Role of the CD40–CD40L expression level pathway in the diagnosis of unexplained recurrent pregnancy loss

Abstract Background Unexplained recurrent spontaneous pregnancy loss (URPL) lacks effective treatment and reliable early diagnosis and prediction. Immunologic dysfunction can be an underlying cause of recurrent pregnancy loss (RPL). Considering the regulatory role of CD40–CD40L in immune responses, we explored its clinical significance in URPL. Methods The 108 women with URPL who were treated in Hebei Yanda Hospital from January 2020 to December 2022 were selected as study subjects, and another 108 healthy women who were not pregnant and matched with the age and body mass index of the study group were selected as the control group. CD40 and CD4 + CD25 + Treg cells in peripheral blood mononuclear cells (PBMCs) and CD40L in peripheral blood platelets were measured by flow cytometry. The predictive value of CD40–CD40L in URPL for the risk of RPL was determined by receiver operating characteristic (ROC) curves. The correlations of CD40–CD40L with CD4 + CD25 + Treg cells and serum pro-inflammatory factors were assessed by Pearson’s analysis. Results CD40 on the surface of PBMCs and CD40L on the surface of platelets were up-regulated in URPL patients. CD40 in combination with CD40L had high predictive value for the risk of RPL in URPL patients. Peripheral blood CD40–CD40L was positively linked to IL-17 and IL-23, and negatively to CD4 + CD25 + Treg cells and IL-10 in URPL patients. Conclusions The CD40–CD40L pathway expression in peripheral blood can help predict the risk of RPL in URPL patients. Plain language summary This study discovered significant increases in the blood levels of two transmembrane glycoproteins, CD40 and CD40L, in patients with two or more consecutive unexplained recurrent spontaneous pregnancy loss, compared to healthy women. The elevated levels of CD40 and CD40L indicated an increased risk of recurrent pregnancy loss in these patients, which could help early and timely management of these patients.


Introduction
Recurrent pregnancy loss (RPL) is perceived as the termination of two or more pregnancies prior to 24 weeks of gestation, which affects approximately 1-5% of women desiring a child (ESHRE Guideline Group on RPL et al. 2018).The consequences of RPL are both physical, such as infection or bleeding, and psychological, including increased risks for post-traumatic stress disorder, anxiety, depression and suicide (Toth et al. 2018, Issakhanova et al. 2023).The aetiology of RPL is complex and lacks specific clinical manifestations, such as parental and embryonic chromosomal abnormalities, endocrine dysfunction, immunological diseases, uterine structural abnormalities, prethrombotic states, infection or environmental factors.However, in addition to known causes, at least 50% of RPL is unexplained, which is classified as unexplained recurrent spontaneous pregnancy loss (URPL) (Homer 2019, Deng et al. 2022).Due to the complex aetiology and pathogenesis of RPL, strategies to predict and reduce RPL remain ineffective.Therefore, elucidating the underlying causes of URPL has become a priority in the field of reproductive medicine research.However, there is currently no safe and effective treatment for URPL and reliable early diagnosis and prediction.
It is broadly acknowledged that a successful pregnancy extremely necessitates a highly regulated and fine-tuned balance between embryonic antigen tolerance and immune activation (Yang et al. 2019).Despite close contact with immunologically foreign foetal-placental alloantigens during pregnancy, the maternal immune system does not reject the foetus but contributes to its implantation and development in the uterus (Mor et al. 2017, Eskandarian andMoazzeni 2019).Describing the characteristics of the maternal immune system during pregnancy is of imperative importance to understand how it maintains tolerance to the allogeneic foetus (Abu-Raya et al. 2020).Indeed, immunological aberrancy is thought to be a root contributor to adverse pregnancy outcomes, and immune dysfunction also accounts for over half of RPL cases (Li D et al. 2021).
CD40 is an acknowledged transmembrane protein belonging to the tumour necrosis factor (TNF) receptor superfamily, which is expressed in immune as well as non-immune cells (Karnell et al. 2019).Of note, CD40 and its ligand CD40L are broadly identified as stimulatory immune checkpoints and confer an essential function in numerous immunological processes conducive to cell-mediated and humoral immune responses (Tang et al. 2021).The CD40-CD40L interaction plays vital roles in B cell formation, immunoglobulin production, isotype switching, antigen presentation and immune activation, which is crucially involved in the progression of autoimmune disease (Laman et al. 2017).CD40-CD40L costimulatory pathway not only amplifies T cell-dependent immune responses and promotes T cell activation, but also underlies the generation of memory CD8 þ T-cells, promotes CD8 þ T cell expansion and pluripotency, and plays a key part in the differentiation of CD4 þ T cells, thereby mediating stronger anti-tumour effects and target cell killing effects (Elizondo et al. 2017, Nishat et al. 2018).Notably, the maturation stage of uterine dendritic cells exerts a vital function in the aetiology of RPL (Eskandarian and Moazzeni 2019).Within the adaptive immune system, an array of T cells emerge as potential participants in maintaining immune tolerance towards the semi-allogeneic foetus, and the abnormal functions and frequencies of immune cells in human decidua have been documented in various obstetrical disorders, including RPL (Liu et al. 2017).Moreover, immune factors associated with URPL are receiving increasing attention, such as plasma cells, natural killer cells, autoantibodies, dendritic cells, human leukocyte antigen system-sharing and regulatory T cells (Treg), accompanied by treatment options such as intravenous immunoglobulin, intralipid injection, corticosteroids, heparin and aspirin (Vomstein et al. 2021).Treg cells secrete the anti-inflammatory cytokine IL-10, which can selectively impede Th1-mediated cellular responses by suppressing the secretion of inflammatory cytokines and hinder the activation of Th17 and Th1 cells (Cai et al. 2016).The predominance of Th2 cytokines (IL-4, IL-5 and IL-10) over Th1 cytokines (TNF-a, IFN-c and IL-2) and of Treg cytokines over Th17 cytokines at the maternal-foetal interface in the context of successful pregnancy (Lee et al. 2011, Grandi andTramontano 2018).To date, however, there are no reports regarding the role of the CD40-CD40L pathway in URPL.Based on this background, the present study investigated the expression levels of CD40-CD40L in peripheral blood of patients with URPL and their clinical significance in assisting URPL screening and predictive diagnosis, and analysed their correlation with Treg cells and inflammatory factors (IL-17, IL-23 and IL10) in order to find more immune factors that can be used for URPL screening and early prevention or treatment.

