A machine learning-driven score predicted diagnostic endomyocardial biopsy with an AUC of 0.91 in internal and 0.82 in external validation cohorts, identifying 19.9% diagnostic yield among heart failure patients of unknown etiology.
Observational (n=775)
Sí
Does a machine learning-driven score based on non-invasive clinical and imaging data predict the diagnostic yield of endomyocardial biopsy in heart failure patients?
A novel machine learning-derived score using non-invasive parameters like CMR LGE and NT-proBNP can accurately predict the likelihood of a diagnostic endomyocardial biopsy, potentially reducing unnecessary invasive procedures.
Estimación del efecto: AUC 0.91 internal validation; AUC 0.82 external validation (95% CI Internal validation 95% CI 0.89–0.96; External validation 95% CI 0.76–0.89)
Abstract Despite its low diagnostic yield, endomyocardial biopsy (EMB) remains the gold standard for establishing a definitive diagnosis in many cardiomyopathies. We developed and validated a machine-learning–based score to predict the likelihood of diagnostic EMB using non-invasive data. We retrospectively analyzed 775 heart failure patients who underwent EMB. A random forest algorithm was selected for score development based on superior discriminative performance. The model was externally validated in an independent cohort ( n = 171). The study population was predominantly male (72.1%), with half of the patients in NYHA class III–IV. EMB yielded a definitive diagnosis in 19.9% of cases, most commonly amyloidosis (50%). A predictive score (0-100 range) was derived from key non-invasive predictors. Right ventricular late gadolinium enhancement (LGE) on cardiac magnetic resonance emerged as the strongest predictor, followed by left ventricular and atrial LGE, NTproBNP levels, and renal function. The model demonstrated excellent discrimination, with an area under the curve of 0.92 (95% CI = 0.89–0.96) in cross-validation and 0.91 (95% CI = 0.86–0.98) in the testing set, with consistent performance on external validation (AUC 0.82, 95% CI = 0.76–0.89). This machine-learning-based score may provide a non-invasive tool to support EMB decision-making in clinical practice.
Basile et al. (Mon,) realizaron un estudio observacional en adultos con insuficiencia cardíaca de etiología desconocida que se sometieron a biopsia endomiocárdica (n=775). Se evaluó un puntaje predictivo impulsado por aprendizaje automático basado en variables clínicas y de imagen frente a la ausencia de puntuación / gestalt clínica sobre el rendimiento diagnóstico de la biopsia endomiocárdica definido por diagnóstico histopatológico (AUC 0.91 validación interna; AUC 0.82 validación externa, IC del 95% validación interna 95% CI 0.89–0.96; validación externa 95% CI 0.76–0.89). Un puntaje impulsado por aprendizaje automático predijo la biopsia endomiocárdica diagnóstica con un AUC de 0.91 en cohortes de validación interna y 0.82 en cohortes de validación externa, identificando un rendimiento diagnóstico del 19.9% entre pacientes con insuficiencia cardíaca de etiología desconocida.