CD19-redirected chimeric antigen receptor-modified T cells: a promising immunotherapy for children and adults with B-cell acute lymphoblastic leukemia (ALL)

Abstract

Relapsed and chemotherapy-refractory B-cell acute lymphoblastic leukemia (B-ALL) remain significant causes of cancer-associated morbidity and mortality for children and adults. Development of new molecularly targeted treatment strategies for patients with high-risk B-ALL is thus a major preclinical and clinical priority. Adoptive cellular therapy with patient-derived human T cells genetically engineered to express CD19 redirected chimeric antigen receptors (CD19 CAR T cells) is one immunotherapeutic modality that has recently demonstrated remarkable efficacy in re-inducing remission in patients with multiply relapsed B-ALL. Investigative teams at several major cancer centers are currently conducting phase I clinical trials in children and/or adults with relapsed/refractory B-ALL to assess the safety and to identify the maximally tolerated dose of each group’s CD19 CAR T-cell product. All groups have reported major clinical toxicities associated with CD19 CAR T-cell treatment, including cytokine release syndrome (CRS) and macrophage activation syndrome, neurologic dysfunction and aplasia of normal B lymphocytes, while CD19 CAR T cells persist in vivo. Toxicities have generally been transient or manageable with supportive care measures. Some patients with life-threatening CD19 CAR T-cell induced sequelae have received anti-cytokine receptor antibody treatment to diminish CRS symptoms and/or corticosteroids to terminate CAR T-cell proliferation. Remarkably, 67–90% of children and adults with B-ALL treated with CD19 CAR T cells in these trials have achieved morphologic leukemia remission with many patients also in molecular remission. The duration of CD19 CAR T cell persistence in vivo has varied appreciably among treated patients and likely reflects differences in the CD19 CAR constructs utilized at each institution. CD19-positive and CD19-negative B-ALL relapses after CD19 CAR T-cell treatment have occurred in some patients. Phase II trials to assess the efficacy of CD19 CAR T-cell immunotherapy in larger cohorts of patients with relapsed/refractory B-ALL are ongoing or planned.

Introduction

B-cell acute lymphoblastic leukemia (B-ALL), formerly called B-precursor acute lymphoblastic leukemia and the most common childhood cancer with approximately 4000 de novo cases diagnosed each year in the US, is caused by genetic mutations that induce aberrant arrest of normal lymphoid maturation, evasion of apoptosis and uncontrolled cellular proliferation [Hanahan and Weinberg, 2000; Teitell and Pandolfi, 2009]. Over 6000 incident cases of ALL occur in adults each year in the US. Increased understanding of the biologic heterogeneity of childhood ALL has facilitated development of risk-stratified chemotherapy regimens to deliver appropriately intensive therapy for each subgroup of patients [Schultz et al. 2007; Pui et al. 2008; Jeha and Pui, 2009]. Lack of rapid response to induction chemotherapy is highly predictive of future ALL relapse, which occurs in 15–20% of children with ALL and remains a leading cause of pediatric cancer mortality [Nguyen et al. 2008; Bhojwani and Pui, 2013; Inaba et al. 2013]. Adults with ALL fare even more poorly with greater than 50% relapse rates and 20–40% overall survival [Fielding et al. 2007; Moorman et al. 2012; Forman and Rowe, 2013]. Nearly half of children with newly diagnosed B-ALL have no prognostic leukemia-associated cytogenetic abnormality and many children who relapse have no distinguishing characteristics from those who achieve remission [Borowitz et al. 2008; Inaba et al. 2013; Loh et al. 2013]. Current curative intent treatment strategies for adults with B-ALL are focused upon induction of remission with multi-agent cytotoxic chemotherapy (plus tyrosine kinase inhibition for patients with BCR-ABL1-rearranged ALL) with subsequent swift hematopoietic stem cell transplantation (HSCT) when medically feasible [Lazarus and Advani, 2012]. The success of such approaches, however, remains suboptimal to date with many adult patients failing to achieve complete remission with frontline therapy or relapsing before or shortly after HSCT. In addition, significant short- and long-term toxicities often result from treatment with chemotherapeutic agents and from HSCT. Alternative therapeutic approaches for children and adults with high-risk B-ALL are necessary to improve overall survival, decrease risk of relapse and/or minimize toxicity.

Significant preclinical and clinical research efforts focused upon investigation of antibody-based and/or adoptive cellular therapeutics for high-risk cancers have been ongoing for over two decades. Various immunotherapeutic modalities and strategies in development for solid tumors and hematologic malignancies have been described comprehensively in other reviews [Park et al. 2011; Restifo et al. 2012; Fry and Mackall, 2013; Sadelain et al. 2013; Mantripragada et al. 2014]. Particular progress has recently been made with an adoptive immunotherapy approach involving the genetic engineering of human T cells with synthetic chimeric antigen receptors (CARs) against tumor-associated antigens expressed on the cell surface. In contrast to T cell receptor-directed T cells, reprogrammed CAR T cells are capable of recognizing and binding to a cell surface antigen of interest in a major histocompatibility complex (MHC) antigen-independent manner. Engagement of the CAR results in intracellular signaling via T cell costimulatory domains and subsequent exponential expansion of the CAR T cells to induce tumor cell killing. However, since antigen expression is rarely restricted to cancer cells, reprogrammed antigen-specific CAR T cells may also bind to those same antigens present on nonmalignant cells and evoke ‘on target/off tumor’ or ‘bystander’ effects that may be detrimental to the host.

