Designing The Next Wave Of T Cell Engagers
A conversation with Adam J. Pelzek, Ph.D., associate director of biology, Abpro
Adam J. Pelzek, Ph.D., is associate director of biology at Abpro, where he has led programs in immuno-oncology research and development with a focus on developing therapeutic T cell engager bispecific antibody agents for the targeting of solid tumors (breast, gastric, liver). We caught up with him to discuss designing novel T cell engagers and how industry can pave a future for the field.
T cell engagers are attracting (renewed?) attention in oncology. From your perspective, what makes them a promising class, and what scientific or clinical hurdles still remain?
T cell engagers (TCEs) for hematological malignancies, specifically those with an arm targeting CD3, have been very successful, with approvals including targeted therapies against CD19, CD20, BCMA, and GPRC5D. Interestingly, in 2024, we saw the first approval in a solid tumor indication for a TCE against DLL3. CD3-directed TCEs are promising as a class of drugs, as they have a proven clinically validated mechanism of action, and they are available off-the-shelf in contrast to current cell-based targeted therapies, like CAR-T. Momentum in the TCE field suggests that we are likely to see more approvals in solid tumor indications in the years to come.
One challenge faced by TCEs is to find a therapeutic window for dosing that does not promote severe toxicity due to cytokine release syndrome, while still potentiating T cell killing of the tumor. Delivering an efficacious but tolerable clinical dose requires that potential toxicities are considered at early stages of design and development, toward differentiated molecules that allow for an improved therapeutic window.
Selectivity is a known challenge for T cell engagers targeting tumor-associated antigens like HER2. What are some strategies being explored across the field to address off-tumor toxicity?
Unless a tumor-associated antigen is onco-fetal (not expressed in normal tissues) in nature, there is some risk of harming on-target, off-tumor tissues with TCEs, even when the expression of the tumor antigen is highly expressed on the tumor cell surface and relatively low on normal tissues. Some past trials for HER2 x CD3 TCEs have failed due to dose-limiting toxicities, and therefore, having fully potent interactions may not be optimal.
Designing TCEs that promote on-target, on-tumor selectivity is one way to address this issue. With a bivalent approach for the tumor antigen, tuning the selectivity of the targeting arms makes the T cell interaction lean more on avidity-driven effects, which may help to mitigate the risk of off-tumor tissue damage and on-target, off-tumor cytokine release. ABP-102/CT-P72, our HER2 x CD3 TCE agent in co-development with Celltrion, takes advantage of this principle and drives cell killing and cytokine release proportional to HER2 density. While killing and cytokine release are reduced on HER2-low cells, the preclinical therapeutic index of ABP-102/CT-P72 suggests it may be sufficient to treat HER2-low tumors.
Other approaches under investigation to promote selectivity include multi-specific approaches with more than one tumor antigen targeting moiety (multiple epitopes on the same target antigen, or multiple distinct antigens likely to be on the same tumor cell). Another approach involves the design of conditionally activated TCEs, such as those with protease-cleavable protein masks for the tumor antigen targeting arm. A more selective TCE that is amenable to higher dosing may help push the concentration of the antibody higher in the tumor microenvironment, which could lead to more T cell engagement with target-expressing tumors.
Cytokine release syndrome (CRS) continues to be a major concern. What are some of the innovative approaches to mitigating CRS in the next wave of T cell engagers?
For CD3-directed TCEs, the potential for systemic CRS seems to be related to the strength of the interaction with CD3 on T cells. Therefore, reduction of the CD3 arm affinity may make a difference for less severe CRS, and these types of modifications may also have some beneficial effects on pharmacokinetics and biodistribution. The valency of the interaction with CD3 also makes a difference, and many therapeutic candidates have a 1 x 1 format with a single CD3-arm. Interestingly, the design of ABP-102/CT-P72 allows for functional monovalency with CD3, which makes it unlikely to cluster T cell receptors independent of tumor antigen, and it does not initiate spontaneous cytokine release from T cells in the absence of target cells.
