Safety and persistence of adoptively transferred autologous CD19-targeted T cells in patients with relapsed or chemotherapy refractory B-cell leukemias

Abstract

We report the findings from the first 10 patients with chemotherapy-refractory chronic lymphocytic leukemia (CLL) or relapsed B-cell acute lymphoblastic leukemia (ALL) we have enrolled for treatment with autologous T cells modified to express 19-28z, a second-generation chimeric antigen (Ag) receptor specific to the B-cell lineage Ag CD19. Eight of the 9 treated patients tolerated 19-28z(+) T-cell infusions well. Three of 4 evaluable patients with bulky CLL who received prior conditioning with cyclophosphamide exhibited either a significant reduction or a mixed response in lymphadenopathy without concomitant development of B-cell aplasia. In contrast, one patient with relapsed ALL who was treated in remission with a similar T-cell dose developed a predicted B-cell aplasia. The short-term persistence of infused T cells was enhanced by prior cyclophosphamide administration and inversely proportional to the peripheral blood tumor burden. Further analyses showed rapid trafficking of modified T cells to tumor and retained ex vivo cytotoxic potential of CD19-targeted T cells retrieved 8 days after infusion. We conclude that this adoptive T-cell approach is promising and more likely to show clinical benefit in the setting of prior conditioning chemotherapy and low tumor burden or minimal residual disease. These studies are registered at www.clinicaltrials.org as #NCT00466531 (CLL protocol) and #NCT01044069 (B-ALL protocol).

Figures

Figure 1

Cytokine analysis of EOP 19-28z–transduced T cells stimulated on CD19+ AAPCs. EOP 19-28z–transduced T cells were plated on 3T3 fibroblast AAPCs expressing CD19. After 48 hours in culture, cell supernatants were assayed for cytokine levels. The cytokine levels were normalized to transduction efficiency. In addition, background cytokine levels measured on 3T3 fibroblasts that did not express CD19 were subtracted.

Figure 2

Patient serum cytokine levels before and after cyclophosphamide and T-cell infusions. Serum cytokines were determined before and after 19-28z+ T-cell infusion. The first set of patients (CLL-1 to CLL-3) did not receive cyclophosphamide treatment before T-cell infusion, whereas the second set (CLL-5 to CLL-7 and ALL-1) received cyclophosphamide before T-cell infusion. The x-axis represents treatment time course with day 0 being the day of T-cell infusion and day −2 being the day of cyclophosphamide treatment.

Figure 3

Marked reduction of peripheral lymphadenopathy was observed after treatment with autologous 19-28z+ T cells. Representative computed tomography scan images of patient CLL-5 before treatment and from 4 weeks and 14 weeks after treatment show mild increase in axillary (top) and pelvic lymphadenopathy (bottom) at 4 weeks but regression of lymphadenopathy at 14 weeks after therapy with modified T cells.

Figure 4

B-cell aplasia in the peripheral blood and BM of patient ALL-1 after infusion with autologous 19-28z+ T cells. Samples before and after infusion of blood (A) and BM (B) were obtained at the time points indicated on the x-axis and used for cell counts and immunophenotyping. Cell concentrations were measured with an AcT diff cell counter (Coulter). Immunophenotyping was performed by gating on CD45 and defining cell populations with the following cell surface markers: CD3+ (T cells), CD19+CD10− (mature B cells), CD19+CD10+ (progenitor B cells or tumor cells), CD19−CD10+ (granulocytes). (A) The plotted value for each cell population was derived as the product of the cell concentration by the frequency. (B) The plotted value for each cell population represents the percentage within the CD45 gate.

Figure 5

Immunohistochemistry staining of tissues with 19E3 anti-CAR Ab. (A) Trafficking of targeted T cells to tumor sites in CLL-4 44 hours after T-cell infusion. Images were acquired at room temperature using Nikon 90i Eclipse microscope system (Nikon) with a Nikon Plan Apo VC 20×/0.75 objective lens (magnification 200×), Nikon DigiSight Digital Camera Head and Nikon NSI-Elements Version 3.10 software. (B) Trafficking and homing of targeted T cells to BM. IHC staining on BM biopsies shows the presence of 19-28z+ T cells 5 and 6 weeks after infusion in patients ALL-1 and CLL-7, respectively (C) Trafficking and homing of targeted T cells to BM ≤ 8 weeks after infusion. ND indicates not done; (−) no positive 19-28z+ T cells found; (+) positive 19-28z+ T cells found occasionally, and (++) positive 19-28z+ T cells readily identifiable.

Figure 6

In vivo persistence of modified T cells. (A) Persistence of 19-28z+ T cells after infusion in the peripheral blood. CD3+ T cells derived from the peripheral blood samples of patients with CLL and ALL at various time points after infusion were activated with Dynabeads. Quantitative PCR analysis was performed, and the average vector copy number (vcn) was determined per 100 cells. The average vcn was below the limit of detection for CLL-2 and CLL-3 patient samples at all available data points and, therefore, are not represented. Samples for ALL-1 were not available on day 0 (at 1 hour after infusion). (B) Expression of 19-28z+ CAR after infusion after restimulation with either AAPCs or Dynabeads. CD3+ T cells derived from ALL-1 peripheral blood 2 days (d2) and 8 days (d8) after infusion were selected and activated with either Dynabeads (top) or AAPCs (bottom). CD3+ T cells derived from CLL-8 peripheral blood 5 weeks (wk 5) after infusion were selected and activated with Dynabeads. Flow cytometric analysis was performed with anti-CD3 and anti-CAR Abs 7 days after restimulation. (C) Inverse correlation between peripheral CD19+ cell counts and maximum vcn detected in the peripheral blood. The peripheral blood CD19+ cell count (obtained by CD19 FACS analysis) is plotted on the x-axis (Table 1). The maximum vector copy number per 100 cells and normalized per 108 infused 19-28z+ T cells detected in the peripheral blood after infusion (day 1-8), as shown in panel A, are reported on the y-axis (supplemental Table 2).