Feasibility and preclinical efficacy of CD7-unedited CD7 CAR T cells for T cell malignancies

Abstract

Chimeric antigen receptor (CAR)-mediated targeting of T lineage antigens for the therapy of blood malignancies is frequently complicated by self-targeting of CAR T cells or their excessive differentiation driven by constant CAR signaling. Expression of CARs targeting CD7, a pan-T cell antigen highly expressed in T cell malignancies and some myeloid leukemias, produces robust fratricide and often requires additional mitigation strategies, such as CD7 gene editing. In this study, we show fratricide of CD7 CAR T cells can be fully prevented using ibrutinib and dasatinib, the pharmacologic inhibitors of key CAR/CD3ζ signaling kinases. Supplementation with ibrutinib and dasatinib rescued the ex vivo expansion of unedited CD7 CAR T cells and allowed regaining full CAR-mediated cytotoxicity in vitro and in vivo on withdrawal of the inhibitors. The unedited CD7 CAR T cells persisted long term and mediated sustained anti-leukemic activity in two mouse xenograft models of human T cell acute lymphoblastic leukemia (T-ALL) by self-selecting for CD7^-, fratricide-resistant CD7 CAR T cells that were transcriptionally similar to control CD7-edited CD7 CAR T cells. Finally, we showed feasibility of cGMP manufacturing of unedited autologous CD7 CAR T cells for patients with CD7⁺ malignancies and initiated a phase I clinical trial (ClinicalTrials.gov: NCT03690011) using this approach. These results indicate pharmacologic inhibition of CAR signaling enables generating functional CD7 CAR T cells without additional engineering.

Keywords: CD7; T cell lymphoma; T cell malignancies; T-ALL; adoptive cell therapy; chimeric antigen receptor; engineered T cells; fratricide.

Graphical abstract

Figure 1

Dasatinib and ibrutinib prevent CD7 CAR T cell fratricide, and the inhibitory effect is reversible (A) A schematic diagram showing the effect of dasatinib and ibrutinib on CAR signaling. (B) An outline of CAR T cell generation. (C) Representative flow plots (left) and summary of percent of CAR⁺ and CD7⁺ cells (right) on day 7 post-transduction for each specified T cell type (mean ± SD, Ctrl CD7 CAR T, n = 4; other conditions, n = 10). (D) Top: ex vivo fold expansion of specified T cell types over time (mean ± SD, n = 8). Bottom: cell viability on day 7 post-transduction determined by flow cytometry (mean ± SD, Ctrl CD7 CAR T, n = 4; other conditions, n = 10). (E) Left: cytotoxicity of specified effector T cells against Jurkat (mean ± SD, n = 10) or CCRF-CEM (mean ± SD, n = 6) target cells at 72 h after coculture setup at a 1:4 effector-to-target ratio. Right: expansion of specified effector T cells when cocultured with Jurkat (left, mean ± SD, n = 10) or CCRF-CEM (right, mean ± SD, n = 6) target cells for 72 h at a 1:4 effector-to-target ratio. (F) Similar coculture setup with (E) at a 1:10 effector-to-target ratio (mean ± SD, n = 4). Statistical differences are calculated by one-way ANOVA with Tukey’s multiple comparisons (C–F). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. n.s, non-significant.

Figure 2

PI CAR T cells showed superior anti-tumor activity and long-term persistence in vivo (A) Schematic of model setup for (B)–(D). (B) Representative IVIS images showing tumor bioluminescence. (C) Change of tumor bioluminescence over time in mice receiving the specified T cell treatment, as measured by IVIS imaging. Each line represents data from one individual animal. (D) Animal survival over time. (E) Schematic of model setup for (F) and (G). (F) Representative IVIS images showing bioluminescence from infused T cells. (G) Change of T cell bioluminescence over time in mice receiving the specified T cell treatment, as measured by IVIS imaging. Each line represents data from one individual animal. (H) Schematic of model setup for (I)–(L). (I) Representative IVIS images showing bioluminescence from infused T cells. (J) Change of T cell bioluminescence over time in mice receiving the specified T cell treatment, as measured by IVIS imaging. Each line represents data from one individual animal. (K) Absolute counts of infused CCRF-CEM tumor cells (left) and T cells (right) in 50 μL peripheral blood on day 15 after T cell infusion (mean ± SD, n = 5 for NT, n = 6 for CD7 KO, n = 6 for PI). (L) Animal survival over time. Statistical differences are calculated by log rank test (D and L) or one-way ANOVA with Tukey’s multiple comparisons (K). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Figure 3

Persisting PI CAR T cells lack CD7 expression and transcriptionally resemble CD7-edited CAR T cells (A) Representative flow plots (left) and summary of percent of CAR⁺CD7⁻ cells (right) pre-infusion and at 27 days post-infusion (mean ± SD, n = 3 for pre-infusion, n = 13 for post-infusion). (B) CD7 protein expression in persisting control (Ctrl) non-transduced (NT) T cells and PI CAR T cells. (C) Quantitative PCR results showing relative CD7 mRNA level of post-infusion PI CAR T cells compared with pre-infusion (mean ± SD, n = 3 for pre-infusion, n = 6 for post-infusion). (D) Percent of CD4⁺ and CD8⁺ cells in the starting cell material (PBMCs), pre-infusion products (pre), and persisting CD7⁻ PI CAR T cells post-infusion (post) (mean ± SD, n = 3 for PBMCs, n = 3 for pre-infusion, n = 13 for post-infusion). (E) Heatmap was plotted using normalized gene expression from each sample. Gene expression was normalized with Trimmed Means of M values (TMM) and log2 transformed counts per million (log2(CPM)). Results of unsupervised clustering are shown. (F) Scatterplot showing significantly high transcriptome profiling correlation between CD7-unedited and -edited CD7 CAR T cells. Mean normalized gene expression was calculated by averaging normalized gene expression (same method as described in E) from three biological replicates in each condition. p value and coefficient were calculated with linear regression. Highlighted are genes involved in regulating the immune function of T cells. Statistical differences are calculated by unpaired two-tailed t test (A and C). ∗∗∗∗p < 0.0001. See Figures 2E, S6A, and S7A for model setup.

Figure 4

Characterization of cGMP-manufactured autologous PI CAR T cells for T-ALL patients (A) Frequency of CD7⁻ normal CD4⁺ and CD8⁺ T cells measured by flow cytometry in PBMCs collected from patients with T cell malignancies. (B) Left: absolute T cell counts for each patient at the time of transduction (transduced) and on the day of cryopreservation (cryopreserved, day 4 post-transduction). Right: fold expansion of PI CAR T cells between transduction and cryopreservation. (C) Viability of PI CAR T cells at the time of cryopreservation. (D) Percent of CAR⁺ T cells at the time of cryopreservation. (E) Vector copy number per transduced T cell at the time of cryopreservation. (F) Cytotoxicity of PI CAR T cells and donor-matched NT T cells on coculture with FFluc-labeled Jurkat T-ALL cell line for 24 h. In all panels, each dot represents data from an individual patient. Means ± SD are shown.