs, with no lymph node metastasis or distant metastasis (DM), treated in 2001–2014. Patient and disease characteristics and treatment outcomes (eg, primary tumor control [PTC], intralobar recurrence–free survival [LRFS], disease-free survival [DFS], overall survival [OS], and toxicity) were examined. Further treatment planning details, including dose, fractionation, and gross tumor volume/internal target volume (GTV/ITV) densities, were obtained in the treatment planning system. RESULTS: A total of 96 patients with 109 lesions who were treated had follow-up information available and were included. There were 42 (44%) men and 54 (56%) women, with a median age at time of diagnosis of 77 years (range: 52–95 yr). A total of 46% of patients had documented chronic obstructive pulmonary disease (COPD), and 66% had at least one prior malignancy (51% had prior NSCLC; range: 0–5). Further, 58% of patients could not undergo surgery due to pulmonary reserve, 17% could not undergo surgery due to other comorbidities, and 22% chose SBRT, even though they were deemed surgical candidates. A total of 80% of patients were stage IA, and median tumor size was 1.7 cm (range: 0.6–6.9 cm); 75% had adenocarcinoma (AC), and 17% had squamous cell carcinomas (SCC). The most commonly used dose and fractionation was 48 Gy in four fractions over 2 weeks (77%). There were seven (6%) grade 1 (four skin, two pain, one cough) and two (2%) grade 2 skin toxicities. With a median follow-up of 25 months, the 2-year OS was 84%, and DFS was 74%, while PTC was 91% (LRFS: 85%). Patients with AC and SCC had similar PTC (P = .939) and DFS rates (P = .909). Compared with stage IA, patients with stage IB disease had similar PTC (P = .462) but were more likely to develop DM (P = .003). The mean GTV/ITV density varied from 0.236– 1.010 g/cm3 (mean: 0.697 g/cm3; standard deviation [SD] = 0.182). Density did not have a statistically significant effect on PTC, but those with denser GTV/ITV values (> 0.7 g/cm3) had inferior DFS (P = .027). CONCLUSIONS: Our institutional experience confirmed that SBRT to primary NSCLC is well tolerated and provides excellent LC, regardless of tumor size or histology. Tumor density did not appear to have a significant effect on PTC, but denser tumors were more likely to have poorer outcomes, likely owing to associated larger tumor burden. (P098) Dosimetric Predictors of Pulmonary Toxicity in Patients With Malignant Pleural Mesothelioma Receiving Radiation Therapy to the Intact Lungs Yaseen Zia, MD, Vishruta Dumane, PhD, Kenneth Rosenzweig; Icahn School of Medicine at Mount Sinai INTRODUCTION: Pleurectomy and decortication is gaining wider acceptance as a standard surgery for malignant pleural mesothelioma. Adjuvant radiation is particularly challenging, due to the need to treat the pleura while sparing the underlying intact lung. Typical dosimetric constraints are of limited value in this patient population. This study aimed to establish whether there was a functional subunit of lung that needed to be spared that was predictive of development of radiation pneumonitis in patients with malignant pleural mesothelioma receiving radiation therapy to the intact lungs. MATERIALS AND METHODS: A total of 18 patients with malignant pleural mesothelioma (MPM) were treated with definitive or adjuvant hemithoracic pleural intensity-modulated radiation therapy (IMRT) after radical pleurectomy and decortification between 2010 and 2014. The dose was prescribed to the planning target volume (PTV), which included the entire parietal and visceral pleura, with a median dose of 4,500 cGy (range: 3,000–5,940 cGy). Specific dosimetric parameters pertaining to both lungs were recorded. Toxicity was assessed during weekly on-treatment visits and in follow-up examinations using Common Terminology Criteria for Adverse Events version 4.0 (CTCAE v4.0) criteria. RESULTS: The median follow-up was 4 months (range: 1–18 mo). Six patients (33%) developed grade 3 pneumonitis or higher, and 12 patients (67%) did not. Mean lung dose and V30 of the total lungs were found to be statistically significant for the development ARS PROCEEDINGS 2015 69 American Radium Society Scientific Papers and Posters 2015 of pneumonitis. Contralateral lung V5 and mean lung dose were found not to correlate with the development of pneumonitis, in contrast to previously published reports. We found that for each 1,000 cc of ipsilateral lung, if 150 cc was treated to less than 20 Gy, there was significantly less pneumonitis. No patients with 150 cc spared per 1,000 cc of normal lung developed pneumonitis vs 43% (6 of 14) of patients who failed to reach that threshold (P < .05). CONCLUSIONS: Sparing the ipsilateral lung of at least a finite functional unit per 1 L of the ipsilateral lung is a predictor of the development of radiation pneumonitis. This represents a new dosimetric measure in plan evaluation and correlates significantly with the development of toxicity in patients with malignant pleural mesothelioma receiving radiation to the ipsilateral lung. This factor might be a more effective and useful parameter in these challenging cases. (P099) Influence of Surveillance PET/CT on Detection of Early Recurrence Following Definitive Radiation in Stage III Non–Small-Cell Lung Cancer Jay Reddy, MD, PhD, Chad Tang, MD, Zhongxing Liao, MD, Daniel Gomez, MD; UT MD Anderson Cancer Center PURPOSE/OBJECTIVES: There are little data to support the use of varying imaging modalities in evaluating recurrence in non–smallcell lung cancer (NSCLC). We compared the efficacy of surveillance positron emission tomography–computed tomography (PET-CT) vs CT scan of the chest in detecting recurrences following definitive radiation for NSCLC. MATERIALS AND METHODS: We retrospectively analyzed 189 patients treated between 2000 and 2011 who met the inclusion criteria of biopsy-proven stage III NSCLC and completion of definitive radiation treatment (dose ≥ 60 Gy). These patients were then grouped based on the ratio of PET-CT scans: CT scans of the chest during the postradiation surveillance window, defined as 2–18 months posttreatment. A ratio of 0 described the CT-only group, and a ratio of > 1 described the PET-high group. We compared survival times from the end of treatment to the date of death or last follow-up utilizing log-rank tests. Multivariate analysis was conducted to identify factors associated with decreased survival. RESULTS: In the entire cohort, median event-free survival (EFS) was 8.7 months, and median overall survival (OS) was 29.9 months. The CT-only group had a median EFS of 9 months vs 8.7 months for the PET-high group (P = .85). There was no difference in OS between the CT-only and PET-high groups (median OS: 35.3 mo and 29.3 mo, respectively; P = .66). There was also no difference in local recurrence-free survival or distant metastases-free survival between the CT-only and PET-high groups (P = .06 and P = .32, respectively). Similarly, on multivariate analysis, stratification into the PET-high group was not associated with improved EFS (hazard ratio [HR] = 0.978; 95% confidence interval [CI], 0.688–1.391; P = .902) or OS (HR = 1.032; 95% CI, 0.723–1.472; P = .864). CONCLUSIONS: In stage III NSCLC patients treated with definitive radiation, increased frequency of PET-CT scan surveillance did not result in decreased time to detection of locoregional or distant 70 recurrence or improved survival. If validated, further investigation is warranted to elucidate the benefit, if any, of NSCLC posttreatment surveillance with PET-CT scan. (P100) Intensity-Modulated Radiation Therapy After Extrapleural Pneumonectomy With and Without Chemotherapy for Malignant Pleural Mesothelioma: No Fatal Pulmonary Toxicity and Long-Term Survival Sameer Jhavar, J. Pruszynski, Y. Liu, A. Gowan, P. Rascoe, N. Thawani, N. Deb, Mehul Patel; Baylor Scott and White Hospital, Texas A&M Medical Sciences Center PURPOSE: Intensity-modulated radiation therapy (IMRT) after extrapleural pneumonectomy (EPP) for malignant pleural mesothelioma (MPM) has been associated with fatal pulmonary toxicity. A single-institution experience with IMRT following EPP for MPM is reported. METHODS: Between 2005 and 2014, a total of 18 patients with MPM were treated with EPP followed by hemithoracic IMRT. IMRT target volume was the entire hemithorax and the thoracotomy and chest tube incision sites. Patients were treated with a median dose of 4,500 cGy in 25 fractions. Kaplan-Meier curves were used to graphically asses the overall survival (OS) and relapse-free survival (RFS). Median survival times are reported for both OS and RFS. RESULTS: Of the 18 patients analyzed, 17 were males, and 11 had right-sided tumors. Median age was 54 years (range: 40–76 yr). The most common histology was epithelioid type. Chemotherapy was neoadjuvant in four and adjuvant in seven patients. A total of 3, 12, and 3 patients had pathological American Joint Committee on Cancer (AJCC) stages II, III, and IV, respectively. Involvement of surgical margin, lymphovascular space, pericardium, and chest wall was seen in 9, 7, 12, and 3 patients, respectively. The highest and lowest mean lung dose (MLD) was 9.3 Gy and 5 Gy, respectively, with a mean of 7.14 Gy. The highest V20 (normal lung volume receiving ≤ 20 Gy) was 7%, and the mean V20 was 2.23% (range: 0%–7%). At a median follow-up of 3 years, 8 patients were alive and 10 patients were dead. Ten (55%) patients experienced disease recurrence. The median RFS and OS were 29.67 months (95% confidence interval [CI], 11.79–78.1 mo) and 38.2 months (95% CI, 17.4–78.1 mo), respectively. No grade 3 acute toxicities were seen. No grade 3 or fatal pulmonary toxicities have been reported. CONCLUSION: In our patient population, strict adherence to lung dose constraints during IMRT resulted in improved outcomes without fatal toxicity. (P101) Does Maximum SUV From F-18 PET Scan Predict Outcomes for Early-Stage Non–Small-Cell Lung Cancer Treated With Stereotactic Body Radiotherapy (SBRT)? Corey J. Hobbs, MD, Stephen J. Ko, MD, Nitesh N. Paryani, MD, Michael G. Heckman, MS, Nancy N. Diehl, BS, Jennifer L. Peterson, MD, Katherine S. Tzou, MD, Robert C. Miller, MD, Laura A. Vallow, MD, Steven J. Buskirk, MD; Mayo Clinic Florida American Radium Society Scientific Papers and Posters 2015 PURPOSE: Prior studies have shown mixed results regarding the predictive value of positron emission tomography (PET) scan following definitive treatment of early-stage lung cancers. The primary outcome of this study was to evaluate the association of maximum standard uptake value (SUVmax) on PET scan with recurrence and survival in patients with lung cancer who were treated with stereotactic body radiotherapy (SBRT). Secondary outcomes were to evaluate associations of baseline patient characteristics with SUVmax, recurrence, and survival. MATERIALS AND METHODS: A total of 61 lung cancer patients with a documented pretreatment SUVmax and treated with SBRT between 2008 and 2014 were included in this study. Baseline information was collected regarding age, gender, lesion size, SUVmax, glucose at time of PET, T stage, histology, nodal evaluation, and total radiation dose. Examined recurrences included local, nodal, ipsilateral lung, contralateral lung, and distant metastases. Time to death and cause of death were also recorded. Median clinical follow-up was 12.7 months (range: 2.8–52.9 mo). The median SUVmax of the cohort was 6.4. A total of 54 patients (89%) had biopsy-proven malignancy. RESULTS: Baseline characteristics between patients with a low (< 6.4) and high (≥ 6.4) SUVmax were similar, except for lesion size in the high-SUVmax group (median: 2.2 cm vs 1.7 cm; P < .001) and higher T stage (1b or 2a: 71.0% vs 36.7%; P = .018). Adenocarcinomas were more common in the low-SUVmax group (76.0% vs 41.1%), while squamous cell carcinomas were less common (8.0% vs 44.8%). A total of 16 patients (26.2%) experienced any recurrence; 9 patients (14.8%) experienced nodal recurrence, 3 patients (4.9%) experienced same lung recurrence, and 5 patients (8.2%) experienced distant recurrence. Also, 18 patients (29.5%) died during follow-up. There was no evidence of a difference in recurrence outcomes between the two groups. An evaluation of associations of baseline patient characteristics with the three most common outcomes of nodal or same lung recurrence, any recurrence, and death was performed. The only baseline variable that was significantly associated with any of these outcomes was serum glucose at time of PET; patients with a glucose value higher than the median of 103 mg/dL had an increased risk of nodal or same lung recurrence (relative risk [RR] = 6.77; P = .017) and any recurrence (RR = 3.76; P = .027). CONCLUSION: We did not find an association of SUVmax with recurrences or death; however, the sample size was relatively small, and the power to detect differences was low. These findings will be further evaluated in a larger multicenter study in the future. (P102) Surgery Improves Survival in 14,228 Patients With Malignant Pleural Mesothelioma Andrea S. Wolf, MD, MPH, Emanuela Taioli, MD, Marlene Camacho-Rivera, ScD, MPH, Kenneth E. Rosenzweig, MD, Raja M. Flores, MD; Mount Sinai Medical Center; North Shore/Long Island Jewish Health System, Hofstra School of Medicine OBJECTIVES: Left untreated, malignant pleural mesothelioma (MPM) has uniformly poor prognosis. Prolonged survival has been reported with surgery-based multimodality therapy, but to date, no trial has demonstrated independent survival benefit of surgery over other therapies for MPM. We evaluated whether cancer-directed surgery independently influenced survival in a large populationbased dataset. METHODS: The Surveillance, Epidemiology, and End Results (SEER) database was explored from 1973 to 2009 to identify all cases of pathologically proven MPM. Age, sex, race, diagnosis year, stage, cancer-directed surgery, radiation, and vital status were analyzed (chemotherapy data not available). The association between prognostic factors and survival was estimated using a Cox proportional hazards model. RESULTS: There were 14,228 patients with pathologically proven MPM. On multivariable analysis, female gender, younger age, early stage, and cancer-directed surgery were independent predictors of longer survival. In comparison with no treatment, surgery alone was independently associated with significantly longer survival, with an adjusted hazard ratio (aHR) for mortality of 0.65 (0.62–0.68), while radiation alone was not (aHR = 1.17 [1.10– 1.25]). The combination of surgery and radiation was associated with a survival outcome similar that with to surgery alone (aHR = 0.69 [0.63–0.75]). In patients diagnosed from 2000–2009, the aHR for mortality with radiation was 1.26, 0.68 for surgery, and 0.63 for surgery plus radiation, with similar results obtained in patients diagnosed from 1973–1999. This suggests that improvements in technique over time have not altered the impact of therapy on patients with MPM. CONCLUSIONS: Despite developments in surgery, perioperative management, and radiotherapy, the prognosis for MPM patients has not improved over the past four decades. In this SEER study of 14,228 patients over 36 years, cancer-directed surgery was associated with better survival in MPM, independent of other prognostic factors. These data support the role of surgery-based therapy as the cornerstone of treatment for this challenging disease. (P103) A New Score Predicts Survival in Patients With Non–Small-Cell Lung Cancer Steven E. Schild, MD, Angelina D. Tan, BS, BA, Jason A. Wampfler, BS, Julian R. Molina, MD, PhD, Helen J. Ross, MD, Ping Yang, MD, PhD, Jeff A. Sloan, PhD; Mayo Clinic PURPOSE: This analysis was performed to create a scoring system to estimate the survival of patients with non–small-cell lung cancer (NSCLC). METHODS: Data from 1,274 NSCLC patients were analyzed to create and validate a scoring system. Univariate (UV) and multivariate (MV) Cox models were used to evaluate the prognostic importance of each baseline factor. Prognostic factors that were significant on both univariate and multivariate analyses were used to develop the score. They included quality of life, age, performance status, primary tumor diameter, nodal status, distant metastases, and smoking cessation. The score for each factor was determined by dividing the ARS PROCEEDINGS 2015 71 