A New Chimeric Antigen Receptor (CAR) Targeting the CD23 Antigen Expressed by Chronic Lymphocytic Leukemia (B-CLL) Cells

Abstract 2446 B-Chronic lymphocytic leukemia (B-CLL) is characterized by a progressive accumulation of B-lymphocytes expressing CD19, CD20 dim and aberrantly expressing the CD5 T-cell marker. Moreover, they over-express the B-cell activation marker CD23. Chimeric Antigen Receptors (CAR) are engineered molecules able to redirect T-cell killing/effector activity towards a selected target in a non MHC-restricted manner. First trials targeting B-CLL were based on both monoclonal antibodies and anti-CD19/anti-CD20.CAR-transduced T cells. However, this approach causes the elimination of normal B-lymphocytes and B-precursors with consequent impairment of humoral immunity. Selective CD23 expression on B-CLL cells renders this molecule an optimal target to design a specific CAR. We have generated a novel CD23-targeting CAR to redirect T cells against CD23 + B-CLL. Transduced T cells were tested for cytotoxicity against different CD23 + -targets, using a classic 51 Chromium release assay, and for specific cytokine release, by multiplex flow cytomix assay. T cells from B-CLL patients were efficiently transduced with the anti-CD23.CAR (average expression 68%, n=10) and redirected specifically toward autologous blasts (average lysis 58%, n=5). On the contrary, anti-CD23 transduced T-cells did not displayed any relevant killing versus normal B cells (average lysis 13%, n=3), differently from anti-CD19.CAR redirected T-cells, which killed tumor and normal B cells in an indistinct manner. Moreover, anti-CD23.CAR redirected T lymphocytes derived from both healthy donors (HD) and B-CLL patients displayed a specific lytic activity against CD23 + EBV-LCLs, even in presence of soluble CD23 enriched plasma without being inhibited (average lysis with no plasma 67%; average lysis with 25% of CD23 enriched plasma 79%; average lysis with 50% of CD23 enriched plasma 88%, n=3). We also demonstrated that the expression of the anti-CD23.CAR caused a significant increase in cytokine release from transduced in vitro activated T cells after a 48h stimulation with CD23 + targets. B-CLL derived CD23.CAR-expressing T cells (n=3) secreted 4-fold more INF-gamma, and 1445-fold more TNF-beta, compared to non transduced T cells. Interleukin-2 was also released (average release 2681 pg/mL, n=3) and sustained the antigen-dependent proliferation of CD23.CAR + T cells. To confirm in vivo the in vitro data , we tested NT and CD23.CAR transduced T cells in a recently published xenograft model of B-CLL (Ref biblio, primo nome, giornale, anno). This model is based on the intravenous or subcutaneous injection of the established human B-CLL cell line MEC1 into Rag2 −/− gamma c −/− mice, which lack not only B and T cells, but also natural killer (NK) cells, thus presenting a profound immunosuppressive environment leading to an high efficiency of both B-CLL and T-cell engraftment. Moreover, this model reproduces the systemic involvement of the disease and it closely resembles the aggressive form human B-CLL, representing an optimal experimental setting to test the efficacy of new therapeutic agents. In the first set of our experiments, 8 weeks-old male mice were subcutaneously injected with 10*10 6 MEC1 cells in the left flank. Then, mice bearing an established tumor were treated intravenously with a single dose of NT or engineered CD23.CAR T cells (2*10 6 ), without any addition of exogenous IL-2. Animals were monitored twice a week for weight and tumor growth (measuring three perpendicular diameters), and sacrificed when the mean tumor volume reached a dimension of 3 1000 mm 3 , before presenting clinical signs and symptoms. Compared with NT-treated mice, the infusion of CD23.CAR + T cells resulted in a significant delay in tumor growth, as measured by tumor volume diameter/day (CD23.CAR + T cells vs NT T cells: p=0.04 at day 12; n=3). In conclusion, our results suggest that CD23.CAR-redirected T cells provide cytotoxic activity against CD23 + B-CLL cells in vitro and in vivo , while sparing normal B lymphocytes, as compared to other available CARs targeting pan-B-cell antigens, such as CD19 and CD20. These results are encouraging and demonstrate the feasibility of generating CD23.CAR + T lymphocytes for adoptive T-cell therapy of patients with B-CLL. Disclosures: No relevant conflicts of interest to declare.