Lynch Syndrome screening in Vancouver: A public health approach

LYNCH SYNDROME SCREENING IN
VANCOUVER
A PUBLIC HEALTH APPROACH
ROBERT WOLBER MD, MPH
LYNCH SYNDROME
ALDRED SCOTT WARTHIN
1866-1931
HENRY T. LYNCH
1928-
KINDRED ANALYSIS 1895-2000: SINGLE PROGENITOR WITH 929 KNOWN
DESCENDANTS FOLLOWED OVER 7 GENERATIONS (MSH2 MUTATION)
LYNCH SYNDROME
HEREDITARY MUTATION OF DNA MISMATCH REPAIR GENE (ONE COPY)
WITH SOMATIC LOSS OF THE 2nd COPY AND ENZYME DELETION IN TUMOUR
CELLS RESULTS IN A 10-TO 50-FOLD INCREASED CANCER RISK
50%-75% LIFETIME RISK FOR COLORECTAL CANCER (VERSUS 6%).
ACCOUNTS FOR 3%-7% OF CANCERS AT THIS SITE
60% LIFETIME RISK FOR ENDOMETRIAL CANCER (VERSUS 4%). ACCOUNTS
FOR 3%-5% OF CANCERS AT THIS SITE
MEAN AGE AT CANCER DIAGNOSIS 42-61 YEARS (VARIABLE: FAMILY G
BRANCHES VARIED FROM 46 TO 59 YEARS)
MODESTLY INCREASED RISK FOR TCC OF RENAL PELVIS AND URETER;
ADENOCARCINOMA OF STOMACH, DUODENUM AND OVARY
30% OF PERSONS WITH AN IDENTIFIED GERMLINE MISMATCH REPAIR
ENZYME MUTATION DO NOT MEET AMSTERDAM CRITERIA
30% OF PERSONS MEETING AMSTERDAM CRITERIA ARE NOT FOUND TO
HAVE AN MMR DEFECT AND DO NOT HAVE MSI-HIGH TUMOURS: “FAMILIAL
COLORECTAL CANCER TYPE X”
DE-NOVO MMR MUTATION PREVALENCE (AMONG MMR MUTATED PATIENTS)
ESTIMATED AT 2%-3%
WELL-KNOWN HEREDITARY CANCER SYNDROMES
SYNDROME
GENE
PENETRANCE (CANCER RISK)
FREQUENCY
XERODERMA
PIGMENTOSUM
XP GENE
COMPLEX
MELANOMA, SQUAMOUS CA:
100%
1/250,000
LILI-FRAUMENI
P53
SARCOMA, LEUKEMIA
BREAST CA: 90%
1/50,000
VON HIPPELHIPPEL-LINDAU
VHL
HEMANGIOBLASTOMA,
RENAL CA: 80%
1/35,000
MULTIPLE
ENDOCRINE
NEOPLASIA (MEN)
MEN / RET
MEDULLARY THYROID CA,
PHEOCHROMOCYTOMA: 95%
1/30,000
RETINOBLASTOMA
(HERITABLE)
RB 1
RETINOBLASTOMA: 50%
(VARIABLE)
1/20,000
FAMILIAL
ADENOMATOUS
POLYPOSIS (FAP)
APC
COLORECTAL CA: 100%
1/15,000
LYNCH SYNDROME*
MLH1, MSH2, PMS2,
MSH6 (98.8%)
COLON + ENDOMETRIUM +
OVARY + UGI + PANCREAS +
URETER CARCINOMA = 80%
1/300
(1/1000(1/1000-1/200)
HEREDITARY
BREASTBREAST-OVARIAN *
BRCA 1 / BRCA 2
BREAST CANCER: 60%
SEROUS OVARIAN CA: 60%
1/300
(1/1000(1/1000-1/100)
* DNA REPAIR DEFECT: MULTIORGAN, MANIFEST IN MID-LIFE, HIGH PENETRANCE AND FREQUENCY
WHY IDENTIFY LYNCH SYNDROME?
