In this issue of Blood, Kochenderfer et al show that engineered CD19–chimeric antigen receptor (CAR) and donor-derived allogeneic T cells can safely treat CD19-positive B-cell malignancies, which have relapsed after allogeneic stem cell transplantation.1
Although allogeneic stem cell transplantation from matched sibling, unrelated umbilical cord blood and increasingly haploidentical donors can cure many patients with advanced hematologic malignancies through a T-cell-mediated graft-versus-tumor effect, relapse of disease remains a major cause for treatment failure. A variable proportion of such patients can achieve further remissions using donor lymphocyte infusions (DLIs) with or without pre-DLI chemotherapy. These remissions can lead to extended survival in up to 60% of patients with chronic myeloid leukemia, but in <20% of patients with acute leukemia, and in an intermediate percentage of patients with lymphoma and myeloma.2 In addition, approximately one-third of patients who received DLI developed clinically significant graft-versus-host disease (GVHD), which can be life threatening or fatal when acute and disabling in its chronic form. Indeed, the development of clinical GVHD is strongly associated with clinical tumor responses. A new strategy for selectively targeting relapsed hematologic malignancies after allogeneic stem cell transplants without inducing GHVD could offer patients a safer and potentially more effective alternative to standard DLI.
Pioneering work by Eshhar and others has demonstrated that T cells can be genetically modified to express novel tumor antigen recognition receptors composed of the variable binding domains of an immunoglobulin molecule joined to the constant, signaling domains of the T-cell receptor (TCR).3 These CAR-expressing T cells then become functionally redirected to the specific tumor antigen, which is recognized by the immunoglobulin portion of the molecule and can proliferate and mediate non–major histocompatibility complex restricted cytotoxicity against cells that express the tumor antigen target. Advances in vector technology, specifically the advent of lentiviral vectors that can efficiently target both dividing and nondividing lymphocytes, have facilitated the translation of genetically engineered T cells to the clinic.4
The most advanced clinic trials have focused on CD19, which is restricted in its expression to normal and malignant B cells. One of the major impediments to the clinical development of CAR technology has been the limited in vivo persistence and expansion of CAR-modified T cells.5 In a pilot clinical trial, autologous T cells, which were genetically engineered to express an anti-CD19 CAR, were used to treat 3 patients with refractory chronic lymphocytic leukemia (CLL), 2 of whom achieved durable complete responses (CR; >2 years) and 1 a stable partial response (PR).6 Two patients with refractory pediatric acute lymphoblastic leukemia (ALL) were successfully treated by a similar approach with 1 remission ongoing at nearly 1 year, although this patient experienced a severe cytokine release syndrome that required treatment with cytokine blockade.7 Five patients with refractory adult ALL, including 2 patients with 63% and 70% marrow blasts at enrollment, achieved a molecular remission after CD19-CAR T-cell therapy, allowing 4 of 5 patients to proceed to allogeneic stem cell transplantation.8 Other groups have reported successful treatment of progressive CD19+ B-cell malignancies, including follicular lymphoma, using CD19-CAR T cells.9
In this issue, Kochenderfer and colleagues report for the first time the results of a pilot study of 10 patients who received CD19-CAR-engineered and donor-derived allogeneic T cells for treatment of persistent or relapsed CD19-positive B-cell malignancies after allogeneic stem cell transplantation and at least 1 standard DLI infusion (see figure).1 This 2-arm study included 6 patients with relapses after matched sibling transplants and 4 patients with relapses after matched URD transplants. Peripheral blood mononuclear cells were obtained from the healthy donors, transduced with a gammaretroviral vector encoding the CD19-CAR construct (which includes a portion of the CD28 costimulatory molecular and the signaling domain of CD3ζ), and infused on day 8 of culture. None of the patients received any pre-CAR T-cell lymphodepletion treatment. Even so, 3 URD patients, including 2 with CLL and 1 with mantle cell lymphoma, exhibited complete or partial regression of disease after CD19-CAR therapy; 1 of the CLL patients has an ongoing CR, and the patient with mantle cell lymphoma has an ongoing PR. Importantly, although 6 of 10 study patients had GVHD earlier in their transplant courses, none of the patients exhibited GVHD after CAR-modified T-cell transfers. However, all of the responders had transient cytokine-mediated toxicities, including hypotension and fever, and 1 patient had depressed cardiac function that lasted for 4 months. Although these early observations are encouraging, enthusiasm for the specific approach taken in this study is tempered by the fact that the majority of patients (7/10) did not exhibit an objective response. The authors provide some clues to explain the low response rate: (1) After peaking at ∼10 days after infusion, the CD19-CAR T cells disappeared from the blood in all the patients before 1 month. (2) Expression of the inhibitory receptor PD-1 increased rapidly (by day +10) on the CAR-modified T cells. Although the objective responses occurred early after T-cell transfers, it is also unclear how well this approach would work for patients with more proliferative B-cell malignancies such as B-cell ALL. Conceivably, the response frequency and duration could be augmented by using cytoreductive and/or lymphodepleting therapies prior to CAR T-cell transfers, serial T-cell infusions, and/or the incorporation of anti-PD1 or anti-CTLA4 antibodies to minimize T-cell exhaustion. In addition, further studies will be needed to establish the best CAR “model” because the inclusion of the CD137 (4-1BB) cytoplasmic signaling domain to the CD3 ζ chain may enhance T-cell persistence, proliferation, and antitumor activity compared with CARs that carry the CD3 ζ chain alone.10 Nonetheless, the work by Kochenderfer and colleagues represents a bold and elegant step forward for CARs and hopefully for their drivers.
Conflict-of-interest disclosure: The author declares no competing financial interests.