Ethics statement
This study was ratified by the ethical review committee of Hebei Yanda Hospital (approval no.2022-0821, 08/21/2022).All participants had voluntarily signed the informed consent.

Study subjects
A total of 225 women with a history of RPL who were treated in Hebei Yanda Hospital from January 2020 to December 2022 were selected as screening subjects.Based on the inclusion criteria of patients with URPL, 108 patients with URPL were finally included in the study, and according to the hospital number, another 108 healthy women without a history of spontaneous pregnancy loss (the control group) who were matched with the age and body mass index (BMI) of the study group were randomly selected as the control group.The STROBE flowchart of the subjects included in the study is shown in Figure 1.All URPL patients experienced pregnancy loss within 8-10 weeks.The median age and interquartile range (IQR) of URPL patients were 28 (26, 32) years, while that of the control group was 29 (26, 31) years.All cases were selected from the Department of Family Planning and the Department of Reproductive Medicine of Hebei Yanda Hospital.The inclusion criteria for URPL patients were as follows: (1) a history of two or more spontaneous pregnancy loss before 20 weeks of pregnancy; (2) with normal menstruation; (3) without uterine abnormalities, endocrine disorders, infectious diseases, chromosomal abnormalities or autoimmune diseases associated with URPL, and no abnormalities during laboratory tests.Specifically, the diagnosis of URPL in all included URPL patients was made according to the following criteria to exclude any verifiable cause: first, pelvic examination and transvaginal ultrasound to confirm an anatomically normal uterus.If abnormalities were suspected, a hysterosalpingogram or diagnostic hysteroscopy was performed for further confirmation; second, normal karyotype of both partners and absence of embryonic chromosomal abnormalities were confirmed; third, cervical mucus was cultured to exclude chlamydia and ureaplasma infection; fourth, endocrine (luteal deficiency, polycystic ovary syndrome, hyperandrogenemia and hyperprolactinemia) and metabolic disorders (insulin resistance, diabetes mellitus, hyperthyroidism and hypothyroidism) were ruled out; fifth, Leiden factor V and prothrombin gene mutations, as well as protein C, protein S and antithrombin III activity were tested to exclude hereditary thrombosis; finally, autoantibodies, including anticardiolipin and antinuclear antibodies, were analysed to rule out autoimmune disease.In addition, partners of all RPL patients were tested for normal semen status.The control group were required to exclude the following diseases: (1) spontaneous pregnancy loss, ectopic pregnancy, preterm or stillbirth; (2) endocrine diseases: thyroid disease, gonadal imbalance; (3) genetic history, autoimmune diseases, diabetes mellitus, tumours and other diseases.
All RPL patients in this study underwent surgical or medical management of miscarriage, and blood HCG levels were measured every other day to monitor for threatened or unavoidable pregnancy loss.If the blood HCG level was prominently lower than normal level, an ultrasound examination was performed.If the ultrasound showed loss of foetal heartbeat and a reduction in HCG levels, RPL was confirmed and it is recommended that the patient should undergo surgical or medical management of miscarriage within 24 h.Missed (delayed) pregnancy loss was not included.Peripheral blood samples were collected within 24 h of the onset of pregnancy loss symptoms from URPL patients.There was no significant difference in blood collection time between the control subjects and the gestational weeks of blood collection in URPL patients.
