Quantitative characterization of platelet count and alanine aminotransferase dynamics to inform personalized dosing in the first-in-human LP-184 trial.

Safety evaluation in first-in-human (FIH) Phase 1 oncology trials is largely descriptive and provides limited quantitative insights. This study characterized the dynamics of platelet count (PLT) and alanine aminotransferase (ALT) concentration in the LP-184 FIH study and explored statistical modeling to support personalized dosing strategies. PLT and ALT data were analyzed to characterize temporal trends, inter-patient variability, and correlations with baseline characteristics and exposure. Firth logistic regression was fitted to identify potential contributors to clinically relevant PLT abnormalities. PLT nadir predominantly occurred in cycle 2, with baseline PLT and total single-infusion dose as critical contributors. Patients with baseline PLT below 200 K/µL were at high risk of developing grade ≥ 2 PLT decreased adverse events. ALT levels mostly peaked in cycle 1, with a moderate association with dose and no evident relationship to baseline liver metastases. Patients with glioblastoma exhibited a higher probability of experiencing ALT increased events at grade ≥ 2. A single infusion dose above 40 mg was associated with increased occurrence of high-grade PLT decreased and ALT increased events. A personalized dose recommendation strategy is proposed. Integration of in-depth data analytics in FIH studies facilitates earlier safety signal detection and informs safety monitoring and dose selection in later studies. ClinicalTrials.gov, TRN: NCT05933265, Registration date: 23 June 2023.