MON-215 Comparison of Ultrasonographic and Biochemical Markers of Polycystic Ovarian Morphology Across the Phenotypic Spectrum of Polycystic Ovary Syndrome

Introduction: Polycystic ovarian morphology (PCOM) manifests from disordered folliculogenesis and can be reliably identified on ultrasonography by estimating mean antral follicle number per ovary (FNPO). However, PCOM can be heterogeneous among women with polycystic ovary syndrome (PCOS) – leading to the hypothesis that the degree of disordered folliculogenesis may differ between clinical phenotypes. To begin to shed light on such differences, we aimed to evaluate the relationships between ultrasonographic and biochemical markers of PCOM across phenotypes of PCOS. Methods: Women with PCOS (n = 119) and Controls (n = 43) were recruited from the general population. PCOS was defined according to the Rotterdam Criteria. Thresholds for menstrual irregularity, androgen excess, and PCOM were obtained from the 2018 International Guideline for the Assessment and Management of PCOS. Phenotypes included: Classic (menstrual irregularity + androgen excess; n = 73), Ovulatory (androgen excess + PCOM; n = 19), and Mild (menstrual regularity + PCOM; n = 27). Women provided their reproductive history and underwent a hirsutism assessment, fasting blood draw, and transvaginal ultrasound scan. Sera were assayed for total testosterone and anti-müllerian hormone (AMH). Scans were evaluated for FNPO and follicle size populations (i.e. 2–9 mm, 2–5 mm, 6–9 mm). Per antral follicle production of AMH was estimated as the ratio of AMH : FNPO. Differences in ultrasonographic and biochemical markers between groups were assessed with the Steel-Dwass Test. Associations between ultrasonographic and biochemical markers within groups were judged with Spearman’s rank correlation. Differences in the utility of each marker to distinguish a given phenotype of PCOS from Controls were determined with logistic regression. Results: Classic, Ovulatory, and Mild PCOS showed similar elevations in the number of total (2–9 mm) and small follicles (2–5 mm) and AMH compared to Controls. Classic PCOS also had a higher number of medium follicles (6–9 mm) than Controls. AMH was moderately correlated with the number of total and small follicles across groups but related to the number of medium follicles only in the non-androgenic women (Mild PCOS and Controls). FNPO offered the best utility to distinguish PCOS from Controls, with clinically relevant gains over AMH, particularly in Mild PCOS. Conclusions: These findings suggest that AMH plays a different role in the degree of disordered folliculogenesis in Classic versus Mild PCOS – one that is not readily explained by differences in its estimated follicular production between phenotypes. Further studies are needed to understand the implications of these findings for the diagnosis and evaluation of the condition. Acknowledgements: This study was supported in part by Ansh Labs. Assays for AMH were performed at its facilities in Webster, TX.