Serum Soluble Suppression of Tumorigenicity-2 Levels Predict Cardiovascular Events in Patients Undergoing Incident Peritoneal Dialysis: A Prospective Cohort Study

Peritoneal dialysis (PD) is a home-based renal replacement therapy for patients with end-stage renal disease (ESRD) [1]. The major cause of death in patients undergoing PD is cardiovascular (CV) disease, which accounts for nearly half of all deaths of a known cause [2]. Therefore, the prediction of CV events is critical in these patients, as it may improve the overall management and prognosis.

Suppression of tumorigenicity-2 (ST2) is a member of the interleukin (IL)-1 receptor family [3]. The two essential forms are a transmembrane form (ST2 ligand, ST2L) and a soluble form (sST2). By binding to ST2L, IL-33 exerts its cellular functions, including several cardioprotective effects. sST2 acts as a decoy receptor to sequester free IL-33, preventing activation of the ST2/IL-33 signaling pathway and subsequently aggravating myocardial fibrosis and myocardial cell apoptosis [4, 5]. It is regarded as a marker of fibrosis, remodeling, and inflammation.

Previous studies have demonstrated that serum sST2 level is independent of renal function and age and is not influenced by hemodialysis (HD) [6‒8]. Therefore, it may be a stable biomarker, and its prognostic value in patients undergoing HD has attracted increasing interest. Studies in HD have found that elevated sST2 levels are independently predictive of all-cause and CV mortality [9, 10]. However, the role of sST2 in PD has not been adequately investigated. Only one study involving patients with prevalent PD has shown that higher serum sST2 levels are associated with increased mortality and CV event risk [11]. In this study, we explored the ability of sST2 to predict CV events in patients receiving incident PD.

In this longitudinal single-center cohort study, all adult patients (≥18 years of age) who initiated PD therapy from March 2020 to June 2021 in the PD center of West China Hospital of Sichuan University and regularly been followed up subsequently were included. Exclusion criteria included a history of kidney transplantation, prior HD, active infections before the first peritoneal equilibration test (PET), malignancy, chronic liver disease, or systemic rheumatic disease. Patients who could not tolerate a 2-L peritoneal fill volume during PET were also excluded.

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of West China Hospital of Sichuan University [2019(1015)]. Written informed consent was obtained from all the participants. This study was registered in the Chinese Clinical Trial Registry (https://www.chictr.org.cn/index.aspx, ChiCTR1900027473).

The patients’ medical histories, demographic characteristics, and dates of PD initiation were collected. Baseline CV disease, smoking history, and type of PD were also recorded. Blood samples to measure sST2 levels were obtained before PD catheter implantation. All patients underwent a standard PET 4–6 weeks after initiation of PD. Blood samples for laboratory parameters and dialysate samples were obtained during the first PET. Complete blood count, biochemical indexes, high-sensitive C-reactive protein, IL-6, β2-microglobulin, intact parathyroid hormone, and N-terminal pro-brain natriuretic peptide (NT-proBNP) were measured in the central clinical laboratory of our hospital. The urea clearance index and weekly creatinine clearance (Ccr) were determined using standard methods and indexed by body surface area to assess dialysis adequacy. Residual renal function was calculated as the average of 24-h urinary urea and Ccr values [12].

All samples (serum after centrifugation) were stored in a deep freezer at −80°C. sST2 levels in both the serum and dialysate were determined using specific enzyme-linked immunosorbent assay (ELISA) kits according to the manufacturer’s recommendations (R&D Systems, Inc. Minneapolis, MN, USA). The results are expressed as ng/mL. The lower limit of detection for sST2 was 31.3 pg/mL.

The clinical outcome measure was the occurrence of CV events. According to the Framingham Heart Study, CV events are defined as those related to coronary heart disease (coronary death, myocardial infarction, coronary insufficiency, and angina) and peripheral artery disease (intermittent claudication), cerebrovascular events (ischemic stroke, hemorrhagic stroke, and transient ischemic attack), and heart failure [13]. All patients were followed until December 2023, censored at death or until the termination of PD treatment, whichever occurred first.

