Abstract
OBJECTIVE: Among women with uterine fibroids (UFs), we assess the extent to which the linzagolix effect on pain alleviation is explained by its effect on reducing heavy menstrual bleeding (HMB) and fibroid volume (FV). DESIGN: Post hoc analysis on the pooled data from two randomised double-blind placebo-controlled phase 3 trials. SETTING: 94 sites in the US (PRIMROSE 1 trial) and 95 sites in Europe/US (PRIMROSE 2 trial). POPULATION: Women aged ≥ 18 years with ultrasound-confirmed UFs and HMB (n = 1012). METHODS: Participants were randomised to linzagolix (100 mg and 200 mg, with and without hormonal add-back therapy) versus placebo. A post hoc mediation analysis was conducted on the pooled PRIMROSE 1 and PRIMROSE 2 data. The effect of linzagolix versus placebo on pain reduction was divided into three components (effect explained by HMB reduction associated with linzagolix, FV reduction associated with linzagolix, and remaining [not yet explained] treatment effect), with proportions of each component reported. MAIN OUTCOME MEASURES: The mediation analysis outcome was clinically significant pain reduction, defined as a change of ≥ 2 pain categories from baseline to Week 24 using the Numeric Rating Scale (pain categories: no pain (0), and mild (1-3), moderate (4-6), severe pain (7-10)). RESULTS: In the mediation analysis, 28%-51% (depending on treatment arm) of linzagolix effect on pain reduction was explained by its effect on HMB reduction, while 2%-8% was explained by its effect on FV reduction. The residual proportion ranged between 44% and 67%, depending on treatment arm, and was statistically significant only in the linzagolix 200 mg without add-back therapy arm (p = 0.002). CONCLUSIONS: This analysis showed that reductions in pain were significantly mediated by reductions in HMB (all doses) and FV (200 mg alone) in linzagolix-treated women with UFs. Further research is needed to identify additional mediating factors. TRIAL REGISTRATION: ClinicalTrials.gov: NCT03070899 and NCT03070951.