Multivariate Predictors of Heart Transplantation Outcomes in the Era of Chronic Mechanical Circulatory Support

Abstract

Determining which pretransplantation (TX) characteristics predict the development of chronic renal dysfunction (CRD) or death after heart TX would enable more accurate risk assessment at the time of candidate evaluation. A cohort of 278 patients underwent TX in three hospitals between 1993 and 2002. Predictive models for CRD (serum creatinine consistently above 2 mg/dL) and allograft loss (death or re-TX) were constructed using logistic and Cox regression, respectively. Using logistic regression, CRD was more likely to develop in TX patients if they had a larger body surface area (odds ratio [OR] = 5.8 per m2, 95% confidence interval [CI] = 1.04 to 31.9, p = 0.04) or were inotrope dependent (OR = 1.8, 95% CI = 0.90 to 3.7, p = 0.09). Notably, the implementation of mechanical circulatory support as bridge to transplantation decreased the risk of CRD (OR = 0.30, 95% CI = 0.12 to 0.72, p = 0.007). Cox analysis demonstrated independent predictive ability of improved survival for males (hazard ratio [HR] = 0.42, 95% CI = 0.21 to 0.83, p = 0.01). Worse survival was observed with prior sternotomy (HR = 3.5, 95% CI = 2.0 to 6.0, p < 0.001), diabetes mellitus (HR = 1.9, 95% CI = 0.98 to 3.9, p = 0.06), and elevated serum creatinine (HR = 2.8 per mg/dL, 95% CI = 1.3 to 5.8, p = 0.007). Certain pretransplant characteristics clearly predispose a patient to the development of CRD or increased mortality after heart transplantation. Interestingly, the risk of CRD after heart transplantation is greater for patients bridged to transplant with inotropes than with mechanical circulatory support. When hemodynamically indicated, timely implementation of pretransplant mechanical circulatory support should be considered.