Phase 1 studies of central memory-derived CD19 CAR T-cell therapy following autologous HSCT in patients with B-cell NHL

Abstract

Myeloablative autologous hematopoietic stem cell transplantation (HSCT) is a mainstay of therapy for relapsed intermediate-grade B-cell non-Hodgkin lymphoma (NHL); however, relapse rates are high. In phase 1 studies designed to improve long-term remission rates, we administered adoptive T-cell immunotherapy after HSCT, using ex vivo-expanded autologous central memory-enriched T cells (TCM) transduced with lentivirus expressing CD19-specific chimeric antigen receptors (CARs). We present results from 2 safety/feasibility studies, NHL1 and NHL2, investigating different T-cell populations and CAR constructs. Engineered TCM-derived CD19 CAR T cells were infused 2 days after HSCT at doses of 25 to 200 × 10(6) in a single infusion. In NHL1, 8 patients safely received T-cell products engineered from enriched CD8(+) TCM subsets, expressing a first-generation CD19 CAR containing only the CD3ζ endodomain (CD19R:ζ). Four of 8 patients (50%; 95% confidence interval [CI]: 16-84%) were progression free at both 1 and 2 years. In NHL2, 8 patients safely received T-cell products engineered from enriched CD4(+) and CD8(+) TCM subsets and expressing a second-generation CD19 CAR containing the CD28 and CD3ζ endodomains (CD19R:28ζ). Six of 8 patients (75%; 95% CI: 35-97%) were progression free at 1 year. The CD4(+)/CD8(+) TCM-derived CD19 CAR T cells (NHL2) exhibited improvement in expansion; however, persistence was ≤28 days, similar to that seen by others using CD28 CARs. Neither cytokine release syndrome nor delayed hematopoietic engraftment was observed in either trial. These data demonstrate the safety and feasibility of CD19 CAR TCM therapy after HSCT. Trials were registered at www.clinicaltrials.gov as #NCT01318317 and #NCT01815749.

Figure 1

Constructs and treatment schema for clinical trials. (A) The CD19R:ζ DNA sequence (optimized by GeneArt) that contains the CAR sequence consisting of the V_H and V_L gene segments (scFv) of the CD19-specific FMC63 monoclonal antibody (mAb), IgG4 hinge-C_H2-C_H3, and signaling domains CD3ζ (CD19CAR) was cloned into a self-inactivating lentiviral vector pHIV7 that has an EF-1 promoter. (B) In addition to CD19 scFv, the CD19R:CD28:ζ/EGFRt⁺ epHIV7 lentiviral construct also contains (1) the cytoplasmic domain of the CD3ζ chain; (2) the CD28 costimulatory domain, (3) the self-cleaving T2A sequence; and (4) the truncated EGFR sequence as indicated. The huEGFRt was synthesized by PCR splice overlap extension to fuse in frame the human granulocyte-macrophage–colony-stimulating factor receptor's leader peptide to domains III and IV and the transmembrane spanning components of huEGFR (base pairs, 1000-2004). This fusion product was then cloned into the epHIV7 vector (in which the cytomegalovirus promoter of pHIV7 was replaced with an EF-1 promoter) along with the CD19CAR and T2A sequences, and the final construct was confirmed by sequence analysis. (C) Leukapheresis for T-cell manufacturing may be drawn before or after cycles of salvage chemotherapy. Research participants received T-cell infusions on either day +2 or +3 after HSCT. *Disease assessments continue every 6 months until 2 years after HSCT (6, 12, 18, and 24 months). PET, positron emission tomography.

Figure 2

Surface phenotype of T_CM-enriched CD19R⁺ cells. Freshly thawed T-cell products of (A) patient NHL1-4 and (B) patient NHL 2-2 were examined for the expression of CD3, CD4, CD8, CD62L, CD28, CD27, CD127, and PD-1 by flow cytometry. Percentages of positive cells (filled histograms) are calculated using the subtraction method on the basis of the gating of isotype-stained cells (open histograms), except CD4 and CD8 using indicated gating strategy. (C) Percentages of positive cells from all the patients are depicted. NHL1 patients are represented by filled circles and NHL2 patients by open circles. Means are presented for each surface phenotype.

Figure 3

CD19 CAR T–cell persistence by flow cytometry. PBMCs from patient blood were analyzed by flow cytometry at different time points after adoptive transfer. (A) Live cells were identified by staining with 7-amino actinomycin D (7-AAD) and then gated on CD3 (either fluorescein isothiocyanate [FITC] or allophycocyanin [APC] labeled), with dump antibodies anti-CD14 (either phycoerythrin [PE] or FITC labeled), and anti-CD16 (either PE or FITC labeled). (B) Anti-CD19scFv CAR Alexa Fluor 647 antibody was used for detection of CD19-specific CAR⁺ T cells. Percentages of the highest CAR-positive cells at the indicated time points in all NHL1 patients are depicted. (C) Erbitux (cetuximab)-biotin/streptavidin-PE was used for detection of CD19 specific CAR⁺ T cells. Percentages of the highest CAR-positive cells at the indicated time points in all NHL2 patients are depicted. (D) Time course for a representative patient sample from NHL1 shows CAR T cells at baseline and various points after adoptive transfer. (E) Time course for a representative patient sample from NHL2 shows CAR T cells at baseline and various points after adoptive transfer.

