Background: The effects of controlled ovarian hyperstimulation (COH) for IVF in terms of breast cancer risk remain controversial, despite the hormone-dependent nature of the latter.
Methods: Eligible studies up to 15 February 2013 were identified and pooled effect estimates for relative risk (RR) were calculated separately for the investigations using the general population and those using infertile women, as a reference group. Fixed- or random-effects models were implemented and subgroup analyses were performed, as appropriate.
Results: Eight cohort studies were synthesized, yielding a total cohort size of 1,554,332 women among whom 14,961 incident breast cancer cases occurred, encompassing 576 incident breast cancer cases among women exposed to IVF. No significant association between IVF and breast cancer was observed either in the group of studies treating the general population (RR = 0.91, 95% confidence interval (CI): 0.74-1.11) or infertile women (RR = 1.02, 95% CI: 0.88-1.18), as a reference group. Of note were the marginal associations, protective for pregnant and/or parous women after IVF (pooled effect estimate = 0.86, 95% CI: 0.73-1.01) and adverse for women <30 years at first IVF treatment (pooled effect estimate = 1.64, 95% CI: 0.96-2.80).
Conclusions: At present, COH for IVF does not seem to impart increased breast cancer risk. Longer follow-up periods, comparisons versus infertile women, subgroup analyses aiming to trace vulnerable subgroups, adjustment for various confounders and larger informative data sets are needed before conclusive statements for the safety of the procedure are reached.
Keywords: IVF; breast cancer; hormones; infertility; meta-analysis.