CHEMOTHERAPY DECISIONS: STAGE II AND III MSI-HIGH/ MMR
DEFICIENT COLON CANCERS MAY HAVE AN ADVERSE OUTCOME
WITH STANDARD 5-FU THERAPY
SURGICAL TREATMENT DECISIONS: PATIENTS WITH GERMLINE
MUTATIONS IN MMR GENES MAY CHOOSE TOTAL/SUBTOTAL
COLECTOMY AT TIME OF DIAGNOSIS
GENETIC COUNSELLING: IDENTIFICATION OF LYNCH KINDREDS
ALLOWS EARLY SCREENING AND PROPHYLACTIC SURGERY:
PROVEN 67% REDUCTION OF CANCER INCIDENCE
THERE IS NO DOWN SIDE TO RULING OUT LYNCH SYNDROME
Summary of Recommendations:
The Evaluation of Genomic Applications in Practice
and Prevention (EGAPP) Working Group: “found
sufficient evidence to recommend offering genetic
testing for Lynch syndrome to individuals with newly
diagnosed colorectal cancer to reduce morbidity and
mortality in relatives” (Genet Med 2009:11(1):35–41)
Report From the Jerusalem Workshop on Lynch
Syndrome: “recommended that all incident CRC and
endometrial cancer cases should be screened for
the disease” (Jerusalem, October 26–27, 2009)
MISMATCH REPAIR
IMMUNOHISTOCHEMISTRY
MLH 1 AND MSH 2 TOGETHER IDENTIFY 90-95% OF
GERMLINE MUTATIONS ASSOCIATED WITH MSIHIGH TUMOURS
MSH 6 AND PMS 2 PICK UP AN ADDITIONAL 3-10%
NORMAL PHENOTYPE = ALL 4 PRESENT
ABNORMAL PHENOTYPE = ANY ONE ABSENT
DETAILS:
– IF MMR IS NOT PRESENT IN BACKGROUND TISSUE, THEN
AN ABNORMAL STUDY IS NOT VALID
– IF MLH1 IS LOST THEN PMS 2 CANNOT BE INTERPRETED
(IT WILL BE NEGATIVE BECAUSE PMS2 PROTEIN IS
UNSTABLE IN THE ABSENCE OF MLH1)
– IF MSH 2 IS LOST THEN MSH 6 CANNOT BE INTERPRETED
– IF MLH1 OR MSH2 LOST WITHOUT LOSS OF PMS2 OR
MSH6, THEN STUDY SHOULD BE REPEATED
MLH 1
NORMAL
DELETED
HOW DOES AN ABNORMAL MMR PROFILE
RELATE TO GERMLINE GENE MUTATION?
A FALSE NORMAL MMR-IHC IN THE PRESENCE OF
MUTATION IS UNCOMMON :
SUBSTITUTION MUTATIONS MAY PRODUCE
IMMUNOREACTIVE MMR PROTEINS THAT ARE NONFUNCTIONAL
5 ADDITIONAL MMR GENES HAVE BEEN IDENTIFIED
(BUT ACCOUNT FOR ONLY 0.2% OF LYNCH
PATIENTS)
EVEN MMR MUTATED PATIENTS MAY DEVELOP
CANCER BY THE MORE COMMON CHROMOSOMAL
INSTABILITY PATHWAY
A FALSE ABNORMAL MMR-IHC IN THE ABSENCE OF
MUTATION IS COMMON:
MLH1 PROTEIN IS INACTIVATED BY SOMATIC B-RAF
V600 MUTATIONS IN UP TO 63% OF MLH1-IHC
DELETED, NON-MUTATED CASES
B-RAF IS MUTATED IN ONLY 1% OF MLH1 MUTATED
CASES: GOOD SCREENING TOOL TO RULE OUT
MLH1 LYNCH SYNDROME
ANY OF THE 4 COMMON MMR GENES MAY BE
INACTIVATED BY PROMOTER HYPERMETHYLATION
WITHOUT MUTATION: AMONG MLH1-IHC DELETED,
NON-MUTATED CASES 47% WILL HAVE “C” REGION
PROMOTER METHYLATION
6% OF MLH1 MUTATED CASES WILL HAVE “C”
REGION METHYLATION: NOT SO GOOD FOR RULING
OUT LYNCH SYNDROME
J Med Genet. 2012;49(3):151-157
HOW GOOD IS MMR-IHC TESTING IN AN
UNSELECTED TUMOUR POPULATION?
