Clinical Features and Management of Suboptimal Ovarian Response During in vitro Fertilization and Embryo Transfer: Analysis Based on a Retrospective Cohort Study.

Abstract

Based on dynamic changes of indicators during controlled ovarian hyperstimulation and of clinical outcomes of suboptimal ovarian response with different protocols, this study aimed to summarize the clinical characteristics of SOR and provide clinical recommendations. Data of 125 patients with SOR and 125 controls who had undergone appropriate protocols for in vitro fertilization-embryo transfer were collected from a single medical center from January 2017 to January 2019. Basic clinical indexes, including age, BMI, antral-follicle count, infertility time, basic follicle-stimulating hormone, luteinizing hormone, LH/FSH ratio, estradiol, progesterone, testosterone, androstenedione, prolactin, anti-Mullerian hormone, and thyroid stimulating hormone levels, were analyzed using T-test. Dynamic indexes during COH, including amount and days of gonadotropin, sex hormone levels, and number of large/medium/small follicles at specified time periods, were analyzed using T-test and joint diagnosis analysis with ROC curves. Indexes of laboratory and clinical indicators were analyzed using the chi-square test. For the SOR group, BMI, duration time, and dosage of gonadotropin used for SOR were significantly higher. In the ultra-long/long group, ROC curve analysis showed that the LH/FSH ratio and BMI yielded cutoff values of 0.61 and 21.35 kg/m2, respectively. A combined diagnosis of the two indexes showed higher sensitivity (90%) and specificity (59%). In the GnRH-ant group, ROC curve analysis showed an LH level, an LH/FSH ratio on COH day 2, and BMI yielded cutoff values of 2.47 IU/L, 0.57, and 23.95 kg/m2, respectively. Combining the two indexes with BMI, both showed increased sensitivity (77%) and specificity (72% and 74%). The estradiol level and progesterone level during the late follicular stage in SOR patients were significantly lower than those in control patients for both protocol groups. At each monitoring time, delayed follicular development was observed. The live-birth rate in fresh cycles of the ultra-long/long group and the live-birth rate in cumulative cycles of the antagonist group in the SOR group were lower than those in the control group. SOR had adverse effects on clinical outcome. We provide some threshold values of basic LH/FSH ratio, BMI, COH day 2 LH, counts of follicles, and levels of estradiol/progesterone to be taken as reference to assist the early recognition of SOR.