Exploratory Unsupervised Analysis of HCC Mass Spectrometry Imaging Data for Biomarker Discovery

Hemi Babu

Mentor: Claire L. Carter, Ph.D, Assistant Member at Hackensack Meridian Center for Discovery and Innovation, Assistant Professor at Hackensack Meridian School of Medicine, Co-Director at MSAP Shared Resource, Lombardi Comprehensive Cancer Center, Georgetown University

Date/Time: April 29th, 2025 at 12:00pm

Abstract: Hepatocellular carcinoma (HCC) is a clinically aggressive and metabolically diverse liver malignancy characterized by pronounced spatial heterogeneity, and is the second leading cause of cancer-related deaths wordwide Treatment for unresectable HCC often involves locoregional therapies such as transarterial chemoembolization (TACE), but to date there is limited evidence of improved survival outcomes following treatment. A greater understanding of both HCC and TACE response within the tumor microenvironment is needed to improve clinical outcomes. Mass Spectrometry Imaging (MSI) has emerged as a powerful tool for spatial metabolomics and lipidomics, offering high-resolution, label-free molecular profiling of tissue microenvironments. This project focuses on the analysis of MSI data acquired from a preclinical model of HCC liver tissue treated with caffeic acid- based TACE in a lipiodol-based emulsion. Data generated using a Bruker Solarix MALDI FT-ICR MS in negative ion mode, yielded over 8 million m/z features across ~39,000 spatial coordinates. The data processing and analysis will be performed using the Cardinal package in R, which provides specialized statistical and spatial tools tailored for MSI datasets. 

Through unsupervised machine learning and spatial segmentation, we aim to identify molecular signatures—particularly metabolites and lipid species—with localized expression patterns relevant to tumor heterogeneity, therapeutic targeting, and oxidative stress pathways. Emphasis will be placed on spatially enriched ions that may reflect drug penetration into the heterogenous tumor, treatment response or biochemical remodeling post-TACE. This work underscores the potential of computational MSI analysis in advancing biomarker discovery and mechanistic insight in liver cancer research.