Clinical development of CAR T cells–challenges and opportunities in translating innovative treatment concepts

Jessica Hartmann ,

Corresponding author. Tel: +49 6103774026; E-mail: [email protected]

Corresponding author. Tel: +49 6103774011; E-mail: [email protected]

Martina Schüßler-Lenz,

Division of Medical Biotechnology, Paul-Ehrlich-Institut, Langen, Germany
German Cancer Consortium (DKTK), Heidelberg, Germany

Search for more papers by this author

Attilio Bondanza,

Innovative immunotherapies, Ospedale San Raffaele, Milano, Italy

Search for more papers by this author

Christian J Buchholz

Corresponding author. Tel: +49 6103774026; E-mail: [email protected]

Corresponding author. Tel: +49 6103774011; E-mail: [email protected]

Very first published: one August two thousand seventeen Utter publication history
DOI: Ten.15252/emmm.201607485 View/save citation
Cited by (CrossRef): zero articles Check for updates

See the Glossary for abbreviations used in this article.

Abstract

Chimeric antigen receptor (CAR) T cell therapy, together with checkpoint inhibition, has been celebrated as a breakthrough technology due to the substantial benefit observed in clinical trials with patients suffering from relapsed or refractory B-cell malignancies. In this review, we provide a comprehensive overview of the clinical trials performed so far worldwide and analyze parameters such as targeted antigen and indication, CAR molecular design, CAR T cell manufacturing, anti-tumor activities, and related toxicities. More than two hundred CAR T cell clinical trials have been initiated so far, most of which aim to treat lymphoma or leukemia patients using CD19-specific CARs. An enhancing number of studies address solid tumors as well. Notably, not all clinical trials conducted so far have shown promising results. Indeed, in a few patients CAR T cell therapy resulted in severe adverse events with fatal outcome. Of note, less than 10% of the ongoing CAR T cell clinical trials are performed in Europe. Taking lead from our analysis, we discuss the problems and general hurdles preventing efficient clinical development of CAR T cells as well as opportunities, with a special concentrate on the European stage.

A subclass of medicinal products encompassing cell therapy, gene therapy, and tissue engineering. CAR T cells belong to this group as well. Specific legislation for ATMPs is valid in the EU.

Chimeric antigen receptor (CAR) T cells

T cells derived from the patient’s own blood (autologous) or derived from a healthy person (allogenic) genetically engineered to express an artificial T cell receptor, through which they are targeted to disease-related cells independently of MHC engagement.

Clinical end points and surrogate end points

There are numerous ways to treatment clinical or surrogate end points, and individual trials may use different definitions. According to guidelines of the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH), a clinical end point is a explore variable to assess the clinically relevant effect of the investigational medicinal product (IMP) in a particular disease, whereas a surrogate end point relates to a clinically significant outcome but does not itself measure a clinical benefit (ICH E8–general considerations for clinical trials).

Clinical response (as defined by the Response Evaluation Criteria In Solid Tumors (RECIST) guideline) The disappearance of all clinical evidence of a disease is called finish response (CR), whereas at least 30% tumor reduction is defined as partial response (PR). Less than 25% increase in tumor size is called stable disease (SD), and patients with more than 25% enhancing tumor mass have progressive disease (PD).

Overall survival (OS) Time from explore enrollment or randomization until death.

Progression-free survival (PFS) Time from enrollment or randomization until disease progression or death.

Event-free survival (EFS) Time from enrollment or randomization until disease progression, death, or discontinuation of treatment.

Duration of response (DoR) Time from confirmation of a response (CR, PR, or SD) until disease progression. Notably, clinical response is often used as surrogate end point in oncology trials, whereas improvement in survival is considered a direct measure of clinical benefit.

European Medicines Agency (EMA)

European authority responsible in the European Union for evaluating marketing authorisations of medicinal products including CAR T cells submitted through the centralized procedure.

Good manufacturing practice (GMP)

Production of medicinal products under defined high-quality standards.

Toxicity associated with CAR T cell therapy

On-target/off-tumor toxicity Side effects caused by killing of healthy tissue by CAR T cells due to target antigen expression outside tumor tissue.

Off-target toxicity Side effects in CAR T cell-treated patients due to cross-reactivity of the engineered antigen trussing domain with a non-related surface protein.

Cytokine-release syndrome (CRS) Systemic inflammatory response resulting in non-infective fever with elevated levels of inflammatory cytokines such as interleukin-6 and interferon-γ.

Neurotoxicity Presence of neurocognitive deficits.

Introduction

For many decades, cancer therapy mainly relied on surgery, chemotherapy, and radiotherapy. In latest years, the concept of stimulating the patient’s immune response and the observed durability of responses has established cancer immunotherapies as a novel treatment option for a series of cancer types. One promising treatment is the adoptive transfer of T cells genetically engineered to express a chimeric antigen receptor (CAR) (Fig 1A). Such CAR T cells recognize surface antigens independently from MHC confinement. When targeted to tumor surface antigens, CAR T cells proliferate and kill tumor cells upon antigen contact (Fesnak et al, 2016).

