A propensity score matching comparison between GnRH agonist and GnRH antagonist protocols in patients in POSEIDON groups 3 and 4.

Abstract

To the Editor: The conflicting definition of poor ovarian response has complicated the comparison of proposed treatments. The Patient Oriented Strategies Encompassing Individualized Oocyte Number (POSEIDON) classification[1] was formulated to guide individual ovarian response management through detailed stratification. POSEIDON groups 3 and 4 were defined as patients with low anti-Müllerian hormone (AMH; <1.2 ng/mL) and/or low antral follicle count (AFC; <5). The key to successful assisted reproductive technology (ART) is choosing a suitable controlled ovarian stimulation (COS) protocol to retrieve more oocytes, which is related to the cumulative live birth rate (CLBR).[1,2] Findings on the merits of gonadotropin releasing hormone (GnRH) agonist (GnRH-a) and GnRH antagonist (GnRH-ant) regimens in patients from POSEIDON groups 3 and 4 remain controversial.[3-5] The propensity score matching (PSM) method was used to balance confounding factors in the study to achieve efficacies comparable to those of non-randomized control trials (RCT). This study aimed to explore differences in CLBR, number of retrieved oocytes, and clinical outcomes between GnRH-a and GnRH-ant protocols using PSM in the first COS cycle of patients undergoing ART procedures. This retrospective cohort study evaluated 494 women representing POSEIDON groups 3 and 4, treated with ART at the First Affiliated Hospital of Sun Yat-sen University between January 2015 and December 2019. Patients with endocrine abnormalities, structural uterine abnormalities, preimplantation genetic testing cycles, or those lost to follow-up were excluded. The study included women representing POSEIDON groups 3 and 4 undergoing ART and in their first COS cycle (n = 397). These included 181 and 216 patients using the GnRH-a and GnRH-ant regimen, respectively. This study was approved by the Medical Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University (No. 2019[379]). The initial dose of recombinant follicle-stimulating hormone (FSH) or human menopausal gonadotropin was dependent on the woman's age, body mass index (BMI), FSH, and AMH levels, and AFC. About 250 μg of human choriogonadotropin (hCG) was administered when ≥2 follicles reached a diameter of ≥18 mm. Transvaginal oocyte retrieval was performed 36 h after hCG injection. Fresh embryos were transferred 2 to 5 days after oocyte retrieval. Frozen-thawed embryos were transferred to a natural cycle with luteal support or a hormonal substitution cycle. The primary outcome was the CLBR, defined as the number of deliveries with at least one live birth following the first COS cycle, including all fresh and/or frozen embryo transfer cycles during the 2-year follow-up. The secondary outcomes were laboratory indicators (such as retrieved oocytes) and clinical outcomes. Clinical pregnancy was defined as viable intrauterine pregnancy confirmed by transvaginal ultrasonography at 7 to 8 gestational weeks. Age is an important factor in live births, and AMH levels are predictors of ovarian response. Thus, subgroups were stratified by age (<35 years or ≥35 years) and AMH levels (≥0.88 pg/mL or <0.88 pg/mL), respectively. The optimal cut-off value of AMH was set to 0.88 pg/mL according to the receiver operating characteristic curve and the area under the curve of the CLBR. As shown in Supplementary Table 1, https://links.lww.com/CM9/B249, the patients in the GnRH-ant group were insignificantly older than those in the long GnRH-a group (37.0 [interquartile range (IQR), 33.0–39.0] years vs. 36.0 [IQR, 32.0–39.0] years, P = 0.099). The groups differed significantly in ovarian reserve markers (AMH, AFC, and FSH). However, other parameters such as basal estradiol and prolactin levels, FSH/LH ratio, BMI, infertility duration, and initial GnRH dosage were similar between the two