RoMa: A Cardiopulmonary Exercise Testing Based Risk Tool in Hypertrophic Cardiomyopathy

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Background: The RoMa classification, based on peak heart rate and oxygen pulse derived from cardiopulmonary exercise testing, was recently proposed to stratify patients with hypertrophic cardiomyopathy by physiological reserve during exercise. We aimed to externally validate RoMa in an independent multicenter cohort with hypertrophic cardiomyopathy and assess its association with long‐term clinical outcomes. Methods: In this retrospective multicenter cohort study patients with hypertrophic cardiomyopathy, undergoing cardiopulmonary exercise testing, were consecutively enrolled. Patients were enrolled regardless of left ventricular outflow tract obstruction and were naïve to disease‐specific therapy (eg, mavacamten). Patients were categorized into RoMa I to IV based on percentage of predicted heart rate and oxygen pulse. The primary end point was a composite of all‐cause and cardiovascular death, sudden cardiac death, or aborted sudden cardiac death, heart failure–related hospitalization, stroke, systemic embolism, surgical myectomy, and heart transplantation. Results: The study included 292 patients (age 51 [36–63] years, 70% male sex, 30% with obstructive left ventricular outflow tract). Functional capacity declined hierarchically across RoMa groups (peak oxygen uptake 29.2 to 17.9 mL/kg/min; P ‐trend P =0.019). In multivariable analysis, RoMa II to IV compared with RoMa I were independently associated with higher hazard ratios (HRs) for the composite outcome (HRs, 3.89–5.37; all P <0.05), whereas genotype, LVEF <50%, male sex, and left ventricular outflow tract obstruction were not predictive. Conclusions: The RoMa classification independently predicts long‐term, clinically relevant outcomes in hypertrophic cardiomyopathy regardless of left ventricular outflow tract obstruction and may provide a novel approach to risk stratification.

This publication has 28 references indexed in Scilit: