Just like a runaway eighteen-wheeler on the interstate is a scary sight to behold, an immune system without brakes can quickly transform from a cherished protector to a host’s worst nightmare. To protect our bodies from threats without destroying our tissues in the process, our immune system not only deploys sophisticated methods of surveillance and response—it also boasts elaborate mechanisms for modulating that response. One key molecular player responsible for this function is a cell surface receptor called programmed cell death protein-1 (PD-1), a B- and T-cell specific receptor first identified in the early 90’s by scientists at Kyoto University in Japan. Upon binding its partner on a different cell (programmed death ligand-1, or PD-L1), PD-1 can trigger a signaling cascade ultimately abrogating the function of that B- or T-cell, thus acting as a ‘brake’ on the adaptive immune response.
The PD-1/PD-L1 system likely exists to prevent autoimmunity or immune overactivation, but it appears that at some point the secret got out to tumor cells, which often express high levels of PD-L1 to evade immune destruction. Since the discovery of this phenomenon, scientists have been hard at work developing therapeutics to target PD-1/PD-L1 and enhance anti-tumor immune responses—a goal which was realized in 2014 with the FDA approval of pembrolizumab, a PD-1 targeting antibody that now forms part of an expanding arsenal of immune checkpoint inhibitors against a variety of cancers. Checkpoint inhibitors can enhance the function of a patient’s native B- and T- cells against a tumor, but they might also bolster the cancer-fighting power of CAR-T cells—T cells genetically engineered to recognize tumor antigens before being infused into a patient (previously discussed here). Now armed with these two potentially transformative cancer therapies, however, clinicians face a new challenge: what is the best way to combine checkpoint inhibitors and CAR-T cells for maximum effect?
Enter Drs. Alexandre Hirayama and Cameron Turtle of the Clinical Research Division at Fred Hutch, two immunotherapy-focused clinician-scientists who led a recent study published in Blood Advances addressing this very question. The team focused on a specific type of hematologic cancer called large B-cell lymphoma (LBCL), which responds to immunotherapies including CAR-T cells but often relapses following treatment—a phenomenon that might be associated with the tumor cells’ expression of PD-L1. “Before this study, there were short reports of using CAR-T cells in combination with immune checkpoint inhibitors,” notes Dr. Turtle, “but the outcomes were mixed, with findings that did not inform the best way to deliver these two treatments to patients.” Indeed, one could imagine that order and timing might matter for two treatments that must work in tandem to fight tumors. “At the time this work was being done, it wasn’t even clear whether ‘cutting the brakes’ off CAR-T cells with a checkpoint inhibitor would lead to immune overactivation and downstream problems,” continues Turtle. To address these fundamental questions, Hirayama and Turtle led a phase 1 clinical trial in LBCL patients combining a specific in-house manufactured CAR-T cell product (targeting a tumor antigen called CD19) with a PD-L1 antibody called durvalumab.
The crux of the study rested on a comparison between two groups of study participants: Group 1, which started receiving durvalumab after CAR-T cell infusion, and Group 2, which started receiving durvalumab before CAR-T cell infusion. After starting the durvalumab regimen, both groups then received monthly doses of durvalumab. As mentioned above, durvalumab targets PD-L1, which was thought to be used by LBCL tumors to deactivate CAR-T cells (which express PD-1). As Dr. Hirayama explains, “our initial hypothesis was that treatment with durvalumab before CAR-T cell therapy would prevent PD-L1 on tumors from interacting with inhibitory PD-1 on CAR-T cells, resulting in improved efficacy compared to delayed durvalumab administration.” Surprisingly, however, the researchers saw the exact opposite: patients given durvalumab before CAR-T cells saw lower response rates than those given durvalumab after! With some more digging, the team correlated this phenomenon to the timing of a spike in soluble PD-L1 (sPD-L1) levels in patients’ blood following durvalumab treatment. Importantly, they also confirmed that sPD-L1 could inhibit T-cell function in vitro. As Hirayama explains, “our observations in this trial suggest that the relative timing of durvalumab initiation and CAR-T cells really matters. CAR-T cell populations peak in patients at a fixed time following infusion, and sPD-L1 levels similarly spike after a certain stereotyped period following durvalumab administration—in the group of patients receiving durvalumab before CAR-T cells, these two peaks overlapped, while in the other group, CAR-T cell counts peaked before sPD-L1 levels did, which may explain the preserved efficacy in this group.”
Like any good study, this one leaves many questions. What is responsible for the spike in sPD-L1 levels following durvalumab administration? How would CAR-T cells interact with checkpoint inhibitor therapies targeting PD-1? What about the possibility of genetically engineering CAR-T cells to lack PD-1 altogether? Although Hirayama and Turtle note a general lack of enthusiasm surrounding CAR-T cells and checkpoint inhibitor therapies due to recent, less-than-promising trial results, they are quick to note that the LBCL patients receiving CAR-T cells and durvalumab in this trial had longer durations of response than those receiving CAR-T cells alone in a previous trial of theirs. They also stress the need for more rigorous studies before discounting a potentially fruitful treatment strategy. As the pair explain, “Here, we find that the precise manner in which these two therapies are combined can make a difference in patient responses; thus, study design may contribute to the mixed findings of other trials which may have not considered the timing of the treatments as closely. Regardless of the ultimate outcome, counterintuitive findings like ours argue for continued, rigorous study of CAR-T cell/checkpoint inhibitor combination therapies against different cancer types before we discount what could end up being a significant therapeutic advancement.”
The spotlighted work was funded by the National Institutes of Health, the Life Science Discovery Fund, the Bezos family, the Anderson Family, the Fred Hutch Immunotherapy Integrated Research Center, Juno Therapeutics (a Bristol Myers Squibb company), and AstraZeneca.
Fred Hutch/University of Washington/Seattle Children’s Cancer Consortium members Drs. Cecilia Yeung, Ryan Cassaday, Aude Chapuis, Damian Green, Hans-Peter Kiem, Filippo Milano, Mazyar Shadman, Brian Till, Stanley Riddell, and David Maloney contributed to this study.
Hirayama, A. V., Kimble, E. L., Wright, J. H., Fiorenza, S., Gauthier, J., Voutsinas, J. M., Wu, Q., Yeung, C. C. S., Gazeau, N., Pender, B. S., Kirchmeier, D. R., Torkelson, A., Chutnik, A. N., Cassaday, R. D., Chapuis, A. G., Green, D. J., Kiem, H.-P., Milano, F., Shadman, M., … Turtle, C. J. (2024). Timing of anti–PD-L1 antibody initiation affects efficacy/toxicity of CD19 CAR T-cell therapy for large B-cell lymphoma. Blood Advances, 8(2), 453–467.