Figure 4

Comparison of CAR T–cell persistence between NHL1 and NHL2 T-cell therapy. gDNA was extracted from frozen aliquots of whole blood and tested for the WPRE copy number by TaqMan qPCR. Average copy numbers are presented if ≥2 of 3 replicates generated a cycle threshold (Ct) value. In the time points where only 1 of 3 was detectible, actual value (not average) was plotted. Starting on day 1, participants were measured for WPRE every 7 days (±3 days) during the first 28 days on study and then monthly thereafter. (A) WPRE copy numbers from 1 mL blood collected from day 1 up to day 158 after T-cell infusion in the NHL1 trial, except patient NHL1-3 (who only had PBMC sample), are plotted (whole blood, N = 7). After day 28, there was no detectable WPRE for any patient. (B) WPRE copy numbers from 1 mL blood collected from day 1 up to day 157 after T-cell infusion in the NHL2 trial are plotted (whole blood, N = 8). After day 27, there was no detectable WPRE for any patient. (C) WPRE copy numbers from blood collected from day 1 through day 28 after T-cell infusion in the NHL1 trial are plotted as a function of log10 copies/µg of gDNA (whole blood, N = 7; PBMCs, N = 1; UPN043). Gray area at bottom of graph is below the lower limit of quantification for the assay. (D) WPRE copy numbers from blood collected from day 1 through day 28 after T-cell infusion in the NHL2 trial are plotted as a function of log10 copies/µg of gDNA (whole blood, N = 8). Gray area at bottom of graph is below the lower limit of quantification for the assay. (E) The figure provides box and whisker plots of CAR copy number over time (AUC) by trial and dose level, with the individual patient AUCs shown in black. Box and whisker plots graphically present the median (heavy gray line), the mean (dashed line), low and upper quartiles (ends of the box), and minimum and maximum values. Outlier values are in red. The AUCs were calculated on the log₁₀ WPRE values from day 1 through day 25 after T-cell infusion (see D-E). If a 25-day WPRE was not available, one was interpolated. Measurements that were considered below the limit of quantification were set to 4 copies/μg of gDNA. WPRE data were available from 8 participants (whole blood, N = 7; PBMCs, N = 1) from NHL1 and all 8 participants from NHL2 (whole blood, N = 8). Each participant had 5 to 8 measurements. The mean AUC for NHL2 was 40.2 copies × days/μg and the mean AUC for NHL1 was 25.4 copies × days/μg, giving a mean difference of 14.8 (95% CI: 7.4-22.3), with a P value for the Welch 2-sample 2-tail Student t test of <.001. (F) Mean peak expansion is the peak WPRE value measured in log10 CAR copies/µg of gDNA. Mean peak expansion for NHL1 was 1.6 copies/μg and for NHL2 was 2.79 copies/μg, giving a mean difference of 1.19 (95% CI: 0.54-1.83), yielding a P value of .002 using a t test. (G) Maximum persistence is the day associated with the last value above the lower limit of detection followed by 2 measurement that were below the limit of detection. Mean maximum persistence for NHL1 was 18.25 days and for NHL2 was 20.5 days, giving a mean difference of 2.25 (95% CI: −8.56 to 13.1), yielding a P value of >.5 using a t test.

Figure 5

Cytokine profiles after HSCT and CD19 CAR T–cell infusion. Patient serum samples were analyzed with the Luminex IS100 bead array technology, and kits were purchased from Life Technologies (Invitrogen). The assays were performed by Clinical Immunobiology Correlative Studies laboratory (CICSL) at COH using commercially available 30-plex cytokine detection assays. Each data point represents the average of triplicate measurement. (A) Interferon-γ is plotted in pg/mL at weekly time points through day 28. (B) IL-6 is plotted in pg/mL at weekly time points through day 28.

Figure 6

B-cell reconstitution following HSCT and CD19 CAR T–cell infusion. PBMCs from patient blood were analyzed by flow cytometry at different time points after adoptive transfer. (A) Samples were analyzed by multiflow cytometry after staining with 7-AAD, CD3 VioBlue, CD19 PE, CD20 FITC, and CD10 APC for the analysis of B-cell reconstitution. Absolute CD19⁺ cells were calculated based on the percentage of CD19⁺ cells, white blood cell counts, and percentage of lymphocytes. (B) Rituximab levels in the serum of the patients were determined with a time-resolved fluorometry assay.