# Limitations of hormonal and clinical markers in predicting GnRH agonist trigger success in polycystic ovary syndrome: A critical reappraisal

## TL;DR

This study finds that current hormonal and clinical markers are unreliable for predicting oocyte yield in PCOS patients using GnRH agonist triggers during IVF.

## Contribution

The study identifies that antral follicle count (AFC) is the only potential predictor, but its predictive power is weak for PCOS patients.

## Key findings

- A high AFC was the only independent predictor of oocyte yield per follicle in PCOS patients.
- AFC's predictive validity was poor with an AUC of 0.561, indicating unreliable prediction of suboptimal oocyte yield.
- BMI and serum LH levels were not significant predictors of oocyte yield in this patient group.

## Abstract

Polycystic ovary syndrome (PCOS) is associated with altered hypothalamic-pituitary-ovarian function, which may affect the success of GnRH agonist triggers used during IVF. Identifying reliable predictors of oocyte yield in these patients remains a clinical challenge.

To evaluate possible predictors of suboptimal oocyte retrieval per aspirated follicle when ovulation is triggered with GnRH agonist in PCOS patients.

Between 30/04/2021 and 30/12/2022, a prospective cohort study was conducted among women with PCOS (n = 104) in which a GnRH agonist trigger was employed during a GnRH antagonist protocol. Hormonal and clinical parameters were tested for their ability to predict the oocyte per aspirated follicle (OPF) rate in in-vitro fertilization (IVF) treatment.

The mean age of the patients, mean number of aspirated follicles, number of collected oocyte cumulus complexes and OPF were 28.6 ± 3.9 years, 31.4 ± 10.2, 21.7 ± 8.9, and 70.2 ± 19.1%, respectively. When patients were stratified into three groups according to their OPF percentiles (Q1:0–25th percentile, Q2:26–75th percentile, Q3:76–100th percentile), body mass index (BMI) and antral follicle count (AFC) were significantly higher in the Q1 group compared to the Q2 and Q3 groups. However, regression analysis revealed that only AFC was an independent predictor of the OPF rate (RR: −0.005, 95% CI: −0.008 to −0.002, p = 0.001), but not BMI or serum LH levels on the day of triggering. Notably, the predictive validity of AFC to recognize a low OPF rate was poor (AUC = 0.561).

A high AFC was the only identifiable predictor of the OPF rate and a suboptimal response when a GnRH agonist trigger was used for final follicular maturation. However, since a low AUC for AFC suggests a poor performance, we conclude that this study was not able to find any reliable prediction markers for the OPF rate in PCOS patients triggered with GnRHa.