Mapping the Tumor Microenvironment With a Blood Test
A newly developed blood test offers the first non-invasive way to map the tumor microenvironment.
The landscape of cells surrounding a tumor, known as the tumor microenvironment (TME), can shape cancer growth, spread, and response to therapies—but its complexity and dynamic nature make it difficult to study.
Now, researchers have identified nine distinct cellular neighborhoods in the TME and used these to develop a novel blood test technique that offers the first non-invasive way to study the TME.
The tumor microenvironment influences therapy response
As cancer cells take root, they exploit normal cell processes to evade immune surveillance, make collagen scaffolds, and build new blood vessels—forming a permissive environment that helps the tumor to grow and spread.
“The tumor microenvironment consists of immune and structural cells that together provide the ‘soil’ in which cancer cells live,” explained Dr. Aaron Newman, an associate professor at Stanford University and senior author of the new research. “The composition of this soil strongly determines how well cancer cells can multiply, spread, and respond to therapy.”
The TME can influence response to chemotherapy, immunotherapy, and radiotherapy through a variety of intrinsic (present before drug administration) or acquired (evolving in response to therapy) resistance mechanisms.
Additionally, for cancer cells with the same drug target, responses to the same therapy can vary widely, depending on the TME.
“Many of the most promising modern therapies, such as immunotherapies, bi-specifics, personalized vaccines, and others, directly involve interactions with the tumor microenvironment,” noted Newman.
The complexity, heterogeneity, and ever-changing nature of the TME make it challenging to design and predict the response to new therapies.
Studying the tumor microenvironment
The TME has traditionally been studied through biopsy samples—using either microscopy or bulk genetic analysis. However, biopsies are invasive and are only performed occasionally throughout a patient’s diagnosis and treatment journey.
Capturing a snapshot of the TME at a single point in time also does not reflect the dynamic nature of the cellular interactions, which change with disease progression and treatment.
Newman and his team therefore set out to develop a non-invasive way to profile the TME.
“A blood test can overcome many of these challenges: it is noninvasive (and thus safe), it can be obtained repeatedly to determine how the tumor microenvironment changes over time and in response to therapy, and it can access information about the patient’s cancer more holistically, especially for patients with metastatic disease,” Newman explained.
By integrating over 10 million single-cell and spatial transcriptomes from diverse cancer samples into a machine learning framework, dubbed Spatial EcoTyper, they identified nine spatial ecotypes (SEs) that were highly conserved across tumor types.
“Spatial ecotypes in cancer are collections of cells that are physically close to each other and that share specific biological characteristics,” explained Newman. “They can be thought of as cellular neighborhoods or ‘social networks’ that form the building blocks of the tumor microenvironment.”
Advancing Biomarker Discovery With Liquid Biopsy Research
This webinar explores how liquid biopsy research is advancing biomarker discovery through extracellular vesicle (EV)-associated biomarkers and circulating tumor DNA (ctDNA).View Webinar
The SEs they identified were dispersed throughout the tumor: some were located outside the tumor, some at its edge, and others found deep within it. Each SE had unique biology, spatial features, and associations with clinical outcomes.
“The SEs span many fundamental processes that are important for cancer growth and therapy response, including blood vessel growth, wound healing, immune suppression, immune activation, and states associated with cancer cell killing,” said Newman.
Methylation patterns linked to spatial ecotypes
As cancer cells or healthy cells die, they release DNA into the blood, which carries distinctive methylation patterns that indicate which genes were actively expressed in the cell of origin. The team realized that they could distinguish between SEs using DNA methylation profiling from cell-free DNA.
“It is well-established that methylation patterns are strong markers of cell identity,” said Newman. “Our interpretable AI approach, Liquid EcoTyper, was used to identify methylation signatures of each spatial ecotype while also accounting for signals from cancer cells and normal cell-free DNA, among other signals.”
As Liquid Ecotyper is not a “black box”, the team could trace the methylation signatures it identified back to specific genetic markers of each SE.
To test whether the model could reconstruct the TME, Newman and his colleagues applied it to a cohort of melanoma patients where cell-free DNA and paired tumor biopsies were available.
Cell and Gene Therapies: The Evolution of Advanced Biologics
This eBook takes a look at the challenges shaping advanced biologics, alongside practical strategies to improve consistency, scalability, and compliance across development and manufacturing.View eBook
They found that the levels of SEs determined by the blood test closely correlated with tumor analysis and were also strongly associated with response to immune checkpoint inhibitors.
Bringing more information on the tumor microenvironment to the clinic
Once validated, the blood test for the TME could be used in numerous ways in the clinic.
“These include forecasting response to therapy prior to treatment initiation and real-time monitoring of the tumor microenvironment during therapy,” said Newman. “Depending on the state of the tumor microenvironment, the latter could provide actionable information about whether a therapy is working or likely to work, and whether a patient should switch to a different therapy.”
This approach to studying the TME holds promise, but larger and more diverse single-cell spatial transcriptomics cohorts will be needed to comprehensively survey SEs across multiple cancer types. The authors also outline that the biological insights from SE identification will require experimental confirmation.
“Our ongoing and future studies include evaluating the utility of the approach in cancer types beyond melanoma, in therapies beyond immunotherapy alone, in early-stage disease (including in patients with precancerous lesions), and in the setting of real-time monitoring of tumor microenvironment dynamics,” concluded Newman.
Reference: Zhang W, Brown EL, Usmani A, et al. Non-invasive profiling of the tumor microenvironment with spatial ecotypes. Nature. 2026. doi: 10.1038/s41586-026-10452-4
About the interviewee:
Dr. Aaron Newman is currently an associate professor in the Department of Biomedical Data Science at Stanford University, where he is a member of the Institute for Stem Cell Biology and Regenerative Medicine and the Stanford Cancer Institute. He is also a Chan Zuckerberg Biohub Investigator. He earned his PhD in the biomolecular science and engineering program from the University of California, Santa Barbara, and completed postdoctoral training in cancer genomics at Stanford University.
Add Technology Networks as a preferred Google source to see more of our trusted coverage.
