The predictive value of ultrasound markers for pregnancy outcome in recurrent miscarriage: a retrospective study

Recurrent miscarriage is a distressing disorder that prompts couples to proactively schedule regular follow-up visits and examinations in subsequent pregnancies, including ultrasound scans and measurements of hCG, P and E2 levels in the first trimester. In the present study, we propose a prediction model using commonly used markers from the fifth to the ninth week for first trimester RPL patients.

In the present study, we found that age and P levels at week 5 were predictive factors for first trimester pregnancy loss. In contrast to previous studies¹¹our study found a linear relationship between age and pregnancy loss in the first trimester. We support the assumption that age is an independent predictor of pregnancy loss¹. In addition, serum progesterone level is also a predictive marker for RPL patients in the fifth week. It is worth noting that all of these patients were administered various forms of progesterone once they became pregnant¹²Progesterone is released in bursts under the influence of luteinizing hormone, but its release from the corpus luteum is determined by the increase in human chorionic gonadotropin levels after implantation.¹³. The role of serum progesterone levels has been a topic of discussion. Coomarasamy et al¹⁴ found in a large cohort study that progesterone therapy did not significantly increase the rate of live births compared with placebo. However, current recommendations support the use of progesterone in women with a history of miscarriage and bleeding during early pregnancy¹⁵. In line with the previous study¹⁶our study also supports its usefulness as a predictive marker in early stages of pregnancy in women with recurrent miscarriages. In particular, in our study, cases of RPL in subsequent pregnancies with serum progesterone levels below 23.9 ng/ml are at higher risk of another miscarriage in the first trimester. However, in other studies, progesterone thresholds were lower^17,18This may be due to the recurrent miscarriages among the participants in our study, which prompted us to administer individual progesterone to our RPL patients. In summary, our results suggest that higher progesterone levels in the fifth week of pregnancy positively influence pregnancy outcome.¹⁹Human chorionic gonadotropin, a glycoprotein derived from the placenta, plays a crucial role in the course of pregnancy²⁰. Lower hCG concentration at 7 weeks of gestation was independently associated with first trimester pregnancy loss. hCG facilitates the embryonic implantation process, promotes blood vessel growth, regulates trophoblast cell development, and plays an important role in regulating immune responses at the maternal-embryonic and fetal interfaces throughout pregnancy.²¹Liu et al. reported that a threshold of 88,000 IU/L at peak hCG can serve as a predictive indicator of early pregnancy outcome in women who have suffered recurrent miscarriages²²In this study, a cut-off value of 69,636.6 mIU/mL is the inflection point for a higher risk of pregnancy loss at 7 weeks of gestation. Therefore, we support the view that hCG plays an essential role in ensuring a good pregnancy outcome.

The yolk sac plays a crucial role in maternal-fetal exchange before placental circulation is established. In the sixth and seventh weeks of pregnancy, YSD was lower in the group that had miscarried in the first trimester. However, after performing a regression analysis, no significant differences were found, which is inconsistent with previous studies.²³. This discrepancy may be attributed to the inclusion of different variables in our regression model. Our model accounted for important demographic factors and commonly used serum hormones (hCG, P and E2). It is worth noting that higher YSD has been identified as an adverse marker in other studies^24.25our study found a lower YSD in the pregnancy loss group, which is consistent with some previous studies⁷Our results indicate a positive correlation between YSD and gestational week, suggesting that YSD increases with pregnancy progression, as reported in previous studies²⁶The gestational sac, a fluid-filled structure that surrounds the embryo in the early stages of embryonic development, can be visualized by ultrasound as early as 4.5 to 5 weeks of pregnancy.²⁷. Its size increases with gestational age. mGSD serves as an early indicator of pregnancy loss²³which is consistent with our research. Based on our RCS analysis, we found that an mGSD < 18.3 mm at the sixth week of pregnancy indicates a higher risk of pregnancy loss in the first trimester. In addition, the RCS analysis performed in our study showed that by the ninth week, an mGSD < 33,3 mGSD > 48.3 mm is associated with a higher risk of first trimester pregnancy loss; however, given the small sample size at ninth week, this result should be interpreted with caution. In addition, CRL was identified as a reliable predictor of pregnancy loss in the present study. Specifically, a small CRL value (less than 2.4 mm, 9.9 mm, 16.9 mm, and 18.6 mm) at sixth, seventh, eighth, and ninth weeks was found to be independently associated with first trimester pregnancy loss. As reported in a previous study²⁸they determined different cutoff values ​​of 6.0, 8.5, and 10.9 mm for the 6th, 7th, and 8th weeks of gestation, respectively. The differences might be due to the different inclusion of participants and the different methods of determining the inflection point. In the present study, we included RPL women and used a restricted cubic spline to determine the inflection point. At the same time, the cutoff values ​​were determined by adjusting for age, BMI, and previous pregnancy losses, which improved the accuracy. In addition, our results demonstrated the strong predictive power of the CRL at the eighth week with an AUC value of 0.871. Low CRL in the first trimester has been reported to be significantly associated with an increased risk of chromosomal abnormalities²⁹Another study^29,30 reported that short CRL was not associated with an increased risk of fetal chromosomal abnormalities. In this study, the majority of patients experienced pregnancy loss before ten weeks, and further investigation is needed to determine the specific underlying causes.