In this review, we describe the development of CD19-redirected CAR T-cell approaches for human B-cell malignancies. We highlight the impressive clinical results described to date in current phase I trials testing CD19 CAR T cells in children and adults with relapsed or chemotherapy-refractory precursor B-ALL, as well as delineate potential toxicities and clinical sequelae of these promising new immunotherapeutic strategies.

CD19-redirected CAR T cells: optimal design

CARs are synthetic receptors comprised of several key components: (1) an extracellular MHC-independent antigen-binding domain usually derived from a monoclonal antibody single chain variable fragment (ScFv); (2) an extracellular spacer domain (in some CARs); (3) a transmembrane linking domain; and (4) an intracellular costimulatory T-cell signaling domain or multiple domains (Figure 1). DNA constructs encoding such CARs may be stably incorporated into human T cells via retroviral or lentiviral transduction. CARs may also be more transiently incorporated into T cells via other modalities, such as electroporation of CAR-encoding messenger RNA constructs. The first ‘modern day’ synthetic CAR was pioneered in 1989 by Eshhar and colleagues [Gross et al. 1989], which spurred tremendous interest in the adoptive cellular therapy field and has led to further refinements of this sophisticated technology during the past 25 years.

Figure 1.

Generations of chimeric antigen receptors (CARs) utilized in clinical testing.

Various iterations or ‘generations’ of CARs have been designed during the ontogeny of CAR T-cell development as researchers have become increasingly aware of the variable potency and efficacy of CARs depending upon individual domain components (Figure 1). Most first generation CARs utilized a CD3ζ intracellular signaling domain (‘first signal’) to activate T cells, which ultimately resulted in limited CAR T-cell activation and expansion, and often in nonsustained in vivo antitumor effects.

With increased understanding of the critical importance of the intracellular ‘second signal’ activation for CAR T-cell efficacy, subsequent generations of CARs have optimized ScFv and linker component design and, more importantly, have incorporated additional intracellular costimulatory signaling domains [e.g. CD27, CD28, CD134 (OX40), CD137 (4-1BB)] in efforts to increase the expansion, persistence and potency of CAR T cells, as well as to prevent cellular exhaustion in vivo (Figure 2).

Figure 2.

Binding of CD19 CAR T cells to CD19-expressing cells can induce potent on target/on tumor and on target/off tumor sequelae.

Most CARs used in current clinical trials of engineered T cells for patients with B-ALL are derived from second generation constructs with CD3ζ and another signaling endodomain. Delivery of a second costimulatory signal has indeed appeared to induce significantly greater T-cell expansion and longer-term persistence in vivo to date in treated patients.

Third generation CAR constructs comprised of CD3ζ and two additional co-stimulatory endodomains transduced into T cells are also under clinical evaluation, but have not thus far resulted in greater efficacy than second generation CARs [Davila et al. 2012; Sadelain et al. 2013; Kenderian et al. 2014; Mackall et al. 2014].

Given the potential for clinically significant CAR T-cell induced toxicities (discussed below), optimal CAR design thus must carefully balance desired antitumor potency with minimization of hazardous side effects [Gardner and Jensen, 2014].

Various groups have focused intensively in the past decade on targeting the B-lymphocyte antigen, CD19, a phosphoglycoprotein ubiquitously expressed on malignant and nonmalignant B-cells. Based upon the clinical efficacy and apparent tolerability of targeting CD20 (another commonly expressed B cell antigen) with the anti-CD20 monoclonal antibody rituximab in patients with B-cell hematologic malignancies [Cramer and Hallek, 2012], it was hypothesized that the CD19 receptor could be similarly targeted with engineered T cells expressing a CD19-redirected CAR and that treatment with these CD19 CAR T cells would also be ultimately tolerable in patients [Cooper et al. 2004; Kochenderfer and Rosenberg, 2013].

In preclinical studies, several research teams observed that co-incubation of CD19 CAR T cells with B-cell leukemia or lymphoma cell lines induced potent T-cell degranulation, cytokine production and tumor cytotoxicity in vitro. Furthermore, treatment of human ALL cell line-xenografted mice with retrovirus- or lentivirus-transduced CD19 CAR T cells (second generation) resulted in systemic trafficking of T cells to sites of CD19+ disease and in successful eradication of leukemia [Brentjens et al. 2007; Milone et al. 2009; Barrett et al. 2011].

A first-in-human clinical study conducted at the City of Hope Medical Center tested CD19- or CD20-redirected CAR T cells in patients with relapsed/refractory B-cell lymphomas observed little toxicity of infused CAR T cells, but also rapid loss of T cells in vivo at 24 hours to 1 week post infusion [Jensen et al. 2010]. Data from this trial and other early studies highlight that a major barrier to therapeutic efficacy is lack of CAR T-cell persistence, which more recent clinical trials appear to have overcome successfully with second and third generation CD19-redirected CARs.

