Cancer immunotherapy has changed the treatment of some cancers by harnessing the body’s own immune system to recognize and attack cancer cells. But getting the immune system to attack cancer is rarely as simple as turning it on.
Cancer cells can hide from immune attack. They can alter the environment around them, making it harder for immune cells to reach them or function effectively. Patients with seemingly similar cancers can respond very differently to the same treatment. And tumors can change as treatment progresses.
Researchers at the University of Pittsburgh are studying these challenges from several directions. Together, their work reflects a broader shift in cancer immunotherapy: Instead of viewing a tumor as a fixed target, researchers are increasingly interested in how cancer and the immune system change over time, and whether those changes can be used to improve treatment.
Making cancer visible
Before the immune system can destroy a cancer cell, it must recognize that the cell is abnormal.
Jing Hong Wang, MD, PhD, studies one way cancer cells can interfere with that process. Cells use molecules called MHC class I to display small pieces of proteins on their surfaces. T cells examine these pieces and can recognize and attack cells displaying abnormal proteins.
Some cancers reduce or eliminate MHC class I, making them harder for T cells to see. Prior research has shown that this can happen through changes to a gene called B2M, which is an essential part of the MHC class I system. Wang’s research identified a subset of CD8 T cells that can be activated by chemo-treated cancer cells independent of their MHC class I.
Her team has also explored ways to restore that visibility. In laboratory models of B-cell lymphoma, drugs called HDAC inhibitors increased MHC class I expression. Combining an HDAC inhibitor with an immune checkpoint inhibitor called anti-PD-1 overcame resistance in tumors that had not responded to anti-PD-1 treatment alone. In a model of head and neck cancer, deleting B2M similarly prevented a combination of radiation and anti-PD-L1 treatment from working.
But making a tumor visible is only part of the problem.
Patients have different collections of T-cell receptors, or TCRs—the proteins or structural proteins that allow T cells to recognize specific targets. So even if two patients’ tumors carry the same cancer-related targets (e.g. mutations in p53), their immune responses may be quite different. One patient may have T cells capable of recognizing the cancer-related targets, while another may have too few of those cells, or none that recognize such targets effectively.
Tumor cells can differ, too, including in how much MHC class I they express and how readily they can increase its production in response to various treatments such as radiation.
Understanding those differences could help researchers develop treatments that do two things at once: make cancer easier for the immune system to see and strengthen the immune cells best equipped to attack it.
Cancer doesn’t grow alone
Even when immune cells can recognize a cancer, they still have to reach it and function in the environment around it.
Wei Du, MD, PhD, studies this problem in leukemia, focusing on the bone marrow, where blood cells are produced. The bone marrow is more than a factory for making blood cells. It is a complex ecosystem of blood-forming stem cells, supporting cells, blood-vessel cells, immune cells, macrophages and other populations that constantly communicate with one another.
When leukemia develops, those relationships can change. Leukemia cells can influence surrounding cells to produce signals that help the cancer survive, grow, settle into the bone marrow and resist treatment. Some of these changes may begin even before leukemia is fully established.
Du’s research also examines what happens when DNA damage triggers immune and inflammatory responses in the bone marrow. Persistent DNA damage can affect both damaged cells and neighboring cells. Her studies have identified TREM1 as one pathway associated with a leukemia-supportive inflammatory environment, while other research has identified NLRP12 in response to DNA damage in aging blood-forming stem cells.
These findings point to a problem that extends beyond the leukemia cell itself. If the surrounding environment helps protect cancer, treatment may need to target that environment, too.
Researchers are investigating ways to disrupt signals that keep leukemia stem cells in protective areas of the bone marrow, where they may be harder for treatment to reach. Other work focuses on leukemia-associated macrophages, immune cells that can be altered by leukemia and take on functions that support cancer-cell survival and growth.
At the same time, restoring a healthy bone marrow environment could help normal blood-forming cells recover after treatment.
For Du, the bone marrow is therefore not simply the place where leukemia happens. It is part of the disease itself and may be part of the solution.
Finding the right immune response
The challenge in cancer immunotherapy is not simply getting the immune system to respond. Researchers also need to direct that response toward cancer while limiting damage to healthy tissues.
Warren Shlomchik, MD, studies this balance in allogeneic hematopoietic stem cell transplantation, or alloSCT. The procedure can produce a powerful graft-versus-leukemia (GVL) effect, in which donor immune cells recognize and attack remaining leukemia cells. But those same cells can also cause graft-versus-host disease (GVHD), attacking healthy tissues.
One of Shlomchik’s research questions began with a basic observation: Different types of T cells may behave differently after transplantation.
In mouse studies, his team found that memory T cells caused less GVHD than naïve T cells—T cells that have not previously encountered their target—while still providing functional immune protection and some anti-leukemia activity.
That finding eventually moved into clinical testing. Patients received stem cell grafts in which naïve T cells had been selectively removed, and early clinical studies showed reduced GVHD. The approach is now being examined in larger randomized clinical trials.
The work gets at a central problem in immunotherapy. More immune activity is not necessarily better. The goal is to produce an immune response that can attack the cancer without causing unnecessary harm to healthy tissue.
