Bioinformatics Analysis of Tumor Infiltrating Immune Cell Subsets in Hepatocellular Carcinoma

Bioinformatics Internship Presentation

Ushna Ahmad (Mentor: Dr. Yuriy Gusev, Innovation Center for Biomedical Informatics, Department of Oncology, Georgetown University)

August 29, 2017, 2:30pm, Room 341, Basic Science

In recent years, immunotherapy has become an important method in treating cancer patients. However, cancer immune editing and heterogeneity within the tumor, between cancers, and between patients has made this a difficult task. By performing molecular profiling of the tumors, it is possible to learn which types of immune cells are present in or nearby the tumors, providing insight into the tumor microenvironment and the impact of infiltrating immune cells.

Hepatocellular carcinoma (HCC) is the second most common cause of cancer death around the world. This project focused on molecular profiling of 39 clinical samples of HCC tumors collected from MedStar Georgetown University Hospital and determining the estimation of fractions of infiltrating immune cells present. This was accomplished by running a pipeline for RNAseq data processing using the Cancer Genome Cloud (CGC), which is powered by the Seven Bridges platform. The gene expression data was then analyzed using Cell type Identification By Estimating Relative Subsets Of known RNA Transcripts (CIBERSORT), a free online tool that quantifies the relative expression of 22 types of immune cell specific signatures by evaluating tumor samples’ gene expression profiles (GEPs).

Patient results were statistically compared to previous similarly processed results for an HCC dataset from The Cancer Genome Atlas (TCGA) database as well as 40 clinical samples for colorectal cancer (CRC). The results were visualized in various plots to compare the molecular profile of patients in the HCC dataset, between datasets, and between cancers using R and Multi-Experiment Viewer (MeV).  Finally, using gene expression data, patients were stratified into groups based on phenotypic information for key biomarkers, including IFNG, GZMA, GZMB, PRF1, IL12, and IL18, and the molecular profiles for each were explored.

This project is important for understanding the tumor-immune cell environment for these patients, and provides information to clinicians to better select targeted therapy options. Among the top five cell types present in the tumor samples are macrophages (M2 and M0). Current research for immunotherapy primarily focuses on lymphocytes (T and B cells), but my results suggest exploring targets for tumor-associated macrophages (TAMs) is also critical, especially for liver cancer, where they are the dominating cell type. These patients will undergo immunotherapy and their tissue samples will be collected for a follow-up analysis to compare with data before therapy.