Results
A total of 562 cycles were included in the study with 257 cycles of fresh ET and 305 cycles of FET. In fresh ET cycles, subgroups were divided according to COS protocols, cycles of ultra-long, long, antagonist and short protocols were respectively 108, 56, 34 and 59. In EFT cycles, subgroups were ascertained according to different embryos origins, cycles of ultra-long, long, antagonist and short protocols were 98, 101, 54 and 52.
The characteristics of adenomyosis population in fresh ET was shown inTable I . Except for BMI, there were significant differences in age, duration of infertility, primary infertility, AFC, FSH, AMH, mean diameter of initial uterus and dysmenorrhea history in intragroup comparisons. The differences mainly concentrated on the comparisons between short protocol and other protocols. Compared with ultra-long, long and antagonist protocols, short protocol had older maternal age (37.00 versus 34.00, 32.50, 33.50, P=0.001), poorer AMH (1.30 versus 1.65, 2.75, 1.31, P<0.001) and lower AFC (9.0 versus 11.00, 13.00, 10.00). Compared with ultralong, antagonist and short protocols, long protocol had the highest AMH (2.75 versus 1.65, 1.32, 1.30, P<0.001) and the minimal uterine volume (5.43 versus 6.22, 6.05, 5.45, P<0.001). The proportion of severe dysmenorrhea in ultra-long protocol was highest (37.96% versus 17.86%, 11.76%, 13.56%). In terms of the comparisons among ultra-long, long and antagonist protocols, no significant differences existed in age, BMI, duration of infertility, primary infertility and FSH.
COS related parameters and pregnancy outcomes were shown inTable II . Compared with long, antagonist and short protocols, gonadotropin duration (11.00 versus 9.00, 9.00, 9.00, P <0.001) and gonadotropin dosage (2775.00 versus 1800.00, 2175.00, 1975.00, P <0.001) obviously increased in ultra-long protocol. Compared with ultra-long, long and antagonist protocols, endometrial thickness on HCG trigger day was thinnest (0.9 versus 1.1, 1.1, 1.0, P=0.002) and the number of retrieved oocytes was lowest (4.00 versus 7.00, 10.00, P<0.001) in short protocol. However, no statistical significance existed among intragroup comparisons of ultra-long, long and antagonist protocols. As to high-quality embryos on day 3, long protocol had more good embryos compared with ultra-long, antagonist and short protocols (4.00 versus 2.00, 2.00, 2.00, P<0.001).
The parameters of pregnancy outcomes included IR, BPR, EPR, CPR, MR, early MR and late MR and LBR. Compared with ultra-long and long protocols, IR (28.2% versus 49.7%, 52.1%, P=0.001) and CPR (35.6% versus 64.3%, 57.4%, P=0.004) in short protocol significantly decreased. The decrease of IR (33.3% versus 49.7%, 52.1%) and CPR (38.2% versus 57.4%, 64.3%) also existed in antagonist protocol, although no statistical significance was detected because of strict P adjustment of Bonferroni method (Padj=0.008). LBR in ultra-long, long, antagonist and short protocols was respectively 39.8%, 48.2%, 26.5% and 20.3%. Compared with long protocol, LBR in short protocol decreased obviously (48.2% versus 20.3%, P<0.001). LBR in antagonist protocol was also pessimistic (48.2% versus 26.5%), although no statistical significance. There were no significant differences in intragroup comparisons of BPR, MR, early MR and late MR.
Considering the differences of baseline parameters among four protocols, we carried out CPR and LBR associated multiple factors logistic analysis, which was shown in Table III and IV. When CPR was set as the outcome variable, after adjusting age, BMI, FSH, AFC, AMH, initial uterus diameter and dysmenorrhea history, we observed that long protocol was a protective factor compared with short protocol (OR 2.414, 95% CI 1.011-1.144, P=0.047). AFC was also a protective factor (OR 1.073, 95% CI 1.007-5.767, P=0.030), and the main risk factor was age (OR 1.073, 95% CI 1.007-5.767, P<0.001). When LBR was set as the outcome variable, after adjusting similar independent factors, ultra-long (OR 2.47, 95% CI 1.005-6.07, P=0.049) and long protocol (OR 2.786, 95% CI 1.055-7.353, P=0.039) were both protective factors compared with short protocol. Secondary infertility (OR 2.088, 95% CI 1.018-4.283, P=0.045) and AFC (OR 1.069, 95% CI 1.001-1.142, P=0.047) were also protective factors. Age (OR 0.809, 95% CI 0.743-0.882, P<0.001) was still a risk factor.
In FET cycles, we summarized baseline data of frozen embryo originating COS cycles, which was shown in Table V . Similar to characteristics in fresh ET cycles, compared with ultra-long, long and antagonist protocols, short protocol had older age (37.00 versus 33.00, 31.00, 33.00, P<0.001) and lower AMH (1.35 versus 2.84, 4.17 and 4.07, P<0.001). Age, BMI, duration of infertility, primary infertility, AFC, FSH and AMH had no statistical difference among ultra-long, long and antagonist protocols. The related parameters of frozen embryo originating COS protocols and pregnancy outcomes were shown in Table VI. Compared with long, antagonist and short protocols, gonadotropin dosage (2512.50 versus 2025.00, 1875.00 and 2025.00, P<0.001) and gonadotropin stimulation duration (11.00 versus 10.00, 9.00, 9.00, P<0.001) significantly increased in ultra-long protocol. Ultra-long and long protocol had higher E2 level on HCG trigger day compared with short protocol (3117.00, 3000.00 versus 2131.50, P<0.001). No significant differences in progesterone level (P=0.542) and endometrial thickness (P=0.676) on HCG trigger day existed among four protocols. The number of retrieved oocytes (6.5 versus 12.00, 12.00, 11.00, P<0.001) and 2PN zygotes (4.00 versus 7.00, 8.00 6.50, P<0.001) in short protocol were significantly lower than those in ultra-long, long and antagonist protocols. Absolutely different from fresh ET cycles, no statistical differences among ultra-long, long, antagonist and short protocols on IR (45.5%, 47.2%, 41.8%, 50.0%, P=0.851), BPR (12.2%, 6.9%, 5.6%,13.5%, P=0.313), CPR (43.9%, 49.5%, 40.7%, 50.0%, P=0.658), MR (42.9%,32.7%, 47.6%, 38.5%, P=0.626), early MR (27.9%, 32.0%,36.4%, 30.8%, P=0.917), late MR (14.3,-,9.5%,7.7%,P=0.76) and LBR (24.7%, 33.0%, 20.8%, 30.8%, P=0.343) were detected.