LAUSR

Female survivors of childhood cancer are at risk of late effects of their cancer treatment including increased risk of premature menopause before age 40 years (1). Premature menopause causes infertility at an early age, and also is linked to other adverse health outcomes such as heart disease and osteoporosis (2). Identifying biomarkers that define risk strata for treatmentassociated premature menopause may allow survivors to partake in targeted medical interventions aimed to reduce or prevent the consequences of premature menopause, such as fertility preservation procedures and prevention programs for cardiovascular diseases and osteoporosis. In this issue of the Journal, Brooke and colleagues aim to identify these potential biomarkers in a report describing the first systematic assessment of the how genetic factors may influence risk of premature menopause among childhood cancer survivors (3). The study uses data from 799 female participants in the St. Jude Lifetime Cohort Study (4) to conduct a genomewide association study of the prevalence of clinically diagnosed premature menopause at study entry. The authors provide evidence that a high-risk haplotype upstream from Neuropeptide Y Receptor Y2 (NPY2R) associates with prevalence of premature menopause, where strongest associations are observed among participants exposed to ovarian radiotherapy. Although attenuated, the association between homozygosity for the haplotype and increased prevalence of premature menopause among survivors exposed to ovarian radiotherapy replicates in analyses using data from 1624 survivors enrolled in the Childhood Cancer Survivor Study. Additional evidence of the genetic association with premature menopause is provided through bioinformatics analyses. The authors examine genotype tissue expression (GTEx) data and find that NPY2R is most highly expressed in hypothalamus tissue, the portion of the brain that controls the release of pituitary hormones. Importantly, NPY2R regulates the gene NPY that other investigators have shown is involved in the luteinizing hormone surge before ovulation in mice (5,6). The results of the study by Brooke et al. (3) support incorporating genetic data into prediction models of treatmentassociated premature menopause. However, the study has several limitations that should be addressed in future validation studies before clinical implementation into prediction models. Specifically, the control participants are on average younger than the case patients, where a large majority of the control participants are younger than age 40 years and are still at risk of developing premature menopause. To address this limitation, the authors conducted an analysis that adjusts for age at clinical assessment for premature menopause and found similar results. Future causal inferences of the association between the high-risk haplotype ’and treatmentassociated premature menopause will be strengthened if the association persists as more participants develop premature menopause. Additionally, the study lacks data on the timing of premature menopause in participants. Information on how genetic variation may influence age at premature menopause would be useful for childhood cancer survivors when making fertility decisions. It is important to mention that the majority of participants are of European ancestry, potentially limiting the generalizability of the study results to other racial groups. Lastly, the study’s small sample size suggests that there may be additional variants to be identified in association with premature menopause induced by childhood cancer treatment. Identifying women at higher risk of premature menopause is valuable for a variety of reasons. Female childhood cancer survivors have more concerns about future fertility than young adults who have not been diagnosed with cancer (7). A biomarker for risk of premature menopause may help to assuage