Figure 1. CAR T cell therapy–principle and clinical trial overview

(A) The CAR T cell therapy process. T cells are isolated from blood of the patient or a donor, activated, and then genetically engineered to express the CAR construct (an example shown in gray above the vector particle in violet). After ex vivo expansion of the CAR T cells, they are formulated into the final product. The patient undergoes either a conditional chemotherapy or the CAR T cell product is directly infused. (B) Schematic representation of a T cell receptor (TCR) and four types of chimeric antigen receptors (CARs) being displayed on the surface of a T cell while contacting their antigen (crimson) on the tumor cell. The single-chain variable fragment (scFv) as ligand-binding domain mediating tumor cell recognition in CARs is shown in light blue with the VH and VL domains being connected via a long pliable linker and transmembrane domain to intracellular signaling domains. Pro-inflammatory cytokines or co-stimulatory ligands voiced by the CAR T cells are depicted for the four th generation. (C) Overview of so-called wise CAR T cells products. Pooled CAR T cell products consist of two or more single-targeting CAR T cell types with distinct antigen specificities. Multi-CAR T cells harbor several CAR molecules with different antigen specificities. A tandem CAR T cell voices a CAR construct harboring two ligand-binding domains with different antigen specificities. In a conditional CAR T cell activation and co-stimulation are separated on two CAR constructs recognizing different target antigens. In the split CAR construct the ligand-binding or signaling domain is physically separated permitting managed CAR T cell activation. iCAR T cells additionally express a receptor engineered to recognize an antigen voiced on normal tissue to provide an inhibitory signal in turn. In addition CAR T cells can be tooled with suicide genes or switches (e.g., iCasp9) permitting ablation of CAR T cells. (D) Left, status of published CAR T cell gene therapy trials or trials registered at ClinicalTrials.gov including long-term follow-up studies. The status of one trial is unknown and not listed. The total number of clinical trials (dark blue bars) is compared to published clinical trials (light blue bars). The asterisk indicates zero trials. Right, phases of CAR T cell gene therapy trials. Long-term follow-up studies are not included. For nine trials, the phase classification is unknown. The asterisk indicates zero trials.

CARs are composed of an extracellular cording domain, a hinge region, a transmembrane domain, and one or more intracellular signaling domains (Fig 1B). Single-chain variable fragments (scFvs) derived from tumor antigen-reactive antibodies are commonly used as extracellular strapping domains. All CARs harbor the CD3ζ chain domain as the intracellular signaling domain. Second- or third-generation CARs also contain co-stimulatory domains, like CD28 and/or 4-1BB, improving proliferation, cytokine secretion, resistance to apoptosis, and in vivo persistence. Third-generation CARs exhibit improved effector functions and in vivo persistence as compared to second-generation CARs, whereas fourth-generation CARs, so-called TRUCKs or armored CARs, combine the expression of a second-generation CAR with factors that enhance anti-tumoral activity, such as cytokines, co-stimulatory ligands, or enzymes that degrade the extracellular matrix of solid tumors (Fig 1B; Chmielewski & Abken, 2015). To enhance the safety of CAR T cell therapy, so-called clever T cells which are either tooled with a suicide gene or include synthetic control devices are under non-clinical and clinical investigation (Fig 1C; Zhang & Xu, 2017).

Thus, CAR T cells are sophisticated medicinal products with the unique feature of being able to self-amplify and persist in treated patients. Their translation from basic and pre-clinical research to clinical trials therefore poses many challenges that slow down clinical development, while many cancer patients despairingly await novel treatment options. With the aim of identifying the hurdles in clinical translation of this therapeutic concept, we have analyzed all available data from ongoing and ended clinical trials. Based on our analysis, we suggest suggestions to facilitate translation of CAR T cell products especially in Europe.

Ended and ongoing CAR T cell clinical trials

As of the end of 2016, two hundred twenty CAR T cell trials are documented of which one hundred eighty eight are ongoing including nine long-term follow-up studies (Fig 1D, Datasets EV1 and EV2, Appendix Table S1). Most of the clinical trials conducted are phase one (128) primarily evaluating safety and dose finding, but phase 1/Two and phase two trials assessing efficacy are catching up especially with CD19 as the CAR antigen (39 of seventy five phase 1/Two or phase two trials; Fig 1D, Datasets EV1 and EV2).

The very first CAR T cell trials initiated about twenty years ago included patients with advanced epithelial ovarian cancer or metastatic renal cell carcinoma and targeted the folate receptor or carbonic anhydrase IX (CAIX), respectively (Kershaw et al, 2006; Lamers et al, 2006). The next two registered clinical trials with published results reported on single patients suffering from neuroblastoma (Dataset EV3) or follicular lymphoma (Dataset EV4) reaching finish response (Park et al, 2007; Till et al, 2008). However, the breakthrough was achieved over the following years with CD19-specific CAR T cells targeting B-cell malignancies. Accomplish or partial response was reported not only for single individuals but also for the majority of patients in some trials (Dataset EV4). From then on, the number of CAR T cell trials substantially enlargened and now grows exponentially (Fig 2A). In two thousand sixteen alone, sixty two fresh CAR T cell clinical trials have been entered into ClinicalTrials.gov.

Figure Two. CAR T cell trials over time and geographical distribution

(A) Timeline of cancer CAR T cell trials as listed in Datasets EV1 and EV2 distinguishing inbetween ongoing number (dark blue bars) and freshly initiated trials in the indicated year (light blue bars). (B) Geographical distribution of worldwide ongoing CAR T cells clinical trials (left) and distribution of trial sites of the ongoing European studies (right). Five studies are multi-centric, of which four are multi-country trials in Europe (Dataset EV5). Long-term follow-up studies are not included. Color code indicates the prevalence of trials from low (green) to high (crimson).

CAR T cell therapy was primarily introduced in the USA, then spreading to the rest of the world (Fig 2B). Presently, eighty nine CAR T cell clinical trials are in progress outside the USA, with highest numbers in China (66 trials) and Europe (14 trials; Dataset EV5). Compared to the USA and China however, Europe is clearly lagging behind. The majority of trials in Europe are performed in UK (8), followed by Germany (Three) and France (Trio) (Fig 2B).

Of the current trials, one hundred thirty three target hematological malignancies and seventy eight solid tumors (Fig 3A and B; Datasets EV1 and EV2). For tumors of the hematopoietic and lymphoid system, seventeen different CAR antigens are under investigation (Fig 3D). The most frequently targeted antigen is CD19 with fifty six ongoing and eight non-active trials. Even more antigens (22) are investigated for the treatment of solid tumors (Fig 3D). Previous trials focused on CEA as antigen-targeting colorectal cancer, breast cancer, gastric cancer, adenocarcinoma as well as liver metastases. Ongoing trials target mesothelin, ErbB2/Her2, GD2 (neuroblastoma or sarcoma), or GPC3 (hepatocellular carcinoma).