Continuous variables are expressed as the mean ± standard deviation or median (interquartile range, IQR) and were compared using Student’s t test or the Mann-Whitney U test. Discrete variables are expressed as percentages and were compared using the χ² test or Fisher’s exact test. Spearman’s rank correlation analysis was performed to analyze the associations between serum sST2 level and other variables. To determine variables independently associated with sST2, a stepwise multiple linear regression analysis was performed. Cox proportional regression analyses were performed to determine the independent prognostic values of sST2 and NT-proBNP for CV events. Cumulative survival was estimated using Kaplan-Meier survival curves and compared using log-rank tests. Receiver operating characteristic curves were generated to determine the prognostic abilities of sST2 and NT-proBNP and identify the optimal cutoff values. All data analyses were performed using SPSS version 26.0 (SPSS Inc. Chicago, IL, USA), and statistical significance was set at p < 0.05.

Figure 1 shows the flowchart of the study. 164 patients underwent serum sST2 measurements and received PD catheter implantation. One patient withdrew from PD for catheter migration, and 26 patients were excluded for the following reasons: peritonitis (5), unqualified samples (8), lost to follow-up (9) and 4 patients could not tolerate 2 L peritoneal fill volume during the PET because of small abdominal cavity.

In total, the study enrolled 137 participants with a median age of 47 years (IQR 36–54); 59.82% of them were men. Primary glomerulonephritis was the most common cause of ESRD in the study population. About 10.22% of the patients had CV disease at baseline. All the participants were prescribed peritoneal dialysate containing 1.5% or 2.5% dextrose. Detailed information on the patients’ baseline characteristics is presented in Table 1.

The median sST2 levels in serum and dialysate were 11.90 ng/mL (IQR 7.29–18.53) and 0.27 ng/mL (IQR 0.14–0.52), respectively. Univariate analysis showed a significant correlation between the sST2 level in serum and that in dialysate (r = 0.338, p = 0.000), but the correlation did not remain after multivariate adjustment. Male sex (r = 0.269, p = 0.001), age (r = 0.180, p = 0.024), IL-6 (r = 0.228, p = 0.006), and β2-microglobulin (r = 0.227, p = 0.006) were found to be significantly correlated with serum sST2 level (Table 2).

During a median follow-up of 17.28 months (IQR 9.25–21.43), 49 (35.76%) patients experienced CV events, which consisted of coronary heart disease (4 patients, 2.92%), heart failure (37 patients, 27.00%), cerebrovascular events (4 patients, 2.92%), and peripheral artery disease (4 patients, 2.92%).

Multivariate Cox regression analysis revealed that both sST2 and serum phosphate levels were significantly associated with the risk of CV events when using continuous approach (Table 3). When using a dichotomized approach (both median and calculated cutoff values), the higher sST2 group had 2.980- and 3.048-fold increased risks of CV events when compared to the lower sST2 group (Table 4). However, only when patients were dichotomized based on the calculated cutoff values, the higher NT-proBNP group had a 2.530-fold increased risk for CV events compared to the lower NT-proBNP group (Table 4). The Kaplan-Meier survival curves revealed that patients with higher sST2 levels showed significantly worse outcomes than those with lower sST2 levels (log-rank test; p < 0.001). Similarly, patients with elevated NT-proBNP levels had a notably higher risk of CV events (log-rank test; p < 0.001) (Fig. 2a, b).

Furthermore, serum sST2 and NT-proBNP levels simultaneously above the calculated cutoff values were associated with a higher risk (hazard ratio = 3.398, 95% CI 1.813–6.367, p < 0.001) of CV events than higher sST2 or NT-proBNP levels alone (Table 4). Kaplan-Meier survival curves revealed that patients with two markers elevated had a markedly higher risk of CV events (log-rank test; p < 0.001) (Fig. 2c).

Receiver operating characteristic curves showed that serum sST2 and NT-proBNP levels could predict CV events; the area under the curve (AUC) values were 0.680 (95% CI 0.590–0.769, p = 0.001) and 0.681 (95% CI 0.590–0.773, p < 0.001), respectively. The calculated optimal cutoff values were identified: 13.17 ng/mL for sST2 and 1,580.50 pg/mL for NT-proBNP. The difference between the AUCs was not statistically significant (p > 0.05) (Fig. 3).

