Myocardial work in Chagas disease: comparative analysis with conventional parameters in different cardiac clinical stages

Chagas disease (CD) is a neglected chronic infectious disease and represents an important cause of morbidity and mortality in endemic areas of Latin America. Chronic Chagas cardiomyopathy (CCC) is the most severe manifestation of the disease and the leading cause of death in these patients. Myocardial deformation, measured by speckle-tracking echocardiography (STE), has been proven to detect subclinical disease and has prognostic value in CCC, but it is load dependent. We aimed to characterize myocardial work (MW) as new, less load-dependent, echocardiographic variables in patients with different clinical stages of CCC and to assess its correlation with left ventricular (LV) ejection fraction (EF) and accuracy to detect LV dysfunction, compared with routinely used parameters. Fifty consecutive patients with CD were included. Clinical assessment, NT-proBNP, and a comprehensive 2D echocardiogram were performed. Patients were divided into four groups, depending on disease stages: Stage A-indeterminate form (<i>n</i> = 9), defined as asymptomatic patients without abnormalities on physical examination, electrocardiogram, or echocardiography; Stage B1, chronic Chagas cardiomyopathy (CCC) with preserved ejection fraction (<i>n</i> = 18); Stage B2, CCC with left ventricular global dysfunction (LVEF <55%, <i>n</i> = 13); and Stage C, CCC with LVEF <55% and clinical heart failure (<i>n</i> = 10). Patients in Groups B2 and C had lower values of global work index (GWI), global constructive work (GCW), and global work efficiency (GWE) than those in patients in Groups A and B1 (<i>p</i> < 0.001). Values of GWI, GCW, and GWE in Groups A and B1 were normal compared with normative data. Among the studied parameters, GCW showed the strongest correlation with LVEF (<i>r</i> = 0.854), followed by GWI (<i>r</i> = 0.848) and global longitudinal strain (GLS) (<i>r</i> = 0.810), while moderate correlations were observed for GWE (<i>r</i> = 0.578) and NT-proBNP (<i>r</i> = 0.634). Pairwise comparisons of correlation coefficients using Steiger's <i>Z</i> test revealed that GCW had a significantly stronger correlation with LVEF than NT-proBNP (<i>p</i> < 0.001), but no significant difference was observed when compared with GLS or GWI (<i>p</i> > 0.05). For the detection of LV global dysfunction as a categorical variable, defined as LVEF <55%, the best accuracy was observed for GCW (AUC = 0.976; 95% CI: 0.927-1.000; optimal cutoff 1,699 mmHg%), followed by GWI (AUC = 0.965; 95% CI: 0.905-1.000; optimal cutoff 1,282 mmHg%) and GLS (AUC = 0.938; 95% CI: 0.856-0.994; optimal cutoff -15.5%). Comparisons between AUCs were performed using DeLong's test for paired ROC curves. GWI and GCW showed good correlation with NT-proBNP levels (<i>r</i> = -0.567 and <i>r</i> = -0.552, respectively, both <i>p</i> < 0.001) in our cohort, similar to LVEF by the Simpson method (<i>r</i> = -0.594, <i>p</i> < 0.001). Myocardial work parameters may be promising markers of early LV systolic dysfunction in Chagas cardiomyopathy and may serve as useful prognostic markers for patients with CCC.