Functional tricuspid regurgitation: A clustering analysis and prognostic validation of three echocardiographic phenotypes in an external cohort
Résumé
BACKGROUND: Functional tricuspid regurgitation (FTR) is an independent risk factor for morbidity and mortality. New pathophysiological concepts but also new therapeutic options are justifying new knowledges for characterizing FTRs and their prognoses. AIM: To study echocardiographic criteria associated with prognosis in FTR-patients using a clustering method in two cohorts. METHODS AND RESULTS: Two hundred forty-one patients with at least severe (≥grade 3) TR were enrolled: 92 in the retrospective cohort (mean age 77.9 ± 13 years) and 149 in the prospective validation cohort. Hierarchical clustering analysis was conducted. Four parameters explained the clustering categorization according to a multinomial regression (right ventricular (RV) end-diastolic mid-cavity diameter, RV free-wall strain, right atrial (RA) volume index, RA strain; p = 0.0039). Three clusters were identified in the retrospective cohort: Cluster 1 had better right ventricular, left ventricular, and right atrial function than Cluster 2 (reduced RV and RA strain despite similar sizes). Cluster 3 included patients with severely dilated heart chambers associated to RV and RA dysfunctions. When applying the model in the validation (external) cohort, the rate of the primary endpoint (hospitalization for heart failure and/or death from any cause) was lowest in Cluster 1 (30.8% versus 48% and 58.8% in Clusters 2 and 3, respectively; p < 0.05). CONCLUSION: In FTR patients, different profiles of RV and RA remodeling are associated with different outcomes. Therefore, the diagnostic work-up in this clinical setting should include RV and RA characteristics. Under noninterventional management, the phenotype corresponding to preserved RV size and preserved RA and RV functions appears to have a better prognosis.