In vitro fertilization, endometriosis, nulliparity and ovarian cancer risk ⁎ ,

Gynecologic Oncology 128 (2013) 260–264
Contents lists available at SciVerse ScienceDirect
Gynecologic Oncology
journal homepage: www.elsevier.com/locate/ygyno
In vitro fertilization, endometriosis, nulliparity and ovarian cancer risk
Louise M. Stewart a,⁎, C. D'Arcy J. Holman a, Patrick Aboagye-Sarfo a, Judith C. Finn b, c,
David B. Preen a, Roger Hart d, e
a
School of Population Health, Faculty of Medicine, Dentistry and Health Sciences, The University of Western Australia, M431, 35 Stirling Highway, Crawley, WA 6009, Australia
School of Primary, Aboriginal and Rural Health Care, Faculty of Medicine, Dentistry and Health Sciences, The University of Western Australia, M516, 35 Stirling Highway, Crawley,
WA 6009, Australia
c
Department of Epidemiology and Preventive Medicine, Monash University, The Alfred Centre, 99 Commercial Road, Melbourne, Vic 3004, Australia
d
School of Women's and Infant's Health, Faculty of Medicine, Dentistry and Health Sciences, The University of Western Australia, King Edward Memorial Hospital, 374 Bagot Road, Subiaco,
WA 6008, Australia
e
Fertility Specialists of Western Australia, Bethesda Hospital, 25 Queenslea Drive, Claremont, WA 6010, Australia
b
H I G H L I G H T S
► Infertile women who remain nulliparous have an increased risk of ovarian cancer.
► In nulliparous, endometriosis increases risk and IVF may also be a risk factor.
► There is little or no increase in risk with IVF or endometriosis in parous women.
a r t i c l e
i n f o
Article history:
Received 3 September 2012
Accepted 25 October 2012
Available online 29 October 2012
Keywords:
In vitro fertilization
Ovarian cancer
Cohort study
Hazard ratios
Risk factors
a b s t r a c t
Objectives. To examine the risk of invasive epithelial ovarian cancer in a cohort of women seeking treatment
for infertility.
Methods. Using whole-population linked hospital and registry data, we conducted a cohort study of
21,646 women commencing hospital investigation and treatment for infertility in Western Australia in
the years 1982–2002. We examined the effects of IVF treatment, endometriosis and parity on risk of ovarian
cancer and explored potential confounding by tubal ligation, hysterectomy and unilateral oophorectomy/
salpingo-oophorectomy (USO).
Results. Parous women undergoing IVF had no observable increase in the rate of ovarian cancer (hazard ratio
[HR] 1.01; 95% confidence interval [CI] 0.35–2.90); the HR in women who had IVF and remained nulliparous was
1.76 (95% CI 0.74–4.16). Women diagnosed with endometriosis who remained nulliparous had a three-fold increase in the rate of ovarian cancer (HR 3.11; 95% CI 1.13–8.57); the HR in parous women was 1.52 (95% CI 0.34–
6.75). In separate analyses, women who had a USO without hysterectomy had a four-fold increase in the rate of
ovarian cancer (HR 4.23; 95% CI 1.30–13.77). Hysterectomy with or without USO appeared protective.
Conclusions. There is no evidence of an increased risk of ovarian cancer following IVF in women who give
birth. There is some uncertainty regarding the effect of IVF in women who remain nulliparous. Parous women
diagnosed with endometriosis may have a slightly increased risk of ovarian cancer; nulliparous women have a
marked increase in risk.
© 2012 Elsevier Inc. All rights reserved.
Introduction
Sadly, ovarian cancer is often diagnosed too late. There remains a
pressing need to identify early warning signs and groups of women at
increased risk.
⁎ Corresponding author. Fax: +61 6488 1188.
E-mail addresses: [email protected] (L.M. Stewart),
[email protected] (C.D.J. Holman), [email protected]
(P. Aboagye-Sarfo), judith.fi[email protected] (J.C. Finn), [email protected]
(D.B. Preen), [email protected] (R. Hart).
