T cells expressing CD19 chimeric antigen receptors for acute lymphoblastic leukaemia in children and young adults: a phase 1 dose-escalation trial

Abstract

Background: Chimeric antigen receptor (CAR) modified T cells targeting CD19 have shown activity in case series of patients with acute and chronic lymphocytic leukaemia and B-cell lymphomas, but feasibility, toxicity, and response rates of consecutively enrolled patients treated with a consistent regimen and assessed on an intention-to-treat basis have not been reported. We aimed to define feasibility, toxicity, maximum tolerated dose, response rate, and biological correlates of response in children and young adults with refractory B-cell malignancies treated with CD19-CAR T cells.

Methods: This phase 1, dose-escalation trial consecutively enrolled children and young adults (aged 1-30 years) with relapsed or refractory acute lymphoblastic leukaemia or non-Hodgkin lymphoma. Autologous T cells were engineered via an 11-day manufacturing process to express a CD19-CAR incorporating an anti-CD19 single-chain variable fragment plus TCR zeta and CD28 signalling domains. All patients received fludarabine and cyclophosphamide before a single infusion of CD19-CAR T cells. Using a standard 3 + 3 design to establish the maximum tolerated dose, patients received either 1 × 10(6) CAR-transduced T cells per kg (dose 1), 3 × 10(6) CAR-transduced T cells per kg (dose 2), or the entire CAR T-cell product if sufficient numbers of cells to meet the assigned dose were not generated. After the dose-escalation phase, an expansion cohort was treated at the maximum tolerated dose. The trial is registered with ClinicalTrials.gov, number NCT01593696.

Findings: Between July 2, 2012, and June 20, 2014, 21 patients (including eight who had previously undergone allogeneic haematopoietic stem-cell transplantation) were enrolled and infused with CD19-CAR T cells. 19 received the prescribed dose of CD19-CAR T cells, whereas the assigned dose concentration could not be generated for two patients (90% feasible). All patients enrolled were assessed for response. The maximum tolerated dose was defined as 1 × 10(6) CD19-CAR T cells per kg. All toxicities were fully reversible, with the most severe being grade 4 cytokine release syndrome that occurred in three (14%) of 21 patients (95% CI 3·0-36·3). The most common non-haematological grade 3 adverse events were fever (nine [43%] of 21 patients), hypokalaemia (nine [43%] of 21 patients), fever and neutropenia (eight [38%] of 21 patients), and cytokine release syndrome (three [14%) of 21 patients).

Interpretation: CD19-CAR T cell therapy is feasible, safe, and mediates potent anti-leukaemic activity in children and young adults with chemotherapy-resistant B-precursor acute lymphoblastic leukaemia. All toxicities were reversible and prolonged B-cell aplasia did not occur.

Funding: National Institutes of Health Intramural funds and St Baldrick's Foundation.

Conflict of interest statement

Declaration of interests

We declare no competing interests.

Copyright © 2015 Elsevier Ltd. All rights reserved.

Figures

Figure 1:. Clinical activity and expansion of CD19-chimeric antigen receptor (CAR) T cells

