Test-driving CARs

Redirecting T-cell specificity through the introduction of a chimeric antigen receptor (CAR) is emerging as a clinically feasible approach for adoptive immunotherapy. In this issue of Blood, Till and colleagues now demonstrate that autologous T cells genetically modified to express a CD20-specific CAR can be safely infused in patients with B-lineage lymphomas.

Approximately 15 years after Zelig Eshhar demonstrated how to redirect the specificity of T cells,¹  investigators are reporting their early clinical experiences with infusing T cells genetically modified to express chimeric antigen receptors (CARs). By combining T-cell therapy with gene therapy in compliance with current good manufacturing practice (cGMP) for phase 1 and 2 trials, investigators have bypassed tolerance to enable clinical-grade T cells to recognize desired cell-surface antigens independent of MHC. T cells are rendered tumor-specific through introduction of a CAR, which is typically composed of the scFv from a monoclonal antibody (mAb) that forms part of the CAR ectodomain and which, upon binding antigen, activates T cells by phosphorylation of conserved immunoglobulin tyrosine activation motifs within a chimeric CD3-ζ or FcϵRI endodomain. To limit potential deleterious off-target effects, the first human trials for hematopoietic malignancies have infused genetically modified CAR⁺ T cells that target lineage-restricted antigens, such as CD19 and CD20 expressed on malignant (and healthy) B cells,^2,3  and generally have used first-generation CARs (activating T cells solely through CD3-ζ⁴ ). Building on clinical experiences of therapeutic mAbs targeting CD20, Till and colleagues demonstrate that an intrapatient dose-escalation study infusing autologous clinical-grade T cells expressing a CD20-specific CAR with or without low-dose IL-2 can be undertaken in patients with non-Hodgkin lymphomas. In the current financial and regulatory climate, publishing the results of a gene-therapy trial is a singular accomplishment. However, like many initial human experiences, the results raise more questions than they answer.

Sustaining the survival of adoptively transferred CAR⁺ T cells is one of the major impediments to achieving significant therapeutic responses. One way to enhance persistence is to infuse a heterogeneous population of T cells so that subpopulations can participate in lymphopenia-induced proliferation. This has been championed by Rosenberg and colleagues⁵  at the National Institutes of Health who demonstrated that clinical responses can be achieved when melanoma-specific T cells are infused after lymphodepleting chemotherapy and when bulk populations of T cells are given rather than T-cell clones. Thus, most infusions of CD20-specific T cells were given after chemotherapy and the trial was altered to infuse populations of genetically modified T cells rather than clones. Cumulatively, up to 4.4 × 10⁹/m² cells were infused within 10 days, and while a limited number of infused T cells could be detected in the peripheral blood of some patients for up to 3 months, the question remains as to why the infused T cells did not persist longer and at increased levels leading to loss of normal CD20⁺ B cells. Perhaps it was due to the cells entering replicative senescence after nonviral gene transfer and ex vivo expansion to clinically meaningful numbers. Perhaps it was due to insufficient T-cell help, despite the use of low-dose IL-2. Perhaps it was due to the level of CAR expression or competency of CAR-dependent signaling, which might have been insufficient to sustain a proliferative T-cell signal. Perhaps it was due to a lack of CAR⁺ central-memory T cells in the inoculum that were capable of long-term in vivo persistence.⁶  Or perhaps it was due to incomplete lymphodepletion resulting from the choice of preinfusion chemotherapies used.

The trial described by Till et al drives home the observation that multiple infusions of autologous CAR⁺ T cells targeting a B-lineage antigen are both safe and feasible. This clinical experience can now serve as a platform for future endeavors to answer questions concerning the improvement of persistence and the resolving of issues regarding homing to tumor deposits, thereby improving the therapeutic potential of their CAR⁺ T cells.

The clinical data in this issue of Blood are among the first reports on the potential of T cells manufactured under cGMP that have been genetically modified to redirect specificity.^7-9  With this publication, the authors have advanced the promising technology of CARs, which combine the specificity of mAbs with the replicative and homing potentials of T cells.