-THE BEST SENSITIVITY OF MMR-IHC FOR DETECTION OF GERMLINE
MMR MUTATION ESTIMATED IN 2009 AT 95%. THIS APPROACHES THE
THEORETICAL LIMIT OF SENSITIVITY FOR MMR IHC TESTING
-MMR-IHC SPECIFICITY FOR MMR MUTATION IS BIOLOGICALLY LIMITED
ON AN AGE-SPECIFIC BASIS (METHYLATION INCREASES WITH AGE)
-IF THE PREVALENCE OF MMR MUTATION IN COLON CANCER IS 3-7%,
THE PREDICTIVE VALUE OF A NORMAL MMR-IHC TEST FOR THE
ABSENCE OF A GERMLINE MMR MUTATION APPROACHES 99%
-PREDICTIVE VALUE OF AN ABNORMAL MMR-IHC FOR THE PRESENCE
OF A GERMLINE MMR MUTATION (ALL AGES) ESTIMATED AT 20% 30%: AT LEAST 2/3 ARE DUE TO B-RAF OR HYPERMETHYLATION
-ALMOST ALL MMR-IHC ABNORMAL TUMOURS ARE MSI-HIGH (EXCEPT
FOR MSH6)
WHERE TO START
WE TESTED MANY CLONES ON MANY CASES...
MARKER CLONE
MLH1
G168-728
ES05
ES05
PMS2
EP51
A16
EPR3947
A16-4
C-20
MRQ-28
MOR4G
MSH2
G2191129
25D12
MSH6
EP49
BC/44
44
PU29
VENDOR
CELL MARQUE
LEICA
DAKO
EPITOMICS
BIOCARE
EPITOMICS
BECTONDICKINSON
SANTA CRUZ
CELL MARQUE
LEICA
CELL MARQUE
LEICA
EPITOMICS
BIOCARE
CELL MARQUE
LEICA
VENDOR ID
285M-14
NCL-L-MLH1
M3640
AC-0049
CM344AK
2858-S
556415
COMMENTS
HAPPY @ 1/25
HAPPIEST @ 1/25
NOT HAPPY
HAPPIEST @ 1/25
HAPPY @ 1/25
DIRTY
NOT HAPPY
WEAK
288M-14
WEAK
NCL-L-PMS2 REALLY WEAK
286M-14
HAPPIEST@1/1000
NCL-MSH2
NOT SO MUCH
AC-0047
HAPPIEST@1/200
CM265A
HAPPY @ 1/200
287M-14
HAPPY@ 1/400
NCL-L-MSH6 NOT HAPPY
N > 1000. MANY THANKS TO DEANNA, DAVID AND BEV
MLH1 AND PMS2 ANTIGENS ARE “ISCHEMIC TIME” LABILE: THEY ARE
MORE STRONGLY EXPRESSED IN BIOPSIES THAN IN RESECTIONS
MLH1 BIOPSY
MLH1 RESECTION
PMS2 BIOPSY
PMS2 RESECTION
MSH2 BIOPSY
MSH2 RESECTION
RESECTION
MSH6 BIOPSY
MSH6 RESECTION
MMR-IHC TESTING PLAN FOR VCH
WHAT WOULD BE THE COST FOR VCH OF SCREENING EVERY
NEWLY DIAGNOSED COLORECTAL, ENDOMETRIAL AND NONSEROUS OVARIAN CANCER?
