School of Medicine and Health Sciences Poster Presentations
Poster Number
170
Document Type
Poster
Keywords
Cancer; Oncology; Bioinformatics; Structural Biology
Publication Date
Spring 2017
Abstract
While BRAF is among the most well-established oncogenes in human cancers, more recently it has garnered attention for its role in suppressing antitumor immunity, especially in melanoma. Because tumor-infiltrating lymphocyte (TIL) density is strongly prognostic in colorectal cancer (CRC)7, we decided to investigate the connection between TIL density and the BRAF-activating V600E mutation in CRC.
We used ESTIMATE to quantify immune infiltrate in samples from the TCGA colon adenocarcinoma (COAD) dataset (n = 216). This is an algorithm that uses the gene-expression signature of 141 immune-related genes to infer the presence of immune cells in the tumor infiltrate. COAD samples with BRAF V600E mutations have more immune infiltrate than wild-type samples (p < 0.05, Figure 1).
The MSI-H subtype of CRC is characterized by epigenetic silencing of DNA mismatch-repair genes and higher mutational load, neoantigen expression and TIL density. Since it has been previously established that BRAF mutation is highly correlated with MSI-H phenotype3, we performed ANOVA to assess if BRAF mutation predicts ESTIMATE score independently of MSI-H phenotype. While the BRAF term was not significant in this multivariate linear model, we included an interaction term that was significant (p < 0.005), suggesting that both variables contribute to ESTIMATE score. To determine the effect of the interaction, we grouped samples into MSI-H and MSS groups and compared the wild-type and BRAF V600E subgroups within them [Figure 2]. In the MSS group only, immune scores were significantly higher in the BRAF-V600E subgroup (p = 0.05, Figure 2).
Next, we sought to reproduce our results in an independent cohort of 619 CRC samples from the BROAD institute that have immune scores based on immunostaining9. We grouped samples into TIL-positive and TIL-negative groups and compared the numbers of BRAF-V600E and BRAF-WT samples within each group. We observed a significant correlation between BRAF mutation and TIL+ immune score within the MSS group (p = 0.01352, odds ratio 5.76) [Figure 3]. Notably, the MSS group is the same group from the TCGA cohort in which we observed a significant correlation between BRAF mutation and ESTIMATE score.
The analysis we present here reveals a significant correlation between the BRAF V600E mutation and increased immune infiltrate in MSS tumors across two independent CRC datasets. Because V600E is an activating mutation, this finding suggests an immunity-stimulating role for BRAF in CRC, in contrast to BRAF’s role in melanoma, which appears largely immunosuppressive.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Open Access
1
Mutations in BRAF Are Associated With Higher Levels of Immune Infiltrates in Microsatellite-Stable Colon Cancer
While BRAF is among the most well-established oncogenes in human cancers, more recently it has garnered attention for its role in suppressing antitumor immunity, especially in melanoma. Because tumor-infiltrating lymphocyte (TIL) density is strongly prognostic in colorectal cancer (CRC)7, we decided to investigate the connection between TIL density and the BRAF-activating V600E mutation in CRC.
We used ESTIMATE to quantify immune infiltrate in samples from the TCGA colon adenocarcinoma (COAD) dataset (n = 216). This is an algorithm that uses the gene-expression signature of 141 immune-related genes to infer the presence of immune cells in the tumor infiltrate. COAD samples with BRAF V600E mutations have more immune infiltrate than wild-type samples (p < 0.05, Figure 1).
The MSI-H subtype of CRC is characterized by epigenetic silencing of DNA mismatch-repair genes and higher mutational load, neoantigen expression and TIL density. Since it has been previously established that BRAF mutation is highly correlated with MSI-H phenotype3, we performed ANOVA to assess if BRAF mutation predicts ESTIMATE score independently of MSI-H phenotype. While the BRAF term was not significant in this multivariate linear model, we included an interaction term that was significant (p < 0.005), suggesting that both variables contribute to ESTIMATE score. To determine the effect of the interaction, we grouped samples into MSI-H and MSS groups and compared the wild-type and BRAF V600E subgroups within them [Figure 2]. In the MSS group only, immune scores were significantly higher in the BRAF-V600E subgroup (p = 0.05, Figure 2).
Next, we sought to reproduce our results in an independent cohort of 619 CRC samples from the BROAD institute that have immune scores based on immunostaining9. We grouped samples into TIL-positive and TIL-negative groups and compared the numbers of BRAF-V600E and BRAF-WT samples within each group. We observed a significant correlation between BRAF mutation and TIL+ immune score within the MSS group (p = 0.01352, odds ratio 5.76) [Figure 3]. Notably, the MSS group is the same group from the TCGA cohort in which we observed a significant correlation between BRAF mutation and ESTIMATE score.
The analysis we present here reveals a significant correlation between the BRAF V600E mutation and increased immune infiltrate in MSS tumors across two independent CRC datasets. Because V600E is an activating mutation, this finding suggests an immunity-stimulating role for BRAF in CRC, in contrast to BRAF’s role in melanoma, which appears largely immunosuppressive.
Comments
Poster presented at GW Annual Research Days 2017.