New CAR T-cell therapy uses shorter linker, leading to more potent therapy for ALL
CD19-directed chimeric antigen receptor T-cell therapies have produced unprecedented response rates among patients with advanced hematologic malignancies, yet many patients who respond to this therapy experience disease relapse.
To combat this problem, several ongoing clinical trials are evaluating new CAR constructs that target novel antigens or use a bispecific approach targeting multiple antigens.
One study set out to evaluate CD22 as a target for CAR T-cell immunotherapy. After analyzing results of two pilot studies of CD22-directed CAR T-cell therapy among six children and three adults with acute lymphoblastic leukemia, researchers found response rates were lower than anticipated, especially compared with results of another study by the NCI that had used CD22 as a target antigen.
Study investigator Marco Ruella, MD — assistant professor of medicine in the division of hematology/oncology and Center for Cellular Immunotherapies and scientific director of the lymphoma program at Hospital of the University of Pennsylvania — spoke with Healio about how modifications to shorten the CAR’s single-chain variable fragment created a more potent and effective therapy and how this knowledge may impact the development of CAR T cells moving forward.
Healio: Can you explain your group’s rationale for the novel CAR design used in this study?
Ruella: We set out to develop a new CAR that targeted CD22 early on, and we applied a CAR design similar to that which we applied to CD19-directed CAR-T. This means we used a standard linker between the heavy and light chain. The concept was pretty similar to other CARs we have developed at University of Pennsylvania. We assumed that this similar design would work efficiently, and preclinical testing showed both in vitro and in vivo responses. We also observed responses to therapy initially in our clinical trial, but the number of patients who responded and their duration of response were suboptimal compared with other groups of patients using similar CAR constructs. Therefore, our group needed to find a way to improve this CAR T-cell therapy to make it more successful.
Healio: What function does the single-chain variable fragment serve for CAR T cells?
Ruella: The single-chain variable fragment is the portion of the CAR that binds to the antigen. The linker between the heavy and light chains bridges the two halves of the single-chain variable fragment. So, it is a very important component that affects CAR conformation and the way it binds to the antigen.
Healio: Did you aim to treat patients who previously progressed on CD19-directed CAR T cells or to verify CD22 as a suitable antigen target in addition to CD19?
Ruella: The way we designed the clinical trial was sort of agnostic but, by the end of the study, most patients treated with CD22 CAR T cells in this trial did not respond or had disease progression after receiving CD19 CAR T cells. Therefore, these patients needed a different approach — either a new type of CAR T cell or a CD19 bispecific antibody. In the long run, a more effective strategy would likely be a dual approach that combines CD22 CAR T cells with CD19 CAR T cells. An ongoing trial at Penn uses the new CAR construct described in our recent publication. In that trial, CD19 CAR is given first, followed by CD22 CAR T cells. We hope to report results from this study soon.
Healio: What is the key takeaway from your study?
Ruella: The CAR design, in particular the design of the antigen binding domain, is very important for driving in vivo antitumor activity in humans. Even minor changes in this portion of the CAR, such as in the length of a linker within the single-chain variable fragment, can have a drastic impact on clinical outcomes. As we described in our study, changing the length of this linker creates a CAR that can dimerize, and the dimerization of the CAR enhances the activation of our T cell. This increased activation leads to better antitumor effects.
Healio: Did any of your findings surprise you?
Ruella: We did not expect that such a small modification would have such a large effect on the function of the CAR T cells. The other surprise was that this shortened-linker approach works in other CAR constructs. We have shown that it works on a CD33-targeting CAR, but we still need to test every target that we evaluate. For example, when we used a CD19 CAR and tried to shorten the linker within the single-chain variable fragment, we found that it did not drive activation of the CAR T cells.
Interestingly, this dimerization characteristically led to a tonic signaling that lowered the threshold of activation for the CAR T cells. In the past, this was believed to be a detrimental approach because it could lead to T-cell exhaustion, but in our case, we have shown that this is beneficial for CD22 CAR T-cells, and perhaps also for CD33 CAR T cells.
Healio: Do the results of your study have any potential impact on clinical practice?
Ruella: They're already having an impact. Once we realized that we had to conduct research to prove the new CAR’s feasibility we did just that, and subsequently we started two clinical trials that are already treating patients. One trial includes CD22 CAR T cells that we gave to both pediatric and adult patients. I referred to the adult portion of this trial earlier, where we are treating patients with a single CD22 CAR T-cell infusion. The study protocol also includes CD19 CAR T cells in a sequential approach, where we give CD19 followed by CD22 CAR T cells. Moreover, our study has already had a pretty huge impact because of the resulting two clinical trials but also because it has the potential to produce more products that will be dual targeting, including the new CD22 CAR T cells that we developed.
For more information:
Marco Ruella, MD, can be reached at Perelman Center for Advanced Medicine, 3400 Civic Center Blvd., Philadelphia, PA 19104; email: firstname.lastname@example.org.