0090-8258/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ygyno.2012.10.023
Subfertile women have been shown to be at increased risk of ovarian
cancer [1,2], and women who undergo treatment with in vitro fertilization (IVF) may be at even greater risk. Only a small number of studies
have examined the relationship between IVF treatment and ovarian
cancer risk [3–9]. Of these, four did not find an association between
IVF and risk of ovarian cancer [4–7], while three found an increased
risk of ovarian cancer in IVF patients; one with extended follow-up
[9], another in a comparison that was restricted to women who gave
birth [3] and a third in women who purchased drugs for IVF, compared
with general population controls [8]. Research in this field is limited by
short periods of follow-up, small numbers of ovarian cancer cases and
L.M. Stewart et al. / Gynecologic Oncology 128 (2013) 260–264
comparisons with the general population which do not allow for adjustment for parity or other important risk factors.
In view of this conflicting evidence and the limitations described and
in light of the poor prognosis for women diagnosed with ovarian cancer
(only around 40% of women survive more than 5 years beyond diagnosis) [10], we believe this important issue warrants further investigation.
The aim of the present study was to examine the risk of invasive epithelial ovarian cancer in a cohort of women seeking infertility treatment,
comparing the risk in women who had IVF with those who did not,
whilst at the same time considering the impact of infertility diagnosis,
in particular endometriosis, and parity, surgical sterilization, hysterectomy, unilateral oophorectomy or salpingo-oophorectomy, age and
socio-economic status.
Materials and methods
The study cohort
The study cohort included all women in Western Australia (WA)
seeking hospital investigation or treatment for infertility during the
years 1982–2002. We have previously described a similar cohort
established using the same methods for a study on IVF and breast cancer
[11]. Women in the study cohort had at least one hospital diagnosis of
infertility or procreative management (ICD-9 628.0 to 628.9 or
ICD-10 N97.0 to N97.9 or ICD-9 V26.1 to V26.9 or ICD-10 Z31.1 to Z
31.9), with the first such diagnosis occurring when they were aged between 20 and 44 years inclusive.
Data sources
The study cohort and outcomes of interest were identified using the
resources of the WA Data Linkage System [12]. Data on exposures and
outcomes was collated in de-identified format from 1980 to 2010 (covering the recruitment period plus 2 years before and 8 years after). Information was extracted from the Hospital Morbidity Data System to
identify the cohort and also to identify relevant diagnoses and surgical
procedures. The Hospital Morbidity Data System collates information
on all inpatient admissions at all hospitals in WA. We retained information on diagnoses of endometriosis and pelvic inflammatory disorders
(PID – ICD-9 codes 614.0 to 614.9; ICD-10 codes N70.1, N70.9, N73.0
to N73.9), and the following procedures: hysterectomy, unilateral oophorectomy or salpingo-oophorectomy (USO), bilateral oophorectomy
or salpingo-oophorectomy (BSO), tubal ligation and IVF treatment as
these factors were believed to potentially influence the risk of ovarian
cancer. IVF cycles were identified using the Hospital Morbidity Data
System and the Reproductive Technology Register. Linkage to the Midwives Notifications System enabled us to identify births; deaths were
identified from the Deaths Register and cancer diagnoses from the
WA Cancer Registry. We attempted to minimise loss to follow-up by
restricting the cohort to women known to be resident in WA. Women
who moved interstate after 1988 (when linkage commenced) were
identified from the WA Electoral Roll; these were excluded from the
study population as were women who had an interstate or overseas address on their hospital records. Socio-economic status using location of
residence was derived from the address recorded on the woman's hospital record at the first infertility diagnosis. The Index of Education and
Occupation was chosen to represent socio-economic status [13].
