Differential expression of immune-regulatory genes associated with PD-L1 display in melanoma: implications for PD-1 pathway blockade

JM Taube, GD Young, TL McMiller, S Chen… - Clinical Cancer …, 2015 - AACR
JM Taube, GD Young, TL McMiller, S Chen, JT Salas, TS Pritchard, H Xu, AK Meeker, J Fan…
Clinical Cancer Research, 2015AACR
Purpose: Blocking the immunosuppressive PD-1/PD-L1 pathway has antitumor activity in
multiple cancer types, and PD-L1 expression on tumor cells and infiltrating myeloid cells
correlates with the likelihood of response. We previously found that IFNG (interferon-
gamma) was overexpressed by tumor-infiltrating lymphocytes in PD-L1+ versus PD-L1 (−)
melanomas, creating adaptive immune resistance by promoting PD-L1 display. This study
was undertaken to identify additional factors in the PD-L1+ melanoma microenvironment …
Abstract
Purpose: Blocking the immunosuppressive PD-1/PD-L1 pathway has antitumor activity in multiple cancer types, and PD-L1 expression on tumor cells and infiltrating myeloid cells correlates with the likelihood of response. We previously found that IFNG (interferon-gamma) was overexpressed by tumor-infiltrating lymphocytes in PD-L1+ versus PD-L1(−) melanomas, creating adaptive immune resistance by promoting PD-L1 display. This study was undertaken to identify additional factors in the PD-L1+ melanoma microenvironment coordinately contributing to immunosuppression.
Experimental Design: Archived, formalin-fixed paraffin-embedded melanoma specimens were assessed for PD-L1 protein expression at the tumor cell surface with IHC. Whole-genome expression analysis, quantitative (q)RT-PCR, IHC, and functional in vitro validation studies were used to assess factors differentially expressed in PD-L1+ versus PD-L1(−) melanomas.
Results: Functional annotation clustering based on whole-genome expression profiling revealed pathways upregulated in PD-L1+ melanomas, involving immune cell activation, inflammation, and antigen processing and presentation. Analysis by qRT-PCR demonstrated overexpression of functionally related genes in PD-L1+ melanomas, involved in CD8+ T-cell activation (CD8A, IFNG, PRF1, and CCL5), antigen presentation (CD163, TLR3, CXCL1, and LYZ), and immunosuppression [PDCD1 (PD-1), CD274 (PD-L1), and LAG3, IL10]. Functional studies demonstrated that some factors, including IL10 and IL32-gamma, induced PD-L1 expression on monocytes but not tumor cells.
Conclusions: These studies elucidate the complexity of immune checkpoint regulation in the tumor microenvironment, identifying multiple factors likely contributing to coordinated immunosuppression. These factors may provide tumor escape mechanisms from anti–PD-1/PD-L1 therapy, and should be considered for cotargeting in combinatorial immunomodulation treatment strategies. Clin Cancer Res; 21(17); 3969–76. ©2015 AACR.
AACR