Prognostic model on pregnancy outcomes for women with recurrent spontaneous abortions treated with cyclosporin A: A single-institution experience

Highlights • The prognostic model for pregnancy outcomes was constructed.• The prognostic factors included age, ANA, and anti-β2 glycoprotein I antibody levels.• The prognostic models showed higher value for predicting pregnancy success.


Introduction
Recurrent Spontaneous Abortion (RSA) is defined as two or more spontaneous pregnancy losses before 20 weeks of gestation with the same sexual partner and is regarded as a common obstetric complication in the early pregnancy stage. 1 The morbidity of RSA ranges from 2−4 %, with 80 % of cases taking place in the first trimester. 2The pathogenesis of RSA occurrence in the first trimester is more complicated than that in the second trimester, and the etiologic factors remain unidentified in nearly 50 % of RSA cases occurring in the first trimester. 3Immunerelated RSA can be divided into alloimmune and autoimmune types based on maternal immunological factors. 4The alloimmune type is defined as Unexplained RSA (URSA) and is related to an imbalance of maternal-fetal immune tolerance, which accounts for a nearly 50 % incidence rate in RSA. 5 In contrast, autoimmune RSA is caused by autoimmune diseases and is significantly related to autoimmune factors. 6he balance between immune tolerance and effector immunity is significantly related to successful pregnancy, and an imbalance in the regulatory system can cause adverse reproductive outcomes.Cyclosporine A (CsA), isolated from metabolites of Trichoderma polysporum and Cladospora, can inhibit autoimmune responses. 7The use of CsA can regulate maternal-fetal immunity by inhibiting maternal immune rejection of embryonic antigens and promoting the growth, movement, and invasion of trophoblast cells, and thus it is considered an effective drug for the treatment of RSA. 8 The use of conventional doses of prednisone and CsA in pregnant patients following organ transplantation has been established as safe for fetuses and newborns. 70][11] A study by Li et al. found that low-dose CsA promotes the proliferation, invasion, and migration of villous trophoblast cells, allowing successful implantation to be achieved. 9Further, Azizi et al. found that the use of CsA in women with RSA with an elevated TH1/TH2 ratio was significantly associated with an elevated live birth rate and improved TH1/TH2 ratio. 10Moreover, the use of CsA has been shown to improve embryo implantation and pregnancy rates in patients undergoing in vitro fertilization and embryo transfer. 11However, the rate of failure in pregnancy outcomes for RSA women treated with CsA remains high, and the prognostic factors require identification for further management and improvement of pregnancy outcomes.Therefore, the current retrospective study was performed to identify the prognostic factors and construct prognostic model for pregnancy outcomes in RSA women treated with CsA.

Study design and patients
This retrospective cohort study recruited 209 patients with RSA who were admitted to the present study's hospital between October 2016 and October 2018.This study was approved by the Ethics Committee of Shengjing Hospital of China Medical University (nº 2022PS014K), and informed consent was waived owing to the retrospective design of the current study.All patients were treated with CsA, with 25 mg bid per day from the preparation for pregnancy until 12 weeks of gestation or abortion.The preparation of pregnancy will be followed by a one-year pre-pregnancy phase, and if participants do not become pregnant within that year, their treatment and participation will be terminated.The additional inclusion criteria are as follows: 1) Continued spontaneous abortion or unexplained fetal arrest and death occurred in utero more than two times; and 2) Neither partner was consanguineous and had normal karyotypes.After removing patients with incomplete clinical data, 154 were ultimately selected for analysis.

Outcome definition
Pregnancy success was defined as the occurrence of uneventful pregnancies that were spontaneously conceived, resulting in the delivery of a live baby following treatment.Pregnancy failure was defined as the termination of pregnancy and failure to deliver a live baby due to abortion, fetal arrest, fetal death, or other reasons following treatment.

Statistical analysis
The characteristics of patients according to pregnancy outcome are shown as means (standard deviation) or medians (quartile) for continuous variables according to data distribution, while categorical variables are shown as events (frequency).The differences between groups were assessed using a t-test and the Kruskal-Wallis test for continuous variables, whereas the Chi-Square test was applied to assess the differences in categorical variables.Multivariate logistic regression was applied to identify the prognostic factors of pregnancy success in women treated with CsA.The area under the receiver operating characteristic curve (AUC) and 95 % Confidence Interval (95 % CI) were calculated to assess the prognostic values of the constructed model.Exploratory analyses were performed according to disease status.All reported p-values were 2-sided, and the inspection level was set at 0.05.Statistical analysis in this study was performed using SPSS version 19.0 (SPSS 19.0,IBM Corp., Armonk, NY, USA).

