Endometriosis and IVF: are agonists really better? Analysis of 1180 cycles with the propensity score matching

Objective: To compare the outcomes of patients with confirmed endometriosis undergoing in vitro fertilization (IVF)-embryo transfer (ET) treated with either gonadotropin-releasing hormone agonist (GnRHa) or gonadotropin-releasing hormone antagonist (GnRHant) using the propensity score (PS) matching. Design: Observational, retrospective analysis from January 2000 to December 2010. Setting: Private tertiary fertility clinic. Patient(s): Patients with endometriosis confirmed by ultrasound or surgery (American Fertility Society; AFS grades I IV) that underwent an IVF-ET, stimulated with standard controlled ovarian hyperstimulation (COH) and GnRHa or GnRHant. Intervention(s): A PS was assigned to all patients, which calculates the conditional probability of receiving a certain treatment; a higher PS (1) meant a higher probability of receiving treatment with GnRHa, and a lower PS (0) meant a higher probability of receiving GnRHant. The PS was calculated with a logistic regression model adjusted specifically for age, follicle stimulating hormone, antral follicle count and previous IVF cycles. All patients were divided into three groups according to their PS. Main outcome measure(s): pregnancy rate (PR) per cycle. Results: 1180 patients were analyzed. Raw PR per cycle was 41.8% and 23.4%, and PR per ET was 44.3 and 27%, respectively. PR per cycle: 41.9 versus 30% in group A; in group B, 39.7% versus 36.4% and in group C, 15.4% versus 18.9%. The overall odds ratio for PR adjusted by PS was 1.10 [0.58-2.19]. Conclusions: After matching patients by PS, PR after COH with either GnRHa or GnRHant may be equally effective.