Endometriosis and ART: A prior history of surgery for OMA is associated with a poor ovarian response to hyperstimulation

Abstract

Background: Many women whose fertility may have been impaired by endometriosis require assisted reproductive technology (ART) in order to become pregnant. However, the influence of ovarian endometriosis (OMA) on ovarian responsiveness to hyperstimulation has not been clearly established.

Objective: To evaluate the risk of a poor ovarian response (POR) to stimulation and ART outcomes in women with OMA.

Materials and methods: We conducted a large observational controlled matched cohort study in a tertiary care university hospital between 01/10/2012 and 31/12/2015. After matching by age and anti-Müllerian hormone (AMH) levels, 201 infertile women afflicted with OMA (the OMA group) and 402 disease-free women (the control group) undergoing an ART procedure were included in the study. The main outcomes that we measured were a POR to hyperstimulation (i.e., ≤ 3 oocytes retrieved, or cancelled cycles), the clinical pregnancy rate, and the live birth rate. All of the women with endometriosis underwent a pre-ART work-up, in order to obtain an accurate diagnosis and staging of their disease. An OMA diagnosis was based on published imaging criteria (obtained by transvaginal sonography or magnetic resonance imaging) or on histological analysis for patients with a prior history of endometriosis surgery. The statistical analyses were conducted using univariate and multivariate logistic regression models.

Results: The incidence of a POR to hyperstimulation was significantly higher for the OMA group than for the control group [62/201 (30.8%) versus 90/402 (22.3%), respectively; p = 0.02]. However, no significant differences were found between the OMA and the control group in terms of the clinical pregnancy rate [53/151 (35%) versus 134/324 (41.3%), respectively; p = 0.23] and the live birth rate [39/151 (25.8%) versus 99/324 (30.5%), respectively; p = 0.33]. By multivariate analysis, a prior history of surgery for OMA was found to be an independent factor associated with a POR to stimulation [OR = 2.1; 95% CI: 1.1-4.0], unlike OMA without a prior history of surgery [OR: 1.5; 95% CI: 0.9-2.2].

Conclusion: The presence of OMA during ART treatment increased the risk of a POR to hyperstimulation, although the live birth rate was not affected. Furthermore, having OMA and having previously undergone surgery for OMA was identified as an independent risk factor for a POR.

Fig 1. Patient inclusion flowchart.

OMA: Ovarian endometrioma; ART: Assisted reproductive technology; AMH: anti-Müllerian hormone.