The uncut version: base-edited allo-CAR T cells

In this issue of Blood, Diorio et al¹ harness base editing technology to develop a potent and complex gene-edited CD7-specific chimeric antigen receptor (CAR) T-cell product for off-the-shelf use in patients with T-cell leukemia and other CD7⁺ malignancies. Their manufacturing platform showcases the potential of base editing for future progress toward safer and more accessible CAR T-cell therapies.

After the success of personalized CAR T-cell therapy in B-cell malignancies, the field expanded to develop CAR technology for other malignant diseases. With some CAR T-cell products moving toward use earlier in treatment, there is an increasing interest in overcoming the logistic hurdles involved in personalized cell manufacturing by establishing allogeneic off-the-shelf solutions.² Herein, gene silencing with programmable nucleases (eg, via Zinc finger nucleases, TALEN, CRISPR-Cas9) has become an essential tool to facilitate engineering of T cells with the desired attributes.

Conventional programmable nucleases allow gene silencing by forcing DNA double-stranded breaks (DSBs) at coding regions of the targeted gene. Repeated cuts and error prone DNA repairs promote small insertions and deletions that induce frameshift mutations and disrupt protein expression. Despite achieving highly efficient gene knockouts, repetitive cutting by nucleases can induce genetic rearrangements such as inversions, larger deletions, and even complete loss of chromosomes.³ To overcome the challenges of safe allogeneic CAR (allo-CAR) T-cell therapy, multiple genetic modifications will be required to eliminate the risk of graft-versus-host disease and allogeneic cell immunogenicity, which limit CAR T-cell persistence and antitumor efficacy.⁴ However, simultaneous targeting of multiple genes with nuclease-assisted gene disruption can create myriad translocations and genetic rearrangements with unknown long-term consequences.

Base editing is a technology that couples the programmable nature of the CRISPR-Cas9 system with avoidance of DNA DSBs during genetic modification. By fusing a single-strand DNA (ssDNA) deaminase enzyme to a catalytically inactive Cas9 variant, only an ssDNA cut (nick) is induced. The Cas9-mediated nicking of the genomic DNA exposes a short stretch of ssDNA to the attached deaminase which then converts the selected bases within their target window. Since the original report on cytosine base editors (CBEs) in 2016, which force targeted cytosine-to-thymine base conversions,⁵ iterative improvements have yielded novel versions of base editors that reduce unwanted byproducts, improve the targeting scope, and allow editing of different bases.⁶ Previous work has demonstrated that CBEs can be used to induce stop codons or disrupt splice sites to thwart effective protein expression in T cells.⁷ 

Diorio et al used a messenger RNA–encoded fourth-generation CBE to efficiently silence 4 different genes in more than 90% of the cells after a single electroporation. Building on previous work published in Blood,⁸ they chose to disrupt CD7 expression to avoid fratricide (self-killing) after subsequent transduction with a CD7-specific CAR-encoding lentivirus. TRAC gene knockout was performed to eliminate potential allo-reactive T-cell receptors and avoid graft-versus-host disease. Additional silencing of the endogenous immune checkpoint programmed cell death protein 1 (PD1; encoded by PDCD1) was introduced to improve antitumor performance in immunosuppressive environments. CD52 deletion should enable conditioning for patients with an anti-CD52 monoclonal antibody to enhance initial CAR T-cell engraftment and synergistically target CD52⁺ malignant blasts. In line with previous studies, multigene editing with CBEs did not induce detectable translocations in karyotyping and targeted sequencing assays in the modified T cells.^7,9

Protocols to generate off-the-shelf CAR T-cell products aim to optimize CAR T-cell yields to result in many doses per production cycle. Diorio et al demonstrate that base editing enabled higher cell viability, better proliferation, and optimized CAR T-cell yields when compared with the conventional CRISPR-Cas9 nuclease system. This may be explained by reduced DNA damage response observed after quadruple editing with CBEs and better cellular fitness in the absence of gross genetic rearrangements. With overall yields between 16 × 10⁹and 17.2 × 10⁹ CAR⁺ T cells, a single manufacturing run may provide up to 200 administrations in adult patients (assuming a treatment dose of 1 × 10⁶ CAR⁺ T cells per kilogram of body weight in a cohort of patients averaging 80 kilograms each). Therefore, we can assume significantly lower cost per dose in contrast to personalized ex vivo manufacturing.

A high degree of reproducibility during allogeneic cell manufacturing is paramount to deliver consistent clinical outcomes. To date, only a few studies with gene-edited allo-CAR T-cell products have investigated donor-to-donor variability regarding product potency. In the article by Diorio et al, 3 independent clinical-scale manufacturing runs displayed low inter-donor variability regarding base editing outcome, cell yields, and even in vivo performance. Donor-dependent differences were observed during in vitro cytokine release assays and CD4:CD8 ratios in the final products. Future studies may identify donor- or product-dependent factors that are associated with better outcome.

Donor-derived CD7-specific CAR T cells have already been reported to induce high rates of complete remission in patients with leukemia.¹⁰ Extensive in vivo modeling using human T-cell leukemia cell lines and patient-derived xenografts in immunodeficient mice attest to the high antitumor efficacy of the base-edited CD7-specific CAR T-cell products. Notably, Diorio et al did not include modifications that alter HLA expression to improve the persistence of the CAR T cells. Instead, they chose a formulation of edits to create a cellular product for potent antitumor attack as an effective means of bridging to curative allogeneic stem cell transplantation. Pending favorable risk-benefit analysis after finalization of a molecular off-target analysis, this new off-the-shelf CD7 CAR T-cell product will complement the clinical trial pipeline, and it may ultimately expand the treatment options for patients who have treatment refractory T-cell malignancies.