Phase I clinical trial results to date

In this section, we review the progress of CD19 CAR T-cell clinical trials for patients with B-ALL that have been reported to date by major cancer centers in the US with an emphasis upon pediatric studies. Results of CD19 CAR T-cell therapy trials for adults with other B-cell malignancies, such as chronic lymphocytic leukemia (CLL) and B-cell lymphomas, are described in detail elsewhere [Kalos et al. 2011; Porter et al. 2011; Davila et al. 2012; Jena et al. 2014; Maus et al. 2014; Ramos et al. 2014].

National Cancer Institute CD19 CAR T cell trial [ClinicalTrials.gov identifier: NCT01593696]

The National Cancer Institute (NCI) phase 1 clinical trial aims to assess the safety, toxicities and response rates of CD19 CAR T-cell immunotherapy in children, adolescents and young adults with relapsed/refractory CD19+ B cell malignancies and to define the maximally tolerated dose (MTD) of CD19 CAR T cells using a dose escalation trial design [Lee et al. 2014b]. Via an intent-to-treat analysis, Lee and colleagues assessed the feasibility and biologic correlates of treatment with retrovirally-transduced CD19 CAR T cells with CD3ζ and CD28 costimulatory domains (CD19-CAR.z.28). All patients received T-cell depleting chemotherapy with fludarabine and cyclophosphamide prior to a single CD19-CAR.z.28 T cell infusion. A total of 19 of the 21 enrolled patients received the prescribed dose of CD19 CAR T cells, demonstrating 90% feasibility. The maximum tolerated dose (MTD) was defined as 1 × 10⁶ CD19-CAR.z.28 T cells/kg of body weight. After the dose escalation phase, patients in the expansion cohort were treated with CD19-CAR.z.28 T cells at the MTD.

The NCI investigators observed that in vivo CD19 CAR T-cell expansion peaked in peripheral blood of treated patients at approximately 14 days post infusion. Engineered T cells were largely absent at 2 months post infusion, which coincided with normal B-cell recovery (and thus lack of prolonged B cell aplasia) in their patients. Lee and colleagues also reported several anticipated toxicities of CD19 CAR T-cell treatment, including high fever and severe hypotension consistent with a cytokine release syndrome (CRS) (described in detail below) occurring in 14% of patients. Grade 3–4 signs and symptoms of CRS (see NCI Cancer Therapy Evaluation Program, http://ctep.cancer.gov) correlated positively with higher serum levels of interleukin-6 (IL-6) and interferon-γ (IFN-γ) and with higher C-reactive protein (CRP) levels. Most patients experienced temporary grade 3–4 myelosuppression that was attributed to prior chemotherapy. Six patients had transient mild neurotoxicity, including visual hallucinations or dysphasia. All of these adverse events were fully reversible with close observation and supportive care, and no patients died of treatment-related toxicity.

Intent-to-treat analysis demonstrated complete responses (CR) in 14 of 21 (67%) patients at 28 days after CD19 CAR T-cell infusion, as measured by bone marrow morphologic remission (20 patients with B-ALL patients) or by imaging (1 patient with diffuse large B-cell lymphoma). A total of 12 of the 20 B-ALL patients (60%) achieved molecular remissions with no flow cytometric evidence of minimal residual disease (MRD < 0.01%) by flow cytometric testing. Of these 12 patients, 10 underwent subsequent allogeneic HSCT and remained in MRD-negative remission at 1 year post CD19 CAR T-cell treatment. The other 2 patients, who did not undergo HSCT, relapsed with CD19-negative B-ALL at 3 and 5 months post CD19 CAR T-cell treatment, respectively. Of note, CD19 CAR T-cell treatment eradicated ALL cells in the central nervous system (CNS) of two patients with active CNS leukemia at time of T-cell infusion.

Lee and colleagues concluded that CD19-CAR.z.28 T cell therapy was feasible and safe in children, adolescents and young adults with relapsed/refractory B-ALL and demonstrated potent anti-leukemic efficacy at the MTD. However, their data also highlight the emerging issue of CD19-negative relapses (immune escape) as a potential mechanism of CD19 CAR T-cell treatment failure [Lee et al. 2014b]. This phase I trial continues to accrue eligible patients.

Children’s Hospital of Philadelphia and University of Pennsylvania CD19 CAR T cell trials (CTL019) [ClinicalTrials.gov identifier: NCT01626495, NCT01029366]

The Children’s Hospital of Philadelphia and University of Pennsylvania (CHOP/Penn) recently published updated results of their phase I trials of lentivirally-transduced CD19 CAR T cells with CD3ζ and 4-1BB (CD137) endodomains (CD19-CAR-CD3ζ-41BB) in 30 children and adults with relapsed/refractory B-cell malignancies [Grupp et al. 2013; Grupp, 2014; Maude et al. 2014b]. A total of 18 of these patients had previously relapsed after HSCT. In this trial, 27 of the 30 patients (90%) achieved complete morphologic remission with 22 of these in molecular remission (MRD < 0.01%) at 1 month after CD19-CAR-CD3ζ-41BB T-cell treatment, including 2 of 3 patients previously treated with the anti-CD19/CD3 bispecific antibody blinatumomab [Topp et al. 2011].