(A) Waterfall plot of the percent change in bone marrow blast frequency from baseline to day 28, response, and cytokine release syndrome (CRS) grading in all 20 patients with B-precursor acute lymphoblastic leukaemia (B-ALL) treated. *One patient with progressive disease (PD) because of a greater than 50% increase in circulating blasts. (B) CAR T cells in the CSF of 17 patients with B-ALL who underwent lumbar puncture within 1 month of CAR infusion annotated according to CNS leukaemia status (table 1 and appendix) and neurotoxicity. Patients who developed neurotoxicity had significantly higher concentrations of CSF CAR T cells (p=0·0039). Three additional patients did not have samples sufficient for analysis. (C) B-cell depletion and rapid recovery in peripheral blood in responding patients (n=14). Patients were followed up until recovery of circulating B cells or day 28, whichever occurred later. Circles designate circulating B cells in responding patients (left axis). Grey bars designate number of patients with evidence for normal B-cell progenitors in the marrow (eg, haematogones) at the designated time point (right axis). (D) Kaplan-Meier plot showing 51·6% overall survival probability after 9·7 months for all patients enrolled (top, n=21, median follow-up 10 months) and 78·8% leukaemia-free survival beginning at 4·8 months in patients with B-ALL who had MRD-negative remission (bottom, n=12). Ten of these 12 patients had a subsequent HSCT and all remain leukaemia free. (E) Disappearance of CSF leukaemia in two patients coincident with CAR T-cell migration to the CSF. (F) Percent (top) and absolute number (middle) of circulating CD19-CAR T cells by flow cytometry (n=21) and qPCR (n=18; bottom). Peripheral blood was analysed in each patient until CAR T cells were no longer detected or day 28, whichever occurred later. Horizontal solid lines show the median at each time point, and dashed lines indicate the lower limit of detection. Circles designate responding patients. Sample days were designated days 3 (range 1–5), 7 (5–9), 14 (12–16), 28 (25–31), 42 (45–49), and 68 (55–81). (G) Absolute number of circulating blasts in peripheral blood of all patients with B-ALL. Horizontal solid lines show the median at each time point, and dashed lines indicate the lower limit of detection. Circles designate responding patients and triangles designate patients who did not respond. Sample days were designated days 3 (range 1–5), 7 (5–9), 14 (12–16), and 28 (24–32). Patient 1 is the responding patient with blasts at day 28 by flow cytometry. (H) Time course of peripheral blood flow cytometry from a representative patient (patient 14) shows circulating leukaemia and low concentrations of non-malignant B cells at day −1, followed by CAR T-cell expansion coincident with clearance of leukaemia and non-malignant B cells by day 10, followed by disappearance of CD19-CAR T cells and B-cell recovery by day 53. Blue dots show CD19-CAR T cells; red dots show leukaemia blasts (CD19+CD34+); green dots represent normal B cells. CR=complete response. MRD=minimum residual disease. SD=stable disease.

Figure 2:. Circulating chimeric antigen receptor (CAR) T-cell numbers are biomarkers of response and cytokine release syndrome (CRS) severity

Responding patients had significantly higher peak circulating CAR T cells measured by (A) flow cytometry or (B) qPCR than non-responders. (C) Peak circulating CART cells correlated with CRS severity. Triangles show non-responding patients. (D) Higher absolute numbers of circulating CD8+ CAR+ T cells (p=0·0087), CD8+ effector memory CAR+ T cells (p=0·0087), and CD4+ effector memory CAR+ T cells (p=0·026) were seen in patients with grade 3 or 4 CRS than those without CRS or grade 1 or 2 CRS. Assessable circulating CAR T cells were not seen in non-responding patients except for one as shown by the triangle. Horizontal dashed lines indicate the lower limit of detection. (E) Responding patients with ALL with higher disease burden were significantly more likely to have grade 3 or 4 CRS than patients with lower disease burdens (p=0·039).

Figure 3:. Interleukin 6 (IL-6), interferon γ (IFNγ), and C-reactive protein (CRP) concentrations are biomarkers of cytokine release syndrome severity

(A) Timecourse of rises in circulating inflammatory cytokines and C-reactive protein in one representative patient with grade 3 cytokine release syndrome (CRS) who did not receive tocilizumab (left) and another patient with grade 4 CRS who was treated with tocilizumab plus hydrocortisone (right). Normal C-reactive protein is r=0·81, 95% CI 0·54–0·92; p<0·0001). Data are available only for patients 1–19. (E) Higher peak C-reactive protein concentrations are associated with increased CRS severity. CAR=chimeric antigen receptor. TNF=tumour necrosis factor. GM-CSF=granulocyte-macrophage colony-stimulating factor.