Explanatory variables
We explored the potential relationship between IVF and ovarian
cancer risk, and considered the variables identified above as possible
confounders in this relationship. All variables were first examined
separately in univariate and age-adjusted models and then, where
appropriate, were included in the final model, according to methods
described by Hosmer, Lemeshow and May [14]. Variables under
261
consideration included IVF, age, socio-economic status, PID, endometriosis, birth, USO, tubal ligation and hysterectomy. Variables that were set
at the start of follow-up (age, socio-economic status, diagnosis of PID or
endometriosis) were entered into the model as fixed covariates; variables that could change status during follow-up (birth, tubal ligation,
USO, hysterectomy and IVF) were entered as time-dependent variables.
In this way, follow-up time was correctly apportioned into time before
and time after exposure. Age was divided into quartiles (20–27, 28–31,
32–35, 36–44 years) and entered into the model as a categorical variable. Socio-economic status was entered into the model as a binary variable with the upper quartile compared to the lower three quartiles
combined.
We captured all births in WA during 1980–2010, as well as previous
births in women confined in WA during this time period, but the small
proportion of births occurring either out of the state or prior to 1980 in
women who did not also give birth in WA were unknown to us. In addition, we found that a small proportion of women had a reversal without prior mention of a tubal ligation. We were able to correctly classify
these women, however it is likely there would also be women who had
a tubal ligation that was unrecorded and no later reversal, opting instead simply to have IVF. We were unable to classify this small proportion of women correctly. Women with PID or endometriosis had these
diagnoses recorded in their hospital records at or before the start of
follow-up. It was possible that other women in the cohort also suffered
from these conditions, though they remained undiagnosed, or were diagnosed later during follow-up.
Outcome variable
The outcome of interest was invasive epithelial ovarian cancer. Borderline ovarian tumours and non-epithelial tumours were excluded as
outcome events, but women who were diagnosed with these were
still included in the study population. Follow-up was censored in
women who were diagnosed with borderline ovarian tumours only if
they underwent a BSO. Otherwise they were considered to be still at
risk of invasive epithelial ovarian cancer and were treated as such in
the analysis.
Data analysis
Hazard ratios (HRs) were estimated using Cox regression models.
Follow-up commenced at the date of first infertility admission and
continued until the date of epithelial ovarian cancer diagnosis, date
of BSO, date of death or censor date (15 August 2010), whichever
came first. Data were analysed using SPSS version 19.
Ethics approval
This study received ethics approval from The University of Western
Australia Human Research Ethics Committee and the Department of
Health WA Human Research Ethics Committee.
Results
The cohort
A total of 22,045 women had a first diagnosis of either infertility or
procreative management between 1982 and 2002 when they were
aged between 20 and 44 years inclusive, and were eligible for inclusion
in the cohort. Of these, 379 were identified as having an interstate address or having moved out of the State and were excluded. A further
20 women were excluded because they were considered not to be at
risk of a diagnosis of ovarian cancer after the start of infertility treatment. These included 13 women who had a BSO before their first infertility admission and 7 women who were diagnosed with ovarian cancer
262
L.M. Stewart et al. / Gynecologic Oncology 128 (2013) 260–264
prior to or within 6 months of their first infertility admission. The final
cohort therefore comprised a total of 21,646 women.
The mean and median ages at the start of follow-up were both
31 years; the mean and median ages at the end of follow-up were both
48 years. The total duration of follow-up was 366,041 person-years
with a mean of 17 years. Ovarian cancer was diagnosed in women in
the cohort when they were aged between 33 and 61 years; the mean
age at diagnosis was 46 years. There were 38 cases of ovarian cancer
in the cohort: 16 in women undergoing IVF and 22 in women not undergoing IVF (Table 1).
Table 2
Potential ovarian cancer risk and protective factors: age-adjusted analysis.