American Radium Society Scientific Papers and Posters 2015 5-year survival rate (%) by 10 and summing these scores to form a total score. Multivariate models and the score were validated using bootstrapping with 1,000 iterations from the original samples. RESULTS: The score for each factor ranged from 1–7 points, with higher scores reflective of better survival. Total scores of 32–37 correlated with a 5-year survival of 8.3% (95% confidence interval [CI], 0%–17.1%), 38–43 correlated with a 5-year survival of 20% (95% CI, 13%–27%), 44–47 correlated with a 5-year survival of 48.3% (95% CI, 41.5%–55.2%), 48–49 correlated to a 5-year survival of 72.1% (95% CI, 65.6%–78.6%), and 50–52 correlated to a 5-year survival of 84.7% (95% CI, 79.6%–89.8%). The bootstrap method confirmed the reliability of the score. CONCLUSIONS: Prognostic factors that were significantly associated with survival on both UV and MV analyses were used to construct a valid scoring system that can be used to predict survival of NSCLC patients. The score can be used for trial stratification or for choosing patients specifically for high-risk trials. Optimally, this score will be helpful when counseling patients and designing future trials. (P104) Esophagus- and Contralateral Lung–Sparing IMRT for Locally Advanced Lung Cancer in the Community Hospital Setting Johnny Kao, MD, Jeffrey Pettit, MS, Shanata S. Ramsaran, Terry Palatt, MD; Good Samaritan Hospital Medical Center INTRODUCTION: The optimal technique for performing lung intensity-modulated radiation therapy (IMRT) remains poorly defined. Due to concerns regarding acute and late toxicity, the potential benefit of dose escalation beyond 60 Gy has not been established. We hypothesize that improved dose distributions associated with normal tissue-sparing IMRT can allow for safe dose escalation that will translate into decreased acute and late toxicity. MATERIALS AND METHODS: We performed a retrospective analysis of 82 consecutive patients with stage II/III or stage IV lung cancer with a single distant metastasis (median age: 69 yr, 53% male, 21% small-cell lung cancer, 83% white, 70% Eastern Cooperative Oncology Group [ECOG] performance status score 0 or 1, 13% stage IV, and 87% receiving concurrent chemotherapy) treated from January 2010 to September 2014. From January 2010 to April 2012 (cohort A), patients were treated with the community standard of predominantly three-dimensional conformal radiotherapy (76%) without specific esophagus or lung constraints. From May 2012 to September 2014 (cohort B), patients were treated with predominantly IMRT (95%) while selectively sparing uninvolved lung and esophagus. The study endpoints were dosimetry, toxicity, and overall survival (OS). RESULTS: Despite higher mean prescribed radiation doses in cohort B (64.5 Gy ± standard deviation [SD] 5.0 vs 60.8 Gy ± SD 6.2; P = .04), patients in cohort B had significantly lower lung V20, V10, V5, mean lung, maximum esophagus, and mean esophagus doses (P ≤ .001). Mean lung V20 was 23.3 Gy ± SD 7.2 in cohort B vs 32.2 Gy ± SD 11.6 for cohort A. Mean esophagus dose was 20.2 Gy ± SD 10.2 in cohort B vs 34.3 Gy ± SD 12.7 for cohort A (P = 72 .001). Patients in cohort B had reduced acute grade ≥ 3 esophagitis (0% vs 11%; P < .001) and late grade ≥ 2 pneumonitis (5% vs 21%; P = .01). The incidence of hospitalization for dehydration and/or pulmonary complaints was 11% for cohort B vs 37% for cohort A (P = .008). Three patients in cohort A who developed grade 5 pneumonitis had lung V20 values of 41%, 48%, and 58% and lung V5 values of 90%, 99%, and 88%, respectively. Median survival in cohort B has not been reached at 24 months vs 13 months for cohort A (P = .13). CONCLUSION: These data provide proof of principle that suboptimal radiation dose distributions are associated with significant acute and late lung and esophageal toxicity that may result in hospitalization or even premature mortality. We propose a relatively simple four-field IMRT technique with strict attention to commonly accepted lung and esophageal dose-volume constraints as a preferred approach for the majority of locally advanced lung cancers. (P105) Community-Based Early-Stage Treatment (BEST) Outcomes for NSCLC Austin N. Arnone, BS, Christopher Biggs, MD, PhD, Daniel Reed, DO, Terry Lee, MD, Cheri Pantoja, CCRP, Kevin Rogers, MS; Arizona Center for Cancer Care PURPOSE/OBJECTIVES: Stereotactic body radiation therapy (SBRT) has been established as an effective treatment for earlystage non–small-cell lung cancer (NSCLC) in patients who are medically inoperable or refuse surgery. This retrospective study analyzes outcomes of lung SBRT in the setting of a communitybased cancer treatment center in Arizona. MATERIALS AND METHODS: Between 2008 and 2013, a total of 146 tumor sites in 126 patients with stage I NSCLC were definitively treated using SBRT administered via the Varian RapidArc, a continuous dynamic IMRT system with cone-beam computed tomography (CT) localization. Simulation was gated, and treatment targeted a static volume. Patients were determined to be medically inoperable (n = 96) or refused surgery (n = 30). Treatment was delivered using a risk (tumor site, size)-adapted dose of 60 Gy (73 sites), 55 Gy (4 sites), or 50 Gy (52 sites) or a lower dose (17) in five fractions. Treatment outcomes were interpreted using KaplanMeier analysis of overall survival (OS), local control (LC), and disease-free survival (DFS). Correlation of patient, tumor, and treatment factors with survivability was analyzed using Cox proportional hazards. RESULTS: Median age was 78 years; 44 patients were male, and 82 were female. The 3-year OS for all patients was 76%, with a median follow-up of 20 months (range: 2–67 mo). Median OS was 57 months. The 3-year LC for the entire cohort was 90.5%, and 3-year DFS was 68%. The 10 sites that developed local recurrence all received less than 60 Gy, yielding a 3-year LC of 81.7% for this subgroup (n = 65). Currently, LC is 100% for sites receiving 60 Gy (n = 73). Multivariate analysis demonstrated that no patient, tumor, or treatment factor was significantly predictive of survival. There was a 7.1% rate of grade 3 toxicity by Radiation Therapy Oncology Group (RTOG) criteria, with no grade 4 or 5 toxicities American Radium Society Scientific Papers and Posters 2015 within 90 days of treatment. CONCLUSIONS: LC, survival, and toxicity of stage I NSCLC treated with SBRT in this community setting are comparable to those reported in university and multi-institutional trials. The efficacy, safety, and convenience of SBRT have been translated to a large cohort of patients in an outpatient community cancer center. These results also indicate that doses < 60 Gy delivered in five fractions may be less effective at achieving LC. As lower doses are examined in central lesions, it will be important to closely evaluate any possible reduction in LC. (P106) Dosimetric Comparisons of Treatment by Different Radiotherapy Techniques for Stage III Non–Small-Cell Lung Cancer Veronica Finnegan, MD, Kiernan May, CMD, Jeffrey Bogart, MD, Paul Aridgides, MD, Varun Chowdhry, MD, Seung Hahn, MD; SUNY Upstate; Massachusetts General Hospital PURPOSE: Standard dose for stage III non–small-cell lung cancer was established 30 years ago with Radiation Therapy Oncology Group (RTOG) 7301, where patients treated to 60 Gy had improved outcomes. Phase I and II studies have shown that higher doses are safe but have not been investigated in a phase III trial until recently. Optimal radiation dose was investigated in RTOG 0617, demonstrating that 60 Gy (standard) was superior to 74 Gy (high dose) with chemoradiotherapy ± cetuximab. We hypothesize that this is due to increased normal tissue toxicity. We performed a comparative dosimetric study between different radiation techniques. MATERIALS AND METHODS: Review of 10 patients treated accord- ing to RTOG 0617. Dosimetric differences to normal organs were compared using three-dimensional (3D), intensity-modulated radiation therapy (IMRT), and volumetric-modulated arc therapy (VMAT). RESULTS: A paired-difference test (two-sample t-test) for various dose parameters was performed. For plans to 74 Gy, there were statistically significant differences for lung V20 (P = .016), mean lung (P = .015), and mean esophagus (P = .046) between 3D and IMRT. There were statistically significant differences for the spinal cord (P = .38), mean lung (P = .021), mean esophagus (P = .001), esophagus V60 (P = .001), heart V60 (P = .031), heart V45 (P = .022), and heart V40 (P = .011) between 3D and VMAT. Comparing IMRT and VMAT, there was a statistically significant difference for mean esophagus dose (P = .013). These results were similar for plans to 60 Gy. There were statistically significant differences for lung V20 between 3D and IMRT. There were statistically significant differences for the spinal cord, mean esophagus, and heart V45 between 3D and VMAT. CONCLUSION: For both the 60-Gy and 74-Gy dose levels, signifi- cant dosimetric advantages (normal tissue sparing) were observed with IMRT, and a larger benefit was observed for 74-Gy dose plans. The dose to the esophagus, heart, and lung are likely to be clinically meaningful in terms of toxicity. VMAT provided benefit over conventional IMRT at the 60-Gy dose level. Additional studies are warranted to further investigate the impact of advanced radiotherapy techniques for the delivery of high-dose thoracic radiotherapy. (P107) Vero SBRT Treatment of Moving Lung Tumors Charles A. Kunos, MD, PhD, John P. Shanahan, MS; Summa Cancer Institute PURPOSE: Intrafraction motion management of moving lung tumors during stereotactic ablative body radiotherapy (SBRT) remains challenging. We investigated whether 40%-thresholded 2-[18F] fluoro-2-deoxy-D-glucose (18F-FDG) positron emission tomography (PET) images enhanced intrafraction tracking capabilities of the Vero SBRT platform. MATERIALS AND METHODS: Three patients with six total moving lung tumors had gross tumor volume (GTV) contours created on free-breathing, end-inspiration, and end-expiration breath-hold computed tomography (CT) images. A thresholded 40% maximum standard uptake value 18F-FDG PET contour was generated. The four individual contours contributed to a single image-guided internal target volume (ITV). Vero SBRT treatment planning was conducted on the free-breathing CT dataset, with the ITV expanded all around by 5 mm for a final planning target volume (PTV). The prescription for each lesion was 40 Gy in four fractions of 10 Gy given every other day. Non-coplanar radiation doses were determined using a Monte Carlo algorithm. T-test statistics were calculated to compare parameter means. RESULTS: The PTV was 23 cm3 on average (standard deviation [SD] = 18 cm3) without the 18F-FDG PET contour and 36 cm3 on average (SD = 32 cm3) with the 18F-FDG PET contour (P = .41). The lung V20 Gy was 2.6% on average (SD = 0.9%) without the 18FFDG PET contour and 2.8% on average (SD = 0.5%) with the 18FFDG PET contour (P = .68). Vero SBRT gimbal pan-and-tilt range of 4 cm was not exceeded by the addition of the 18F-FDG PET contour to the ITV. CONCLUSIONS: Overall, 40%-thresholded 18F-FDG PET contours nonsignificantly enlarge PTVs when multiphase free-breathing, inspiration, and expiration breath-hold scans are used for Vero SBRT. Whether 18F-FDG PET contours improve local control of moving lung tumors treated by Vero SBRT needs further study. (P108) Stereotactic Body Radiation Therapy for the Treatment of Large (> 5 cm) Primary Non–Small-Cell Carcinoma Michael C. Roach, MD, Sana Rehman, MD, Dan Mullen, DDS, Jeff D. Bradley, MD, Cliff G. Robinson, MD; Washington University PURPOSE: Patients with inoperable large primary non–small-cell lung carcinoma (NSCLC) present a therapeutic challenge, given the concern for radiation delivered to a large volume of the lung. As such, these patients have so far been excluded from most prospective trials of stereotactic body radiation therapy (SBRT). We evaluated the outcomes of SBRT in the treatment of large primary NSCLC at our institution. ARS PROCEEDINGS 2015 73 American Radium Society Scientific Papers and Posters 2015 MATERIALS AND METHODS: A total of 25 patients with biopsyproven large NSCLC treated with SBRT alone with definitive intent were identified from an institutional review board (IRB)approved prospective thoracic SBRT registry. Tumors were defined as large if they were greater than 5 cm in diameter on computed tomography (CT) (American Joint Committee on Cancer [AJCC] T2b or T3). All had positron emission tomography (PET) scans without evidence of nodal metastasis. Patients were treated to 45–60 Gy in three or five fractions. Patients were reviewed for overall survival (OS), local control (LC), progression-free survival (PFS), and toxicity, with survival and control calculated from completion of therapy using the Kaplan-Meier method. Toxicity was graded according to Common Terminology Criteria for Adverse Events version 4.03 (CTCAE v4.03). RESULTS: Mean follow-up was 21 months (range: 2–83 mo). Median tumor size was 5.5 cm; 23 patients had T2b, and 2 had T3 tumors. Actuarial 2-year OS was 39%, and median survival was 20.1 months. The 2-year PFS was 68%. One patient failed in the mediastinum, two failed locally and in the mediastinum, one failed distantly, and one failed both locally and distantly. Both distant failures occurred in the two patients with T3 tumors. Two patients developed second primaries in different lobes. At 2 years, the actuarial rates of local failure and of distant failure were both 9%. Treatment was well tolerated, with 24% developing any chest wall toxicity (12% grade 1, 8% grade 2, and 4% grade 3). A single patient (4%) developed a rib fracture, and another developed a chest wall ulcer. A single patient (4%) required steroids for pneumonitis. CONCLUSIONS: The use of SBRT for T2bN0 primary NSCLC is a safe, effective, and well-tolerated treatment. (P109) Standard Immunochemotherapy Plus Radiation vs Dose-Intense Chemotherapy With Rituximab in Stage I/II Primary Mediastinal Large B-Cell Lymphoma: A Single-Institution Experience Michael S. Binkley, BA, Susan M. Hiniker, MD, Sharon Wu, MD, Yaso Natkunam, MD, PhD, Erik S. Mittra, MD, PhD, Ranjana H. Advani, MD, Richard T. Hoppe, MD; Stanford University BACKGROUND: Primary mediastinal large B-cell lymphoma (PMBCL) is an uncommon variant of diffuse large B cell lymphoma, representing 2% to 3% of cases of non-Hodgkin lymphoma. Given the limited prospective data and predominantly single-center retrospective reports, the optimal chemotherapy regimen remains undefined, as does the role of radiation therapy (RT) to the mediastinum. We sought to investigate factors associated with outcomes among patients who received rituximab with standard chemotherapy (cyclophosphamide, doxorubicin, vincristine, and prednisone [R-CHOP] or etoposide, doxorubicin, cyclophosphamide, vincristine, prednisone, and bleomycin [R-VACOP-B]) with RT, compared with those who received dose-adjusted etoposide, vincristine, doxorubicin, cyclophosphamide, and rituximab (DA-EPOCH-R). METHODS: We retrospectively analyzed patients with stage I/II PMBCL treated in the rituximab era at our institution. Exclusion 74 criteria included stage III/IV disease and follow-up of less than 3 months. Response to chemotherapy was assessed with interim or postchemotherapy positron emission tomography-computed tomography (PET-CT) using the Deauville score. We used the Kaplan-Meier method to determine freedom from progression (FFP) and overall survival (OS). Patient characteristics were compared using Fisher’s exact test for dichotomous variables and Wilcoxon rank-sum test for continuous variables. RESULTS: A total of 28 patients with stage I/II PMBCL were identified and treated at our institution from 2003–2012. Pretreatment characteristics included age (median 37.5, range: 20–68 yr), female gender (n = 14), performance score (median 1, range: 0–2), international prognostic indicator (IPI) score (median 1, range: 0–2), B symptoms (n = 8), elevated lactate dehydrogenase (LDH) (n = 19), bulk of disease (median 10 cm, range: 4–19.4 cm), and extranodal disease (n = 10). A total of 14 patients received six cycles of R-CHOP or 12 weeks of R-VACOP-B and RT (median 36 Gy, range: 36–45 Gy) and had a median follow-up of 94 months. Following R-CHOP/R-VACOP-B