Durable regression of Medulloblastoma after regional and intravenous delivery of anti-HER2 chimeric antigen receptor T cells

Anandani Nellan, Christopher Rota, Robbie Majzner, Cynthia M. Lester-McCully, Andrea M. Griesinger, Jean M. Mulcahy Levy, Nicholas K. Foreman, Katherine E. Warren, Daniel W. Lee

Standard-of-care therapies for treating pediatric medulloblastoma have long-term side effects, even in children who are cured. One emerging modality of cancer therapy that could be equally effective without such side effects would be chimeric antigen receptor (CAR) T cells. Knowing that human epidermal growth factor receptor 2 (HER2) is overexpressed in many medulloblastomas and has been used as a CAR T target before, we sought to evaluate the efficacy of more sophisticated anti-HER2 CAR T cells, as well as the feasibility and efficacy of different routes of delivering these cells, for the treatment of pediatric medulloblastoma. Daoy, D283 and D425 medulloblastoma cell lines were characterized by flow cytometry to evaluate HER2 expression. Anti-tumor efficacy of HER2-BBz-CAR T cells in vitro was performed using cytokine release and immune cytotoxicity assays compared to control CD19 CAR T cells. In vivo, Daoy and D283 tumor cells were orthotopically implanted in the posterior fossa of NOD.Cg-Prkdc scid Il2rg tm1Wjl /SzJ (NSG) mice and treated with regional or intravenous HER2-BBz-CAR T cells or control CD19 CAR T cells. Non-human primates (NHPs) bearing ventricular and lumbar reservoirs were treated with target autologous cells bearing extracellular HER2 followed by autologous HER2-CAR T cells intraventricularly. Cerebrospinal fluid and blood were collected serially to measure the persistence of delivered cells and cytokines. HER2-BBz-CAR T cells effectively clear medulloblastoma orthotopically implanted in the posterior fossa of NSG mice via both regional and intravenous delivery in xenograft models. Intravenous delivery requires a log higher dose compared to regional delivery. NHPs tolerated intraventricular delivery of autologous cells bearing extracellular HER2 followed by HER2-BBz-CAR T cells without experiencing any systemic toxicity. HER2-BBz-CAR T cells show excellent pre-clinical efficacy in vitro and in mouse medulloblastoma models, and their intraventricular delivery is feasible and safe in NHPs. A clinical trial of HER2-BBz-CAR T cells directly delivered into cerebrospinal fluid should be designed for patients with relapsed medulloblastoma.