Figure Three. Indication, age, CAR generation, and targeted antigen in clinical trials

(A) Solid tumors versus tumors of the hematopoietic and lymphoid system. The number of ongoing trials (dark blue bar) is compared to the number of non-active trials (light blue bar). (B) Patient age distribution for solid tumors (light blue bars) and hematological malignancies (dark blue bars). (C) Generation of the CAR constructs applied. (D) Targeted antigens separated for tumors of hematopoietic or lymphoid origin (upper panel) and for solid tumors (lower panel).

Most clinical trials have used autologous, unselected peripheral blood mononuclear cells (PBMC) as the beginning material and IL-2 for stimulation resulting in a CAR T cell product consisting of CD4 and CD8 T cells with an activated effector T-cell phenotype. More recently, methods to isolate defined T cell subsets or to drive T cells into a certain phenotype have been developed (Xu et al, 2014; Ramos et al, 2016; Turtle et al, 2016). In addition, automated manufacturing might be an option to simplify the process and enhance the robustness of CAR T cell production. CD19-CAR T cells generated using a closed automated GMP cell processing system have been shown to be comparable to CD19-CAR T cells produced by the conventional processes in terms of transduction efficiency, phenotype, function, and overall yield (Mock et al, 2016; Priesner et al, 2016).

Typically, CAR T cells are infused intravenously. However, intra-tumoral (You et al, 2016), intracranial (Brown et al, 2015) or intra-peritoneal injection (Koneru et al, 2015), hepatic artery (Katz et al, 2015), pleural (Petrausch et al, 2012), or transcatheter arterial infusion are being investigated as well (Datasets EV1 and EV2). To increase the tolerability of the treatment and to lower the risk of side effects, the given CAR T cell dose is often split over numerous injections (e.g., three injections each one day apart; Datasets EV3 and EV4). The total treatment dose is in the range of 7.Five × ten 7 –Three.Four × ten 8 CAR T cells if a motionless dose is applied (Fig 4A). However, the majority of trials use an inter- or intra-patient dose escalation regime (Datasets EV1 and EV2). Dose escalation usually covers 2-log steps embarking somewhere inbetween one × ten 6 and one × ten 9 CAR T cells (Fig 4B). Notably, the total number of infused cells depends on the percentage of CAR-positive T cells within the product, which is very variable, not only inbetween different studies (Fig 4C) but also within single trials (Fig 4D and Datasets EV3 and EV4).

Figure Four. CAR T cell dose and percentages of CAR-positive cells within CAR T cell products

(A, B) CAR T cell dose indicated in the explore description of published CAR T cell gene therapy trials or trials registered at ClinicalTrials.gov. The CAR T cell dose is normalized to seventy five kg or 1.72 m two per dose. CAR T cells are either administered as a immobile dose (A) or in a dose escalation regimen (B). Each dot represents a single trial. (C, D) The reported number of CAR-positive cells as given in Datasets EV3 and EV4 in column «%CAR + cells (median; range)» was used to identify the median amount of CAR-positive T cells in the various cell products within one clinical trial (C) and the range of variability inbetween cell products within one clinical trial (lowest (min) and highest (max) percentage CAR-positive T cells per trial) (D).

For the generation of CAR T cells, a slight preference for the use of gamma-retroviral vectors (RVs), directly followed by lentiviral vectors (LVs), can be observed. Only very few clinical studies used electroporation for the transfer of the CAR construct (Datasets EV3 and EV4). In the majority of all trials, second-generation CARs were transferred (Fig 3C). Third- or fourth-generation CARs are being tested especially when targeted to CD19 (Dataset EV1).

Clinical benefit for many cancer patients

The most famous case of a patient who benefitted from CAR T cell therapy is very likely that of Emily Whitehead, a child suffering from recurrent acute lymphoblastic leukemia (ALL), who has now reached five years of cancer-free survival (http://emilywhitehead.com). Emily was part of the NCT01626495 trial, and thus, one of the more than three hundred patients with hematological malignancies in addition to the about one hundred fifty patients with solid tumors for which published data are available (Datasets EV3 and EV4).

CAR T cell therapy emerges to be especially active against B-cell malignancies. This is due to the tumor cell selective and homogenous expression of CD19 or CD20 as well as the lighter access for CAR T cells. Taking into account the dismal prognosis for late-stage patients, even when balanced against the observed toxicities (see below), the clinical response observed in CAR T cell trials treating CD19-positive malignancies is substantial. From the two hundred forty three patients (199 adult, forty four pediatric) treated with CD19-CAR T cells, objective response has been observed for more than 60% while only 20% did not react (Fig 5A, Dataset EV4). Of note, in trials including pediatric and adult patients, the clinical outcome appeared to be independent from age (Cruz et al, 2013; Maude et al, 2014; Lee et al, 2015; Zhang et al, 2016). Whether this holds true for overall survival will be established once long-term monitoring data become available.

Figure Five. Clinical outcome

(A, B) Best clinical outcome for hematological malignancies (A) and solid tumors (B) dependent on the targeted antigen. The number of treated patients is provided in brackets below the targeted antigen. CR, accomplish response; PR, partial response; SD, stable disease; PD, progressive disease; NR, no response; NE, not evaluable.

In five trials, more than 85% of treated patients reached finish response (CR) as best clinical outcome (NCT00968760, NCT01865617, NCT01815749, NCT01626495, NCT01044069) (for details, see Dataset EV4). In these trials, the time points for evaluation range from four weeks (Turtle et al, 2016) to thirty months (Kebriaei et al, 2016). They included patients suffering from ALL or non-Hodgkin lymphoma (NHL) with different grades of detectable disease in bone marrow (BM), extramedullary sites, or cerebrospinal fluid. Patients with minimal residual disease (MRD) or morphologic disease, defined by the percentage of blasts in BM, were also included (Dataset EV4). Interestingly, there shows up to be no major difference in remission rates inbetween patients with morphologic disease and patients with MRD (Davila et al, 2014; Maude et al, 2014; Turtle et al, 2016; Park et al, 2017). However, patients with MRD lived significantly longer (18-month follow-up) (Park et al, 2017). Thus, low disease cargo seems to improve the durability of CAR T cell therapy, at least in ALL.