HIGH ESTIMATE:
1000 INCIDENT VCH CASES PER YEAR x REAGENT COST
$80/CASE (VENTANA DETECTION SYSTEM) = $80,000
IF 7% OF COLORECTAL AND 5% OF ENDOMETRIAL CANCERS
ARE LYNCH, WE MIGHT EXPECT 70 “AT RISK” PATIENTS TO
BE IDENTIFIED PER YEAR AT A COST OF $1200/ PATIENT (FOR
US)
TESTING IS ONLY DONE ONCE AT THE TIME OF CANCER
DIAGNOSIS
NOBODY GETS MISSED: AMSTERDAM CRITERIA WILL MISS
30% OF LYNCH PATIENTS AND ALL “EPIGENETIC” MSI-HIGH
TUMOURS
NO “LOOK BACK” TESTING UNLESS THE PATIENT HAS A
RECURRANCE OR AT CLINICIAN REQUEST
DO THE RIGHT THING
TEST EVERY NEW COLORECTAL, ENDOMETRIAL AND NONSEROUS OVARIAN CANCER
NO INFORMED CONSENT- THIS IS NOT A GERMLINE DNA
TEST
TEST DIAGNOSTIC BIOPSY SAMPLES (100% CORRELATION
WITH RESECTION)
STANDARDIZED REPORTING: MMR NORMAL, MMR
ABNORMAL (DEFINED), OR UNDETERMINED
REPEAT STUDIES DONE ON RESECTION SPECIMENS ONLY IF
THE BIOPSY STUDY FAILS
NO USE OF THE TERM MICROSATELLITE INSTABILITY OR MSI
(BECAUSE THAT IS NOT WHAT WE ARE TESTING)
NO TESTING OF ADENOMAS OR HYPERPLASIAS (NO LOH)
UPPER GI AND UROTHELIAL CANCERS MAY BE TESTED ON
REQUEST
IMPLEMENT B-RAF TESTING FOR MLH1 DELETED TUMOURS
DEVELOP AND PARTICIPATE IN A TMA-BASED PROFICIENCY
TESTING PROGRAM
NOTIFY THE BC HEREDITARY CANCER PROGRAM OF THIS
INITIATIVE (1 YEAR IN ADVANCE)
VCH MMR REPORTING
-MMR normal:
“Immunohistochemical stains for mismatch repair proteins (MLH1, PMS2, MSH2 and
MSH6) reveal a normal MMR profile. All proteins tested are present.”
-MMR abnormal – MLH1 deleted:
“Immunohistochemical stains for mismatch repair proteins (MLH1, PMS2, MSH2 and
MSH6) reveal an abnormal MMR profile. MLH1 is deleted. PMS2 cannot be
interpreted due to linkage with MLH1. This patient may benefit from referral to the
BCCA Hereditary Cancer Program.”
-MMR abnormal – MSH2 deleted:
“Immunohistochemical stains for mismatch repair proteins (MLH1, PMS2, MSH2 and
MSH6) reveal an abnormal MMR profile. MSH2 is deleted. MSH6 cannot be
interpreted due to linkage with MSH2. This patient may benefit from referral to the
BCCA Hereditary Cancer Program.”
-MMR abnormal – MSH6 deleted:
“Immunohistochemical stains for mismatch repair proteins (MLH1, PMS2, MSH2 and
MSH6) reveal an abnormal MMR profile. MSH6 is deleted. This patient may benefit
from referral to the BCCA Hereditary Cancer Program.”
-MMR abnormal – PMS2 deleted:
“Immunohistochemical stains for mismatch repair proteins (MLH1, PMS2, MSH2 and
MSH6) reveal an abnormal MMR profile. PMS2 is deleted. This patient may benefit
from referral to the BCCA Hereditary Cancer Program.”
-MMR inconclusive:
“Immunohistochemical stains for mismatch repair proteins (MLH1, PMS2, MSH2 and
MSH6) are technically inconclusive. Repeat studies should be performed on the
subsequent resection specimen.”
