Q: What are some of the biggest challenges in pancreatic cancer?
Dr. Hwang: I would start with the unique tumor microenvironment. Stromal cells create high interstitial pressure, which can limit the efficacy of drug therapy and create a hypoxic environment. Hypoxia is one reason we think that radiotherapy isn’t very effective in pancreatic cancer.
Most patients do not respond to immunotherapy and we see an abundance of M2-polarized macrophages and myeloid-derived suppressor cells. Of the T cells that are present, many lack the markers of activation and exhaustion that are associated with response.
On the detection side, we lack reliable strategies to detect pancreatic cancer early. Late detection puts us behind the eight ball when trying to control this aggressive disease.
Q: What are the major areas of focus for your laboratory?
Dr. Hwang: We look at the interface between cell-intrinsic and -extrinsic mechanisms of therapeutic resistance. We use single-cell technologies to identify intrinsic cancer cell features and gene expression programs that play a functional role in mediating resistance to chemotherapy and in modulating the immune system to drive resistance to immunotherapies.
At the same time, we leverage spatial technologies to look at the interactions between specific cancer cell subtypes and their local microenvironment to understand how those interactions lead to emergent properties like metastasis and therapeutic resistance.
Q: What approaches are important in your studies?
Dr. Hwang: We identify cell states in patient tumors using single-cell transcriptomics and examine how the proportions of distinct cell states change with treatment strategies. However, we are increasingly interested in understanding the determinants of cell states, and have turned to single-cell multiomic approaches wherein we combine transcriptome information with concurrent genomic and/or epigenomic data. Cancer cells do not exist in isolation, and much of their biological behavior is driven by interactions with other cells in the tumor microenvironment. To dissect these interactions, we are also using spatial proteomics and transcriptomics.
We typically start our research from patient-derived specimens to ensure that the biology we discover is relevant to human disease. As we are limited in our ability to understand the mechanisms underlying these behaviors, it is critical that we test our human findings in model systems. For the latter, we use genetically engineered mouse models as well as 3D co-cultures of various tumor cell types to mimic the microenvironment.
Q: How are you implementing single-cell approaches into your work?
Dr. Hwang: In an early example of our single-cell work, we compared two cohorts of pancreatic cancer patients. One cohort received treatment before surgery, while a second cohort, with comparable stage and grade, did not. We then compared tumor specimens at the single-cell level to identify each cell type and state and their proportions in each cohort. One of the cancer cell states, which we termed “neural-like progenitor,” was strongly enriched in the treated cohort, suggesting a potential role in treatment resistance and/or tumor proliferation/regeneration after therapy. These findings led to experiments that allowed us to determine if certain genes play a role in treatment resistance, which is paramount in our efforts to improve treatment options and patient outcomes.
Q: How do you foresee combined single-cell whole genome and transcriptome solutions impacting oncology research?
Dr. Hwang: We’ve seen some early applications of combined single-cell genome and transcriptome approaches, but they have significant limitations. BioSkryb’s ResolveOME, which provides unified genomic and transcriptomic data in the same individual cells at the highest resolution to date, is for the first time opening a window into identifying and tracking cancer cell clones without inference.
This ability to do lineage tracing with single-cell genomic information and concurrent cell state identification with single-cell transcriptomics allows us to look how metastases are derived from the primary tumor and the genetic contributions to cell state. Now we can ask: “How heterogeneous are the cells that form the metastasis in terms of clonality and cell state? What does this tell us about the evolution of treatment resistance?”
The combination of having a comprehensive readout of cell state with concurrent clonal tracking is powerful. I can imagine this playing a big role in helping us understand how different cancers develop and evolve under treatment. At the end of the day, the question becomes, “Which components of the cell states that are able to survive our therapies are hardwired into the genome as opposed to representing more plastic epigenetic and phenotypic changes?” The way you would target these processes would be very different.
William L. Hwang, MD, PhD is a physician-scientist in the Center for Systems Biology, Center for Cancer Research and Department of Radiation Oncology at the Massachusetts General Hospital; associate faculty at the Broad Institute of MIT and Harvard; and assistant professor at Harvard Medical School. He specializes in the treatment of gastrointestinal cancers, and his research program is focused on elucidating the interactions between pancreatic cancer cells and their microenvironment at high resolution using single-cell and spatial genomics, genetically-engineered mouse models, and 3D organoid co-cultures.
With its novel combination of single-cell methods, BioSkryb Genomics is unlocking a deeper understanding of tissue health. To learn more, visit www.bioskryb.com