Clinical and molecular features associated with glucolipid metabolic disorders and cardio-/cerebro-vascular adverse events in CML patients receiving olverembatinib therapy

Abstract Background Olverembatinib, a third-generation BCR::ABL1tyrosine kinase inhibitor (TKI), has demonstrated promising efficacy in resistant and/or intolerant chronic myeloid leukemia (CML). However, cardio- and cerebro-vascular adverse events (CCVAEs) and metabolic disturbances are emerging concerns. Whether baseline clinical and molecular features contribute to these adverse events remains unclear. We investigated the associations among clinical characteristics and somatic mutations at baseline and the risk of glucolipid metabolic disorders and CCVAEs in TKI-resistant CML patients receiving olverembatinib-therapy. Methods We retrospectively analyzed 167 TKI-resistant CML patients in chronic or accelerated phase who received olverembatinib therapy and had baseline somatic mutation data from targeted DNA sequencing and glucolipid metabolic laboratory data including fasting glucose, total cholesterol and lipid profile (low-density lipoprotein, high-density lipoprotein and triglycerides). Glucolipid metabolic disorders were defined as new-onset or worsened hyperglycemia, hyperlipidemia, or hypercholesterolemia assessed according to CTCAE criteria during olverembatinib therapy. CCVAEs were defined as new-onset or aggravated hypertension, arterial or venous occlusion, heart failure or pulmonary hypertension. Results 167 patients with chronic (n = 108) or accelerated phase (n = 59) TKI resistant CML treated with olverembatinib were included in this study. Median age was 43 (interquartile range [IQR], 31 - 50) years, and 116 (69%) were male. 22 (13%) patients experienced imatinib failure only; 50 (30%), imatinib, dasatinib and/or flumatinib therapy failure; and 95 (57%) had a history of nilotinib exposure. 135 (81%) patients received ≥2 prior TKIs. With a median follow-up was 56 (IQR, 39 - 72) months, 67 (40%) patients developed new-onset or worsened metabolic disorders including hyperglycemia (n = 30), hypercholesterolemia (n = 21), hyperlipidemia (n = 17) alone or combined. Median interval from starting olverembatinib therapy to the onset of metabolic disorders was 11 (IQR, 5 - 33) months. Additionally, 58 (35%) patients experienced CCVAEs including hypertension (n = 28), arterial/venous occlusion (n = 29), intracerebral hemorrhage (n = 3), pulmonary hypertension (n = 2) and heart failure (n = 1) alone or combined. Median interval from starting olverembatinib therapy to the onset of CCVAEs was 34 (IQR, 18 - 56) months. Among them, arterial/venous occlusive events, the most clinically notable events, occurred at a median of 40 (IQR, 23 - 60) months. In multi-variable Cox regression, higher total cholesterol (HR = 1.40 [95%CI, 1.08 - 1.81] per mmol/L, p = 0.01), the increasing number of prior TKI lines (HR = 1.43 per line [95%CI, 1.01 - 2.02], p= 0.04) and the presence of ASXL1mutation (HR = 2.07 [1.17 - 3.67], p= 0.01) before olverembatinib therapy were significantly-associated with glucolipid metabolic disorders. Prior nilotinib exposure (HR = 4.14 [1.28 - 13.37], p= 0.02) and mutations in FAT4(HR = 6.31 [1.76 - 22.65], p = 0.005) and TET2 (HR = 6.64 [2.13 - 20.72], p = 0.001) were significantly-associated with increased risk of CCVAEs. Moreover, prior nilotinib exposure (HR = 12.12 [1.68 - 87.64], p = 0.01), higher total cholesterol (HR = 4.43 [1.80 - 10.93], p = 0.001) and fasting glucose levels (HR = 1.76 [1.15 - 2.69], p = 0.009), as well as the presence of ASXL1(HR = 4.23 [1.44 - 12.42], p = 0.009), TET2(HR = 15.31 [2.44 - 96.03], p = 0.004), and FAT4(HR = 39.22 [7.89 - 195.03], p < 0.001)mutations were significantly-associated with arterial/venous occlusive events. Conclusions Glucolipid metabolic disorders and CCVAEs are common olverembatinib-induced adverse events in CML, occurring in 40% and 35% of patients, respectively. Higher total cholesterol, more prior TKI-therapy lines, particularly prior nilotinib exposure, the presence of ASXL1, FAT4and/orTET2mutations were the independent predictors for glucolipid metabolic disorders, CCVAEs and arterial/venous occlusive events. These clinical and molecular features may guide early risk identification and targeted surveillance.