Tyrosine kinase inhibitor (second-generation)-related pancreatic toxicity in chronic myeloid leukemia patients analyzed using the FAERS database

OBJECTIVE: To analyze pharmacovigilance data for second-generation tyrosine kinase inhibitor (TKI)-related acute pancreatic toxicity in chronic myeloid leukemia (CML) patients contained in the FAERS database. MATERIALS AND METHODS: This observational pharmacovigilance study examined reports for the period Q1 2016 to Q3 2025. Disproportionality analysis was performed using four procedures: i) reporting odds ratio (ROR), ii) proportional reporting ratio (PRR), iii) Bayesian confidence propagation neural network (BCPNN), and iv) multi-item gamma Poisson shrinker (MGPS). RESULTS: Of the 16,057,190 adverse events (AEs) retrieved, 24,803, 31,487, and 11,404 were linked to nilotinib, dasatinib, and bosutinib, respectively. Nilotinib exhibited the strongest and broadest risk signals, including significant associations with pancreatitis (ROR = 10.84), acute pancreatitis, elevated amylase (ROR = 19.94) and lipase (ROR = 14.25), and increased bilirubin (ROR = 11.40), indicating clear pancreatic toxicity with potential hepatic involvement and abdominal symptoms. Bosutinib primarily signaled intense and upper abdominal pain (ROR = 5.04 - 5.44), with weaker signals for pancreatitis. Dasatinib showed the weakest signals, mainly elevated amylase (ROR = 4.79) and lipase (ROR = 8.57). Four algorithms yielded 12 positive signals (17.39% of total). Median age was 52 - 58 years; onset of AEs typically occurred within 30 days of commencing treatment. CONCLUSION: Pharmacovigilance data in the FAERS database showed strongest pancreatic/gastrointestinal toxicity signals for nilotinib, indicating the need for close monitoring, followed next by bosutinib, with dasatinib exhibiting the weakest signals of the 3 agents. These findings arising from patient therapy serve as a valuable guide for decisions on drug selection, risk management, and future research.