SUMMARY OF VCH TESTING JANUARY 2012
TO MARCH 2013
747 PATIENTS TESTED, 3 SITES, 20 PATHOLOGISTS
747 INFORMATIVE RESULTS (NO INCONCLUSIVES)
10 REPEATS
154 ABNORMAL (20.6%)
118 MLH1 (77%)
20 MSH2 (13%)
11 MSH6 (7%)
5 PMS2 (3%)
1 CASE MLH1 + MSH6
MEAN AGE OF NORMAL AND ABNORMAL GROUPS = 65 YEARS
ACTUAL COST PER TEST PANEL (VENTANA DETECTION
SYSTEM) = $70/ CASE x 685/ YEAR = $48,000
MMR AGE DISTRIBUTION
180
160
140
CASES
120
100
abnormal
normal
80
60
40
20
0
21-30
decade
%abnormal
21-30
50%
31-40
41-50
31-40
39%
51-60
41-50
13%
61-70 71-80 81-90 91-100
51-60
18%
61-70
20%
71-80
22%
81-90
22%
91-100
0%
RECOMMENDATIONS TO OUR SURGEONS,
GYNECOLOGISTS, GASTROENTEROLOGISTS
MMR ABNORMAL = REFER TO BC HEREDITARY
CANCER PROGRAM FOR BRAF AND/OR GERMLINE
MUTATION ANALYSIS
STRONG FAMILY HISTORY, MMR NORMAL =
REFER TO BC HEREDITARY CANCER PROGRAM
FOR MSI AND POSSIBLE GERMLINE ANALYSIS
MUTATION IDENTIFIED = POSSIBLE SCREENING
OF BLOOD RELATIVES (NOT CHILDREN)
NO FAMILY HISTORY, MMR NORMAL = NO
FURTHER ACTION
TO DATE, NOT A SINGLE MMR ABNORMAL
PATIENT REFERRED TO THE BCHCP HAS HAD
CONFIRMATORY TESTING
LETTER FROM VCH ADMINISTRATION
DECEMBER 2012
“We have been requested by the BCCA (British Columbia
Cancer Agency) to discontinue the current protocol of
immunohistochemical testing being used at VCH and to apply
the current BCCA eligibility criteria (age<50, Amsterdam).”
“There are some significant issues related to “one off”
implementation of new testing criteria, including: differing
standards being applied to the province, cost effectiveness of
new testing modalities in a resource constrained environment,
the ethics and appropriateness of testing when there has been
made no provision for recommended follow-up.”
WHAT ARE THE IMPLICATIONS OF THIS
PROGRAM?
1/300 = 16,000 INDIVIDUALS IN BC LIKELY HAVE A GERMLINE
MISMATCH REPAIR MUTATION (4000 IN VCH): THEY ARE
STALKED BY A KILLER
SCREENING OF 4.8 MILLION BY GENE SEQUENCING WILL
NEVER HAPPEN IN OUR LIFETIME
BY SCREENING ONLY PATIENTS WITH CANCER, 1/300
BECOMES 1/15
USE OF MMR IHC ($70/CASE) ENRICHES POOL TO 1 IN 3
SUPPLEMENTAL B-RAF q-PCR ($120/CASE) WILL IMPROVE
THIS RATIO TO 1 IN 2
PROVINCE WIDE, MOLECULAR TESTING MIGHT HAVE TO BE
PERFORMED ON 1000 PATIENTS AND KIN PER YEAR X 20
YEARS
THE COST OF GENOMIC TESTING IS DROPPING ($1200/CASE)
ANNUAL COST $1.5 MILLION OVER THE NEXT 20 YEARS = $30
MILLION (OR LESS)
LYNCH SCREENING SIMPLIFIED
1/300 PEOPLE
CANCER
IN 5%
1/15 PATIENTS
MMR-IHC
DELETED
IN 20%
1/3 MMR ABNORMALS
1/2 BRAF NORMALS
B-RAF ON
MLH1 DELETED
CASES
DNA
DONE!
