Mechanistic and Translational Insights of CARv3-TEAM-E in Glioblastoma

Introduction

Glioblastoma (GBM) remains among the most lethal adult solid tumors, with a median overall survival of approximately 15 months despite maximal therapy and virtually no effective salvage options at recurrence.¹ The current standard-of-care paradigm—maximal safe resection followed by temozolomide-based chemoradiotherapy and adjuvant temozolomide—has remained fundamentally unchanged since the landmark EORTC-NCIC study more than two decades ago.¹ Although tumor-treating fields demonstrated an incremental survival benefit in the EF-14 trial, the broader therapeutic trajectory of GBM has remained largely unaltered.²

Attempts to translate the success of immunotherapy into GBM have thus far produced disappointing results. Immune checkpoint inhibition, which transformed the treatment landscape across multiple solid tumors, has failed to demonstrate meaningful survival benefit in phase 3 GBM studies.³ This resistance reflects several converging biological barriers, including profound intratumoral immunosuppression, extensive spatial and molecular heterogeneity, enrichment of regulatory T-cell (Treg) and myeloid-derived suppressor cell populations, and the blood-brain barrier (BBB) as an additional anatomic constraint limiting immune trafficking and therapeutic penetration.³ Collectively, these features create a tumor microenvironment (TME) that is both immunologically hostile and evolutionarily adaptable.

Chimeric antigen receptor (CAR) T-cell therapy has fundamentally reshaped the treatment of refractory hematologic malignancies through its ability to redirect cytotoxic lymphocytes with high antigen specificity.⁴ However, the biological requirements for effective CAR T-cell therapy in solid tumors are substantially more complex. Unlike hematologic cancers, where target antigens are often uniformly expressed and readily accessible, GBM exhibits marked intratumoral antigen heterogeneity, enabling antigen-loss escape under selective immune pressure.⁵ Simultaneously, infiltrating effector populations encounter a profoundly suppressive TME that limits persistence, trafficking, and sustained cytotoxic activity before durable tumor eradication can be achieved.

Within this context, epidermal growth factor receptor variant III (EGFRvIII) emerged as an attractive therapeutic target. EGFRvIII is a tumor-specific neoantigen generated through an in-frame deletion mutation that is absent from normal brain tissue, thereby creating a theoretically favorable therapeutic window for CAR T-cell targeting.⁶ However, early clinical experience also revealed the limitations of monospecific targeting strategies in GBM. In the phase 1 CART-EGFRvIII study, peripheral infusion of EGFRvIII-directed CAR T cells produced evidence of on-target biologic activity but failed to induce objective radiographic responses.⁶ Post-treatment tumor analyses demonstrated EGFRvIII antigen loss alongside dense Treg infiltration, establishing both antigen escape and adaptive immunosuppression as dominant resistance mechanisms.

These findings crystallized two central engineering challenges for next-generation GBM cellular therapy platforms: first, the need to overcome antigen heterogeneity through dual- or multi-antigen targeting strategies; and second, the need to neutralize or repurpose immunosuppressive cellular populations within the TME. The CARv3-TEAM-E construct, evaluated in the INCIPIENT trial, was specifically engineered to address both.⁷ By combining EGFRvIII-directed CAR signaling with local secretion of a T cell–engaging antibody molecule (TEAM) against wild-type EGFR, this platform attempts to simultaneously mitigate antigen-loss escape while reshaping the local immune architecture within the CNS tumor microenvironment.

Here, we review the mechanistic rationale, early clinical findings, and translational implications of the INCIPIENT study, contextualizing CARv3-TEAM-E within the broader evolution of CNS-directed cellular immunotherapy and the next phase of engineered immune therapies for solid tumors.

Mechanistic Rationale: Engineering Beyond a Single Antigen

The CARv3-TEAM-E construct incorporates three distinct transgenes within a single lentiviral vector. The primary CAR component is a second-generation anti-EGFRvIII construct containing a 4-1BB costimulatory domain linked to CD3-zeta, designed to enhance T-cell persistence and functional durability relative to CD28-based architectures.⁸ The second component is a T cell–engaging antibody molecule (TEAM) targeting wild-type EGFR, which is expressed across the majority of GBMs but absent from normal brain parenchyma, thereby creating a potential therapeutic window for CNS-directed EGFR targeting.⁷ A truncated CD19 molecule serves as the third transgene and functions as a manufacturing marker without direct therapeutic activity.

