Delayed implantation might be an early signal for poor reproductive outcomes in assisted reproductive technology.

Delayed implantation is strongly associated with adverse reproductive outcomes, including higher rates of biochemical pregnancy and ectopic pregnancy as well as a reduced ongoing pregnancy rate. Our results provide novel insights into early hCG interpretation, thereby supporting more precise prognosis assessment and personalized counseling in ART.

A total of 1475 patients with single blastocyst transfer with frozen-thawed cycles in the Reproductive and Genetic hospital of CITIC Xiangya from January 2022 to May 2024 were recruited. All patients received standard management before embryo transfer, including surgical removal of intrauterine adhesions, hydrosalpinx, endometrial polyps, and cessation of smoking/alcohol for at least 3 months. Fresh cycles and women with repeated implantation failure (RIF) were excluded. All patients underwent four hCG tests on days 6, 8, 10, and 12 following single blastocyst transfer. A total of 309 patients were negative for all hCG tests. For patients with a positive hCG result, initial detection of hCG ≥ 5 IU/L on day 6 after blastocyst transfer was defined as normal implantation, while initial detection ≥5 IU/L on days 8, 10, or 12 was classified as delayed implantation, according to previous literature [ 3 ]. There were 455 patients enrolled in the normal implantation group and 60 patients enrolled in the delayed implantation group. It should be noted that no patients in this cohort had their first positive hCG (≥5 IU/L) on day 12. This retrospective study was approved by the CITIC Xiangya Reproductive and Genetic Hospital Ethical Committee (No.LL-SC-2025-003). In frozen-thawed cycles, hormone replacement therapy (HRT) was employed for women with irregular or anovulatory cycles. Oral estradiol valerate (6 mg daily) was started on menstrual cycle day 3. After 10 days, endometrial thickness was assessed via transvaginal ultrasound. If necessary, the estradiol dose was increased to a maximum of 8 mg per day. Once the endometrial thickness exceeded 8 mm, luteal phase support was initiated with oral dydrogesterone (10 mg twice daily) and vaginal progesterone (200 mg twice daily). Single blastocyst transfer was performed 5 days after progesterone administration, as previously recommended [ 8 ]. For patients with regular menstrual cycles, a natural protocol was used. When the leading follicle reached 16 mm in diameter, serum luteinizing hormone (LH) levels were monitored daily alongside transvaginal ultrasound. Single blastocyst transfer was performed 5 days after confirmed follicular rupture. Luteal support was initiated following confirmed ovulation with oral dydrogesterone (10 mg twice daily). Prior to embryo transfer in both protocols, endometrial thickness was confirmed by ultrasound. Cycles were canceled if the thickness was less than 8 mm. The Gardner blastocysts grading system was used to assess degree of blastocoel expanding and inner cell mass (ICM) and trophectoderm (TE) morphology [ 9 ]. We used A, B, or C to represent the quality of ICM and TE separately. The combination of developmental stage and ICM and TE quality formed the grading of each blastocyst. We classified blastocysts with an A or B in both ICM and TE as high-quality blastocysts (AA, AB, BA or BB), those with a C in either ICM or TE as low-quality blastocysts (AC, CA, BC, CB). Very-low grade blastocysts referred to CC blastocysts would not be transferred. The serum hCG concentrations were sequentially measured at 8:00–9:00 AM on blastocyst transfer days 6, 8, 10 and 12 from peripheral blood and results were expressed in IU/L. Samples were rested for 15 min and centrifuged at 3000 rpm for 10 min. The serum on the upper layer was used for analysis. The measurements were performed using electrochemical luminescence (ECLIA) on a Cobas 8000 (Roche Diagnostics, Germany). Daily internal quality control and yearly external quality control were performed on request. The intra-assay coefficient of variation (CV) of serum hCG was 2.3%, and the interassay CV value was 3.7%. All the data obtained from our laboratory in IVF center. The reproductive outcomes included biochemical pregnancy, early miscarriage and ectopic pregnancy as well as ongoing pregnancy. For the biochemical pregnancies, serum hCG concentration temporarily increased and then gradually decreased, and there was no gestational sac detected either intrauterine or extrauterines under ultrasound examination. For early miscarriages, serum hCG concentration increased, there was a gestational sac inside uterus, but there was yet no fetal heart beat or heart beat stopped within 12 weeks of pregnancy. For the ectopic pregnancies, serum hCG concentration increased and there was a gestational sac outside uterus. For ongoing pregnancy, pregnancy persisted more than 12 weeks and growth of intrauterine fetus was normal. All statistical analyses were conducted using the IBM SPSS Statistics Version 25.0 (IBM Corp., USA). Means and standard deviations were calculated for continuous variables, and Student’s t test was used for comparisons. We used the chi-square test to compare the frequencies and proportions, as appropriate. Multivariate logistic regression analysis was conducted to analyze the factors associated with delayed implantation after adjusting for confounding variables. The following variables were used in the multivariate analysis: female age, body mass index (BMI), endometrium (Em) thickness, natural/HRT cycles and preimplantation genetic testing analysis (PGT-A) cycles as well as the quality of embryos. All statistical tests used a two-tailed α of 0.05.

