Phenotypic heterogeneity and adverse pregnancy outcomes in polycystic ovary syndrome patients undergoing ART: exploring potential links for prenatal surveillance

The relationship between phenotypic heterogeneity in PCOS and pregnancy/neonatal outcomes in women undergoing ART remains insufficiently explored across phenotypes, and further research is needed to understand its relevance for antenatal monitoring. This retrospective analysis included 1,377 PCOS patients treated between January 2018 and July 2023. According to the Rotterdam criteria, patients were stratified into four phenotypic groups: A (oligo-anovulation [OA] + hyperandrogenism [HA] + polycystic ovarian morphology [PCOM]), B (OA + HA), C (HA + PCOM, ovulatory phenotype), and D (OA + PCOM, non-hyperandrogenic phenotype). We analyzed demographic characteristics, embryological parameters, pregnancy and neonatal outcomes. Four phenotypes (A: n =473, B: n =72, C: n =168, D: n =664) were comparable in maternal age, baseline FSH, and oocyte yield, but differed significantly in duration of infertility, BMI, baseline LH, testosterone, and AMH levels. No significant interphenotype differences were observed in clinical pregnancy outcomes (implantation rate, clinical pregnancy rate, miscarriage rate, live birth rate, or singleton birth weight). Significant interphenotype differences were observed in the incidences of gestational hypertension, placenta-related complications, and other complications. After multivariable adjustment, PCOS phenotype was not independently associated with gestational hypertension or placenta-related complications. Notably, PCOS phenotypes were identified as significant contributors to other complications, although the limited number of events warrants cautious interpretation. In twin pregnancies, fetal weight and length also differed significantly across phenotypes, however, after multivariable adjustment, PCOS phenotype was no longer significantly associated with these outcomes. In summary, among PCOS patients undergoing ART, clinical pregnancy rates, live birth rates, and neonatal outcomes appear broadly comparable across phenotypes, however, the risk profiles for obstetric complications suggest a possible phenotype-specific pattern, but this should be interpreted with caution given the limited sample size. Accordingly, further prospective studies are warranted to clarify these phenotype-specific risk profiles and to explore their implications for antenatal management.