The CHOP/Penn investigators specifically noted an 82% CR rate in children and adults with high levels of disease burden (i.e. >50% leukemia marrow involvement), a population of patients in whom blinatumomab may have diminished efficacy [Topp et al. 2014, 2015]. Two patients with active CNS leukemia at time of T-cell infusion had no subsequent evidence of ALL in cerebrospinal fluid (CSF). Four patients subsequently withdrew from the study to pursue other anticancer treatment, including HSCT. Remission status was re-assessed in all patients at 6 months post T cell infusion with a 67% event-free survival (EFS) and 78% overall survival (OS). A total of 7 patients who achieved CR subsequently relapsed (3 with CD19-negative B-ALL) between 6 weeks and 8.5 months following CD19 CAR T-cell infusion, and all of these patients subsequently died of disease progression. The first patient treated on this trial remains in CR at 3 years post CD19 CAR T cells at the time of this review [Grupp et al. 2013; Maude et al. 2014b].

Differing from data reported by the NCI team, Maude and Frey and colleagues noted prolonged persistence of their CD19 CAR T cells in a majority of treated patients, which appeared to correlate positively with favorable clinical responses. These patients also experienced on target/off tumor toxicity with aplasia of normal/nonmalignant B lymphocytes, which had persisted for months to years in many patients and has required supportive care with monthly intravenous immunoglobulin replacement. Conversely, those patients who ‘lost’ CD19-CAR-CD3ζ-41BB cells (as evidenced by lack of circulating CD19 CARs T cells and/or loss of T-cell induced B-cell aplasia) within a few months of T cell treatment had greater incidence of relapse. The authors maintain that persistence of CD19 CAR T cells for at least 6 months is likely necessary to achieve sustained remission, particularly for patients who do not undergo subsequent HSCT [Maude et al. 2014b].

The CHOP/Penn investigators also reported several T-cell therapy associated toxicities, including CRS and neurologic abnormalities. In their trial, all treated patients experienced clinical and laboratory stigmata of CRS (e.g., fever, hypotension, hyperferritinemia, elevated cytokine levels, elevated CRP levels), which was classified as mild or moderate in 22 of 30 patients. CRS tended to be of greater severity in patients with highest leukemia burdens. The group also noted particular sensitivity of elevated serum levels IL-6 as a potential biomarker of CRS. They and others proposed that severe CRS can be safely managed in some patients with treatment with the anti-IL-6 receptor monoclonal antibody tocilizumab [Teachey et al. 2013; Barrett et al. 2014; Lee et al. 2014a; Maude et al. 2014a]. A total of 13 patients also experienced transient neurologic dysfunction, such as delirium or global encephalopathy with associated aphasia, confusion and/or hallucinations [Maude et al. 2014b].

This phase I trial continues to accrue eligible patients. Multi-institution phase II trials to investigate the efficacy of CD19-CAR-CD3ζ-41BB T-cell treatment in larger cohorts of children and adults with relapsed/refractory B-ALL are also ongoing.

University of Washington/Seattle Children’s Hospital and Fred Hutchinson Cancer Research Center CD19 CAR T cell trials, PLAT02 [ClinicalTrials.gov identifier: NCT02028455] and FH2494 [ClinicalTrials.gov identifier: NCT01475058]

The Seattle Children’s Hospital (SCH) phase I trial aims to determine the feasibility of manufacturing CD19 CAR T-cell products of defined CD4:CD8 composition and transgene expression, as well as to assess the safety and toxicity of infused CD19 CAR T cells in children and young adults with B-ALL who have relapsed after HSCT [Gardner et al. 2014]. Preliminary trial results of 13 patients treated with lymphodepleting chemotherapy and lentivirally-transduced CD19 CAR T cells with CD28, CD3ζ and 4-1BB domains and an epidermal growth factor receptor ‘termination switch’ (CD19-CAR-CD28-41BB-CD3ζ-EGFRt; see below) at 5 × 10⁵ – 5 × 10⁶ CD19 CAR T cells/kg body weight were recently reported by Gardner and colleagues at the American Society of Hematology 2014 Annual Meeting (Gardner et al. 2014). A total of 11 of 13 patients (85%) achieved molecular remissions with MRD < 0.01% and one patient had a partial response (PR). All responding patients demonstrated robust CD19 CAR T-cell expansion in vivo at 1–2 weeks post infusion, as well as T-cell persistence at >40 days with concomitant B-cell aplasia. One patient subsequently relapsed. The SCH investigators also observed CD19 CAR T-cell associated toxicities, including CRS in the 12 responding patients and generally reversible encephalopathy in 4 patients. Gardner and co-investigators also reported detectable CD19 CAR T cells in the CSF of treated patients. No patient has required termination of CD19-CAR-CD28-41BB-CD3ζ-EGFRt cells with an anti-EGFR antibody such as cetuximab. This phase I trial continues to accrue eligible patients in the dose escalation phase. Once the MTD is determined, the SCH investigators plan to open a phase II expansion arm to test the efficacy and durability of their CD19 CAR T cells in a larger cohort of patients with relapsed/refractory B-ALL regardless of prior HSCT status.