Ovarian cancer risk factors
a
Hazard ratios derived from separate Cox regression models, adjusted for
age at the start of follow-up and including only the variable listed. Each hazard
ratio compares women in the exposed group with all other women in the cohort who did not have the exposure.
b
Women who gave birth to at least one child were compared with women
who did not give birth.
c
PID and endometriosis were both diagnosed at, or prior to the first infertility admission.
d
Women who had a hysterectomy without USO in the same or any other admission.
e
Women who had both a hysterectomy and a USO in the same or separate
admissions.
f
Women who had a USO without hysterectomy in the same or any other admission.
g
Women in the upper quartile of the Index of Education and Occupation
were compared with women in the lower three quartiles combined.
The association between IVF, potential confounding factors and
ovarian cancer rate was examined in age-adjusted analysis (Table 2)
and then in multivariate analysis (Table 3).
The age-adjusted rate of ovarian cancer in women in the cohort
undergoing IVF, compared with women seeking infertility treatment
but not IVF, was 1.40 (95% confidence interval [CI] 0.73–2.68) (Table 2).
Women who gave birth had a reduced rate of ovarian cancer
(Table 2).
The most common diagnoses at baseline in women seeking infertility investigation and treatment were PID and endometriosis. We
did not find an association between PID and ovarian cancer, but we
did note an increased age-adjusted rate of ovarian cancer in women
diagnosed with endometriosis (Table 2).
Women in the upper quartile of socio-economic status, as measured
by the Index of Education and Occupation, had a reduced rate of ovarian
cancer (Table 2).
Of the surgical procedures we identified, we found that tubal ligation was associated with a reduced rate of ovarian cancer. Hysterectomy
performed with or without USO also appeared protective, though CIs
around all these estimates included one. In contrast, USO performed
without hysterectomy was associated with an increased rate of ovarian
cancer (Table 2).
We retained the following variables in the final model: IVF, birth, endometriosis, age and socio-economic status. We examined the relationship between these variables and the rate of ovarian cancer firstly in the
whole cohort, and then separately in women who gave birth and
Exposure
Hazard ratio
(95% CI)a
IVF
Birthb
PIDc
Endometriosisc
Tubal ligation
Hysterectomyd
Hysterectomy with USOe
USOf
High socioeconomic statusg
1.40
0.48
1.02
2.23
0.66
0.55
0.72
4.23
0.52
(0.73–2.68)
(0.25–0.93)
(0.42–2.43)
(0.97–5.12)
(0.26–1.68)
(0.13–2.32)
(0.10–5.27)
(1.30–13.77)
(0.22–1.25)
women who did not. Results of the multivariate analyses are presented
in Table 3.
Consistent with the age-adjusted results presented in Table 2, in the
multivariate model we found that having given birth was protective,
undergoing IVF was associated with a slightly increased rate of ovarian
cancer and the diagnosis of endometriosis was associated with an increased rate of ovarian cancer (Table 3, first column of results).
When we examined women separately according to whether
they had given birth, we found that the effect of both IVF and endometriosis were more pronounced in women who did not give birth
than in women who did (Table 3, second and third columns of results). In women who remained childless, IVF treatment was associated with a 76% increase in the rate of ovarian cancer which was not
Table 1
Characteristics of the study population.a
Characteristic
All women in the cohort
Women undergoing infertility
treatment but not IVF
Women undergoing
IVF treatment
Number of women
Number of women who gave birth
Number of women diagnosed with ovarian cancer
Mean length of follow-upb (years)
Total length of follow-up (years)
Mean age at start of follow-up (years)
Mean age at first birth (years)
Mean age at ovarian cancer diagnosis (years)
Mean age at end of follow-up (years)
Number of women diagnosed with PIDc
Number of women diagnosed with endometriosisc
Number of women who underwent tubal ligation
Number of women who had a hysterectomyd
Number of women who had a hysterectomy with USOe
Number of women who had a USOf
Number of women in the highest quartile of socioeconomic status (%)
21,646
14,907
38
16.9 ± 5.9
366,041
31.2 ± 5.2
29.6 ± 6.0
46.0 ± 7.0
48.0 ± 7.1
3,890
2,978
3,740
2,188
696
500
5,268 (24%)
14,098
10,032
22
17.0 ± 5.9
240,203
30.8 ± 5.3
28.3 ± 5.8
46.7 ± 8.2
47.7 ± 7.2
2,576
1,914
2,856
1,599
488
276
2,928 (21%)
7,548
4,875
16
16.7 ± 5.9
125,837
32.1 ± 4.8
32.2 ± 5.4
44.9 ± 4.8
48.7 ± 6.9
1,314
1,064
884
589
208
224
2,340 (31%)
a
The study cohort comprised a total of 21,646 women commencing hospital investigation and treatment for infertility between 1982 and 2002. Information on exposures and
outcomes was collected over a period of 30 years, from 1980 to 2010.