Chemotherapy can be applied to reduce tumor cargo before CAR T cell therapy. After CAR T cell therapy, patients who achieved a finish response can be suggested allogeneic hematopoietic stem cell transplantation (HSCT) to provide extra curative potential. In the ALL studies, the remissions induced by CAR T cells have been variously consolidated with transplantation [three out of 30, 11%; (Maude et al, 2014); seven out of 14, 50% (Davila et al, 2014); ten out of 14, 71% (Lee et al, 2015); thirteen out of 27, 48% (Turtle et al, 2016)]. Despite these differences and with all the limitations of short-term follow-up, the durability of responses inbetween the studies emerges to be remarkably similar. This suggests that CAR T cells might provide a substantial clinical benefit regardless of HSCT consolidation. Moreover, it is often overlooked that in the same studies, a substantial fraction of patients going into remission after CAR T cells had been previously transplanted, indicating that diseases insensitive to the graft-versus-leukemia effect can be instead sensitive to engineered T cells.

Overall, there is a tendency for CD19-CAR T cells to be most effective in patients suffering from ALL, slightly less so for NHL, and the least for chronic lymphocytic leukemia (CLL), suggesting an influence of the disease type on efficacy. Albeit preliminary, the results of CAR T cells in NHL emerge promising. A mixed patient population included nine high-grade chemorefractory NHL cases, of which four achieved finish remission and extra two partial remission (Kochenderfer et al, 2015). The initial result of a subsequent trial sponsored by Kite Pharma with the same CAR T cell product confirmed these results, with four out of seven patients treated achieving accomplish remission, which is continuing after twelve months (Locke et al, 2017). Considering that the prognosis of high-grade chemorefractory NHL is particularly dismal with a median survival of few weeks, these results prompted filing of the Kite Pharma CAR T cell product with the Food and Drug Administration (FDA). For other tumor antigens of hematological malignancies, clinical benefit is less pronounced (Fig 5A, Dataset EV4), albeit final results from these mostly ongoing trials are still to come.

At variance with the results obtained for hematological malignancies, no encouraging data have been published for solid tumors (12 different antigens targeted), besides anecdotal evidence for remissions in single patients (Fig 5B, Dataset EV3). An exception is the case of CAR T cells specific for GD2 which resulted in more than 50% CR in a phase I clinical trial addressing neuroblastoma patients (Louis et al, 2011).

The high remission rates obtained with CD19-CAR T cells in B-ALL compare favorably to standard chemotherapy as well as to recently approved antibody-based therapies, such as blinatumomab, a bispecific T cell engager (BiTE) directed against CD3 and CD19. Blinatumomab induced accomplish remission in 42.9% of patients with Philadelphia chromosome-negative relapsed or refractory ALL, with a median relapse-free survival of Five.9 months according to the European public assessment report (EPAR; EMA/CHMP/469312/2015). The single-arm phase II examine was the basis for conditional marketing authorization in the EU. The favorable outcome and superior efficacy of blinatumomab (34% accomplish remission rate at twelve weeks, median OS 7.7 months) over standard chemotherapy (16% CR rate, Four.0-month median OS) have been confirmed in a randomized managed trial (Kantarjian et al, 2017). Direct comparison of finish remission rates inbetween blinatumomab and CAR T cells in B-ALL is misleading, however, due to differences in time points of outcome assessment (12 weeks in the case of blinatumomab and twenty eight days in the case of CAR T cells). Yet, the reported durable remissions and event-free survival rates following CAR T cell administration are particularly promising, especially since these patient populations included cases refractory to blinatumomab (Maude et al, 2014). Very recently, Novartis announced that the corresponding CAR T cell product CTL019 (tisagenlecleucel) received recommendation for approval by the FDA Oncologic Drugs Advisory Committee for the treatment of relapsed or refractory pediatric and youthful adult patients with B-cell ALL (www.novartis.com/news/media-releases/novartis-car-t-cell-therapy-ctl019-unanimously-10-0-recommended-approval-fda). The final decision about marketing approval will be taken by the FDA and is expected within the next months.

As can be lightly verified in the examples listed above, CAR T cell trials differ in many parameters including disease entity, disease cargo, CAR construct design, production and amplification of CAR T cells, patient pre-conditioning and administered doses, to name just a few. Given this complexity, the identification of the most relevant parameters for a positive clinical outcome is the prime concentrate of ongoing research. In those CD19-CAR T cell trials which resulted in substantial benefit for the patients, second-generation CARs were usually used (Dataset EV4). However, many other CD19-CAR T cell trials that also used second-generation CARs demonstrated a less promising outcome. Obviously, other still unknown parameters impair a positive outcome.

Lymphodepletion was demonstrated to be beneficial for enhanced in vivo CAR T cell expansion and persistence (Dai et al, 2015; Turtle et al, 2016), while IL-2 co-administration was not recommended (Zhang et al, 2015). Another significant factor influencing in vivo expansion is the CAR T cell phenotype. Products containing higher amounts of CAR T cells with a central (CD62L + ) and/or stem cell memory phenotype (CD45RA + ) demonstrated enhanced in vivo expansion (Xu et al, 2014). Notably, this phenotype can be preserved during manufacturing by using IL-7 and IL-15 instead of IL-2 for cultivation (Casucci et al, 2013). Such CAR T cells may become less tired after repeated antigen-specific stimulation (Xu et al, 2014). In a similar direction, administration of a defined CD4:CD8 CAR T cell composition (ratio 1:1) showcased for the very first time a correlation inbetween cell dose and the time point of peak CAR T cell expansion (Turtle et al, 2016). Interestingly, the absolute number of CD8 + CAR T cells was higher than that of CD4 + CAR T cells at the peak of expansion, despite CD4 + and CD8 + CAR T cells having been infused in a 1:1 ratio. Also, the used co-stimulatory domain influences CAR T cell activity. The CD28 domain confers higher anti-tumoral activity (Zhao et al, 2015), whereas longer persistence of CAR T cells is observed for the 4-1BB domain [<1 month for CD28 (Brentjens et al, 2013) versus up to four years for 4-1BB (Porter et al, 2015)].

