Studies in recent years have established that the spatial arrangements of cells hold information crucial to understanding immunotherapy response. In a study published in the Journal for ImmunoTherapy of Cancer (JITC), Lu et al. found that multiplex imaging-based approaches like multiplex immunohistochemistry (mIHC) and immunofluorescence (mIF) had much higher predictive power compared to single-plex PD-L1 IHC or single-cell sequencing.
Those findings were reflected at this year’s Society for Immunotherapy of Cancer (SITC) Annual Meeting. Aside from our excitement at physically attending our first major conference since the COVID-19 pandemic began, we were awed by the output of research recognizing the power of spatial biology for immunotherapy.
There were over a dozen abstracts submitted to SITC which featured Akoya’s technology alone. We were also pleased to host a well-attended dinner symposium that brought together leaders from the academic, pharma, and contract research organization (CRO) spaces for a discussion on the importance of spatial biomarkers in this space.
Symposium panelists and moderator. Left to right: Michelle Poulin, PhD, Akoya Biosciences; Michael Surace, PhD, AstraZeneca; Houssein Sater, MD, Cleveland Clinic Martin Health; Qingyan “Sandy” Au, PhD, Neogenomics.
An introduction from Akoya CEO Brian McKelligon.
Single-cell, spatial context is essential for decoding the tumor microenvironment
The advent of immuno-oncology has led to a dramatic transformation in the way we approach cancer treatment. “The therapeutic agent is no longer the drug that you’re putting in the patient,” noted one symposium panelist, Dr. Michael Surace, Associate Director at AstraZeneca.
As the effectiveness of immunotherapies can vary widely between patients, researchers have increasingly focused their efforts on finding biomarkers to distinguish the responders from the non-responders, so the right therapy reaches the right patient.
Techniques like single-plex immunohistochemistry and next-generation sequencing have helped make progress in this area, identifying biomarkers such as PD-L1, tumor mutational burden, and microsatellite instability. Yet these approaches cannot capture the full complexity of the tumor microenvironment.
“Spatial biology is going to be able to help us dissect what is really important”
Scientists often make general classifications about the tumor microenvironment based on levels of immune activity, such as “immune desert”, “immune-excluded”, and “inflamed”. However, as panelist Dr. Houssein Sater, Lead Physician Scientist at Cleveland Clinic Martin Health, observed, these classifications don’t always reflect reality. Some parts of a tumor may be inflamed, while others have no immune cells at all—spatial context is essential to making these distinctions. “Spatial biology is going to be able to help us dissect what is really important,” said Dr. Sater. “I hope that will be the answer to the next biomarker.”
Aside from capturing these high-level spatial patterns, scientists are looking to dive deep into the tumor biology, down to the single-cell level. “Single-cell is very important because spatial analysis at the single-cell level can enable unambiguous immune profiling,” said panelist Dr. Qingyan “Sandy” Au. As Principal Scientist and Director of Multiplexing Operations at NeoGenomics, she brought a CRO perspective to the discussion.
Many biomarkers can co-localize in multiple cell subtypes. With single-cell imaging, it’s possible to see exactly which cells are expressing which markers and phenotype them accordingly. Combining spatial context with single-cell resolution allows researchers to study cells within the context of their tissue and the cell-cell interactions taking place, an approach known as spatial phenotyping.
Considerations for validating spatial biomarker assays
How can insights from spatial phenotyping translate into biomarker assays that can be applied to clinical studies? A common theme that emerged during the panel discussion was the need for validation in mIF. A new spatial biomarker assay will need to be reproducible, said Dr. Sater, and generate comparable data across research groups, requiring multi-institutional collaboration.
Immuno-oncologists are already on the case. Recent studies indicate that the research community is recognizing the need for spatial phenotyping and has been working to make this technology more viable for clinical trials and diagnostics. The Multiplexed Immunofluorescence Reproducibility Evaluation (MITRE) study, published in JITC, measured the inter- and intra-site variability of a 6-plex panel developed on the Phenoptics™ platform and showed high concordance. Several other research groups are also tackling the challenge of validation.
At Johns Hopkins (JHU), the astrophysics department collaborated with the cancer center to develop a platform called AstroPath™, which relies on Phenoptics for data generation. In a study published in Science, the JHU team described how they built an mIF panel and identified a spatial phenotypic signature highly predictive of immunotherapy response in advanced melanoma.
“You need large numbers of samples, big swaths of tissue, and you need to be able to image them reliably”
Phenoptics is a high-throughput platform, which makes it well-suited for these types of studies that seek to develop assays for large-cohort clinical research. “You need large numbers of samples, big swaths of tissue, and you need to be able to image them reliably,” said Dr. Surace.
However, spatial biology is applicable across the research continuum, from discovery to translational and clinical research. Whereas Phenoptics excels at addressing narrower questions that require fewer markers and higher throughput, ultrahigh-plex methods, like CODEX®, are ideal for answering broader questions and digging deep into the tumor microenvironment. These approaches are complementary and can be used effectively together to identify and validate spatial biomarkers.
“If you want to start with exploratory studies, you can use a high-plex [platform], and then down-select to identify a subset of biomarkers for potential late-stage use,” said Dr. Au. “Then we can use Polaris—since it’s high-throughput—to validate them and support late-stage development, patient management, and potentially commercialization.”
As we learned at SITC, it’s evident that spatial biology is transforming our approach to biomarker discovery and development. At Akoya, we’re committed to partnering with leaders in immuno-oncology to support research that ultimately improves treatment outcomes and patient care. Learn more about our spatial biology solutions
