Immune signature as predictive marker for response to checkpoint inhibitor immunotherapy and overall survival in melanoma

Abstract

Background: Malignant melanoma is an immunogenic skin cancer with an increasing global incidence. Advanced stages of melanoma have poor prognoses. Currently, there are no reliable parameters to predict a patient's response to immune checkpoint inhibitor (ICI) therapy.

Methods: This study highlights the relevance of a distinct immune signature in the blood for response to ICI therapy and overall survival (OS). Therefore, the immune cell composition in the peripheral blood of 45 melanoma patients prior to ICI therapy was analyzed by flow cytometry and complete blood count.

Results: Responders to ICI therapy displayed an abundance of proliferating CD4⁺ T cells, an increased lymphocyte-to-monocyte ratio, a low platelet-to-lymphocyte ratio, low levels of CTLA-4⁺ Treg, and (arginase 1⁺ ) polymorphonuclear myeloid-derived suppressor cells (PMN-MDSC). Nevertheless, non-responders with similar immune cell compositions also benefited from therapy displaying increased long-term OS.

Conclusions: Our study demonstrated that the observed immune signature in the peripheral blood of melanoma patients prior to treatment could identify responders as well as non-responders that benefit from ICI immunotherapies.

Keywords: checkpoint inhibitor; immunotherapy; malignant melanoma; myeloid-derived suppressor cells; platelets.

FIGURE 1

Immunomonitoring of lymphocytes, monocytes, platelets, and the melanoma biomarkers, S100 and LDH. (A) Overall survival of responders and non‐responders to immune checkpoint inhibitor (ICI) therapy (Log‐rank (Mantel‐Cox) test, p = 0.0253 *). (B) Lymphocytes. The bar diagrams depict the percentage of lymphocytes in viable cells. Lymphocytes were identified in FSC‐A/SSC‐A plots via flow cytometry. Doublets and dead cells were excluded from analysis. The dashed line marks the median non‐responder value (two‐tailed Mann‐Whitney test, p = 0.0486 *). (C) Platelets, platelet‐to‐lymphocyte ratio (PLR), and survival analysis. Complete blood count (CBC) provided lymphocyte and platelet count to calculate PLR. The dashed lines mark the normal range (two‐tailed Mann‐Whitney test, p(platelets) = 0.0395 *, p(PLR) = 0.0552; Log‐rank (Mantel‐Cox) test, survival analysis, p(platelets, nR >271 vs. nR <271) = 0.0251 *, p(platelets, nR >271 vs. R) = 0.0006 ***, p(PLR, nR >289 vs. R) = 0.0028 **, p(PLR, nR >289 vs. nR <289) = 0.0353 *). (D) LDH, S100, and survival analysis. CBC provided data on LDH and S100 levels. The dashed lines mark the normal range (two‐tailed Mann‐Whitney test, p(LDH) = 0.0569, p(S100) = 0.2532); Log‐rank (Mantel‐Cox) test, survival analysis, p(LDH, nR >333 units/L vs. nR <333 units/L) = 0.0063 **, p(LDH, nR >333 units/L vs. R) = 0.0002 ***; p(S100, nR >0.1 μg/L vs. nR <0.1 μg/L) = 0.0188 *, p(S100, nR >0.1 μg/L vs. R) = 0.0004 ***). Medians with interquartile range. CTLA‐4, cytotoxic T‐lymphocyte–associated protein 4; ICI, immune checkpoint inhibitors; LDH, lactate dehydrogenase; nR, non‐responder; PD‐1, programmed cell death protein 1; R, responder; S100, S100 protein