The major finding of this study was that baseline serum sST2 level was an independent predictor of the risk of CV events in patients receiving incident PD. To our knowledge, this is the first study to investigate the role of sST2 on CV outcomes in an incident PD population. Furthermore, we found that the combination of sST2 and NT-proBNP improved the prediction accuracy for these patients.

sST2 has been confirmed as a potential marker for monitoring cardiac fibrosis and remodeling in patients with chronic kidney disease [5, 14]. sST2 has been studied mainly in patients undergoing HD, where it has shown a significant predictive value for all-cause and CV mortality [9, 10, 15, 16]. However, relatively few clinical studies have reported the predictive value of sST2 in ESRD patients undergoing PD. Only one study by Choi et al. [11] prospectively investigated the predictive ability of sST2 to clinical outcomes in patients undergoing prevalent PD with a median dialysis vintage of 30 months. The median sST2 level was 70.9 ng/mL, which was higher than that of our study. As dialysis vintage increases, volume management may become more difficult, and PD patients may be chronically in a hypervolemia status. Meanwhile, other CV disease risk factors are also increasing. Thus, sST2 levels might increase with prolonged dialysis vintage. Choi et al. [11] compared the predictive value of sST2 and sLR11 for all-cause mortality and major adverse cardiac and cerebrovascular events. However, we evaluated the predictive value of baseline sST2 for CV events in incident PD patients and compared it with NT-proBNP, a more widely used biomarker in clinical practice. Measuring baseline sST2 levels may help identify individuals at high risk of CV events earlier and initiate prompt management, which may improve their clinical outcomes.

NT-proBNP is widely used for diagnosing heart failure and predicting CV events [17], and its application in patients with renal impairment has been reported [15]. However, it is partially cleared by the kidneys or dialysis [18]. In clinical settings, the interpretation of serum biomarker concentrations affected by kidney function or dialysis is complex and thus not easily applicable. In this study, we found that serum sST2 levels were independent of renal function, urea clearance index, and Ccr in the studied population, which is consistent with previous reports in clinical settings [5, 6, 19, 20]. We also confirmed that, only when using calculated cutoff values, NT-proBNP showed predictive power for predicting CV events. Meanwhile, regardless of using the continuous or dichotomized (both median and calculated cutoff values) approach, sST2 showed better predictive power than NT-proBNP. We speculate that sST2 might be a more stable marker for the early risk stratification of potential CV events in patients receiving incident PD.

Furthermore, we found that the predictive value of the combination of serum sST2 and NT-proBNP for CV events in incident PD patients was higher than that of sST2 or NT-proBNP alone, which is consistent with the findings of studies on maintenance HD [10]. During a median follow-up of 17.28 months, the incidence of CV events in our study reached 35.76%, reaffirming the high incidence of CV events in patients on dialysis. Cardiac disease-related mortality remains high [18], responsible for 36% of all-cause mortality in patients receiving PD, and CV events are common among this population (25–51% by county) [21, 22]. Therefore, it is essential to identify high-risk individuals among these patients and initiate appropriate intervention strategies to improve their survival. The combined use of sST2 and NT-proBNP may improve the predictive value of CV events in this population.

This study had some limitations. First, it was limited to a single center, we studied a relatively small cohort, and our findings may not be generalizable to other ethnicities. Second, echocardiography, which assesses cardiac structure and function, was not routinely performed. Third, medications that affect the CV risk were not documented and may have influenced clinical outcomes. Further research in a larger population is needed to elucidate the predictive value of sST2 for adverse CV events in patients receiving incident PD.

Baseline serum sST2 level is an independent predictor of the risk of CV events in patients undergoing incident PD, and its combination with NT-proBNP can provide a more accurate predictive value for these patients. These findings may contribute to the early identification of individuals at high risk of CV events and early intervention, which may improve the clinical outcomes of patients receiving PD.

The authors thank all the participating patients and staff of the PD Center, Department of Nephrology, West China Hospital of Sichuan University.

The study was approved by the Ethics Committee of the West China Hospital of Sichuan University, approval No. 2019(1015). It was conducted in accordance with the ethical standards set forth in the Declaration of Helsinki and its later amendments. Written informed consent was obtained from all participants.

The authors have no conflicts of interest to declare.

This research was supported by Sichuan Science and Technology Program (No. 2022YFS0148).

Research idea and study design: Zi Li; sample collection and data acquisition: Xiaofang Wu, Zhiyun Zang, Yunyun Zhang, Qijiang Xu, Li Pu, Xiaoxiao Xia, and Niya Ma; statistical analysis and interpretation: Yunyun Zhang, Xiaofang Wu, and Li Pu; data supplementation and revision of the manuscript: Qijiang Xu; supervision or mentorship: Zi Li. All authors read and approved the final manuscript.

Yunyun Zhang, Qijiang Xu and Xiaofang Wu contributed equally to this work.

The datasets generated or analyzed during this study cannot be publicly shared due to ethical considerations. Access to the data can be requested from Zi Li upon reasonable request.