‘Immune Archetypes’ Of Cancer Could Help Tailor Treatments To Tumors

February 10, 2022

By Deborah Borfitz

February 10, 2022 | Immunotherapies vary widely in their ability to effectively treat patients with cancer, and the reason is the differing immune states of their tumor. Only a finite number of “immune archetypes” exist across all cancers—and these distinct environments are inhabited by more than the T cells targeted by currently marketed immunotherapy drugs, according to Alexis Combes, Ph.D., director of the Disease to Biology (D2B) CoLab at the University of California, San Francisco (UCSF) and incoming assistant professor in UCSF’s department of pathology. 

The findings, recently published in Cell (DOI: 10.1016/j.cell.2021.12.004), emerged from a holistic survey of the immune microenvironments of 364 tumor biopsies using flow cytometry and RNA sequencing. Clustering of 10 immune cell features identified 12 unique tumor archetypes spanning cancer types.

To intelligently act on the immune system requires knowing the immune and tumor gene-expression patterns of patients, Combes says. As demonstrated in the study, these immune archetypes each represent a different subversion of the body’s immune defenses and are variably drawn from a few or many kinds of cancer. That makes some melanomas quite like some lung tumors but more biologically different than other melanomas.

Multiple data sources—including genetic sequencing, cell surface markers, imaging, and patient clinical data—were used for the unbiased interrogation. Immunotherapies tend to work well for melanoma and some lung tumors, but certain cancer patients don’t respond at all, he says. Therefore, the first order of business was to create an atlas of the types of immune cells present in as many types of tumors as possible to look for recurring patterns.

The findings offer a new way of looking at cancer immunotherapy that matches the immune environment around the tumor and points the way to personalized immunotherapies, says Combes. “This work will help clinicians find the right biology to target and avoid targeting cells that aren’t present in the tumor.”

Concerted Effort

Interest in the microenvironment of tumors has been on a precipitous rise since a Nobel Prize was awarded jointly to two cancer immunotherapy researchers in 2018 for their work in uncovering ways to activate the immune system to attack cancer, says Combes. But the search for ways to match patients to therapies based on their immune system behavior has been ongoing for more than two decades.

The big unknown has been why only some proportion of patients—sometimes 20%, sometimes 50%, and sometimes none of them—respond to an immunotherapy drug, he continues. That is the focus of the UCSF Immunoprofiler consortium, where Combes serves as scientific lead.

The consortium was started in 2015 by Matthew (Max) Krummel, Ph.D., in whose lab Combes was previously a postdoctoral researcher. Its mission is to understand the nature of the immune response as a pathway to treating disease and identifying new immunotherapy targets.

For the Cell study, UCSF Immunoprofiler scientists were actively involved in every step from patient consent to the generation and analysis of the data. The solid tumor samples profiled were provided by 78 UCSF clinicians working in various cancer operating rooms and outpatient clinics.

The classification exercise was approached without any preconceived ideas about what might be found, Combes stresses. “We didn’t know how many [archetypes] we were going to find… but it turns out we could only find a defined number of classes, which in this case was 12.”

Larger studies with more patient tumor samples are now needed to confirm the finding, he says, but the suggestion is that cancers have a limited number of immune archetypes that can guide therapeutic decision-making.

Multiomics Approach

To be “very precise” in profiling the immune system of tumors, the research team began by using flow cytometry to look at the protein expressions of every single cell, says Combes. RNA sequencing was then used to understand what they do. “Every cell type is able to change what they do based on their pattern of RNAs expression.”

From one archetype to the next, the researchers could see that immune cells were “doing their own thing” based on the genes and transcriptome of the tumor microenvironment. That activity could also be grouped based on 10 key cell types including CD4+ and CD8+ T cells as well as macrophages, monocytes, and subtypes of dendritic cells.

Most of the immunotherapies being used in the clinic currently are focused on T cells expressing CTLA-4, PD-1, and its ligand PD-L1, but treatment response varies because “there is almost never just T cells in a tumor,” Combes says. The new pan-cancer census found two main classes where T cells express strongly for PD-1 and CTLA-4, but the surrounding cells in the environment determine whether they can respond.  

While it was previously suggested that different cancer types might have the same tumor immune microenvironment, this is the first study that proved it out by looking at so many parameters and so many tumor types, he notes.

The research team has now started to look at biopsies from treated patients and how they responded based on their immune archetype, adds Combes. The archetypes for PD-1 and CTLA-4 are likely to be of the most immediate interest, since it is the prevailing target of current immunotherapies.

But as new therapies are developed in the future, they hope the archetypes can be discovered by a simple blood draw and used for enrolling patients into clinical trials based on who would most benefit from the investigational treatment.


With any disease, the immune system has some involvement, Combes says. An overarching goal of the D2B CoLab is to foster collaborative-based research on behalf of all UCSF researchers by providing centralized access to the assays and expertise needed to tease out the immunological underpinnings in their therapeutic area of interest.

The UCSF CoLabs initiative comprises a set of labs, some of which specialize in microscopy and others in the coding and computational work needed to understand the data, he explains. The same team who worked on the immune archetypes for cancer also last year described the pattern of immune cell expression in the blood of people with COVID-19, as reported in Nature (DOI: 10.1038/s41586-021-03234-7).

“UCSF is a pioneer on this front,” says Combes, enabling projects that would otherwise be unachievable due to the competition for funding and required knowledge sharing. Science at large is among the beneficiaries—both the identified cancer immune archetypes that can now be further studied in new tumors and patients and the replicable framework that can be applied to an assortment of different diseases.