How to train your T cell: genetically engineered chimeric antigen receptor T cells versus bispecific T-cell engagers to target CD19 in B acute lymphoblastic leukemia

Abstract

Antigen-specific T cell–based immunotherapy is getting its day in the sun. The contemporaneous development of two potent CD19-specific immunotherapeutic modalities for the treatment of B-cell malignancies provides exciting opportunities for patients, physicians and scientists alike. Patients with relapsed, refractory or poor-risk B-cell acute lymphoblastic leukemia (ALL) previously had few therapeutic options and now have two potential new lifelines. Physicians will have the choice between two powerful modalities and indeed could potentially enroll some patients on trials exploring both modalities if needed. For scientists interested in tumor immunology, the advent of chimeric antigen receptor T-cell therapy and of bispecific T-cell engagers (BiTEs) provides unprecedented opportunities to explore the promise and limitations of antigen-specific T-cell therapy in the context of human leukemia. In this article, we compare chimeric antigen receptor T cells and BiTEs targeting CD19 in B-cell ALL in the setting of the available clinical literature.

Keywords:

• adoptive cell therapy
• acute lymphoblastic leukemia
• B-cell acute lymphoblastic leukemia
• bispecific T-cell engager
• blinatumomab
• CART19
• CD19
• chimeric antigen receptor
• CTL019
• immunotherapy
• lymphoma
• non-Hodgkin lymphoma

1. Introduction

Immunotherapy was the ‘breakthrough of the year’ in 2013, [1], occurring on a background of multiple reports of novel strategies that harness the power of the immune system against cancer in patients with an otherwise poor prognosis [2,3]. The most exciting results were seen in patients with B-cell acute lymphoblastic leukemia (ALL) using anti-CD19 immunotherapy [4-7].

CD19 is an excellent tumor target because it is universally expressed on malignant B cells. CD19 expression on normal tissues is restricted to B-lymphocytes, and extensive clinical experience shows that treatment-related B-cell aplasia is generally well tolerated. A number of anti-CD19 agents are currently in clinical trials: naked or drug-conjugated monoclonal antibodies (mAbs or ADC, respectively), bispecific T-cell engagers (BiTEs) and chimeric antigen receptor T cells (CARTs). Naked mAbs rely on antibody-dependent cellular cytotoxicity, complement-dependent cytotoxicity or direct cytotoxicity, whereas ADC rely on specific internalization of a toxic payload with subsequent drug-like activity. In contrast, BiTEs and CARTs recruit the ‘Swiss Army Knife’-like polyfunctionality of T cells. Therapies that rely on T-cell function, therefore, do not employ a single effector pathway for their activity. The most promising clinical results to date have been observed using these T cell–based immunotherapies. Although both approaches target the same molecule and engage T-cell effector functions, their underlying biology differs significantly. Analyzing the characteristics and mechanism of action of these two highly effective strategies is important to direct their best clinical use.

2. Bispecific monoclonal antibodies

Bispecific mAbs are synthetic proteins that include two antigen-recognizing domains derived from the variable regions of mAbs. Typically one domain recognizes a tumor antigen (e.g., CD19) while the other binds to an immune cell (e.g., CD3, T cell) [8]. The rationale is to transiently redirect a nontumor-specific, polyclonal T-cell population against CD19-positive neoplasms by circumventing the native T-cell receptor-MHC-based recognition. The leading CD19-CD3 BiTE (blinatumomab; Amgen) was recently approved in the US by the FDA for the treatment of ALL. Blinatumomab is rapidly eliminated from the body and therefore must be administered by continuous infusion via a central line. Typically, patients are admitted for at least the first week of therapy and each cycle lasts 4 weeks, followed by a 2-week washout phase. Up to 4 – 5 cycles are administered. In ALL, blinatumomab was tested in two Phase II studies in adult patients with minimal residual disease (MRD). The results of the last, larger confirmatory trial showed that 90/113 patients became MRD negative (80%). At last follow-up (February 2014), 74 of 106 patients ending the program received the complete treatment (4 cycles or 1 cycle followed by allogeneic stem-cell transplantation [ASCT]), while 32/106 discontinued for adverse effects (AEs), disease relapse or investigator decision; 79/106 patients were alive and monitored. Serious AEs occurred in 60% of patients, with 59% of patients having grade 3 and 27% grade 4 AEs, including pyrexia (15%), tremor (7%), aphasia (5%), encephalopathy (5%) and overdose (5%). Two fatal AEs occurred on treatment: subdural hemorrhage and atypical pneumonia (the latter was considered treatment-related) [9-11].

