Discovering New Biomarkers Using Spatial Approaches

Spatial approaches to biomarker discovery are manifold, but all share the examination of biomarkers in a spatial context within cells and tissues. The burgeoning field of spatial biology encompasses many methods—such as spatial transcriptomics and proteomics, single-cell RNA-seq, next-generation sequencing (NGS), and fluorescence in situ hybridization—and is a potent investigative tool in the search for new biomarkers. Such techniques, coupled with human creativity, is accelerating the pace of biomarker discovery today. Here’s a brief glimpse into the busy world of biomarker discovery using spatial biology approaches.

High-throughput, hypothesis-free discovery

For high-throughput biomarker discovery, 10x Genomics’ Visium platform allows for spatial, whole transcriptome analysis across continuous areas of tissue sections. Nucleic acids are targeted on the spatial array with spatially barcoded capture probes, then analyzed by NGS. The high throughput of NGS supports a discovery approach, where you can analyze the whole transcriptome across the whole tissue section, or a portion of it. “This allows for hypothesis generation, where researchers can capture transcriptomic information in a sample of interest, look at some unbiased preliminary analyses like clustering analysis to get a sense for what's happening in the tissue overall, and then probe different signatures within the spatial context,” says Jacob Stern, Director of Product Management at 10x Genomics.

Researchers from the University of Texas MD Anderson Cancer Center recently used Visium to analyze samples from patients with an aggressive type of ovarian cancer, a subpopulation of which can exceed survival expectations. “This is a good example of what you can do with Visium, because they looked across a broad swath of biopsies taken from patients at tumor-stroma interactions, comparing short-term and long-term survivors,” says Stern. They found that a specific subtype of cancer-associated fibroblasts, and its spatial location relative to a subtype of ovarian cancer cell in the tumor microenvironment (TME), correlated with long-term survival. Then they validated their findings using multiplex immunofluorescence.

The results suggest that signaling networks in the TME, namely cross-talk between cancer-associated fibroblasts and tumor cells, play a role in the long-term survival of these cancer patients. “The fact that the researchers can go in and analyze everything everywhere all at once allowed them to do this hypothesis-free biomarker discovery, finding indicators of what could be happening in these two physiologically different patient populations,” says Stern. “And they came up with signatures that you wouldn't have necessarily thought to look for.”

Leveraging heterogeneity

Zeroing in on heterogeneity among cells can also point researchers toward new biomarkers. “Cancer cells are notoriously heterogeneous at genetic and also phenotypic levels,” says Ilayda Hasakiogullari, Global R&D Product Manager at CellCarta. “Understanding this heterogeneity is key to designing truly personalized therapy approaches, which we strive for at CellCarta.” Discovering new biomarkers is trickier in smaller cell populations—with fewer cells the heterogeneity among them can be missed or drowned out using bulk analysis methods,  which also omit essential contextual information. “Cells do not act alone—tissue context is critical to understand the biology of the disease and response or resistance to therapy,” says Hasakiogullari. “Spatially resolved transcriptome and proteome information can enlighten us further on cellular interactions, state, and function.”

In a recent collaboration, CellCarta used spatial transcriptomics to stratify cancer patients based on response to treatment with neoadjuvant immunotherapy, followed by tumor removal. “We used spatial transcriptomics to investigate heterogeneity in the pre-treatment samples to discover biomarkers that might indicate a good response after the immunotherapy,” says Hasakiogullari. “We also tested the post-treatment samples to analyze and compare ‘leftover’ tumor cells, which were missed by bulk methods due to high background signals.” Using spatial transcriptomics results, they categorized patients based on full or partial response to therapy. “In another study, using a whole transcriptome spatial transcriptomics assay, we were able to detect and characterize different tumor clones in non-responder patients that were probably missed with standard IHC assays,” she says. “Both results will be published later this year.”

Single-cell spatial analysis

Resolve Biosciences’ spatial analysis platform enables both multiplexed, multi-omic, and single-cell studies. “With single-molecule detection capabilities and assay design built for focused studies of relevant biomarkers, Resolve Biosciences’ Molecular Cartography helps scientists understand what is real and what is noise,” says Kieren Patel, Chief Business Officer at Resolve Biosciences. “This helps speed biomarker discovery by improving data interpretation and overcoming limitations of other single-cell technologies.”

Researchers from the University of Pittsburgh recently used Resolve Biosciences’ platform to validate a biomarker and potential therapeutic target for liver regeneration, for which no targeted therapies are currently available. The Wnt-β-catenin signaling pathway is known to be important in metabolic liver zonation and liver regeneration after injuries. Using Molecular Cartography, the researchers mapped metabolic zones in the liver, and showed rescued liver zonation and regeneration in knockout mice using an antibody that activates Wnt signaling. “This was only possible through an assay that provided spatially resolved, single-cell analysis with the specificity and sensitivity of the Resolve Biosciences system,” says Patel. “This study shows the power of Molecular Cartography technology to examine the levels and spatial organization of this biomarker in specific tissue environments in the liver.”

Spatial biomarkers

Initial efforts to predict responses to cancer immunotherapies using the absence or presence of a biomarker were generally lukewarm. “We know today that the mere presence of a biomarker is not sufficient to determine whether a patient will respond to a treatment or have a good outcome,” says Oliver Braubach, Director of Applications at Akoya Biosciences. “[However,] spatial biology changes the definition of a biomarker. The notion of a spatial biomarker denotes not just the presence of a molecule, but a pattern of cells expressing certain biomarkers within a tissue."

Researchers from Gary Nolan's lab at Stanford using Akoya’s platform found that distinct enrichment of certain cell types exists in unique cellular neighborhoods, which can be causative or suppressive of cancer. “The presence and quantity of these cellular neighborhoods is a nice predictive biomarker that outperforms conventional biomarker analyses that look at just the presence of a molecule,” says Braubach. “They’ve also shown that cellular neighborhoods identified on our platform are better at predicting whether a cancer patient will respond to immunotherapy, compared to conventional biomarker assays.”

Spatial biomarkers using Akoya’s platform are also being used as a companion diagnostic for a cancer drug in clinical trials, to determine whether the cancer drug will be effective. “If that gets FDA approval, it will be one of the first times in which spatial biomarkers are used to determine whether a patient should receive a particular, treatment,” says Braubach. But he’s even more excited to use the power of Akoya’s spatial biomarker platform to re-analyze and “rescue” failed drugs—drug candidates in which companies invested huge amounts of resources, but which failed to win FDA approval. “There's a good reason why the drugs should work, but we didn't have assays then that were able to tell us in which patients they would work,” he says. Now with the increased acuity of spatial biology, we’re better able to figure out which patient will benefit from this over another one.”

Biomarkers that predict response to cancer immunotherapy continue to be a fierce area of research. Immunotherapies work remarkably well for a lucky subset of patients, but matching therapy to patient remains a huge challenge. “That really is the magic trick right now, finding the right therapy for each particular tumor,” says Braubach. “Immunotherapy is so targeted and sensitive that it depends on this nuance, and we think spatial biology can do this better than anything else right now.”