Premature ovarian insufficiency in patients with systemic lupus erythematosus on cyclophosphamide: a systematic review and meta-analysis

Background Premature ovarian insufficiency (POI) is a major long-term adverse effect of cyclophosphamide (CYC) therapy in systemic lupus erythematosus (SLE). Reported prevalence varies due to differences in populations, protocols, and diagnostic criteria. Objective: To estimate the prevalence of POI in women with SLE exposed to CYC and explore sources of heterogeneity. Methods MEDLINE, Embase, Web of Science, Scopus, and LILACS were searched from inception through October 2024. Twenty-one studies were included in the systematic review; 12 CYC-exposed cohorts contributed to the primary meta-analysis. Two reviewers independently extracted data and assessed study quality using the Joanna Briggs Institute checklist. Results The pooled prevalence of POI in SLE treated with CYC was 15% (95% CI 7-26%), with substantial heterogeneity (I 2 = 85%; Cochran’s Q = 73.98 p 0.0001; tau 2 = 0.0353). In contrast, prevalence among unselected SLE populations was markedly lower, confirming the potential gonadotoxic effect of CYC. Temporal analysis showed reduced prevalence in studies published after 2005 (9%) compared with pre-2005 (29%), coinciding with adoption of Euro-Lupus low-dose protocols. Meta-regression identified age at exposure as a significant predictor (β = 0.21, 95% CI: 0.04-0.38, p = 0.017). Cumulative CYC dose was not independently associated with prevalence, likely reflecting the widespread adoption of lower-dose regimens after 2005. Methodological sensitivity analyses using logit transformation (20.0%, 95% CI 12.7-30.1) and GLMM (14.5%, 95% CI 7.1-27.2) were added to assess model robustness. Funnel-plot and the Egger’s test indicated no evidence of publication bias (p = 0.91). Conclusions Women with SLE treated with CYC show a substantial burden of POI, with an estimated prevalence of approximately 15%. These results underscore the need for standardized diagnostic criteria, fertility counseling, and individualized treatment decisions. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/ , identifier CRD42025641213.