Serum human chorionic gonadotropin can predict live birth 7 days after frozen-thawed blastocyst transfer.

Nobuyuki Kidera: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Tomonori Ishikawa: Writing – review & editing, Conceptualization. Toshihiro Kawamura: Writing – review & editing, Conceptualization. Naoyuki Miyasaka: Writing – review & editing, Supervision.

We investigated 18,491 FET cycles performed at our hospital from January 2016 to December 2022, excluding embryos at the day 3 cleavage stage and multiple pregnancies in line with previous studies, considering the effect on hCG levels. The 18,491 cycles included embryos on days 5 and 6 of the blastocyst stage with 1 or 2 transferred embryos and included in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI). Serum hCG concentrations and cutoff values were investigated from days 12–16 after ovulation (i.e., from days 7–11 after FET) for natural cycle FET and addition of progesterone to hormone replacement cycle FET (HRC-FET). Serum hCG concentration was measured from the 12th to 16th days after ovulation on the day the patients opted to visit and was performed only once during the same cycle. In addition, we investigated the personal and population characteristics (i.e., maternal age, paternal age, history of pregnancy, rate of ICSI, history of delivery, number of embryo transfers, blastocysts of grade A inner cell mass (ICM), blastocysts of grade A trophectoderm (TE), HRC-FET, cause of infertility, history of endometrial operation, history of cesarean section, history of myomectomy, and day 5 blastocyst transfer) of clinical pregnancy and live birth in all 18,491 cases. Morphological evaluation of the embryos was performed using Gardner Schoolcraft criteria ( 20 ). Endometrial surgery includes endometrial curettage, manual vacuum aspiration, septal incision, submucosal myomectomy, and transcervical polypectomy. Patients with serum hCG concentration 1.0 IU/L as assayed were included in the study. Clinical pregnancy was defined as the presence of a gestational sac detected on transvaginal ultrasonography. Live births were defined as deliveries after 22 weeks of pregnancy. All procedures conformed to the ethical standards of the institutional and national research committees and the 1964 Declaration of Helsinki and its later amendments. This retrospective study was approved by the Clinical Ethics Committee of Denentoshi Ladies Clinic (25030731). Informed consent was obtained from all participants. The serum hCG concentration was performed using an AIA-CL2400 and CL AIA-PACK hCG TEST CUP (Tosoh Corporation). The calibration curve for use with the CL AIA-PACK hCG assay has been standardized against the WHO 5 th International Standard Chorionic Gonadotropin, National Institute for Biological Standards and Control Code 07/364. Ovarian stimulation was performed on the 3rd day after menstruation. For this, clomiphene citrate was prescribed at 50 mg/d and used with recombinant follicle-stimulating hormone and human menopausal gonadotropin unless the antral follicle count and antimüllerian hormone levels showed poor ovarian reserve. Clomiphene citrate alone was administered when ovarian reserve was low. Furthermore, oocyte retrieval was performed 35–37 hours after administering recombinant hCG hormone (Ovidrel, EMD Serono, Switzerland) or gonadotropin-releasing hormone agonist (Buserecur, Fujifilm Pharmaceuticals, Tokyo, Japan) when one or more ovarian follicles reached ≥18 mm in diameter on ultrasound examination. Spermatozoa were obtained by density gradient centrifugation, and a universal IVF medium with phenol red (Origio, Yokohama, Japan) was used as the fertilization medium. The insemination concentration was set at 1.0–1.5 × 10 5 /mL. After dropping the sample into a Makler chamber, spermatozoa were analyzed using a sperm motility analysis system (SMAS, DITECT, Tokyo, Japan). Embryos were exposed to the global total medium (Astec, Japan). Intracytoplasmic sperm injection was performed using standard techniques. Fertilization was evaluated 18–19 hours after IVF or ICSI and was considered normal when two distinct pronuclei were evident. The embryos were assessed using a stereoscope (SZX16; Olympus Corporation, Tokyo, Japan). Embryo cryopreservation was performed using a