The mechanistic innovation of this platform lies in how the TEAM functions within the tumor microenvironment. Rather than circulating systemically, the secreted TEAM becomes concentrated locally at sites where CAR T cells engage cognate antigen, thereby generating a spatially confined, high-effector-density microenvironment while limiting systemic exposure.⁹ This localized amplification addresses a central limitation of conventional systemically administered bispecific antibodies, which must balance adequate tumor-site concentration against off-target toxicity in EGFR-expressing tissues such as skin, gastrointestinal mucosa, and pulmonary epithelium.

Equally important is the TEAM’s demonstrated ability to redirect Treg populations against tumor targets in vitro.¹⁰ If operative in vivo, this mechanism would represent a functional inversion of one of the dominant immunosuppressive populations within GBM. Rather than merely overcoming Treg-mediated immune suppression, CARv3-TEAM-E may potentially repurpose these cells into antitumor effectors within the local CNS microenvironment.

Collectively, the construct represents more than a dual-targeting platform. It reflects an evolution toward compartmentalized, locally amplified cellular immunotherapy designed specifically to address the biologic constraints that have historically limited CAR T-cell efficacy in solid tumors.

The INCIPIENT Trial: Study Design and Delivery Strategy

INCIPIENT (NCT05660369) is a nonrandomized, open-label, single-site phase 1 study conducted at Massachusetts General Hospital evaluating intraventricular administration of CARv3-TEAM-E T cells in patients with recurrent or newly diagnosed GBM.⁷ Eligible patients required WHO grade 4 GBM with EGFRvIII-positive disease and measurable lesions; prior EGFRvIII-targeted therapy was exclusionary. The initial safety cohort enrolled three participants between March and July 2023, each receiving a single intraventricular infusion of 10×10⁶ CAR-positive CARv3-TEAM-E T cells through an Ommaya reservoir.

The intraventricular route represents a deliberate pharmacokinetic and biologic strategy. By delivering effector cells directly into the cerebrospinal fluid (CSF) compartment, this approach circumvents BBB trafficking limitations that have historically restricted the efficacy of peripherally infused CNS-directed immunotherapies.¹¹ In addition, intraventricular delivery enables convective distribution across leptomeningeal surfaces and infiltrative parenchymal disease, potentially enhancing regional exposure throughout the CNS compartment.

This strategy builds on prior CNS CAR T-cell experience. IL13Rα2-directed CAR T cells previously demonstrated the ability to induce complete response in a patient with leptomeningeal GBM, while early-phase intraventricular B7-H3–targeted CAR T-cell studies established biologic activity and acceptable safety in diffuse intrinsic pontine glioma.^12,13 INCIPIENT extends this paradigm further by integrating compartmental delivery with dual-targeting immune engineering in infiltrative parenchymal GBM.

The manufacturing workflow evolved during the study. In Participant 1, Ommaya placement occurred separately from tumor tissue acquisition. Subsequent protocol modifications allowed Participants 2 and 3 to undergo concurrent craniotomy, tissue acquisition, and reservoir placement during a single operative procedure, thereby reducing procedural burden and streamlining vein-to-vein time in a disease where rapid progression may narrow therapeutic opportunity.

Clinical Results: Radiographic Responses and Correlative Biology

The radiographic responses observed in INCIPIENT were rapid and, in several cases, dramatic, representing a qualitative departure from prior GBM immunotherapy experience. Participant 1, a 74-year-old man with recurrent EGFRvIII-positive, MGMT-methylated GBM, demonstrated radiographic tumor regression within 24 hours of infusion, with continued improvement on serial imaging over the subsequent two weeks.⁷ Although the response was ultimately transient, progression occurring at day 72 was accompanied by post-treatment biopsy demonstrating EGFRvIII loss with retained EGFR copy-number gain, consistent with antigen modulation as an adaptive escape mechanism analogous to that observed in the prior CART-EGFRvIII study.⁶

Participant 2, a 72-year-old man with recurrent EGFRvIII-positive GBM following chemoradiotherapy and tumor-treating field therapy, demonstrated an 18.5% reduction in cross-sectional tumor area by day 2. By day 69, this response deepened to a 60.7% reduction from baseline and remained durable beyond 150 days after a single infusion without glucocorticoid or antiangiogenic support.⁷ Within the historical context of recurrent GBM, such durability following a single cellular therapy infusion is highly notable.

Participant 3 provided perhaps the most biologically informative observation in the study. Despite recurrent tumor tissue demonstrating loss of EGFRvIII expression at enrollment—with persistence of wild-type EGFR amplification and transcript expression—the patient nonetheless experienced near-complete tumor regression by day 5 following infusion.⁷ This finding provides in-human evidence that the TEAM component can independently mediate antitumor activity through wild-type EGFR targeting even in the absence of meaningful EGFRvIII expression. Mechanistically, this substantially broadens the implications of the platform beyond strictly EGFRvIII-driven disease.

