The association of myocardial strain with cardiac magnetic resonance and clinical outcomes in patients with acute myocarditis

Abstract

Introduction: The role of myocardial strain in risk prediction for acute myocarditis (AMC) patients, measured by cardiac magnetic resonance (CMR), deserves further investigation. Our objective was to evaluate the association between myocardial strain measured by CMR and clinical events in AMC patients.

Material and methods: This was a prospective single-center study of patients with AMC. We included 100 patients with AMC with CMR confirmation. The primary outcome was the composite of all-cause mortality, heart failure and AMC recurrence in 24 months. A subgroup analysis was performed on a sample of 36 patients who underwent a second CMR between 6 and 18 months. The association between strain measures and clinical events or an increase in left ventricular ejection fraction (LVEF) was explored using Cox regression analysis. Global peak radial, circumferential and longitudinal strain in the left and right ventricles was assessed. ROC curve analysis was performed to identify cutoff points for clinical event prediction.

Results: The mean follow-up was 18.7 ± 2.3 months, and the composite primary outcome occurred in 26 patients. The median LVEF at CMR at baseline was 57.5% (14.6%). LV radial strain (HR = 0.918, 95% CI: 0.858-0.982, p = 0.012), LV circumferential strain (HR = 1.177, 95% CI: 1.046-1.325, p = 0.007) and LV longitudinal strain (HR = 1.173, 95% CI: 1.031-1.334, p = 0.015) were independently associated with clinical event occurrence. The areas under the ROC curve for clinical event prediction were 0.80, 0.79 and 0.80 for LV radial, circumferential, and longitudinal strain, respectively. LV longitudinal strain was independently correlated with prognosis (HR = 1.282, CI 95%: 1.022-1.524, p = 0.007), even when analyzed together with ejection fraction and delayed enhancement. LV and right ventricle (RV) strain were not associated with an increase in LVEF. Finally, when the initial CMR findings were compared with the follow-up CMR findings, improvements in the measures of LV and RV myocardial strain were observed.

Conclusion: Measurement of myocardial strain by CMR can provide prognostic information on AMC patients. LV radial, circumferential and longitudinal strain were associated with long-term clinical events in these patients.

Keywords: diagnosis; magnetic resoance imaging; myocardial strain; myocarditis; pericarditis.

Figure 1

Study design and flowchart. CMR, cardiac magnetic resonance; AMC, acute myocarditis.

Figure 2

In (A,B) cine short axis in medial cut in diastole (A) and systole (B). (C,D) Bullseye radial strain and circumferential strain, respectively. The endocardial and epicardial were delimited manually in diastole in short-axis cine images in medial location of the left ventricle with automatic propagation to the entire cardiac cycle. Circumferential strain in the short axis cine in medial location of the LV in diastole (E) and systole (F) with contractile defect in the inferior, inferolateral and attenuated contraction in the septal and anterior region, corroborating with the bullseye (C,D). In (G) T2-weighted image in medial section of the LV showing edema (white arrows) and quantification (blue areas). Image (I) showed delayed short-axis enhancement with myocardial fibrosis (white arrows) and quantification (yellow areas).

Figure 3

ROC curve and boxplot diagrams relating combined events with (A) LV radial strain; (B) LV circumferential strain; and (C) LV longitudinal strain related. LV, left ventricle; AUC, area under curve; sens, sensitivity; spec, specificity.

Figure 4

Boxplot diagrams presenting differences between myocardial strain in initial vs. evolutionary CMR: (A) LV radial strain; (B) LV circumferential strain; (C) LV longitudinal strain related and; (D) RV longitudinal strain. CMR, cardiac magnetic resonance; LV, left ventricle; RV, right ventricle.