vitrification kit. In addition, FETs were performed using a hormone replacement cycle or natural cycle. The transfer date for blastocysts was 5 days after ovulation or 5 days after the start of luteal support. In HRC-FETs, estradiol administration usually begins on the 3rd day of the menstrual period. When the endometrial thickness was ≥7 mm on the 11th to 19th days of the menstrual period, luteal support was initiated and continued until the 9th week of gestation. For natural cycle FET , a progesterone concentration of ≥1.00 ng/mL was defined as indicating ovulation. From the 3rd day after ovulation until the 4th week of gestation a progesterone vaginal suppository was used. Thawing medium (KITAZATO BioPharma, Tokyo, Japan) was used for the embryo thawing. If an appropriately increasing serum hCG level was observed, a transvaginal ultrasound was performed at 5–6 weeks of gestation to determine the presence of a gestational sac, which indicates clinical pregnancy. On the basis of the questionnaires obtained from all the patients, the delivery procedure, complications, gestational age, infant weight, and sex were investigated. For P values, univariate analysis was performed for each characteristic (i.e., maternal age, paternal age, rate of ICSI, history of pregnancy, history of delivery, first time for embryo transfer, blastocysts of grade A TE according to Gardner criteria ( 20 ), blastocysts of grade A ICM according to Gardner criteria, HRC-FET, cause of infertility, history of endometrial operation, history of cesarean section, history of myomectomy and day 5 blastocyst) and live birth. History of delivery included vaginal birth, cesarean section, term and preterm delivery. If there were significant differences, logistic regression analysis was performed. The backward elimination method was applied to select variables in the predictive logistic models. A χ 2 test was performed to calculate the P values for the clinical pregnancy, miscarriage, and live birth rates from the 7th to 11th day after embryo transfer on day 5 and day 6 embryos. A receiver operating characteristic (ROC) curve analysis was used to compare serum hCG levels, clinical pregnancy, and live birth in early pregnancy after FET. In addition, the serum hCG cutoff values for predicting clinical pregnancy and live birth were identified using Youden index and ROC analysis to identify the points of optimal sensitivity and specificity. Statistical analyses were performed using SPSS version 26.0 software (SPSS, Inc., Chicago, IL). The Kolmogorov–Smirnov test was used to evaluate the normal distribution. The results are expressed as the median, interquartile range (IQR), 95% confidence interval (CI), and P values. Statistical significance was set at P <.05 for all tests.

The characteristics of live births in the FET group are shown in Table 1 . The univariate analysis revealed no significant differences in HRC-FET, history of endometrial operation, and cause of infertility (male factor, tubal problems, endometriosis, and unexplained infertility) between the groups. Meanwhile, maternal age, history of cesarean section, history of myomectomy, day 5 blastocyst transfer, first-time FET, history of pregnancy, blastocysts of grade A TE, and grade A ICM were significantly (all P <.001) higher in live births. On the basis of these results, we performed a logistic regression analysis, adding previous delivery history to these characteristics. Paternal age, history of pregnancy, and ICSI did not differ significantly between the live births. In contrast, maternal age ( P <.001, 95% CI: 0.860–0.875) and history of cesarean section ( P <.001, 95% CI: 0.427–0.518) were significantly lower in live births. Meanwhile, day 5 blastocyst transfer ( P <.001, 95% CI: 1.682–2.017), first-time FET ( P =.001, 95% CI: 1.055–1.221), and history of delivery ( P <.001, 95% CI 1.248–1.450) were significantly higher in live births. Blastocysts of grade A TE ( P <.001, 95% CI: 2.002–2.368) and grade A ICM ( P <.001, 95% CI: 1.228–1.472) were significantly associated with a higher live birth rate. Table 1 Characteristics of live birth in 18,491 FET excluding day 3 embryos and multiple pregnancies Overall Live birth a P values b , c Yes No Univariate analysis Logistic regression analysis Number 18,491 5,788 12,668 