Predictive value of clinical examination parameters for cardiovascular adverse events during treatment of chronic myeloid leukemia with tyrosine kinase inhibitors

Abstract

Treatment of chronic myelogenous leukemia (CML) requires management of long-term use of tyrosine kinase inhibitors (TKIs). Although cardiovascular adverse events (CAEs) caused by off-target effects of TKIs can be life-threatening, the optimal method of monitoring for CAEs has not been established. Here, we comprehensively evaluated the clinical utility of various cardiovascular parameters, including ankle-brachial blood pressure index (ABI), cardiac ankle vascular index (CAVI), and carotid ultrasonography and electrocardiogram measurements, for monitoring and predicting CAEs in 74 patients with CML receiving TKIs. Based on concordance statistics, the predictive value of established risk factor models was significantly improved by addition of both ABI and CAVI, as follows: model 1 (hypertension, smoking history, and dyslipidemia), 0.680 vs. 0.817 (p = 0.041); model 2 (hypertension, dyslipidemia, and diabetes mellitus), 0.685 vs. 0.830 (p = 0.047); and model 3 (age, hypertension, dyslipidemia and diabetes mellitus) 0.737 vs. 0.818 (p = 0.044). However, no single cardiovascular parameter independently improved the predictive value of established risk factor models. In conclusion, addition of combined assessment of ABI and CAVI to assessment of established risk factors can improve prediction of future CAEs and may enable better clinical management of patients 3 with CML receiving TKIs.