The immunostaining of platelets was performed in concert with prior research (Michelson et al. 2000, Garlichs et al. 2001).Briefly, the fixed blood was diluted (1:200) in phosphate-buffered saline (PBS) (the collected peripheral blood was placed in a 3.2% sodium citrate vacuum container (Becton-Dickinson, San Jose, CA), and citrate anticoagulant blood was immediately fixed with 1% formaldehyde (1:1, v/ v)) and then incubated for 30 min with antibodies of saturating concentrations (anti-CD40L and anti-CD41a platelet-specific, directed against the membrane glycoproteins IIb/IIIa on platelets) at room temperature.Cells were detected with a flow cytometer (FACSort, Becton-Dickinson, San Jose, CA) within 2 h after sampling and then analysed by CellQuest software (Becton-Dickinson, San Jose, CA).Calibration beads (Becton-Dickinson, San Jose, CA) were used for fluorescence calibration.The results were expressed as net mean fluorescence intensity (MFI).The MFI value of each treatment method was acquired by subtracting the MFI value of the isotype control group from the MFI value of the positively stained sample.The identification of platelets was performed by gating on CD41a-phycoerythrin positivity as well as their characteristic light scatter.The evaluated platelet population was observed to be �98% positive for CD41a.
To assess CD40 on monocytes, fixed blood was diluted at a ratio of 1:5 and cultivated with anti-CD40 and anti-CD14 (antiendotoxin receptor) at room temperature for 30 min at RT. Erythrocytes were eliminated by the addition of 2 mL of FACS lysis solution (Pharmingen, Franklin Lakes, NJ) for 10 min at RT.After two PBS washes, leukocytes were fixed in formaldehyde (1%) in PBS.The identification of monocytes was conducted by gating on CD14þ cells.The results were displayed MFI.
The isolation of peripheral blood mononuclear cells (PBMCs) was implemented utilising Ficoll density gradient centrifugation.After the cells were washed three times with PBS, 3 lL of FITC-CD4 (eBioscience, San Diego, CA) and PE-CD25 (eBioscience, San Diego, CA) antibodies were added into 100 lL adsorbent to adjust the cell density to 1 � 10 6 / mL under dark conditions, followed by three PBS washes.Afterwards, flow cytometry (FACSCanto II from BD, Franklin Lakes, NJ) was implemented to test the cell suspension.

Statistical analysis
SPSS 21.0 (IBM Corp., Armonk, NY) and GraphPad Prism 8 (GraphPad Software Inc., San Diego, CA) were utilised for data analyses and plotting.Kolmogorov-Smirnov's test was performed to assess the distribution normality of continuous variables.For normally distributed continuous variables, data were depicted as mean ± standard deviation (SD).For non-normally distributed variables, data were exhibited as median and IQR.When variables were normally distributed, an independent sample t-test was conducted to compare measurement data between the two groups, otherwise Mann-Whitney's U-test was used.Receiver operating characteristic (ROC) curves were adopted to analyse the predictive value of CD40 and CD40L for the risk of RPL in URPL patients.Pearson's analysis was performed to analyse the correlation between CD40-CD40L with different parameters.The p value was acquired from a twosided test and p < .05indicated statistical significance.

Clinical baseline data
As presented in Table 1, there was no evident difference in age and BMI between URPL patients and healthy women in the control group (p > .05),but there was a prominent difference in the number of gravidity (p < .001).The median and IQR for URPL patients with a previous pregnancy loss was 8 (7, 8.75) weeks and for the number of spontaneous pregnancy loss was 4 (2, 6).

