The impact of unexplained spontaneous miscarriage history on cycle outcomes in patients undergoing in vitro fertilization: an analysis of 6535 patients.

Over a span of four and a half years, 6535 patients and their first oocyte retrieval cycles were recruited in our study. By June 2022, 4236 of them with viable embryos have completed a total of 7915 embryo transfer cycles and Fig.  1 shows a summary profile of the study. Fig. 1 Flow chart of the study Flow chart of the study As shown in Table  1 , the average number of spontaneous miscarriages in Group B was 1.33 ± 0.77. Compared with patients without miscarriage history (Group A), couples with a history of spontaneous miscarriage (Group B) were older, while their duration of infertility was shorter and a smaller proportion of births. Accordingly, the basal FSH and E2 level were higher in Group B, suggesting a relatively decreased ovarian reserve. For the rest of the basic characteristics, there were no significant differences between the two groups, including the female BMI, basal LH, and basal P4. Due to the older age and poorer ovarian reserve, fewer follicles and lower total E2 level were observed on trigger day in Group B, whilst the average E2 value was comparable. Subsequently, together with a lower oocyte retrieval rate, patients in Group B presented with fewer oocytes retrieved/matured, less cleaved embryos, and fewer viable embryos. The mature oocyte rate, fertilization rate, and cleavage rate were broadly similar between the two groups, but the oocyte utilization rate was obviously lower in patients with miscarriage histories. Table 1 Baseline profiles and COH cycle outcomes of patients with and without miscarriage histories Cycle characteristics and outcomes Group A No miscarriage history ( n  = 5023) Group B A history of miscarriage ( n  = 1512) P value a Group B1 ( n  = 1166) Group B2 ( n  = 230) Group B3+ ( n  = 116) P value b Maternal age (year) 34.77 ± 3.49 35.52 ± 3.24 <0.001 35.31 ± 3.22 35.77 ± 3.14 37.16 ± 3.18 <0.001 Paternal age (year) 36.62 ± 4.99 37.27 ± 4.53 <0.001 37.02 ± 4.56 37.57 ± 4.26 39.16 ± 4.34 <0.001 Maternal BMI (kg/m 2 ) 21.20 ± 1.48 21.20 ± 1.50 0.839 21.18 ± 1.48 21.07 ± 1.48 21.68 ± 1.70 0.001 Duration of infertility (year) 2.59 ± 2.58 2.22 ± 2.37 <0.001 2.32 ± 2.36 1.91 ± 2.52 1.83 ± 2.16 0.011 Number of previous miscarriages (n) 0.00 ± 0.00 1.33 ± 0.77 <0.001 1.00 ± 0.00 2.00 ± 0.00 3.38 ± 1.23 <0.001 Primary infertility (n, %) 0 (0.0%) 0 (0.0%) / 0 (0.0%) 0 (0.0%) 0 (0.0%) / Pluriparous (n, %) 1015 (20.2%) 208 (13.8%) <0.001 178 (15.3%) 21 (9.2%) 8 (6.9%) <0.001 Cause of infertility (n/%) <0.001 <0.001 Tubal factor 2613 (52.02) 1095 (72.42) 796 (68.27) 189 (82.17) 110 (94.83) Ovulatory dysfunction 250 (4.98) 13 (0.86) 12 (1.03) 1 (0.43) 0 (0) Male factor 458 (9.12) 12 (0.79) 11 (0.94) 1 (0.43) 0 (0) Endometriosis 154 (3.07) 17 (1.12) 15 (1.29) 1 (0.43) 1 (0.86) Unexplained factor 141 (2.81) 43 (2.84) 38 (3.26) 4 (1.74) 1 (0.86) Mixed factors 1407 (28.01) 332 (21.96) 294 (25.21) 34 (14.78) 4 (3.45) AFC (n) 10.87 ± 6.33 10.63 ± 6.74 0.212 10.93 ± 6.80 10.17 ± 6.56 8.51 ± 6.17 0.001 Basal FSH (IU/L) 5.67 ± 1.62 5.78 ± 1.61 0.018 5.76 ± 1.60 5.88 ± 1.58 5.77 ± 1.75 0.595 Basal LH (IU/L) 3.25 ± 1.95 3.36 ± 2.27 0.081 3.39 ± 2.28 3.32 ± 2.26 3.25 ± 2.18 0.784 Basal E2 (pg/ml) 33.22 ± 14.58 35.83 ± 19.94 <0.001 34.58 ± 15.71 35.31 ± 15.68 49.41 ± 45.06 <0.001 Basal P4 (ng/ml) 0.28 ± 0.13 0.28 ± 0.13 0.678 0.28 ± 0.13 0.30 ± 0.13 0.26 ± 0.13 0.073 Duration of cycle treatment (d) 11.32 ± 2.34 11.22 ± 2.26 0.133 11.27 ± 2.20 11.06 ± 1.85 11.03 ± 3.31 0.278 HMG dose (IU) 1871.25 ± 569.69 1845.60 ± 591.30 0.128 1861.84 ± 572.11 1817.61 ± 572.24 1737.93 ± 781.07 0.073 E2 on trigger day (pg/ml) 2863.69 ± 1528.62 2734.05 ± 1564.31 0.004 2803.10 ± 1551.32 2684.77 ± 1603.74 2137.72 ± 1496.38  10 mm on trigger day (n) 11.62 ± 7.38 11.17 ± 7.74 0.040 11.56 ± 7.75 10.91 ± 8.17 7.78 ± 5.62  10 mm) (pg/ml) 270.21 ± 104.24 269.52 ± 106.97 0.823 267.04 ± 103.55 273.12 ± 110.88 287.27 ± 129.79 0.130 Number of punctured follicles (n) 13.21 ± 9.14 12.75 ± 9.71 0.099 13.17 ± 9.84 12.56 ± 9.90 8.90 ± 6.73 <0.001 Oocytes retrieved (n) 9.61 ± 7.04 9.11 ± 7.31 0.015 9.42 ± 7.45 8.95 ± 7.28 6.27 ± 5.03 <0.001 MII oocytes (n) 8.25 ± 6.08 7.83 ± 6.38 0.020 8.09 ± 6.46 7.67 ± 6.51 5.54 ± 4.61 <0.001 Fertilized oocytes (n) 6.63 ± 5.09 6.28 ± 5.39 0.021 6.50 ± 5.46 6.19 ± 5.52 4.26 ± 3.80 <0.001 Cleaved embryos (n) 6.48 ± 4.99 6.14 ± 5.29 0.019 6.36 ± 5.36 6.03 ± 5.37 4.16 ± 3.77 <0.001 Viable embryos (n) 3.59 ± 3.17 3.26 ± 3.36 0.001 3.35 ± 3.47 3.28 ± 3.20 2.29 ± 2.20 0.005 Oocyte retrieval rate (%) 0.73 ± 0.22 ( n  = 5023) 0.72 ± 0.22 ( n  = 1512) 0.022 0.72 ± 0.22 0.72 ± 0.23 0.69 ± 0.28 0.243 Mature oocyte rate (%) 0.87 ± 0.17 ( n  = 4991) 0.87 ± 0.18 ( n  = 1492) 0.728 0.87 ± 0.18 0.87 ± 0.18 0.87 ± 0.21 0.934 Fertilization rate (%) 0.81 ± 0.21 ( n  = 4953) 0.80 ± 0.24 ( n  = 1476) 0.231 0.80 ± 0.22 0.80 ± 0.22 0.82 ± 0.42 0.742 Cleavage rate (%) 0.98 ± 0.08 ( n  = 4869) 0.98 ± 0.08 ( n  = 1436) 0.434 0.98 ± 0.09 0.98 ± 0.07 0.97 ± 0.10 0.633 Viable embryo rate per oocyte retrieved (%) 0.40 ± 0.26 ( n  = 4991) 0.38 ± 0.27 ( n  = 1492) 0.006 0.37 ± 0.27 0.39 ± 0.27 0.39 ± 0.30 0.594 Note: P value a referred to the P value for Group A and Group B. P value b referred to the P value for Group B1, Group B2 and Group B3+ Oocyte retrieval rate was calculated by dividing the total number of retrieved oocytes by the number of punctured follicles. Fertilization/cleavage rate was calculated by dividing the number of fertilized/cleavage embryos by the total number of mature/fertilized oocytes. Viable embryo rate per oocyte retrieved was calculated by dividing the number of viable embryos by the number of retrieved oocytes Baseline profiles and COH cycle outcomes of patients with and without miscarriage histories Note: P value a referred to the P value for Group A and Group B. P value b referred to the P value for Group B1, Group B2 and Group B3+ Oocyte retrieval rate was calculated by dividing the total number of retrieved oocytes by the number of punctured follicles. Fertilization/cleavage rate was calculated by dividing the number of fertilized/cleavage embryos by the total number of mature/fertilized oocytes. Viable embryo rate per oocyte retrieved was calculated by dividing the number of viable embryos by the number of retrieved oocytes Table  1 also describes a phenomenon that with the increasing number of previous miscarriages, most of the above-mentioned trends in Table  1 became more apparent. In other words, when the miscarriage times increased from 1 to more than 3, patients in the subgroups showed advanced age, shorter infertility duration, less pluriparous, higher basal E2, lower AFC, and consequently fewer follicles/oocytes/embryos. Notably, although women with a history of miscarriages had a lower oocyte utilization rate, as the number of miscarriages increased, this indicator would not decrease further. That was, those with 1, 2, or ≥ 3 times of miscarriage had comparable oocyte utilization rate. To further explain where the difference in oocyte utilization rate between Group A and Group B came from, especially to assess whether the miscarriage history was an independent influencing factor on the oocyte utilization rate, a multiple linear regression logistic analysis was performed (Table  2 ). The results revealed that the history of spontaneous miscarriage had a significantly negative impact on the oocyte utilization rate. Table 2 Multiple linear regression analysis of the influencing factors for viable embryo rate per oocyte retrieved Variables β SE 95%CI P value Maternal age (year) -0.003 0.0012 -0.005, 0.000 0.026 Paternal age (year) 0.001 0.0008 0.000, 0.003 0.163 Duration of infertility (year) -0.002 0.0013 -0.005, 0.000 0.070 Previous miscarriage history (n, %) -0.021 0.0079 -0.036, -0.006 0.008 Pluriparous (n, %) 0.017 0.0082 0.001, 0.033 0.038 Basal FSH (IU/L) 0.001 0.0021 -0.003, 0.005 0.761 Basal E2 (pg/ml) 3.335E-5 3.7831E-5 -4.080E-5, 0.000 0.378 E2 on trigger day (pg/ml) -7.302E-6 2.0328E-6 -1.129E-5, -3.317E-6 0.000 Follicles > 10 mm on trigger day (n) 0.000 0.0006 -0.001, 0.033 0.694 Oocytes retrieved (n) -0.006 0.0006 -0.007, -0.005 <0.001 Note: SE, standard error Multiple linear regression analysis of the influencing factors for viable embryo rate per oocyte retrieved Note: SE, standard error Pregnancy outcomes in embryo transfer cycles are detailed in Tables  3 and 4 . As displayed in Table  3 , patients in Group B had an average of 1.30 ± 0.68 times of previous miscarriage. While in Group B the couples’ ages were older, the infertility duration was shorter and the proportion of parity was smaller, there were no differences in the maternal BMI, single embryo transfer rate, FET rate, and cleavage-stage embryo transfer rate between the two groups. The biochemical pregnancy rate, the clinical pregnancy rate, the ongoing pregnancy rate, the live birth rate, the implantation rate, and the cumulative pregnancy rate per woman were all lower in Group B than those in Group A. An unexpected but impressive result was that no statistical difference was observed regarding the miscarriage rate. Moreover, the ectopic pregnancy rate was also comparable between the two groups. Table 3 Embryo transfer cycle characteristics and outcomes of patients with and without miscarriage histories Cycle characteristics and outcomes Group A No miscarriage history ( n  = 6193) Group B A history of miscarriage ( n  = 1722) P value a Group B1 ( n  = 1347) Group B2 ( n  = 266) Group B3+ ( n  = 109) P value b Maternal age (year) 35.56 ± 3.52 36.29 ± 3.27 <0.001 36.05 ± 3.22 36.76 ± 3.04 38.12 ± 3.82 <0.001 Paternal age (year) 37.45 ± 5.14 38.06 ± 4.52 <0.001 37.87 ± 4.58 38.42 ± 4.07 39.49 ± 4.45 0.001 Maternal BMI (kg/m 2 ) 21.18 ± 1.48 21.19 ± 1.52 0.719 21.20 ± 1.50 21.01 ± 1.47 21.52 ± 1.80 0.010 Duration of infertility (year) 2.56 ± 2.64 2.19 ± 2.39 <0.001 2.27 ± 2.34 1.94 ± 2.55 1.83 ± 2.54 0.029 Number of spontaneous miscarriages (n) 0.00 ± 0.00 1.30 ± 0.68 <0.001 1.00 ± 0.00 2.00 ± 0.00 3.29 ± 0.97 <0.001 Primary infertility (n, %) 0 (0.0%) 0 (0.0%) / 0 (0.0%) 0 (0.0%) 0 (0.0%) / Pluriparous (n, %) 1277 (20.6%) 228 (13.3%) <0.001 196 (14.6%) 17 (6.4%) 15 (13.8%) <0.001 Endometrium thickness (mm) 10.35 ± 1.21 10.26 ± 1.42 0.278 10.30 ± 1.30 10.21 ± 1.45 9.58 ± 1.52 0.107 Single embryo transfer rate (%) 25.8 (1596/6193) 25.8 (445/1722) 0.952 24.6 (332/1347) 27.4 (73/266) 33.0 (36/109) 0.531 FET rate (%) 88.3 (5467/6193) 89.3 (1537/1722) 0.260 88.9 (1198/1347) 90.2 (240/266) 90.8 (99/109) 0.711 Cleavage stage transplantation rate (%) 82.9 (5137/6193) 82.5 (1420/6193) 0.636 82.8 (1115/1347) 80.5 (/214266) 83.5 (99/109) 0.633 Biochemical pregnancy rate (%) 51.6 (3193/6193) 47.5 (818/1722) 0.003 48.3 (650/1347) 46.6 (124/266) 40.4 (44/109) 0.270 Clinical pregnancy rate (%) 46.1 (2852/6193) 42.3(729/1722) 0.006 43.1 (580/1347) 42.5 (113/266) 33.0 (36/109) 0.125 Ongoing pregnancy rate (%) 40.5 (2508/6193) 37.0 (637/1722) 0.009 38.1 (513/1347) 35.7 (95/266) 26.6 (29/109) 0.052 Ectopic pregnancy rate (%) 1.9 (120/6193) 1.5 (25/1722) 0.184 1.6 (21/1347) 1.1 (3/266) 0.9 (1/109) 0.771 Miscarriage rate (%) 14.9 (425/2852) 15.9 (116/729) 0.497 14.8 (86/580) 18.6 (21/113) 25.0 (9/36) 0.189 Implantation rate (%) 33.0 (3560/10790) 30.4 (912/2999) 0.008 31.4 (740/2359) 29.8 (137/459) 28.2 (51/181) 0.575 Cumulative pregnancy rate (%) 80.4 (2682/3337) 76.5 (688/899) 0.011 77.8 (542/697) 75.2 (109/145) 64.9 (37/57) 0.081 Live birth rate (%) 38.5 (2382/6193) 35.0 (603/1722) 0.009 35.9 (484/1347) 34.6 (92/266) 24.8 (27/109) 0.062 Note: P value a referred to the P value for Group A and Group B. P value b referred to the P value for Group B1, Group B2 and Group B3+ Embryo transfer cycle characteristics and outcomes of patients with and without miscarriage histories Note: P value a referred to the P value for Group A and Group B. P value b referred to the P value for Group B1, Group B2 and Group B3+ Table 4 Binary logistic regression analysis of the influencing factors for pregnancy outcomes Cycle characteristics and outcomes Adjusted OR value P value Miscarriage rate Maternal BMI (kg/m2) 1.103 (1.039, 1.171) 0.001 Spontaneous miscarriage history 1.013 (0.866, 1.185) 0.869 Clinical pregnancy rate Maternal age (year) 0.954 (0.938, 0.971) <0.001 Paternal age (year) 0.981 (0.969, 0.993) 0.001 Blastocyst transfer compared to cleavage-stage embryo transfer 1.205 (1.062, 1.367) 0.004 Frozen-thawed embryo transfer compared to fresh embryo transfer 1.428 (1.218, 1.675) <0.001 Spontaneous miscarriage history 0.922 (0.850, 1.001) 0.054 Endometrial thickness on embryo transfer day (mm) 1.039 (1.022, 1.057) <0.001 Number of transferred embryos (n) 1.416 (1.271, 1.578) <0.001 Ongoing pregnancy rate Maternal age (year) 0.952 (0.935, 0.968) <0.001 Paternal age (year) 0.976 (0.965, 0.988) <0.001 Blastocyst transfer compared to cleavage-stage embryo transfer 1.150 (1.012, 1.307) 0.032 Frozen-thawed embryo transfer compared to fresh embryo transfer 1.432 (1.215, 1.687) <0.001 Spontaneous miscarriage history 0.960 (0.904, 1.019) 0.182 Endometrial thickness on embryo transfer day (mm) 1.037 (1.020, 1.055) <0.001 Number of transferred embryos (n) 1.487 (1.339, 1.672) <0.001 Live birth rate Maternal age (year) 0.958 (0.942, 0.975) <0.001 Paternal age (year) 0.976 (0.964, 0.988) <0.001 Blastocyst transfer compared to cleavage-stage embryo transfer 1.370 (1.161, 1.617) <0.001 Frozen-thawed embryo transfer compared to fresh embryo transfer 1.144 (1.006, 1.301) 0.040 Spontaneous miscarriage history 0.921 (0.846, 1.003) 0.057 Number of transferred embryos (n) 1.493 (1.334, 1.670) <0.001 Note: The covariates selected into the binary logistic regression analysis included maternal age, parental age, maternal BMI, parity, infertility duration, cause of infertility, endometrial thickness on embryo transfer day, fresh or frozen-thawed embryo transfer, the number and stage (cleavage or blastocyst) of embryos transferred and spontaneous miscarriage history (0, 1, 2, 3+). Only the previous spontaneous miscarriages and the variates with P  < 0.05 were presented in the table Binary logistic regression analysis of the influencing factors for pregnancy outcomes Note: The covariates selected into the binary logistic regression analysis included maternal age, parental age, maternal BMI, parity, infertility duration, cause of infertility, endometrial thickness on embryo transfer day, fresh or frozen-thawed embryo transfer, the number and stage (cleavage or blastocyst) of embryos transferred and spontaneous miscarriage history (0, 1, 2, 3+). Only the previous spontaneous miscarriages and the variates with P  < 0.05 were presented in the table Table  3 also specifies the pregnancy outcomes of women with a history of spontaneous miscarriage, which were further split into three subgroups according to their number of previous miscarriages. The results suggested that with the increase in the number of miscarriages prior to IVF-ET treatment, the age of couples increased and the infertility duration decreased. It was also observed that the patients in Group B3 + showed the highest BMI value. And patients with different miscarriage number had different proportion of parity. When comparing the pregnancy outcomes of these three subgroups, it was found that the miscarriage rate, the biochemical pregnancy rate, the clinical pregnancy rate, the ongoing pregnancy rate, the implantation rate, the ectopic pregnancy rate, the live birth rate, and the cumulative pregnancy rate per women were not significantly different, in a context where the singe embryo transfer rate, the FET rate, and the cleavage stage embryos transfer rate were generally similar among the three subgroups. To further clarify the confounding impact factors of miscarriages after embryo transfer of patients with different numbers of previous miscarriages, a multiple logistic regression was conducted (see Table  4 for details). In terms of the miscarriage rate, advanced paternal age and higher maternal BMI were associated with increased chances of miscarriages after IVF-ET, while most importantly, the number of previous miscarriages were turned out not related to miscarriages in IVF-ET cycles. Additionally, there was also no association between history of miscarriage and the other pregnancy outcomes (clinical pregnancy, ongoing pregnancy rate and live birth rate).

