Document resource
Introduction Immune checkpoint inhibitors (CPI) have become the leading cause of drug-induced liver injury (DILI) in national registries. Current evidence on checkpoint inhibitor-induced liver injury (ChILI) is limited to studies with small sample sizes and lacks systematic evaluation. Here, we present a national real-world analysis of incidence, clinical characteristics, and outcomes of ChILI in all patients who received CPI over 5 years.Methods Using the prospective electronic prescribing and management system for Systemic Anti-Cancer Therapy (SACT) in the UK, all adult patients who received CPI for melanoma or renal cell carcinoma with or without targeted therapy between January 2018 and December 2022 at 20 hospitals were identified.Patients were systematically screened for acute liver injury defined as ≥ grade 2 by the Common Terminology Criteria for Adverse Events (CTCAE v5). Detailed clinical data, including CPI, investigations, treatment and outcomes were collected. Causality assessment and formal adjudication were performed. ChILI pattern and severity grading were assessed using the international expert working group criteria and CTCAE.Results Over the 5-year study period, out of 8080 patients who received CPI and were screened, acute liver injury occurred in 9% (724 patients). The overall risk of ChILI among patients receiving CPI was 7% (535 patients). The highest incidence was among melanoma patients receiving combination therapy of ipilimumab and nivolumab, 19% developed ChILI.Liver injury was not attributed to ChILI in 26% of cases. Progressive liver metastases were the most common alternative diagnosis, followed by DILI due to other drugs.Among patients who developed ChILI, hepatocellular was the most common pattern occurring in 41% (220 patients). Three-quarters met the severe liver injury by CTCAE, whereas 2% met the severe DILI criteria. 10 patients had features of cholangiopathy (2%), and 2 had sinusoidal obstruction syndrome (SOS) (0.4%). None of the patients progressed to acute liver failure (ALF) or died due to ChILI.46 out of 535 ChILI patients improved without steroids (9%). Among the 489 who received steroids, 143 had over 3 months of steroids (29%), 122 received additional immunosuppression (25%), and 89 had significant steroid-related side effects (18%).Conclusions In this largest real-world cohort of ChILI to date, no patients developed ALF or died due to ChILI. Acute liver injury is due to an alternative aetiology in over 25% of patients. Therefore, early investigations are crucial to avoid unnecessary immunosuppression.ChILI has multiple phenotypes; cholangiopathy and SOS are rare but require clinical suspicion and early investigations. CTCAE overestimates the clinical severity of ChILI, and the management varies significantly across the UK. Current guidelines need to be revised to harmonise clinical practice.