| Abstract: | Background: The development of tyrosine kinase inhibitor (TKI) improves the survival of chronic myeloid leukemia (CML). Adverse effects following long-term treatment are of concern. Current results regarding metabolic effects of TKI are limited. Objectives: To investigate the risk of type 2 diabetes mellitus (T2DM) and hyperlipidemia (HLD) in CML patients with TKI therapy using real-world data setting. Methods: This retrospective cohort study identified CML patients with aged 20 years old or older from Taiwan Cancer Registry (2010–2018) and extracted their corresponding TKI therapy information from nationwide claim database (2010–2020). Patients with other cancers and with DM or HLD history before TKI therapy were excluded. The DM and HLD cohorts were generated separately. Patients were grouped by their first-line TKI therapy including imatinib, dasatinib and nilotinib. They were followed from TKI initiation to new-onset T2DM or HLD, death, or up to 90 days after treatment discontinuation or switched to another TKI, or the end of database (2020 DEC 31). Multivariable Cox regression was performed to estimate the risks of T2DM or HLD. Results: There were 1359 and 1307 patients in T2DM (41.4% imatinib, 28.7% dasatinib, 29.9% nilotinib) and HLD (41.5% imatinib, 28.7% dasatinib, 29.8% nilotinib) cohorts, respectively. For the risk of T2DM, the comparison among the first-line dasatinib, imatinib and nilotinib users showed no significant difference. The risk of HLD was occurred in nilotinib users, which was 2.83 (HR, 95%CI = 1.96–4.09, p< 0.001) times that of imatinib users and 1.69 (95%CI = 1.10–2.60, p = 0.017) times that of dasatinib users. There was no significant difference for HLD occurrence among the 3 drugs when patients with history of hypertension, diabetes, higher comorbidity scores (2 or more), or older age. In patients without history of hypertension (nilotinib vs imatinib, HR = 3.74, 95%CI = 2.38–5.87, p< 0.001; nilotinib vs dasatinib, HR = 1.87, 95%CI = 1.24–2.82, p = 0.003), without history of coronary artery disease/ischemic heart disease (nilotinib vs imatinib, HR = 3.0; 95%CI = 2.05–4.41, p< 0.001; nilotinib vs dasatinib, HR = 1.67; 95%CI = 1.15–2.43, p-value = 0.008) or without diabetes (nilotinib vs imatinib, HR = 2.96, 95% CI = 2.00–4.37, p< 0.001; nilotinib vs dasatinib, HR = 1.80, 95%CI = 1.21–2.68, p = 0.004), significantly higher risks of HLD were found in the nilotinib group. Conclusions: The first-line nilotinib users had significantly higher risk of HLD, followed by dasatinib as compared to imatinib users. There was no significantly difference in the risk of T2DM among three TKI users. Our results suggest that a routine monitoring of lipid profile might be necessary for CML patients with nilotinib or dasatinib. |
