Background and Aim Tyrosine kinase inhibitors (TKIs) have substantially improved the prognosis of chronic myeloid leukemia (CML). However, TKIs can cause serious vascular adverse events (VAEs) such as ischemic heart disease (IHD), peripheral artery occlusive disease (PAOD), cerebral infarction (CI), and pulmonary embolism (PE). The Framingham general cardiovascular risk scores (FRS), Suita score (SS), and SCORE (Systematic Coronary Risk Evaluation) system are commonly used to predict the incidence of cardiovascular events, but their usefulness as evaluation methods for patients with CML is unknown. We aimed to determine the appropriateness of these evaluation methods for a Japanese CML cohort. Methods We retrospectively surveyed patients with CML who visited our institute between January 2003 and March 2020. All patients were evaluated using the three risk assessment tools (FRS, SS, low-risk SCORE chart). To predict the 5-year event-free survival, we determined the cut-off values of each scoring system from the receiver operating characteristic analysis (ROC) curve. Results Sixty-nine patients with CML (38 males) were registered. Their median age was 66 years (range, 25–91years), and the median follow-up time was 48 months. Eight patients (2 males) had VAEs (3 IHD, 2 PAOD, 1 CI, 1 PE, and 1 sudden death). Four patients were treated with nilotinib, 2 with imatinib,1 with dasatinib, and 1 with bosutinib. The cut-off values obtained from the ROC curve for the SS and SCORE chart were 45 and 2, respectively. The sensitivity, specificity, and area under the curve (AUC) were 0.574, 0.875, and 0.713 for the SS and 0.689, 0.750, and 0.755 for the SCORE chart, respectively. We were unable to calculate the FRS because of the small number of men with thrombosis. However, the cut-off value was 11 points when limited to the female patients. The 5-year cumulative incidence rates of VTEs were 14.9% and 0% in patients with SS of ≥45 and Conclusion For this Japanese CML cohort, we propose that the SS and SCORE chart are valid evaluation methods for VAE assessment.
Read full abstract