Severe side effects and toxicities

While CAR T cell therapy has shown outstanding clinical benefit, it is sometimes associated with a multiplicity of toxicities that can be life-threatening (see Datasets EV3 and EV4 for published adverse events). Several death cases were reported, especially in the last year. These were due to neurotoxicity caused by cerebral edemas in the CD19-CAR trials sponsored by Juno Therapeutics. After the very first reported deaths, the trial was interrupted and the conditioning regimen was switched from cyclophosphamide and fludarabine to cyclophosphamide alone. However, soon after reinitiation, two more fatal cases excluded the conditioning regimen as causative. These fatal outcomes are on the one forearm surprising when considering that other ongoing CD19-CAR T cell trials did so far not report cumulating fatal cases of cerebral edemas (Table 1) (DeFrancesco, 2017). On the other arm, there had been at least one fatal outcome of neurotoxicity in another CD19-CAR T trial (NCT01865617) one hundred twenty two days after CAR T cell infusion (Turtle et al, 2016). Furthermore, reversible symptoms of neurotoxicity including confusion, delirium, expressive aphasia, encephalopathy, and seizures were reported in several other studies (Brentjens et al, 2011; Maude et al, 2014; Kochenderfer et al, 2015; Lee et al, 2015; Turtle et al, 2016). In some patients, CD19-CAR T cells have been found in cerebrospinal fluid (CSF; Brentjens et al, 2011; Lee et al, 2015). Whether neurological toxicities are solely restricted to CD19-specific CAR T cells or generally associated with CAR T cell therapy remains to be elucidated. Indeed, the potential causes for the occurrence of neurotoxicity are under debate. The postulated pathophysiological mechanisms include cytokine diffusion and/or translocation of activated CAR T cell across the blood–brain barrier.

Clinical development of CAR T cells – challenges and opportunities in translating innovative treatment concepts – Hartmann – two thousand seventeen – EMBO Molecular Medicine – Wiley Online Library

Clinical development of CAR T cells–challenges and opportunities in translating innovative treatment concepts

Jessica Hartmann ,

Corresponding author. Tel: +49 6103774026; E-mail: [email protected]

Corresponding author. Tel: +49 6103774011; E-mail: [email protected]

Martina Schüßler-Lenz,

Division of Medical Biotechnology, Paul-Ehrlich-Institut, Langen, Germany
German Cancer Consortium (DKTK), Heidelberg, Germany

Search for more papers by this author

Attilio Bondanza,

Innovative immunotherapies, Ospedale San Raffaele, Milano, Italy

Search for more papers by this author

Christian J Buchholz

Corresponding author. Tel: +49 6103774026; E-mail: [email protected]

Corresponding author. Tel: +49 6103774011; E-mail: [email protected]

Very first published: one August two thousand seventeen Total publication history
DOI: Ten.15252/emmm.201607485 View/save citation
Cited by (CrossRef): zero articles Check for updates

See the Glossary for abbreviations used in this article.

Abstract

Chimeric antigen receptor (CAR) T cell therapy, together with checkpoint inhibition, has been celebrated as a breakthrough technology due to the substantial benefit observed in clinical trials with patients suffering from relapsed or refractory B-cell malignancies. In this review, we provide a comprehensive overview of the clinical trials performed so far worldwide and analyze parameters such as targeted antigen and indication, CAR molecular design, CAR T cell manufacturing, anti-tumor activities, and related toxicities. More than two hundred CAR T cell clinical trials have been initiated so far, most of which aim to treat lymphoma or leukemia patients using CD19-specific CARs. An enlargening number of studies address solid tumors as well. Notably, not all clinical trials conducted so far have shown promising results. Indeed, in a few patients CAR T cell therapy resulted in severe adverse events with fatal outcome. Of note, less than 10% of the ongoing CAR T cell clinical trials are performed in Europe. Taking lead from our analysis, we discuss the problems and general hurdles preventing efficient clinical development of CAR T cells as well as opportunities, with a special concentrate on the European stage.

A subclass of medicinal products encompassing cell therapy, gene therapy, and tissue engineering. CAR T cells belong to this group as well. Specific legislation for ATMPs is valid in the EU.

Chimeric antigen receptor (CAR) T cells

Clinical end points and surrogate end points

There are numerous ways to treatment clinical or surrogate end points, and individual trials may use different definitions. According to guidelines of the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH), a clinical end point is a probe variable to assess the clinically relevant effect of the investigational medicinal product (IMP) in a particular disease, whereas a surrogate end point relates to a clinically significant outcome but does not itself measure a clinical benefit (ICH E8–general considerations for clinical trials).

Overall survival (OS) Time from explore enrollment or randomization until death.

Progression-free survival (PFS) Time from enrollment or randomization until disease progression or death.

Event-free survival (EFS) Time from enrollment or randomization until disease progression, death, or discontinuation of treatment.

European Medicines Agency (EMA)

European authority responsible in the European Union for evaluating marketing authorisations of medicinal products including CAR T cells submitted through the centralized procedure.

Good manufacturing practice (GMP)

Production of medicinal products under defined high-quality standards.

Toxicity associated with CAR T cell therapy

On-target/off-tumor toxicity Side effects caused by killing of healthy tissue by CAR T cells due to target antigen expression outside tumor tissue.