This retrospective observational study consisted of 6535 patients, who underwent their first IVF or ICSI (intracytoplasmic sperm injection) treatment between January 2018 and June 2022. This study was completed in the Department of assisted reproduction, the Ninth People’s Hospital, Shanghai Jiao Tong University School of medicine. The inclusion criteria included: infertile women (those with normal sexual life, who cannot conceive with more than 1 year without contraception), age less than 40 years old, body mass index (BMI) between 18.5 and 24 kg/m 2 , normal ovarian reserve (basal FSH<10IU/L), and no previous IVF history. If a patient underwent more than one IVF/ICSI cycle in the above period, only the first cycle would be analyzed. The exclusion criteria were considered as follows: (1) couples or one of them had abnormal karyotype (polymorphism not included); (2) abnormal uterus (deformity, large uterine fibroid, adenomyoma, etc.); (3) endocrine diseases (including thyroid disease, high prolactiniummia, diabetes, etc.). According to whether the patients had a history of spontaneous miscarriage, they were divided into two groups: Group A with no previous miscarriage history and Group B with a history of miscarriage. Then patients in Group B were further stratified into three subgroups based on their number of miscarriages, namely Group B1 (only once), Group B2 (two times), and Group B3+ (three or more times). The history of spontaneous miscarriage was obtained from patient reports and medical records. All patients with pregnancy count = 0 were excluded and all patients included in this study were secondary infertility patients. All patients and their spouses had normal chromosomes. All embryos in this study were not subjected to embryo aneuploidy screening. Patients generally started controlled ovarian stimulation (COH) on day 3 of the menstrual cycle. On the day, transvaginal ultrasound (TVS) was performed to check the antral follicle count (AFC), and serum FSH, LH, E2, and P4 were assayed. Then on the basis of patients’ individual profiles, ovarian stimulation protocols were applied and medications were prescribed from that day, including PPOS protocol and GnRH-ant protocol. In the PPOS protocol, patients underwent daily intramuscular injection of 150 or 225 IU of human menopausal gonadotropin (hMG) and simultaneously administered with 10 mg of medroxyprogesterone acetate (MPA) until the trigger day. In the GnRH-ant protocol, patients were injected daily with 150 or 225 IU hMG intramuscularly. Cycle monitoring starts 4–5 days after the start of stimulation. During each monitoring, the number and the size of follicles were recorded by TVS, serum steroid hormones were measured as well. GnRH-ant was added when the maximum follicle diameter was ≥ 14 mm until the trigger day. When the diameter of one dominant follicle reached 20 mm or at least three follicles reached diameters of 18 mm, the final oocyte maturation was triggered by 0.1 mg of triptorelin (decapeptyl, Ferring Pharmaceuticals) alone or co-triggered with 5,000 IU of hCG (Lizhu Pharmaceutical Trading Co., China). All follicles with diameters of over 10 mm were retrieved within 32–36 h following maturation induction under transvaginal ultrasound guidance. According to semen parameters, conventional insemination or ICSI can be used for fertilization. The number of blastomeres, regularity and degree of embryo fragmentation were evaluated on the third day after oocyte retrieval, and the embryos were scored according to Cummins’s criteria. The embryos with high quality (including grade 1 and grade 2, 6–8 cells blastomere embryos) on the examination day were either fresh transferred or frozen, taking various factors of individuals into consideration (GnRH-ant protocol was used for all fresh embryos and PPOS protocol was used for all frozen embryos). The rest embryos that were not so ideal of quality were placed in extended culture to day 5 or day 6 until the blastocyst stage, among which only good-morphology blastocysts were selected for fresh transfer or freeze. For patients who prepared inadequately for fresh embryo transfer, all highest quality embryos were frozen by vitrification. Frozen-thawed embryo transfer (FET) was performed for these cases later. Endometrial preparation was performed under the hormone replacement cycle. Embryos aged 5–6 days old were thawed for transfer on the fifth day of endometrium translation. The luteal phase was supplemented with an oral progesterone pill, together with a vaginal progesterone pill. The routine luteal support until the pregnancy test was negative; Or, if pregnant, it will last until the 12th week of pregnancy. For the newborn follow-up, highly trained nurses at our center surveyed the patients by telephone during each stage of pregnancy