Cumulative Multidomain Vulnerability and Early Mortality After Heart Transplantation: Development and Temporal Evaluation of the RIGHT-HTx Framework
In the authors' words
BACKGROUND Risk models from high-income transplant programmes may not generalise to resource-constrained settings with different recipient acuity, donor characteristics and access to mechanical circulatory support. We developed RIGHT-HTx, a parsimonious framework integrating clinical and physiological vulnerabilities, and evaluated its prognostic performance in a subsequent same-centre cohort. METHODS This retrospective single-centre study included 436 consecutive adult heart transplant recipients (2012-2026). Of 357 development recipients, 341 (95.5%) were score-evaluable. RIGHT-HTx incorporated three vulnerability domains: clinical acuity (INTERMACS 1-2), graded renal dysfunction (eGFR 30-59 or <30 mL/min/1.73 m2) and haemodynamic-donor mismatch (TPG >12 mmHg combined with donor-to-recipient weight ratio <1.0). A seven-point Cox-derived score underwent internal validation using 1,000 bootstrap resamples and was applied unchanged to 79 subsequent recipients. Low-, intermediate- and high-risk categories represented zero, one and 2 affected domains, respectively. RESULTS Among development recipients, 75/341 (22.0%) died within one year. Mortality increased across low-, intermediate- and high-risk categories (8.1%, 20.3% and 55.6%; log-rank P<0.001). Elevated TPG was not significantly associated with mortality individually, whereas its combination with donor undersizing was independently associated with mortality (adjusted HR 2.87, 95% CI 1.53-5.40; P=0.001). The full model improved discrimination beyond renal dysfunction alone (optimism-corrected C-index 0.067), and the simplified score achieved an optimism-corrected C-index of 0.732. The temporal cohort more often had 1 affected domain (75.9% versus 60.4%; P=0.010). Discrimination was comparable (C-index 0.738, 95% CI 0.643-0.833), although observed Kaplan-Meier one-year mortality exceeded original predictions (35.4% versus 25.4%; calibration slope 0.929). In pooled analyses (n=420), mortality increased from 8.4% (13/154) with no affected domain to 57.3% (47/82) with two or more. CONCLUSIONS RIGHT-HTx identifies cumulative multidomain vulnerability associated with early mortality after heart transplantation. Temporal discrimination was comparable, but absolute-risk calibration differed. The framework may support risk recognition and mitigation, not transplant eligibility or allocation; independent multicentre validation is required before clinical implementation.
Appeared: Friday, November 6. medRxiv. Preprint, not yet peer-reviewed.