Fertility treatment and risk of ovarian cancer in a large nationwide cohort of 151,821 infertile Danish women

Abstract Study question Do use of fertility drugs increase the risk of ovarian cancer among infertile women Summary answer Use certain types of fertility drugs may modify the risk of ovarian cancer among infertile women, primarily for serous ovarian tumors and among parous women. What is known already Even though most previous studies on the association between fertility treatment and ovarian cancer have not been able to show a convincing association, some studies have found an increased risk of this malignancy among women undergoing fertility treatment. However, many of the previous studies have had methodological limitations including selection bias and potential confounding due to missing information of important factors, such as parity status and hormonal contraceptive use, small sample size as well as short- and incomplete follow-up time. Furthermore, only few studies have assessed the association between fertility treatment and ovarian cancer according to histological type. Study design, size, duration This retrospective register-based cohort study included virtually all 20-45 year old infertile women in Denmark between 1971 and 2017 as identified in the Danish Infertility Cohort. All women were followed from entry in the cohort (i.e. the date of first infertility diagnosis) to occurrence of ovarian cancer, any other cancer (except non-melanoma skin cancer), death, emigration, bilateral oophorectomy or end of follow-up (December 31, 2017), whichever occurred first. The median follow-up length was 10.3 years Participants/materials, setting, methods In total, 332 women were diagnosed with ovarian cancer during the follow-up period. Information on the use of specific fertility drugs (clomiphene citrate, gonadotropins, hCGs, GnRH receptor modulators and progesterone), ovarian cancer, covariates and vital status was obtained from the Danish Infertility Cohort and various Danish national registers. Cox proportional hazard regression models were used to calculate hazard ratios (HRs) and 95% CIs for ovarian cancer overall and for serous ovarian tumors. Main results and the role of chance After adjustment for attained age (as the underlying time scale), calendar year of study entry, highest obtained level of education, maternal age at first childbirth, number of childbirths, hormonal contraceptive use as well as mutual adjustment for treatment with any other specific fertility drugs, ever use of hCG (HR 0.62, 95% CI 0.44-0.89) and GnRH receptor modulators (HR 0.63, 95% CI 0.40-1.00) were associated with a decreased risk of overall ovarian cancer. In contrast, ever use of gonadotropins (HR 1.43; 95% CI 0.91-2.24) and especially progesterone (HR 1.81; 95% CI 1.18-2.78) increased the risk of overall ovarian cancer. No marked association was observed for clomiphene citrate (HR 1.07, 95% CI 0.80-1.44)). The rates for serous ovarian cancer generally resembled those observed for overall ovarian cancer; however only the association for progesterone reached statistical significance (HR 2.89; 95% CI 1.67-4.99). The observed associations existed mainly among parous women and did not vary with time since first use of the fertility drug in question and no statistically significant associations were observed with cumulative dose of the specific fertility drugs. Limitations, reasons for caution The median age at the end of follow-up was only 42.5 years, which is markedly lower than the usual peak age for ovarian cancer in Denmark (mid 60s). Hence, we were not able to capture the true, potential long-term association between use of fertility drugs and ovarian cancer. Wider implications of the findings Use of certain specific types of fertility drugs in fertility treatment may modify the risk of ovarian cancer among subgroups of women. However, although this study is by far the largest to date, additional large epidemiological studies with longer follow-up time are needed to further clarify the observed associations. Trial registration number Not applicable.