until delivery. A series of information was collected through a standardized questionnaire including gestational weeks, mode of delivery, newborn gender, birth date, locality, birth length, and birth weight. If the nurses failed to contact the couples, information was tried to obtain from the local family planning service agencies as an alternative. The primary outcome measure was the miscarriage rate. The secondary indicators were clinical pregnancy rate, continuing pregnancy rate and implantation rate, live birth rate, cumulative pregnancy rate, and oocyte utilization rate per oocyte retrieved (oocyte utilization rate). Normal fertilization was defined as 2PN embryos observed on the first day after conventional fertilization or ICSI. The oocyte retrieval rate was calculated as the total number of oocytes retrieved divided by the total number of follicles punctured; the mature oocyte rate was calculated as the total number of mature oocytes divided by the total number of oocytes retrieved; the fertilization rate was calculated as the total number of fertilized oocytes divided by the total number of mature oocytes; and the cleavage rate was calculated as the total number of cleaved embryos divided by the total number of fertilized oocytes. The viable embryo rate per oocyte (oocyte utilization rate) was defined as the number of viable embryos divided by the number of oocytes retrieved. Biochemical pregnancy was defined as a pregnancy 14 days after embryo transfer, when ultrasound examination does not detect pre gestational sac pregnancy and blood β - human chorionic gonadotropin (hCG) is positive (> 5IU/L). Clinical pregnancy is confirmed by ultrasound showing an intrauterine gestational sac. Ectopic pregnancy referred to a pregnancy in which an embryo is implanted outside the uterine cavity. Implantation rate was defined as the number of gestational sac divided by the number of transferred embryos. The abortion rate was defined as the pregnancy rate of spontaneous termination of pregnancy. The ongoing pregnancy rate was referred to the proportion of pregnancies more than 12 weeks of gestational age. The live birth rate was identified as the number of live-birth dividing the number the embryo transfer cycle. The cumulative pregnancy rate denoted the final number of pregnant patients divided by the number of patients who underwent their embryo transfer cycles during the period. Perform statistical analysis using SPSS 22.0 software (SPSS, Inc.). Quantitative data is expressed as mean ± standard deviation; Qualitative data is presented in percentages. Use t-test to compare the means between two groups, and one-way ANOVA to compare the means between three groups. Perform ratio comparison through chi square test or Fisher’s exact test (depending on the situation). Use binary logistic regression to analyze the impact of spontaneous miscarriage history (0, 1, 2, ≥ 3 times) and other covariates (patient basic information and ART parameters) on the miscarriage rate of embryo transfer cycles. The covariates were selected according to literature and biological plausibility, including maternal/parental age, maternal BMI, parity, infertility duration, cause of infertility, endometrial thickness on embryo transfer days, fresh or frozen-thawed embryo transfer, as well as the number and stage of embryos transferred. Additionally, the same set of potential confounders was introduced in the regression models to calculate the impact of miscarriage history on other pregnancy outcomes (implantation, biochemical/clinical/ongoing pregnancy, ectopic pregnancy, and live birth), all of which using the enter method. Furthermore, a multiple linear regression was performed to assess the effects of spontaneous miscarriage history on oocyte utilization rate. Statistically significant factors were included in the logistic regression equation and an adjusted odds ratio (aOR) and CI were displayed. A P value was considered statistically significant if < 0.05.

To summarize, in this study, we found that for patients with unexplained spontaneous miscarriage histories, the oocyte utilization rate was lower than that of patients without miscarriage histories. In the subsequent embryo transfer cycles, the miscarriage rate was demonstrated not associated with previous history of miscarriage, once the patients had satisfied embryos for transfer. Moreover, the biochemical/clinical/ongoing pregnancy rate, live birth rate, implantation rate, and ectopic pregnancy rate were all not related to the number of previous spontaneous miscarriages. Additionally, once high-quality embryos are achieved, the history of spontaneous miscarriage does not have a negative impact on the pregnancy outcomes. These may provide some guidance for IVF-ET treatment of patients with a miscarriage history.