Maternal age, year, median (IQR) 38 (32–44) 36 (30–42) 39 (33–45) <.001 <.001 Paternal age, year, median (IQR) 39 (31–47) 37 (30–44) 40 (32–48) <.001 .073 History of pregnancy: Yes (%) 58.4 54.6 60.2 <.001 .414 History of delivery d : Yes (%) 32.8 32.6 32.9 .669 <.001 First time for embryo transfer (%) 36.4 44.9 32.5 <.001 .001 ICSI e (%) 56.6 54.6 57.5 <.001 .610 Blastocyst of grade A inner cell mass f (%) 72.2 83.1 67.3 <.001 <.001 Blastocyst of grade A trophectoderm f (%) 66.4 82.1 59.2 <.001 <.001 HRC–FET g (%) 81.0 80.8 81.1 .563 Causes of infertility  Male factor (%) 35.6 36.2 35.4 .589  Tube (%) 14.3 14.8 14.0 .158  Endometriosis (%) 15.4 15.1 15.5 .301  Unexplained infertility (%) 39.4 39.0 39.6 .492 History of endometrial operation: Yes h (%) 4.8 4.5 5.0 .165 History of cesarean section: Yes (%) 18.1 11.8 21.1 <.001 <.001 History of myomectomy: Yes (%) 5.9 4.4 6.5 <.001 .241 Day 5 blastocyst transfer i (%) 71.6 84.8 6.5 <.001 <.001 Note. FET = frozen-thawed embryo transfer, ICSI = intracytoplasmic sperm injection, IQR = interquartile range, HRC-FET = hormone replacement cycle frozen-thawed embryo transfer. a 35 cases of induced abortion were excluded from the 18,491 cases. b For P values, univariate analysis was performed for each characteristic and live birth, and if there was a significant difference, logistic regression analysis was performed. c Statistical significance was set at P <.05. d Including vaginal birth, cesarean section, term, and preterm delivery. e No. of ICSI/total No. of FET. f Based on Gardner Schoolcraft criteria. g No. of HRC-FET/total No. of FET. h Including endometrial curettage, manual vacuum aspiration, septal incision, submucosal myomectomy and transcervical polypectomy. i No. of day 5 blastocyst transfer/No. for Day 5 and 6 blastocyst transfers. Characteristics of live birth in 18,491 FET excluding day 3 embryos and multiple pregnancies Note. FET = frozen-thawed embryo transfer, ICSI = intracytoplasmic sperm injection, IQR = interquartile range, HRC-FET = hormone replacement cycle frozen-thawed embryo transfer. 35 cases of induced abortion were excluded from the 18,491 cases. For P values, univariate analysis was performed for each characteristic and live birth, and if there was a significant difference, logistic regression analysis was performed. Statistical significance was set at P <.05. Including vaginal birth, cesarean section, term, and preterm delivery. No. of ICSI/total No. of FET. Based on Gardner Schoolcraft criteria. No. of HRC-FET/total No. of FET. Including endometrial curettage, manual vacuum aspiration, septal incision, submucosal myomectomy and transcervical polypectomy. No. of day 5 blastocyst transfer/No. for Day 5 and 6 blastocyst transfers. Table 2 shows pregnancy outcomes and serum hCG cutoff values from 7th day to 11th day after FET for clinical pregnancy and live births in 18,491 cycles, excluding day 3 embryos and multiple pregnancies from among 18,814 FET cycles. There were no significant differences among each day of clinical pregnancy ( P =.628), miscarriage ( P =.181), or live birth rates ( P =.071). Table 2 Pregnancy outcomes and serum hCG cutoff value of clinical pregnancy and live birth in 18,491 cycles from 7th to 11th after FET. Date of investigated serum hCG 7th day a 8th day a 9th day a 10th day a 11th day a P values b Number 194 1,139 6,879 8,691 1,588  Maternal age c , year, median (IQR) 37 (31–43) 37 (30–44) 37 (31–43) 37 (31–43) 38 (32–44)  Clinical pregnancy rate (%) 46.9 43.9 42.5 43.1 43.7 .628  Miscarriage rate (%) 16.4 25.7 26.8 25.4 26.8 .181  Live birth rate d (%) 39.1 32.5 30.5 31.6 31.4 .071 Serum hCG cutoff value of clinical pregnancy (mIU/mL) 21.35 30.0 40.35 58.0 64.8  PPV (%) 93.6 92.40 93.5 95.90 97.6  NPV (%) 97.9 96.10 97.8 97.60 97.8  Sensitivity 0.978 0.952 0.972 0.969 0.972  Specificity 0.942 0.939 0.951 0.970 0.982  Youden index e 0.920 0.891 0.922 0.938 0.955 Serum hCG cutoff value of live birth (mIU/mL) 31.05 46.6 72.05 95.4 159.9  PPV (%) 81.1 75.4 74.9 75.9 77.7  NPV (%) 97.1 97.7 97.9 98.7 98.4  Sensitivity 0.961 0.960 0.960 0.977 0.970  Specificity 0.855 0.848 0.859 0.859 0.872  Youden index e 0.816 0.809 0.819 0.836 0.842 Note. hCG = human chorionic gonadotropin, FET = frozen-thawed embryo transfer, IQR = interquartile range, NPV = negative