Soluble ST2 Is a Sensitive and Specific Biomarker for Fulminant Myocarditis

Abstract

Background The aim of the study was to identify biomarkers that can facilitate early diagnosis and treatment of fulminant myocarditis (FM) in order to reduce mortality. Methods and Results First, the expression profiles of circulating cytokines were determined in the plasma samples from 4 patients with FM and 4 controls using human cytokine arrays. The results showed that 39 cytokines from patients with FM were changed at admission. Among them, 8 cytokines returned to normal levels at discharge, including soluble ST2 (sST2), which showed the most marked dynamic changes from disease onset to resolution. Then, in a cohort of 76 patients with FM, 57 patients with acute hemodynamic dysfunction attributable to other causes, and 56 patients with non-FM, receiver operating characteristic curve analyses suggested that plasma sST2 level was able to differentiate FM from non-FM or other FM-unrelated acute heart failure more robustly N-terminal pro-B-type natriuretic peptide or cardiac troponin I. Moreover, longitudinal analysis of plasma sST2 was performed in 10 patients with FM during hospitalization and 16 patients with FM during follow-up. Finally, the diagnostic value was validated in an additional 26 patients with acute onset of unstable hemodynamics. The cutoff value of plasma sST2 for optimal diagnosis of FM was established at 58.39 ng/mL, where a sensitivity of 85.7% and specificity of 94.7% were achieved. Conclusions Elevated sST2 level was associated with mechanical stress or inflammation. Especially, sST2 might be used as a potential biomarker for the rapid diagnosis of FM, which was characterized by strong mechanical stretch stimulation and severe inflammatory response. Registration URL: https://www.clinicaltrials.gov; Unique identifier: NCT03268642.

Keywords: adult; biomarker; fulminant myocarditis; inflammatory; soluble ST2 (sST2).

Figure 1. Overall study design and the scheme of the validation cohort.

Detailed population information and the corresponding objectives are shown in the Supplemental Tables. AHF indicates acute heart failure; AMI, acute myocardial infarction; DCM, dilated cardiomyopathy; FM, fulminant myocarditis; NFM, nonfulminant myocarditis; and VHD, valvular heart diseases.

Figure 2. Levels of 122 human cytokines in the patients with fulminant myocarditis (FM) and controls.

A, Volcano plot of the expression of human cytokines in 4 controls and 4 patients with FM at admission. LIMMA test was used to calculate the significance. The red plots represent differentially expressed proteins with an adjusted P value <0.05, whereas the black plots represent insignificant changes. The horizontal and vertical dotted lines in volcano plots represented the threshold value for the significance used to define upregulation or downregulation of cytokines was a fold change >2 (or <0.5), as well as with an adjusted P value of <0.05. B, Expression heatmap of cytokines that significantly changed in 4 patients with FM at admission compared with 4 controls, which correspond to the red plots in Figure 2A. The normalized levels of cytokines log2 (fold change) were indicated by a different color code (right). C, Volcano plot of the expression of human cytokines in 4 patients with FM at admission and 4 patients with FM at discharge. The red plots represent differentially expressed proteins with a P<0.05, whereas the black plots represent insignificant changes. The horizontal and vertical dotted lines in volcano plots represent the threshold value for the significance used to define upregulation or downregulation of cytokines was a fold change >1, as well as with a P value of <0.05. D, Expression heatmap of cytokines that significantly changed in 4 patients with FM at discharge vs 4 patients with FM at admission. Those cytokines corresponded to the red plots in Figure 2C. A different color code (right) represented normalized level of cytokines log2 (fold change). CRP indicates C‐reactive protein; DKK, Dickkopf protein; DPPIV, dipeptidyl peptidase IV; IFN‐γ, interferon γ; IL, interleukin; MIF, macrophage migration inhibitory factor; OPN, osteopontin; PAI‐1, plasminogen activator inhibitor 1; PDGF, platelet‐derived growth factor; TGFb, transforming growth factor β; TNFb, tumor necrosis factor β; uPAR, urokinase plasminogen activator receptor; and VEFG‐C, vascular endothelial growth factor C.

Figure 3. Diagnostic performance of plasma soluble ST2 (sST2) for the detection of fulminant myocarditis (FM).

A, Circulating concentrations of sST2 in 76 patients with 8 control individuals (data are presented as medians and quartile 1 to quartile 3 [Q1–Q3], and Mann‐Whitney test was used to elevate the differences, **P<0.05). B, Circulating concentrations of sST2 in 76 patients with FM and 57 patients with acute heart failure (AHF; data are presented as medians and Q1 to Q3, and Kruskal‐Wallis test was used to elevate the differences, **P<0.05). C, Receiver operating characteristic (ROC) curves of plasma sST2, NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide), and cardiac troponin I (cTnI) in 76 patients with FM and 57 patients with AHF (the Delong test was used to calculate significance). D, Circulating concentrations of sST2 in 56 patients with nonfulminant myocarditis (NFM) and 76 patients with FM (data are presented as medians and Q1 to Q3, and Mann‐Whitney test was used to elevate the differences, **P<0.05). E, ROC curves of plasma sST2, NT‐proBNP, and cTnI in 56 patients with NFM and 76 patients with FM (the Delong test was used to calculate significance). AMI indicates acute myocardial infarction; AUROC, area under the receiver operating characteristic; DCM, dilated cardiomyopathy; and VHD, valvular heart disease.

Figure 4. Correlations between plasma soluble ST2 (sST2) level and heart damage.

Correlation analysis of plasma sST2 with high‐sensitivity C‐reactive protein (hs‐CRP) (A), ejection fraction (EF) (B), NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) (C), and cardiac troponin I (cTnI) (D) value in 76 patients with fulminant myocarditis (FM; Spearman correlation and linear regression analysis were used to calculate significance). b indicates regression coefficient.

Figure 5. Plasma levels of soluble ST2 (sST2) in patients with fulminant myocarditis (FM) during hospitalization.

A, Circulating concentrations of sST2 in 10 patients with FM during hospitalization. Correlation analysis of plasma sST2 with NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) (B), cardiac troponin I (cTnI) (C), and ejection fraction (EF) (D) values during hospitalization (Spearman correlation and linear regression analysis were used to calculate significance). b indicates regression coefficient.

Figure 6. Receiver operating characteristic (ROC) curves of circulating soluble ST2 (sST2) in the validation cohort.

A, Patients with fulminant myocarditis (FM) vs all other patients compared with NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide), cardiac troponin I (cTnI), and sST2. B, Patients with FM vs all other patients or those with acute myocardial infarction (AMI) using sST2. AUROC indicates area under the receiver operating characteristic.