Of 515 eligible patients, 455 were in the normal implantation group, while 60 were in the delayed implantation group. Among the delayed implantation group, 56 patients were first found hCG ≥ 5 IU/L on embryo transfer day 8, 4 patients were first found hCG ≥ 5 IU/L on embryo transfer day 10. No patient was first found hCG ≥ 5 IU/L on embryo transfer day 12 (see Fig.  1 ). In all pregnancies, the normal implantation rate was 88% (455/515), while delayed implantation occurred in 12% of cases. Specifically, delayed implantation was observed in 11% (56/515) on day 8, 1% (4/515) on day 10, and 0% on day 12 after blastocyst transfer (see Fig.  2 ). Fig. 1 Flow chart of case selection. RIF = repeated implantation failure; hCG = human chorionic gonadotropin; ET = embryo transfer Fig. 2 Distribution of implantation timing for all pregnancies by early sequential hCG tests. Day 6 normal implantation; D8–12 delayed implantation Flow chart of case selection. RIF = repeated implantation failure; hCG = human chorionic gonadotropin; ET = embryo transfer Distribution of implantation timing for all pregnancies by early sequential hCG tests. Day 6 normal implantation; D8–12 delayed implantation Table 1 presents the basic characteristics of the patients. Female/male age, AMH, BMI, basal FSH, LH, primary/secondary infertility, infertility cause and endometrium thickness as well as the ratio of natural/HRT cycles did not show significant differences between the normal and delayed implantation group. However, the delayed implantation group had less PGT-A cycles (18.3% vs. 51.9%, P  = 0.000). In addition, the incidence of high quality embryo was significantly lower in the delayed implantation group (48.3% vs. 75.2%, P  = 0.000). Table 1 Comparison of patients between delayed and normal implantation groups Delayed implantation Normal implantation P value n  = 60 n  = 455 Age (years) Female 33.7 ± 4.1 33.6 ± 4.5 0.851 Male 35.5 ± 5.8 36.0 ± 5.8 0.637 BMI (kg/m 2 ) 22.0 ± 2.3 21.9 ± 2.6 0.919 FSH(U/L) 6.4 ± 2.7 5.9 ± 2.4 0.147 LH (U/L) 4.7 ± 6.1 4.9 ± 4.9 0.731 Fertility (%) Primary 17 (28.3) 153 (33.6) 0.467 Secondary 43 (71.7) 302 (66.4) AMH (ng/ml) 3.5 ± 2.8 4.3 ± 3.5 0.106 Cause of infertility (%) Male factor 4 (6.7) 38 (8.4) 0.805 Tubal factor 38 (63.3) 265 (58.2) 0.488 Ovulation disorder 10 (16.7) 109 (24.0) 0.255 Other 8 (13.3) 43 (9.5) 0.357 Frozen-thawed cycles (%) Natural 23 (38.3) 174 (37.7) 0.989 HRT 37 (61.7) 281 (62.3) PGT-A cycles (%) 11 (18.3) 236 (51.9) 0.000 Embryo quality (%) Low 31 (51.7) 113 (24.8) 0.000 High 29 (48.3) 342 (75.2) BMI = body mass index; AMH = anti-Müllerian hormone; Em = endometrium; ET = embryo transfer; HRT = hormone replacement therapy; PGT-A = preimplantation genetic testing for aneuploidy Comparison of patients between delayed and normal implantation groups BMI = body mass index; AMH = anti-Müllerian hormone; Em = endometrium; ET = embryo transfer; HRT = hormone replacement therapy; PGT-A = preimplantation genetic testing for aneuploidy In Table  2 , the delayed implantation group demonstrated significantly lower hCG level compared with normal implantation group on day 8 (17.7 ± 16.0 vs. 77.4 ± 43.9, P  = 0.000), day 10 (40.7 ± 31.5 vs. 182.8 ± 117.1, P  = 0.000) and day 12 (87.0 ± 113.5 vs. 495.1 ± 346.3, P  = 0.000). After calculating, the hCG rising ratio between day 8 and day 10 was comparable between the two groups (2.4 ± 1.6 vs. 2.3 ± 0.8, P  = 0.322). However, the hCG rising ratio was significantly decreased when day12: day 10 (1.9 ± 2.4 vs. 2.5 ± 1.2, P  = 0.002) in the delayed implantation group. Table 2 Comparison of dynamics of hCG rising between delayed and normal implantation groups Delayed implantation Normal implantation P value n  = 60 n  = 455 HCG levels Day 6 N/A 25.1 ± 13.0 N/A Day 8 17.7 ± 16.0 77.4 ± 43.9 0.000 Day 10 40.7 ± 31.5 182.8 ± 117.1 0.000 Day 12 87.0 ± 113.5 495.1 ± 346.3 0.000 HCG rising ratio Day 8:day 6 N/A 3.2 ± 1.4 N/A Day 10:day 8 2.4 ± 1.6 2.3 ± 0.8 0.322 Day 12:day 10 1.9 ± 2.4 2.5 ± 1.2 0.002 N/A due to no data in delayed implantation group on day 6 Comparison of dynamics of hCG rising between delayed