predictive value, PPV = positive predictive value. a after embryo transfer. b P values calculated among date of investigated serum hCG with the use of χ 2 tests. c Maternal age on oocyte retrieval. d Live birth rate was calculated from the number of live births/18,491 cycles. e Youden index was calculated as sensitivity+specificity-1. Pregnancy outcomes and serum hCG cutoff value of clinical pregnancy and live birth in 18,491 cycles from 7th to 11th after FET. Note. hCG = human chorionic gonadotropin, FET = frozen-thawed embryo transfer, IQR = interquartile range, NPV = negative predictive value, PPV = positive predictive value. after embryo transfer. P values calculated among date of investigated serum hCG with the use of χ 2 tests. Maternal age on oocyte retrieval. Live birth rate was calculated from the number of live births/18,491 cycles. Youden index was calculated as sensitivity+specificity-1. Figure 1 shows the ROC curve of serum hCG concentration by day after embryo transfer for predicting live birth in FET. The ROC analysis of serum hCG level for predicting live birth after FET showed high discrimination on days 7–11 after FET (all P <.001): 7th day area under curve (AUC) 0.962 (95% CI: 0.938–0.985), positive predictive value (PPV) 81.1%, negative predictive value (NPV) 97.1%; 8th day AUC: 0.958 (95% CI: 0.947–0.968), PPV: 75.4%, NPV 97.7%; 9th day AUC: 0.954 (95% CI: 0.949–0.959), PPV: 74.9%, NPV: 97.9%; 10th day AUC: 0.957 (95% CI: 0.954–0.961), PPV: 75.9%, NPV: 98.7%; 11th day AUC: 0.954 (95% CI: 0.944–0.963), PPV: 77.7%, NPV: 98.4%. In addition, serum hCG cutoff values from the 7th to 11th days after embryo transfer were 31.05 mIU/mL (sensitivity 0.961, specificity 0.855, Youden index 0.816) on the 7th day, 46.6 mIU/mL (sensitivity 0.960, specificity 0.848, Youden index 0.809) on the 8th day, 72.05 mIU/mL (sensitivity 0.960, specificity 0.859, Youden index 0.819) on the 9th day, 95.4 mIU/mL (sensitivity 0.977, specificity 0.859, Youden index 0.836) on the 10th day, and 159.9 mIU/mL (sensitivity 0.970, specificity 0.872, Youden index 0.842) on the 11th day. The probability of live birth was 81.1% when serum hCG was ≥31.05 mIU/mL on day 7 after embryo transfer, 75.4% at ≥46.6 mIU/mL on day 8, 74.6% at ≥72.05 mIU/mL on day 9, 75.9% at ≥95.4 mIU/mL on day 10, and 73.9% at 159.9 mIU/mL on day 11. In contrast, the probability was markedly lower when serum hCG values were below these thresholds: 2.88% on day 7, 2.24% on day 8, 2.01% on day 9, 1.24% on day 10, and 1.55% on day 11. Figure 1 ROC curve of serum hCG concentration by day after embryo transfer for predicting live birth in FET. ( A ) 7th day a ; ( B ) 8th day a ; ( C ) 9th day a ; ( D ) 10th day a ; ( E ) 11th day a . FET = frozen-thawed embryo transfer; hCG = human chorionic gonadotropin; ROC = receiver operating characteristic. a Date serum hCG was investigated after blastocyst transfer. ROC curve of serum hCG concentration by day after embryo transfer for predicting live birth in FET. ( A ) 7th day a ; ( B ) 8th day a ; ( C ) 9th day a ; ( D ) 10th day a ; ( E ) 11th day a . FET = frozen-thawed embryo transfer; hCG = human chorionic gonadotropin; ROC = receiver operating characteristic. a Date serum hCG was investigated after blastocyst transfer. Meanwhile, ROC analysis of serum hCG level for predicting clinical pregnancy after FET showed excellent discrimination on days 7–11 after FET (all P <.001): day 7, AUC: 0.990 (95% CI: 0.981–0.999), PPV: 93.6%, NPV: 97.9%; day 8, AUC: 0.989 (95% CI: 0.984–0.993), PPV: 92.4%, NPV: 96.1%; day 9, AUC: 0.993 (95 % CI: 0.992–0.995), PPV: 93.5%, NPV: 97.8%; day 10, AUC: 0.995 (95 % CI: 0.994–0.996), PPV: 95.9%, NPV: 97.6%; and day 11, AUC: 0.998 (95 % CI: 0.997–0.999), PPV: 97.6%, NPV: 97.8%; respectively. In addition, serum hCG cutoff values from the 7th to 11th days after embryo transfer were 21.35 mIU/mL (sensitivity 0.978, specificity 0.942, Youden index 0.920) on the 7th day, 30.0 mIU/mL (sensitivity 0.952, specificity 0.939, Youden index 0.891) on the 8th day, 40.35 mIU/mL (sensitivity 0.972, specificity 0.951, Youden index 0.922) on the 9th day, 58.0 mIU/mL (sensitivity 0.969, specificity 0.970, Youden index 0.938) on the 10th day, and 64.8 mIU/mL (sensitivity 0.972, specificity 0.982, Youden index 0.955) on the 11th day.

