The Incidence of Immune Checkpoint Inhibitor-Related Hepatotoxicity (ICH) in a Real-World Setting: A Retrospective Cohort Study

초록

Background: Recent meta-analyses of clinical trials have identified liver injury as a common adverse event among cancer patients treated with immune checkpoint inhibitors (ICIs). Objectives: To estimate the incidence and risk factors of ICH among cancer patients between those with normal and abnormal liver function at baseline. Methods: We utilized electronic health records from Pusan National University Hospital transformed to OMOP-CDM, involving adult patients treated with ICIs between January 2011 and June 2022. Excluded were patients lacking liver function test results and those with liver cancer but normal liver function at baseline. We stratified the patients into two groups based on the baseline liver function: the normal liver function, defined by AST and ALT ≤ 40 IU/L and total bilirubin ≤ 1.2 mg/dL, and the abnormal liver function group. Grade 3 hepatotoxicity defined by the CTCAE v5.0 was the outcome measure. We censored patients at 30 days after the last ICI dose or the data cut-off date (June 30, 2022), whichever came first. Baseline characteristics were compared between stratified groups using t-test or chi-square test, as appropriate. Within each group, incidence of ICH as well as the associations between baseline characteristics and ICH were estimated using a multivariable Cox regression model. Results: Of 840 ICI-treated adult patients, 803 were eligible for this study. Mean age was 67.1 years (SD: 10.1) and the majority were male (n = 563, 70.1%). The most common ICIs were atezolizumab (n = 275, 34.2%) and pembrolizumab (n = 264, 32.9%). About 32.4% (n = 260) received a chemotherapy during follow-up, primarily carboplatin (n = 141, 54.2%) and etoposide (n = 57, 21.9%). The incidence of ICH was 19.0 cases / 100 person-years overall, with higher rates in the abnormal group (9.1 vs. 45.4 cases/100 person-years). Baseline characteristics were significantly different between groups except for the type of chemotherapy. Compared to males, females were more likely to develop ICH in the normal group (HR: 2.54, p value: 0.045); and patients with liver cancer (HR: 2.99, p value: 0.034) or receiving etoposide (HR: 24.0, p value: 0.039) were more likely to develop ICH in the abnormal group. Conclusions: Abnormal liver function at baseline exhibited a higher ICH rate; liver cancer and concurrent use of etoposide were identified as unique risk factors for ICH within this group, indicating the underlying causes of ICH development vary according to the baseline liver function. These results underscore the importance of tailoring ICH prevention strategies to the baseline liver function of patients treated with ICIs.