and normal implantation groups N/A due to no data in delayed implantation group on day 6 Table 3 showed potential factors associated delayed implantation. After adjusting for confounders, including female age, BMI, endometrium thickness, natural/HRT cycles and preimplantation genetic testing analysis (PGT-A) cycles as well as the quality of embryos, higher incidence of good quality embryos (adjusted OR, 0.293; 95% CI 0.165–0.520, P  = 0.000) and PGT-A cycles (adjusted OR, 0.183; 95% CI 0.008–0.381, P  = 0.000) might be protective factors avoiding delayed implantation. However, the natural/HRT cycles and age above or lower than 35 years were not associated with delayed implantation. Table 3 Factors associated with delayed implantation after multivariable logistic regression analysis OR for delayed implantation Crude OR Adjusted OR [95% CI] P [95% CI] P Non-PGT-A 1 1 PGT-A 0.208 [0.106; 0.411] 0.000 0.183 [0.088; 0.381] 0.000 Low quality embryo 1 1 High quality embryo 0.267 [0.154; 0.464] 0.000 0.293 [0.165; 0.520] 0.000 HRT cycle 1 1 Natural cycle 0.926 [0.530; 1.618] 0.788 0.835 [0.459; 1.519] 0.554 <35 years 1 1 ≥35 years 0.800 [0.456; 1.405] 0.438 0.445 [0.163; 1.216] 0.114 PGT-A = preimplantation genetic testing for aneuploidy, HRT = hormone replacement therapy Factors associated with delayed implantation after multivariable logistic regression analysis PGT-A = preimplantation genetic testing for aneuploidy, HRT = hormone replacement therapy As shown in Table  4 , the biochemical pregnancy rate was significantly higher in the delayed implantation group compared with normal implantation group (61.7% vs. 15.4%, P  = 0.000). The early miscarriage rate was comparable between groups (5.0% vs. 10.1%, P  = 0.249). The ectopic pregnancy rate was also significantly higher in delayed implantation group (3.3% vs. 0.4%, P  = 0.016). Moreover, the ongoing pregnancy rate was significantly lower in the delayed implantation group in contrast to the normal implantation group (30.0% vs. 74.1%, P  = 0.000). Table 4 Comparison of reproductive outcomes between delayed and normal implantation groups Delayed implantation Normal implantation P value n  = 60 n  = 455 Biochemical pregnancy (%) 37 (61.7) 70 (15.4) 0.000 Early miscarriage (%) 3 (5.0) 46 (10.1) 0.249 Ectopic pregnancy (%) 2 (3.3) 2 (0.4) 0.016 Ongoing pregnancy (%) 18 (30.0) 337 (74.1) 0.000 Comparison of reproductive outcomes between delayed and normal implantation groups

Human chorionic gonadotropin (hCG) is a glycoprotein hormone secreted by both the cytotrophoblast and syncytiotrophoblast, which could be detected as early as 6–7 days after fertilization [ 1 ]. It is widely used to predict the pregnancy outcome in natural conception and assisted reproductive technology. In ART, anxious couples often hope to know the pregnancy outcomes as early as possible. It has been reported that a surum positive hCG ≥ 5 IU/L indicated the presence of implantation [ 2 ]. A study from Yadid et al. found that a single early test of hCG ≥ 4.0 IU/L at day 5 after blastocyst transfer could be used for the diagnosis of biochemical pregnancy [ 3 ]. Furthermore, previous study reported that hCG ≥ 5.0 IU/L and ≥16 IU/L could predict ongoing pregnancy/multiple pregnancy 5 days after two fresh blastocysts transfer [ 4 ]. In addition, Yuan and colleagues showed that hCG > 5.34 IU/L might predict single ongoing pregnancy and hCG > 17.95 IU/L predict multiple pregnancies on the seventh day after two cleavage embryo transfer [ 5 ]. In addition, a previous retrospective cohort study on hCG rising kinetics demonstrated that the average rate of rise of hCG level resulted in live birth was 50% in 1 day and 124% in 2 days [ 6 ]. However, among patients receiving two or more embryos, only 6.1% had abnormal hCG rises [ 7 ], implying asynchronous implantation events during embryo development. Nevertheless, there is few studies performed on the influence of implantation timing of embryo and subsequent reproductive outcomes in ART. In this study, we aimed to investigate pregnancy outcomes of delayed and normal implantation by early sequential hCG tests.