The expression levels of the CD40-CD40L pathway in peripheral blood
In this study, we compared the CD40-CD40L pathway expression levels in peripheral blood of URPL patients and healthy women in the control group.Flow cytometry showed that CD40 expression on the surface of monocytes was noticeably upregulated in URPL patients compared with healthy women in the control group (p < .001, Figure 2(A)).Meanwhile, CD40L expression on the surface of platelets was increased in URPL patients (p < .001, Figure 2(B)).

The CD40-CD40L pathway expression had predictive value for the RPL in RPL patients
We further plotted the ROC curves to evaluate the auxiliary diagnostic value of CD40 and CD40L expression for URPL, respectively.The area under the curve (AUC) for CD40 in the diagnosis of URPL was 0.8982 (p < .0001, Figure 3(A)), with a cut-off value of 17.47 (84.26% specificity and 80.56% sensitivity).The AUC for CD40L in diagnosing URPL was 0.9531 (p < .0001, Figure 3(B)), with a cut-off value of 13.67 (94.44% specificity and 81.48% sensitivity).Subsequently, ROC curve was used to analyse the predictive value of a combination of CD40 and CD40L for the risk of RPL in URPL patients (Figure 3(C)).The AUC was 0.9634 (p < .0001),with 98.15% specificity and 81.48% sensitivity.Overall, when CD40 level in monocytes of RPL patients was higher than 17.47 or CD40L level was higher than 13.67, the possibility of RPL might increase, and CD40 combined with CD40L had a better predictive value for the risk of RPL in URPL patients than CD40 or CD40L alone.

The number of CD4 1 CD251 Treg cells was reduced in peripheral blood of URPL patients and negatively correlated with CD40-CD40L expression
Unlike cytotoxic T lymphocytes, Treg cells are a subset of CD4þ T helper cells that confer essential functions in For normally distributed continuous variables, data were depicted as mean ± standard deviation (SD).For non-normally distributed variables, data were exhibited as median and interquartile range (IQR).

The correlation between CD40-CD40L and inflammatory factors in URPL patients
The course of pregnancy is strongly associated with inflammatory events (Zhang X and Wei 2021).ELISA elicited that URPL patients had higher levels of pro-inflammatory factors IL-17 and IL-23 and lower anti-inflammatory factor IL-10 level in peripheral blood serum than women undergoing normal pregnancy (p < .001,Supplementary Figure 1A).Subsequently, Pearson's analysis was used to evaluate the association between peripheral blood CD40-CD40L levels and serum inflammatory factors in URPL patients.CD40 level on the PBMC surface was evidently inversely correlated with IL-10 level (r ¼ −0.3317, p ¼ .0005)and positively correlated with IL-17 and IL-23 levels (r ¼ 0.3517, p ¼ .0002;r ¼ 0.3499, p ¼ .0002,Supplementary Figure 1B).Similarly, CD40L level on the platelet surface was negatively related to IL-10 (r ¼ −0.3272, p ¼ .0005)and positively linked with IL-17 and IL-23 (r ¼ 0.3475, p ¼ .0002;r ¼ 0.3595, p ¼ .0001,Supplementary Figure 1C).