A phase I/II trial testing these CD19-CAR-CD28-41BB-CD3ζ-EGFRt cells in adults with B-cell malignancies, including B-ALL, who have relapsed after HSCT is also ongoing at the Fred Hutchinson Cancer Research Center (FHCRC). Turtle and colleagues reported preliminary findings at the American Society of Hematology 2014 Annual Meeting (Turtle et al., 2014). Of 13 patients with relapsed/refractory B-ALL enrolled to date, 11 have been treated on the dose escalation phase of this trial at doses of 2 × 10⁵, 2 × 10⁶ or 2 × 10⁷ CD19 CAR T cells/kg. A total of 9 of these patients (82%) achieved morphologic and molecular CR. One patient with B-ALL treated at the 2 × 10⁷ CD19 CAR T cells/kg dose died of severe CRS-associated complications.

Similar to results reported by other investigators, the FHCRC team has observed less CRS and overall toxicity in adults with relapsed/refractory non-Hodgkin lymphoma (NHL) or CLL who have been treated with CD19-CAR-CD28-41BB-CD3ζ-EGFRt cells [Turtle et al. 2014].

Memorial Sloan Kettering Cancer Center CD19 CAR T-cell trial [ClinicalTrials.gov identifier: NCT01044069]

Investigators at the Memorial Sloan Kettering Cancer Center (MSKCC) have published data from several phase I trials testing CD19 CAR T cells in adults with relapsed/refractory B-cell malignancies, including B-ALL, CLL and other NHL [Brentjens et al. 2011, 2013; Davila et al. 2014]. In a recent report, Davila and colleagues treated 16 adults with relapsed/refractory B-ALL with lymphodepleting chemotherapy and retrovirally-transduced CD19-CAR T cells with CD28 and CD3ζ endodomains (CD19-CAR-CD28-CD3ζ). As in other institutions’ trials, the investigators observed that CD19 CAR T-cell levels also peaked at 1–2 weeks in the peripheral blood of treated patients, coinciding with signs and symptoms of CRS as below. A total of 14 of the 16 (88%) patients demonstrated morphologic CR or CR with incomplete blood count recovery (CRi) at 1 month post CD19 CAR T-cell treatment with 12 patients in molecular remission by flow cytometric or polymerase chain (PCR) reaction-based MRD assessment. A total of 7 of 10 transplant-eligible patients subsequently underwent allogeneic HSCT. The MSKCC investigators noted CD19 CAR T-cell persistence until approximately 3 months post infusion in most patients [Davila et al. 2014].

The MSKCC team also noted CRS and reversible neurologic complications in several patients with statistically significant prolongation of hospitalization in those patients with clinically severe CRS. Davila and colleagues further correlated serum cytokine and CRP levels with CRS signs and symptoms, maintaining that CRP elevation was a rapidly obtainable and sensitive biomarker of severe CRS in their population and could be used to guide clinical management of CD19 CAR-cell-infused patients. As in other studies, some patients were treated with high-dose steroids and/or tocilizumab to mitigate severe toxicities.

MSKCC phase I trials testing CD19-CAR-CD28-CD3ζ in patients with relapsed/refractory B-ALL continue to accrue, involving adults [ClinicalTrials.gov identifier: NCT010440690]and children and young adults [ClinicalTrials.gov identifier: NCT01860937]. A phase II trial to investigate the efficacy of CD19 CAR T cells in a larger cohort of patients is also planned.

MD Anderson Cancer Center CD19 CAR T-cell trials [ClinicalTrials.gov identifier: NCT00968760, NCT01497184]

CD19 CAR T cells are also under clinical testing in several phase I trials at the MD Anderson Cancer Center (MDACC) in adults with relapsed/refractory B-cell malignancies, including B-ALL. The current MDACC CD19 CAR T-cell product utilizes a second generation CAR construct with CD3ζ/CD28 signaling moieties (CD19RCD28) and a Sleeping Beauty (SB) transposon nonviral gene transfer manufacturing system [Kebriaei et al. 2013, 2014; Singh et al. 2013].

At the American Society of Hematology 2014 Annual Meeting, Kebriaei and colleagues reported initial results of 10 adults with B-ALL treated with adjuvant CD19RCD28 following HSCT at doses of 1 × 10⁶–5 × 10⁷ modified T cells/m² body surface area. A total of 3 (30%) of these patients were in remission at a median of 5 months post CD19RCD28 infusion. In addition, 8 patients with relapsed B-ALL were treated with CD19RCD28 at doses of 1 × 10⁶–5 × 10⁷ modified T cells/m², 3 of whom (37.5%) were in remission at 3 months post T cell infusion.

The MDACC investigators reported few acute or late toxicities with their SB-modified CD19RCD28 T cells and suggest that treatment of patients with CD19 CAR T cells in the adjuvant post HSCT setting may be an effective method to eradicate MRD-level disease [Kebriaei et al. 2014]. These trials continue to accrue patients to determine the safety and to define the toxicity of CD19RCD28 T cells in adults with relapsed/refractory B-cell leukemias and lymphomas.

