Prospective Evaluation of Hepatitis B Virus Reactivation in Cancer Patients Receiving Tyrosine Kinase Inhibitors

Data on hepatitis B virus reactivation (HBVr) during tyrosine kinase inhibitor (TKI) therapy remain limited. We prospectively enrolled patients with chronic or resolved HBV receiving TKIs for lung cancer or chronic myeloid leukemia at one Taiwanese center (January 2023-March 2025). HBV DNA and liver enzymes were monitored 12-weekly and HBsAg 24-weekly. HBVr was defined by AASLD criteria, with APASL and EASL definitions applied secondarily. Among 79 patients (20 HBsAg-positive, 59 resolved HBV), none received antiviral prophylaxis. Median follow-up was 42 weeks (IQR, 19-65). HBVr occurred in 7 (8.9%; 95% CI, 4.4%-17.2%): 5 HBsAg-positive (25.0%; 95% CI, 11.2%-46.9%), and 2 with resolved HBV (3.4%; 95% CI, 0.9%-11.5%). APASL and EASL definitions yielded 30.0% and 35.0% in HBsAg-positive patients, with no additional events in resolved HBV. No HBVr-associated hepatitis flare meeting the study definition was observed, although transient flares between assessments could not be excluded and deterioration in one fatal case was attributed clinically to sepsis. Median HBVr detection time was 19 weeks (IQR, 14-30). Week-48 Kaplan-Meier HBVr incidence was 37.5% (95% CI, 17.0%-69.5%) in HBsAg-positive and 5.6% (95% CI, 1.4%-21.2%) in resolved HBV (log-rank p < 0.001); Aalen-Johansen estimates were 28.3% (95% CI, 9.5%-50.9%) and 4.3% (95% CI, 0.8%-13.2%), respectively. In exploratory Firth-corrected regression, HBsAg positivity was associated with HBVr (HR, 6.95; 95% CI, 1.60-40.19), while osimertinib exposure was not. HBVr was common among HBsAg-positive patients and less common but not negligible among those with resolved HBV receiving TKIs.