Shlomchik’s laboratory is also studying why some leukemias are more vulnerable to the graft-versus-leukemia response than others. His team has developed laboratory models of leukemia that differ in their sensitivity to immune attack, allowing researchers to investigate what makes some cancers susceptible to GVL while others resist it.
Engineering the immune response
Another approach is to change the immune cells themselves.
CAR-T cell therapy, which involves genetically modifying a patient’s T cells to recognize cancer, has transformed treatment for several B-cell and plasma-cell cancers. But developing CAR-T therapies for other cancers can be more difficult, in part because potential targets may also be present on healthy cells that are essential to survival.
Researchers are exploring several ways around these challenges. One involves combining CAR-T therapy with gene editing of blood-forming stem cells. In this approach, researchers modify the stem cells so they are no longer recognized by an engineered immune cell. The protected stem cells can then be returned to the patient, potentially allowing the immune cells to attack cancer cells carrying the target while protecting the blood-forming system.
Another approach uses bispecific antibodies, which recognize both a cancer target and a T cell, effectively bringing the two together.
Sawa Ito, MD, PhD, is studying another form of engineered cellular therapy: TCR-T cells, which are T cells modified to express a T cell receptor (TCR) which recognizes a specific target. CAR-T cells recognize targets on the outside of cancer cells whereas TCR-T cells can recognize proteins whether they are expressed outside or inside cancer cells. This expands the range of cancer-related proteins that can be targeted.
Dr. Ito’s lab has worked on TCRs that recognize transplant antigens (minor histocompatibility antigens) that can target leukemia cells in the context of an allogeneic hematopoietic stem cell transplant. In collaboration with Warren Shlomchik, MD, whose laboratory demonstrated the safety and efficacy of using such products to target minor histocompatibility antigens in mouse models, she cloned TCRs targeting human minor histocompatibility antigens (also in collaboration with BlueSphere Bio). One of those TCRs advanced to clinical testing, wherin two patients were treated TCR-T expressing it.
Their clinical trial with interferon-gamma, or IFN-γ, also grew out of research in Shlomchik’s laboratory, where he found in mouse models that GVL against myeloblastic leukemias requires IFN-γ. Ito and Shlomchik co-developed a clinical trial testing IFN-γ combined with T cells from the initial stem cell donor in patients whose leukemia has returned after an HLA-matched stem cell transplant.
In the first-in-human study, four of six patients achieved deep remissions with no measurable residual disease, and two remained in remission more than three years later. Those results led to an FDA-R01-sponsored multicenter Phase II clinical trial that is now evaluating the approach further.
When resistance changes
All of this work raises a larger question about what it means for a cancer to be “resistant” to treatment.
Riyue Bao, PhD, studies cancer as a spatially complex, constantly changing ecosystem. Tumors can change internally, interact differently with the immune system, remodel surrounding tissue and evolve in response to treatment and other environmental influences. Her group is particularly interested in how tumor-intrinsic programs can reshape the surrounding microenvironment and create barriers to effective immune attack.
From that perspective, resistance may not always be a permanent characteristic of a tumor. Tumor cells can transition between biological states during treatment, and some resistant states may potentially be reversible.
Bao’s laboratory is using genomic data, single-cell measurements, spatial information, imaging and longitudinal clinical outcomes to study these changes. Artificial intelligence and machine learning can help researchers work through the enormous amount of information generated by these approaches and identify relationships that might otherwise be difficult to see.
But finding a pattern is only the beginning. The more important question is why that pattern exists, whether it reflects a causal mechanism, and whether changing that mechanism could alter the cancer’s response to treatment.
Her laboratory is developing immune-competent patient-derived organoids—three-dimensional models grown from patient tissue that can be used to study how tumors respond to immunotherapy. Researchers can alter biological features in these models and observe whether those changes affect treatment response.
That takes the research beyond simply identifying a biomarker associated with a particular outcome. It allows researchers to ask a more direct question: If we change this feature, can we change what happens to the cancer?
Treating a moving target
Cancer does not remain static. As the disease changes, so can the way cancer cells interact with the immune system and the way they respond to treatment.
Much of precision oncology has relied on biopsies or other samples taken at a particular point in a patient’s disease. But a sample can show researchers what a cancer looks like at one moment; it may not explain how the disease got there or what it will look like after treatment.
Bao sees an opportunity to combine information collected over time—including clinical samples, blood-based biomarkers, imaging, single-cell and spatial measurements, and functional patient-derived models—to build a more complete picture of an individual patient’s cancer.
The goal is to identify resistance earlier, understand why it is happening and select treatment combinations based on the biology of the cancer at that point in time.
Researchers are pursuing a range of strategies to do this. These may include making cancer easier for the immune system to see, breaking down barriers that keep immune cells out, disrupting the environment that protects cancer cells or engineering immune cells to recognize cancer more precisely.
The common thread is that treatment may become less static. Instead of asking only which therapy matches a cancer, researchers are increasingly interested in how the disease is changing—and whether treatment can change with it.
For Pitt researchers, answering that question brings together cancer biology, immunology, cellular engineering, computational science and clinical medicine. The goal is not simply to make the immune system stronger, but to make its response more precise, more durable and better suited to the cancer in front of it.