This study identified the hCG cutoff values for clinical pregnancy and live births after single-blastocyst FET singleton pregnancies. These findings support the use of early serym hCG testing to reduce medical resource utilization and patient burden. Future studies could investigate these findings in a multicenter cohort or validate these findings in day 3 embryos or multiple pregnancies.

We identified serum hCG cutoff values for clinical pregnancy and live births from the 7th to 11th days after FET. Compared with previous reports ( 2 , 11 , 12 , 13 , 14 ), our findings indicate that live births can be identified earlier. As previously reported, serum hCG concentration increases up to approximately the 10th week of gestation; in our results, the serum hCG cutoff value also increased from days 7 to 11 after FET. If the hCG level is detectable but below the cutoff value for clinical pregnancy, possibilities include chemical abortion or ectopic pregnancy. Hence, careful follow-up with symptom assessments and physical examinations is warranted. As shown in Table 1 , in addition to embryo grade, maternal age, history of delivery, and first embryo transfer, a history of cesarean section also had an impact on the live birth rate, suggesting that prior cesarean delivery may affect endometrial function. Many studies have examined early prediction of pregnancy and delivery outcomes. Some reports predicted clinical pregnancies on the 14th day after oocyte retrieval and live births on the 10th day after embryo transfer. Lambers et al. ( 10 ) assessed 204 patients who underwent IVF or ICSI, showing their hCG cutoff levels were 76, 142, and 223 mIU/mL on days 14, 15, and 16, respectively. Meanwhile, Qiu et al. ( 2 ) reported that the AUC for clinical pregnancy and live birth were 0.986 and 0.922 with the corresponding hCG cutoff values of 113.28 and 146.37 mIU/mL on the 10th day after 772 single-blastocyst transfers ( 2 ). In this report, live births were predicted from days 7–11 after FET (i.e., from the 12th to 16th day after ovulation or progesterone supplementation), which was earlier than in previous reports. In addition, the sensitivity and specificity were comparable even at 7 and 8 days after FET, compared with 9–11 days after FET. We showed the serum hCG cutoff values for live births on the 7th to 11th days after FET and suggested live births can be predicted as early as on the 7th and 8th days after FET. Considering the need to preserve precious medical resources and reduce patient burdens, measuring serum hCG levels during early pregnancy may be necessary. This study has two limitations. First, the sample size on the 7th day after FET was small; therefore, we believe it is necessary to consider a greater number of cases in future studies. Second, this study did not include multiple pregnancies or day 3 embryos. These were excluded because a previous study reported that they affect serum hCG levels in early pregnancy ( 3 , 5 , 6 , 11 , 15 , 16 ). We will investigate the cutoff values for multiple pregnancies and day 3 embryos as soon as we accumulate more cases. Despite these limitations, the sample size in this study was larger than that of previous studies, and it was a single-center study; hence, we suggest that there is little variation in the data because of procedures and culture methods.

N.K. has nothing to disclose. T.I. has nothing to disclose. T.K. has nothing to disclose. N.M. has nothing to disclose.