Baylor College of Medicine CD19 CAR T-cell trial [ClinicalTrials.gov identifier: NCT00840853]

Investigators at Baylor College of Medicine (BCM) have developed virus-specific CD19 CAR T cells that are also under current study in phase I trials for adults with relapsed/refractory B cell malignancies. Given its prior success with treating patients with severe cytomegalovirus, Epstein–Barr virus and adenovirus infections after HSCT with engineered trivirus-specific T cells (VSTs) [Leen et al. 2006; Micklethwaite et al. 2010], the BCM team subsequently hypothesized that CD19-redirected VSTs (CD19.CAR-VSTs) could be activated and expanded by endogenous viral antigens in patients with relapsed B-cell malignancies.

Two patients with relapsed B-ALL following HSCT and two patients with clinically high-risk B-ALL in sustained CR2 post HSCT were treated with CD19.CAR-VSTs [Cruz et al. 2013]. T cells persisted in vivo for 8–12 weeks after infusion. The 2 patients treated with CD19.CAR-VSTs in CR2 remained in morphologic remission at 2 and 8 months post infusion, respectively. One patient had evidence of circulating normal B cells, while the other had B cell aplasia. A third patient, who had active relapsed BCR-ABL1-rearranged ALL at time of CD19.CAR-VST treatment, achieved CR at 2 weeks post infusion. He subsequently relapsed during a time when dasatinib dosing was increased, which the authors hypothesized may have been toxic to CD19.CAR-VSTs and possibly contributed to relapse [Cruz et al. 2013]. The BCM trials are also ongoing to identify tolerable dosing, safety, and toxicity profiles of CD19.CAR-VSTs in adults with relapsed/refractory B cell malignancies.

Summary

These pivotal phase I clinical trials in the US testing various CD19 CAR T-cell products in children and adults with multiply relapsed or extremely chemorefractory B-ALL have generated tremendous clinical interest. Each team of investigators is in process of identifying the MTD of their CD19 CAR T cells and has reported various T-cell immunotherapy-associated toxicities that are discussed below in detail. The investigators also demonstrate that even severe toxicities can generally be managed with close observation, supportive care and (in some cases) intervention with steroids or cytokine receptor blockade medications. Surprisingly, even in the phase I setting, most investigative teams to date have reported remarkable antileukemia efficacy of their CD19 CAR T cells with 67–90% CR rates in extremely high-risk patients. As mentioned above, appropriately powered phase II trials are planned or ongoing at each institution to determine the efficacy of CD19 CAR T-cell treatment in patients with relapsed/refractory B-ALL.

Other early-phase trials of CD19 CAR T cells in children and/or adults with relapsed/refractory B-cell leukemias or other malignancies are also ongoing at: BCM [ClinicalTrials.gov identifier: NCT00586391, NCT01853631, NCT02050347; Xu et al. 2014]; City of Hope Medical Center [ClinicalTrials.gov identifier: NCT02146924]; MDACC [ClinicalTrials.gov identifier: NCT01362452]; FHCRC [ClinicalTrials.gov identifier: NCT01865617]; Chinese PLA General Hospital [ClinicalTrials.gov identifier: NCT01864889] CHN-PLAGH-BT-005; NCI [ClinicalTrials.gov identifier: NCT01087294, NCT00924326]; University College, London [ClinicalTrials.gov identifier: NCT01865617]; University of Pennsylvania [ClinicalTrials.gov identifier: NCT02167360, NCT02030847]; and Uppsala University [ClinicalTrials.gov identifier: NCT02132624].

Potential clinical sequelae of CD19-CAR T-cell therapy

A noteworthy issue for many current molecularly targeted anti-cancer agents, including CAR T cells, is their inability to discriminate between antigen-expressing tumor cells and antigen-expressing normal cells. Indeed, infusion of CD19 CAR T cells into patients with relapsed/refractory B-ALL has resulted a number of ‘on target/on tumor’ and ‘on target/off tumor’ side effects of varying severity (Figure 2). The most commonly reported toxicities observed in early phase clinical testing are delineated below.

Tumor lysis syndrome (TLS)

TLS refers to the cluster of metabolic disturbances (e.g. hyperkalemia, hyperphosphatemia, hyperuricemia) that may result from cellular death byproducts of rapidly growing tumors or from initiation of anti-cancer therapy [Howard et al. 2011]. TLS most frequently occurs in patients with large tumor burdens and/or in patients with highly proliferative malignancies, particularly acute leukemias. While traditionally a sequela of conventional cytotoxic chemotherapy treatment, TLS has also been observed in a small number of CD19 CAR T-cell treated patients. Perhaps due to the gradual, then exponential kinetics of in vivo stimulated CD19 CAR T-cell expansion with peak proliferation at 1–2 weeks, TLS does not appear to be a major toxicity reported in the initial CD19 CAR T cell B-ALL phase I trials even in patients with high leukemia burdens [Davila et al. 2014; Gardner et al. 2014; Lee et al. 2014b; Maude et al. 2014b].