Off-target toxicity Side effects in CAR T cell-treated patients due to cross-reactivity of the engineered antigen trussing domain with a non-related surface protein.

Cytokine-release syndrome (CRS) Systemic inflammatory response resulting in non-infective fever with elevated levels of inflammatory cytokines such as interleukin-6 and interferon-γ.

Neurotoxicity Presence of neurocognitive deficits.

Introduction

Figure 1. CAR T cell therapy–principle and clinical trial overview

(A) The CAR T cell therapy process. T cells are isolated from blood of the patient or a donor, activated, and then genetically engineered to express the CAR construct (an example shown in gray above the vector particle in violet). After ex vivo expansion of the CAR T cells, they are formulated into the final product. The patient undergoes either a conditional chemotherapy or the CAR T cell product is directly infused. (B) Schematic representation of a T cell receptor (TCR) and four types of chimeric antigen receptors (CARs) being displayed on the surface of a T cell while contacting their antigen (crimson) on the tumor cell. The single-chain variable fragment (scFv) as ligand-binding domain mediating tumor cell recognition in CARs is shown in light blue with the VH and VL domains being connected via a long limber linker and transmembrane domain to intracellular signaling domains. Pro-inflammatory cytokines or co-stimulatory ligands voiced by the CAR T cells are depicted for the four th generation. (C) Overview of so-called brainy CAR T cells products. Pooled CAR T cell products consist of two or more single-targeting CAR T cell types with distinct antigen specificities. Multi-CAR T cells harbor several CAR molecules with different antigen specificities. A tandem CAR T cell voices a CAR construct harboring two ligand-binding domains with different antigen specificities. In a conditional CAR T cell activation and co-stimulation are separated on two CAR constructs recognizing different target antigens. In the split CAR construct the ligand-binding or signaling domain is physically separated permitting managed CAR T cell activation. iCAR T cells additionally express a receptor engineered to recognize an antigen voiced on normal tissue to provide an inhibitory signal in turn. In addition CAR T cells can be tooled with suicide genes or switches (e.g., iCasp9) permitting ablation of CAR T cells. (D) Left, status of published CAR T cell gene therapy trials or trials registered at ClinicalTrials.gov including long-term follow-up studies. The status of one trial is unknown and not listed. The total number of clinical trials (dark blue bars) is compared to published clinical trials (light blue bars). The asterisk indicates zero trials. Right, phases of CAR T cell gene therapy trials. Long-term follow-up studies are not included. For nine trials, the phase classification is unknown. The asterisk indicates zero trials.

CARs are composed of an extracellular roping domain, a hinge region, a transmembrane domain, and one or more intracellular signaling domains (Fig 1B). Single-chain variable fragments (scFvs) derived from tumor antigen-reactive antibodies are commonly used as extracellular cording domains. All CARs harbor the CD3ζ chain domain as the intracellular signaling domain. Second- or third-generation CARs also contain co-stimulatory domains, like CD28 and/or 4-1BB, improving proliferation, cytokine secretion, resistance to apoptosis, and in vivo persistence. Third-generation CARs exhibit improved effector functions and in vivo persistence as compared to second-generation CARs, whereas fourth-generation CARs, so-called TRUCKs or armored CARs, combine the expression of a second-generation CAR with factors that enhance anti-tumoral activity, such as cytokines, co-stimulatory ligands, or enzymes that degrade the extracellular matrix of solid tumors (Fig 1B; Chmielewski & Abken, 2015). To enhance the safety of CAR T cell therapy, so-called wise T cells which are either tooled with a suicide gene or include synthetic control devices are under non-clinical and clinical investigation (Fig 1C; Zhang & Xu, 2017).

Thus, CAR T cells are complicated medicinal products with the unique feature of being able to self-amplify and persist in treated patients. Their translation from basic and pre-clinical research to clinical trials therefore poses many challenges that slow down clinical development, while many cancer patients despairingly await novel treatment options. With the aim of identifying the hurdles in clinical translation of this therapeutic concept, we have analyzed all available data from ongoing and ended clinical trials. Based on our analysis, we suggest suggestions to facilitate translation of CAR T cell products especially in Europe.

Ended and ongoing CAR T cell clinical trials

The very first CAR T cell trials initiated about twenty years ago included patients with advanced epithelial ovarian cancer or metastatic renal cell carcinoma and targeted the folate receptor or carbonic anhydrase IX (CAIX), respectively (Kershaw et al, 2006; Lamers et al, 2006). The next two registered clinical trials with published results reported on single patients suffering from neuroblastoma (Dataset EV3) or follicular lymphoma (Dataset EV4) reaching finish response (Park et al, 2007; Till et al, 2008). However, the breakthrough was achieved over the following years with CD19-specific CAR T cells targeting B-cell malignancies. Finish or partial response was reported not only for single individuals but also for the majority of patients in some trials (Dataset EV4). From then on, the number of CAR T cell trials substantially enlargened and now grows exponentially (Fig 2A). In two thousand sixteen alone, sixty two fresh CAR T cell clinical trials have been entered into ClinicalTrials.gov.

Figure Two. CAR T cell trials over time and geographical distribution

Figure Three. Indication, age, CAR generation, and targeted antigen in clinical trials

Most clinical trials have used autologous, unselected peripheral blood mononuclear cells (PBMC) as the kicking off material and IL-2 for stimulation resulting in a CAR T cell product consisting of CD4 and CD8 T cells with an activated effector T-cell phenotype. More recently, methods to isolate defined T cell subsets or to drive T cells into a certain phenotype have been developed (Xu et al, 2014; Ramos et al, 2016; Turtle et al, 2016). In addition, automated manufacturing might be an option to simplify the process and enhance the robustness of CAR T cell production. CD19-CAR T cells generated using a closed automated GMP cell processing system have been shown to be comparable to CD19-CAR T cells produced by the conventional processes in terms of transduction efficiency, phenotype, function, and overall yield (Mock et al, 2016; Priesner et al, 2016).

