Background: Cardiac allograft vasculopathy (CAV) represents a major prognostic factor in long-term survivors of heart transplantation (HTx). Reliable diagnosis of CAV late after HTx is important but remains the domain of invasive techniques such as coronary angiography.
Methods: To test alternative approaches, 54 consecutive HTx recipients (mean time since HTx: 52 months) were studied with intravascular ultrasound (IVUS), angiography, dobutamine stress echocardiography and immunofluorescence staining against anti-thrombin III (AT-III) in endomyocardial biopsies. Univariate and multivariate predictors as well as receiver-operating-characteristic (ROC) curves of different sets of predictors were calculated.
Results: Using IVUS as reference standard, CAV was present in 80% of subjects. Coronary angiography identified CAV correctly in only 44% of cases. If AT-III staining alone was used as a diagnostic criterion, CAV was correctly identified in 77% of subjects. In a multivariate analysis, only AT-III, donor age and echocardiography at rest emerged as independent predictors of CAV (p < 0.05 for all), yielding an excellent discriminative power.
Conclusions: With almost equal reliability when compared with IVUS, CAV can be identified using information on donor age, wall motion score at rest and AT-III staining late after HTx. Coronary angiography may be limited to patients with a high probability score and should not be used routinely for surveillance of CAV.