Longitudinal extracellular-vesicle (EV) RNA analysis from CSF and peripheral blood provided important pharmacodynamic correlates. Across all participants, EGFRvIII and wild-type EGFR copy numbers within CSF-derived EV-RNA declined following infusion and ultimately became undetectable in post-treatment samples.^7,14 Peripheral blood EV-RNA demonstrated analogous but lower-magnitude reductions. These findings suggest that EV-RNA profiling may emerge as a minimally invasive biomarker strategy for monitoring CAR T-cell antitumor activity, treatment response, and early antigen escape in future GBM cellular therapy studies.

Safety Profile and Pharmacokinetic Correlates

No dose-limiting toxic effects were observed in the initial safety cohort. Treatment-related adverse events included grade 3 encephalopathy in Participant 1 and grade 3 fatigue in Participant 3. Participants 2 and 3 developed transient cyclic fevers and asymptomatic pulmonary nodules with ground-glass opacities that resolved spontaneously over several weeks.⁷ Intravenous anakinra was used intermittently for fever management, and notably, no glucocorticoids were administered for treatment-related toxicity.

This steroid-sparing approach is clinically meaningful. Corticosteroids, while often necessary during conventional CAR T-cell toxicity management, may impair CAR T-cell persistence and functional activity through broad lymphosuppressive effects.¹⁵ The use of IL-1 blockade through anakinra instead reflects an evolving supportive care paradigm designed to preserve CAR T-cell functionality while mitigating inflammatory toxicity.

Pharmacokinetic analyses revealed additional mechanistic insights. CAR-positive and TEAM-positive T cells remained detectable within the CSF for several weeks following infusion but declined substantially by week 4 in all participants. Similarly, CSF cytokines—including interferon-γ, TNF-α, and interleukin-6—peaked early before normalizing over the same interval, suggesting that the dominant effector phase is acute and temporally constrained in the absence of persistence-enhancing strategies.

Importantly, compartmental analysis of TEAM binding demonstrated profound localization within the CNS compartment. At day 21, circulating CAR T cells in peripheral blood demonstrated minimal TEAM surface binding, whereas CSF-derived CAR-positive cells showed markedly higher binding percentages.⁷ This compartmentalization provides mechanistic reassurance regarding systemic EGFR-directed toxicity and supports the hypothesis that local amplification within the CNS compartment enables therapeutic wild-type EGFR targeting without the systemic toxicities typically associated with peripheral EGFR inhibition.

The Persistence Problem: Toward Durable Response

The central challenge emerging from INCIPIENT is persistence. In two of three participants, tumor progression temporally correlated with decline and eventual loss of detectable CARv3-TEAM-E cells within the CSF compartment. This pattern—rapid initial cytoreduction followed by effector exhaustion or disappearance—mirrors early experiences with first-generation CAR T-cell therapies in hematologic malignancies and reinforces that initial potency alone is insufficient without sustained persistence.¹⁶

Several strategies warrant further exploration. Lymphodepleting preconditioning chemotherapy, routinely employed in hematologic CAR T-cell therapy, may enhance engraftment and persistence through depletion of endogenous lymphocyte populations competing for homeostatic cytokines such as IL-7 and IL-15.¹⁷ However, application within the CNS-directed setting requires careful neurologic risk assessment.

Repeat scheduled dosing may represent a more immediately translatable strategy. Participant 1 received a second infusion at day 37 following initial response, providing proof of principle for iterative dosing approaches. Additional next-generation engineering strategies—including IL-15 or IL-21 armoring, dominant-negative TGF-β receptor incorporation, and transcriptional programs enforcing stem-like T-cell phenotypes—remain active areas of investigation aimed at overcoming exhaustion and improving durability.¹⁶

Antigen heterogeneity also remains an unresolved challenge. While Participant 1 demonstrated EGFRvIII loss under selective immune pressure, Participant 3 simultaneously demonstrated that wild-type EGFR targeting through TEAM secretion can retain meaningful biologic activity even after EGFRvIII loss. This observation raises an important translational question: whether future iterations of this platform may ultimately expand eligibility beyond EGFRvIII-positive disease into broader wild-type EGFR–expressing GBM populations.¹⁸