Blinatumomab was also tested in adults with active relapsing/refractory B-cell acute lymphoblastic leukemia (B-ALL). In this more advanced setting, 43% of the 189 patients achieved complete remission (CR). About 80% of responses occurred within the first cycle and 63% were MRD negative. The median relapse-free survival was 5.9 months (5 – 8.4 months) and the median overall survival (OS) was 6.1 months (4.2 – 7.5 months). Importantly, 2% of patients had grade ≥ 3 cytokine release syndrome (CRS). Grade ≥ 3 nervous system AEs were headache (4%), encephalopathy (3%) and ataxia (2%). Three (2%) patients had grade 5 AEs considered treatment-related (sepsis, n = 2; candida infection, n = 1) [12]. Another recently published study of 36 patients with relapsed or refractory B-precursor ALL (43% prior ASCT) showed 69% of patients achieving CR, again mostly MRD-negative. In this study, median RFS was 7.6 months (4.5 – 9.5) and median OS 9.8 months (8.5 – 14.9) [5]. Seven out of 14 patients who achieved MRD response and received subsequent ASCT following blinatumomab immunotherapy achieved long-term survival longer than 2 years [13]. Similar clinical results have been obtained in pediatric cohorts, with CR rate about 30% and similar toxicity profile, requiring dose escalation of the drug. In one study, median OS was 5.7 months (3.3 – 9.7 months) with a median follow-up time of 12.4 months [14,15]. Due to the significant rates of severe CRS in the initial studies, currently blinatumomab is administered with a dose escalation of the drug during the first cycle of administration [16].

Important insights into the mechanism of action of BiTE were gleaned from correlative studies. After blinatumomab infusion, peripheral B-cell counts drop to ≤ 1 cell/μL in < 1 week and remain undetectable throughout treatment; conversely, T cells show redistribution with early loss (first few hours) from the peripheral blood (PB) and then recovery to baseline in several days. Interestingly, some patients show expansion of T cells in the PB, while in most patients T-cell counts remain at least stable. Interestingly, PB T-cell expansion was not shown to be a prerequisite for treatment efficacy. Serum cytokines such as IL-10, IL-6 and IFN-γ increased in most patients shortly after the first infusion; however, no cytokine release was observed in the following cycles [17].

3. Anti-CD19 chimeric antigen receptors

CARTs are typically autologous patient T cells that are collected by leukapheresis and then subjected to an ex vivo manufacturing process that involves activation, gene transfer, numerical expansion and cryopreservation prior to reinfusion into the patient. Gene transfer is typically carried out using lentiviral or retroviral transduction of DNA encoding a synthetic protein (chimeric antigen receptor, CAR) that is constructed from the fusion of a single-chain variable fragment of a mAb with a costimulatory domain (41BB or CD28) and the T-cell receptor CD3zeta (CD3z) chain [18,19]. The expression of the CAR19 protein redirects the T cell against CD19+ cells, leading to cytolysis.

Anti-CD19 CAR T cells have demonstrated potent antileukemia activity in early-phase clinical trials, and > 100 patients have now been treated [20]. Upon stimulation by CD19, CAR T cells generate potent cytolytic activity and are induced to proliferate and release inflammatory cytokines. Moreover, after clearing the tumor, a subset of the infused CART19 cells become memory T cells and provide long-lasting immunosurveillance [3,21].

Our group has recently shown that CART19 T cells in adults and children with r/r ALL produce a 90% CR rate a with 6-month event-free survival of 67% and 6-month OS of 78%. Most of these patients had failed multiple therapies before receiving CART19 (60% post-ASCT and 2 patients were refractory to blinatumomab). At 6 months, 68% of patients showed persistence of CART19. All the patients experienced CRS. Severe CRS occurred in 27% of the patients and was associated with a higher disease burden. CRS could be effectively treated with the anti–interleukin-6 receptor antibody tocilizumab [4]. In a more recent report, focusing on adult patients, severe CRS occurred in 11 of 12 patients. It was self-limited in two, rapidly reversed with anti-IL-6–directed therapy in six and was refractory to therapy, contributing to death in three patients [22].