The etiology of RPL is complex and multifactorial, highlighting the need for a robust prediction model for pregnancy outcomes. According to the findings of Cecilia et al.⁶the final scoring system, which takes into account maternal age, bleeding score, mGSD, mYSD and the presence of fetal heartbeat, achieved an AUC of 0.901. In contrast, using only demographic variables (maternal age and bleeding severity) yielded an AUC of 0.724 to predict progressing pregnancy. Similarly, using only ultrasound variables (mGSD, mYSD and presence of fetal heartbeat) yielded an AUC of 0.873. In our model, logistic regression was used to include more important factors for each week of pregnancy, resulting in a prediction model with only one to three variables. Despite the complexity of the RPL investigated in this study, the discriminatory power shows good performance. In addition, in comparison with another study¹⁷ For serum markers to predict miscarriage, dual markers (estradiol and progesterone; AUC = 0.871) or three markers (hCG, estradiol and progesterone; AUC = 0.869) showed similar performance at the seventh and eighth weeks of gestation in our results. However, since comparisons across each week of gestation are not possible, this may not be suitable for RPL women who require accurate assessment at each stage. In our study, the model performed better at the seventh (AUC = 0.872, 95% CI: 0.814–0.930) and eighth (AUC = 0.871, 95% CI: 0.789–0.953) week of pregnancy than at the fifth (AUC = 0.671, 95% CI: 0.601–0.740), sixth (AU = 0.796, 95% CI: 0.734–0.857) and ninth (AUC = 0.813, 95% CI: 0.679–0.947) week. This is inconsistent with a previous study⁹who reported a better model at the sixth week, and they suggested that this might be related to the morphological changes of the gestational sac during embryo development. In the present study, the variability of different performance at different weeks of gestation might be because most pregnancy losses occur at the seventh and eighth weeks of gestation.

Women with RPL are a special population that deserves more attention. RPL causes considerable confusion and psychological distress for affected families, especially for the women affected.¹. A simplified and practical prediction model could alleviate their stress and anxiety in subsequent pregnancies and ultimately improve pregnancy outcomes. Maternal stress during pregnancy may be associated with an increased risk of various adverse birth outcomes¹. Given the frequent visits of RPL patients, our study provided a specific predictive marker for different weeks of pregnancy, which may be more practical and simple in clinical settings. In addition, we used RCS to identify the inflection point for higher risk of pregnancy loss. However, it is important to acknowledge that our study has limitations because it was retrospective and conducted in a single center. The results are limited by potential selection and confounding factors; therefore, further validation through multicenter studies is needed. In addition, the inclusion of RPL women in this study is not limited to participants without explanation. However, we performed a sensitive analysis to further mitigate the potential selection bias, and the results suggest similar predictors and similar AUC at each week of pregnancy. Therefore, these models could be robust and generalizable to most RPL women. Third, despite administering appropriate individualized treatment for patients with recurrent pregnancy loss in our hospital, there may be underlying factors that affect reproductive outcomes. Preventive measures for subsequent pregnancy loss in the RPL prediction model should also be considered in the future. In addition, there may be inherent differences among the different ultrasound operators in our study despite their shared experience with ultrasound, but these may reflect the actual clinical setting. Therefore, future research should consider a well-designed prospective study with a consistent and experienced ultrasound operator. Finally, the women in the present study were Chinese, which limits the generalizability of the current results to other populations.