PREDICTIVE MODEL TO ASSESS RISK FOR CARDIAC ALLOGRAFT VASCULOPATHY - AN INTRAVASCULAR ULTRASOUND STUDY

Objectives. This study was performed to assess the influence and interdependence of immunologic and nonimmunologic risk factors in the development of cardiac allograft vasculopathy. Another primary objective was to establish a clinically useful model for risk assessment of cardiac allograft vasculopathy that would facilitate identifying those heart transplant recipients likely to have severe intimal proliferation and thereby at greatest risk for adverse clinical events. Background. To our knowledge, no comprehensive intravascular ultrasound study has assessed the relative influences of both nonimmunologic and immunologic factors in the development of cardiac allograft vasculopathy, currently the major limitation to long-term cardiac allograft survival. Methods. Using a computer-assisted model of stepwise logistic regression, immunologic and nonimmunologic risk factors were evaluated to help identify the development of severe intimal thickening in 101 subjects who underwent intravascular ultrasound. Prospective validation of the fmdings was performed in a separate consecutive cohort of 37 heart transplant recipients, and the accuracy of this model to predict a relative risk >1 for the development of severe intimal hyperplasia was assessed. Results. Significant independent predictors of severe intimal hyperplasia in this model included a donor age >35 years, a first-year mean biopsy score >1 (a measure not only of severity of rejection, but also of frequency of insidious rejection) and hypertrigIyceridemia at two incremental levels of risk (150 to 250 mg/dl [1.70 to 2.83 mmol/liter] and >250 mg/dI [2.83 mmol/liter]). Based on the absence (0) or presence (1) of these factors, 12 individual categories of risk were ascertained with increasing relative risks and predicted probabilities for severe intimal hyperplasia. Prospective validation of this model revealed a sensitivity and specificity of 70% and 90%, respectively, and the positive and negative predictive values were 85% and 80%, respectively. Additionally, subjects with severe intimal thickening had a fourfold higher cardiac event rate than those without severe intimal proliferation on intravascular ultrasound. Conclusions. This study establishes a clinically useful predictive model that can be applied to individual heart transplant recipients to assess their risk for developing significant cardiac allograft vasculopathy and, thus, aids in the identification of patients at risk for cardiac events in whom closer surveillance and risk factor modification may be warranted.