Abstract 692P
Background
The treatment paradigm for clear cell renal cell carcinoma (ccRCC) has been transformed by the advent of immunotherapies. However, the composition and role of cells making up the ccRCC tumor immune microenvironment (TIME) has yet to be comprehensively described. Here, we leveraged a genomic data driven approach to characterize TIME subtypes in ccRCC.
Methods
Whole transcriptome data from patients with localized disease in the Cancer Genome Atlas KIRC (TCGA-KIRC) project was utilized (n=382). CIBERSORT was used for immune cell deconvolution, and unsupervised hierarchical clustering divided the cohort based on similar immune profiles. Survival of each cluster was analyzed, and gene set enrichment analysis was performed among clusters. The tumor immune dysfunction and exclusion (TIDE) tool, which uses a genomic signature validated on immunotherapy-treated melanoma patients to model tumor immune evasion, was then used to predict response to immune checkpoint blockade (ICB) in the clusters.
Results
A distinct M0hi cluster demonstrated decreased survival, more aggressive disease, and enrichment of epithelial to mesenchymal transition (EMT) hallmark genes [Enrichment Score = 0.69, p=0.001]. This cluster also showed an increase in myeloid derived suppressor cell (MDSC) and cancer associated fibroblast (CAF) gene signatures and a lower predicted response to ICB using the TIDE tool (Table). Table: 692P
| CL1 (M2mod) | CL2 (CD4 Memoryhi) | CL3 (M2hi) | CL4 (M0hi) | CL5 (CD8hi) | |
| Median OS (mo., 95% CI) | NR (77.0 – NR) | NR (90.5 – NR) | NR (84.3 – NR) | 45.3 (31.3 – NR) | 118.8 (93.0 – NR) |
| Median PFS (mo., 95% CI) | NR (89.9 – NR) | 123.8 (106.8 – NR) | NR (NR – NR) | 40.4 (20.1 – NR) | NR (NR – NR) |
| Stage III/IV (%) | 28 | 35 | 20 | 48 | 35 |
| PD-L1* | 0.01 | -0.01 | -0.07 | -0.93 p = 0.0001** | 0.30 |
| T-Cell Exclusion* | 0.19 | 0.21 | 0.32 | 0.69 p= 6.3x10-10** | -0.10 |
| Predicted ICB Response (%) | 27 | 23 | 20 | 4 | 34 |
| CAF* | 0.03 | 0.04 | 0.06 | 0.09 p= 2.2x10-16** | -0.03 |
| MDSC* | 0.02 | 0.02 | 0.02 | 0.05 p=6.3x10-5** | 0.02 |
NR: not reached CAF: Cancer Associated Fibroblast MDSC: Myeloid Derived Suppressor Cells ∗Signatures from TIDE tool, Z scores ∗∗Global p-value, Kruskal-Wallis
Conclusions
Comprehensive characterization of the TCGA-KIRC cohort led to the identification of a distinct cluster of ccRCC defined molecularly by decreased PD-L1 and increased EMT gene expression and cellularly by enrichment of M0 macrophages, CAFs, MDSCs, and an exclusion of T Cells. Future work will translate immune cells defining this cluster more broadly into a scoring paradigm that can be applied prospectively to better identify and treat early-stage patients with this aggressive TIME subtype.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
T. Morgan: Financial Interests, Personal, Funding: GenomeDx; Financial Interests, Personal, Funding: Myriad Genetics; Financial Interests, Personal, Funding: MDx Health; Non-Financial Interests, Personal, Advisory Board: TerumoBCT; Non-Financial Interests, Personal, Advisory Board: Myriad Genetics; Non-Financial Interests, Personal, Advisory Board: MDx. G. Palapattu: Financial Interests, Personal, Ownership Interest: Semper Therapeutics; Financial Interests, Personal, Research Grant: Minomic International Ltd; Financial Interests, Personal, Stocks/Shares: NantKwest; Financial Interests, Personal, Expert Testimony: Dupont; Financial Interests, Personal, Expert Testimony: Takeda Pharmaceuticals; Financial Interests, Personal, Invited Speaker: Astellas; Financial Interests, Personal, Advisory Role: Bristol Myers Squibb; Financial Interests, Personal, Advisory Role: Tokai Pharmaceuticals; Financial Interests, Personal, Advisory Role: Genome Dx. A. Udager: Financial Interests, Personal, Funding: Ventana Medical Systems. All other authors have declared no conflicts of interest.