SEQUENCING
“The cost-effectiveness of genetic testing strategies for Lynch syndrome
among newly diagnosed patients with colorectal cancer”
Office of Public Health Genomics, Centers for Disease Control
Genetics in Medicine (2010) 12, 93–104
COMPARISON OF INCREMENTAL COST PER QUALITY
ADJUSTED LIFE-YEAR SAVED ($ per QALY) FOR UNIVERSAL
MMR-IHC + DNA, AGE-TARGETED MMR IHC + DNA AND MSI +
DNA:
TREATING HYPERTENSION = $85,000
BIANNUAL PAP + HPV TEST= $40,000
HPV VACCINATION = $26,000
SEASONAL FLU VACCINATION = $25,000
CHILDHOOD VACCINATION IN DEVELOPING WORLD = $200
WHO: ANYTHING < $70,000 IS WORTH FUNDING
WORLD BANK: ANYTHING < PER CAPITA GNP ($40,000 IN
CANADA)
Genetics in Medicine (2010) 12, 93–104
UNIVERSAL MMR IHC SCREENING IS THE LOWEST
COST STRATEGY
$23,300 / LIFE -YEAR SAVED FOR UNIVERSAL IHC
TESTING FOLLOWED BY DNA SEQUENCING
$22,500 / LIFE- YEAR SAVED FOR UNIVERSAL IHC
FOLLOWED BY BRAF-V600 (MLH1) FOLLOWED BY
DNA SEQUENCING
$38,500 / LIFE-YEAR SAVED FOR IHC TESTING
RESTRICTED TO PATIENTS YOUNGER THAN 50
YEARS FOLLOWED BY SEQUENCING
TARGETING YOUNGER PATIENTS WILL MISS
NEARLY HALF OF ALL LYNCH SYNDROME
PATIENTS
“From an ethical standpoint, discontinuing testing is an
untenable position. As these specimens pass through our
laboratories, we have one chance at identifying those
individuals and kindreds at high risk for Lynch syndrome,
unless VCH would want to fund and implement an open-ended
retrospective initiative at some future date. This is not likely.”
“Universal MMR testing is not a “one-off”; it is the completion
of a pathology report. Epidemiologically and economically, this
strategy has been validated as the most clinically effective and
cost effective.”
“VCH laboratories do not need the permission of the BCCA to
implement procedures that bring it in line with internationally
recognized guidelines, particularly if the Agency lags behind in
this area. The BCCA will not dictate that VCH pathologists
practice beneath what we know to be the professional standard
of care.”
WE CONTINUE TO TEST
“COST EFFECTIVE” IS NOT THE SAME AS COST SAVING
OPEN BAR MODEL OF HEALTH CARE
ECONOMICS
POTENTIAL LONG TERM COST SAVINGS IN B.C. WITH
LYNCH SYNDROME SCREENING
16000 X 0.8 = 12,800 ADVANCED STAGE CANCERS
PREVENTED OVER THE NEXT 20-30 YEARS
$30 MILLION INVESTMENT
ALL OTHER POTENTIAL EXPENDITURES AND
SAVINGS ASIDE, THE COST OF CHEMOTHERAPY
DRUGS ALONE FOR A PATIENT WITH ADVANCED
COLORECTAL OR GYNECOLOGIC CANCER IS
$150,000 OVER 2 YEARS
12,800 x $150,000 = $1.92 BILLION POTENTIAL
SAVINGS IN DRUG COSTS ALONE
$64 SAVED FOR EVERY $1 INVESTED
THIS IS NOT ROCKET SCIENCE
WHY WOULDN’T WE DO THIS?
PUBLIC HEALTH MEASURES ARE OFTEN CONTROVERSIAL
BECAUSE THEY HAVE AN ECONOMIC IMPACT
THE INSTITUTIONS OR INDUSTRIES THAT PAY THE PRICE MAY
NOT BE THE ONES WHO BENEFIT FROM THE NEW
PROTECTIONS
COSTS ARE USUALLY MORE CONCRETE THAN BENEFITS
THE PRICE MAY NEED TO BE PAID NOW WHILE THE BENEFIT IS
REALIZED LATER
IN TIMES OF ECONOMIC CONSTRAINT, PEOPLE ARE OFTEN
UNWILLING TO PAY SHORT-TERM COSTS TO OBTAIN LONG
TERM BENEFITS