Fragility Index Analysis for Robustness of Evidence in Randomized Controlled Trials Leading to FDA Approval of Pharmacological Therapies for Endometriosis

INTRODUCTION: The U.S. Food and Drug Administration (FDA) has approved several pharmacological therapies for endometriosis, based on randomized controlled trial (RCT) data. Although these results have shaped current practice, reliance on this threshold alone may obscure the robustness of RCT evidence. The Fragility Index (FI) is a statistical tool that quantifies the stability of trial outcomes. OBJECTIVE: This study aims to evaluate the FI for RCTs of FDA-approved endometriosis therapies. METHODS: A systematic review aimed at analyzing the FI of all RCTs supporting FDA-approved endometriosis treatments up to August 2025 was performed. Eligible trials included patients with endometriosis, tested an FDA-approved drug, and reported dichotomous primary endpoints with sufficient data for FI calculation. Two reviewers independently extracted trial data, including sample size, intervention, comparator, outcome events, and patients lost to follow-up (LFO). The primary outcome was the FI of the primary efficacy endpoint. Secondary outcomes included FI distribution across drug classes and comparator type, as well as FI compared with the number of patients LFO. RESULTS: Seven RCTs enrolling 4,082 patients were included in the analysis. Trials were published between 1990 and 2022. The median sample size per trial was 299. The most studied drug classes were GnRH agonists (four trials, 57.1%), GnRH antagonists (two trials, 28.6%), and progestins (one trial, 14.3%). Trials comparing active therapy to placebo (n=3) had a median FI of 50 with a median of 23 patients lost to follow-up, whereas active-comparator trials (n=4) had a median FI of 0 with 75 patients LFO. Elagolix vs placebo demonstrated the highest robustness (FI=72, LFO=31), followed by relugolix vs placebo (FI=50.5, LFO=104), and Leuprolide vs placebo (FI=14, LFO=30). In contrast, studies comparing goserelin, depot medroxyprogesterone, and nafarelin to danazol all yielded an FI of 0 (Figures 1A, B). CONCLUSIONS: RCTs supporting FDA-approved medical therapies for endometriosis vary widely in statistical robustness. Placebo-controlled trials demonstrated high FI values, suggesting stability, while active-comparator designs were consistently fragile. These findings offer the value of integrating FI into trial reporting and interpretation to contextualize statistical significance and guide evidence-based treatment decisions in endometriosis.Figure 1