2246P Tumor-infiltrating Lymphocites and Gene Expression Across Multiple Cancer-Types: A Translational Analysis from the SOLTI-1904-ACROPOLI Study

Tumor-infiltrating lymphocytes (TILs) are a mixed population of immune cells that infiltrate the tumor microenvironment (TME) and play a crucial role in recognizing and attacking cancer cells. The presence, localization and composition of TILs can be used as a biomarker for predicting clinical outcomes and response to immunotherapy in different types of cancer. Our aim was to investigate the correlation between the presence of TILs and gene expression across multiple cancer types. Patients (pts) with advanced solid cancer who participated in the molecular prescreening of the SOLTI-1904-ACROPOLI study (NCT04802876) were included. A pathological analysis was performed to determine the proportion of TILs (%TILs) on a hematoxylin-eosin section. On the same FFPE tumor block, gene expression (GE) was evaluated using a panel of 72 genes, including 10 immune genes (i.e., CD19, CD274, CD3, CD4, CD68, CD8, CD84, CD86, CXCL8 and PDCD1). An exploratory analysis of the association between TILs and GE was performed. 1,003 pts were included across 39 different cancer-types. TILs were successfuly evaluated in 922 pts (91.9%). Median (m) %TILs was 5% (IQR: 2-15, range: 0-80). Cancer-types with higher %TILs were duodenal (m=25%, IQR: 20-27.5), head and neck (m=22.5%, IQR: 9-30) and cervical carcinoma (m=20, IQR: 3-35). %TILs was higher in samples from primary tumors when compared to metastatic samples (m=6 vs m=3%, p<0.001). Significant differences of %TILs were also seen according to the site of metastasis (M1): m=8% in lymph nodes vs m=7.5% en lung M1 vs m=3% in liver M1 (p<0.001). A quantitative SAM analysis identified 31 genes significantly upregulated in tumors with higher %TILs (FDR<1%). Among the top 10 most upregulated genes, nine were immune (i.e., CD8, PDCD1 and CD19) and the remaining one MYC. However, correlation between TILs and immune genes was moderate (I,e., coefficient=0.49 between CD8 and TILs). Immune genes or signatures may provide additional information not captured by TILs alone. Combining both measures could provide a more comprehensive understanding of the immune response in the TME and help predict response to immunotherapy in cancer patients.