Contextualizing INCIPIENT Within CNS CAR T-Cell Development

INCIPIENT should be interpreted within the broader trajectory of CNS-directed CAR T-cell development rather than as an isolated signal. Previous studies established that CAR T cells can be safely delivered through both peripheral and intracranial routes in glioma patients.^12,13,19,20 The IL13Rα2-directed CAR T-cell experience demonstrated that intracranial delivery could produce complete response in recurrent leptomeningeal GBM, albeit requiring multiple sequential infusions.12 GD2-directed CAR T cells subsequently demonstrated biologic activity in H3K27M-mutated diffuse midline gliomas.20 B7-H3–targeted intraventricular CAR T cells further established early evidence of safety and bioactivity in diffuse intrinsic pontine glioma.¹³

Collectively, these studies established the feasibility of compartmental CNS cellular immunotherapy. INCIPIENT advances this framework further by integrating dual-targeting immune engineering with intraventricular delivery and demonstrating that even a single infusion can access infiltrative parenchymal disease beyond leptomeningeal surfaces.

The comparison with the earlier CART-EGFRvIII experience is particularly instructive. Monospecific EGFRvIII-directed CAR T cells induced biologic activity but failed to generate objective radiographic responses, with antigen escape emerging as the dominant resistance mechanism.⁶ CARv3-TEAM-E was specifically engineered to address this limitation, and the INCIPIENT findings suggest that the underlying design hypothesis translates into clinically meaningful activity.

The advance is therefore not simply the addition of a second antigenic target. Rather, it is the development of a compartmentalized, locally amplified immune architecture capable of addressing intratumoral heterogeneity while simultaneously limiting systemic toxicity through spatial confinement of biologic activity.

Translational Implications for Clinical Practice and Trial Design

Although the sample size remains small, several translational implications emerge from INCIPIENT.

First, EGFRvIII positivity may not ultimately represent a strict prerequisite for therapeutic activity. Participant 3 demonstrated meaningful response despite recurrent disease lacking significant EGFRvIII expression, suggesting that wild-type EGFR expression itself may serve as a clinically actionable biomarker for future trial enrollment.¹⁸ Given the broad prevalence of wild-type EGFR expression across GBM, this could substantially expand the addressable patient population.

Second, EV-RNA liquid biopsy from CSF represents a promising pharmacodynamic monitoring platform. Longitudinal tracking of EGFRvIII and EGFR copy-number dynamics may provide a minimally invasive approach for assessing treatment response, identifying antigen escape, and potentially informing timing of retreatment or therapeutic transition.¹⁴

Third, the steroid-sparing supportive care approach used in INCIPIENT warrants attention. The use of anakinra rather than glucocorticoids aligns with broader efforts within the CAR T-cell field to preserve T-cell persistence and functionality while managing inflammatory toxicity.¹⁵

Finally, workflow optimization remains critically important. The protocol modifications enabling concurrent tumor tissue acquisition and Ommaya placement reduced procedural burden and shortened the path from diagnosis to infusion. In a rapidly progressive disease such as GBM, minimizing manufacturing and logistical delays may directly affect therapeutic feasibility and clinical outcome.

Conclusion

The INCIPIENT trial establishes CARv3-TEAM-E as a mechanistically sophisticated and biologically active platform for recurrent GBM. The rapid radiographic regressions observed—occurring in some cases within 24 to 72 hours of a single intraventricular infusion—represent a biologic signal unlike prior GBM immunotherapy experience. Equally important, the response observed in an EGFRvIII-negative recurrence provides proof of principle that locally secreted TEAM molecules can independently mediate clinically meaningful antitumor activity through wild-type EGFR engagement.

The defining challenge moving forward is persistence. Current data suggest that the dominant effector phase remains acute and temporally constrained, with recurrence emerging alongside loss of detectable CAR T-cell populations. Addressing this through optimized lymphodepletion, repeat dosing strategies, or next-generation engineering approaches designed to enhance T-cell durability and resistance to TME-mediated suppression will likely determine the long-term therapeutic viability of this platform.

INCIPIENT does not establish CARv3-TEAM-E as a standard of care. What it establishes is something potentially more important at this stage of development: a compelling proof of principle that compartmentalized, locally amplified cellular immunotherapy can generate rapid and meaningful biologic activity in a disease that has historically remained resistant to immunotherapeutic intervention.