In this cohort study of 6535 first oocyte retrieval cycles and 7915 subsequent embryo transfer cycles collected from 6535 patients, we found that compared with patients without a history of spontaneous miscarriages, patients with a miscarriage history had lower oocyte utilization rate, while this indicator did not further decrease with the increase in the number of previous miscarriages. Furthermore, patients with a miscarriage history also presented with lower biochemical pregnancy rate, clinical pregnancy rate, ongoing pregnancy rate, live birth rate, and implantation rate than those of patients without a history of spontaneous miscarriages. Most importantly, further multivariate analysis showed that the miscarriage rate was not related to the number of previous spontaneous miscarriages, once the patients had viable embryos for transfer. In addition, the other pregnancy outcomes, including the biochemical/clinical/ongoing pregnancy, live birth, ectopic pregnancy, and implantation, were all demonstrated to be irrespective of a history of miscarriages. About 50% of patients with RPL can be etiologically diagnosed and receive appropriate reproductive treatment [ 9 , 17 ]. However, for the remaining unexplained RPL patients, whether they should receive IVF-ET treatment has been controversial. Anyhow, IVF-ET provides a method to investigate both earlier and later stages of reproduction performance in patients with unexplained miscarriages, which may be helpful for the diagnosis and treatment of these patients in the future [ 12 , 16 , 18 , 19 ]. Early stages involve oocyte retrieval, fertilization, cleavage, and blastocyst formation, whereas later stages refer to implantation, pregnancy, and neonatal outcomes. In this study, we investigated the cycle performances of patients with unexplained miscarriage histories in IVF-ET treatment by analyzing the clinical treatment data in our center over the past four years, which is barely explored comprehensively in previous studies. In the oocyte retrieval cycles, we found that patients in the miscarriage group were older and had relatively poor ovarian reserves, and there were no differences in the mature oocyte rate, fertilization rate, cleavage rate, and average E2 level on trigger day between the two groups. For the most important indicator, the oocyte utilization rate, an obvious decrease was detected in the miscarriage history group, suggesting that the oocytes qualities of patients in this group are relatively inferior, which may in turn, also be a possible explanation for the previous spontaneous miscarriages in this group. When further stratifying the patients in the miscarriage group according to their previous number of miscarriages, it was revealed that the increase in the number of miscarriages would not lead to a decrease in the oocyte utilization rate. The multiple linear regression analysis further confirmed that the history of miscarriages, rather than the number of spontaneous miscarriages, was associated with the oocyte utilization rate. However, when we reviewed the published studies [ 1 , 12 , 18 – 20 ], little literature was available regarding the performances of patients with different numbers of miscarriages during oocyte retrieval cycles. The only study concentrating on the oocyte retrieval cycle outcomes of patients with miscarriage history reported a similar finding to ours, in which the oocyte utilization rate was 43.39% in the abortion group and 44.54% in the non-abortion control group [ 21 ]. This difference was not so significant as ours, which we speculate is due to the small sample size in the previous study. Hence, our study on the oocyte retrieval cycle indicates that miscarriage history does affect the oocyte utilization rate negatively in IVF, however, once patients with unexplained previous miscarriages can obtain oocytes, their probabilities of obtaining viable embryos are not associated with the exact number of miscarriages. In the embryo transfer cycles, it was worth note that there was no difference in the miscarriage rate between the two groups, although the rate of biochemical pregnancy, clinical pregnancy, ongoing pregnancy, cumulative pregnancy, live birth, and implantation in the miscarriage group were significantly lower than those of patients without a history of miscarriage. Based on the literature available, the miscarriage rate of recurrent miscarriage patients was found to be higher than that of patients without miscarriage histories, both in patients conceived naturally and through IVF/ICSI [ 1 , 18 ]. Though with discrepancies (similar miscarriage rate in our study and higher miscarriage rate in previous ones), we make interpretations as follows: (1) The COH process with external hormone use may alter the endocrine situations where follicles grow, thus some unknown negative factors interfere with the development of follicles might be avoided; (2) Through in vitro culturing and grading of embryos, only the embryos of good qualities are selected for transfer, eliminating those of poor developmental competences, which is in accordance with the decreased oocyte utilization rate in the spontaneous miscarriage group. This selection process in our study is achieved by the operation of IVF itself, instead of extra intervention such as preimplantation genetic testing for aneuploidies (PGT-A). (3) We could obtain improved endometrium by appropriate endometrium preparation schemes, and preferable uterine environment by hysteroscope in some individuals before embryo transfer. Nonetheless, the underlying mechanism remains to be studied. The comparisons among the subgroups of the spontaneous miscarriage group showed that the number of spontaneous miscarriages did not affect the miscarriage rate, as well as other pregnancy-related indicators, including the rate of implantation, biochemical pregnancy, clinical pregnancy, ongoing pregnancy, and live birth. Further multiple logistic regressions also presented that the occurrence of miscarriage and the success of pregnancy in IVF-ET cycles were not related to the number of previous miscarriages. That is to say, patients with or without previous miscarriage histories are equally fertile in the embryo transfer cycles. This result suggests that as long as high-quality embryos can be obtained, even if there is a history of spontaneous miscarriages in the past, it will not have a negative impact on their way to achieving pregnancy and live birth. Nevertheless, it was indeed an unexpected result that these pregnancy-related indicators did not decrease further with the increase in previous miscarriage number. Moreover, it was noteworthy that no statistical