Predictive value of soluble suppression of tumorigenicity 2 in atrial fibrillation: a systematic review and meta-analysis

Purpose Atrial fibrosis is the main pathological basis for the pathogenesis and progression of atrial fibrillation (AF). Soluble suppression of tumorigenicity 2 (sST2) is involved in fibrosis. Recent studies have explored its predictive value in AF outcomes. We performed this study to assess whether sST2 is an independent biomarker of AF outcomes and explore the potential mechanism. Methods PubMed, Web of Science, EMBASE, and Cochrane Library databases were searched systematically from inception through July 1, 2023, to identify relevant studies. Outcomes of interest included occurrence, recurrence, and major adverse cardiac events (MACEs) of AF. This meta-analysis was reported following the criteria outlined in PRISMA 2020, and the protocol was registered in PROSPERO (number: CRD42023459789). All statistical analyses were performed using the STATA version 16. Result Twenty four studies with 14,755 patients were included in the meta-analysis. The meta-analyses found that sST2 was significantly associated with the risk of occurrence [HR:1.04, 95% CI: 1.02–1.07, P < 0.01; I2 = 67.8%], recurrence [HR:1.09, 95% CI: 1.02–1.16, P < 0.01; I2 = 89.5%], and MACEs (HR:1.60, 95% CI: 1.13–2.27, P < 0.01; I2 = 82.0%) of AF. Furthermore, patients with AF showed higher sST2 than controls without AF (SMD: 0.41, 95% CI: 0.27–0.54, P < 0.01; I2 = 0%), and AF patients with recurrence after catheter ablation (CA) showed significantly higher sST2 than those without recurrence (SMD: 0.81, 95% CI: 0.33–1.28, P < 0.01; I2 = 83.9%). Sensitivity analyses showed that the outcomes were stable. Conclusions Higher sST2 was association with an increased risk of occurrence, recurrence, and MACEs of AF. Assessing sST2 can be used as a potential screening method to predict AF outcomes. Systematic Review Registration PROSPERO (CRD42023459789).


Introduction
Atrial fibrillation (AF) is one of the most common clinical arrhythmias, affecting more than 46.3 million individuals worldwide (1).The prevalence of AF is expected to double in the next 30 years to more than 17 million in Europe alone.AF leads to peripheral embolism, stroke, heart failure (HF), and is associated with high mortality and hospitalization rates (2).In addition, AF recurrence is also a challenging problem, with a recurrence rate of up to 30% (3).The pathophysiology of AF is complex and is thought to involve pro-inflammatory responses leading to electrophysiological remodeling, which in turn leads to atrial fibrosis and structural remodeling.The end result is to provide an arrhythmogenic substrate for AF triggers (4)(5)(6).
As with other diseases, blood markers have been used for the purpose of risk stratification for AF (7).The suppression of tumorigenicity 2 (ST2) is a member of the interleukin-1 (IL-1) receptor family, which exists in two forms: transmembrane receptor (ST2l) and soluble decoy receptor (sST2) (8).sST2 is relates to markers of hemodynamic load, released from the myocardium and vascular endothelial cells in response to pressure or volume overload, and is involved in fibrosis and remodeling through pathways related to inflammation (8,9).sST2, mainly a well-known HF biomarker, and is also associated with worsening outcomes after myocardial infarction (MI) (10,11).Recent studies have identified the predictive value of sST2 in AF (12-14).However, these studies are small and contradictory.Therefore, we aimed to assess whether sST2 is an independent biomarker of AF outcomes and explore the potential mechanism.

Search strategy
Two researchers performed a systematic literature search using four electronic databases (PubMed, Web of Science, EMBASE, and Cochrane Library) with MESH terms and keywords (("Atrial fibrillation" OR "AF") AND ("Biomarkers" OR "Soluble Suppression of Tumorigenicity 2" OR "Soluble ST2" OR "sST2" OR "ST2").We also conducted a hand-searching of relevant articles.The disagreement was resolved by consulting a senior reviewer (Dazhuo Shi).

Literature inclusion and exclusion criteria
The inclusion criteria were: (1) study design was observational study (included prospective cohort, retrospective cohort, casecontrol, and cross-sectional study); (2) target population was AF patients; (3) there were measured sST2 at least two groups in one study; (4) outcomes of interest included occurrence, recurrence, and major adverse cardiac events (MACEs) of AF.The MACEs, which was defined as a composite outcome of fatal or non-fatal cardiovascular events, such as death, MI, HF, stroke, rehospitalization, and revascularization.

Data extraction
The following data were extracted: (1) information on the publication: first author's name, publication year, location; (2) demographic characteristics: sample size, age, gender; (3) study details: study design, follow-up period, measurement methods of sST2, data on the diagnostic analysis (definition of the control group, sample size, mean ± standard deviatio (SDs) or median interquartile ranges (IQR) values, the optimal cut-off value, area under the curve (AUC) for the receiver operating characteristic curve (ROC), sensitivity, and specificity), and data on the prognostic analysis (clinical outcomes, unadjusted and/or multivariable-adjusted HRs/ORs, 95% CIs, the optimal cut-off value, AUC for the ROC, sensitivity, and specificity).

Bias assessment
The Newcastle-Ottawa Scale (NOS) (15) was used to assessed each study quality.NOS focused on three major aspects: participant selection (0-4 stars), comparability (0-2 stars), and exposure (0-3 stars).Studies were regarded as moderate-to-high quality with the total score ≥6, and <6 for low quality.
The Grading of Recommendations Assessment, Development and Evaluation (GRADE) (16) approach was used to assess the quality of outcome evidence.The quality were categorized into four levels (high, moderate, low, or very low).This meta-analysis only included observational studies, which start with a "low quality", and other factors may then upgrade or downgrade the quality level.

Statistical analysis
All data of sST2 were pooled analysis by means ± SDs or HRs/ ORs.The Cochrane Q-test and the I 2 -value were used to assess the statistical heterogeneity, where P < 0.1 or I 2 > 50% suggested significant heterogeneity.A random-effect model were selected for this meta-analysis, considering the potential heterogeneity across studies.The publication bias was evaluated by employing the funnel plots and Egger's test, where a P-value higher than 0.1 indicated no significant publication bias.For sensitivity analysis, omitted one study at a time to assess the robustness.STATA version 16 was used for all statistical analyses.

Study search
The flowchart of literature screening was shown in Figure 1.The initial database search identified 292 records, 189 of which were duplicates.According to the analysis of titles and abstracts, 65 studies were excluded, and 38 were included.

Study quality
The mean NOS scores was 7.54 (range 6-9), indicating moderate to high quality.The details of the quality assessment were shown in Supplementary Table S1.
The GRADE grade showed that 2 evidences were mediate, 1 was low, and 2 were very low.The results of the GRADE assessment were shown in Supplementary Table S2.

Predictive value of sST2 in occurrence of AF
Table 2 diaplayed the characteristics of the included studies for predictive value of sST2 in occurrence of AF.A total of 12 studies were analyzed with 11,961 participants (including 1,707 AF patients and 10,254 controls).The pooled analysis of 10 studies (12,21,24,25,27,28,32,34,36,37) as a continuous variable indicated that sST2 was associated with the risk of AF occurrence (HR:1.04,95% CI: 1.02-1.07,P < 0.01; I 2 = 67.8% Figure 2).Sensitivity analyses showed that the outcomes were stable (Supplementary Figure S2A).The funnel plot and Egger's test revealed publication bias in the results (Egger's test, P = 0.006, Supplementary Figures S2B,C).

