Questionable Validity of Cardiac Risk Score on the Basis of the NSABP B-31 Model

Abstract

TO THE EDITOR: Romond et al 1 presented the results of a 7-year follow-up of the cardiac function in the NSABP B-31 clinical trial. The authors concluded that the late development of chronic heart failure (CHF) after the addition of trastuzumab to paclitaxel after doxorubicin/cyclophosphamide chemotherapy is uncommon; furthermore, they developed a cardiac risk score that includes age and baseline left ventricular ejection fraction (LVEF) assessed by multigated acquisition scans for the prediction of the absolute risk of CHF in individual patients. Although a number of limitations have already been pointed out in the editorial by Mayer and Lin, 2 we would like to discuss some points in greater detail. Currently, there is no universally accepted gold standard for measuring LVEF that is widely used to monitor the cardiac systolic function after chemotherapy. Multigated acquisition scan is performed by labeling a patient’s RBCs with technetium 99m pertechnetate and measures changes in radioactivity in the left ventricle between end-diastole and end-systole rather than truly measuring left ventricle volumes. However, because the acquisition of counts is gated with the patient’s electrocardiogram, the accuracy of LVEF calculations in patients with arrhythmias will lead to artificially reduced counts and reduce the accuracy of LVEF calculations. In addition, poor labeling of the RBCs can occur. In one study, the average LVEF showed considerable variation (up to 8% differences in calculated means between institutions), using the same raw data but differ