difference was observed regarding the miscarriage rate among the subgroups with different previous miscarriage number. Although p  > 0.05, however, the miscarriage rate did increase with the increase in the number of previous miscarriages (14.8% vs. 18.6% vs. 25.0%). We speculate that these results could be partially due to the limited sample size in the subgroups and we might observe a statistical difference with a larger cohort. For this point, even if the duration of in vitro culture and embryo self-selection in IVF-ET treatment may be a reason to decrease miscarriage rate, there is still some unknown reasons which lead to the higher miscarriage rate (25%). In addition, we compared our data with the probability of miscarriage after the spontaneous conception of the patients with miscarriage histories in the published literature (Figure S1 ) [ 1 ]. We found that the probability of recurrent miscarriage in patients treated with IVF-ET in our center was obviously lower than those in the population with natural pregnancies in the literature (Our data vs. published data: 14.8% vs. 19.8% / n  = 5051 for once, 18.5% vs. 27.7% / n  = 890 for twice and 25.0% vs. 41.9% / n  = 273 for three or more times). Though not obtained from the same center/population, this comparison of large samples suggests that IVF-ET treatment might be helpful for patients with a history of miscarriage to reduce the incidence of recurrent miscarriage. This effect is probably achieved by ovarian stimulation, embryo culturing/grading, or improved endometrial/intrauterine conditions, as mentioned above. However, we still should be cautious when making this speculation and the more confirmed results awaits for following designed study to explore. To date, our study is one of the largest retrospective ones focused on the IVF/ICSI outcomes of patients with unexplained miscarriage histories. Moreover, our study is among the first to report the oocyte retrieval cycle characteristic and performances of patients with unexplained miscarriages. We also preliminarily speculate that the early stages of the reproductive process (involving oocyte retrieval, fertilization, cleavage, and blastocyst formation) might participate in the beneficial effects of IVF (similar miscarriage rate) on patients with miscarriage histories, by observing an inferior oocyte utilization rate in this population. Admittedly, there are also some limitations in our study. First, the baseline characteristics of the patients were not completely comparable. Although we used multivariate analysis for correction, we still do not rule out that there may be some bias. Second, exclusion of patients with abnormal uterine anatomy, endocrine disorders, or genetic abnormalities may lead to selection bias and limit external validity. Third, subgroup trends (e.g., miscarriage rate rising from 14.8 to 25.0%) suggest a possible Type II error due to underpowered subgroups, and conclusions about this should be drawn with caution. Finally, the retrospective nature and single-center design introduce inherent limitations, particularly in generalizability, thus a multi-center prospective cohort study with larger sample size is expected to be carried out in the future.

Miscarriage is a common adverse outcome of pregnancy, with a reported prevalence of 12–15% among clinically recognized pregnancies by 20 weeks of gestation [ 1 ]. Recurrent miscarriage, as an important part of recurrent pregnancy loss (RPL), defined as two or more miscarriages, not necessarily consecutive, represents a significant health problem [ 2 ]. The exact incidence of recurrent miscarriage is difficult to estimate and varies widely between reports, but most studies report that RPL affects 1–5% of all couples attempting conception [ 2 – 4 ]. Spontaneous miscarriage following spontaneous pregnancy, especially recurrent miscarriage, potentially results from a complex interplay between parental age, genetic, hormonal, immunological, uterine, and environmental factors [ 3 , 5 – 8 ]. These explanations allow for treatment interventions in the following pregnancy. Nonetheless, even after a comprehensive evaluation, a cause for recurrent miscarriage is identified in less than 50% of couples [ 9 ]. Empirical treatments are usually recommended for unexplained recurrent miscarriage, including emotional support, nutritional optimization, close monitoring, and ultimately rely on natural conception. Most of these treatments show limited effects and offers little benefit to patients. Therefore, at present, unexplained recurrent miscarriage is a frustrating health problem, without effective treatments to improve the live birth rate. Moreover, the number of previous spontaneous miscarriages is associated with the increased risk of subsequent pregnancy and perinatal complications, including early abortion, preterm birth, and perinatal death, which may also damage fertility [ 10 , 11 ]. Therefore, even though the empirical therapies are not well defined for their effectiveness, they are still offered for patients with unexplained miscarriage history. Among them, assisted reproductive technique (ART), mainly referring to in vitro fertilization and embryo transfer (IVF-ET), usually serves as the most common treatment option. Meanwhile, patients with unexplained miscarriage history sometimes suffered from concurrent secondary infertility, in which IVF-ET becomes the compelling choice. To date, whether couples with unexplained miscarriage histories can benefit from IVF-ET treatment is still a controversial issue and under a heated debate. Some studies believe that for patients with unexplained miscarriage history, especially for those diagnosed with RPL, IVF-ET treatment alone cannot improve the pregnancy efficiency, and may even prolong the time of the next pregnancy occurrence [ 12 – 15 ]. Other studies have shown that, although there are some confounding factors, the implementation of IVF-ET treatment can significantly reduce the miscarriage rate [ 16 ]. Moreover, the overall reproductive performance of women with unexplained miscarriage histories who have received IVF-ET treatment is unclarified purposefully and waiting for further investigations. In this study, we focus on the relationship between the miscarriage history and IVF outcomes to explore the reproductive performances of patients with different number of previous miscarriages in IVF-ET treatment.

Below is the link to the electronic supplementary material. Supplementary Material 1: Fig. S1: Probability of spontaneous abortion after natural or IVF pregnancy Supplementary Material 1: Fig. S1: Probability of spontaneous abortion after natural or IVF pregnancy