Cardiovascular magnetic resonance-derived metrics as diagnostic markers for left ventricular outflow tract obstruction in hypertrophic cardiomyopathy

Abstract Background Left ventricular outflow tract obstruction (LVOTO) may be present in up to 70% of hypertrophic cardiomyopathy (HCM) patients and is related to symptoms and worse outcomes. Doppler echocardiography is currently the gold standard method for diagnosing LVOTO, as cardiovascular magnetic resonance (CMR) does not reliably evaluate outflow hemodynamics. Previous studies have aimed to detect LVOTO indirectly with CMR by capturing the underlying pathophysiological and spatial alterations. However, there is a need for easy-to-measure, reliable metrics. Purpose This study aimed to assess the utility of CMR-derived metrics as diagnostic indicators for LVOTO in HCM patients, using Doppler echocardiography as the reference method. Methods In this single-center retrospective cross-sectional study, adult HCM patients with transthoracic echocardiography (TTE) and CMR within a 6-month time window were enrolled. Doppler LVOT gradient was measured both at rest and under provocative maneuvers. LVOTO was defined as a gradient ≥30 mmHg. In CMR, all measurements were performed in the 3-chamber balanced steady-state free precession cine images (Figure 1). Results A total of 448 HCM patients (median age 55 (45-62) years, 301 (67%) males) were included. LVOTO (either at rest or under provocation) was present in 186 patients (42%) of which 65 patients (35%) only had provocable LVOTO. The final multivariable logistic regression model for discriminating LVOTO included the following CMR metrics: 1) the minimum distance between the mitral leaflet tip and the interventricular septum (MV-IVS) indexed to body surface area (BSA) (minimum MV-IVSi distance), 2) LV stroke volume, 3) signal intensity ratio LVOT/LV, and 4) the presence of flow artifacts during systole in LVOT. The multivariable model has an outstanding discriminatory ability with an area under the curve (AUC) of 0.94 (95% confidence interval (CI), 0.92-0.96). Indexing minimum MV-IVS to BSA normalizes values for differences in body size, accounts for sex differences, and improved the diagnostic accuracy for predicting LVOTO. The univariable model of the minimum MV-IVSi distance has an AUC of 0.91 (95% CI 0.89-0.94). Specifically, a minimum MV-IVSi distance ≤6.5 mm/m² yielded a sensitivity of 62%, specificity of 95%, and positive predictive value of 89% for the detection of LVOTO. In comparison, 9.0 mm/m² had a sensitivity of 95%, specificity of 68%, and negative predictive value (NPV) of 95% to rule out LVOTO (Figure 2). NPV increased to 98% when the threshold of LVOTO was set at ≥50 mmHg. The intra- and inter-observer variability was excellent with an intraclass correlation of ≥0.95. Conclusion CMR-derived metrics, particularly the minimum MV-IVS distance indexed to BSA, can accurately identify LVOTO in HCM patients and can easily be integrated into a standard CMR analysis.Figure 1 Figure 2