with irradiation, 5-year FFP and OS were both 100%. Also, 14 patients received four to eight cycles DA-EPOCH-R and had a median follow-up of 38 months, with one patient receiving RT (36 Gy) with postchemotherapy PET-CT Deauville score 4. Following DA-EPOCH-R, 3-year FFP and OS were both 100%. Univariate analysis found no significant association between FFP or OS with pretreatment patient characteristics, chemotherapy type, or PET-CT Deauville score ≥ 4. Acute toxicities included neutropenia (absolute neutrophil count [ANC] < 500 cells/mm3, n = 10), hospitalization for neutropenic fever (n = 6), and grade 2 radiation pneumonitis (n = 2). Late effects following RT included hypothyroidism (n = 2). CONCLUSION: Both R-CHOP/R-VACOP-B with RT and DA-EPOCH-R demonstrate excellent outcomes. Long-term follow-up will be required to assess potential complications of contemporary RT. Our data support use of standard immunochemotherapy with RT or DA-EPOCH-R as monotherapy. (P110) Hodgkin Lymphoma and Pregnancy: Treatment Patterns and Survival Outcomes of Women Treated With Modern Chemotherapy and Radiotherapy Eleanor Osborne, MD, Michelle Fanale, MD, Leslie Ballas, MD, Yasuhiro Oki, MD, Andrea Milbourne, MD, Grace Smith, MD, PhD, Sarah Milgrom, MD, Valerie Reed, MD, Isidora Arzu, MD, Bouthaina Dabaja, MD, Chelsea Pinnix, MD, PhD; UT MD Anderson Cancer Center; University of Southern California PURPOSE: Hodgkin lymphoma (HL) is the fourth most common malignancy diagnosed during pregnancy. However, the appropriate management of HL during pregnancy is disputed, and the effect of suboptimal staging imaging and modified treatment regimens is unclear. Here, we report treatment approaches and survival outcomes of pregnancy-associated Hodgkin disease at our institution. METHODS: We performed a single-institution, retrospective analysis of 36 women diagnosed with HL during pregnancy between 1991 and 2014. Kaplan-Meier and chi-square analyses were used to determine survival outcomes. American Radium Society Scientific Papers and Posters 2015 RESULTS: Of the initial 36 charts reviewed, 6 patients were excluded due to inadequate long-term follow-up. Among the 30 remaining patients, 24 patients (80%) had stage I/II disease. Six patients had B symptoms (20%). The median gestational age at diagnosis was 20 weeks (range: 2–37 wk), with most patients (60%) being diagnosed in the second trimester. A total of 19 patients (63%) initiated treatment while pregnant (4 with radiation, 15 with chemotherapy). Two women terminated their pregnancies to initiate treatment, and there were two spontaneous abortions in women who initiated chemotherapy at 4 weeks and 15 weeks of gestation. Further, 18 women (60%) had full-term pregnancies (> 37 wk gestation); the median gestational age at delivery was 37 weeks (range: 26–42 wk). A total of 15 women (50%) received both chemotherapy and radiation, 10 (33%) received chemotherapy alone, and 5 (17%) received radiation alone. The majority of patients had a complete response after finishing therapy, but six women (20%) had progressive or primary refractory disease, and six women (20%) relapsed. After a median follow-up of 57.5 months, the mean progressionfree survival (PFS) was 43.4 months (range: 1–261 mo), and the mean overall survival (OS) was 80.8 months (range: 8–273 mo). Patients with progressive or primary refractory disease had poorer outcomes (mean PFS: 10.3 mo vs 49.3 mo, P < .0001; mean OS: 31.2 mo vs 93 mo, P = .0031). Initiating treatment during pregnancy was not associated with improved outcomes (mean PFS: 13.4 mo vs 46.7 mo, P = .89; mean OS 41.5 mo vs 85.3 mo, P = .81). CONCLUSION: Overall, outcomes in patients diagnosed with HL in pregnancy in the modern era of chemotherapy and radiation are good. As in HL not diagnosed during pregnancy, outcomes are better in patients with a complete response following initial therapy. Delaying all therapy until the postpartum period is appropriate in properly selected patients and is not associated with poorer outcomes. (P111) Single-Isocenter Frameless VolumetricModulated Arc Radiosurgery for Multiple Intracranial Metastases Steven Lau, MD, PhD, Kaveh Zakeri, MD, Xiao Zhao, MD, Ruben Carmona, MAS, Erik Knipprath, Daniel R. Simpson, MD, Sameer K. Nath, MD, Gwe-Ya Kim, PhD, Parag Sanghvi, MD, Jona A. Hattangadi, MD, Clark C. Chen, MD, PhD, Kevin T. Murphy, MD; University of California, San Diego; University of California, Davis; Yale University PURPOSE: Stereotactic radiosurgery is a well-accepted treatment for patients with intracranial metastases, but outcomes with volumetric-modulated arc radiosurgery (VMAR) are poorly described. We report our initial clinical experience applying a novel singleisocenter technique to frameless VMAR for simultaneous treatment of multiple intracranial metastases. METHODS: Between 2009 and 2011, a total of 15 patients underwent frameless VMAR for multiple intracranial metastases using a single, centrally located isocenter. Among them, three patients were treated for progressive or recurrent intracranial disease. A total of 62 metastases (median 3 per patient, range 2–13) were treated to a median dose of 20 Gy (range: 15–30 Gy). Three patients were treated with fractionated SRS. Follow-up, including clinical examination and magnetic resonance imaging (MRI), occurred every 3 months. RESULTS: Median follow-up for all patients was 7.1 months (range: 1.1–24.3 mo), with 11 patients (73.3%) followed until death. For the remaining four patients alive at the time of analysis, median followup was 19.6 months (range: 9.2–24.3 mo). Overall survival (OS) at 6 months was 60.0% (95% confidence interval [CI], 40.3%–88.2%). Local control rates at 6 and 12 months were 91.7% (95% CI, 84.6%– 100.0%) and 81.5% (95% CI, 67.9%–100.0%), respectively. Regional failure was observed in nine patients (60.0%), and seven patients (46.7%) received salvage therapy. Grade ≥ 3 treatment-related toxicity was not observed. Median total treatment time was 7.2 minutes (range: 2.8–13.2 min). CONCLUSIONS: Single-isocenter, frameless VMAR for multiple intracranial metastases can produce clinical outcomes comparable with those of conventional radiosurgery techniques. (P112) Spine Stereotactic Radiosurgery in the Treatment of Metastatic Pheochromocytoma: A Case Series Brian J. Deegan, MD, PhD, Amol J. Ghia, MD, Mary Frances McAleer, MD, PhD, Xin A. Wang, PhD, Paul D. Brown, MD, Jing Li, MD, PhD; UT MD Anderson Cancer Center PURPOSE: Metastasis occurs in approximately 10% of all cases of pheochromocytoma and can cause significant morbidity and mortality. The axial skeleton is the most frequent site of these metastases. Since good systemic therapy options are lacking, local therapy remains the cornerstone of palliation for many of these patients. Due to the poor response of pheochromocytoma to standard fractionated radiotherapy, stereotactic radiosurgery (SRS) is an attractive option to overcome potential radioresistance and provide more durable local control (LC) of these tumors. Here, we report our institutional experience in the treatment of spine metastases from pheochromocytoma with spine SRS (SSRS). METHODS: The available records of patients with metastatic pheochromocytoma treated with SSRS between 2000 and 2014 were retrospectively reviewed. Four patients with nine treated metastatic spinal lesions were identified. Follow-up spine magnetic resonance imaging (MRI) was used to evaluate LC. Pain and symptom data were assessed to evaluate toxicity. Kaplan-Meier method was used to assess LC and overall survival (OS). RESULTS: Median age at time of SSRS was 55 years (range: 46–63 yr). Median follow-up for each treated site was 11 months (range: 4.5–54.6 mo). Treated areas included C-spine (22%), T-spine (33%), L-spine (33%), and sacrum (11%). The most common fractionation scheme was 27 Gy in three fractions (55.6%), followed by 24 Gy in one fraction (22.2%), 16 Gy in one fraction (11.1%), and 18 Gy in three fractions (11.1%). Crude LC rate was 100%, with no local treatment failures observed in the follow-up period. There were two patient deaths in the group (50%). The Kaplan-Meier OS from SSRS at 1 year was 75% and 50% at 2 years. Median time to SSRS from metastatic presentation was 17.6 months (range: 9.2– 72.7 mo). Toxicities included pain (two cases), fatigue (two cases), and vertebral fracture (one case). ARS PROCEEDINGS 2015 75 American Radium Society Scientific Papers and Posters 2015 CONCLUSION: To our knowledge, this work is the first study describing the utility of SSRS in the treatment of metastatic pheochromocytoma. Our data suggest that SSRS is an effective, safe, and durable treatment option. Given the robust tumor control of SSRS and the possibility for metastatic foci to serve as sources for further systemic spread, proactive treatment of spinal metastasis earlier in the disease course may offer therapeutic benefit to these patients. Larger patient numbers and longer follow-up are required to address this issue more fully. (P113) Posttreatment Sequelae for Long-Term Survivors of Brain Metastases Jared R. Robbins, MD, Raphael L. Yechieli, MD, Parag Sevak, MD, M. Salim Siddiqui, MD, PhD; Medical College of Wisconsin; University of Miami; Henry Ford Hospital PURPOSE/OBJECTIVES: Brain metastases typically portend a poor outcome, but long-term survival is possible. The purpose of this study is to evaluate intracranial progression after initial therapy, the need for salvage treatments, and rates of late toxicity in long-term brain metastases survivors. MATERIALS AND METHODS: In this institutional review board (IRB)-approved study, we identified 44 patients with brain metastases surviving > 3 years treated at a single institution from 1987 to 2009. All outcomes were calculated starting at the time of initial brain metastases. Univariate and multivariate logistic regression modeling, as well as Kaplan-Meier and log-rank test methods, was used to characterize outcomes. RESULTS: Median age was 56 years (range: 28–83 yr). Primary sites were lung 52%, breast 23%, melanoma 11.4%, and other 13.6%. Initial treatments were 71% surgery followed by radiosurgery to the cavity (Sx+SRS), 16% whole-brain radiation therapy (WBRT), 9% radiosurgery alone (SRS), and 5% surgery followed by WBRT (Sx+WBRT). The 5-year, 10-year, and 15-year overall survival (OS) was 60%, 34%, and 19%, respectively (range: 3–26.5 yr). Seven failures at the site of initial brain metastasis were observed, with 1-year, 3-year, and 5-year local control of 98%, 93%, and 84%, respectively. Salvage for local failures included Sx (43%), Sx+SRS (29%), SRS (14%), and hospice (14%). New brain metastases developed in 52%, with 1-year, 3-year, and 5-year freedom from new brain metastases rates of 79%, 63%, and 44%, respectively. Initial salvage treatments for new metastases were SRS (45%), WBRT (14%), surgery (14%), Sx+SRS (18%), Sx+WBRT (5%), and hospice (5%). After initial treatment, additional salvage therapies included 20% of patients receiving surgery, 40% receiving SRS, and 25% receiving WBRT. Radiation necrosis (RN) occurred in 27% of patients, with 1-year, 3-year, and 5-year rates of 5%, 19%, and 32%, respectively. The median time to RN was 24 months (range: 7–54 mo), and the median survival after RN was 39 months (range: 2–280 mo). Overall, 41% of patients experienced no intracranial progression, but 22% of them did develop RN. On multivariate analysis, higher initial Karnofsky performance status (KPS) (P = .008; hazard ratio [HR] = 0.943) and freedom from new brain metastases after initial treatment (P ≤ .001; HR = 0.183) were associated with improved survival. 76 CONCLUSIONS: For patients with brain metastases, long-term survival is possible, but the majority of long-term survivors will still experience intracranial progression and require salvage therapies after initial treatment. While salvage therapies may control disease, the development of new metastasis seems to adversely affect survival. RN risk increases over time and may affect a significant portion of long-term survivors. (P114) Radiotherapeutic Care Within the Veterans Health Administration of US Veterans With Metastatic Cancer to the Brain: Supportive Measures (Part 1 of 2 Reports) George A. Dawson, MD, Shruti Jolly, MD, Helen Fosmire, MD, Maria Kelly, MD, Stephen Lutz, MD, Micheal Hagan, MD, PhD, Ruchika Gutt, MD, Drew Moghanaki, MD, MPH, Lori Hoffman-Hogg, MS, RN, CNS, AOCN, Mitchell Ancher, MD, Alice Cheuk, MD; US Veterans Healthcare Administration National Palliative Radiotherapy Task Force Metastatic cancer to the brain is estimated to occur in 170,000 Americans annually. Of them, over 600 cases occur in US veterans. Management of brain metastases is complex. Prognostic scoring criteria and evidence-based guidelines have been developed by societies, including the American Society of Radiation Oncology (ASTRO), to provide guidance in the care of patients with metastatic brain cancer. Patterns of care among Veterans Health Administration radiation oncologists (VHA ROs) in the treatment of brain metastases are not known. METHOD: An electronic survey was sent to all (82) VHA ROs at 38 active VHA radiation oncology centers. Follow-up phone calls were made to nonresponders. The survey inquired about supportive measures, including the number of brain metastases consults seen per year, steroid usage and dosing, use of the Radiotherapy Oncology Group (RTOG)-recursive partitioning analysis (RPA) or diagnosis-specific graded prognostic assessment (GPA) prognostic score, onsite availability of stereotactic radiosurgery (SRS), and demographics. Additional information about VHA ROs regarding their employment status, years in practice, and academic appointments was also obtained. RESULTS: A total of 62 of 82 VHA ROs responded to the questionnaire (76%). Most respondents had academic appointments (70%). Most respondents (79%) received more than 10 consults annually. A total of 89% only used steroids in the setting of neurological changes or edema, while ~10% always used steroids for brain metastases. Prognostic scores for brain metastases were used routinely by only 42% of VHA ROs. The RTOG-RPA classification was used the most (73%). Physicians in practice for less than 5 years (P = .024) and full-time employees (P < .001) were more likely to use prognostic scoring. Also, 69% of surveyed sites referred to another VA or VA-contracted facility for SRS services when required (31% + SRS onsite). The most common dexamethasone dose was 16 mg/day, used by 54% of the respondents. CONCLUSIONS: Veterans with brain metastases treated at VHA American Radium Society Scientific Papers and Posters 2015 radiation oncology centers receive appropriate care. Still, the use of prognostic indices in treatment decision-making is statistically significantly more likely in cases treated by recent training program graduates. Given the relatively recent development of these scoring systems, educational efforts need to be devoted to increasing their use in the clinic. (P115) The Experience of a Radiation Oncology Center in Gauging the Use of Single-Fraction Radiotherapy for Painful Bone Metastases George A. Dawson, MD, Ignat Glushko, MBA, Alice V. Cheuk, MD; Veterans Health Administration BACKGROUND: There is mounting category 1 evidence of the effectiveness of single-fraction radiotherapy (SFRT) in the treatment of uncomplicated, painful bone metastases [Jones, CA Cancer J Clin 2014]. Surveys indicate that physicians in the United States still prefer multiple fractions of RT for uncomplicated bone metastases compared with their Canadian or European colleagues [Popovic et al, Radiother Oncol 2014]. In this report, we used current procedure terminology (CPT) code tracking to determine the scope of use of SFRT for painful malignant bone lesions compared with multiple-fraction RT (MFRT). METHOD: We searched for encounters with an International Classification of Diseases (ICD) code 198.5—Secondary malignancy of bone and/or bone marrow either in primary or secondary positions—and current procedural terminology (CPT) code 77431—Physician weekly visit for 1–2 treatments. Data points below represent 77431 usage only. From 2002 to 2008, usage was as follows: 2002 (0), 2003 (0), 2004 (5), 2005 (1), 2006 (7), 2007 (4), and 2008 (0). From 2009 to October 13, 2014, usage was as follows: 2009 (0), 2010 (0), 2011 (4), 2012 (12), 2013 (10), and 2014 (8). These periods were chosen to reflect a surge in interest in SFRT usage at the end of life, such as the Choosing Wisely Campaign of the American Society of Radiation Oncology in 2013. Data limitations/assumptions: Not all providers code appropriately for both procedures and diagnosis. RESULTS: We noted a 133% increase in the use of SFRT for painful bone lesions at our center from 2002–2008 to 2009–October 13, 2014. On average, CPT code 77431 was used 2.43 times a year for patients with ICD code 198.5 of bone metastases from 2002–2008, and 5.67 times (8.50 times, excluding 2009 and 2010) from 2009– October 13, 2014. Of note, in 2002, 2003, 2008, 2009, and 2010, CPT code 77431 was not used at all. CONCLUSION: We observed an increase in the use of SFRT for bone metastases over the time period covered. Tracking the encounters by ICD codes and CPT codes, when properly coded, served as a useful tool in providing a snapshot view of SFRT usage. Additionally, physician education is a prerequisite for the proper use of a CPT 77431 to capture the true rate of usage of SFRT in clinical practice. (P116) Pain and Radiographic Control After Stereotactic Radiosurgery for Spinal Metastases From Hepatocellular Carcinoma: A Comparison With Other Radioresistant Histologies Todd J. Carpenter, MD, Michael H. Buckstein, MD, PhD, Eddie Zhang, BS, Seth Blacksburg, MD, MBA, Isabelle M. Germano, MD, Sheryl Green, MBBCH; Department of Radiation Oncology, Department of Neurosurgery, Icahn School of Medicine at Mount Sinai; Department of Radiation Oncology, Winthrop University Hospital PURPOSE: We sought to determine the relative radioresistance of hepatocellular carcinoma (HCC) by evaluating the clinical efficacy of stereotactic radiosurgery (SRS) for spinal metastases from HCC compared with outcomes in patients with classically radioresistant histologies in terms of local control (LC) and pain control. MATERIALS AND METHODS: We performed a retrospective review of all patients treated at our institution with spine SRS for metastatic HCC, renal cell carcinoma, melanoma, and sarcoma from January 2007 through May 2014. Radiographic control and patient-reported pain control were analyzed as a function of various patient- and treatment-related parameters. RESULTS: Of 134 lesions in 96 patients treated with spine SRS during the study period, 41 were radioresistant histologies, including 18 HCC, 1 mixed HCC/cholangiocarcinoma, 15 renal cell carcinoma, 6 melanoma, and 1 leiomyosarcoma. Median age was 61 years. Extraosseous disease was present in 63% overall (74% and 55% in HCC and non-HCC patients, respectively; P = NS). Spinal cord compression was present in 29% (32% and 27% in HCC and non-HCC patients, respectively; P = NS), and 24% had decompressive surgery prior to SRS (26% and 23% in HCC and non-HCC patients, respectively; P = NS). Median dose was 18 Gy (range: 14–18 Gy), with no difference between groups. Follow-up imaging was available for 35 patients. With a median follow-up of 6.4 months, actuarial 3-, 6-, and 12-month LC rates for HCC and non-HCC patients were 74%, 65%, and 35% and 94%, 94%, and 85%, respectively (P = .0383). Median time to local failure was 3.5 months for HCC patients and 10.7 months for non-HCC patients. On multivariate analysis (MVA), there was significantly worse LC with HCC histology (P = .0257). Of the 29 patients reporting pretreatment pain, initial pain relief was achieved in 27 (93%); both patients who did not experience initial pain relief had HCC. Actuarial 3-, 6-, and 12-month pain control rates for HCC and non-HCC patients were 73%, 61%, and 23% and 100%, 90%, and 90%, respectively (P = .0405). This interaction remained significant on MVA (P = .0414). CONCLUSIONS: When treated with SRS to the spine, metastatic HCC has worse pain and radiographic control than other highly radioresistant histologies, suggesting that HCC should be included in the category of highly radioresistant tumors. Whether these lesions may benefit from further dose escalation and/or alternate treatment strategies will be the subject of future studies. ARS PROCEEDINGS 2015 77 American Radium Society Scientific Papers and Posters 2015 (P117) Outcomes, Patterns of Failures, and Toxicity for Patients Diagnosed With Pulmonary Metastases Treated With Stereotactic Body Ablative Radiotherapy (SABR) Quynh-Nhu Nguyen, MD, William C. Chance, MD, Peter Balter, PhD, Jim Welsh, MD, Daniel Gomez, MD, Ritsuko Komaki, MD, Zhongxing Liao, MD, Joe Chang, MD, PhD, Reza J. Mehran, MD; UT MD Anderson Cancer Center PURPOSE: To report patterns of failure, outcomes, and toxicity for patients with pulmonary metastases treated with stereotactic body ablative radiotherapy (SABR). METHODS: From 2007 to 2014, a total of 49 patients with 61 pulmonary metastases were irradiated with SABR doses. Primary histology included: lung (n = 14), gastrointestinal (GI) (n = 15), sarcoma (n = 5), head/neck (n = 6), genitourinary (GU) (n = 3), melanoma (n = 3), endometrial (n = 2), and adrenal (n = 1). SABR doses were prescribed by tumor location and proximity to critical structures (central vs peripheral). The majority of metastases received 50 Gy in four fractions (n = 48), 70 Gy in 10 fractions (n = 3), 40 Gy in four fractions (n = 4), 50 Gy in five fractions (n = 1), or 45 Gy–60 Gy in 10 fractions (n = 5). Local failures were defined as recurrence within the high-dose region; locoregional failures included recurrences outside of the high-dose region, within the same lobe, or lymph nodes; and distant failures occurred in a separate lobe or outside of the thorax. RESULTS: The 1-year and 2-year overall survival (OS) rates were 88% and 66%, respectively, with a median OS time of 29.1 months. The 1-year and 2-year progression-free survival (PFS) rates were 64% and 45%, respectively, with a median PFS time of 19.9 months. The 1-year and 2-year freedom from local failure rates were 94% and 84%, respectively. The 1-year and 2-year freedom from locoregional failure rates were 80% and 63%, respectively. The 1-year and 2-year distant metastases–free survival rates were 72% and 51%, respectively, with a median time to new distant metastases of 26.9 months. The toxicities included brachial plexopathy (grade 2, n = 2), radiation pneumonitis (grade 3, n = 1), and chest wall toxicity (grade 2, n = 7). CONCLUSION: SABR is an effective treatment modality for patients with pulmonary metastases, with excellent local control. Further studies are warranted to elucidate which patients with pulmonary metastases would benefit from the local control with SABR and determine when to treat with systemic therapy due to quick progression of distant metastases. (P118) Radiotherapeutic Care Within the Veterans Health Administration of US Veterans With Metastatic Cancer to the Brain: Part 2 Clinical Treatment Patterns Alice V. Cheuk, MD, Ruchika Gutt, MD, Drew Moghanaki, MD, MPH, Michael Hagan, MD, PhD, Stephen Lutz, MD, Shruti Jolly, MD, Helen Fosmire, MD, Maria D. Kelly, MD, Lori Hoffman-Hogg, MS, RN, CNS, AOCN, Mitchell Anscher, MD, George A. Dawson, MD; US Veterans Healthcare Administration National Palliative Radiotherapy Taskforce PURPOSE: Optimal radiation treatment (RT) for brain metastases 78 must be individually tailored. Treatment guidelines by the American Society for Radiation Oncology (ASTRO) were used as the basis for a study of practice patterns among Veterans Health Administration (VHA) radiation oncologists (ROs). Radiotherapeutic interventions for three clinical scenarios were correlated with ASTRO guidelines. METHODS: A survey was sent to all VHA ROs (n = 82). ROs were asked for treatment recommendations, which were analyzed by employment status, academic appointment, and years in practice. SCENARIOS: (1) Uncontrolled non–small-cell lung cancer (NSCLC) with hemiplegia, > 10 brain metastases, Karnofsky performance status (KPS) 30, and life expectancy (LE) < 3 months. (2) Controlled NSCLC with hemiplegia, 4–6 brain metastases, KPS 70, and LE > 4 months. (3) New NSCLC without symptoms, 2 small brain lesions, KPS 90, and LE > 6 months. Treatment options for scenarios 1 and 2 were: supportive care alone, steroids only, whole-brain RT (WBRT), and stereotactic radiosurgery (SRS). Options for scenario 3 were: resection with postoperative WBRT, SRS plus WBRT, SRS alone, and WBRT alone. ROs were asked to provide a WBRT dose fraction scheme for each selected case. RESULTS: The survey response rate was 76%. Scenario 1: The majority (66%) of respondents chose WBRT, with 58% prescribing 3,000 cGy in 10 fractions and 42% delivering 2,000 cGy in 5 fractions. The others (34%) chose no intervention or steroids only. Full-time (FT) employees were more likely to choose WBRT (P < .001) compared with part-time (PT) employees and contractors. Those with an academic appointment and greater number of years in practice (P < .001) also chose WBRT more frequently. Scenario 2: Nearly all (98%) ROs recommended WBRT; 3,000 cGy in 10 fractions was the most common fractionation. There was an association with having an academic appointment, being an FT employee, being in practice less than 5 years, and choosing WBRT (P < .001). Scenario 3: FT employees (64%) and those with academic appointments were more likely to choose SRS alone (P < .001) compared with others. ROs in practice for 6–20 years preferred to do SRS with WBRT (P < .001). CONCLUSIONS: ROs in practice less than 5 years, with academic appointments, or with an FT status had statistically significant associations with WBRT/SRS choice. In poor-prognosis patients, consideration of best supportive care measures is done, and shortcourse RT is often recommended. For patients with good KPS and limited small brain metastases, SRS with or without WBRT is recommended by most practitioners in accordance with ASTRO guidelines. (P119) Electronic Brachytherapy Management of Atypical Fibroxanthoma: Report of Seven Cases Stephen W. Doggett, MD; Aegis Oncology PURPOSE: To evaluate the suitability of treating an uncommon skin malignancy, atypical fibroxanthoma (AFX), with electronic brachytherapy. American Radium Society Scientific Papers and Posters 2015 MATERIALS AND METHODS: From February 2013 to September 2014, we were referred a total of seven cases of AFX, all involving the scalp. All were treated with electronic brachytherapy 50 kV radiation (Xoft Inc., Fremont, California). All lesions received 40 Gy in two fractions per week; 5-mm margins were utilized. RESULTS: As of October 2014, there has been one local recurrence, which was the only lesion that was not debulked surgically prior to electronic brachytherapy. CONCLUSIONS: AFX is a rare skin cancer that is believed to be of mesenchymal origin, likely histiocytic. A less likely possibility is that it is derived from dedifferentiated squamous cell cancer. This is the largest reported series of AFX treated with radiation therapy in the literature. No contraindication to the use of radiation is found in the literature. Prior series all utilized surgery, likely due to the clinically rapid progression of this tumor. Risk of recurrence is mitigated with surgical debulking prior to brachytherapy. Electronic brachytherapy appears to be a safe and effective treatment for AFX. (P120) Pregnancy and Parenthood in Radiation Oncology, Views and Experiences Survey (PROVES): Results of a Blinded Prospective Trainee Parenting and Career Development Assessment Emma B. Holliday, Awad A. Ahmed, Reshma Jagsi, Natalie Clark, Wendy A. Woodward, Clifton D. Fuller, Charles R. Thomas, Jr; UT MD Anderson Cancer Center; UT Southwestern Medical Center; University of Michigan; Oregon Health and Science University Knight Cancer Institute BACKGROUND: Medical school and residency span nearly a decade, during which many students and residents traditionally begin families. As the number of women entering medicine increases, the effects of pregnancy and childbearing on residency experience, research productivity, and career aspirations should be explored. Although men increasingly share childrearing responsibilities in the modern era, recognizing gender differences in time spent in childcare duties and the potential effect on residency experience is paramount to understanding the generational evolution of modern family dynamics for physician-led families. METHODS: An anonymous, voluntary, 102-item online survey was distributed via email to 540 current radiation oncology residents and 2014 graduates. Respondents were asked about demographics, marital and parental status, pregnancy during residency, publication productivity, career aspirations, and experiences working with pregnant co-residents. Respondents with children were additionally asked about childcare arrangements. Women who had been pregnant during residency were further asked questions regarding radiation safety, maternity leave, and breastfeeding experiences. RESULTS: A total of 190 respondents completed the survey—107 (56.3%) men and 84 (43.7%) women—for a 35.2% response rate; 97 respondents (51.1%) were parents, and 84 (44.2%) reported that they or their spouse/partner had become pregnant during residency. Also, 52 women (54.2%) and 31 men (33%) (P = .003) reported delaying starting a family due to residency-related reasons. Respondents with children were more often male (65% vs 47.3%; P = .014), were in a higher level of training (79.3% vs 54.8% of respondents were PGY4 or higher; P = .001), were older (median age 32 yr, interquartile range [IQR]: 31–35 yr vs 30, IQR: 29–33 y; P < .001), had a PhD (33% vs 19.3%; P = .033), were married (99% vs 43%; P < .001), and had a spouse/partner who did not work (24.7% vs 1.9%; P < .001). There was no difference in the number of manuscripts published or expressed likelihood of pursing an academic career by parental status. Among parents, men more often had spouses/partners who did not work (38.1% vs 0%; P < .001), reported that their spouse/partner performed a greater percentage of childcare duties (70%, IQR: 60%–80% vs 35%, IQR: 20%–50%; P < .001), and reported that their spouse/partner was more likely to take care of unexpected childcare duties (74.6% vs 31%; P < .001). Common concerns expressed by women with children included radiation safety, maternity leave, breastfeeding, and effects of pregnancy on clinical training, research experience, and their co-residents’ workload. CONCLUSIONS: Pregnancy and parenthood are common during residency. Women are responsible for more childcare duties than men but have similar research productivity and career aspirations. Further follow-up is necessary to determine the relationship between pregnancy and parenthood over time. (P121) Image-Guided Radiation Therapy Utilization and Practice Patterns: Results From a National Survey of ASTRO Membership Nima Nabavizadeh, MD, David A. Elliott, MD, Aaron Kusano, MD, Yiyi Chen, PhD, Timur Mitin, MD, PhD, John M. Holland; Oregon Health and Science University; University of Washington INTRODUCTION: Image-guided radiation therapy (IGRT) practices differ widely across institutions, with no consensus regarding ideal pretreatment imaging modality, frequency, or verification process. The purpose of this study is to survey clinical IGRT practice patterns and their impact on clinical workflow. METHODS: A total of 5,979 surveys were emailed to the membership of the American Society of Radiation Oncology (ASTRO). The disease site–specific survey consisted of questions pertaining to planning target volume (PTV) margins, pretreatment image guidance modality/frequency, and method of image verification, as well as questions regarding the utility and value of IGRT. Online image verification was defined as images checked and corrected prior to the day’s treatment. Offline image verification was defined as images obtained prior to treatment and then verified prior to the following day’s treatment. Associations between IGRT practice patterns and PTV margin size were examined using a linear regression model. RESULTS: Of 671 responses (11%), 70 were nonphysician, resulting in 601 evaluable responses. The majority of respondents used IGRT (99%) for at least one fraction, with cone-beam computed tomography (CBCT) being the most commonly used modality (85%). Daily CBCT was obtained most frequently for intact prostate (63%), followed by prostatic fossa (60%), head and neck (H/N) ARS PROCEEDINGS 2015 79 American Radium Society Scientific