The CAR T is a mature T cell that has been modified by the incorporation of the CAR construct into its genome with the purpose to give it a new antigen specificity independent of the antigen presenting cell (APC). The CAR construct is designed to include the essential components needed to give that T cell a new antigen specificity. The method of insertion of the gene into the cell uses a gamma retrovirus or lentivirus that has been rendered replication-incompetent.
Once the inserted gene is transcribed, various components of the CAR-T construct are made and strung together using the T cell's existing machinery for the production and modification of proteins. The final product consists of an extracellular antigen-binding portion, a hinge or transmembrane part, and an intracellular part that is capable of downstream signaling to activate the T cell to mount an immune response against the antigen for cytotoxicity.
The CAR T cells that are approved are targeted towards CD19, which is a protein expressed on the surface of B cells. However, the same engineering process can make CAR T cells directed towards other proteins, DNA, RNA, polysaccharides, and so on. This underscores the great potential of this technology.
Early Single Institution Trials
Several milestones provided the essential landmarks for the development of CAR technology. Eshhar (Proc Natl Acad Sci U S A 1989;86(24):10024-10028) and Hood (Cell 1990;60(6):929-939) demonstrated the new antigen specificity and downstream signaling. This and other works formed the essential milestones towards making functional CARs.
The first generation of CAR Ts had a single signaling domain such as CD3zeta. They were able to engage CD19 but lacked the ability to mount the type of sustained activation necessary for effective cytotoxicity because they lacked the costimulatory signal that would have been provided by the APC. Since the antigen specificity is done independent of the APC, the latter cannot be relied upon to provide the needed costimulatory signal. The costimulatory signal sequence (CD28 or 4-1BB) was then added, giving rise to second-generation CAR Ts. The advantage of the costimulatory signal was that the CAR T cell could function completely independent of the APC and serially engage CD19 B cells for cytotoxicity.
The first phase I human trials provided valuable insight into this new therapeutic initiative. For example, Brentjens and colleagues reported a phase I trial of eight patients with chronic lymphocytic leukemia (Blood 2011;118(18):4817-4828). The first four patients had no meaningful benefit and the authors rightly concluded that a lack of lymphodepleting chemotherapy was part of the reason. They then added cyclophosphamide for lymphodepletion prior to the administration of CAR T cells and better T-cell expansion and clinical responses were noted.
CAR Ts demonstrated clinical efficacy in the single-site phase I setting as described by several resarchers (N Engl J Med 2014;371(16):1507-1517; Sci Transl Med 2014;6:224; Lancet 2015;385(9967):517-528). Across the three studies, CR rate was 68-90 percent and MRD-negative rate was 57-73 percent in patients, many of whom had relapsed refractory disease and had failed a previous allogeneic stem cell transplant. The toxicity profile included cytokine release syndrome (CRS), neurological toxicity, and B-cell aplasia. These and other phase I trials paved the way for carefully conducted single-arm phase II clinical trials to move CAR Ts forward into the real-world setting.
Formal Multicenter Clinical Trials
Several phase II trials were conducted to showcase the efficacy of CAR Ts in the multicenter clinical trial setting.
Maude reported on 75 children and young adults with ALL infused once with CTL019 in a multicenter fashion between 2015 and 2017 of whom 81 percent achieved a CR (N Engl J Med 2018;378:439-448). Seventy-three percent had CRS and 40 percent had neurotoxicity of any grade. CTL019 was approved by the FDA and is being marketed as tisagenlecleucel.
Neelapu also reported a multicenter phase II trial in 101 patients with DLBCL and PMBCL infused once with KTE-C19 in 2015 to 2016 in 22 study centers (N Engl J Med 2017;377(26):2531-2544). ORR was 82 percent and CR 54 percent, which appeared durable after a medial follow up of 15 months. KTE-C19 is now FDA approved and being marketed under the pharmacological name axicabtagene ciloleucel.
