Tyrosine Protein Kinase SYK-Related Gene Signature in Baseline Immune Cells Associated with Adjuvant Immunotherapy-Induced Immune-Related Adverse Events in Melanoma

Abstract

Purpose: Immune checkpoint inhibition (ICI) shows benefits in adjuvant (AT) and neoadjuvant melanoma treatments. However, ICI frequently induces severe immune-related adverse events (irAE). Unlike metastatic disease, in which irAEs are a clinical trade-off for treatment that improves survival, the toxicity burden from ICI in the AT setting is a substantial clinical problem urging for irAE-predictive biomarkers.

Experimental design: We assessed postsurgical, pre-ICI treatment peripheral CD4+ and CD8+ T cells from clinical trial patients (CheckMate 915) treated with AT nivolumab (n = 130) or ipilimumab/nivolumab (COMBO, n = 82). Performing RNA sequencing differential gene expression analysis, we tested baseline differences associated with severe (grades 3-5) irAEs and constructed an irAE-predictive model using least absolute shrinkage and selection operator-regularized logistic regression.

Results: The analysis of predicted protein-protein interactions among differentially expressed genes in peripheral CD4+ cells revealed significant enrichment of the spleen tyrosine kinase (SYK) pathway, associated with severe irAEs in COMBO-treated patients. This gene expression signature predicted severe-irAE COMBO patients (χ2P value = 0.001) with 73% accuracy and was independent of disease recurrence (P = 0.79). The irAE-predictive model incorporating this gene expression signature demonstrated 82% accuracy (χ2P value = 8.91E-06).

Conclusions: We identified baseline gene expression differences in key immune pathways of peripheral blood T cells from COMBO-treated patients with grades 3 to 5 irAEs and defined a SYK-related gene signature correctly identifying ∼60% of COMBO-treated patients with grades 3 to 5 irAEs. This finding aligns with our previous work linking anti-CTLA4 irAEs with a germline variant associated with high SYK expression. This gene signature may serve as a baseline biomarker of severe grade 3 to 5 irAE risk, which is especially important in AT treatment.

Figure 1.

Heatmaps showing the significantly differentially expressed genes (DEGs) defined as p<0.01, ∣log2FC∣ >1, from RNA-seq DEG analysis of peripheral CD4⁺ and CD8⁺ T cells, comparing high-grade (3-5) vs. low-grade (0-2) irAEs. a. DEGs from COMBO CD4⁺ analysis (n=78 samples); 87 genes (n=63 upregulated in irAE; n=24 downregulated in irAE). b. DEGs from NIVO CD4⁺ analysis (n=124 samples); 43 genes (n=7 upregulated in irAE; n=36 downregulated in irAE). c. DEGs from COMBO CD8⁺ analysis (n=78 samples); 92 genes (n=34 upregulated in irAE; n=58 downregulated in irAE). d. DEGs from NIVO CD8⁺ analysis (n=121 samples); 75 genes (n=29 upregulated in irAE; n=46 downregulated in irAE).

Figure 2.

Pathway enrichment analysis of DEG in peripheral CD4⁺ and CD8⁺ from patients in Checkmate-915. a. The overlap in DEGs between the treatments (NIVO/COMBO) and cell types (CD4⁺/CD8⁺), comparing high-grade (3-5) vs. low-grade (0-2) irAEs. The Intersection Size (y-axis on the vertical bar chart) indicates the number of unique DEGs in each comparison. The Set Size (x-axis on the horizontal bar chart) indicates the total number of DEGs in each category. b. Top enriched pathways (adjusted p<0.01) from GO analysis using the overlapping DEGs between cell types and treatments. Each bar is labeled with the number of DEGs from the pathway that are present (e.g., COMBO CD4⁺ and CD8⁺ cells shared 9 DEGs out of a total of 847 genes in the IL-2 signaling pathway; CD4⁺ cells in COMBO and NIVO share 5 DEGs in the same pathway).

Figure 3.

Protein-protein interaction volcano plot illustrating proteins encoded by significantly enriched gene sets from Combo CD4⁺ DEGs between patients with high-grade (3-5) vs. low-grade (0-2) irAE. The odds ratio (OR) for the odds of observing the given DEGs at random is plotted on the x-axis. The y-axis plots the −log(p-value) for the OR (a larger −log(p-value) means the genes are less likely to be randomly observed). The blue points represent proteins associated with significant gene sets (OR >1.0, p-value <0.05; darker blue signifies greater significance), indicating the DEGs associated with these proteins are unlikely to be randomly observed.

Figure 4.

Differential expression of genes expressed in CD4⁺ T cells that encode proteins associated with SYK identify the majority of COMBO-treated patients with high-grade irAEs. a. Semi-supervised hierarchical clustering using SYK-associated genes in protein-protein interaction (PPI) analysis for COMBO CD4⁺ differential expressed genes (DEGs) between patients with high-grade (3-5) vs. low-grade (0-2) irAE (24 out of 136 SYK-interacting genes; p-value 9.96E-04, adj. p-value 0.04). The cluster to the far right comprises the 24 patients with the most similar mean gene expression of these 24 genes and contains the majority of patients with severe irAEs. b. Network diagram showing the Immune response Fc epsilon Receptor I pathway identified from GO on these 24 genes.

Figure 5.

Bubble plot showing irAEs shared among COMBO-treated patients in the SYK-pathway cluster (SYK-irAEs). To be included, the irAE was present in at least one TOX^high (above dotted line; n=13) and one TOX^low (below dotted line; n=11) patient. Each row represents a patient with the SYK-GEP and each column represents an SYK-irAE. The size of the bubble corresponds to the number of events of each toxicity that the patient experienced while on treatment with larger circles indicating more frequent events. The color scale shows the CTCAE grade for the irAEs: orange and red events are more severe (grade 3-4) while green and blue events are less severe (grade 1-2).

Figure 6.

Model performance for logistic regression model predicting severe irAEs (grade 3-5) using age, sex, stage, and 5-gene SYK GEP expression (CD22, PAG1, CD33, HNRNPU, FCGR2C) in COMBO-treated patients. a. Receiver operating characteristic (ROC) curve and corresponding area under the curve (AUC) plotting sensitivity and specificity of the prediction model. b. Waterfall plot for severe irAE prediction using the regression model parameters, scaled so that prediction scores <0 are predicted to have no severe irAEs, and scores >0 are predicted to have severe irAEs. The actual toxicity status of the patients is indicated by color (blue indicates no severe irAE, salmon indicates severe irAE).