Papers and Posters 2015 (52%), pelvic intensity-modulated RT (51%), lung (50%), esophagus (40%), central nervous system (39%), and breast (7%). Regardless of imaging modality, daily online or offline image verification was the most common schedule (range: 72%–96% daily, 4%–24% weekly, and 1%–3% first few fractions only). Online image verification was most common for H/N (92%) and least common for breast (77%), with first-few-fractions–only online schedules most common for all disease sites except breast. The majority of respondents felt comfortable with therapists verifying IGRT independent of a physician (54%) and did not believe IGRT techniques negatively impacted clinical productivity (53%) or the physician-patient relationship due to excessive interruptions (57%). Additionally, the majority of respondents agreed that for pediatric cases, the benefits of IGRT outweighed the risks of additional radiation exposure (85%). No association was seen between IGRT frequency or method of verification and PTV margin size (P > .05 for all comparisons). CONCLUSION: IGRT use is widespread, without standardization of pretreatment imaging modality, frequency, or verification process. Additionally, PTV margin size selection does not appear to be based on IGRT frequency or method of verification. Further research aimed at optimizing IGRT techniques is needed to ensure accurate, safe, timely, and cost-effective treatment delivery. (P122) Older African Americans’ Use of Religious Music to Cope With Cancer Jill B. Hamilton, PhD, Angelo D. Moore, PhD, FNP, Kayoll Galbraith, BSN, Peter A. Johnstone, MD; Johns Hopkins University School of Nursing; Center for Nursing Science and Clinical Inquiry, Womack Army Medical Center, US Army; University of North Carolina, Chapel Hill; Moffitt Cancer Center OBJECTIVE: Among Americans, African Americans are more like- ly to pray at least daily, report affiliations with a religious group, indicate that religion is very important in their lives, and indicate that they are certain God exists. An extensive body of literature on African-American spirituals informs regarding the use of religious songs to manage life stressors, such as a cancer diagnosis. METHODS: A total of 65 African-American men and women residing in the southeastern United States were interviewed. Inclusion criteria included: African-American ethnicity by selfreport, age at least 50 years, and having experienced the loss of a loved one or a life-threatening illness. Semistructured interviews lasting 15–60 min were conducted between 2008 and 2010. These interviews were held in participants’ homes or private rooms located in their churches. Participants were given a $25 gift card for participating. All interviews were audiotaped and transcribed verbatim. Initial steps of content analysis were to construct a table that organized the data collected, including participants’ responses on whether a song, scripture, or prayer was used; the words of the songs, scriptures, and prayers; the personal meanings of the songs, scriptures, or prayers; and the outcomes derived from using any of these religious expressions. Five categories of religious songs derived from the data were: Thanksgiving and Praise, Instructive, Memory of Forefathers, Communication With God, and Belief in Life After Death. 80 RESULTS: Of the participants interviewed, 23 indicated that cancer was their most stressful life event. The most frequent type song used was Thanksgiving and Praise (n = 9, 39%), followed by Instructive (n = 8, 35%), Communication with God (n = 7, 30%), Belief in Life after Death (n = 5, 22%), and Memory of Forefathers (n = 3, 13%). The most frequently reported outcomes were comfort (n = 9, 39%), hope (n = 9, 39%), and strength (n = 9, 39%). The least frequently reported outcomes were peace of mind (n = 8, 35%), support (n = 6, 26%), protection (n = 1, 4%), and guidance (n = 1, 4%). CONCLUSIONS: Religion and particularly the use of religious songs are important aspects of coping with the cancer experience among older African Americans. During diagnosis and treatment, a religious song is likely to be a viable complement to therapies for symptom reduction and mood elevation among this population. (P123) Reducing Time From Patient CT Simulation (CT SIM) Appointment Time to Start of Actual CT Scan: Lean Thinking in the VA System Alice V. Cheuk, MD, George A. Dawson, MD, Jorge H. Restrepo, AS; James J. Peters VA Medical Center PURPOSE: Computed tomography simulation (CT SIM) is vital to planning a patient’s radiation treatment (RT) and often takes a considerable amount of time to perform. This is particularly true of prostate patients, who have to drink until their bladders are full for the scan. CT SIM appointments are scheduled for 60 minutes, but preliminary observations have found that images are often acquired after the hour mark. The goal of the study was to reduce the time from CT SIM appointment/registration time to actual CT image acquisition for prostate CT SIMs, using Lean methodology. METHODS: Observation of the process was performed (Gemba walking) to map out the current process and to document time from appointment/registration to actual CT image acquisition. Each step in the process was diagrammed, timed, and recorded. The current map underwent workflow analysis, and areas of delay/ redundancy were modified or eliminated. A new process map was generated and put into practice. Five CT SIMs were timed and recorded as before. A sustain plan was implemented, involving measures to maintain use of the new map, and CT SIMs were timed to ensure continuation of the new process. RESULTS: A total of 10 CT SIMs were timed with the old process, and the average time was 46.6 minutes. On analysis, the old process map had 16 steps with major barriers, which included time needed to consume water before CT SIM, time for consent, and delays caused by patients needing to urinate before the scan, requiring the process to be restarted from the beginning. Several steps were eliminated/consolidated, including waiting for the consent to be completed before starting the drinking process (consolidated), having the front desk call therapists to notify them that the patient is ready (eliminated), and having therapists tell the patient to start drinking (eliminated). After these changes were made, there were 10 steps in the new process map, and the average time for CT acquisition was shortened to 36.2 minutes, a 22% improvement. These changes were maintained with the sustain plan, as revealed by an acquisition time of 32 minutes during the sustain phase. American Radium Society Scientific Papers and Posters 2015 CONCLUSIONS: By applying Lean thinking, we were able to reduce the time for CT image acquisition for prostate cancer patients requiring radiotherapy planning by 22%. By reducing the time for CT SIM, veterans’ experiences and satisfaction were improved, flow through the radiation oncology department was optimized, and capacity for CT SIMs was increased. (P124) Dosimetric Comparison of Three-Dimensional Conformal Proton Radiotherapy and IntensityModulated Proton Therapy for Treatment of Pediatric Hodgkin Lymphoma Jamie M. Pinckard, BS, William W. Chance, MD, Cody N. Crawford, CMD, Shengpeng Jiang, Xiaodong Zhang, PhD, Anita Mahajan, MD; UT Health Science Center, San Antonio; UT MD Anderson Cancer Center OBJECTIVES: Thoracic irradiation in pediatric patients is associated with serious long-term adverse effects. We compared the target coverage and dose distribution in normal tissues in pediatric patients with Hodgkin lymphoma using three-dimensional conformal proton radiotherapy (3DPRT) and intensity-modulated proton therapy (IMPT). METHODS: From 2009–2014, seven pediatric patients with Hodgkin lymphoma with intrathoracic involvement received proton therapy to 21–25.2 Cobalt Gray Equivalents (CGE) in 15 fractions. Five patients received 3DPRT, and two patients received IMPT. To compare dosimetric endpoints between the modalities, an additional 3DPRT or IMPT plan was generated for each patient (n = 14 plans). Mean doses to target and nontarget tissues were then compared between 3DPRT and IMPT using the Wilcoxon matched-pair signed-rank test. RESULTS: The planning target volume was adequately treated using both techniques. When compared to 3DPRT, IMPT resulted in a lower mean total lung dose (632 cGy vs 737 cGy; P < .05) and mean heart dose (797 cGy vs 972 cGy; P < .05). The mean dose to the breasts, thyroid gland, esophagus, pancreas, stomach, and liver were similar between the modalities. CONCLUSIONS: In pediatric patients with intrathoracic involvement, IMPT produced significantly lower doses to the lung and heart compared with 3DPRT. Additional studies will be needed to determine the clinical benefit of these findings. (P125) Glioblastoma Multiforme Outcome Comparison Between Pediatrics and Adults: Is There a Difference? Dayssy A. Diaz, MD, Duran Mitchell, Isildinha Reis, PhD, Joseph Panoff, MD; Department of Radiation Oncology, Division of Biostatistics, Department of Public Health Sciences, University of Miami PURPOSE: There is a paucity of data regarding the clinical outcome of glioblastoma multiforme (GBM) in the pediatric population. Previous data suggest that there is a difference in overall survival (OS) between pediatric patients and adult patients with GBMs; however, there have been no direct comparisons in the literature. We compared pediatric and adult GBM outcome mea sures using the Surveillance, Epidemiology, and End Results (SEER) database, a program of the National Cancer Institute that collects cancer incidence and survival data from approximately 28% of the US population. MATERIALS AND METHODS: Patients with pathologically confirmed grade 4 gliomas diagnosed between 1980 and 2011 were included in this analysis. Patients aged older than 60 years were excluded from the analysis. Pediatric cases were classified as those aged ≤ 21 years. Cause-specific survival (CSS) and overall survival (OS) were estimated by Kaplan-Meier (product-limit) method. Cox proportional hazards analyses were performed, evaluating potential predictors of CSS and OS. RESULTS: There were 6,351 patients available for comparison; 6,099 patients were adults, and 252 patients were pediatric. The median age for the pediatric patients was 13 years, and the median age for the adults was 52 years. The OS time was the same as the CSS for pediatric patients (16 mo). The same observation was seen for adults (12 mo). Pediatric patients had a significantly better OS (P < .001) and CSS (P < .0001) when compared with adults. The 3-year OS was 23.4% for pediatrics vs 11.8% for adults, and the 3-year CSS was 24.3% for pediatrics vs 12.9% for adults. These results persisted when stratified by decade of diagnosis (1980s: P = .0008; 1990s: P = .0229; and 2000s: P < .0001). Univariate analysis indicated female sex (P = .03), use of radiation (P < .0001), surgical resection (P < .0001), age (P < .0001), recent year of diagnosis (P < .0001), and pediatric patients (P < .0001) to be significant predictors of improved OS. Multivariate analysis demonstrated that pediatric status, use of surgery, use of radiation, and recent decade of diagnosis were significant predictors of OS and CSS (P < .0001 for all variables). Race was found to be a significant predictor of CSS (P = .0132) but not of OS (P = .667). CONCLUSION: This is the first report to directly compare pediatric and adult patients with GBM. Pediatric patients had significantly superior OS and CSS when compared with adults, and these differences remained significant over time and on multivariate analysis. The underlying cause of these survival differences between pediatrics and adults requires exploration, with attention to molecular biological tumor differences. (P126) Long-Term Volumetric Follow-Up of Juvenile Pilocytic Astrocytomas Treated With Proton Beam Radiotherapy Edward M. Mannina, MD, MPH, MS, Greg Bartlett, CMD, Peter A. Johnstone, MD, Kevin P. Mcmullen, MD; Indiana University School of Medicine; University Of South Florida INTRODUCTION: Juvenile pilocytic astrocytomas (JPAs) are World Health Organization (WHO) grade 1 glial neoplasms treated by resection with radiotherapy reserved for inoperable cases or following subtotal resection. Proton radiotherapy minimizes integral dose and is thus preferred in children. We analyzed the magnetic resonance imaging (MRI) follow-up of patients with JPA treated with proton radiotherapy to define the volume changes, response rate, need for postradiotherapy intervention, and survival. ARS PROCEEDINGS 2015 81 American Radium Society Scientific Papers and Posters 2015 METHODS: A total of 15 pediatric patients histologically diagnosed with JPA made up this retrospective report. From August 2005 through March 2012, patients were treated to a median dose of 5,400 cGy (relative biological effectiveness [RBE]) using proton radiotherapy and then followed with serial MRIs for 3 years. MRIs were imported into Eclipse 11 treatment planning software, where contours of the T1 contrast-enhancing volumes, including cystic components, were generated by one clinical radiation oncologist (EMM). Volume in cm3 was plotted against time since completion of therapy to track volumetric changes. Demographics, prior therapies, and postradiotherapy interventions were cataloged. RESULTS: This is a retrospective review of 15 patients with a mean age of 10.9 years (range: 4–20 yr) and mean number of 8.9 MRIs (range: 4–12). A total of 10 of 15 (67%) patients had prior R2 resections, with 3 patients having two R2 resections. Further, 12 of 15 (80%) patients had cerebrospinal fluid (CSF) shunts prior to radiotherapy, and 9 of 15 (60%) patients received prior chemotherapy, all with at least a platinum agent and vincristine. With a median follow-up of 55.3 months, 14 of 15 (93%) patients were alive, for an estimated 5-year overall survival (OS) of 93.3%. Median event-free survival (EFS) was 86.6 months, with an estimated 5-year EFS of 72.2%. A total of 11 of 15 (73.3%) patients were declared to be responders by 6 months, with 3 of 11 (27%) demonstrating pseudoprogression (increase in volume followed by spontaneous regression), with a mean time to maximum volume of 177 days. Also, 4 of 15 (26.7%) patients were nonresponders, including one who died of progression 9 months after radiotherapy and another who restarted chemotherapy. Three patients underwent shunt revisions, while two received hyperbaric oxygen, one for presumed radionecrosis and another for biopsy-proven radionecrosis (the only patient with prior radiation). Stereotactic cyst aspiration was required in one case. One heavily pretreated patient developed a hematologic malignancy requiring further chemotherapy. A total of 8 of 15 (53.3%) patients required no further therapy or intervention after radiotherapy. Gross tumor volume changes ranged from a 91% reduction to a 207% increase during the evaluation period. CONCLUSIONS: Pediatric patients with PAs can have extended survival following proton beam radiotherapy. This volumetric study illustrates that responders are declared within 6 months, but vigilant surveillance is necessary due to the potential need for postradiotherapy interventions. (P127) Stereotactic Accelerated Partial Breast Irradiation (SAPBI) for Early-Stage Breast Cancer: Rationale, Feasibility, and Early Experience Using the CyberKnife Radiosurgery Delivery Platform Olusola Obayomi-Davies, MD, Bridget Oppong, MD, Sonali Rudra, MD, Erini Makariou, MD, Lloyd D. Campbell, CMD, Kaiguo Yan, PhD, Leonard Chen, MD, PhD, Simeng Suy, PhD, Sean P. Collins, MD, PhD, Keith Unger, MD, Eleni Tousimis, MD, Brian T. Collins, MD; Georgetown University Hospital BACKGROUND: The efficacy of accelerated partial breast irradiation (PBI) utilizing brachytherapy or conventional external beam radiation has been studied in early-stage breast cancer treated with 82 breast-conserving surgery. Data regarding stereotactic treatment approaches are emerging. The CyberKnife linear accelerator enables excellent dose conformality to target structures while adjusting for target and patient motion. We report our institutional experience on the technical feasibility and rationale for stereotactic accelerated PBI (SAPBI) delivery using the CyberKnife radiosurgery system. METHODS: From February 2013 to December 2013, a total of 10 patients