Other groups have obtained comparable results using similar anti-CD19 CAR constructs. At Memorial Sloan Kettering Cancer Center, 16 adult patients with r/r B-ALL were treated with CART19 (retroviral, costimulated with CD28) and achieved a CR rate of 88%. Most of these patients subsequently underwent ASCT [6]. At the NIH, 20 children and young adults with r/r ALL were treated with CART19 (retroviral, costimulated with CD28). By intention-to-treat analysis, the rates of CR and MRD-negative CR were found to be 70 and 60%, respectively. OS at a median follow-up of 10 months was 52%. Leukemia-free survival of the 12 patients who achieved an MRD-negative CR was 78.8% beginning at 4.8 months. All 10 patients who underwent ASCT in a CAR-induced MRD-negative state remained disease-free with no unexpected peritransplant toxicities. In contrast, two patients who achieved MRD-negative complete response but were judged ineligible for ASCT both relapsed with CD19-negative leukemia at 3 and 5 months. CART19-related toxicities were reversible, with the most severe being grade 4 CRS in 3/21 of ALL patients [6].

Hence, CART19 therapy is potently antileukemic but is associated with noteworthy AEs. B-cell aplasia is an expected on-target effect, occurs in all responding patients and persists for as long as CART19 are detectable. B-cell aplasia is generally well tolerated and the resultant hypogammaglobulinemia is routinely managed with the infusion of pooled immunoglobulins, a strategy that successfully prevents opportunistic infections [23]. Tumor lysis syndrome is observed especially in patients with high tumor burden and correlates with the number of CD8+ cells in the infused product [4]. The most concerning AE is CRS. CRS is characterized initially by high fevers and malaise but can progress to capillary leak, hypoxia and hypotension. A massive release of cytokines including IL-6, TNFa, IL-10 and IFNg is usually observed between day 1 and day 10. In most cases, CRS is self-limited or rapidly reversed with anti-cytokine–directed therapies [21,24].

4. Expert opinion

Anti-CD19 BiTE and CART are two very effective approaches to treat poor-prognosis B-ALL. The absence of head-to-head comparisons of BiTE with CART precludes any definitive statements on their relative roles and merits. However, careful review of the extant clinical trial reports leads to several conclusions (Table 1).

Table 1. Comparison of the main characteristics of blinatumomab and CART19 for the treatment of relapsing/refractory B-ALL.

	Blinatumomab [5,12]	CART19 [4,7]
Clinical results in r/r B-ALL
CR rate	30 – 69%	70 – 90%
EFS (median or at last FU)	5.9 – 7.6 months	66% (at 6 months)
OS (median or at last FU)	6.1 – 9.8 months	51.6 (at 9.7 months) – 78% (at 6 months)
MRD-negative CR (% of CR)	63 – 69%	60 – 88%
Severe CRS (grade 3 – 4)	2 – 6%	27 – 28%
Other grade ≥ 3 AES	Encephalopathy, B-cell aplasia	Encephalopathy, B-cell aplasia
Long-term antitumor effect	Few patients (most received ASCT)	Multiple patients
Advantages	Off-the-shelf, predictable PK, known dose–response, able to turn off at will	Long-lasting effect after a single infusion, no need for months of onerous continuous infusion
Disadvantages	Continuous infusion via central line; theoretical induction of peripheral tolerance due to recruitment of Tregs	Personalized therapy; intensive manufacturing process; long-term follow-up required given use of lenti- or retroviral techniques

ASCT: Allogeneic stem-cell transplantation; CRS: Cytokine-release syndrome; PK: Pharmacokinetic.

The most important comparison is the CR rate: in adult and pediatric r/r ALL, blinatumomab leads to 30 – 69% CR; in similar populations, CART19 leads to 77 – 90% CR. Long-term survival comparisons are problematic because different percentages of patients underwent ASCT after blinatumomab or CART. Thus, the two modalities appear to engender broadly equivalent efficacy. The potent immune pressure exerted by both these modalities is illustrated by the potential for antigen-loss mutation relapses, as noted in ∼ 10 – 15% of patients treated with either modality [10,25]. The remainder of the relapses occurs with CD19-positive disease. In the setting of CART19, this occurs upon loss or lack of persistence of CART19 cells. In the setting of blinatumomab: when 16 patients with MRD achieved a response, 4 relapses occurred (all after completing drug treatment), of which 2 were CD19- and 2 were CD19+ [9]. In a more recent study, 10 out of 25 patients with active r/r ALL reaching CR after blinatumomab relapsed, with 3/10 relapses being CD19 negative [5].

