Clinical Phenotypes in Heart Transplantation: Implications for Prognosis and Personalized Follow-up

Le résumé fourni par la source

BACKGROUND: Outcomes after heart transplantation remain highly variable, and traditional risk models perform poorly in this heterogeneous population. We sought to identify clinically meaningful recipient subgroups and assess their prognostic implications using latent class analysis (LCA). METHODS: We analyzed all adult recipients of a first isolated heart transplantation in Spain. LCA based on pretransplant characteristics identified recipient phenogroups, whose 5-y all-cause mortality was compared using Cox models and internally validated with bootstrap resampling and Harrell's C-statistics. RESULTS: Among 3683 recipients (74.5% male), 3 phenogroups were identified. Phenogroup 1 (41.3%) included younger patients, more women, dilated cardiomyopathy, and fewer comorbidities. Phenogroup 2 (22.1%) comprised older men with ischemic cardiomyopathy and high comorbidity burden. Phenogroup 3 (36.6%) represented urgent transplants with frequent circulatory support, ventilation, and longer ischemia. Five-year mortality was 24.0%, 28.5%, and 33.0% across groups ( P < 0.001). Compared with phenogroup 1, mortality risk was higher in phenogroup 2 (HR = 1.20; 95% confidence interval, 1.02-1.41) and phenogroup 3 (HR = 1.47; 95% confidence interval, 1.28-1.69). Causes of death differed, graft failure predominating in phenogroup 1 and infection in phenogroups 2 and 3. Internal validation confirmed high model stability (entropy 0.99) and modest discrimination (optimism-adjusted C-statistic 0.55). CONCLUSIONS: LCA identified 3 reproducible clinical phenogroups with distinct prognoses and mechanisms of death. Recognizing these patterns may refine posttransplant surveillance and guide phenotype-based management strategies.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.