received CyberKnife SAPBI. Four gold fiducials were implanted around the lumpectomy cavity prior to treatment under ultrasound guidance. Fiducials were tracked in real time using the Synchrony tracking system. The clinical target volume (CTV) was defined on contrast-enhanced computed tomography (CT) scans using surgical clips and postoperative changes. Inverse CyberKnife plans were generated to deliver 30 Gy in five fractions to a 5-mm uniform expansion around the CTV (PTV30Gy). PTV and organs at risk (OARs) dosimetry were assessed as per the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-39 study, and cosmesis was assessed using the Harvard breast cosmesis scale. RESULTS: At least three fiducials were tracked in 100% of cases. The mean PTV30Gy was 70 cm3, and the mean prescription isodose line was 80%. A total of 100% of the PTV30Gy received the prescription dose, with a mean maximum dose of 36 Gy. The mean ipsilateral breast V30Gy and V>15Gy were 14% and 31%, respectively. The ipsilateral lung V9Gy was 3%, and the contralateral lung V1.5Gy was 8%. For left-sided breast cancers, the heart V1.5Gy was 31%, with a mean heart dose of 1 Gy. The mean contralateral breast maximum dose was 3 Gy. The maximum skin and chest wall doses were 32 Gy and 33 Gy, respectively. One patient experienced grade 1 skin induration 3 weeks after treatment, which resolved with observation. At a median follow-up of 46 weeks, all patients have experienced good/excellent cosmetic outcomes, and no significant adverse outcomes have been reported. CONCLUSIONS: SAPBI via CyberKnife is a suitable platform for partial breast irradiation, offering improvements over existing APBI techniques, with excellent normal tissue-sparing. Our early findings indicate that CyberKnife SAPBI delivered in five fractions is a feasible, well-tolerated, and reliable platform for delivering PBI. (P128) Statistically Significant Calibration Curve for Ir-192 HDR Source Using Multichannel Gafchromic Film Analysis J. Barbiere, J. Napoli, A. Ndlovu; HUMC PURPOSE: This work presents a unique methodology to create statistically valid calibration curves required to validate high-doserate (HDR)–planned dose distributions during brachytherapy treatment planning commissioning. A robust calibration curve is essential to accurately convert measured planar Gafchromic film density to dose for comparison with calculated dose distributions using standard metrics. The process presented herein is appropriate for widespread application since it can be used with any planning/ analysis system that is capable of DICOM (Digital Imaging and COmmunication in Medicine) dose import/export, does not American Radium Society Scientific Papers and Posters 2015 require specialized phantoms, and can be performed rapidly on the actual film sheet used for plan validation in a single exposure. MATERIALS AND METHODS: A section of Gafchromic EBT3 film was positioned in contact with an HDR bronchial catheter surrounded by water-equivalent material. The film was irradiated for 1,350 Cisec with a Nucletron Ir-192 HDR line source and then digitized with an Epson 10000 XL color scanner. The remaining section of the film can be used for plan evaluation. Film analysis was performed in MatLab with the imaging toolbox. Transmission values for the red, green, and blue channels were converted into optical density (OD) as a function of perpendicular distance from the center of the line source. A corresponding theoretical dose curve for each exposure was calculated using the TG43 formalism. The calculated dose was fitted to the measured OD in the clinically significant range to a fourth-order polynomial. A single dwell position plan, normalized for the appropriate irradiation time, was created using the Oncentra treatment planning system. The three-dimensional (3D) DICOM dose matrix was imported into a Varian Eclipse external beam treatment planning system, and the two-dimensional (2D) dose in a coronal plane was extracted for comparison in MatLab with the corresponding measured film dose plane. RESULTS: (1) The green channel had greater range, contrast scale, and accuracy than either the blue or red channel and was therefore used for analysis. (2) The calibration curve, consisting of nearly 200 points, was statistically valid and robust. (3) Agreement between the measured film and plan dose was within +/− 1.5% throughout the clinical range in 0.1-mm increments. (4) Our results provide a significant improvement in both dose and distance to agreement over current practice, which typically uses fewer than 10 calibration points. CONCLUSION: An accurate and robust calibration curve using the green channel of Gafchromic film can be easily created for use in validating HDR planning systems. (P129) Bone Marrow Aspiration Under CT Guidance: Technique and Value Cory Pfeifer, MD, Tyler Ternes, Shannon St. Clair, William Palko, MD, Christopher Dakhil; University of Kansas, Wichita; Lenox Hill Hospital; Wesley Medical Center; Cancer Center of Kansas PURPOSE: (1) Describe the quality basis for our institution’s transition toward computed tomography (CT)-guided bone marrow aspiration. (2) Identify the equipment and approach used in bone marrow aspiration, as well as its pitfalls. (3) Outline the steps employed at our institution. (4) Emphasize the role of the partnership between radiologists, pathologists, and oncologists in the successful execution of the procedure. BACKGROUND: The shift toward quality-based health care delivery models has underlined the need to provide accurate diagnoses with as few complications as possible. These changes in our health care structure parallel procedure-complicating factors, such as changes in typical patient body habitus and clinician availability to perform standard bone marrow aspiration. At our institution, radiologists have partnered with oncologists to address these challenges by providing high-quality service with minimal complication rates utilizing CT guidance. PROCEDURE DETAILS: Standard CT guidance is utilized to localize the bilateral posterior iliac crests, and a posterior technique is employed with the patient in the prone position. We approach the patient from the contralateral side of the iliac target. Our experience has shown that pathologists prefer to supply T-handle twisttype biopsy needles for sampling, as opposed to the hammer-based needle setup. Feedback from the pathology service has suggested that the hammer method results in greater contamination by peripheral blood. Once the marrow cavity is accessed, we aspirate separate marrow aliquots for slide fixation with a histotechnologist at the bedside who prepares the fresh specimens. A sample core is then obtained and preserved separately. Communication with the ordering oncologist is essential for specimen acquisition, as flow cytometric analysis and specific marker assays require disparate bedside preservation techniques. Our stepwise technique will be described pictorially. CONCLUSION: Bone marrow aspiration under CT guidance is a safe procedure that has become commonplace at our institution. Pain is minimal, even in the absence of sedation. All biopsies performed at our institution with participation from the pathology department have resulted in diagnostic specimens, and relationships with healthcare stakeholders have improved. (P130) An Exploratory Pilot Study of Perfusion Patterns in Locoregionally Advanced Head and Neck Cancer Using a Novel Analysis Technique of Dynamic ContrastEnhanced (DCE)-MRI Jonathan Verma, MD, Radka Stoyanova, PhD, N. Andres Parra, PhD, Kyle Padgett, PhD, Michael Samuels, MD, Nagy Elsayyad, MD; University of Miami Sylvester Comprehensive Cancer Center PURPOSE: Recently, an unsupervised pattern recognition (UPR) technique was developed by one of the investigators to analyze dynamic contrast-enhanced magnetic resonance imaging (DCEMRI). The technique utilizes non-negative matrix factorization (NMF) and has hitherto not been studied in head and neck cancers (HNCs). This study aims at implementing this technique in HNCs in order to explore its potential value in characterizing perfusion patterns within these tumors. The presence of subvolumes in the HNC that are differentially perfused may have future therapeutic implications with respect to dose painting, intensification, or deintensification of treatment. METHODS: Eight patients who received definitive chemoradiotherapy for locoregionally advanced HNC were randomly selected. All received 70 Gy/33–35 fractions/6–7 weeks with concurrent cisplatin. As part of the radiotherapy planning, MRI had been obtained for each patient, and DCE sequences were acquired, in addition to the usual T1 and T2 sequences. DCE-MRI datasets with contoured gross target volumes (GTVs) were imported for image analysis into MIM software (MIM Corporation, Cleveland, Ohio). The NMF ARS PROCEEDINGS 2015 83 American Radium Society Scientific Papers and Posters 2015 (non-negative matrix factorization) algorithm was implemented in MIM. There are three characteristic patterns of signal-vs-time curves that UPR identifies within each GTV—namely, Pattern A, which identifies subvolumes with fast contrast uptake and fast washout; Pattern B for subvolumes with delayed uptake and washout; and Pattern C, with slow or no contrast uptake. RESULTS: The aforementioned three patterns identified corresponding subvolumes in each GTV. The mean GTV volume was 56.8 cc (range: 22–146.6 cc). Subvolumes for each of Patterns A, B, and C averaged 31% ± 3.6%, 30% ± 5.3%, and 39% ± 8.4% of the total GTV, respectively. The standard deviations showed relatively tight dispersion of these values around their respective means, suggesting relative constancy throughout the patients studied. Although no clear geographically consistent relationship between the subvolumes emerged, necrotic regions identified by computed tomography (CT) or MRI were consistently located within Pattern C subvolumes. All three patterns were qualitatively identical to those reported in a previously published preclinical study by our group, wherein Pattern B subvolumes corresponded to regions of tumor hypoxia on 18F-fluoro-misonidazole (MISO) images and radiolabeled pimonidazole histological slices. CONCLUSION: UPR is a novel technique that has been developed in a preclinical tumor model and that is potentially applicable in HNCs. This exploratory study appears consistently to identify subvolumes within the GTV that may have therapeutic implications and that may be promising areas for further research (eg, correlation with hypoxia and clinical outcome). (P131) Improved Prostate Delineation in Prostate HDR Brachytherapy With TRUS-CT Deformable Registration Technology Xiaofeng Yang, Peter Rossi, MD, Ashesh Jani, MD, Hui Mao, PhD, Tomi Ogunleye, Walter J. Curran, Tian Liu; Emory University PURPOSE: The main challenge for computed tomography (CT)- based prostate high-dose-rate (HDR) treatment planning is to accurately define prostate volume in CT images due to the poor soft-tissue contrast. To address this issue, we have developed a novel approach that integrates intraoperative transrectal ultrasound (TRUS)-based prostate volume into HDR treatment planning through TRUS-CT deformable registration (TCDR) based on the catheters. This study’s purpose is to assess the clinical feasibility of the TCDR-based approach and compare its performance with the standard physician’s CT contours in the setting of HDR prostate brachytherapy. METHODS: To improve prostate contour accuracy, the proposed TCDR-based approach utilizes two unique features in the HDR procedures: (1) Accurate prostate volume obtained from threedimensional (3D) TRUS images—during HDR procedure, 3D TRUS images of the prostate are acquired with 1- or 2-mm step size right after the catheters are inserted. (2) Accurate TRUS-CT image registration using HDR catheters—both TRUS and CT images are obtained after the catheter insertions. The catheters are reconstructed on the TRUS and CT images and 84 subsequently used as landmarks for the TRUS-CT image registration using a fuzzy-to-deterministic algorithm. The intraoperative TRUS-based prostate volume is then deformed to the CT image to obtain the CT prostate volume, which is used for HDR treatment planning. In a pilot study, 16 prostate HDR brachytherapy procedures were evaluated retrospectively. CT-based prostate volumes were contoured by two experienced physicians. Both TCDR-based prostate volume and the standard CT-based prostate volumes are compared with the MRI-defined prostate contours (gold standard). A t-test was used to examine the significance of the difference between the two physicians’ CT-based contours, the TCDR-defined contours, and the MRI-defined prostate volumes. RESULTS: A 20% to 40% improvement in prostate volume accuracy can be achieved with the TCDR-based approach as compared with the standard CT-based prostate volumes. TCDR-defined prostate volumes match closely to the MRI-defined prostate volumes (mean difference: 0.9 ± 7.3%; P = .54). CT-based contours assessed by the two physicians overestimated the size of the prostate gland by 7.3% (7.3 ± 18.7%) and 59.7% (59.7 ± 37.8%) and were associated with significant variability, even between these two experienced physicians (P < .01). CONCLUSIONS: We have developed a novel TCDR-based approach to improve prostate delineation utilizing intraoperative TRUSbased prostate volume in prostate HDR brachytherapy and demonstrated its improvement in prostate volume accuracy over the standard CT-defined prostate volumes. Integration of TRUS into the HDR brachytherapy treatment planning workflow has the potential to enable accurate dose planning and delivery and enhance prostate HDR treatment outcome. (P132) Evaluation of the Chest Wall Skin Dose Associated With Bolus Application in Postmastectomy Radiation Therapy (PMRT) Using Nanodot OSLD Yuenan Wang, PhD, Janelle Park, MD; Florida Cancer Specialists PURPOSE: A wide variation in the use of bolus has been observed in postmastectomy radiotherapy (PMRT), which affects dose delivery to skin and may have an impact on locoregional recurrence. However, even treatment planning systems (TPSs) with advanced algorithms do not provide accurate dosimetry in the buildup region. This study aims to perform in vivo measurements of the chest wall (CW) skin dose using optically stimulated luminescence dosimeters (OSLDs) and to ensure that adequate skin dose is obtained for all patients undergoing PMRT. MATERIALS AND METHODS: In total, 37 patients undergoing PMRT were treated with either conformal technique or volumetricmodulated RT (RapidArc, Varian Medical Systems). Each patient had two plans for CW radiation: one with 0.5-cm bolus and the other without bolus, originally prescribed as every other day (QOD). During the first treatment of bolus and no-bolus plans, the physicist placed four to six nanoDots (Landauer, Glenwood, IL), a type of OSLD suitable for accurate in vivo dosimetry, at the scar and the superior, inferior, medial, and lateral locations of the CW field. For each patient, there were 8–12 nanoDots sent to Landauer for absolute dose reading. The skin dose difference between the American Radium Society Scientific Papers and Posters 2015 no-bolus and bolus plans was analyzed at our clinic for each patient. Furthermore, if the composite CW skin dose was less than the clinically desired dose (ie, 45 Gy), then we would increase the number of fractions for the bolus plan and reduce the number of fractions for the no-bolus plan to ensure that CW skin obtains the clinically desired dose coverage. RESULTS: A total of 8 out of 37 patients had skin dose below 45 Gy with the 0.5-cm QOD bolus application, and we had to revise the plan fractionation. Among them, two were conformal plans and seven were RapidArc plans. For the conformal planned patients, the average skin dose increase for bolus use compared with no bolus was 34% ± 12%. In contrast, for RapidArc plans, the average skin dose increase for the bolus plan was 57% ± 27%. This was probably due to the oblique incidence of tangential photon beams in the three-dimensional (3D) conformal technique already providing relatively high skin dose. CONCLUSION: Skin dose is dependent on patient anatomy, the incident beam geometry, and planning techniques. For example, RapidArc plans with tissue expanders should be monitored for skin dose