b
All means expressed ± SD.
c
PID and endometriosis were both diagnosed at, or prior to the first infertility admission.
d
Women who had a hysterectomy without USO in the same or any other admission.
e
Women who had both a hysterectomy and a USO in the same or separate admissions.
f
Women who had a USO without hysterectomy in the same or any other admission.
L.M. Stewart et al. / Gynecologic Oncology 128 (2013) 260–264
263
Table 3
Multivariate regression models describing the relationship between ovarian cancer and identified risk and protective factors.a
Factor
Hazard ratio (95% CI)
In the whole cohort (n = 21,646)
Hazard ratio (95% CI)
In women who gave birth (n = 14,907)
Hazard ratio (95% CI)
In women who did not have a recorded birth (n = 6,739)
Birth
IVF
Endometriosis
0.49 (0.25–0.95)
1.36 (0.71–2.62)
2.33 (1.02–5.35)
–
1.01 (0.35–2.90)
1.52 (0.34–6.75)
–
1.76 (0.74–4.16)
3.11 (1.13–8.57)
a
Model 1 (first column of results) includes all women in the cohort; model 2 (second column of results) includes only the sub-group of women who gave birth; model 3 (third
column of results) includes only women who did not give birth. Hazard ratios in each are derived from multivariate Cox regression models that include all the variables listed and
are adjusted for age at the start of follow-up and socio-economic status.
statistically significant (HR 1.76; 95% CI 0.74–4.16) while a diagnosis of
endometriosis was associated with a three-fold increase in the rate of
ovarian cancer (HR 3.11; 95% CI 1.13–8.57) (Table 3, third column of
results).
Discussion
The results of this study suggest that IVF treatment has no effect on
the risk of ovarian cancer in women who give birth, but uncertainty still
surrounds the question of whether IVF contributes to an increased risk
in women who remain childless. Parous women diagnosed with endometriosis may have a slight increase in the risk of ovarian cancer; nulliparous women face a three-fold increase in risk.
Women in our study who had IVF and gave birth had the same risk
of ovarian cancer as women who had non-IVF infertility treatment (HR
1.01; 95% CI 0.35–2.90). Nulliparous women observed in our study may
have had a modest increase in the risk of ovarian cancer after IVF (HR
1.76); however, there is uncertainty over statistical inference and interpretation. With the small numbers of ovarian cancer cases typical of
such studies, this estimate was imprecise and potentially consistent
with a wide range of true values including the null value of 1 (95% CI
0.74–4.16). Parous women diagnosed with endometriosis may have
had a slight increase in the risk of ovarian cancer (HR 1.52; 95% CI
0.34–6.75); nulliparous women had a three-fold increase (HR 3.11;
95% CI 1.13–8.57).
Previous studies have rarely examined nulliparous women separately from parous women, despite the clear association between parity
and ovarian cancer risk. None of the published papers on IVF and ovarian cancer considered parous and nulliparous women separately; indeed, none adjusted for confounding by parity, perhaps because it was
not possible in studies that involved comparisons with the general population and had only small numbers of ovarian cancer cases (between 1
and 13) [4–8]. Even in a recent study of a similar size to this [9], which
compared women undergoing IVF with a historical cohort of unexposed
women, the authors did not adjust for parity in their analysis of invasive
epithelial ovarian cancer. With regard to fertility drug exposure and
ovarian cancer risk [15], three studies identified an increased risk in
women who did not give birth [2,16,17], though another large study
did not [18]. Studies of endometriosis and ovarian cancer have sometimes adjusted for parity [19]; to our knowledge none have considered
nulliparous women as a separate group, though Brinton et al. [20] found
a four-fold increase in risk of ovarian cancer in women with primary infertility and a diagnosis of endometriosis, compared with the general
population.
