Assessing the reproductive lifespan of female cancer survivors using AMH: a meta-synthesis of longitudinal studies

Abstract BACKGROUND Anti-Müllerian hormone (AMH) is widely used to assess gonadotoxicity related to anti-cancer treatments and to measure ovarian function in cancer survivors. The extent of AMH decline across malignancies and treatment regimens and the degree of subsequent recovery have not been systematically quantified. OBJECTIVE AND RATIONALE The primary objective of this review was to quantify the change in AMH following cancer treatment and to model longitudinal AMH trajectories in survivors. The secondary objective was to compare AMH trajectories in cancer survivors with those of age-matched non-cancer controls. Data on long-term ovarian reserve after cancer treatment remain limited. This study integrates recent evidence and age-matched control data to better distinguish treatment-related AMH decline from normal reproductive ageing. SEARCH METHODS A systematic search was performed across CENTRAL, MEDLINE, Google Scholar, Embase, and MedRxiv from inception until 24 January 2024, with an updated search on 19 March 2026. Both MEDLINE and Embase were searched via the Ovid platform, whilst Emtree index terms were used for EMBASE. The search strategy combined both index terms and free-text terms: (gonadotoxicity OR premature menopause OR premature ovarian insufficiency (POI) OR premature ovarian failure (POF)) AND (AMH OR Antimullerian* OR Anti-Mullerian) AND cancer. Studies were included if they involved female cancer patients of any age, including paediatric, adolescent and young adult (AYA), and adult populations, who received chemotherapy and who reported either pre- and post-treatment AMH levels or AMH measurements at defined time points after treatment. Studies were excluded if they did not report quantitative AMH data, involved male patients or non-human subjects, or included studies unrelated to cancer treatment-induced gonadotoxicity. The primary outcome was percentage change in AMH from baseline to 12 and 24 months from the end of treatment. Secondary outcomes included long-term AMH trajectories in adult and paediatric cancer survivors compared to the age-related decline in healthy controls. Risk of bias was assessed using the Joanna Briggs Institute (JBI) study-design-specific critical appraisal tools, and certainty of evidence was evaluated using GRADE. Analyses were performed in GraphPad Prism v10.6. OUTCOMES Across all four cohorts (breast, lymphoma, thyroid, and paediatric/AYA mixed cancers), AMH levels were higher at 12 and 24 months compared with 6 months from the end of treatment. With breast cancer (median age 37.8 years, n = 2259), AMH was reduced by 76.4% at 12 months (95% CI: −93.5 to −55.7; 24 studies; GRADE: moderate-certainty evidence). With lymphoma, ABVD treatment regimens (median age 24 years, n = 146) were associated with a smaller reduction in AMH at 12 months (−21.1%; 95% CI: −81.6 to +14.8; four studies; GRADE: low-certainty evidence), whereas non-ABVD regimens (median age 26.3 years, n = 188) were associated with a larger reduction (−90.1%; 95% CI: −96.1 to −68.3; six studies; GRADE: moderate-certainty evidence). With thyroid cancer (median age 32.5 years, n = 181), radioactive iodine treatment was associated with a 57.9% reduction in AMH at 12 months (95% CI: −65.2 to −36.4; 4 studies; GRADE: high-certainty evidence). In the paediatric and AYA mixed cancers cohort (median age 9, n = 176), AMH was reduced by 63.1% at 12 months (95% CI: 64.3–61.9; four studies; GRADE: very low-certainty evidence). Among the adult and paediatric survivors who retained ovarian function, AMH trajectories remained relatively stable and closely aligned with age-matched healthy controls. WIDER IMPLICATIONS AMH levels consistently improved between 6 and 12–24 months from the end of cancer treatment. We provide generalizable estimates of AMH trajectories in female survivors of both childhood and young adult cancer, indicating that their long-term AMH declines are similar to those of age-matched controls. REGISTRATION NUMBER PROSPERO registration number: CRD420251000939.