| Abstract: | Background: The development of tyrosine kinase inhibitors (TKIs) significantly improved the prognosis of chronic myeloid leukemia (CML) patients. Adverse effects following long-term treatment are of concern; especially, cardiovascular events associated abnormal glucose-lipid metabolism. Objectives: This retrospective cohort study investigates the onset of high hemoglobin A1C (HbA1C), hypercholesterolemia, hypertriglyceridemia, and hyper-low density lipoprotein (LDL)-cholesterolemia in the patients treated with first-line TKI. Methods: CML patients aged over 20 years who had related laboratory data in the 6 months prior to the first-line TKI and within one year after TKI were selected. We extracted information from Taiwan National Health Insurance Database and laboratory data (2015–2020). Patients with other cancers history at baseline were excluded, and they were further grouped by the first-line TKI including imatinib, dasatinib and nilotinib. Time-to-event analysis and multivariable Cox regression was performed to analyze the probability of laboratory data over high cut-off values, including high HbA1C (≥ 6.5 %), hypercholesterolemia (≥ 240 mg/dL), hypertriglyceridemia (≥ 200 mg/dL), and hyper-LDL-cholesterolemia (≥ 160 mg/dL). Death, treatment discontinuation, switched to another TKI or the end of 1-year follow-up period were considered censored. Results: Nilotinib (N) users had higher risk than dasatinib (D) and imatinib (I) users in terms of hypercholesterolemia (N vs. I, hazard ratio, HR, [95% CI] = 3.90 [1.71–8.88]; N vs. D, HR = 2.60 [1.50–4.53]), hyper-LDL-cholesterolemia (N vs. I, HR = 2.91 [1.16–7.31]; N vs. D, HR = 2.15 [1.14–4.05]) and high HbA1C level (N vs. I, HR = 2.38 [1.50–3.78]; N vs. D, HR = 2.14 [1.39–3.28]). In patients without history of hyperlipidemia (N vs. I, HR = 7.61 [1.61–35.98]; N vs. D, HR = 4.33 [1.96–9.59]), without history of ischemic heart disease (N vs. I, HR = 4.36 [1.74–10.90]; N vs. D, HR = 2.76 [2.53–4.99]) or without diabetes (N vs. I, HR = 5.38 [1.87–15.47]; N vs. D, HR = 3.09 [1.67–5.71]), significantly higher risks of hypercholesterolemia were found in the nilotinib group. There was no significant difference for hypercholesterolemia onset in those with histories mentioned above. In patients with history of diabetes (N vs. I, HR = 2.3 [1.43–3.69]; N vs. D, HR = 2.18 [1.39–3.41]), significantly higher risks of high HbA1C level were found in the nilotinib group. No significant difference for high HbA1C level was found in those without diabetes history. Conclusions: Compared with imatinib and dasatinib, nilotinib is more likely to present adverse effects on blood glucose and lipids profile. Therefore, the selection of TKIs would need to consider patient's baseline comorbidity history. |