CRS and macrophage activation syndrome (MAS)

CRS is a major toxicity that has been reported by all CD19 CAR T-cell B-ALL trial investigators to date. CRS is thought to arise from immune hyperactivation and multicytokine elevation resulting from rapid T-cell stimulation and proliferation in response to CAR engagement of the target antigen. Clinically, CRS usually manifests as mild-to-moderate fever and myalgias during the first 7–10 days post T-cell infusion and often can be managed with close observation and supportive care. However, some patients experience significant coagulopathy and vascular leak with hypotension, pulmonary edema, and occasionally multisystem organ failure that can be life-threatening (severe CRS) and requires inpatient hospitalization and intensive medical management. Clinical data reported to date suggest that higher disease burden at time of CD19 CAR T-cell infusion correlates with greater CRS severity. The magnitude of CRS symptoms may also correlate with CD19 CAR T-cell dose [Barrett et al. 2014; Lee et al. 2014a, 2014b; Maude et al. 2014a, 2014b; Maus et al. 2014]. Additional studies are necessary to assess these potential associations more fully.

Laboratory data from CD19 CAR T-cell treated patients with B-ALL who experience severe CRS have shown elevated blood levels of inflammatory cytokines (particularly IL-10, IL-6, IFN-γ), CRP and ferritin. Some patients with life-threatening CRS have been treated with systemic corticosteroids in an effort to lessen or eliminate the hyperproliferative activated CAR T cells [Davila et al. 2014; Lee et al. 2014a; Maude et al. 2014a]. However, steroid treatment in such patients has been associated with rapid ablation of engineered T cells, which may abrogate the antileukemia efficacy of CD19 CAR T cells and may contribute to subsequent disease progression or relapse [Davila et al. 2014].

With recognition of IL-6 elevation as a possible biomarker of CRS, some institutions have also treated patients with severe CRS with tocilizumab in an effort to dampen the immune activation and associated sequelae without terminating CAR T cells with lymphotoxic steroids [Barrett et al. 2014; Lee et al. 2014a; Maude et al. 2014a]. Adjuvant use of tocilizumab in CD19 CAR T-cell treated patients who experience severe CRS has correlated with decreased IL-6 levels and with clinical improvement, but has not appeared to diminish engineered T-cell persistence or anti-CD19 efficacy in reports to date [Davila et al. 2014; Lee et al. 2014b; Maude et al. 2014b].

Based on these collective data, many investigators suggest that tocilizumab should be used for first-line management of severe CRS and that steroids be reserved for patients with life-threatening CRS unresponsive to IL-6 receptor blockade. At this time, however, there is no clear indication for ‘prophylactic’ use of tocilizumab or other anticytokine receptor antibodies in patients prior to CAR T-cell treatment to mitigate or prevent potential CRS.

Some CD19 CAR T-cell treated patients with severe CRS have also developed clinical stigmata resembling those of MAS – or hemophago-cytic lymphohistiocytosis (HLH) – with massive hyperferritinemia, severe coagulopathy with hypofibrinogenemia, hepatomegaly and/or splenomegaly, prolonged fever, marked elevation of a broader spectrum of inflammatory cytokines and/or altered mental status [Maude et al. 2014a; Maus et al. 2014]. The precise mechanism(s) by which CD19 CAR T cells (as well as other CD19-targeted agents such as the bispecific CD19/CD3 T cell-engaging antibody blinatumomab) can induce CRS and MAS remain unclear. While the CRS/MAS severity thus far appears to correlate with disease burden in patients with relapsed/refractory B-ALL, it is not yet known if patients who experience greater CAR T-cell associated toxicity will also have superior antileukemia responses.

CNS trafficking of CAR T cells

Several groups of trial investigators have reported detectable CD19 CAR T cells in CSF cytospins from treated patients [Gardner et al. 2014; Lee et al. 2014b; Maude et al. 2014b]. While current trials have restricted eligibility to patients without overt CNS involvement with B-ALL, these preliminary data suggest that CAR T cells traffic to the CNS in the absence of appreciable CD19 target antigens and may effectively treat CNS leukemia. This hypothesis requires prospective evaluation, however. Strata for patients with CNS involvement with leukemia (CNS-positive or CNS3) are planned for inclusion in subsequent trials to assess the efficacy of CD19 CAR T cells specifically in CNS-positive patients.

The observed CNS trafficking of CAR T cells may also provide some explanation for the neurologic abnormalities (e.g., seizures, aphasia, encephalopathy) experienced by some CD19 CAR T cell-treated patients. CNS dysfunction has been particularly prominent in patients with significant CRS/MAS. It is plausible that stimulated systemic cytokines cross the blood–brain barrier and bind to cytokine receptors in the CNS or are produced directly within the CNS via trafficked CAR T cells, but such postulations necessitate further study.

Prolonged B-cell aplasia

The expected on target/off tumor toxicity of nonmalignant B-lymphocyte depletion has been reported in nearly all patients treated in CD19 CAR T cell phase I trials. Not surprisingly, B-cell aplasia appears to correlate with in vivo persistence of the CD19 CAR T cells, which has been quite prolonged in some patients. Indeed, ‘early’ loss of CAR T cells and concomitant return of detectable normal B cells in the peripheral blood of some patients has prompted some investigators to administer additional doses of CD19 CAR T cells to these patients in an effort to prevent B-ALL recurrence, which has had variable success in maintaining leukemia remission [Maude et al. 2014b].

