Discussion

The occurrence of clinical recognized pregnancy loss accounts for approximately 15%–25% of all pregnancies[2]. A considerable number of women suffer from recurrent pregnancy loss. In early pregnancy inspection, especially for women who present with threatened miscarriage, obstetricians and gynaecologists are committed to identifying serum biological markers in order to predict pregnancy outcomes. Several studies[3-6] have focused on various biomarkers such as serum β-HCG, progesterone, estradiol, PAPP-A, inhibin, CA125 and combination of serum biomarkers and ultrasound features to predict pregnancy viability. Ideal biological markers are highly predictive, detected conveniently and cost-effective. Nowadays, the most widely used indicators are serum β-HCG and progesterone. Progesterone is necessary in maintaining early pregnancy, playing an important role in sustaining decidualization, reducing uterine excitability, inhibiting uterine contraction, suppressing inflammatory response and promoting maternal immune tolerance to the fetal semi-allograft[7-9]. However, due to large individual differences and insufficient evidence regarding appropriate progesterone cut-off levels for risk stratification of spontaneous miscarriage, its clinical value in miscarriage risk assessment remains controversial.
HCG is a specific marker of pregnancy, synthesized in syncytiotrophoblast cells and can be detected as early as 8–11 days following ovulation[10]. Serum HCG level reflects quantity of trophoblast cells, rapidly increasing at the early stage of gestation, slowing down later, and reaching the peak at 8-10 weeks of pregnancy. Previous studies[3,11] have proven quantitative determinations of HCG as a valuable tool in the clinical assessment of early pregnancy outcome. A systematic review summarized 8 studies with a total of 584 women that investigated either intact HCG or β-HCG to predict the outcome in women with threatened miscarriage, showing a sensitivity of 44% and a specificity of 86% through further analysis using HSROC[6]. Liu et al. reported that the optimal cut-off value of peak β-hCG was 88468IU/L, with a sensitivity, specificity, positive predictive value, and negative predictive value for successful pregnancy of 95.6%, 88.0%, 95.6%, and 89.0%, respectively[12]. Estrogen is another highly important hormone in establishing and maintaining pregnancy. Studies investigated that estrogen stimulated VEGF production and blood vessel formation to enable maternal-fetal circulation[13], and showed functions in immune regulation[14]. It modulates the immune response by inducing peripheral T cells to secrete the proinflammatory cytokines[14]. Xu et al. compared sex hormone and sex hormone metabolite levels of women in early pregnancy with and without threatened miscarriage, and reported that serum E2 levels were lower in women with threatened miscarriage[15]. They proposed that abnormal levels of sex hormone metabolites and reduction of estrogen activity might result in bleeding during the first trimester of pregnancy. Compared to progesterone and HCG, estrogen is not as widely used in pregnancy assessment. Previous study has found that the sensitivities of estradiol(80%) and HCG(85%) in predicting pregnancy outcome at week 8 of gestation were better than that of serum progesterone (56%)[16]. Pillai’s review found similar results, but the sensitivity was heterogeneous among previous reported studies[6].
As the levels of serum β-HCG and estradiol change over each week of gestation, most of the previous studies did not take this into consideration. Few reports have examined changes of E2 in early pregnancy. The present study focused on dynamic variation of E2 and β-HCG in early pregnancy women, examining their levels and growth rates to help us predict the miscarriage risk accurately and timely. The results presented that β-HCG and E2 levels of the ongoing pregnancy group were significantly higher and increased more rapidly than those of the inevitable abortion group. β-HCG in normal pregnancy increased drastically within the beginning period of gestation, and from the 5th week to 8th week, medians of the growth rate were 5.82, 1.96 and 0.75 per week. Low growth rate may predict bad pregnancy outcome. Quantitative value and growth rate are both meaningful in early pregnancy assessment. When β-HCG combined with E2, the prediction of pregnancy outcome will be more accurate. Miscarriage women showed lower E2 levels. E2 in normal pregnancy women increased 40%-60% weekly on average. Slow growth rate and decreased E2 level are likely to predict bad outcome. Results of our study agreed with most other studies, moreover we provided more information about the variations and growth rates of women with different early pregnancy outcomes.
The major limitation of our study is the relatively small sample size, especially in the inevitable abortion group. It is a little difficult to get large amount of data of miscarriage women due to a relatively small number of miscarriage patients and lack of hormone values after their abortions. So in the future study we will make more efforts to collect larger samples to strengthen the results and hope to find more highly predictive and cost-effective biological markers in predicting pregnancy outcome.
In conclusion, dynamic monitoring of maternal serum E2 and β-HCG levels has important clinical implication for early pregnancy assessment and pregnancy outcome prediction. Low values and low growth rates of E2 and β-HCG probably indicate bad pregnancy outcomes.