Our data also confirm previous findings of a reduction in risk of
ovarian cancer with tubal sterilization and with hysterectomy, with
or without USO. Tubal ligation is known to protect against ovarian
cancer and in recent meta-analyses, Cibula et al. [21] estimated the
relative risk (RR) of ovarian cancer after tubal ligation to be 0.66
(95% CI 0.60–0.73) while Rice et al. [22] derived a RR of 0.70 (95% CI
0.64–0.75). The estimate from our study was 0.66 (95% CI 0.26–1.68).
Hysterectomy with or without USO has generally been shown to protect
against ovarian cancer [21] and our results were consistent with this.
We found a HR of 0.55 (95%CI 0.22–1.25) for hysterectomy alone, and
0.72 (95% CI 0.10–5.27) for hysterectomy with USO; very similar to
the estimates reported in Rice et al.'s meta-analysis [22] of 0.62 (95%
CI 0.49–0.79) and 0.60 (95% CI 0.47–0.78). We had also expected to
find a protective effect from USO performed without hysterectomy,
but this was not the case, and we were surprised to find the exact opposite. Women who underwent USO without hysterectomy had a
four-fold increase in ovarian cancer risk, with ovarian cancer diagnosed
more than 10 years after USO. This observation derives some support
from a paper published in 1997 by Krieger et al. [23]. These authors
also found an increased risk of ovarian cancer in women who had undergone USO without hysterectomy; however, in their study, most of
the increased risk was concentrated in the first few years after USO.
Nevertheless, we believe this is an important observation that warrants
further investigation. It may be that the reason for performing a USO is
tied with an inherent risk of ovarian cancer in some women, and hence
they are more likely to be subsequently diagnosed with ovarian cancer,
and it may be possible to target these women at increased risk of disease many years before it would normally be diagnosed.
Our study had a number of limitations. Our estimates were imprecise due to the relatively small number of ovarian cancer cases, and
their confidence intervals frequently included one [24,25]. Nevertheless, where possible we verified our estimates by comparison with the
literature and found good agreement between our results and other
published estimates. Even though we recruited women from 1982,
when IVF was in its infancy, and followed them for up to 29 years
through to 2010, the mean age at the end of follow-up was only
48 years, well short of the average age at ovarian cancer diagnosis of
63 years. Future studies, which could take advantage of longer periods
of follow-up, would likely see more ovarian cancer cases diagnosed
and consequently more precise estimates of risk. The fact that
follow-up ended at an average age of 48 years may explain why the average age at ovarian cancer diagnosis was only 46 years. However, it is
also conceivable that this infertile population had a higher incidence of
BRCA1 mutations with possibly an increased risk and earlier development of ovarian cancer [26].
A further limitation of our study was our inability to categorise IVF
exposure according to types and doses of fertility drugs used, as this information was unavailable to us. In addition, we had no information on
the use of fertility drugs, which may influence the risk of ovarian cancer,
other than those used as part of an IVF cycle. We also had no information on the use of oral contraceptive preparations prior to infertility
treatment. Oral contraceptive use is known to lead to a reduction in
the rate of ovarian cancer [27].
