Effect of recombinant LH supplementation on cumulative live birth rate compared with FSH alone in poor ovarian responders: a large,real-world study

Research questionThe benefit of LH supplementation (LHS) over sole use of FSH during controlled ovarian stimulation (COS) remains controversial. Meta-analyses have provided some evidence that the benefit of LHS is limited to women with poor ovarian response (POR). This study aimed to assess the effectiveness of LHS on cumulative live birth rate (CLBR) in POR using a large controlled study in a real-world context.DesignThis retrospective multicentre controlled study used data from registries at 12 French ART centres. All instances of POR undergoing ovarian stimulation and treated with follitrophin-alfa (FSH-α) with or without lutrophin-α were selected following an intention-to-treat principle. POR was defined according to the ESHRE Bologna criteria, and classified into three categories (Mild, Moderate and Severe) according to the Poor Responder Outcome Prediction (PROsPeR) score. The primary end-point was the CLBR associated with fresh and frozen embryos originating from the same ovarian stimulation.ResultsA total of 9787 instances of ovarian stimulation (5218 LHS, 4569 FSH-α only) were analysed, 33.0%, 52.4% and 14.6% being allocated to the Mild, Moderate and Severe PROsPeR categories, respectively. Using a mixed logistic model and adjusting for matched subclasses and baseline POR severity, it was found that the benefit of LHS compared with use of FSH alone differed between baseline severity categories (interaction test, P = 0.007): a significant benefit of LHS for CLBR was found for patients in the Moderate (14.3% versus 11.3%, odds ratio [OR] = 1.37, 95% confidence interval [CI] 1.07–1.75, risk ratio [RR] = 1.29, P = 0.013) and Severe (9.8% versus 4.4%, OR = 2.40, 95% CI– 1.48–3.89, RR = 1.89, P < 0.001) categories, but not for the Mild category (18.8% versus 19.6%, OR = 0.95, 95% CI 0.78–1.15, RR = 0.95, P = 0.60).ConclusionLHS has a significant effect on increasing CLBR in moderately and severely poor ovarian responders.