Low Endogenous LH on the COS Initiation Day of a GnRH-Agonist Regimen Increases the Risk of Early Pregnancy Loss and Adverse ART Outcomes

Abstract

Objective: To assess the impact of serum luteinizing hormone (LH) levels on the day of initiation of controlled ovarian stimulation (COS) after pituitary suppression on early pregnancy loss and assisted reproductive technology (ART) outcomes. Design: Retrospective cohort study. Setting: University-affiliated hospital. Patients: A total of 9540 normogonadotrophic patients were treated with a GnRH agonist for in vitro fertilization (IVF). Based on the serum concentration of LH on the COS initiation day, patients were divided into low (<1 mIU/mL, n=2838), medium (1–1.49 mIU/mL, n=3357), or high (≥1.5 mIU/mL, n=3345) LH groups and received either fresh embryo transfer (ET) or frozen ET (women with high ovarian response, insufficient endometrial thickness, or requesting frozen ET). A total of 6279 cycles were fresh ET (1960, 2222, and 2097 in the low, medium, and high LH groups, respectively). Intervention(s): During IVF/ICSI, a GnRH agonist was used to suppress pituitary function in the midluteal phase or follicular phase, and then gonadotropin was used to induce COS. Main Outcome Measure(s): The early pregnancy loss rate (ePLR) and live-birth rate (LBR) for fresh ET, as well as the cumulative ePLR and LBR for the entire ovarian stimulation cycle, were compared. Results: In the fresh ET cycles, the high, medium and low LH groups had an ePLR of 8.6%, 11.9% and 12.5%, respectively, and LBR of 42.1%, 37.9% and 37.5%, respectively. There were no significant differences in terms of clinical pregnancy rate (CPR), late pregnancy loss rate (lPLR), and ectopic pregnancy rate (EPR) among the three LH groups. For the entire ovarian stimulation cycle, the high LH group had a greater number of retrieved oocytes compared with the low and medium LH groups. Among the groups of high, medium and low LH, the cumulative CPR were 72.8%, 69.8% and 68.8%, respectively, and the cumulative LBR were 63.4%, 60.4% and 58.5%, respectively. There were no significant differences in the cumulative ePLR, lPLR, or EPR. After multivariable logistic regression, compared with the high LH group, the adjusted odds ratio of early pregnancy loss in the low and medium LH group were 1.429 (1.065-1.919, P = 0.018) and 1.389 (1.041-1.853, P = 0.026). Conclusions: After pituitary suppression by a GnRH-agonist during IVF, a low LH level (<1.5 mIU/mL) on the COS initiation day was associated with adverse ART outcomes—including fewer oocytes, higher ePLR and lower LBR in fresh ET—and lower cumulative CPR and LBR in the entire ovarian-stimulation cycle. And LH on the COS initiation day was an independent factor affecting ePLR after multivariate regression.