FIGURE 2

Immunomonitoring of CD4⁺ and CD8⁺ T cells in the peripheral blood. (A) CD4⁺ and CD8⁺ T cells (two‐tailed Mann‐Whitney test, p(CD4⁺ T cells) = 0.6274, p(CD8⁺ T cells) = 0.3274). (B) CTLA‐4⁺ CD4⁺ and CD8⁺ T cells (p(CTLA‐4⁺CD4⁺ T cells) = 0.0319 *, p(CTLA‐4⁺CD8⁺ T cells) = 0.0615). (C) PD‐1⁺ CD4⁺ and CD8⁺ T cells (p(PD‐1⁺CD4⁺ T cells) = 0.1550, p(PD‐1⁺CD8⁺ T cells) = 0.3431). (D) Proliferating (Ki‐67⁺) CD4⁺ and CD8⁺ T cells (p(Ki‐67⁺CD4⁺ T cells) = 0.0420 *, p(Ki‐67⁺CD8⁺ T cells) = 0.2065). (E) Survival analysis of patients depending on CD4⁺ and CTLA‐4⁺CD4⁺ T cells (Log‐rank (Mantel‐Cox) test, CD4⁺ T cells, p(nR >62.8% vs. R) = 0.0063 **, p(nR <62.8% vs. R) = 0.0251 *; CTLA‐4⁺CD4⁺ T cells, p(nR >13.9% vs. R) = 0.0025 **). Medians with interquartile range. CTLA‐4, cytotoxic T‐lymphocyte–associated protein 4; nR, non‐responder; PD‐1, programmed cell death protein 1; R, responder

FIGURE 3

Immunomonitoring of regulatory T cells (Treg). (A) Treg. The bar diagram depicts the percentage of CD4⁺ CD25⁺CD127^lowFoxp3⁺ Treg in CD4⁺ T cells (two‐tailed Mann‐Whitney test, p = 0.1681). (B) CTLA‐4⁺ Treg (p = 0.2101). (C) GARP⁺ Treg (p = 0.4525). (D) HLA‐DR⁺ Treg (p = 0.2154). (E) Survival analysis of patients depending on GARP⁺ and HLA‐DR⁺ Treg (Log‐rank (Mantel‐Cox) test, GARP⁺ Treg, p(nR >25.1% vs. R) = 0.0011 **; HLA‐DR⁺ Treg, p(nR >53.5% vs. nR <53.5%) = 0.0069 **, p(nR <53.5% vs. R) = 0.0006 ***). Median with interquartile range. CTLA‐4, cytotoxic T‐lymphocyte–associated protein 4; Foxp3, forkhead protein 3; GARP, glycoprotein A repetitions predominant; HLA‐DR, human leukocyte antigen‐DR isotype; nR, non‐responder; PD‐1, programmed cell death protein 1; R, responder; Treg, regulatory T cells

FIGURE 4

Immunomonitoring of myeloid‐derived suppressor cells (MDSC). Analysis distinguished CD15⁻CD14⁺CD33^highHLA‐DR^low mononuclear/monocytic MDSC (M‐MDSC), CD15⁺CD33⁺ polymorphonuclear MDSC (PMN‐MDSC), and HLA‐DR^lowCD11b⁺CD33⁺ early MDSC (E‐MDSC). The MDSC gating strategy is depicted in Figure S2A. (A) MDSC (two‐tailed Mann‐Whitney test, p(E‐MDSC) = 0.6658, p(M‐MDSC) = 0.9872, p(PMN‐MDSC) = 0.1042). (B) Arginase 1⁺ MDSC (p(arginase 1⁺ E‐MDSC) = 0.9630, p(arginase 1⁺ M‐MDSC) = 0.8405, p(arginase 1⁺ PMN‐MDSC) = 0.7900). (C) Survival analysis of patients depending on (arginase 1⁺) PMN‐MDSC (Log‐rank (Mantel‐Cox) test, PMN‐MDSC, p(nR >0.5% vs. R) = 0.0064 **, p(nR <0.5% vs. R) = 0.0422 *; arginase 1⁺ PMN‐MDSC, p(nR >51.5% vs. R) = 0.0018 **). Median with interquartile range. CTLA‐4, cytotoxic T‐lymphocyte–associated protein 4; E‐MDSC, early MDSC; HLA‐DR, human leukocyte antigen‐DR isotype; M‐MDSC, monocytic/mononuclear myeloid‐derived suppressor cells; nR, non‐responder; PD‐1, programmed cell death protein 1; PMN‐MDSC, polymorphonuclear MDSC; R, responder