Till date, a single hCG measurement has not been used for the most likely prediction of viability and determining the pregnancy location. However, if single hCG values are available, they can provide clinicians with valuable information about the potential status and trends of pregnancy. Herein, we measured the serum hCG levels 14 days after FRET and FET to determine the differences between these two groups and their impact on pregnancy outcomes.

This study showed that hCG measurements were significantly higher in FET than in FRET in all patients, those with viable and clinical pregnancies, and those who delivered live neonates. Although several studies have evaluated the differences between FRET and FET, they included transfers of > 1 embryo [7,8,9,10]. The resultant multiple pregnancies and vanishing twins are associated with higher hCG values; therefore, the results from these studies should be interpreted cautiously. One study investigated the differences in hCG measurements between sFRET and sFET [11]. However, the hCG levels were measured on the 11th day after embryo transfer in this study. Although higher hCG levels were reported in FET than in FRET in the clinical pregnancies, there was no overall difference [11]. One explanation for this could be that differences in hCG levels on day 14 may be better identified owing to the naturally higher hCG values compared to than on day 11. Another study showed higher median hCG levels in sFET than in sFRET, where singleton live births occurred [12]; hCG was measured on day 19 of the luteal phase (equivalent to 14 days after IVF/ICSI and transfer of a day 5 embryo) but included both, cleavage stage with transfer on day 3 as well as blastocyst stage (day 5 and 6 blastocysts) transferred on day 5 after IVF/ICSI. In addition, blastocysts were transferred significantly more often in the FET group than in the FRET group in this study, and recombinant hCG was administered serially for up to 4 days before blood levels were sampled for hCG levels. A study by Zhu et al. also showed higher hCG levels in sFET than in sFRET on day 9 and 16 in viable pregnancies with gestational age > 13 weeks. In faster-growing blastocysts, higher hCG concentrations were observed on day 9 in sFRET, whereas no difference was observed on day 16 in sFRET and on days 9 and 16 in sFET [13].

During the window of implantation, E2 plays a major role in the process of embryo nidation into the receptive endometrium. Uterine glands are stimulated by nidatory E2 to secrete leukaemia inhibitory factor (LIF) which enables embryo adherence to the luminal epithelium. Uterine stromal cells differentiate into decidual cells and synthesise E2 which makes the uterus refractory to conceive and closes the window of implantation. Expression levels of the fibroblast growth factor subtypes during cell proliferation in the endometrial luteal phase differ with high- and low-dose E2 therapy [14]. E2 levels are reportedly higher in FRET than in FET due to ovarian hyperstimulation. These supraphysiologic E2 levels and premature progesterone elevation may affect the implantation window and embryo-endometrial synchronisation by slowing down the implantation process [12, 13]. These differences may be the cause for the different hCG levels in early pregnancy after MAR, higher hCG values in FET, and higher birth weights in FET [15]. Finally, high responders after IVF/ICSI with high E2 levels have lower cumulative live birth rates and profit from a “freeze all” strategy [16].

To date, no existing data on the optimal time point for serum hCG measurements after IVF are available. Currently, initial hCG levels after ET have been measured over a wide range: 9 days [13, 17], 10 days [18], 11 days [11], 12 days [8], 13 days [19, 20], 14 days [21], 15 days [22], 16 days [13], and 17 days [7]. After 9 days, pregnancies with “lower than excepted” hCG values after FRET and with slow-growing characteristics do not always have a poorer prognosis and can have ‘normal’ hCG values one week later [13]. Furthermore, hCG and recombinant hCG alfa are commonly used to induce ovulation in both FRET and FET (MNC); serum hCG levels increase immediately after the injection. Due to the half-life value of hCG (30–37 h) and depending on the initial administered dosage, serum hCG levels can be elevated up to 14 days after exogenous hCG administration [23]. The authors of this study highlighted, that too early performed quantitative pregnancy tests may be associated with false- positive results and should be interpreted cautiously.

Additionally, there are also no existing recommendations regarding the number of hCG blood samples to be obtained after IVF. Some clinicians measure hCG levels once, while others serially measure hCG levels at intervals ranging from 48 h [18, 22] to 168 h [13, 17]. It should be kept in mind that, in principle, the fewer hCG measurements are done the higher would be the risk to miss early miscarriages and in particular ectopic pregnancies. For this reason, working with single hCG values always requires ultrasound examinations and careful inquiry of clinical symptoms. The combined approach of hCG value and ultrasound is generally accepted [11, 12] and standard care in our tertiary referral centre.

In this study, we determined the threshold hCG values in viable and clinical pregnancies; higher hCG thresholds were observed in viable pregnancies. This could be explained by the fact that early miscarriages were classified as clinical pregnancies; miscarriages have naturally lower hCG values due to their disturbed vitality. This inclusion of early miscarriages as clinical pregnancies also explains the higher specificities in clinical pregnancies and the differences between the two ROC curves with a greater distance from the midline seen in clinical pregnancies (Fig. 2). The calculated threshold hCG values can be used to optimize care in patients undergoing IVF. Therefore, higher hCG values in FET should be considered.

According to the divergent findings in hCG levels analyzed on day 11 after blastocyst transfer [11] compared to day 14 in this study, we hypothesized that day 11 may be too early to determine a difference in hCG levels and the fact that too early taken blood samples may be influenced by exogenous hCG [23], together with the findings of this study with high sensitivities (90.1% in FRET and 90.5% in FET) determining a viable pregnancy, it is indicated that a single day 14 serum hCG value indeed might be sufficient for monitoring pregnancies resulting from MAR.

Multivariable regression analysis showed a potential correlation between “hCG levels” and “mode of transfer” when FET was performed. This was expected because of the higher hCG values observed in the different subgroups and in all patients. Additionally, our findings show that AMH potentially affected the hCG levels. Qui et al. showed that older adults undergoing IVF/ICSI had a negative correlation with serum hCG levels when FET was performed [18]. They suggest determining different cut off values for different age groups due to this. Younger patients reportedly have higher natural AMH levels; thus, the link between AMH as an effector on hCG levels was expected.

The restricted “number of transferred embryos” at our IVF clinic and the resultant high number of sETs performed over the last decade are strengths of this study. Furthermore, the fact that blastocysts of comparable quality (BB or better) were transferred in FRET and FET minimized the potential influence of embryo quality on hCG value and implantation.

The limitations of this study are its retrospective design and the high number of patients who were excluded because the blood samples were not obtained on day 14 (n = 1301). Blood sample collection was normally scheduled 14 days after IVF/ICSI. However, if day 14 fell on a Saturday or Sunday, the blood samples were collected on the next working day, which was 15 or 16 days after IVF/ICSI. There was also a certain overlap of hCG values amongst the different subgroups (e.g. viable and non-viable pregnancies) which highlights the importance of performing time-shifted transvaginal ultrasound in all patients when using single hCG levels for pregnancy monitoring. Certainly, exclusive reliance on single hCG values bears the risk to miss extrauterine pregnancies. In case of clinical symptoms (abdominal pain., bleeding), immediate gynecological examination for the early diagnosis of ectopic pregnancy is indispensable.

In conclusion, significantly higher serum hCG values were found in the sFET group than in the sFRET group on day 14 after ET. Higher hCG values do not imply higher pregnancy or live birth rates; however, they may be linked to higher birth weights in infants. It is suggested that a single serum hCG value (in combination with obligate subsequent ultrasound examination to exclude ectopic pregnancies) could be used for monitoring pregnancies resulting from IVF/ICSI.