Our study had a number of strengths. We were able to examine the
risk of ovarian cancer in a large population-based cohort of women
seeking infertility treatment, with information on exposures and outcomes available over a period of 30 years. We made comparisons within the cohort, rather than comparing infertile women with the general
population, thus reducing the potential for confounding by unknown
variables. We had minimal loss to follow-up as the cohort was restricted
to women known to be resident in WA. Accurate information on exposures and outcomes was obtained from hospital and registry data, thus
eliminating the possibility of non-response. The diagnosis of endometriosis was recorded at the start of follow-up, so there was a clear
264
L.M. Stewart et al. / Gynecologic Oncology 128 (2013) 260–264
temporal relationship between the earlier diagnosis of endometriosis
and later diagnosis of ovarian cancer, with ovarian cancer diagnosed,
on average, 14 years after the diagnosis of endometriosis. A further
strength was that we were able to measure time at risk accurately and
censor follow-up in women who had a BSO who afterwards experience
a dramatic reduction in the risk of ovarian cancer.
We suggest two avenues for future research. Both involve a
narrowing of focus. Firstly, that known and potential risk factors
are investigated separately in parous and nulliparous women, and
secondly that women who require a USO without hysterectomy are
investigated further to determine if the conditions that underlie
the need for this procedure can predict early-stage ovarian cancer.
Conflict of interest statement
LMS, CDJH, PAS, JCF and DBP have nothing to declare. RH is a member of the Medical
Advisory Board of Schering-Plough, Australia and the Medical Advisory Board of
Merck-Serono, Australia and has received travel and accommodation support from
the above to attend conferences. RH is a Medical Director of Fertility Specialists of
Western Australia and holds shares in Western IVF.
Acknowledgments
The authors wish to thank the staff at the Western Australian Data
Linkage Branch and the Hospital Morbidity Data Collection; the Reproductive Technology Register; the Western Australian Cancer Registry;
the WA Deaths Registry; the Midwives Notifications System and the
WA Electoral Roll.
We also thank Dr Allan Jensen of the Danish Cancer Society Research
Centre for generously providing advice on data analysis.
This work was supported in part by a capacity building grant from the
National Health and Medical Research Council, Australia, project no.
573122. Louise Stewart is the recipient of an Australian Postgraduate
Award from the Australian Government Department of Innovation,
Industry, Science and Research.
References
[1] Jensen A, Sharif H, Olsen JH, Kjær SK. Risk of breast cancer and gynecologic cancers
in a large population of nearly 50,000 infertile Danish women. Am J Epidemiol
2008;168:49-57.
[2] Brinton LA, Lamb EJ, Moghissi KS, Scoccia B, Althuis MD, Mabie JE, et al. Ovarian
cancer risk after the use of ovulation-stimulating drugs. Obstet Gynecol 2004;103:
1194-203.
[3] Kallen B, Finnstrom O, Nygren KG, Otterblad Olausson P, Wennerholm UB. In vitro
fertilisation in Sweden: obstetric characteristics, maternal morbidity and mortality.
Br J Obstet Gynecol 2005;112:1529-35.
[4] Venn A, Watson L, Bruinsma F, Giles G, Healy D. Risk of cancer after use of fertility
drugs with in-vitro fertilisation. Lancet 1999;354:1586-90.
[5] Venn A, Watson L, Lumley J, Giles G, King C, Healy D. Breast and ovarian cancer
incidence after infertility and in vitro fertilisation. Lancet 1995;346:995–1000.
[6] Dor J, Lerner-Geva L, Rabinovici J, Chetrit A, Levran D, Lunenfeld B, et al. Cancer
incidence in a cohort of infertile women who underwent in vitro fertilization.
Fertil Steril 2002;77:324-7.
[7] Lerner-Geva L, Geva E, Lessing JB, Chetrit A, Modan B, Amit A. The possible association between in vitro fertilization treatments and cancer development. Int J Gynecol
Cancer 2003;13:23-7.
[8] Yli-Kuha AN, Gissler M, Klemetti R, Luoto R, Hemminki E. Cancer morbidity in a cohort of 9175 Finnish women treated for infertility. Hum Reprod 2012;27:1149-55.
[9] van Leeuwen FE, Klip H, Mooij TM, van de Swaluw AM, Lambalk CB, Kortman M,
et al. Risk of borderline and invasive ovarian tumours after ovarian stimulation for
in vitro fertilization in a large Dutch cohort. Hum Reprod 2011;26:3456-65.
