Therapy of B Cell Malignancies with CD19-Specific Chimeric Antigen Receptor-Modified T Cells of Defined Subset Composition

BACKGROUND:

The adoptive transfer of CD19-specific chimeric antigen receptor-modified (CD19 CAR) T cells is a promising strategy for treating patients with CD19⁺ B cell acute lymphoblastic leukemia (ALL), chronic lymphocytic leukemia (CLL), and non-Hodgkin lymphoma (NHL). Dramatic responses have been observed in a subset of patients receiving CD19 CAR T cell therapy, and prior studies suggest that persistence of transferred T cells may correlate with the extent of tumor regression. The use of unselected T cells to prepare CAR T cells results in variation in the phenotypic composition of the infused product in individual patients, making it difficult to determine whether particular T cell subsets contribute to efficacy and/or toxicity. Studies in our lab demonstrated that genetically modified effector T cells derived from purified T cell subsets differ in the capacity to persist in vivo after adoptive transfer, and that a combination of CAR-modified CD8⁺ central memory (T_CM) and CD4⁺ T cells provides optimal antitumor activity in tumor xenograft models. Based on these data, we designed the first clinical trial in which patients with CD19⁺ B cell malignancies receive CD19 CAR T cells comprised of a defined composition of CD8⁺ T_CM and CD4⁺T cells engineered to express a CD19 CAR.

METHODS:

Patients with relapsed or refractory CD19⁺ ALL, CLL or NHL are eligible for this phase I/II study. CD8⁺ T_CM and CD4⁺ T cells were separately enriched by immunomagnetic selection from a leukapheresis product from each patient, and cryopreserved. The CD8⁺ T_CM and CD4⁺ T cells were stimulated in independent cultures with anti-CD3/anti-CD28 paramagnetic beads, and transduced with a lentivirus encoding the murine FMC63 anti-CD19 scFv, 4-1BB and CD3 zeta signaling domains. After in vitro expansion, the cell product for infusion was formulated in a 1:1 ratio of CD4⁺:CD8⁺ CAR⁺ T cells. A truncated non-functional human epidermal growth factor receptor (EGFRt) encoded in the transgene cassette allowed identification of transgene-expressing T cells by flow cytometry. Lymphodepleting chemotherapy was administered followed by infusion of EGFRt⁺ CAR T cells at one of three dose levels (2 x 10⁵ EGFRt⁺ cells/kg, 2 x 10⁶ EGFRt⁺ cells/kg, 2 x 10⁷ EGFRt⁺cells/kg).

RESULTS:

Twenty patients with relapsed or refractory ALL (n = 9), NHL (n = 10) or CLL (n = 1), including those who failed prior autologous (n = 4) or allogeneic (n = 4) stem cell transplant have been treated on the trial. Fifteen of 20 treated patients received a product that conformed to the prescribed CD8⁺ T­_CM:CD4 composition. Five patients received a product manufactured using a modified strategy either due to low blood lymphocyte counts (n = 3) or due to failure to propagate T cells in culture (n = 2). CD8⁺ T_CM and CD4⁺ T cells have been isolated from 12 additional patients and cryopreserved for therapy. Patients have been treated at all three dose levels without acute infusional toxicity. Severe cytokine release syndrome (sCRS) consisting of fever, hypotension, and reversible neurotoxicity associated with elevated serum IFN-γ and IL-6 was only observed in ALL patients with a high tumor burden. One ALL patient treated at the highest cell dose died of complications associated with sCRS. None of the NHL patients had sCRS. Of patients who are >6 weeks after CD19 CAR T cell therapy, best responses included complete (n=1) or partial (n=5) remission in 6/9 patients with NHL and complete remission in 5/7 patients with ALL. Both CD4⁺ and CD8⁺ CAR-T cells expanded in vivo and could be detected in blood, marrow and CSF. The peak level and duration of persistence of both CD4⁺ and CD8⁺ EGFRt⁺ T cells were associated with clinical response. TCRBV gene sequencing of flow sorted CD4⁺ and CD8⁺ EGFRt⁺CAR T cells from 2 patients showed that proliferating CAR T cells were polyclonal. A subset of NHL patients in whom CAR T cells became undetectable developed a T cell immune response to sequences in the murine CD19-specific scFv component of the CAR transgene.

CONCLUSION:

Adoptive immunotherapy with CD19 CAR T cells of defined subset composition is feasible and safe in a majority of heavily pretreated patients with refractory B cell malignancies and has potent anti-tumor activity. Persistence of CAR-T cells may be limited in some patients by transgene product immunogenicity. Data from this ongoing clinical trial will be updated at the meeting.