References

1. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005;352(10):987-996. doi:10.1056/NEJMoa043330
2. Stupp R, Taillibert S, Kanner A, et al. Effect of tumor-treating fields plus maintenance temozolomide vs maintenance temozolomide alone on survival in patients with glioblastoma: a randomized clinical trial. JAMA. 2017;318(23):2306-2316. doi:10.1001/jama.2017.18718
3. Lim M, Xia Y, Bettegowda C, Weller M. Current state of immunotherapy for glioblastoma. Nat Rev Clin Oncol. 2018;15(7):422-442. doi:10.1038/s41571-018-0003-5
4. Neelapu SS, Locke FL, Bartlett NL, et al. Axicabtagene ciloleucel CAR T-cell therapy in refractory large B-cell lymphoma. N Engl J Med. 2017;377(26):2531-2544. doi:10.1056/NEJMoa1707447
5. Majzner RG, Mackall CL. Tumor antigen escape from CAR T-cell therapy. Cancer Discov. 2018;8(10):1219-1226. doi:10.1158/2159-8290.CD-18-0442
6. O'Rourke DM, Nasrallah MP, Desai A, et al. A single dose of peripherally infused EGFRvIII-directed CAR T cells mediates antigen loss and induces adaptive resistance in patients with recurrent glioblastoma. Sci Transl Med. 2017;9(399):eaaa0984. doi:10.1126/scitranslmed.aaa0984
7. Choi BD, Gerstner ER, Frigault MJ, et al. Intraventricular CARv3-TEAM-E T cells in recurrent glioblastoma. N Engl J Med. 2024;390(14):1290-1298. doi:10.1056/NEJMoa2314390
8. Guedan S, Ruella M, June CH. Emerging cellular therapies for cancer. Annu Rev Immunol. 2019;37:145-171. doi:10.1146/annurev-immunol-042718-041407
9. Choi BD, Yu X, Castano AP, et al. CAR-T cells secreting BiTEs circumvent antigen escape without detectable toxicity. Nat Biotechnol. 2019;37(9):1049-1058. doi:10.1038/s41587-019-0192-1
10. Choi BD, Gedeon PC, Herndon JE 2nd, et al. Human regulatory T cells kill tumor cells through granzyme-dependent cytotoxicity upon retargeting with a bispecific antibody. Cancer Immunol Res. 2013;1(3):163-167. doi:10.1158/2326-6066.CIR-13-0049
11. Ahmed N, Brawley V, Hegde M, et al. HER2-specific chimeric antigen receptor-modified virus-specific T cells for progressive glioblastoma: a phase 1 dose-escalation trial. JAMA Oncol. 2017;3(8):1094-1101. doi:10.1001/jamaoncol.2017.0184
12. Brown CE, Alizadeh D, Starr R, et al. Regression of glioblastoma after chimeric antigen receptor T-cell therapy. N Engl J Med. 2016;375(26):2561-2569. doi:10.1056/NEJMoa1610497
13. Vitanza NA, Wilson AL, Huang W, et al. Intraventricular B7-H3 CAR T cells for diffuse intrinsic pontine glioma: preliminary first-in-human bioactivity and safety. Cancer Discov. 2023;13(1):114-131. doi:10.1158/2159-8290.CD-22-0750
14. Batool SM, Muralidharan K, Hsia T, et al. Highly sensitive EGFRvIII detection in circulating extracellular vesicle RNA of glioma patients. Clin Cancer Res. 2022;28(19):4070-4082. doi:10.1158/1078-0432.CCR-22-0945
15. Amini L, Silbert SK, Maude SL, et al. Preparing for CAR T cell therapy: patient selection, bridging therapies and lymphodepletion. Nat Rev Clin Oncol. 2022;19(6):342-355. doi:10.1038/s41571-022-00607-3
16. Guedan S, Ruella M, June CH. Emerging cellular therapies for cancer. Annu Rev Immunol. 2019;37:145-171. doi:10.1146/annurev-immunol-042718-041407
17. Amini L, Silbert SK, Maude SL, et al. Preparing for CAR T cell therapy: patient selection, bridging therapies and lymphodepletion. Nat Rev Clin Oncol. 2022;19(6):342-355. doi:10.1038/s41571-022-00607-3
18. Hobbs J, Nikiforova MN, Fardo DW, et al. Paradoxical relationship between the degree of EGFR amplification and outcome in glioblastomas. Am J Surg Pathol. 2012;36(8):1186-1193. doi:10.1097/PAS.0b013e31825b05bb
19. Ahmed N, Brawley V, Hegde M, et al. HER2-specific chimeric antigen receptor-modified virus-specific T cells for progressive glioblastoma: a phase 1 dose-escalation trial. JAMA Oncol. 2017;3(8):1094-1101. doi:10.1001/jamaoncol.2017.0184
20. Majzner RG, Ramakrishna S, Yeom KW, et al. GD2-CAR T cell therapy for H3K27M-mutated diffuse midline gliomas. Nature. 2022;603(7903):934-941. doi:10.1038/s41586-022-04489-4