coverage. In conclusion, the skin dose should be measured to ensure that the clinically desired dose is obtained for the various bolus applications in patients undergoing PMRT. (P133) Spinal Cord Dose Comparison With and Without Contrast Density Correction on CT Myelogram Simulations for Patients Treated With Stereotactic Body or Conventional Radiotherapy Sukhjeet S. Batth, MD, MS, Raymond Chiu, CMD, Nicholas Trakul, MD, PhD; USC Norris Cancer Center PURPOSE/OBJECTIVES: Patients treated with stereotactic body radiotherapy (SBRT, also known as stereotactic ablative radiotherapy) or conventional RT to spinal targets often undergo a computed tomography (CT) simulation with a myelogram to accurately identify the spinal cord. Treatment planning algorithms use Hounsfield units and a CT density table to calculate radiation dose. Accuracy of these calculations may be improved by correcting for the contrast’s actual density, which is not present at the time of treatment delivery. We hypothesize that the spinal cord receives a higher radiation dose than calculated after correction of the contrast’s physical density. MATERIALS AND METHODS: CT simulations with a myelogram for treatment of 10 spinal targets from nine patients were retrospectively reviewed. Seven of these targets were treated with SBRT, one was treated with intensity-modulated RT, and two were treated with three-dimensional conformal RT. The contrast from the myelogram was contoured in the treatment planning software, and the physical density was adjusted to 1 g/cm3. The uncorrected spinal cord maximum and mean doses were then compared with the doses after correction of the density. A two-tailed, paired t-test was used to compare the calculated doses. RESULTS: The corrected spinal cord doses were all significantly higher than the uncorrected doses based on paired t-test for the maximum dose (P = .0073) and the mean dose (P = .00087). The mean maximum cord dose was 2,065 cGy and 2,077 cGy after correction. The mean of the average cord dose was 1,302 cGy and 1,315 cGy after correction. CONCLUSIONS: These data suggest that the use of a CT myelogram with simulation significantly underestimates the spinal cord maximum and mean doses. The clinical significance of this finding is uncertain, and this study is limited by its small sample size. Density correction should be further investigated in a larger study, given the high doses used in SBRT and potential morbidity of spinal cord injury. (P134) An Overview of the First One Hundred Patients Treated With a New Cutting-Edge, Frameless, ImageGuided, Stereotactic Radiosurgery System Ning Wen, PhD, Karen Chin-Snyder, MS, Haisen Li, PhD, Kwang Song, PhD, James Gordon, PhD, Indrin Chetty, PhD, Munther Ajlouni, MD, Farzan Siddiqui, MD, Benjamin Movsas, MD, Salim Siddiqui, MD; Henry Ford Health System PURPOSE: To summarize our experience with treating our first 100 patients treated using an innovative linear accelerator (Linac)based stereotactic radiosurgery (SRS) system. MATERIALS AND METHODS: A novel platform for Linac-based SRS treatment, the Edge (Varian Medical Systems, Palo Alto, CA), offers multiple imaging modalities for treatment localization, including an optical surface monitoring system (OSMS) for surface tracking, 2.5 MV portal imaging, triggered monoscopic kV imaging to track intrafractional motion, four-dimensional (4D) conebeam computed tomography (CBCT) to evaluate tumor motion offline, extended CBCT images by stitching multiple CBCT scans together, and a Calypso/Varian electromagnetic beacon-based tracking system. The new robotics couch supports six-degrees-offreedom corrections from multiple imaging modalities for precise patient setup. We have comprehensively commissioned this system and have worked out a clinical workflow for treating patients with different cancer types. RESULTS: The breakdown of tumor sites treated with SRS/ stereotactic body radiation treatment (SBRT) techniques was central nervous system (CNS) (31%), lung (29%), spine (23%), GI (9%), adrenal glands (7%), and head and neck (1%). The treatment techniques used were as follows: volumetric-modulated arc therapy (39%), intensity-modulated RT (35%), dynamic conformal arc (21%), and three-dimensional (3D) conformal RT (5%). For CNS sites except glioblastoma multiforme (GBM), the dose was delivered in a single fraction (range: 10–18 Gy), and the target volume (TV) ranged from 0.10 cc (acoustic neuroma) to 21.05 cc (metastasis). GBM patients were treated with four fractions using the simultaneous integrated boost technique (32 Gy to the enhanced lesion and 24 Gy to the edema). Volume sizes of the enhanced lesions ranged from 7.86 cc to 128.05 cc. All lung patients were treated with 48 Gy in four fractions, except three recurrent treatments with 36 Gy in three fractions. Median TV volume was 36.53 cc (range: 4.09–85.12 cc). V20 for the total lungs was 3.59% on average (range: 0.50%–7.37%). All T-spine metastases were treated in a single fraction, with the dose ranging from 14 Gy (primary: ARS PROCEEDINGS 2015 85 American Radium Society Scientific Papers and Posters 2015 multiple myeloma) to 18 Gy (all other primary diagnoses). The median TV volume was 43.82 cc (range: 11.80–121.77 cc). The maximum dose to 0.03 cc of the spinal cord was 10.86 cc (range: 7.90–15.15 cc). The clinical flow designed for each treatment site using multiple imaging modalities and treatment techniques will be presented. The setup time, imaging fusion time, and beam on time will be categorized according to the treatment site. exceptionally high, although significantly lower agreement was noted postoperatively. Technique and dose prescription between experts were substantively consistent, and these preliminary results will be utilized to create an expert-consensus contouring atlas to aid the nonexpert radiation oncologist in the planning of these challenging, rare tumors. CONCLUSION: The Edge RS system possesses the high-accuracy localization and delivery requirements for safely treating patients with tumors in the CNS and other sites using SRS/SBRT-based techniques. (P136) Quantitative Measurement of Contrast Enhancement in Myxoid Liposarcomas: Response to Neoadjuvant Therapy With Volumetric and Pathologic Correlates (P135) Quantitative Assessment of Target Delineation Variability for Thymic Cancers: Agreement Evaluation of a Prospective Segmentation Challenge Emma B. Holliday, MD, Clifton D. Fuller, MD, PhD, Jayashree KalpathyCramer, PhD, Daniel Gomez, MD, Andreas Rimner, MD, PhD, Ying Li, MD, PhD, Suresh Senan, MD, PhD, Lynn D. Wilson, MD, MPH, Jehee Choi, MD, Ritsuko Komaki, MD, Charles R. Thomas, MD; UT MD Anderson Cancer Center; Martinos Center for Biomedical Imaging Center, Massachusetts General Hospital, Harvard Medical School; Memorial Sloan-Kettering Cancer Center; UT Health Science Center, San Antonio; VU Medical Center; Yale University; Loyola University; Oregon Health and Science University Knight Cancer Institute PURPOSE: We sought to quantitatively determine the relative interobserver variability of expert target volume delineation as part of a larger standardization effort to develop an expert-consensus contouring atlas to complement radiotherapy recommendations for thymic cancers. METHODS: A pilot dataset was created, consisting of a standardized case presentation with anonymized pre- and postoperative DICOM computed tomography (CT) image sets from a single patient with Masaoka-Koga stage III thymoma. Participating expert thoracic radiation oncologists delineated tumor targets on the pre- and postoperative scans as they would for a definitive and adjuvant case, respectively. Respondents then completed a survey detailing the dose prescription and planning target volume (PTV) margins that they would recommend in definitive and postoperative (ie, R1 vs R2) scenarios. Interobserver variability was analyzed quantitatively with Warfield’s simultaneous truth and performance-level estimation (STAPLE) algorithm and the Dice similarity coefficient (DSC). RESULTS: Seven users completed the contouring tasks for definitive and adjuvant cases; of these users, five completed online surveys. Segmentation performance was assessed, with high mean ± SD STAPLE-estimated segmentation sensitivity for definitive-case gross tumor volume (GTV) and clinical target volume (CTV) at 0.77 and 0.80, respectively, and postoperative CTV sensitivity of 0.55; all volumes had a specificity of ≥ 0.99. Interobserver agreement was markedly higher for the definitive-case target volumes, with mean ± SD DSC of 0.88 ± 0.03 and 0.89 ± 0.04 for GTV and CTV, respectively, compared with postoperative CTV DSC of 0.69 ±0 .06 (Kruskal-Wallis: P < .01). CONCLUSION: Expert agreement for definitive-case volumes was 86 Tobias R. Chapman, MD, MPharmacol, George Jour, MD, Darrin J. Davidson, MD, MHSc, FRCSC, Robin L. Jones, MD, MRCP, Ben Hoch, MD, Gabrielle Kane, MD, Edward Y. Kim, MD; University of Washington PURPOSE: Clinical assessment of response to neoadjuvant therapy (NeAT) in soft tissue sarcoma (STS) is often based on Response Evaluation Criteria in Solid Tumors (RECIST) criteria, which may not reflect changes in tumor biology. Myxoid liposarcomas (MLs) have favorable responses to NeAT by RECIST and exhibit contrast enhancement on T1-weighted, postcontrast magnetic resonance imaging (MRI) (T1+). Tumor response may also manifest as changes in contrast enhancement. We present a quantitative assessment of contrast enhancement and correlate these findings with volumetric and pathologic data. MATERIALS AND METHODS: After institutional review board (IRB) approval, 67 patients from 1999–2013 were identified with ML. Five patients received neoadjuvant radiation therapy (RT) with MRI prior to each NeAT and surgery. Tumor volumes (TVs) were contoured on the relevant T1+ MRI. Mean contrast enhancement (MCE) was calculated by subtracting the T1 precontrast MRI signal from the T1+ MRI after normalization to normal muscle. Percent change was calculated using MIM software. Specimen analysis was performed by an STS pathologist. RESULTS: The majority of patients was male, and most tumors were of the limb, low-grade, large (> 5 cm), and deep (T2b). Three patients received neoadjuvant chemotherapy. Pre- and postchemotherapy mean MCE values were 71% (range: 54%–82%) and 24% (range: 2%–47%), respectively. The mean postchemotherapy change in TV was −0.2%. All patients received neoadjuvant RT, with pre- and post-RT mean MCE of 50% and 47%, respectively. Mean post-RT change in TV was −64%. Two patients received neoadjuvant RT alone, with pre- and post-RT mean MCE of 95% and 23%, respectively. The mean post-RT-alone change in TV was −72%. All patients had significant pathological responses to NeAT, with > 80% treatment effect. With median follow-up of 23.5 months, there were no local or distant recurrences. CONCLUSION: Patients with ML treated with neoadjuvant chemotherapy exhibit decreased MCE, with no associated change in TV. Subsequent RT postchemotherapy causes no additional change in MCE but does cause tumor shrinkage. Treatment with RT alone causes changes in both. These data from ML patients with excellent pathologic responses to NeAT suggest that differential tumor responses may not be captured using size criteria alone and that responses to chemotherapy may be better evaluated using MCE. American Radium Society Scientific Papers and Posters 2015 CLINICAL RELEVANCE/APPLICATION: MLs exhibit favorable clinical responses to NeAT by size criteria; however, this may not account for biological tumor changes. We present a technique for quantifying changes in contrast enhancement that may better predict biological tumor response. treat patients. Based on the results of ongoing machine and patientspecific tests and with independent confirmation, the treatments are being delivered accurately and safely. (P137) Seeing Is Believing: Concerns and Solutions in Implementing Magnetic Resonance Image–Guided Radiation Therapy Andrew W. Orton, MD, John D. Gordon, MS, DABR, Tyler P. Vigh, Allison E. Tonkin, MD, George M. Cannon, MD; Huntsman Cancer Institute; Intermountain Medical Center Olga L. Green, PhD, H.O. Wooten, PhD, Yanle Hu, PhD, Rojano Kashani, PhD, Lakshmi Santanam, PhD, Harold Li, PhD, Sasa Mutic, PhD; Washington University School of Medicine PURPOSE: We describe the potential problems and solutions developed in initiating first treatments with a commercially available magnetic resonance image–guided radiation therapy (MR-IGRT) system. Specifically, we address the following: (1) effect of magnetic fields on dose distributions, (2) patient and staff safety in the presence of magnetic and radiation fields, and (3) quality assurance for MR-IGRT. MATERIALS AND METHODS: The foremost problem of integrating an MRI system with a radiation delivery system has been solved by combining a 0.35-T split-doughnut MRI with a gantry carrying three 60 Co sources. This system was installed in our institution in 2012, and since then, we have conducted ongoing work toward its clinical implementation. The problem of the effect of magnetic fields on dosimetric distributions was investigated by comparing measurements with an in-house phantom to Monte Carlo calculations. The problem of patient and staff safety was addressed by measuring the specific absorption ratio (SAR) to ensure that it was safe for patients to be imaged continuously while being treated and by implementing safety checks in our daily workflow. The problem of quality assurance for MR-IGRT was solved by establishing the accuracy of measurement devices (ionization chambers, detector arrays) in the presence of the magnetic field and developing tests that checked both the geometric and dosimetric accuracy of the system on a daily, weekly, monthly, and annual basis. Independent confirmation of dosimetric accuracy was provided by the Imaging and Radiation Oncology Core (IROC) service—reference dosimetry via optically stimulated dosimeters and overall delivery quality via the head-and-neck phantom (film and thermoluminescent dosimeters). RESULTS: The 0.35-T magnetic field was found to have a negligible effect on dose distributions for the type of patient plans used clinically. The measured SAR value was 1.14 W/kg, which ensured that patients do not experience excessive heating during the ongoing real-time imaging at four frames per second. MR training was conducted for all staff, and procedures were implemented to ensure that patients were cleared to be MR-safe before every treatment. Quality assurance included spatial integrity and homogeneity of the MR field, MR and RT isocenter coincidence, and pre- and posttreatment patient-specific QA. Satisfactory results were reported by the IROC for four independent OSL (Optically Stimulated Luminescence) reference dosimetry checks and the head-and-neck phantom. CONCLUSION: The system has been used since January 2014 to (P138) Dosimetric Evaluation of Parotid Dose in Whole-Brain Radiation Plans Covering C1 vs C2 BACKGROUND AND PURPOSE: Dosimetric analysis has demonstrated a higher-than-anticipated dose to the parotid glands during whole-brain radiation therapy (WBRT). Even in the computed tomography (CT) planning era, WBRT fields continue to be designed based on anatomic landmarks identified on digitally reconstructed radiographs (DRRs) of the skull. The dose of radiation delivered incidentally to the parotid glands is influenced by the location of the inferior border of the treatment field. The purpose of this study is to compare the dose delivered to the parotid glands and associated normal tissue complication probabilities (NTCPs) in plans covering the C1 vs C2 vertebral level. MATERIALS AND METHODS: A total of 15 patients underwent CT simulation of the brain, followed by dosimetric evaluation of parotid dose. Two treatment plans were produced for each patient using the CT simulation images: the first extended the inferior field border to the inferior edge of the C1 vertebra, and the second extended the field to the inferior edge of C2. Two dose prescriptions were compared: 30 Gy and 37.5 Gy. NTCPs were estimated using the Lyman-Burman-Kutcher model with parameters obtained from studies published by Eisbruch, Emami, and Roesink. Mean dose to the parotids and NTCPs were compared between the two groups, and statistical significance was determined using a patient-matched two-sided t-test. RESULTS: The mean parotid dose was significantly higher when C2 was covered. For the 30-Gy prescription, coverage to C2 increased the parotid dose from 14.3 Gy to 18.3 Gy (P < .01). For the 37.5-Gy plan, coverage to C2 increased the parotid dose from 18.5 Gy to 23.4 Gy (P < .01). NTCPs for the parotid gland were also higher when covering C2 vs C1. At 30 Gy, NTCP estimates for C1 coverage ranged from 0.0 to 0.08 compared with a range of 0.0 to 0.12 in plans covering C2. At 37.5 Gy, C1 coverage resulted in predicted rates of 0.0 to 0.12 vs 0.01 to 0.21 in plans covering C2. CONCLUSIONS: Clinicians who choose to cover C2 with WBRT should be aware of the increased radiation dose delivered to the parotid glands. NTCP modeling predicts an increased rate of treatment-associated xerostomia as a result. (P139) Clinical Outcomes and Patient-Reported Outcomes After Local Treatment for High-Risk, Localized Prostate Cancer Lora S. Wang, MD, Colin T. Murphy, MD, Tianyu Li, MS, Marc C. Smaldone, MD, Matthew E. Johnson, MD, Mark Hallman, MD, Yu Ning Wong, MD, Daniel Geynisman, MD, Mark L. Sobczak, MD, David Chen, MD, Eric Horwitz, MD; Fox Chase Cancer Center ARS PROCEEDINGS 2015 87 American Radium Society Scientific Papers and Posters 2015 INTRODUCTION: Men with high-risk prostate cancer (CaP) have low rates of disease control and long-term survival compared with their low-risk counterparts. We sought to investigate CaP-specific and patient-reported outcomes for high-risk CaP men treated with prostate-directed therapy. METHODS: From 1989–2011, a total of 1,091 patients with localized, high-risk CaP treated at our institution were analyzed. All patients were staged clinically (prostate-specific antigen [PSA] > 20 ng/mL, biopsy Gleason score > 8, or > clinical T3) and were included if they received radical prostatectomy (RP) plus radiation (RT), RT plus androgen deprivation therapy (ADT), and RT alone. Cox proportional hazards regression models and Kaplan-Meier estimates were used to assess the risk of biochemical failure (BF), distant metastasis (DM), cause-specific mortality (CSM), and overall mortality (OM). Patient-reported outcomes included International Prostate Symptom Score (IPSS) results, Sexual Health Inventory for Men (SHIM) scores, and ADT side effects. RESULTS: There were 315 patients who received RT alone, 681 who received RT + ADT, and 95 who received RP + RT identified. Median follow-up was 68 months. Men who underwent surgery were significantly younger than those who received RT alone or RT + ADT (P < .001). Patients in the RT + ADT cohort had significantly more advanced T-stages and fewer patients with only one high-risk feature compared with men who received RP + RT (P < .001). Men in the RT + ADT group had significantly lower 5-year BF rates (19% vs 36% for RT alone vs 40% for RP + RT), but patients in the RP + RT arm had significantly lower 5-year OM rates (0% vs 14% for RT alone vs 15% for RT + ADT). After adjusting for covariates, patients in the RT + ADT cohort were less likely to have BF compared with the RP + RT group (P < .001), with no significant difference in the development of DM or CSM. Receipt of RT, either alone (P = .006) or with ADT (P = .010), was associated with a significantly increased risk of OM compared with men who received surgery. There were significantly fewer men in the RT-alone group who experienced any ADT side effects (P < .001) compared with men who received RT + ADT or RP + RT. There were 446 patients with evaluable IPSS scores, with median baseline scores of 7, 7, and 4 for RT alone, RT + ADT, and RP + RT, respectively (P = .002). The median baseline SHIM scores were 5, 19, and 15 for RP + RT, RT alone, and RT + ADT, respectively (P < .001), with 96 evaluable patients. CONCLUSIONS: Long-term CaP-specific survival is equally high after RT + ADT and RP + RT in clinically high-risk CaP patients. Further investigation should be aimed at integrating quality of life measures when considering the optimal treatment for men with high-risk CaP. (P140) Clinical Predictors of Survival for Patients With Stage IV Cancer Referred to Radiation Oncology Emily Copel, DO, Johnny Kao, MD, Kenneth Gold, MD, Gina Zarilli, MD, Samuel Ryu, MD, David Yens, PhD; Good Samaritan Medical Center; State University of New York, Stony Brook; New York College of Osteopathic Medicine 88 PURPOSE: To develop a clinically useful predictive model for survival in a heterogeneous group of patients with metastatic cancer referred to radiation oncology. PATIENTS AND METHODS: From May 2012 to August 2013, a total of 143 consecutive patients with stage IV cancer were prospectively evaluated by a single radiation oncologist. We retrospectively analyzed the effects of 29 patient-, laboratory-, and tumor-related prognostic factors on overall survival (OS), using univariate analysis. Variables that were statistically significant on univariate analysis were entered into a multivariable Cox regression model to identify independent predictors of long-term survival. RESULTS: The median OS was 5.6 months. Five prognostic factors significantly predicted survival on multivariable analysis: Eastern Cooperative Oncology Group (ECOG) performance status (0/1 vs 2 vs 3/4), number of active tumors (1–5 vs ≥ 6), albumin levels (≥ 3.4 mg/dL vs 2.4–3.3 mg/dL vs < 2.4 mg/dL), altered mental status (no vs yes), and primary tumor site (breast, kidney, or prostate vs other). Risk group stratification was performed by assigning points for adverse prognostic factors, resulting in favorable-, standard-, and poor-risk groups. The median survival was 21.8 months for favorable risk vs 5.5 months for standard risk and 1.5 months for poor risk (P < .001). CONCLUSIONS: These data suggest that a model that considers performance status, extent of disease, serum albumin, mental status, and primary tumor site represents a clinically relevant tool in radiation oncology to predict survival for patients with stage IV cancer. (P141) Rectal Spacer Injection in Postprostatectomy Patients Undergoing High-Dose Salvage External Beam Radiation Jekwon Yeh, MD, Kenneth Tokita, MD, Jeffrey Chien, John Ravera, MD; Cancer Center of Irvine OBJECTIVE/PURPOSE: There is now increasing literature to support the use of a rectal spacer to decrease rectal side effects during radiation therapy (RT) for patients with intact prostate cancer. There have been no reports on its usage in patients who are undergoing salvage RT after a prostatectomy. This study aims to report on the gastrointestinal toxicity of patients who have had a prostatectomy and received high-dose (>72 Gy) salvage RT with the rectal spacer. MATERIALS AND METHODS: From January 2010 to October 2013, a total of 32 patients had the rectal spacer placed via transperineal injection posterior to the residual Denonvillier’s fascia under ultrasound guidance. All patients were treated to a minimum of 72 Gy to the prostatic fossa with intensity-modulated RT (IMRT) with daily cone-beam computed tomography (CT) to ensure treatment accuracy. All patients had at least 6 months of follow-up. Radiation Therapy Oncology Group (RTOG) scoring for gastrointestinal morbidity was assessed at the end of treatment and 6 months afterward. RESULTS: At the end of treatment, 23 patients (72%) had no change American Radium Society Scientific Papers and Posters 2015 in rectal symptoms. Nine patients (28%) developed grade 1 gastrointestinal (GI) toxicity. No patients developed grade ≥ 2 GI toxicity. At 6 months after treatment, 29 patients (91%) were back to their baseline GI function, with only 3 patients (9%) with residual grade 1 GI toxicity. No patients developed grade ≥ 2 GI toxicity. CONCLUSIONS: Our study shows that rectal spacer placement in patients who have had a prostatectomy is feasible. Despite the high dose of salvage RT, GI toxicity rate was low, and no patient developed grade ≥ 2 GI toxicity with the rectal spacer. (P142) Older Patients Derive Greater Benefit From Adjuvant and Neoadjuvant Radiotherapy in Diverse Solid Malignancies Noah K. Yuen, MD, Arta Monjazeb, MD, PhD, Chin-Shang Li, PhD, Dariusz Borys, MD, Richard Bold, MD, Robert Canter, MD; University of California, Davis Medical Center BACKGROUND: Radiation therapy (RT) is a standard component in the contemporary multimodality management of numerous solid malignancies. Increasing studies have shown that age-related immunologic changes may impact the bioactivity of RT. We hypothesized that outcomes of RT would be influenced by age across various solid cancers. METHODS: Using Surveillance, Epidemiology, and End Results (SEER) data (1990–2011), we identified 959,731 adult patients (aged > 18 yr) with common nonmetastatic solid malignancies, including breast, lung, and rectal cancer, undergoing surgery ± RT. We compared patient demographics, tumor characteristics, and treatments by age. Multivariable analyses were used to examine the effect of these variables on overall survival (OS) and disease-specific survival (DSS). Hazard ratios were calculated based on multivariable Cox proportional hazards models and logistic regression analysis. RESULTS: The study cohort consisted of 70.0% breast, 13.5% lung, 11.9% rectal, 3.4% sarcoma, and 1.2% esophageal cancer. Mean age at diagnosis was 61.1 ± 13.8 years, and 42% of patients were aged ≥ 65 years. A total of 43.2% received either adjuvant (39.0%) or neoadjuvant (4.2%) RT. With the exception of lung cancer (OS hazard ratio [HR] = 1.21; 1.19–1.23), RT was associated with improved survival in patients of all ages: breast (OS HR = 0.72; 0.71–0.73), rectal (OS HR = 0.81; 0.79–0.83), esophageal (OS HR = 0.83; 0.75– 0.91), and sarcoma (OS HR = 0.63; 0.56–0.70). These positive effects were amplified in elderly patients (aged ≥ 65 yr), with breast (OS HR = 0.66; 0.65–0.66) and rectal cancers (OS HR = 0.74; 0.72– 0.76) having the most benefit. On logistic regression, the HR for risk of death per year of age was lower in patients receiving RT vs those not receiving RT: breast (1.057–1.060 vs 1.072–1.074), rectal (1.043–1.048 vs 1.081–1.085), esophageal (1.006–1.023 vs 1.026– 1.046), and lung cancer (1.039–1.047 vs 1.050–1.054). CONCLUSIONS: This is the largest analysis to date examining the effects of RT in a broad population of solid tumor patients. Although we were unable to control for chemotherapy use in this cohort, RT was associated with superior oncologic outcome compared with surgery alone among patients with multiple solid malignancies. This positive effect was most pronounced in older patients with breast and rectal cancer, suggesting age-dependent effects of RT. Further investigation into the mechanism of these age-related effects is indicated. (P143) Multisite Review of Twenty-Six Head and Neck Cancer Patients Who Have Developed Osteoradionecrosis: Location, Etiology, and Treatment Rex Hoffman, MD, Daniel Copps, DDS, Earl Freymiller, MD, DDS, Sopirios Tetradis, DDS, PhD; Roy and Patricia Disney Family Cancer Center, Providence Saint Joseph Medical Center; private practice; UCLA School of Dentistry INTRODUCTION: Osteoradionecrosis (ORN) is felt to be secondary to decreased blood flow through the inferior alveolar artery to the mandible. It has been well documented that if the dose of radiation exceeds a certain threshold, patients are at increased risk for this devastating treatment-related side effect. The effect that systemic therapy has on this altered blood flow is less clear. The medical and dental literature has conflicting reports that the rate of ORN increases or decreases when systemic therapy (chemotherapy or Erbitux) has been given with radiation. Here, the authors review 26 head and neck cancer patients who have developed ORN to see if systemic therapy increased their risk during radiation. MATERIALS: A total of 26 head and neck cancer patients, treated at six different radiation oncology departments, were diagnosed with ORN by a single maxillofacial prosthodontist (DC), a dental radiologist, and a maxillofacial oral surgeon. ORN was defined as a condition of nonviable bone in the site of radiation injury. Diagnosis of ORN was made by both physical exam and radiographic imaging, which consisted of both a Panorex and a conebeam computed tomography (CT) scan. For the purpose of this study, the patients’ records were reviewed, looking specifically at how radiation and systemic therapy had been delivered (induction vs concurrent), as well as what systemic agent(s) they received during radiation. RESULTS: Median time to diagnosis of ORN after treatment by both physical exam and radiographic imaging was 77 months (range: 12–147 mo). Of the 26 patients, 5 had received radiation therapy alone, 7 had received induction chemotherapy followed by concurrent chemotherapy and radiation, and 14 had been treated with concurrent therapy (either Erbitux or platinum-based chemotherapy) and radiation. In each case, necrosis of the bone occurred in the posterior aspect of the mandible. None of the cases was iatrogenic. After the diagnosis was made, treatment consisted of either pharmacologic treatment or hyperbaric oxygen, followed by conservative oral surgery or hemimandibulectomy. CONCLUSION: The medium time from the completion of treatment to diagnosis of ORN was over 6 years. All patients in this review developed ORN in the posterior aspect of the mandible. Neither the specific systemic agent (Erbitux or chemotherapy) nor the manner in which the agent was delivered (induction vs concurrent) appeared to increase the risk of ORN. It is hoped that having a better understanding as to the location, etiology, and treatment of ORN will help to minimize this potentially devastating complication for future generations of head and neck cancer patients. ARS PROCEEDINGS 2015 89 American Radium Society Scientific Papers and Posters 2015 (P144) Radiation-Associated Toxicities in Obese Women With Endometrial Cancer: More Than Just BMI? Savita Dandapani, MD, PhD, Ying Zhang, MD, Richard Jennelle, MD, Yvonne Lin, MD; City of Hope; USC PURPOSE: The current study characterizes the impact of obesity on postoperative radiation-associated toxicities in women with endometrial cancer (EC). MATERIALS AND METHODS: A retrospective cohort study identified 96 women with EC referred to a large urban institution’s radiation oncology practice for postoperative whole-pelvic radiotherapy (WPRT) and/or intracavitary vaginal brachytherapy (ICBT). Demographic, clinicopathologic, and patient-reported toxicity data were obtained from medical records. Anthropometric information, including body mass index (BMI), was collected from nursing intake records at each clinic visit. Radiation-related toxicities were graded according to Radiation Therapy Oncology Group (RTOG) Acute Radiation Morbidity Scoring Criteria. The follow-up period ranged from 1 month to 11 years (median 2 yr). The data were analyzed by chi-square, logistic regression, and recursive partitioning analyses. RESULTS: A total of 68 evaluable EC patients who received WPRT and/or ICBT were included in the primary analysis. The median age was 52 years (range: 29–73 yr). The majority of patients were Hispanic (48, 71%), with 6 (9%) Caucasian, 1 (1%) AfricanAmerican, and 13 (19%) Asian. The median BMI at diagnosis was 34.5 kg/m2 (range: 20.5–56.6 kg/m2). A total of 58 patients (85%) had abdominal hysterectomies, and 10 (15%) had laparoscopic hysterectomies; 43(63%) had a pelvic lymphadenectomy, and 15 (22%) had para-aortic lymphadenectomies. BMI was independently associated with reported radiation-related cutaneous toxicities (P = .022) and gynecologic toxicities (P = .027). Younger women also reported more gynecologic toxicities (P = .039). Adjuvant radiation technique was associated with increased gastrointestinal- and genitourinary-related toxicities but not gynecologic toxicity. There was no association of International Federation of Gynecology and Obstetrics (FIGO) stage, use of adjuvant chemotherapy, or hysterectomy type with reported radiation toxicities. CONCLUSIONS: The impact of obesity on adjuvant treatment is poorly understood. Increasing BMI was associated with increased frequency of gynecologic and cutaneous radiation-associated toxicities. Additional studies to critically evaluate the radiation treatment dosing and treatment fields in obese EC patients are warranted to identify strategies to mitigate the radiation-associated toxicities in these women. 90