Immune checkpoint inhibitors have changed the treatment of melanoma, but many tumors do not respond or eventually develop resistance. This problem is especially urgent in acral and mucosal melanomas, rare subtypes with distinct biology and poorer responses to current immunotherapies.
The Roy Lab studies how tumor-intrinsic programs and the spatial and temporal organization of the tumor immune microenvironment create resistance. Our research program follows three connected pillars: mapping tumors, modeling treatment responses, and matching tumors with therapies.
Acral melanoma arises on the palms, soles, and nail units. Mucosal melanoma arises from melanocytes in mucosal sites. Both have been historically understudied and remain difficult to treat. We seek to understand the biology that distinguishes these tumors, determine why immune responses fail, and identify strategies that can improve treatment.
Dr. Roy also leads an international rare skin cancer biobank based at the McGill University Health Centre. The biobank supports collaborative studies of uncommon tumors that no single center can investigate at sufficient scale.
Pillar 1
What creates immunotherapy resistance in rare melanomas?
We map the tumor immune microenvironment using digital pathology, spatial multiomics, and cancer immunology. These approaches show which cells are present, what molecular programs they express, and how they are organized within tissue.
Our computational work integrates tissue morphology with spatial transcriptomics, single-cell RNA sequencing, ATAC-seq, and other genomic and epigenomic measurements. We use machine learning to identify cellular neighborhoods, connect histologic patterns with molecular states, and study features associated with immune exclusion, treatment response, and clinical outcomes.
We favor interpretable models that can be examined by pathologists and biologists. The goal is to produce measurements that support biological discovery and can eventually be tested in clinically relevant cohorts.
TODO(simon): one to three approved images with mono captions
Pillar 2
What can living tumor tissue reveal that a static measurement cannot?
Tumors are dynamic systems, but most tissue measurements capture only one moment. We use organotypic precision-cut tumor slice cultures to study how living patient tissue responds to therapy over time. These cultures preserve tissue architecture and important components of the tumor immune microenvironment while allowing controlled perturbation outside the patient.
We combine slice cultures with live imaging to track immune-cell migration, tumor-immune interactions, cell death, and treatment response. Related studies in mouse models extend these observations to intact biological systems. Together, these approaches reveal response trajectories that may be missed by static baseline or endpoint measurements.
Machine learning and quantitative image analysis help us extract patterns from time-resolved imaging and connect dynamic behavior with spatial and molecular readouts. This creates a functional view of treatment response that can be compared across patient samples, experimental systems, and therapies.
TODO(simon): one to three approved images with mono captions
Pillar 3
Which therapies can overcome resistance?
We connect biological discovery with functional therapeutic testing. Patient-derived slice cultures and complementary models allow us to compare immunotherapies, antibody-drug conjugates, emerging immune-modulating agents, and rational drug combinations while monitoring their effects on the tumor immune microenvironment.
We integrate these results with pathology, spatial multiomics, live imaging, and clinical annotation. This framework helps us identify biomarkers of response, determine which mechanisms are reproducible across model systems, and prioritize therapies for preclinical development. Our longer-term aim is to support precision oncology strategies and the design of phase I clinical trials grounded in tumor biology.
TODO(simon): one to three approved images with mono captions
We study how epigenetic programs shape melanoma cell states and immune exclusion, with a focus on rare melanomas. This includes work on DNA methylation, chromatin accessibility, alternative splicing, and other regulatory mechanisms that may create therapeutic vulnerabilities.
Our V Foundation-supported work examines the alternative splicing and isoform-specific functions of BRD4 in acral melanoma. We are investigating how distinct BRD4 isoforms regulate melanoma cell states, tumor-immune interactions, and resistance to immune checkpoint blockade.
We use live imaging to study how T cells and myeloid cells migrate, interact with tumor cells, and change during immunotherapy. Integrating dynamic imaging with spatial and molecular measurements allows us to study immune responses in both space and time.
We develop machine-learning approaches that use tissue morphology to identify and predict cellular niches within the tumor immune microenvironment. We then link these niches to molecular programs, immune-cell states, therapeutic response, and clinical outcomes.
The Roy Lab is part of PathSpace, a Montreal-based collaborative network of pathologist-scientists interested in spatial pathobiology. Dr. Simon Roy, Dr. Basile Tessier-Cloutier, and Dr. Vincent Quoc-Huy Trinh meet monthly with their teams to discuss data, methods, and shared research questions.
Our projects connect pathology, oncology, experimental biology, computation, and imaging. When scientifically aligned, trainees may also contribute to collaborations with medical technology and pharmaceutical partners. These experiences can support careers in academic research, biotechnology, pharmaceutical development, or data science.