Schuster and colleagues presented an updated trial of tisangenlecleucel in June 2018 at the European Hematology Association conference. They treated 111 patients with one CAR-T infusion. Bridging therapy was allowed. ORR was 52 percent and CR 40 percent, which appeared durable after more than 1 year follow-up.
These trials show a high rate of complete remissions but relapses, some of which are CD19-negative, have been documented. There is little known about the outcome of patients who failed to respond to CAR T.
One valid criticism of these single arm phase II trials is that the design was based on a modified intention-to-treat paradigm. This design, though a standard and widely accepted way to conduct a phase II trial, is nonetheless biased because outcomes are calculated based on the patients that actually received the product and ignored those not treated.
An additional criticism for the CTL019 study was that the length of time it took to manufacture the product was longer at times and may have allowed for exclusion of those with a more aggressive disease phenotype due to progression. One criticism of the KTE-C19 study was that bridging chemotherapy was not allowed and, although a large percentage of patients was dosed, it is possible that lack of bridging therapy may have allowed some to progress and become ineligible to be dosed. Neither published study was clear as to the number of patients rendered ineligible by the respective study design.
There are several lessons learned from these pivotal phase II trials. The fact that they were done across states and continents indicated that the CAR T cells can be safely taken through customs and distribution channels without loss of efficacy. Secondly, the Kaplan-Meier curves appeared to have a plateau suggesting that CAR Ts might indeed be a curative option at least in B-cell malignancies in patients who previously had no real options.
Some of the patients who were treated earlier at the NIH and University of Pennsylvania are still alive many years out without disease relapse and likely cured with the single CAR-T infusion they received (Mol Ther 2017;25(10):2245-2253). This pattern of rapid response after a single dose of therapy is almost unheard of in lymphoma in the relapse setting. The erstwhile standard of care in large B-cell lymphoma in relapse is autologous stem cell transplant, which unfortunately many are unable to receive. Gisselbrecht described response rate of only 48 percent in those that relapse less than 1 year from diagnosis, suggesting that disease can be chemotherapy-resistant even in first relapse evidenced by the 3-year OS of only 39 percent (J Clin Oncol 2010;28(27):4184-4189).
If CAR-T therapy seems to have a plateau and long-term remission rate of 40-50 percent in patients after two or more lines of therapy, it stands to reason to explore options to expand access to those in relapse after one line of therapy who are likely harboring a chemotherapy-resistant disease. It is this point that has garnered significant press and formed the impetus to commence clinical trials aimed towards moving CAR-T treatment further upfront and make it more readily available to those with relapse/refractory disease.
Randomized Control Trials & the Future
The gold standard for any therapeutic intervention is a randomized control trial done in an intention-to-treat fashion. CAR Ts also need to be subject to this in an effort to clearly and unequivocally cement their role in the treatment of B-cell malignancies.
Several phase III randomized clinical trials are underway to answer this question. Zuma7 (NCT03391466) is open and enrolling while Transform (NCT03575351) and Belinda (NCT03570892) are yet to start enrolling as of Aug. 5, 2018. Phase I and II trials are summarized in Table 1.
CAR therapy has the promise of a potential for cure in patients with relapsed lymphoma who otherwise had none. The process of testing this new therapeutic option is working its way through clinical trials and rapidly becoming part of our armamentarium in the treatment of B-cell neoplasms.
Carefully conducted phase II trials have demonstrated efficacy in the multicenter setting and across continents and catalogued its toxicity profile. Ongoing phase II trials will test this new therapy against standard of care and enable us to determine where CAR Ts lie in the treatment paradigm for B-cell malignancies.
Lastly, efforts should be made to produce next-generation multi-directed CAR T cells to address the question of antigen-negative relapses, and the lessons learned from CD19 CARs should be carefully applied to expand this technology to solid tumors.
OLALEKAN OLUWOLE, MBBS, MPH, is Assistant Professor of Medicine in the Department of Medicine at Vanderbilt University Medical Center, Nashville, Tenn.