[10] AIHW. Ovarian cancer in Australia: an overview, 2010. Cat. no. CAN 48. Australian
Institute of Health and Welfare; 2010 [Available at http://www.aihw.gov.au/
publication-detail/?id=6442468331 Accessed July 25, 2012].
[11] Stewart LM, Holman CDJ, Hart R, Bulsara M, Preen DB, Finn JC. In vitro fertilization
and breast cancer: is there cause for concern? Fertil Steril 2012;98:334-40.
[12] Holman CD, Bass AJ, Rouse IL, Hobbs MS. Population-based linkage of health records in Western Australia: development of a health services research linked database. Aust N Z J Public Health 1999;23:453-9.
[13] Information Paper: An Introduction to Socio-Economic Indexes for Areas (SEIFA),
2006. ABS catalogue no 2039.0. Australian Bureau of Statistics; 2008 [Available at
http://www.abs.gov.au/ausstats/[email protected]/mf/2039.0/ Accessed July 25, 2012].
[14] Hosmer DW, Lemeshow S, May S. Applied survival analysis: regression modeling
of time-to-event data. 2nd ed.Hoboken, NJ: John Wiley and Sons; 2008.
[15] Brinton LA. Breast cancer risk after use of fertility drugs: stimulating new controversy. J Natl Cancer Inst 2012;104:962-4.
[16] Whittemore AS, Harris R, Itnyre J. Characteristics relating to ovarian cancer risk:
collaborative analysis of 12 US case-control studies. II. Invasive epithelial ovarian
cancers in white women. Collaborative Ovarian Cancer Group. Am J Epidemiol
1992;136:1184-203.
[17] Ness RB, Cramer DW, Goodman MT, Kjaer SK, Mallin K, Mosgaard BJ, et al. Infertility, fertility drugs, and ovarian cancer: a pooled analysis of case-control studies.
Am J Epidemiol 2002;155:217-24.
[18] Jensen A, Sharif H, Frederiksen K, Kjaer SK. Use of fertility drugs and risk of ovarian cancer: Danish Population Based Cohort Study. BMJ 2009;338:b249.
[19] Sayasneh A, Tsivos D, Crawford R. Endometriosis and ovarian cancer: a systematic
review. ISRN Obstet Gynecol 2011;2011:140310.
[20] Brinton LA, Lamb EJ, Moghissi KS, Scoccia B, Althuis MD, Mabie JE, et al. Ovarian
cancer risk associated with varying causes of infertility. Fertil Steril 2004;82:
405-14.
[21] Cibula D, Widschwendter M, Majek O, Dusek L. Tubal ligation and the risk of ovarian cancer: review and meta-analysis. Hum Reprod Update 2010;17:55-67.
[22] Rice MS, Murphy MA, Tworoger SS. Tubal ligation, hysterectomy and ovarian cancer: a meta-analysis. J Ovarian Res 2012;5:13.
[23] Kreiger N, Sloan M, Cotterchio M, Parsons P. Surgical procedures associated with
risk of ovarian cancer. Int J Epidemiol 1997;26:710-5.
[24] Stang A, Poole C, Kuss O. The ongoing tyranny of statistical significance testing in
biomedical research. Eur J Epidemiol 2010;25:225-30.
[25] Stang A, Rothman KJ. That confounded P-value revisited. J Clin Epidemiol
2011;64:1047-8.
[26] Oktay K, Kim JA, Barad D, Babayev SN. Association of BRCA1 mutations with occult
primary ovarian insufficiency: a possible explanation for the link between infertility and breast/ovarian cancer risks. J Clin Oncol 2010;28:240-4.
[27] The Cancer and Steroid Hormone Study of the Centers for Disease Control and the
National Institute of Child Health and Human Development. The reduction in risk
of ovarian cancer associated with oral-contraceptive use. N Engl J Med 1987;316:
650-5.