General characteristics
A total of 154 patients were analyzed, with a mean age of 33.24 years.The characteristics of patients according to disease status are shown in Table 1.There were significant differences between URSA and RRSA for ANA (p < 0.001), anti-β2 glycoprotein-I-antibodies (p = 0.002), anti-SSA antibodies (p < 0.001), and anti-SSB antibody (p = 0.043).However, there were no significant differences between URSA and RRSA for ACA (p = 0.

The prognostic factors for pregnancy success in URSA and RRSA
The pregnancy success rate following treatment with CsA in RRSA patients was higher than that in URSA patients (75.68 % vs. 66.67 %), while the difference in this value between patients with RRSA and URSA was not statistically significant.
The characteristics of the included URSA patients according to pregnancy outcomes are shown in Table S1.There were significant differences between pregnancy failure and pregnancy success in URSA patients with respect to age (p < 0.001) and ANA (p = 0.002), while no other significant differences between groups were observed.Moreover, multivariate logistic regression analysis suggested that increased age (OR = 0.809; 95 % CI 0.725-0.902;p < 0.001), positive ANA (OR = 0.158; 95 % CI 0.051-0.487;p = 0.001), and positive ACA (OR = 0.089; 95 % CI 0.010-0.797;p = 0.001) were associated with a reduced incidence of pregnancy success (Table S2).The AUC of combining these variables for predicting pregnancy success in patients with URSA was 0.779 (95 % CI 0.689-0.870);(Fig. S1).
The characteristics of the included RRSA patients according to pregnancy outcomes are shown in Table S3.The pregnancy failure group was significantly older than the pregnancy success group (p = 0.004), while no other significant difference was observed.Moreover, multivariate logistic regression analysis indicated that older age was associated with a lower incidence of pregnancy success (OR = 0.683; 95 % CI 0.512-0.910;p = 0.009; Table S4).The AUC of combining these variables for predicting pregnancy success in RRSA patients was 0.823 (95 % CI 0.655-0.990);(Fig. S2).
In the planning stage, changes in NK should be analyzed for patients before and after treatment with CsA.However, only 49 patients reported NK before and after CsA treatment.The percentage of NK cells before and after treatment with CsA was 20.31 % and 15.38 %, respectively, and the difference was statistically significant (−4.93 %; p < 0.05).