Predictive value of sST2 in MACEs following AF
Table 4 diaplayed the characteristics of the included studies for predictive value of sST2 in MACEs following AF.Five studies (17, 18, 23, 30, 37) with 1,614 AF patients examined the relationship between sST2 and the risk of MACEs following AF.The pooled analysis of the estimates as a continuous variable indicated that sST2 was significantly associated with the risk of MACEs (HR:1.60,95% CI: 1.13-2.27,P < 0.01; I 2 = 82.0%Figure 6).Sensitivity analyses showed that the outcomes were stable (Supplementary Figure S6A).The funnel plot and Egger's test revealed publication bias in the results (Egger's test, P = 0.004, Supplementary Figures S6B,C).

Discussion
Our study found that sST2 was associated with the risk of occurrence, recurrence, and MACEs of AF.For every 1 unit increase in sST2, the risk of occurrence, recurrence, and MACEs increased by 4%, 9%, and 60%, respectively.Furthermore, patients with AF showed higher sST2 than controls without AF, and AF patients with recurrence showed significantly higher sST2 than those without recurrence.Sensitivity analyses showed that the outcomes were stable.
AF is partially explained by atrial remodeling determined by myocardial hypertension, dilatation, infiltration, inflammation and fibrosis (38,39).Atrial remodeling acts in concert with the arrhythmia itself to enhance atrial vulnerability to AF, and might be both cause and consequence of AF (40,41).Atrial fibrosis is a defining structural feature of atrial remodeling, a process that includes extracellular matrix (ECM) and collagen deposition accumulation and is thought to be initiated and regulated by immune cells (42,43).Identifying biomarkers associated with atrial fibrosis will increase our understanding of the pathophysiological mechanisms of AF and could be used to develop pharmacological pathways for the prevention of AF, in addition, adding these biomarkers to the risk scale may lead to more accurate predictions of AF risk (44).sST2, as a biomarker closely related to fibrosis and inflammation, plays an important role in the pathogenesis and progression of AF, and also has certain value in predicting the occurrence, progression, recurrence and prognosis of AF.
Studies have found that sST2 reflect fibrosis may be related to the inhibition of IL-33/ST2l pathway (45,46,9).IL-33 is the functional ligand of ST2l, and the IL-33/ST2l signaling pathway has been shown plays a role in anti-myocardial fibrosis and cardiomyocyte hypertrophy (47,48).sST2 is a receptor for IL-1 and competitively binds to IL-33 to inhibit the protective effect of IL-33/ST2l on myocardium (49).When myocardium induced by pressure or volume overload produce a large amount of sST2, Forest plots show the relationship between sST2 (continuous variables) and the risk of AF recurrence after CA.
Forest plots show the difference in sST2 values between patients with and without AF recurrence after CA.
Chen et al. 10.3389/fcvm.2023.1308166 Frontiers in Cardiovascular Medicine high concentrations of sST2 prevent IL-33/ST2l effects and may therefore lead to atrial fibrosis, which is associated with AF via atrial structural remodeling (49,50).Early atrial dilation in AF patients can lead to physiological stretching of the atrium, causes myofibroblasts to release IL-33, which binds ST2l to myocardial cell membranes and promotes cell integrity and survival.However, during long-term lesions, local and adjacent cells can increase the release of the IL-33 decoy sST2, thereby blocking IL-33/ST2l binding and promoting tissue fibrosis (51, 52).Notably, the novel aspect of IL-33/ST2l signaling mediating cardiac fibrosis represents some novel biomolecular targets for prevention and treatment of AF.In addition, sST2 can also cause myocardial damage by promoting oxidative stress and inflammation.sST2 affects mitochondrial fusion of human cardiac fibroblasts and increases oxidative stress production and inflammatory marker secretion by reducing mitofusins-1 (MFN-1) expression (53).
Fibrosis is associated with impaired cellular coupling, enhanced heterogeneity of intra-atrial conduction, and dispersion of refractory periods, which provide substrates for sustained reentry to drive AF (54,43).We think that sST2 can reflect fibrosis as well as the degree of fibrosis.Studies have found that the expression of sST2 in patients with persistent AF is higher than that in patients with paroxysmal AF, which may explain that sST2 reflects the degree of fibrosis and the progression of AF (55,56).Recurrence after AF ablation is directly related to atrial fibrosis and is positively correlated (57).Our study found that sST2 level is an independent predictor of AF recurrence after CA, and every 1 unit increase in sST2, the risk of recurrence increased by 9%.sST2 is also directly associated with MACEs of AF patients.Increased abnormal hemodynamic load leads to atrial dilation, which is a wellknown cause of the development of AF, and may also stimulate sST2 and BNP secretion.In addition, sST2 levels are elevated during AF, possibly because the heart rate and atrial pressure in patients with AF are higher than normal, thereby increasing cardiac burden and increasing the risk of MACEs (58, 59).IL-33/ST2l signaling is thought to play an important role in regulating the myocardial response of stretched cardiac fibroblasts and cardiomyocytes to biomechanical overload (60).Loss of IL-33/ST2l signaling leads to hypertrophy of cardiomyocytes, fibrosis, and deterioration of left ventricular function, further aggravating ventricular myocardial remodeling, and increasing the risk of death from HF.Our results show that sST2 has a stronger correlation with the risk of MACEs (HR = 1.60) than the occurrence (HR = 1.04) and recurrence (HR = 1.09) of AF, indicating that sST2 has notable prognostic performance, but low diagnostic performance.We believe that sST2, as a new biomarker of inflammation, fibrosis and cardiac stress, may have a more direct correlation associated with cardiac damage and MACEs.And the prognostic information provided by sST2 is in addition to that provided by other well stablished biomarkers, such as BNP and troponins.However, the mechanism of AF occurrence and recurrence is complex, and sST2, although a marker of fibrosis, is not specific to atrial fibrosis, thus showing a weak clinical correlation.As a new biomarker, galectin-3 (Gal-3) and sST2 have been found to play an important role in fibrosis.Gal-3 can specifically bind to ECM and fibroblasts, thereby promoting fibroblast proliferation, inflammatory cell infiltration, and remodeling (57).However, the current research is not very clear about the mechanism of Gal-3 and sST2 on atrial fibrosis, and future research should be performed to reveal their role and mechanism, and provide more new basis for the prevention and treatment of atrial fibrosis.

