Use of HER2-Specific Chimeric Antigen Receptor-Modified Virus-Specific T Cells as a Potential Therapeutic for Progressive HER2-Positive Glioblastoma

Despite multimodality chemoradiotherapy treatment, the prognosis for glioblastoma multiforme (GBM) remains bleak with a 5-yr overall survival (OS) of less than 4% for adults.1 The tenacious resistance of GBM to traditional cytotoxic mediated chemotherapies has sparked clinical investigations in novel T cell-mediated immunotherapies with chimeric antigen receptors (CARs) targeted to cell surface proteins such as IL13Rα2, which is a glioma-linked antigen.2 Previously, in 2010, Ahmed et al3 showed that in preclinical xenograft models with patient derived GBMs, CAR T cells targeting human epidermal growth factor receptor-2 (HER2), killed “bulk” glioma cells and primary GBM stem cells.3 However, this strategy's limitation was the lack of necessary costimulatory signals to the T cell during interaction with its CAR.4 Furthermore, there was a lack of a clinical model that could reliably demonstrate T cell persistence or propagation in HER2 positive patients treated with HER2-CAR T cells.5 Thus, in phase 1 dose escalation trial, Ahmed et al,6 published in JAMA Oncology on April 20, 2017, used polyclonal virus-specific T cells (VSTs), enriched for cytomegalovirus (CMV), Epstein–Barr virus, and adenovirus, that also expressed HER2-CARs with a CD28. ζ-signaling endodomain to treat HER2-positive progressive GBM. These polyclonal VSTs not only provide antitumor activity through their CARs but also received costimulation following engagement by latent virus antigens. HER2/neu positive GBM are not commonly reported, but are associated with worse survival.7,8 The clinical trial conducted by Ahmed et al6 addresses an under-reported subgroup of GBMs by optimizing an immunotherapy with a specific target, HER2/neu. The primary endpoints included feasibility and safety. The secondary end points were cell persistence and antiglioblastoma activity. Investigators had a cohort size of 3 patients per dose level and they received 1 or more intravenous infusions of autologous HER2-CAR VSTs with 4 dose escalations starting at 1 × 106/m2 and then increasing to 3 × 106/m2, 1 × 107/m2, 3 × 107/m2, and 1 × 108/m2. Patients who showed an objective response at 6 wk or later received up to 6 additional doses of T cells at 6- to 12-wk intervals at the same dose level. Infusions were administered between July 25, 2011 and April 21, 2014. The median follow-up period was 8 mo. Patients who were eligible (HER2-positive GBM, CMV seropositivity, normal left ventricular ejection fraction, Karnofsky/Lansky performance score of 50 or more, and life expectancy of 6 wk or more at the time of T cell infusion) were allowed to receive temozolomide (TMZ) up to 2 d prior to infusion. There was no dose-limiting toxicity reported, but 2 patients had grade 2 seizures/headaches. Fifteen of the 17 patients had their highest frequency of HER2-CAR VSTs 3 h after the infusion, and at 6 wk 7 patients had qPCR-positive blood samples. However, this frequency declined, but results suggest that HER2-CAR VSTs can persist for 1 yr at a low frequency and do not expand. Infusion of multiple doses did not alter the in vivo fate of HER2-CAR VSTs. One patient received chemotherapy within 6 wk prior to T cell infusion, so 16 patients were evaluated for antiglioblastoma activity of HER2-CAR VSTs. One patient had a partial response (6%) and 7 (44%) had stable disease for 8 wk to 29 mo after the first T cell infusion. A notable 17-yr-old patient with an unresectable right thalamic GBM received 1 × 106/m2 of HER2-CAR VSTs and had a partial response for 9.2 mo. After a second infusion of 1 × 106/m2 HER2-CAR VSTs, this patient had stable disease and survived for 26.9 mo from the first infusion. Three patients (18%) are alive with stable disease for 29.0, 28.8, and 24.0 mo. Although 8 patients had progressive GBM, 6 of these patients survived for more than 6 mo. In this cohort, the median time to progression was 3.5 mo, the median OS was 11.1 mo after the first infusion, and 24.5 mo after diagnosis. The only correlation found was that in patients who did not receive salvage therapy prior to infusion had a significantly longer median OS (27.2 mo) vs patients who received infusions after salvage therapy (6.7 mo; P = .02). Infusion of up to 1 × 108/m2 was well tolerated without dose-limiting toxicities. This report shows some early safety and efficacy of HER2-CAR VSTs, but no definitive conclusions can be made regarding survival benefit. Limitations of this study include the small sample size, as well as attrition from screening patients for eligibility. Additionally, the monotherapy was shown to not expand in peripheral blood, so there is a need to enhance the antiglioblastoma activity of HER2-CAR VSTs. This clinical trial opens the doors to the possibilities of using combinatorial forces of immune system targets to better eradicate glioma cells, such as checkpoint inhibitors that target programmed cell death protein 1 (NCT02017717) or transforming growth factor β.9 Using HER2-CAR VSTs as a monotherapy has shed light on a new treatment strategy for a deadly subtype of an already aggressive malignancy. Further evaluation of HER2-CAR VSTs is necessary likely in combination with other immunomodulatory approaches to enhance the potency, growth, and survivability of HER2-CAR VSTs.