Fertility treatment and risk of melanoma: a certainty‐rated systematic review and meta‐analysis of 3,733,499 women

Background: Cancer safety after fertility treatment is well studied for breast and gynaecological tumours, but melanoma risk remains poorly characterized. We synthesized internally controlled evidence on this association. Methods: PubMed/MEDLINE, Scopus, Cochrane Library, and ClinicalTrials.gov (2000–2025) were searched for comparative observational studies of melanoma incidence. The primary outcome was incident melanoma; treatment-pathway/SIR analyses were secondary. Hazard ratios and case–cohort relative risks (rare-disease approximation) were pooled using random-effects meta-analysis (REML, Hartung–Knapp) in R; SIRs were synthesized separately. Risk of bias was assessed with ROBINS-E and certainty graded using GRADE. Results: Twelve observational studies (3,733,499 women; 219,357 exposed) were included. ROBINS-E rated no study overall low risk of bias (83% some concerns, 17% high risk), reflecting incomplete ultraviolet-exposure adjustment. Melanoma risk was modestly elevated (HR 1.24, 95% CI 1.02–1.50; I2 = 0%; prediction interval 0.97–1.58); drug-only ovulation induction showed a stronger association (HR 1.39, 95% CI 1.01–1.92), while ART/IVF was near-null (HR 1.09, 95% CI 0.81–1.48). Corresponding absolute increases were approximately 3–8, 5–13, and 1–3 per 10,000 women over 10 years (baseline 13.3–33.1/100,000/year). Certainty was very low for any fertility treatment and drug-only treatment, and low for ART/IVF. SIR-based analyses were non-significant and highly heterogeneous (SIR 1.25, 95% CI 0.66–2.38; I2 = 80%). Conclusions: Observed associations are small and of very low to low certainty, reflecting incomplete ultraviolet-exposure and risk-factor adjustment; causality cannot be established. Current evidence does not justify changes in fertility-treatment practice. Future studies require cycle-level exposure data and rigorous ultraviolet-exposure control.