Medically Assisted Reproduction and Hormone-Related Cancers

It is critical that women undertaking medically assisted reproductive (MAR) treatment and their clinicians know whether the treatments are associated with an increased risk of hormone-related cancers.

To determine the risk of hormone-related cancers following MAR treatment.

This cohort study used an emulated target trial design including Australian health registries and administrative datasets. Participants included women enrolled in Medicare, Australia's universal health insurance scheme, aged 18 to 55 years between January 1, 1991, and December 31, 2018. Data were analyzed from April 2024 to July 2025.

Exposures were defined from Medicare records: assisted reproduction therapy, intrauterine insemination or ovarian stimulation, and ovulation induction with clomiphene citrate.

Hormone-related invasive cancers (identified in the Australian Cancer Database) included breast, ovarian, uterine, thyroid, colorectal cancers and melanoma; in situ cancers included breast cancer and melanoma. Three cancers with no hormonal links-pancreatic, lung, and hematological-were included as negative controls. Flexible parametric survival models ascertained hazard ratios (HRs) and cumulative marginal survival differences in incident cancers per 100 000 population. E-values assessed the risk of bias due to unmeasured confounding.

A total of 1 748 927 women were identified, including 396 661 with history of MAR exposure. Although elevated risk of most hormone-related cancers was observed after MAR treatment (HRs, 1.09-1.64), E-value analysis suggested confounding due to underlying infertility conditions (ie, endometriosis, polycystic ovarian syndrome) could account for this observed elevation for uterine, ovarian, and thyroid cancers. For any specific invasive cancer, fewer than 20 extra cancers per 100 000 women each year were estimated for treated vs comparator groups. Emulated trials on the 6 hormone-related cancers and pancreatic and hematological cancers showed increased cancer risk in the first years after treatment, suggesting detection bias. Increased risk of hematological cancers was observed after MAR treatment (HRs, 1.18-1.27), indicating uncontrolled confounding by race and ethnicity may account for observed excess risk seen for several cancers. Some treatments were associated with decreased lung cancer risk (HRs, 0.72-0.82).

In this cohort study of MAR and cancer using a target trial emulation design, although associations between MAR and some hormone-related cancers were observed, the estimated difference in the number of expected cancers was small and may be explained by unmeasured confounding and detection bias.