Typically, CAR T cells are infused intravenously. However, intra-tumoral (You et al, 2016), intracranial (Brown et al, 2015) or intra-peritoneal injection (Koneru et al, 2015), hepatic artery (Katz et al, 2015), pleural (Petrausch et al, 2012), or transcatheter arterial infusion are being investigated as well (Datasets EV1 and EV2). To increase the tolerability of the treatment and to lower the risk of side effects, the given CAR T cell dose is often split over numerous injections (e.g., three injections each one day apart; Datasets EV3 and EV4). The total treatment dose is in the range of 7.Five × ten 7 –Trio.Four × ten 8 CAR T cells if a immobile dose is applied (Fig 4A). However, the majority of trials use an inter- or intra-patient dose escalation regime (Datasets EV1 and EV2). Dose escalation usually covers 2-log steps kicking off somewhere inbetween one × ten 6 and one × ten 9 CAR T cells (Fig 4B). Notably, the total number of infused cells depends on the percentage of CAR-positive T cells within the product, which is very variable, not only inbetween different studies (Fig 4C) but also within single trials (Fig 4D and Datasets EV3 and EV4).

Figure Four. CAR T cell dose and percentages of CAR-positive cells within CAR T cell products

(A, B) CAR T cell dose indicated in the investigate description of published CAR T cell gene therapy trials or trials registered at ClinicalTrials.gov. The CAR T cell dose is normalized to seventy five kg or 1.72 m two per dose. CAR T cells are either administered as a motionless dose (A) or in a dose escalation regimen (B). Each dot represents a single trial. (C, D) The reported number of CAR-positive cells as given in Datasets EV3 and EV4 in column «%CAR + cells (median; range)» was used to identify the median amount of CAR-positive T cells in the various cell products within one clinical trial (C) and the range of variability inbetween cell products within one clinical trial (lowest (min) and highest (max) percentage CAR-positive T cells per trial) (D).

Clinical benefit for many cancer patients

Figure Five. Clinical outcome

In five trials, more than 85% of treated patients reached accomplish response (CR) as best clinical outcome (NCT00968760, NCT01865617, NCT01815749, NCT01626495, NCT01044069) (for details, see Dataset EV4). In these trials, the time points for evaluation range from four weeks (Turtle et al, 2016) to thirty months (Kebriaei et al, 2016). They included patients suffering from ALL or non-Hodgkin lymphoma (NHL) with different grades of detectable disease in bone marrow (BM), extramedullary sites, or cerebrospinal fluid. Patients with minimal residual disease (MRD) or morphologic disease, defined by the percentage of blasts in BM, were also included (Dataset EV4). Interestingly, there emerges to be no major difference in remission rates inbetween patients with morphologic disease and patients with MRD (Davila et al, 2014; Maude et al, 2014; Turtle et al, 2016; Park et al, 2017). However, patients with MRD lived significantly longer (18-month follow-up) (Park et al, 2017). Thus, low disease cargo seems to improve the durability of CAR T cell therapy, at least in ALL.

Chemotherapy can be applied to reduce tumor cargo before CAR T cell therapy. After CAR T cell therapy, patients who achieved a accomplish response can be suggested allogeneic hematopoietic stem cell transplantation (HSCT) to provide extra curative potential. In the ALL studies, the remissions induced by CAR T cells have been variously consolidated with transplantation [three out of 30, 11%; (Maude et al, 2014); seven out of 14, 50% (Davila et al, 2014); ten out of 14, 71% (Lee et al, 2015); thirteen out of 27, 48% (Turtle et al, 2016)]. Despite these differences and with all the limitations of short-term follow-up, the durability of responses inbetween the studies shows up to be remarkably similar. This suggests that CAR T cells might provide a substantial clinical benefit regardless of HSCT consolidation. Moreover, it is often overlooked that in the same studies, a substantial fraction of patients going into remission after CAR T cells had been previously transplanted, indicating that diseases insensitive to the graft-versus-leukemia effect can be instead sensitive to engineered T cells.

Overall, there is a tendency for CD19-CAR T cells to be most effective in patients suffering from ALL, slightly less so for NHL, and the least for chronic lymphocytic leukemia (CLL), suggesting an influence of the disease type on efficacy. Albeit preliminary, the results of CAR T cells in NHL show up promising. A mixed patient population included nine high-grade chemorefractory NHL cases, of which four achieved accomplish remission and extra two partial remission (Kochenderfer et al, 2015). The initial result of a subsequent trial sponsored by Kite Pharma with the same CAR T cell product confirmed these results, with four out of seven patients treated achieving accomplish remission, which is continuing after twelve months (Locke et al, 2017). Considering that the prognosis of high-grade chemorefractory NHL is particularly dismal with a median survival of few weeks, these results prompted filing of the Kite Pharma CAR T cell product with the Food and Drug Administration (FDA). For other tumor antigens of hematological malignancies, clinical benefit is less pronounced (Fig 5A, Dataset EV4), albeit final results from these mostly ongoing trials are still to come.

The high remission rates obtained with CD19-CAR T cells in B-ALL compare favorably to standard chemotherapy as well as to recently approved antibody-based therapies, such as blinatumomab, a bispecific T cell engager (BiTE) directed against CD3 and CD19. Blinatumomab induced accomplish remission in 42.9% of patients with Philadelphia chromosome-negative relapsed or refractory ALL, with a median relapse-free survival of Five.9 months according to the European public assessment report (EPAR; EMA/CHMP/469312/2015). The single-arm phase II examine was the basis for conditional marketing authorization in the EU. The favorable outcome and superior efficacy of blinatumomab (34% finish remission rate at twelve weeks, median OS 7.7 months) over standard chemotherapy (16% CR rate, Four.0-month median OS) have been confirmed in a randomized managed trial (Kantarjian et al, 2017). Direct comparison of finish remission rates inbetween blinatumomab and CAR T cells in B-ALL is misleading, however, due to differences in time points of outcome assessment (12 weeks in the case of blinatumomab and twenty eight days in the case of CAR T cells). Yet, the reported durable remissions and event-free survival rates following CAR T cell administration are particularly promising, especially since these patient populations included cases refractory to blinatumomab (Maude et al, 2014). Very recently, Novartis announced that the corresponding CAR T cell product CTL019 (tisagenlecleucel) received recommendation for approval by the FDA Oncologic Drugs Advisory Committee for the treatment of relapsed or refractory pediatric and youthful adult patients with B-cell ALL (www.novartis.com/news/media-releases/novartis-car-t-cell-therapy-ctl019-unanimously-10-0-recommended-approval-fda). The final decision about marketing approval will be taken by the FDA and is expected within the next months.