Other approaches under exploration for TCEs include the use of cleavable protein masks for the CD3 arm, and some progress has been made in the use of non-CD3 T cell engagement strategies to enhance T cell activity and avoid the clinical complications of prior immunostimulatory agents (4-1BB, CD28, etc.) or block immune checkpoints (PD-1, CTLA-4, etc.).
Looking at the field as a whole, what do you see as the key features that define a “next-generation” T cell engager — and what might ultimately separate clinical winners from the rest?
Developers of next-generation TCEs have the benefit of past successes and failures to inform new designs. Engineering toward enhanced patient safety while maintaining anti-tumor potential is an approach that is likely to persist, which should allow broadening of the therapeutic index from what was seen with some of the past TCEs. We have approached the design of ABP-102/CT-P72 with these lessons in mind. For example, ABP-102/CT-P72 was well-tolerated at doses exceeding 180 times the maximum tolerated dose of its parental antibody in non-human primate studies, suggesting a broader therapeutic window that could translate to safer dosing in patients.
Design of a novel TCE requires that key choices be made on format, valency, size, presence or absence of an Fc region, Fc modifications to reduce effector functionality, additional design features in conditionally active molecules, and even delivery mechanisms (with some companies attempting novel nucleic acid-based approaches to deliver genes or mRNA encoding TCEs for in-organism production and delivery). These design features and approaches can have profound effects on TCE half-life, tumor/tissue penetrance, and pharmacodynamics. Tumor escape (i.e., immune checkpoint) and resistance mechanisms related to tumor-associated antigen biology are also emerging challenges for TCEs, in addition to intra-tumor heterogeneity of antigen expression.
You recently presented preclinical data on ABP-102/CT-P72. How does this program reflect some of the design and development principles you think are important for future T cell engagers?
In ABP-102/CT-P72, we have attempted to incorporate selectivity as a key feature into the design of the molecule; specifically, higher densities of tumor target antigen should drive avidity-based interactions mediated by our antibody, allowing T cells to become activated to kill tumor cells. Our approach is geared toward avoiding on-target, off-tumor toxicity. In addition, reduced affinity and functional monovalency of the CD3 arm are features that appear to align with trends for improved safety, addressing challenges that have hampered prior HER2 TCEs. We believe these features contributed to significantly reduced cytokine release in HER2-low models while maintaining cytotoxicity in HER2-high models, underscoring the program’s potential for tumor-selective activity and safety. Some other novel TCEs have features that allow for regulation of safety through conditional activity. As we look to initiate first-in-human clinical testing of ABP-102/CT-P72 in the first half of 2026, we are interested in seeing whether our approach for TCE design will translate to clinical efficacy, and whether a similar approach could be applicable to other solid tumor targets.
As the space evolves, what types of partnerships, technologies, or development strategies will be essential to push T cell engagers forward — particularly into solid tumors?
In the TCE development space, design considerations are abundant, and the questions for how to best design novel TCEs for specific solid tumor indications are wide open, as they have largely not been addressed yet in the clinic. Clues will emerge with each clinical success. Innovative technologies for drug design (structural modeling, AI de novo design of potential agents) will take advantage of the groundwork from many decades’ worth of hard-won structural biology data and are likely to continue to enhance discovery and molecular design efforts. However, drug development is a long game, and for biotechnology companies, co-development partnerships can help accelerate next-gen drugs into the clinic.
About The Expert:
Adam J. Pelzek, Ph.D., is associate director of biology at Abpro, where he has led programs in immuno-oncology research and development with a focus on engineering, validation, and pharmacology workflows toward novel therapeutic T cell engager bispecific antibody agents for the targeting of solid tumors (breast, gastric, liver). He earned a doctorate in pathology in the Immunology & Inflammation program at New York University School of Medicine, with a focus on B cells and human antibodies in autoimmunity and infectious disease, and a BS in molecular biology at the University of Wisconsin-Madison. He has over 10 years of preclinical, translational, and clinical research and development experience in immunology. Pelzek’s published work and scientific contributions include manuscripts and patents spanning the fields of immunology, microbiology and infectious diseases, autoimmunity, drug development, and oncology.