Liver injury during durvalumab-based immunotherapy is associated with poorer patient survival: A retrospective analysis

Linnea A Swanson; Ihab Kassab; Irene Tsung; Bryan J Schneider; Robert J Fontana

doi:10.3389/fonc.2022.984940

Liver injury during durvalumab-based immunotherapy is associated with poorer patient survival: A retrospective analysis

Front Oncol. 2022 Oct 24:12:984940. doi: 10.3389/fonc.2022.984940. eCollection 2022.

Authors

Linnea A Swanson¹, Ihab Kassab¹, Irene Tsung², Bryan J Schneider², Robert J Fontana³

Affiliations

¹ Department of Internal Medicine, University of Michigan, Ann Arbor, MI, United States.
² Division of Hematology and Oncology, University of Michigan, Ann Arbor, MI, United States.
³ Division of Gastroenterology and Hepatology, University of Michigan, Ann Arbor, MI, United States.

Abstract

Background: Durvalumab is approved for the treatment of lung cancer, advanced biliary tract cancers, and is also being evaluated in many other solid organ tumors. The aim of our study is to define the incidence, etiology, and outcomes of liver injury in consecutive patients receiving durvalumab-based immunotherapy.

Patients and methods: Durvalumab treated patients between 1/2016 - 7/2020 were identified from the electronic medical record. Liver injury was defined as serum AST or ALT ≥ 5x upper limit of normal (ULN), ALP ≥ 2x ULN, bilirubin ≥ 2.5 mg/dl, or INR ≥ 1.5. Potential drug induced liver injury (DILI) cases were adjudicated using expert opinion scoring and confirmed with Roussel Uclaf Causality Assessment Method (RUCAM).

Results: Amongst 112 patients, 58 (52%) had non-small cell lung cancer, the median age was 65 years, and 60% were male. The 21 (19%) liver injury patients were significantly more likely to harbor hepatic metastases (52% vs 17%, p=<0.001), experience tumor progression (67% vs 32%, p=0.01) or die (48% vs 11%, p<0.001) during follow-up compared to the 91 without liver injury. Using multivariate regression analysis, the development of liver injury during treatment as well as baseline hepatic metastases were independently associated with mortality during follow-up. Six of the 21 (29%) liver injury cases were adjudicated as probable DILI with four attributed to durvalumab and two due to other drugs (paclitaxel, pembrolizumab). Durvalumab was permanently discontinued in two DILI patients, three received corticosteroids, and one was successfully rechallenged. Only one patient with DILI developed jaundice, and none required hospitalization. Liver biochemistries normalized in all 6 DILI cases, while they only normalized in 27% of the 15 non-DILI cases (p=0.002). The 6 DILI patients also had a trend towards improved survival compared to those with other causes of liver injury.

Conclusion: Liver injury was observed in 19% of durvalumab-treated patients and is associated with a greater likelihood of tumor progression and death during follow-up. The four durvalumab DILI cases were mild and self-limited, highlighting the importance of causality assessment to determine the cause of liver injury in oncology patients receiving immunotherapy.

Keywords: Roussel Uclaf Causality Assessment Method; biliary tract cancer; checkpoint inhibitors; drug hepatotoxicity; drug induced liver injury; immune-mediated liver injury; immunotherapy; lung cancer.