Fig 1: (A) TIDE analysis was used to predict patient response to ICI in 7 independent datasets of skin melanoma patients. Higher TIDE score (blue) denotes non-responders to immune-checkpoint blockade, whereas lower TIDE score (red) denotes responders. (B) Percentage (%) of high or low cytotoxic T-cell lymphocytes (CTLs) among responders or non-responders to ICI therapies, across all 7 melanoma datasets. Clearly, higher CTL levels were found among ICI-responders. The numbers of ICI-responders or non-responders with high or low CTL levels, are indicated within each bar. (C) Indicative immunophenogram depicting the four categories of markers, the expression of which, was used to calculate the immunophenoscores (IPS) for each TMB subgroup of patients. These include: MHC molecules (MHC), immunomodulators (CP), effector cells (EC) and suppressor cells (SC). The outer part of the circle includes individual factors; whereas, the inner part illustrates the weighted average z-scores of the factors included in each category. Sample wise z-scores were positively weighted according to stimulatory cells and negatively weighted according to inhibitory cells and averaged. (D) The boxplots indicate the average IPS across the three TMB subgroups, treated with combination ICI therapy [CTLA-4 (+)/PD1 (+)] or with each ICI therapy, alone [CTLA-4 (+)/PD1 (–) or CTLA-4 (–)/PD1 (+)] or none [CTLA-4 (-)/PD1 (-)]. Similar IPS scores were found across all tumors, suggesting that ICI therapy can be applied independently of the patient’s TMB status.
Fig 2: The expression of CD8, PD-1, CTLA-4, IDO1, LAG3, HAVCR2, TIGIT, ILT2 and ILT4 was significantly higher in skin melanomas; whereas, C10orf54 (VISTA) and VTCN1, were expressed at markedly lower levels in the tumor samples compared to normal skin samples. Red asterisks (*) denote significant differences (|log2FC>1| and p<0.01) between skin melanomas from the TCGA-SKCM dataset and matched normal samples from TCGA and GTEx. One-way ANOVA, using disease state (skin melanoma or normal sample) was used to calculate differential expression. The expression data were first log2(TPM+1) transformed for differential analysis and the log2FC was defined as median (skin melanoma) - median (normal skin).
Fig 3: (A) Indicative immunohistochemistry (IHC) staining for the inhibitory receptors IDO1, PD-L1, PD-1, LAG3, CD8A/B (marker for cytotoxic T-cells) and FOXP3 (marker for Tregs) in an independent cohort of 11 cutaneous melanomas. H&E, hematoxylin and eosin staining. (B) Overall, the protein expression of these markers was either not detected (ND) or low and probably did not differ between TMBhigh and TMBlow tumors. (C) Immune-cell fractions across TMBhigh, TMBint and TMBlow skin melanomas, using extracted data (quanTIseq) from The Cancer Immunome Database. (D) The scatterplots depict the percentage of lymphocytes (%), average number of TIL patches and clusters (with standard deviation, SD) in TMBhigh (>30 mut/Mb) and TMBlow (<7.4 mut/Mb) skin melanomas. Neither of these differed significantly between the two subgroups of tumors. (E) The expression of CD8A (log2(TPM+1)) did not correlate with the neoantigen load in either TMB subgroup. Expression of CD8A, PDCD1 (PD-1), CD274 (PD-L1), PDCD1LG2 (PD-L2), IDO1 and CTLA-4 across different immune (F) and molecular (G) subtypes in skin melanoma. Immune subtypes: C1, wound healing (n=41); C2, IFN-gamma dominant (n=27); C3, inflammatory (n=14); C4, lymphocyte depleted (n=19); C5, immunologically quiet (n=0); C6, TGF-b dominant (n=2). Molecular subtypes: BRAFmut (n=150), NF1mut (n=27), RASmut (n=92), tripleWT (n=46).
Fig 4: Multiplex Immunohistochemistry and Cell Types Associated with Poor Response to Neoadjuvant Chemoradiotherapy. Representative overview of the T-cell antibody panel (A) and the TME panel (B), and exemplary images of cell types with association to tumor response to neoadjuvant chemoradiotherapy (C–E). Nuclei were counterstained with DAPI (blue). (A) T-cell panel: CD163 (cyan), CD4 (green), PD-1 (yellow), CD8 (orange), CD3 (red), FoxP3 (white); (B) TME panel: Pan-CK (cyan), aSMA (green), Vimentin (yellow), CD45 (orange), PD-L1 (red), Ki67 (white); (C) PD-1 positive macrophage; (D) PD-1 positive cytotoxic T-cell; (E) Cancer-associated fibroblast.
Fig 5: A: The distribution plot for the ratio of macrophages and Tregs shows individual samples colored according to the assigned K-means cluster. B: Parallel coordinate diagram of mean cluster values for macrophages and Tregs. Cluster 1 (red) is characterized by high Treg infiltration, cluster 2 (green) is characterized by high macrophage infiltration, cluster 3 (blue) has low Treg and low macrophage infiltration. C: Each line shows a representative case of the three clusters with HE-staining (magnification 7.2x) and multiplex immunohistochemistry for T-cell and TME markers: Nuclei were counterstained with DAPI (blue). T-cell panel: CD163 (cyan), CD4 (green), PD-1 (yellow), CD8 (orange), CD3 (red), FoxP3 (white); (B) TME panel: PanCK (cyan), aSMA (green), Vimentin (yellow), CD45 (orange), PD-L1 (red), Ki67 (white)
Supplier Page from MilliporeSigma for Anti-PDCD1 antibody produced in rabbit