B-cell aplasia can be monitored by following immunoglobulin G (IgG) blood levels and has been managed clinically with monthly infusions of intravenous IgG (IVIg) products to decrease risk of opportunistic infection in otherwise hypogammaglobulinemic patients. The long-term sequelae of B-cell aplasia and IVIg replacement remain unknown. Ideally, a ‘therapeutic window’ of CD19 CAR T-cell therapy can be identified in which the T cells persist sufficiently to eradicate cancer definitively, then senesce to allow normal immune recovery. The desired length of T-cell persistence may be relatively short to induce leukemia remission prior to swift HSCT. Greater duration of T-cell persistence may be necessary to achieve durable remission without subsequent transplantation. Such strategies require continued optimization of CD19 CAR constructs and continued testing via clinical trials.

Immune escape

Several phase I trial teams have reported CD19-negative relapses in a small number of patients with B-ALL treated with CD19 CAR T cells. These relapses remain perplexing phenomena. A few patients who relapsed with CD19-negative disease after CD19 CAR T-cell treatment had previously received blinatumomab. It has been hypothesized that repetitive CD19 targeting may facilitate selective emergence of a CD19-negative leukemic clone [Grupp et al. 2013; Lee et al. 2014b; Maude et al. 2014b]. Emerging data suggest that CD19 proteins may be lost or truncated with therapeutic targeting by CD19-redirected T cells or bispecific CD19/CD3 T-cell engaging antibodies, which remains an active area of preclinical and clinical research [Grupp et al. 2013; Maus et al. 2014; Topp et al. 2014].

A recent case report of an adult with CLL treated with CD19 CAR T cells who subsequently developed a CD19-negative plasmablastic lymphoma also highlights the complexity and potential sequelae of selective pressure by CD19-targeted therapies. These ‘escape’ phenomena are perhaps reminiscent of the development of resistance mutations in patients with BCR-ABL1-rearranged chronic myeloid leukemia treated with ABL1-targeted tyrosine kinase inhibitors or patients with FMS-like tyrosine kinase 3 (FLT3) mutant acute myeloid leukemia treated with FLT3 inhibitors [Shah et al. 2002; Grunwald and Levis, 2013].

CD19 CAR T cells: putting on the brakes

Early phase clinical trials of CD19 CAR T cells in patients with relapsed/refractory B-ALL have already demonstrated the dramatic potential efficacy of this promising immunotherapy, as well as its potential for significant toxicities. Strategies to terminate permanently modified (transduced DNA) CAR T cells via incorporated ‘suicide switches’ or to develop shorter lived T cells via transfection of antigen-redirected CAR RNA constructs are thus in development. Current genetic suicide switch approaches include herpes simplex virus thymidine kinase (HSV-TK), inducible caspase 9 (iC9), or EGFRt inclusion in the CAR construct. Some of these approaches may be quite immunogenic in vivo (HSV-TK) or will require administration of a suicide-inducing drug to patient to terminate CAR T cells (iC9, EGFRt).

CD19 CAR T cells developed by the SCH and FHCRC groups are the only product with an incorporated suicide gene in current clinical testing for B-ALL to our knowledge, although a trial of CAR T cells with an iC9 suicide switch for children and adults with refractory solid tumors is currently recruiting [ClinicalTrials.gov identifier: NCT02107963]. Preclinical studies have also demonstrated the efficacy of ‘biodegradable’ CAR T cells transfected with CD19 messenger RNA that have limited in vivo persistence [Barrett et al. 2013]. A clinical trial of RNA-modified CD19 CAR T cells will open soon for patients with relapsed/relapsed Hodgkin lymphoma [ClinicalTrials.gov identifier: NCT02277522], but these more transiently modified T cells have not been tested in patients with B-ALL.

Conclusion

The development of current CD19 CAR T cells is arguably one of the most exciting collaborative bench-to-bedside efforts in modern cancer research and has generated a tremendous amount of enthusiasm among scientists, clinicians and patients. Given the exciting clinical responses reported in the phase I setting, multisite and phase II trials are already underway or planned to assess the efficacy of this experimental treatment formally in appropriately powered cohorts of patients with relapsed/refractory B-ALL.

Many academic institutions conducting these early trials have partnered with biotechnology and pharmaceutical companies as they prepare to ‘scale up’ the engineering and manufacturing of CAR T cells and to decrease the time interval from T-cell collection from the patient to CAR T-cell infusion into the patient. However, initial experience and data from the phase I trials highlight the significant potential toxicities of CD19 CAR T cells.

As we contemplate the future of CD19 CAR T-cell treatment of patients with B-ALL, we call attention to several new challenges and questions that remain to be answered: (1) What are the optimal population(s) of patients and the optimal disease status for CAR T-cell administration? (2) What is the role of HSCT in CAR T-cell treated patients? (3) Is there an ideal therapeutic window of CAR T-cell persistence? (4) Can the potency (and thus the ‘dose’) of CAR T cells be further fine-tuned to maximize anti-cancer efficacy, but minimize toxicity? (5) What are the economics of larger-scale implementation of CAR T-cell treatment for patients with cancer?