The toxicity profile of these two approaches is similar. B-cell aplasia persists for the duration of the anti-CD19 immune pressure. Hence, in the case of blinatumomab, this lasts the duration of the infusion cycles, and in the case of CART19 this is for the duration of T-cell persistence (typically longer with lenti-41BB-CART19 and shorter for retro-28-CART19) [4,5,7].

CRS is observed in both modalities and is associated in both with tumor burden and with dose. Overall, in the CART19-treated patients, the severity of the CRS seems to be slightly higher. It is important to underline however that the presence of CRS appears correlated with the clinical activity and may be needed in order to maximize the T cell–mediated antitumor effect. However, given that the severity of CRS correlates with tumor burden, the optimal time to treat a patient with CART19 may be during a state of MRD, if one can be achieved; this, however, needs prospective study.

There are other factors to consider when comparing BiTE to CART. BiTEs are conventional off-the-shelf drugs while CART19 is a patient-specific tailored cellular therapy. Blinatumomab engages endogenous T cells, while CART19 are T cells that have undergone extensive ex vivo manipulation. Whether the endogenous T cells activated by blinatumomab are so-called ‘tumor infiltrating lymphocytes,’ circulating nonspecific T cells or both is currently unknown. However, it is important to point out that regulatory T cells are also engaged by blinatumomab and this appears to correlate with lack of response [26]. In addition, whether blinatumomab-mediated engagement can overcome tumor-associated exhaustion of ALL-associated tumor-infiltrating lymphocytes is unknown. In contrast, CART19 cells are activated ex vivo during their manufacturing process and CD19-mediated signaling is provided with a built-in costimulatory signal (typically 41BB or CD28).

Anti-CD19 activity of blinatumomab is rapidly terminated upon cessation of the infusion. In contrast, CART19 cells are able to persist long term and establish an antitumor memory. Based on data from other trials, transgenic T cells could persist for 10 years or longer, a significant benefit given the importance of maintenance therapy in ALL [27].

Interestingly, three patients relapsing after blinatumomab therapy have been successfully treated with CART19 [4]. It is unknown whether those patients would have responded to repeat blinatumomab. To our knowledge, no patient failing CART19 has been treated with blinatumomab.

The observation that T cell–based modalities are particularly effective in ALL is noteworthy. ALL has not, till now, been thought to be a particularly immune-responsive leukemia. The use of BiTE and CART indicates that this is not due to inherent resistance to T-cell killing or to proliferation kinetics that is unfavorable to control by T cells, but rather suggest that in the presence of appropriate antigen recognition, ALL is very susceptible to T cell–based attack.

In summary, CART19 and blinatumomab represent the vanguard of antigen-specific T cell–based immunotherapy. Both appear efficacious and they are associated with remarkably similar toxicity profiles. Only head-to-head comparisons will establish the relative roles of these modalities. However, until these aspirational trials are performed, we think that the most important factors to be considered include ease of manufacturing and administration, rates of toxicity, the likelihood of orchestrating a secondary immune response including epitope spreading, and, of increasing importance, the costs and utilization of resources incurred by each modality. We are biased toward the importance of persistent tumor surveillance, a feature that is thought to be important in patients undergoing successful ASCT and that occurs in patients with long-term engraftment of CART19 cells, but that does not occur upon discontinuation of blinatumomab infusions. Thus, in our view, despite the broadly equivalent short-term antileukemic effect of the BiTE and CART platforms, CART19 is preferred as it is the modality that can best assure long-term anti-CD19 immune pressure. Expanded use of the BiTE and CART platforms beyond the treatment of B-ALL and their careful comparison will show how and in what setting T cells can best be ‘trained’ to recognize tumor antigens.

Declaration of interest

M Ruella and S Gill receive research funding from Novartis. M Ruella is partly supported by the EMD-Serono Clinical Immunotherapy Fellowship from the Society for Immunotherapy of Cancer (SITC). M Ruella has a research grant from the Gabrielle Angel’s Foundation. S Gill career development grant from the American Society of Hematology, and research funding from the Leukemia & Lymphoma Society. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.