An important late effect among female survivors of childhood cancer is treatment-related ovarian damage and impaired fertility. While chemotherapy and radiation therapy are associated with impaired fertility, few other risk factors have been identified. Furthermore, little is known about the role of genetic susceptibility to these late effects. Objective: As Anti-Müllerian Hormone (AMH) is a demonstrated marker of ovarian reserve; our objective was to identify clinical and genetic predictors of AMH levels in female survivors of childhood cancer. Methods: Female childhood cancer survivors (n = 181) were recruited from the Texas Children's Cancer Center Long Term Survivor Program (LTSP). AMH (ng/mL) was measured using an enzyme-linked immunosorbent assay. Information on age at diagnosis and enrollment, race, ethnicity, cancer diagnosis, pelvic radiation, and use of alkylating agents was abstracted from medical records. The following single nucleotide polymorphisms (SNPs) were selected based on known or suspected function: AMHR2 rs2002555; CYP2C9*4 rs56165452; CYP2C19*2 rs4244285; and CYP2C19*3 rs4986893. SNPs were genotyped using TaqMan assays. Linear regression was used to determine the association between selected factors and AMH levels. As AMH levels were not normally distributed, the dependent variable was expressed as log10(AMH+1). Results: The mean age at enrollment in the LTSP was 12.4 years. The most common cancer diagnosis was acute lymphoblastic leukemia (47.2%), and a substantial proportion of the population was Hispanic (41.0%). The following variables were significantly associated with lower AMH levels: pelvic radiation (beta = −0.61, P < 0.001) and treatment with alkylating agents (beta = −0.09, P = 0.04). Age at enrollment was associated with higher AMH levels (beta = 0.01, P = 0.04). Hispanic ethnicity was marginally associated with lower AMH levels (beta = −0.08, P = 0.08). There were no significant genetic associations. Conclusions: In one of the largest studies of its kind, our results confirm previous associations between treatment-related factors and ovarian damage. While the SNPs evaluated were not predictive of AMH levels, more work is needed to explain why some survivors experience impaired fertility, while others do not, despite similar therapy.
