PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
March 29, 2026Journal for ImmunoTherapy of Cancer3 citationsOpen Access

Circulating exhausted CD8+ effector memory cells differentiate immune checkpoint inhibitor-induced liver injury from other acute immune-mediated liver injuries

View Full Paper
SAStuart AstburyEAEdmond AtallahJGJane I Grove

Key Points

  • To identify biomarkers that distinguish checkpoint inhibitor-induced liver injury (ChILI) from other liver injury types like drug-induced and autoimmune hepatitis.
  • Recruited patients with cancer on immunotherapy who developed ChILI and those who did not.
  • Collected samples during acute liver injury and at 12 weeks post-treatment if no toxicity occurred.
  • Utilized mass cytometry, flow cytometry, single-cell RNA sequencing, and bulk RNA sequencing of liver biopsies for immune profiling.
  • Conducted cytokine profiling of plasma and validated findings across multiple cohorts.
  • Identified a CD8+ effector memory T-cell subset associated with elevated liver enzyme levels.
  • Significant markers such as CD38, HLA-DR, and CXCR3 were linked to ChILI.
  • Flow cytometry showed increased granzyme expression and exhaustion markers in ChILI T-cells compared to other subsets.
  • Bulk RNA-seq indicated a higher presence of resident CD8+ T-cells in ChILI versus DILI and AIH.

Abstract

Background Checkpoint inhibitor-induced liver injury (ChILI) is an immune-related adverse reaction, occurring in patients with cancer receiving immune checkpoint inhibitors (CPI). ChILI is currently managed with high doses of corticosteroids which carry their own risks and potential side effects, and the lack of available biomarkers makes monitoring patients at risk of developing ChILI a challenge. There is no specific test that distinguishes ChILI from other competing diagnoses such as acute autoimmune hepatitis (AIH) and idiosyncratic drug-induced liver injury (DILI) due to other medications. Methods Patients with cancer taking immunotherapy who did and did not develop ChILI were recruited. Patients gave samples during the acute phase of liver injury, before CPI treatment and at 12 weeks following the start of CPI therapy if no toxicity developed. Healthy controls were recruited as well as patients with DILI and AIH. Whole blood was taken for broad immune phenotyping using mass cytometry, peripheral blood mononuclear cells were isolated for validatory flow cytometry and single-cell RNA sequencing (RNA-seq), and plasma was used for cytokine profiling. Samples from a second ChILI cohort were used for validation. Snap-frozen liver biopsies were used for bulk RNA-seq to compare the immune response in ChILI to DILI and AIH and correlate with peripheral immune signals. Formalin-fixed paraffin-embedded liver biopsies were used to visualize liver CD8 + T-cell infiltration using confocal microscopy. Results We have identified a circulating CD8 + effector memory T-cell subset, expressing high levels of CD38, HLA-DR and CXCR3 and significantly correlated with alanine transaminase. Flow cytometry and single-cell RNA sequencing revealed an increase in granzyme expression, the liver residency marker CD69 and exhaustion markers CTLA-4, PDCD1 and HAVCR2 relative to other CD8 + effector subsets. Liver tissue bulk RNA-seq and immune cell deconvolution showed a significant increase in resident CD8 + T cells in ChILI compared with DILI and AIH, and a significant upregulation of genes related to CXCR chemokine receptor binding. Plasma cytokine profiling highlighted soluble CD27 and PD-1 as significantly elevated in ChILI relative to controls on CPI. Conclusions We have shown that circulating CD8 + T cells provide a potential biomarker to distinguish ChILI from DILI and AIH, and highlight different mechanistic pathways between ChILI and other immune-mediated liver injuries.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Astbury et al. (2026) studied this question.

synapsesocial.com/papers/69c8c2d1de0f0f753b39d383https://doi.org/10.1136/jitc-2025-014178
Ask AI
Helpful
Bookmark
Share
View Full Paper