DVincent Quoc-Huy Trinh and his team use imaging and cancer tissue characterization techniques to develop new diagnostic and therapeutic tools. The techniques are centered on his training as a pathologist and medical knowledge associated with various visual manifestations identified in tumors in patients and animals. His team uses multiplex imaging, secretome studies, artificial intelligence applied to images, transgenic mouse models, and spatial transcriptomics, among other techniques.

Research Theme

Liver and pancreatic cancers have very low survival rates, among the lowest of all cancers affecting patients. This low survival rate is due to the lack of early diagnostic tools, leading to advanced stages of the disease at the time of diagnosis. Our approach is to develop diagnostic and therapeutic tools to identify the disease early and treat it at this stage. If the disease is detected and treated at an earlier stage, the survival rate approaches 100%.

Unfortunately, there is no treatment for these early-stage diseases other than morbid and sometimes fatal surgery. We therefore need to 1) model the disease at this early stage, 2) identify signaling pathways that maintain early tumor cells, and 3) develop therapies specific to this stage of the disease.

Instead of targeting tumor cells directly, we are targeting the many cells found around them, particularly fibroblasts. We have demonstrated that fibroblasts release large amounts of molecules that increase cancer progression. Our goal is therefore to identify these molecules and block their pro-tumor action.

Research objective

The first objective is to analyze the molecules released by fibroblasts and their function on cancer cells. We participated in the discovery of a new category called “supermere” and are demonstrating the crucial importance of this category in signaling between fibroblasts and tumor cells.

The second objective is to create an early-stage pancreatic tumor model with fibroblast depletion through adoptive transfer of cytotoxic T lymphocytes. We previously created the most effective fibroblast depletion model published in the literature and are now testing it in pancreatic tumors. Once fibroblasts are eliminated from the tumor, we will look at the transcriptomic pathways and proteins deregulated by the absence of fibroblasts. Thus, by negativity, we can understand how fibroblasts maintain cancer cells.

The third objective involves creating imaging tools for liver and pancreatic cancer tissue using multiplex imaging methods, training deep learning (artificial intelligence) algorithms for cell classification and detection, and deploying multispectral methods such as spatial transcriptomics. These methods are used in particular to identify tumor cells central to the initiation and progression of the disease, in addition to complementing the other study objectives.