内田 雅士

Introduction. The aim of this study was to evaluate the practicality of the signal transducer and activator of transcription (STAT) 3 polymorphisms as a predictive biomarker and sorafenib trough concentration as a monitoring biomarker for hand-foot skin reaction (HFSR) in patients with hepatocellular carcinoma (HCC). Methods. In total, 43 Japanese HCC patients were included. Sorafenib concentrations were measured, if possible, on days 8, 29, 35, and 57. The sorafenib concentration on day 8 (Cday8) was used for the analysis of HFSR occurring up to day 29. The median concentration for each patient (Cmedian) was used for HFSR occurring up to day 57 (study period). The STAT3 single nucleotide polymorphism (SNP) rs4796793 was determined using cell-free DNA extracted from plasma. Result. The Cday8 tended to be higher in the HFSR onset or grade ≥ 2 HFSR severity group than in the non-HFSR or grade ≤ 1 HFSR severity group. The Cmedian was significantly higher in the HFSR onset or grade ≥ 2 group than in the non-HFSR or grade ≤ 1 HFSR group. The Cmedian thresholds for predicting HFSR onset and severity were 3.62 μg/mL and 6.10 μg/mL, respectively. There was no association between STAT3 rs4796793 and HFSR onset or severity. In multivariate analysis, Cmedian values ≥ 3.62 μg/mL and >6.10 μg/mL were associated with the increased risk of HFSR onset (odds ratio: 16.6, p < 0.01) and severity (odds ratio: 15.7, p < 0.01), respectively. Conclusion. Monitoring of the sorafenib trough concentration may be practical for avoiding HFSR.