protocols. Therefore, PSM using a logit model was used to adjust for age and ovarian reserve markers, the correction threshold was set to 0.02, and 1:1 was used for matching. We then compared the differences in the basic characteristics, COS parameters, laboratory indicators, and clinical outcomes between the GnRH-a and GnRH-ant protocols before and after PSM in each subgroup. According to age stratification [Table 1 and Supplementary Table 2, https://links.lww.com/CM9/B249], before PSM of POSEIDON group 4, basic characteristics (AMH and AFC), COS parameters (GnRH dosage regimen, number of follicles ≥14 mm in diameter, and estradiol concentration on the trigger day), laboratory indicators (number of retrieved oocytes, MII oocytes, two-pronuclear [2PN] embryos, and cleavage embryos, and CLBR in the GnRH-ant group were lower than those in the GnRH-a protocol (P < 0.05). The remaining indices were similar between the groups. After PSM, the laboratory indicators in the GnRH-ant group remained significantly lower than those in the GnRH-a group, while the basic characteristics and CLBR were similar in both groups. Similar results were obtained for POSEIDON group 3. Table 1 - Comparison of the treatment groups before and after PSM for patients in each group stratification by age and AMH levels. Before PSM of POSEIDON 4 group After PSM Parameter GnRH-a protocol GnRH-ant protocol P value GnRH-a protocol GnRH-ant protocol P-value POSEIDON 4 group No. of cases 107 145 93 93 Age (years) 39.00 (37.00, 41.00) 39.00 (37.00, 42.00) 0.55 39.00 (37.00, 41.00) 39.00 (36.00, 42.00) 0.63 AMH (ng/mL) 0.87 (0.67, 1.01) 0.73 (0.47, 0.96) 0.004 0.87 (0.63, 1.01) 0.77 (0.53, 1.01) 0.24 AFC (number) 4.00 (3.00, 7.00) 4.00 (3.00, 6.00) 0.03 4.00 (3.00, 6.00) 4.00 (3.00, 6.00) 0.62 Basal FSH (IU/L) 6.62 (5.60, 8.64) 6.74 (5.74, 9.61) 0.39 6.87 (5.66, 9.17) 6.62 (5.80, 9.36) 0.97 No. of retrieved oocytes 5.00 (4.00, 8.00) 4.00 (3.00, 5.00) <0.001 5.00 (3.00, 7.00) 4.00 (3.00, 5.50) 0.00 Cumulative live births, n (%) 0.001 0.13 Yes 30 (28.00) 18 (12.40) 21 (22.60) 14 (15.10) No 77 (72.00) 127 (87.60) 72 (77.40) 79 (84.90) POSEIDON 3 group No. of cases 74 71 46 46 Age (years) 31.50 (29.00, 33.00) 31.00 (29.00, 33.00) 0.89 32.00 (28.75, 33.00) 31.00 (29.00, 33.00) 0.82 AMH (ng/mL) 0.93 (0.75, 1.02) 0.73 (0.52, 0.91) <0.001 0.88 (0.55, 0.99) 0.84 (0.66. 0.95) 0.79 AFC (number) 6.00 (4.00, 8.00) 5.00 (3.00, 6.00) 0.002 5.00 (4.00, 7.00) 5.00 (4.00, 7.00) 0.97 Basal FSH (IU/L) 6.67 (5.18, 8.43) 7.51 (5.87, 10.14) 0.11 7.01 (5.65, 9.07) 7.10 (5.56, 9.88) 0.95 No. of retrieved oocytes 7.00 (4.00, 10.00) 5.00 (3.00, 6.00) <0.001 6.00 (4.00, 9.00) 5.00 (3.00, 6.00) 0.02 Cumulative live births, n (%) 0.15 0.39 Yes 37 (50.00) 27 (38.00) 20 (43.5) 16 (34.8) No 37 (50.00) 44 (62.00) 26 (56.5) 30 (65.2) High AMH levels group No. of cases 93 68 47 47 Age (years) 36.00 (32.00, 39.00) 37.50 (34.00, 40.75) 0.05 37.00 (32.00, 40.00) 38.00 (33.00, 40.00) 0.47 AMH (ng/mL) 1.01 (0.96, 1.08) 1.02 (0.94, 1.09) 0.97 1.04 (0.95, 1.12) 1.00 (0.94, 1.08) 0.26 AFC (number) 5.00 (4.00, 8.00) 4.00 (3.00, 6.00) <0.001 5.00 (4.00, 6.00) 4.00 (4.00, 6.00) 0.93 Basal FSH (IU/L) 6.46 (5.19, 7.83) 6.70 (5.58, 10.08) 0.13 6.72 (5.65, 8.25) 6.30 (5.54, 9.01) 0.59 No. of retrieved oocytes 7.00 (4.50. 9.50) 4.50 (3.00, 6.00) <0.001 6.00 (4.00, 8.00) 5.00 (4.00, 8.00) 0.01 Cumulative live births, n (%) <0.001 0.13 Yes 43 (46.20) 14 (20.60) 19 (40.40) 12 (25.50) No 50 (53.80) 54 (79.40) 28 (59.60) 35 (74.50) Low AMH levels group No. of cases 88 148 80 80 Age (years) 37.00 (33.00, 40.00) 37.00 (33.00, 40.00) 0.71 37.00 (33.00, 40.00) 38.00 (32.50, 41.00) 0.48 AMH (ng/mL) 0.69 (0.50, 0.80) 0.60 (0.43, 0.75) 0.01 0.66 (0.48, 0.80) 0.67 (0.48, 0.80) 0.94 AFC (number) 4.50 (3.00, 7.00) 4.00 (3.00, 6.00) 0.06 4.00 (3.00, 6.00) 5.00 (3.00, 6.00) 0.94 Basal FSH (IU/L) 7.03 (5.68, 9.45) 7.07 (5.83, 9.56) 0.71 7.12 (5.79, 9.60) 6.59 (5.61, 9.04) 0.25 No. of retrieved oocytes 5.00 (3.00, 