T Cells Shed Exhaustion Marker, Escape Tumors, and Prevent Cancer Relapse

Cancer-fighting T cells that appear to be exhausted may retain a surprising ability to leave tumors and establish long-lasting immune protection, according to a study from researchers at the University of Pittsburgh School of Medicine. The findings challenge the prevailing view that T cells expressing the inhibitory receptor LAG3 are functionally inert or destined to die inside the tumor microenvironment. Instead, the work suggests that a subset of these cells can shed LAG3, migrate through the body, and contribute to immune memory capable of preventing cancer from returning. The study, published in the Journal of Experimental Medicine, also introduces a tracing system that allowed scientists to follow the fate of LAG3-expressing T cells after they changed their surface identity.

The immune system relies heavily on CD8+ T cells to recognize and destroy abnormal cells, including cancer cells displaying tumor-derived antigens. Yet tumors can impose persistent antigenic stimulation, metabolic stress, nutrient deprivation, and suppressive signaling that gradually alter T-cell behavior. This process, commonly known as T-cell exhaustion, is associated with reduced effector activity and the sustained expression of inhibitory receptors such as PD1, TIM3, and LAG3. Exhausted T cells are not simply inactive; they represent a complex and heterogeneous population containing cells with different developmental states and functional capacities. Understanding which exhausted cells remain capable of producing durable immunity has become a central goal of modern cancer immunology.

LAG3, or lymphocyte activation gene 3, is an inhibitory receptor found on chronically stimulated T cells. It binds major histocompatibility complex class II and other ligands, transmitting signals that can restrain T-cell activation. Because of this activity, LAG3 has become an important target for immunotherapy, either alone or in combination with inhibitors of the PD1 pathway. However, the biological consequences of LAG3 expression inside tumors have remained incompletely understood. Dario A.A. Vignali, chair and distinguished professor of immunology at Pitt, and his colleagues set out to determine what happens to tumor-reactive T cells after they acquire LAG3, rather than examining only whether the receptor is present at a single point in time.

To follow these cells, the researchers developed a genetically engineered mouse model that permanently records LAG3 expression. The animals were treated with tamoxifen, a drug that activates a genetic labeling system in cells expressing LAG3 at that moment. Once activated, the system caused those cells to produce tdTomato, a red fluorescent protein. Crucially, the fluorescent label remained in the cells even if they later stopped producing LAG3. This created a molecular time stamp: researchers could distinguish cells that had expressed LAG3 in the past from cells that were still expressing it when tissue samples were analyzed. The approach enabled the team to track the movement and changing phenotype of exhausted T cells over time in melanoma-bearing mice.

After labeling LAG3-expressing cells within melanoma tumors, the investigators observed that the population did not remain uniform. Some red fluorescent cells continued to express LAG3, producing a double-positive profile that marked their historical and current receptor status. These cells were found within the tumor and appeared to remain embedded in the local cancer microenvironment. Other tdTomato-labeled cells no longer displayed LAG3 on their surface. These cells, described as single-positive because they retained the historical fluorescent label but lacked current LAG3 expression, were detected in tumor-draining lymph nodes and in additional tissues outside the tumor. Their distribution indicated that at least some T cells associated with an exhausted state can undergo a transition and regain the capacity to circulate beyond the tumor.

The difference in location between the two populations was particularly important. Cells that continued to express LAG3 were largely confined to the tumor, whereas cells that had lost the receptor appeared able to exit the malignant tissue. The results suggest that LAG3 expression may be linked not only to inhibitory signaling but also to the physical retention of T cells within the tumor microenvironment. Alternatively, the loss of LAG3 may identify cells that have entered a distinct differentiation pathway with altered adhesion, chemokine responsiveness, or migratory behavior. The study does not establish that LAG3 alone directly locks cells inside tumors, but it reveals a strong association between receptor loss and the emergence of mobile, long-lived tumor-reactive T cells.

The researchers next tested whether these departing cells contributed to immune memory. They removed melanoma tumors surgically from mice and allowed the animals to recover for approximately one month, giving the immune system time to establish a response against tumor antigens. When the same melanoma cells were later introduced again, untreated animals were able to control or eliminate the secondary tumors, demonstrating protective immunological memory. However, when the investigators selectively removed the tdTomato-labeled single-positive T cells—the cells that had expressed LAG3 previously but no longer did so—growth of the recurrent tumors accelerated substantially. This experiment showed that the population was not merely relocating without consequence. Its presence was necessary for a robust and durable response against the returning cancer.

The findings provide a possible explanation for why some exhausted T cells can support long-term protection while others remain terminally dysfunctional. Tumor-reactive LAG3+CD8+ T cells may diverge into at least two developmental outcomes: one population remains in the tumor and progresses toward terminal exhaustion, while another loses LAG3, leaves the tumor, and becomes a long-lived memory-like population in peripheral tissues and lymphoid organs. Such cells could provide surveillance after the primary tumor has been removed, rapidly expanding or reactivating when they encounter the same tumor antigens again. Their ability to persist outside the hostile tumor environment may also protect them from the intense metabolic and suppressive pressures that drive terminal dysfunction.

The work has potential implications for therapies that target LAG3. Blocking LAG3 is intended to release inhibitory constraints and improve T-cell activity, but the new findings raise an additional possibility: manipulating this pathway might influence where tumor-reactive T cells reside and whether they can seed systemic immune memory. If LAG3 inhibition promotes the exit of selected T cells from tumors, combination treatments could potentially be designed to enhance both immediate tumor destruction and long-term protection against relapse. The researchers emphasize that this question remains under investigation. Human tumors contain complex mixtures of T-cell states, and a result observed in melanoma-bearing mice will need to be validated in patient samples and clinical studies before it can guide treatment decisions. Even so, the study changes the way scientists may interpret LAG3 expression, showing that a history of exhaustion does not necessarily mark the end of a T cell’s usefulness.

Cancer recurrence remains a major challenge because successful removal or suppression of a primary tumor does not always generate sufficiently durable immune surveillance. The Pittsburgh study suggests that the immune system’s future ability to recognize a returning malignancy may depend on a specialized population that temporarily bears the hallmarks of exhaustion before adopting a more mobile, memory-associated state. By permanently marking cells at the moment they express LAG3, the investigators made it possible to see this transition rather than treating T-cell identity as fixed. The discovery offers a new framework for understanding exhaustion, migration, and immunological memory, and it may eventually help researchers develop immunotherapies that do more than shrink tumors: they could also train the immune system to remain prepared for cancer’s return.

News Publication Date: 18-Aug-2026

Web References: Journal of Experimental Medicine article and DOI: https://doi.org/10.1084/jem.20241968; University of Pittsburgh Department of Immunology: https://www.immunology.pitt.edu/

References: Aggarwal, V., Sun, Y. (Claudia), Liu, C., and colleagues. “Tumor-reactive LAG3+CD8+ T cells diverge into terminally exhausted cells and long-lived memory T cells.” Journal of Experimental Medicine. DOI: 10.1084/jem.20241968.