Selecting CD7⁻ T cells for CAR T-cell therapy

In this issue of Blood, Freiwan et al demonstrate the feasibility of using selected, naturally occurring CD7⁻ T cells to generate CD7-targeting chimeric antigen receptor (CAR) T cells without CD7-directed fratricide and show that CAR^CD7− T cells have favorable biological characteristics.¹ The potential of exploiting CD7⁻ T cells can facilitate the manufacturing of CD7 CAR T cells for T-cell acute lymphoblastic leukemia (T-ALL). In addition, the authors report that beyond T-ALL targeting, CD19-specific CAR^CD7− T cells show improved immunotherapeutic properties leading to better antitumor function, and thus, CD7⁻ T cells are an interesting effector population for CAR T-cell therapy of hematological malignancies.

The overlapping antigen expression between healthy T cells and cancer cells limits the development of CAR T-cell therapy for T-cell–derived malignancies. CD7 is one of the most attractive target antigens since it is highly expressed in T-ALL blasts and T-cell lymphomas. However, its expression on normal T cells results in fratricide of CD7 CAR T cells, which compromises the successful and efficient product manufacturing. Several strategies have been proposed to mitigate the self-elimination of CD7 CAR T cells. These mainly include the disruption of the CD7 gene using gene editing methods (CRISPR-Cas9 or base editing)^2-4 or the cytoplasmic sequestration of CD7 protein using a protein expression blocker.⁵ Both these approaches rely on additional genetic modifications, which further complicates the manufacturing process and increases its cost.

Freiwan et al attempted to avoid CD7 CAR T-cell fratricide by suggesting an approach that does not involve extra genetic modifications but relies on harnessing naturally occurring CD7⁻ T cells for the generation of CD7 CAR^CD7− T cells. Indeed, they found that CD7⁻ T cells were present in the peripheral blood of healthy donors (0.72% to 19.5%) and patients with T-ALL and B-cell acute lymphoblastic leukemia (B-ALL) (3% to 12.5%). Despite their low frequency, the authors described the feasibility of efficient selection of the naturally occurring CD7⁻ T cells by a 2-step bead separation process and further transduction with a CD7 CAR to generate CD7 CAR^CD7− T cells. The production and expansion of CD7 CAR^CD7− T cells were possible without fratricide in contrast to bulk CD7 CAR T cells.

The CD7 CAR^CD7− T cells elicited specific antitumor activity in in vitro and in vivo T-ALL models. Interestingly, the authors observed that the CD7 CAR^CD7− T cells showed phenotypic and functional properties that have been associated with optimal immunotherapeutic function. Specifically, they had an enriched CD4⁺ effector memory phenotype; they maintained their cytotoxic activity, along with cytokine secretion, for several serial stimulations and showed no increase in the expression of checkpoint inhibitory receptors. Also, CD7 CAR^CD7− T cells persisted for long in vivo and protected mice from tumor rechallenge.

Based on these findings, Freiwan et al further investigated whether CAR^CD7− T cells possess favorable immunotherapeutic properties compared with bulk CAR T cells, and thus, naturally occurring CD7⁻ T cells could confer a preferable population for CAR T-cell therapy in general. They produced CD19 CAR^CD7− T cells and compared them with bulk CD19 CAR T cells. Although CD19 CAR^CD7− T cells elicited a slightly lower tumor lytic capacity at low effector-to-target ratios, cytotoxicity and cytokine secretion were again sustained for more repeated stimulations compared with bulk CD19 CAR T cells. In a B-ALL in vivo xenograft model, CD19 CAR^CD7− T cells outperformed conventional CD19 CAR T cells in controlling tumor growth and showed increased expansion. Notably, neither the enrichment in CD4⁺ cells nor a possible predisposition for clonal expansion of CD7⁻ T cells could explain the functional superiority of CD19 CAR^CD7− T cells. Transcriptional comparison of CD7⁻CD4⁺ vs CD7⁺CD4⁺ CD19 CAR T cells revealed differences in several activation pathways, which are prominent in T-cell biology (such as interleukin-2/STAT5, MTORC1, PI3K/AKT, etc). Freiwan et al suggested the hypothesis that CD7 expression on the surface influences the activation dynamics of CAR T cells. Further analysis of single-cell RNA sequencing data of clinical CD19 CAR T-cell products confirmed that, indeed, the CD7⁻ population corresponded to functional effector CD4⁺ cells or dysfunctional cells. In addition, the analysis supported the favorable immunotherapeutic potential of CAR^CD7− T cells, showing the low concentration of CD7 expression on CD4⁺ effector cells and the percentage of CD7⁻CD4⁺ cells correlated with clinical response to CD19 CAR T therapy.

The study from Freiwan et al contributes toward developing effective CAR T-cell therapy for patients with T-cell lineage malignancies who have otherwise limited treatment options. Exploiting the naturally occurring CD7⁻ T cells to avoid fratricide would have an impact not only on the success of manufacturing of CD7 CAR T cells but also on the cost-effectiveness of the therapy since no additional genetic engineering is involved. With a similar aim, recently, Lu et al⁶ described a clinical study with another method of naturally overcoming CD7-directed fratricide by using a CD7 CAR that causes CD7 epitope masking or intracellular sequestration. In this study, Freiwan et al used healthy donor CD7⁻ T cells to produce CAR^CD7− T cells. Although the authors envision future clinical translation by including a donor-derived approach in the posttransplant setting, the proof of the feasibility of the method using patient-derived CD7⁻ T cells is still important. In addition, given the low frequency of CD7⁻ T cells in peripheral blood and the impact of long-term ex vivo expansion on the functional quality of CAR T cells, additional studies are warranted to confirm the viability of the method in clinical-scale manufacturing.

Most interestingly, the study of Freiwan et al introduced a previously not acknowledged role of CD7⁻ T cells in CAR T-cell therapy beyond the limits of T-cell malignancies. CD7 is involved in T-cell activation, and CD7⁻ T cells have been investigated in several pathological situations.⁷ In this study, CAR^CD7− T cells had superior antitumor function and persistence than conventional CAR T cells and had a distinct transcriptional activation profile. Further investigation is certainly warranted to assess in more depth the distinct molecular background defining the improved function of CD7⁻ T cells. In addition, more detailed studies involving CARs with various costimulatory designs will contribute to evaluating whether CD7⁻ T cells may have a biologically significant role in the general context of CAR T-cell therapy for hematological malignancies.

Conflict-of-interest disclosure: M.T. is an inventor of patents and patent applications related to CAR T-cell therapy.