8.00) 4.00 (3.00, 5.00) <0.001 5.00 (3.00, 8.00) 4.00 (3.00, 5.75) 0.03 Cumulative live births, n (%) 0.27 0.44 Yes 24 (27.30) 31 (20.90) 19 (23.80) 15 (18.80) No 64 (72.70) 117 (79.10) 61 (76.20) 65 (81.20) Non-normally distributed data are presented as n (%) or median (interquartile ranges).AFC: Antral follicle count; AMH: Anti-Müllerian hormone; FSH: Follicle-stimulating hormone; GnRH: Gonadotropin releasing hormone; GnRH-a: GnRH agonist; GnRH-ant: GnRH antagonist; POSEIDON: Patient Oriented Strategies Encompassing Individualized Oocyte Number; PSM: Propensity score matching. Before PSM, in the high AMH levels group [Table 1], the AFC levels, GnRH dosage regimen, estradiol concentration on the hCG administration day, number of follicles ≥14 mm, retrieved oocytes, 2PN embryos, the cumulative clinical pregnancy rates (CCPR), and CLBR were significantly lower in the GnRH-ant group than those in the GnRH-a group. However, the number and quality of available embryos were similar in both the groups. After PSM, the difference between the groups was limited to GnRH dosage regimen, the number of retrieved oocytes, and 2PN embryos. Before PSM, in the low AMH levels group [Table 1], the GnRH dosage was lower and the administration time was shorter, resulting in lower estradiol concentrations on the trigger day, lesser follicles ≥14 mm, retrieved oocytes, and 2PN embryos in the GnRH-ant group than those in the GnRH-a group. However, the number and quality of available embryos, CCPR, and CLBR were similar between the two treatment groups. After PSM, only GnRH dosage regimen, the number of follicles ≥14 mm, retrieved oocytes, and 2PN embryos were significantly different. The POSEIDON group introduced a new method to evaluate successful ART treatment by determining the number of retrieved oocytes required to obtain at least one euploid embryo for transfer.[1] Since the number of retrieved oocytes is related to CLBR,[2] maximizing this number is critical. In summary, post-PSM, the number of retrieved oocytes in the GnRH-a group was significantly higher than that in the GnRH-ant group in each subgroup; however, the difference in CLBR between the two groups was insignificant. Therefore, oocyte quality after treatment with the GnRH-a protocol may be slightly inferior to that with GnRH-ant. However, the long GnRH-a protocol insignificantly improved the CLBR by increasing the number of retrieved oocytes in each subgroup. These results are consistent with those of previous studies.[4,5] Two large retrospective studies[6,7] have recently applied POSEIDON stratification to explore the CLBR among patients who underwent repeated ART cycles. However, they did not consider the difference between the GnRH-a and GnRH-ant protocols. In addition, the POSEIDON criteria used by Shi et al[7] were based on the values of age and AFC, but did not include AMH levels. To our knowledge, this is a rare study to compare the differences between the GnRH-ant and GnRH-a protocols in POSEIDON groups 3 and 4 with their first COS cycles based on the PSM method. However, the study had several limitations, including its retrospective nature, small sample size, and the selection of protocols based on clinicians’ preferences. Therefore, a large RCT should be conducted to provide a more reasonable COS protocol in POSEIDON groups 3 and 4. In conclusion, although the increase in CLBR with the GnRH-a protocol was insignificant, it is preferable for patients in POSEIDON groups 3 and 4, to significantly increase the number of retrieved oocytes. Acknowledgements The authors thank Jinzhao Xie (School of Public Health, Sun Yat-sen University) for the support of statistical analyses. Funding The study was supported by grants from the Natural Science Foundation of Guangdong Province (No.2021A1515010290), National Natural Science Foundation of China (No.81871159) and Guangdong Basic and Applied Basic research (No.2022A1515012599). Conflicts of interest None.