Cytokine Signatures of Immune Checkpoint Inhibitor Myocarditis and the Modifying Role of Clonal Hematopoiesis: A Machine-Learning and Multi-Omics Study

Background Immune checkpoint inhibitor (ICI) myocarditis affects up to 2% of patients on ICI and has up to a 50% mortality. Clonal Hematopoiesis of Indeterminate Potential (CHIP) is a risk factor for ICI myocarditis. Objectives: To leverage a multi-omics approach to evaluate cytokines associated with ICI myocarditis, immunosuppression resistance, CHIP in ICI myocarditis and implicated pathways and cell type contributions. Methods 236 cardio-oncology patients were enrolled, including 88 who were on ICI. Cytokine profiling from patients on ICI were sampled and analyzed using machine learning and explainable artificial intelligence. Single-cell multi-omics of peripheral blood from Yale (n=19), Stanford (n=25), and MGH (n=63) evaluated cell-type contributions of cytokines and implicated pathways. Validation of Yale proteomic analyses was done using the Stanford CYTOF dataset. Results Top features associated with ICI myocarditis included IL13, CXCL10, PDGFA, CCL13, CCL17 and KITlg. Top KEGG pathways associated with ICI myocarditis and immunosuppression resistant ICI myocarditis included JAK STAT signaling, IL17 signaling and lipid/atherosclerosis pathways. At the single-cell level, the consensus signals distinguishing myocarditis from non-myocarditis across cohorts was predominantly driven by module-level JAK STAT, T cell, and Toll like receptor activation. Similar transcription gene programs were activated in CHIP carriers including modules implicated in the pro-inflammatory immune response, such as NF kappa B, IL17, MAPK, and TNF pathways within CHIP cells spanning lymphoid (cytotoxic CD8+ NKT, and natural killer cells) and myeloid lineages. Conclusions Across cohorts, JAK STAT signaling emerges as a convergent and potentially actionable axis in ICI myocarditis that is further amplified in CHIP carriers, supporting cytokine-guided risk stratification and JAK STAT directed immunosuppression as testable strategies.

Authors

Institutions

Publication Details

Journal
medRxiv
Published
2026-11-12
DOI
https://doi.org/10.64898/2026.11.11.26367634
Primary Topic
Cancer Immunotherapy and Biomarkers
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Cytokine Signatures of Immune Checkpoint Inhibitor Myocarditis and the Modifying Role of Clonal Hematopoiesis: A Machine-Learning and Multi-Omics Study

Neal Patrick Smith, Eiman Ibrahim, Jaiveer Singh, Ritujith Jayakrishnan et al.
medRxiv
Cancer Immunotherapy and Biomarkers
preprint

Cytokine Signatures of Immune Checkpoint Inhibitor Myocarditis and the Modifying Role of Clonal Hematopoiesis: A Machine-Learning and Multi-Omics Study

Neal Patrick Smith, Eiman Ibrahim, Jaiveer Singh, Ritujith Jayakrishnan, Alokkumar Jha, Yunju Im, Ross L. Levine, Carlos Felipe Matute Martinez, Han Zhu, Evaline Cheng, Jennifer M. Kwan, Stephanie Halene, Anastasia Ardasheva, Jennifer VanOudenhove, Nancy Manchanda, Giulia Biancon, John Hwa, Anthos Christofides, Avnith Vijayram, Dat Hong, James L Cross, Rachel Jaber Chehayeb, Chloe Villani, Akarvin Jeyashanmugaraja, Derrick Lin, Emily Yuan Shuan Lin
preprint en

Abstract

Background Immune checkpoint inhibitor (ICI) myocarditis affects up to 2% of patients on ICI and has up to a 50% mortality. Clonal Hematopoiesis of Indeterminate Potential (CHIP) is a risk factor for ICI myocarditis. Objectives: To leverage a multi-omics approach to evaluate cytokines associated with ICI myocarditis, immunosuppression resistance, CHIP in ICI myocarditis and implicated pathways and cell type contributions. Methods 236 cardio-oncology patients were enrolled, including 88 who were on ICI. Cytokine profiling from patients on ICI were sampled and analyzed using machine learning and explainable artificial intelligence. Single-cell multi-omics of peripheral blood from Yale (n=19), Stanford (n=25), and MGH (n=63) evaluated cell-type contributions of cytokines and implicated pathways. Validation of Yale proteomic analyses was done using the Stanford CYTOF dataset. Results Top features associated with ICI myocarditis included IL13, CXCL10, PDGFA, CCL13, CCL17 and KITlg. Top KEGG pathways associated with ICI myocarditis and immunosuppression resistant ICI myocarditis included JAK STAT signaling, IL17 signaling and lipid/atherosclerosis pathways. At the single-cell level, the consensus signals distinguishing myocarditis from non-myocarditis across cohorts was predominantly driven by module-level JAK STAT, T cell, and Toll like receptor activation. Similar transcription gene programs were activated in CHIP carriers including modules implicated in the pro-inflammatory immune response, such as NF kappa B, IL17, MAPK, and TNF pathways within CHIP cells spanning lymphoid (cytotoxic CD8+ NKT, and natural killer cells) and myeloid lineages. Conclusions Across cohorts, JAK STAT signaling emerges as a convergent and potentially actionable axis in ICI myocarditis that is further amplified in CHIP carriers, supporting cytokine-guided risk stratification and JAK STAT directed immunosuppression as testable strategies.

medRxiv
Boston University (US), Yale New Haven Hospital (US), Memorial Sloan Kettering Cancer Center (US), The University of Texas MD Anderson Cancer Center (US), Yale Cancer Center (US), Yale University (US), Massachusetts General Hospital (US), Stanford Medicine (US), Weill Cornell Medicine (US), University of Nebraska Medical Center (US), Carnegie Mellon University (US), University of Missouri (US), University of Pennsylvania (US), Stanford University (US)
Good health and well-being
Cancer Immunotherapy and Biomarkers
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon