Immunotherapy’s effectiveness depends largely on immune cells—specifically CD8+ “killer” T cells—being able to infiltrate tumors. The presence of these cells within tumors is one of the best predictors of immunotherapy’s effectiveness in patients. Biopsies can provide a sense of the situation to help guide doctors’ decisions, but current methods have significant limitations. Fortunately, thanks to CRI researcher Mohammad Rashidian, Ph.D., doctors might soon have a better tool at their disposal.
The tool is called immuno-PET (which stands for positron emission tomography), and Dr. Rashidian showed that it could be used noninvasively to visualize killer T cells within several different types of tumors in mice. The work was published in The Journal of Experimental Medicine, and Rashidian—a postdoctoral fellow currently working under Hidde L. Ploegh, Ph.D., at the Boston Children’s Hospital and Harvard Medical School—was the study’s first author.
Immuno-PET uses antibodies or antibody fragments that bind specific molecules on the surfaces of cells. In this case, Rashidian and his colleagues used antibody fragments known as “nanobodies” that were specific for the CD8 molecule expressed by killer T cells. Through the use of PET imaging, the CD8-bound nanobodies were able to reveal the precise positions of killer T cells within tumors. While biopsies requires invasive surgery to determine immune cell infiltration into tumors, immuno-PET only requires a simple infusion.
Even when a patient has a single tumor, “part of the tumor might be different than another part of the tumor, because the tumor environment is very dynamic,” according to Rashidian. Additionally, in patients with metastatic cancer, their multiple lesions can each have different immune landscapes. For comprehensive analysis, surgical biopsies would need to be taken from each lesion independently, whereas immuno-PET could visualize all lesions non-invasively at the same time.
In other words, as the authors noted in their paper, “invasive procedures such as biopsies cannot provide this type of global information for an entire tumor mass or its metastases and may yield less reliable or even misleading data when correlating immune cell infiltration status with the outcome of immunotherapy.”
First, Rashidian and his colleagues demonstrated immuno-PET’s visualization capabilities with subcutaneously injected pancreatic tumors (as seen in the image to the right), the density of which makes it notoriously hard for T cells to infiltrate. This established the proof-of-principle that this approach could be suitable for other cancer types.
Then, with a melanoma model, they used immuno-PET to examine the association between immune infiltration and responses to immunotherapy, specifically to anti-CTLA-4 checkpoint immunotherapy. That’s where things got really interesting.
The researchers found that the critical factor for immunotherapy’s effectiveness wasn’t how many killer T cells had infiltrated a tumor, but how they infiltrated a tumor. “If you have killer T cells distributed [evenly] throughout the tumor, then that tumor has a higher likelihood of responding to immunotherapy,” Rashidian said.
The mice that responded best had tumors characterized by a single, uniform cluster of killer T cells within them. In contrast, tumors that were characterized by two or more clusters of killer T cell infiltration were much less likely to respond and these mice had worse overall survival.
Next, the researchers experimented with two mouse models of breast cancer that are associated with different immune landscapes—one is characterized by uniform killer T cell infiltration, while the other was characterized by sporadic infiltration that was limited mainly to the edges of tumors.
When treated with immunotherapy, the response pattern observed in the melanoma model also applied to these breast tumors: those with homogeneous infiltration had much higher response rates compared to those with heterogeneous infiltration. Based on these results, which “agree well with … analysis of human biopsy specimens taken from patients receiving immunotherapy,” the authors concluded that “immuno-PET might be useful as a predictor of the response.”
The authors also noted another important advantage of immuno-PET. To determine if patients are benefiting from therapy, doctors have typically relied on tumor size because, with traditional treatments like chemotherapy, if a tumor got bigger it meant the treatment wasn’t working, at which point the patient would be taken off treatment.
However, with immune-based therapies it’s a little more complicated. Sometimes tumors appear to get bigger initially after immunotherapy before eventually shrinking. This is known as pseudoprogression and is a result of tumors swelling due to immune infiltration, which is actually associated with positive responses.
Currently, radiographic scans—such as CT (computerized tomography) scans as well as glucose-based PET imaging—are used to determine tumor size and guide the next clinical steps, but they cannot distinguish between a tumor that is truly progressing versus one that only appears to be progressing. Because of that, there’s a risk that patients could be taken off immunotherapy prematurely.
“Pseudoprogression, something that CT cannot distinguish, can be figured out with this method,” Dr. Rashidian explained. Consequently, doctors would have a much better idea of what’s actually going on inside tumors of patients treated with immunotherapy, and therefore could potentially make better treatment decisions to improve patient outcomes.
While thus far this nanobody-based immuno-PET approach has only been used to visualize immune infiltration in mouse tumors, Rashidian believes it has immense potential for human patients. Belgian investigators have already used a HER2-binding nanobody construct to determine whether the tumors of breast cancer patients express HER2, an important marker whose presence signals that an anti-HER2 therapy might be a good option for a particular patient.
With respect to CD8-binding nanobodies, Rashidian explained that a “humanized” version suitable for treating patients needs to be developed. “Other than that, basically they can be translated to the clinic. Immuno-PET wouldn’t replace current diagnostic tools, but it would be a great addition, to the benefit of many patients.” When asked whether he had any plans to humanize these nanobodies, he replied, “I would be very interested of course…as soon as we can find the necessary resources to move the technology into the clinic.”
Killer T cells aren’t the only immune cells immuno-PET can observe. Rashidian also discussed his interest in using it to study other immune cells known to be involved in cancer, including helper T cells via CD4-binding as well as myeloid immune cell populations using their relevant markers. “Moreover,” he added, “we need to develop methods to visualize and track the activity and dynamics of tumor-specific T cells, which would clearly be a breakthrough for cell-based immunotherapy approaches.”
Due to this flexibility, immuno-PET could provide doctors with an entirely new way to explore how various factors influence the interactions between immune cells and cancer cells in the tumor microenvironment. This in turn could aid identification of “new mechanisms that drive the different types of immune recruitment,” according to the authors, and potentially open up new therapeutic avenues in cancer treatment.