Abstract 18860: The Diagnostic Utility of Risk Factors Defined by the AHA/ACCF Expert Consensus Statement on the Evaluation of Cardiac Disease in Liver Transplantation Candidates

Abstract

Introduction: The 2012 AHA/ACCF consensus statement regarding cardiac evaluation of liver transplant (LT) candidates specifies 7 risk factors for identifying candidates for cardiac evaluation prior to LT. These include age >60 yrs, hypertension, diabetes, smoking, dyslipidemia, prior cardiovascular disease, and left ventricular hypertrophy. The document issued a Class IIb, LOE C recommendation for noninvasive testing in asymptomatic LT candidates who have “multiple” risk factors but did not define the number of risk factors required to prompt evaluation, suggesting that ≥3 risk factors may be “reasonable”. Hypothesis: Risk factors set forth by the AHA/ACCF can predict significant coronary artery disease (CAD) in LT candidates. We sought to identify a threshold sum of risk factors that should prompt CAD evaluation. Methods: Consecutive LT recipients who underwent pre-operative coronary angiography were analyzed. Logistic regression model and ROC curves were used to determine the predictive value of the sum of risk factors in predicting presence of CAD. Results: Among 45 consecutive patients (44% female; mean age 57±8 yrs) with coronary angiogram prior to LT, 15 had significant CAD (≥50% stenosis) and 13 had severe CAD (≥70% stenosis). The sum of risk factors was strongly predictive of significant CAD (OR, 7.12 per 1 risk factor increment; CI, 2.13-23.73; P = 0.001). ROC analyses demonstrated that the sum of AHA/ACCF risk factors was associated with AUC of 0.89 and 0.88 for significant and severe CAD, respectively (Fig 1). A threshold of ≥2 risk factors provided 100% sensitivity and 60% specificity for severe CAD, while a threshold of ≥3 had inferior sensitivity (Fig 1). Conclusion: This study represents a diagnostic validation of the risk factors set forth by the 2012 AHA/ACCF consensus statement. A threshold of ≥2 risk factors seems optimal for identifying candidates for noninvasive cardiac evaluation in asymptomatic LT candidates.