Significant secondary tricuspid regurgitation (STR) is associated with poor prognosis, but its heterogeneity makes predicting patient outcomes challenging. Our objective was to identify STR prognostic phenogroups.

We analyzed 758 patients with moderate-to-severe STR: 558 (74±14 years, 55% women) in the derivation cohort and 200 (73±12 years, 60% women) in the external validation cohort. The primary endpoint was a composite of heart failure hospitalization and all-cause mortality.

We identified 3 phenogroups. The low-risk phenogroup (2-year event-free survival 80%, 95%CI, 74%-87%) had moderate STR, preserved right ventricular (RV) size and function, and a moderately dilated but normally functioning right atrium. The intermediate-risk phenogroup (HR, 2.20; 95%CI, 1.44-3.37; P<.001) included older patients with severe STR, and a mildly dilated but uncoupled RV. The high-risk phenogroup (HR, 4.67; 95%CI, 3.20-6.82; P<.001) included younger patients with massive-to-torrential tricuspid regurgitation, as well as severely dilated and dysfunctional RV and right atrium. Multivariable analysis confirmed the clustering as independently associated with the composite endpoint (HR, 1.40; 95%CI, 1.13-1.70; P=.002). A supervised machine learning model, developed to assist clinicians in assigning patients to the 3 phenogroups, demonstrated excellent performance both in the derivation cohort (accuracy=0.91, precision=0.91, recall=0.91, and F1 score=0.91) and in the validation cohort (accuracy=0.80, precision=0.78, recall=0.78, and F1 score=0.77).

The unsupervised cluster analysis identified 3 risk phenogroups, which could assist clinicians in developing more personalized treatment and follow-up strategies for STR patients.