Summary
Our findings elucidated that the CD40-CD40L levels were elevated in URPL patients and could assist in the predictive diagnosis of URPL.RPL is a complex, multifactorial and distressing pregnancy disorder (Li J et al. 2021).Among the established causes of RPL, the disruption of immune response to the embryo seems to be the most dominant (Wu H et al. 2021).Foetal chromosomal aberrations were found in 29-60% of recurrent miscarriage patients (Ogasawara et al. 2000, Philipp et al. 2003, Larsen et al. 2013).Although not accurately quantified, a large part of URPL has something to do with immune cause (Weinbaum et al. 1989, Ticconi et al. 2019).In these cases, pregnancy loss may occur through continuous interference of several immune pathways (Wang et al. 2016).As well-known,  the embryo expresses paternally derived antigens foreign the mother, which are considered analogous to allografts; therefore, inducting maternal immunosuppression of the embryo and foetus is of considerable importance in maintaining maternal-foetal tolerance (Cai et al. 2016).It can be ascertained that immune dysregulation at the maternal-foetal interface is indeed critical in the etiopathogenesis of RPL.Particularly, CD40-CD40L interaction is currently an essential immune system required for the regulation of host immune response against various pathogens and diseases, the failure of which can lead to autoimmune disorders (Chand Dakal et al. 2020).Thereby, the current study preliminarily explored the clinical significance of the CD40-CD40L pathway expression in peripheral blood of URPL patients.
Hereditary or acquired thrombosis is bound up with RPL (Bennett et al. 2012).Antithrombotic therapy and antiplatelet drugs have been proposed as potential tools for RPL prevention, and monocyte-platelet aggregates formed by activated platelets binding to P-selectin glycoprotein ligand-1 on monocytes are important as markers of platelet and monocyte activation, and exert crucial effects in promoting thrombosis and inflammatory responses (Zaldivia et al. 2017, Allen et al. 2019).Platelets can directly initiate inflammatory responses in the vascular wall that involves CD40-CD40 ligand interactions, where CD40 is a transmembrane glycoprotein belonging to the TNF receptor superfamily constitutively expressed on monocytes, B cells, dendritic cells, smooth muscle cells, macrophages, platelets and endothelial cells, whereas CD40 ligand (CD40L or CD154) is a tumour-associated transmembrane glycoprotein identified on activated T cells, polymorphonuclear cells, basophils, mast cells, macrophages, natural killer cells, smooth muscle cells and endothelial cells (Freedman 2003).CD40L is also detected in platelets, which are translocated to the platelet surface upon stimulation, and activated platelet surface-expressed CD40L interacts with CD40 on monocytes or endothelial cells, which ultimately induces the synthesis of pro-inflammatory cytokines, adhesion molecules, tissue factors and chemokines (Henn et al. 1998).Based on the previous study (Lukanov et al. 2010), the expression of CD40L in platelets and the expression of CD40 in monocytes were detected in this study.It was found that the expression of CD40 on the surface of monocytes and CD40L on the surface of platelets were both up-regulated in URPL patients.In agreement with our findings, CD40 molecule is upregulated in splenic B cells in pregnancy loss-prone mice, antigen-presenting cells (APCs) activation via CD40 causes reduced implantation events, and overexpression of CD40 on APCs probably adversely impacts the pregnancy outcome (Lorek et al. 2019).Treatment with CD40L monoclonal antibody can definitely diminish the embryo resorption rate and induce maternal hyporesponsiveness to the paternal antigens in pregnancy loss-prone CBA/ J � DBA/2 mice, preventing maternal immune rejection of allogeneic embryos and resulting in maternal immune tolerance (Li G et al. 2020).The previous researchers are mostly based on in vivo animal studies to investigate the role of CD40-CD40L on pregnancy loss.Evidence favouring the clinical significance of CD40-CD40L by performing clinical studies is, however, sparse, which reflects the novelty of our research.Importantly, our results uncovered the expression levels of both CD40 and CD40L had assistant values of predictive diagnosis for URPL, and their combination presented better diagnostic efficacy.
Treg cells are imperative in regulating maternal immune tolerance, and altered function and lower numbers of Treg cells are related to adverse pregnancy outcomes in animal and human studies, indicating the necessity of sufficient numbers and normal function of Treg cells for pregnancy success (Prins et al. 2014).Prior to implantation, the accumulation of Treg cells is noted in the uterine lymph node in response to hormonal together with paternal antigen stimulation, which are expanded and transported to the pregnant uterus through human chorionic gonadotropin-induced chemoattractant mechanisms and resident in the decidua (Deshmukh and Way 2019).Treg cells, particularly the CD4 þ CD25þ subtype, are reduced in decidual tissues and peripheral blood from patients with spontaneous pregnancy loss (Ebina et al. 2016, Quan andYang 2017).Our results found that the level of CD4 þ CD25þ Treg cells was diminished in the peripheral blood of URPL patients and showed a negative correlation with CD40-CD40L levels.Lymphocyte immunotherapy (LIT) is considered an effective treatment for URPL.A prior study have reported that LIT is effective in inducing blocking antibody