Strengths and limitations
Our study performed a detailed meta-analysis and mechanism analysis.First, this is the first meta-analysis to summarize the predictive value of sST2 in AF.Second, this study covers a large sample size from different countries, so the results are relatively stable and reliable.Third, the pooled analysis was based on the most adequately adjusted HRs, so the finding may be independent of potential confounders.Fourth, sensitivity analyses did not significantly affect the results, indicating the results were credible.
There were some limitations in our study.First, the metaanalysis only included observational studies, it carries inherent study design limitations.Second, the heterogeneity in metaanalyses were significant.We used sensitivity analysis and publication bias to explore the source of heterogeneity.Third, some residual factors may affect the results.Fourth, we did not find some suitable case reports to prove our study.

Conclusions
Higher sST2 was association with an increased risk of occurrence, recurrence, and MACEs in AF.Assessing sST2 can be used as a potential screening method to predict AF outcomes.Further well-designed cohort studies and randomized clinical trials are warranted to confirm this finding.

FIGURE 3 Forest
FIGURE 3Forest plots show the difference in sST2 values between patients with and without AF.

TABLE 1
Baseline characteristics of the 24 selected research.

TABLE 2
Characteristics of the 12 studies for predictive value of sST2 in occurrence of AF.
3.4.2Predictive value of sST2 in AF recurrence after CA

TABLE 3
Characteristics of the 8 studies for prognostic analysis of AF recurrence after CA.

TABLE 4
Characteristics of the 5 studies for prognostic analysis of MACEs after AF.