Discussion
CsA is widely used as an immunosuppressant for organ transplantation and autoimmune diseases.Animal studies have shown that low doses can induce maternal-fetal immune tolerance and enhance trophoblast invasion.CsA is used to prevent and resist graft-vs.-hostreactions and no adverse reactions to fetuses have been found in pregnant patients who take conventional doses of CsA for an extended time following organ transplantation. 12Previous studies have addressed the treatment effects of CsA in women with RSA. 11,13For example, Fu et al. recruited 168 refractory immune RSA patients treated with CsA (2-4 mg/kg), and maintenance of 80-150 mg/L CsA was associated with an elevated live birth rate. 13Ling et al. identified 86 URSA patients treated with 100 mg/day for 12 weeks, which increased the live birth rate, and no adverse events were detected. 11These results suggest that CsA should be considered an effective treatment for URSA.However, the prognostic factors for pregnancy success in women taking CsA remain controversial.The current retrospective study recruited 154 women treated with CsA, and the characteristics of the included patients varied widely.The pregnancy success rate in RSA patients treated with CsA was 68.83 %, and the incidence of pregnancy success can be affected by age, ANA, and anti-β2 glycoprotein I antibody, and the prognostic performance of the constructed model based on these factors was relatively higher.Subgroup analyses found the prognostic factors for pregnancy success in URSA women included age, positive ANA, and positive ACA, and the prognostic value of the prognostic model based on these factors was moderate.Moreover, the prognostic factors for pregnancy success in RRSA women included increased age, and the prognostic value was higher for predicting further pregnancy success.
Several studies have previously addressed the prognostic factors of pregnancy outcomes in patients with RSA. 14,15One retrospective study by Yang et al. recruited 492 singleton pregnant women and found that a history of first-trimester RSA was an independent risk factor for cesarean section and pregnancy complications. 14Additionally, Liu et al. conducted a retrospective study of 1240 pregnant women with a history of RSA and found that a cutoff value of serum β-subunit of human chorionic gonadotropin of 88,000 IU/L could predict early pregnancy outcomes. 15Maesawa et al. recruited 175 women with RSA and found that two or more biochemical pregnancies were associated with an increased risk of reproductive failure and spontaneous abortion with normal chromosomes, together with a lower chance of live birth. 16In another study, Caetano et al. performed a case-control study and found that age > 40years, immunological factors, and two or more concomitant factors were significantly related to poor gestational outcomes in women. 17Kruse et al. recruited 217 women with URSA and found that low maternal serum mannan-binding lectin levels could affect pregnancy outcomes for women with URSA. 18However, the prognostic factors for pregnancy success in women with RSA treated with CsA have not yet been identified.Therefore, the authors performed this retrospective study to identify prognostic factors for pregnancy outcomes for RSA women treated with CsA.
Increased age and ANA positivity were associated with a reduced incidence of pregnancy success in this study, which aligns with the existing evidence that maternal age is an independent predictor of poor maternal and fetal outcomes during pregnancy. 19The prevalence of ANA has previously been demonstrated to be higher in recurrent pregnancy loss, and positive ANA results are associated with an increased risk of miscarriage. 20Moreover, the disappearance of ANA in early pregnancy is associated with ongoing pregnancy, which indicates an improvement in systemic autoimmunity. 21The pregnancy success group Table 1 The baseline characteristics of included patients according to disease status.presented with higher positive anti-β2 glycoprotein-I-antibody rates in this study.As β2-glycoprotein functions as a physiologic anticoagulant by inhibiting the key procoagulant activities of thrombin, 22 pregnancy outcomes could be affected.The results of the present study found that positive anti-β2 glycoprotein-I-antibody patients treated with CsA were associated with an increased incidence of pregnancy success, which suggests the use of CsA could offer more benefit for patients who presented positive anti-β2 glycoprotein-I-antibody.The result might not have been stable owing to the marginal 95 % CI, and further prospective studies are needed to verify this finding.
Another study previously found that an imbalance of Th/Ts plays an important role in the progression of RSA. 23Villous trophoblast cells at the maternal/fetal interface initiate and complete uterine placental circulation, secreting a variety of cytokines to maintain the Th1 /Th2 cell balance.Once this balance is disrupted, the fetus is attacked and miscarried. 24IFN-γ, IL4, IL13 and other cytokines secreted by Th1 and Th2  cells interact with each other to maintain maternal immune tolerance to the fetus and ensure fetal growth and development during pregnancy, while Th1 cells oversecrete IFN-γ, which can damage trophoblast cells by inducing activated NK cells, hindering embryo implantation and placental growth and development, and thus lead to abortion. 25The use of CsA may restore the balance of Th/Ts, which could explain why Th/Ts was not associated with the incidence of pregnancy success in the present study. 26everal limitations of this study should be acknowledged.First, it was designed as a retrospective cohort, and the results could, therefore, have been affected by selection and recall biases.Second, the background treatment strategies for RSA patients differ, which could affect the pregnancy outcomes.Third, the severity of RSA could affect pregnancy outcomes, which were not addressed in this study.Fourth, it remains unclear whether the RSA is caused by alloimmunological factors, which should be further explored.Finally, stratified analyses were only performed based on disease status, whereas based on several other important factors were not performed.

Conclusions
The findings of this study showed that age, ANA, and anti-β2 glycoprotein I antibody levels were significantly associated with the incidence of pregnancy success, and the prognostic models for these variables showed higher value for predicting pregnancy success in women treated with CsA.Further large-scale prospective studies should be performed to verify the results of this study and validate the constructed prognostic model.

Fig. 1 .
Fig. 1.The area under the receiver operating characteristic curve for predicting pregnancy success.

Table 2
Logistic regression for the prognostic factors for pregnancy success.