production and enhancing live birth and pregnancy rates in URPL patients (Liu et al. 2021).Paternal immunotherapy is an effective treatment for URPL and urinary incontinence, and paternal immunotherapy diminish the number of in vitro fertilisations required to achieve pregnancy (Fainboim et al. 2021).In URPL patients, immunotherapy with mononuclear cells from the infant's father can impact the balance of Th17/Treg and Th1/Th2 and Th2/Treg imbalance is favourable to pregnancy (Wu L et al. 2014).Prominent progress has been made in CD40/CD40Ltargeted therapies recently.Various types of drugs such as cellular vaccines, agonistic/antagonistic monoclonal antibodies, protein antagonists and adenoviral vectors have been developed and are being applied in early clinical trials for the treatment of autoimmune diseases, malignant tumours and allograft rejection.Overall, these drugs are safe and some of them have good clinical efficacy (Tang et al. 2021).Our findings provided a reference for the possible application of CD40/Cd40L-targeted therapy in the treatment of URPL.Moreover, in our experiments, URPL patients showed elevated IL-17 and IL-23 and decreased IL-10 levels in peripheral blood serum.Notably, CD40-CD40L levels were positively linked with IL-17/IL-23 and inversely related to IL-10.To our knowledge, CD40 and CD40L are pivotal regulators in initiating and sustaining inflammatory responses, and the CD40-CD40L interaction is capable of upregulating key proinflammatory molecules (Zhang B et al. 2013, Ots et al. 2022).The inhibition of CD40L expands the peripheral CD4 þ CD25þ T cell population, increases IL-4 expression, and decreases TNF-a and INF-c in pregnancy loss-prone models (Li G et al. 2020).Based on the aforementioned evidence, it could be summarised that the regulatory role of the CD40-CD40L pathway in URPL involved the Treg cells and inflammatory cytokines.
In conclusion, this present study highlighted that CD40L levels were increased in peripheral blood of URPL patients and could assist in URPL diagnosis, which were also associated with Treg cells and inflammatory factors.This study provided more reference for finding more immune factors related to RPL that can be utilised for URPL screening and early prevention or treatment.However, there were still some limitations in our study.First, the number of enrolled patients is insufficient, which may lead to the contingency of experiments and result error.Second, the incidence of intrauterine adhesion is increased due to repeated pregnancy loss and curettage, which also results in a decrease in the ability of the endometrium to accept embryos, possibly causing another miscarriage.Future studies should expand the sample size to further confirm the validity of the CD40-CD40L pathway expression levels in aiding the predictive diagnosis of URPL patients.It is also worthwhile exploring the association of various immune cells with the CD40-CD40L pathway.Third, more and more risk factors for URPL have been found.When exploring the value of peripheral blood CD40-CD40L level in aiding the predictive diagnosis of URPL, it is more important to include clinical indicators including more URPL risk factors to further analyse the independent correlation between CD40-CD40L expression and URPL occurrence.Fourth, levels of all markers were acquired from peripheral blood, not from the endometrium or placenta, and the immune reactions were achieved between foetus and maternal tissues.Fifth, we chose CD4 þ CD25þ labelling in detecting Treg cells.However, it is more accurate to add Foxp3 to differentiate Treg cells.Sixth, in our study, CD40L expression in platelet surface and CD40 expression in monocytes were boosted in URPL patients.Platelet-derived CD40L may have the ability to stimulate resting platelets by binding to CD40 during direct cell-cell contact, and the interaction of CD40L on activated platelets with CD40 on monocytes induces proinflammatory alterations, as well as the synthesis of adhesion molecules, cytokines, chemokines and tissue factors.Thus, the association between excessive platelet activation and inflammation of injury in URPL can be explained by a sustained upregulation of the CD40-CD40L system.Our current results do not yet fully demonstrate the CD40-CD40L interaction, which promotes the formation of monocyte-platelet aggregates, and further validation remains to be carried out by detecting changes in platelet-monocyte aggregate levels in URPL.

Figure 2 .
Figure 2. The expression levels of the CD40-CD40L pathway in peripheral blood.(A) CD40 expression on the surface of monocytes was measured by flow cytometry; (B) CD40L expression on the platelet surface was measured by flow cytometry.Data comparisons between two groups were carried out using an independent sample t-test.��� p < .001.

Figure 3 .
Figure 3.The CD40-CD40L pathway expression could help predict the risk of RPL in URPL.(A) ROC curve analysis of CD40 expression on the surface of monocytes for URPL diagnosis; (B) ROC curve analysis of platelet CD40L for URPL diagnosis; (C) ROC curve analysis of CD40 combined with CD40L for URPL diagnosis.

Figure 4 .
Figure 4.The number of CD4 þ CD25þ Treg cells was reduced in peripheral blood of URPL patients and negatively correlated with CD40-CD40L expression.(A) Flow cytometry determined the level of CD4 þ CD25þ Treg cells in peripheral blood; (B, C) Pearson's analysis assessed the correlation of CD40 and CD40L levels with CD4 þ CD25þ Treg cells.An independent sample t-test was used for comparisons between two groups in panel A. ��� p < .001.