To our knowledge, this is the first study to investigate the effect of implantation timing by early sequential measurement of hCG levels in ART. We found that the hCG level and dynamics were different in the women with delayed implantation compared with normal implantation, which was associated with higher incidence of biochemical pregnancy and ectopic pregnancy as well as decreased ongoing pregnancy rate. The delayed implantation might be a strong signal associated with poor prognosis of pregnancy. Implantation is a process, whereby the embryo attaches itself to the luminal surface of the endometrium followed by migration via the luminal epithelium and invasion into the deep layer of the endometrium. It is well-known to us successful implantation requires a receptive endometrium, a functional embryo and a synchronized dialogue between them [ 10 , 11 ]. Since both the embryo and endometrial quality can vary, the outcome of implantation will be determined by their synchronism [ 12 ]. The delayed implantation events might be compromised results of embryo and uterine dialogue. Based on previous studies, estrogen and progesterone are the most important hormones necessary to prepare the endometrium for implantation [ 13 ]. In humans, the optimal timing for embryo implantation is called “window of implantation” (days 20–24) during the secretory phase in a normal menstrual cycle (~28 days) [ 14 – 16 ]. Accordingly, this window in ART ranges from 6 to 10 days after follicular rupture and 5–9 days after the administration of progesterone in artificial cycles [ 17 – 19 ]. As a result, the blastocysts were generally suggested transferred on day 5 post follicular rupture in natural cycles or 5 days after progesterone administration in HRT cycles [ 8 ]. Besides proper transfer timing, intrinsic uterine factors such as congenital uterine anomalies, submucous fibroids, endometrial polyps, endometriosis, intrauterine adhesions and thickness of endometrium might disrupt the uterine receptivity [ 20 , 21 ]. In addition, uncontrolled frequent uterine contractions might also contribute to the delayed implantation [ 22 ]. For the embryo factors, aneuploid embryo transfer is inevitable in non-PGT-A cycles. The embryos that blastulate slowly also had lower sustained implantation rates, indicating the asynchrony caused by embryogenic factors [ 23 ]. Moreover, the transfer of single low-grade blastocysts has been reported resulting in a reduced live birth rate [ 24 ]. In the present study, the incidence of delayed implantation remained 12% in all pregnancies. Even many known adverse factors have been removed before embryo transfer, there were still problems to be solved. It could be supposed that the one-size-fits-all luteal support strategy could not satisfy all infertile women in ART. It has been reported temporal delays of endometrium receptivity could be observed in both natural an HRT cycles [ 25 ]. Previous research demonstrated that the progesterone level <9.2 ng/ml or <10.6 ng/ml in luteal phase would negatively affect reproductive outcomes, indicating a minimal requirement of progesterone for establishing the endometrial receptivity [ 26 , 27 ]. In addition, there is a requirement of proper estrogen threshold ranged from 292 to 409 pg/ml in luteal phase [ 28 ]. In addition, embryo factors such as non-PGT-A transfers and low quality embryos has been shown to be associated with delayed implantation in this study, which might contribute to the late detection of hCG. Another possible hypothesis cannot be ruled out is that embryo may normally implant, while secretion of hCG was impaired, which could not be detected as the control. The antiphospholipid antibodies has been shown to down-regulate secretion of hCG from protein to RNA levels [ 29 ]. In addition, it has been reported that the reduction of IGF-1 could down-regulate the expression of PI3K/AKT signaling pathway, increasing the apoptosis of trophoblast cells and leading to decreased secretion of hCG [ 30 ]. The primary limitations of this study are its retrospective design and the limited cohort size from a single IVF center. To exclude interference of asynchronous development of multi embryos, single blastocyst transfer cycles were recruited. Fresh cycles were also excluded to avoid possible exogenous interference by hCG trigger. In addition, this study did not identify any cases with first positive hCG detected on day 12. This finding cannot rule out the possibility of even later implantation, which is likely due to our limited sample size. Hence, the findings in this study might limit its significance to other clinical scenarios. In conclusion, the delayed implantation is associated with poor prognosis of pregnancies in ART, regardless of causes. Late detection of hCG and abnormal hCG rising ratio within 48 h might be a sign of impaired embryo development or ectopic pregnancy. These findings might provide further insights into the extremely early embryo development and evaluation of hCG variations in ART.