As can be lightly verified in the examples listed above, CAR T cell trials differ in many parameters including disease entity, disease cargo, CAR construct design, production and amplification of CAR T cells, patient pre-conditioning and administered doses, to name just a few. Given this complexity, the identification of the most relevant parameters for a positive clinical outcome is the prime concentrate of ongoing research. In those CD19-CAR T cell trials which resulted in substantial benefit for the patients, second-generation CARs were usually used (Dataset EV4). However, many other CD19-CAR T cell trials that also used second-generation CARs displayed a less promising outcome. Obviously, other still unknown parameters impair a positive outcome.

Lymphodepletion was demonstrated to be beneficial for enhanced in vivo CAR T cell expansion and persistence (Dai et al, 2015; Turtle et al, 2016), while IL-2 co-administration was not recommended (Zhang et al, 2015). Another significant factor influencing in vivo expansion is the CAR T cell phenotype. Products containing higher amounts of CAR T cells with a central (CD62L + ) and/or stem cell memory phenotype (CD45RA + ) displayed enhanced in vivo expansion (Xu et al, 2014). Notably, this phenotype can be preserved during manufacturing by using IL-7 and IL-15 instead of IL-2 for cultivation (Casucci et al, 2013). Such CAR T cells may become less tired after repeated antigen-specific stimulation (Xu et al, 2014). In a similar direction, administration of a defined CD4:CD8 CAR T cell composition (ratio 1:1) displayed for the very first time a correlation inbetween cell dose and the time point of peak CAR T cell expansion (Turtle et al, 2016). Interestingly, the absolute number of CD8 + CAR T cells was higher than that of CD4 + CAR T cells at the peak of expansion, despite CD4 + and CD8 + CAR T cells having been infused in a 1:1 ratio. Also, the used co-stimulatory domain influences CAR T cell activity. The CD28 domain confers higher anti-tumoral activity (Zhao et al, 2015), whereas longer persistence of CAR T cells is observed for the 4-1BB domain [<1 month for CD28 (Brentjens et al, 2013) versus up to four years for 4-1BB (Porter et al, 2015)].

Severe side effects and toxicities

While CAR T cell therapy has shown awesome clinical benefit, it is sometimes associated with a multitude of toxicities that can be life-threatening (see Datasets EV3 and EV4 for published adverse events). Several death cases were reported, especially in the last year. These were due to neurotoxicity caused by cerebral edemas in the CD19-CAR trials sponsored by Juno Therapeutics. After the very first reported deaths, the trial was interrupted and the conditioning regimen was switched from cyclophosphamide and fludarabine to cyclophosphamide alone. However, soon after reinitiation, two more fatal cases excluded the conditioning regimen as causative. These fatal outcomes are on the one arm surprising when considering that other ongoing CD19-CAR T cell trials did so far not report cumulating fatal cases of cerebral edemas (Table 1) (DeFrancesco, 2017). On the other mitt, there had been at least one fatal outcome of neurotoxicity in another CD19-CAR T trial (NCT01865617) one hundred twenty two days after CAR T cell infusion (Turtle et al, 2016). Furthermore, reversible symptoms of neurotoxicity including confusion, delirium, expressive aphasia, encephalopathy, and seizures were reported in several other studies (Brentjens et al, 2011; Maude et al, 2014; Kochenderfer et al, 2015; Lee et al, 2015; Turtle et al, 2016). In some patients, CD19-CAR T cells have been found in cerebrospinal fluid (CSF; Brentjens et al, 2011; Lee et al, 2015). Whether neurological toxicities are solely restricted to CD19-specific CAR T cells or generally associated with CAR T cell therapy remains to be elucidated. Indeed, the potential causes for the occurrence of neurotoxicity are under debate. The postulated pathophysiological mechanisms include cytokine diffusion and/or translocation of activated CAR T cell across the blood–brain barrier.

Clinical development of CAR T cells–challenges and opportunities in translating innovative treatment concepts

Jessica Hartmann ,

Corresponding author. Tel: +49 6103774026; E-mail: [email protected]

Corresponding author. Tel: +49 6103774011; E-mail: [email protected]

Martina Schüßler-Lenz,

Division of Medical Biotechnology, Paul-Ehrlich-Institut, Langen, Germany
German Cancer Consortium (DKTK), Heidelberg, Germany

Search for more papers by this author

Attilio Bondanza,

Innovative immunotherapies, Ospedale San Raffaele, Milano, Italy

Search for more papers by this author

Christian J Buchholz

Corresponding author. Tel: +49 6103774026; E-mail: [email protected]

Corresponding author. Tel: +49 6103774011; E-mail: [email protected]

Very first published: one August two thousand seventeen Total publication history
DOI: Ten.15252/emmm.201607485 View/save citation
Cited by (CrossRef): zero articles Check for updates

See the Glossary for abbreviations used in this article.

Abstract

A subclass of medicinal products encompassing cell therapy, gene therapy, and tissue engineering. CAR T cells belong to this group as well. Specific legislation for ATMPs is valid in the EU.

Chimeric antigen receptor (CAR) T cells

Clinical end points and surrogate end points

Overall survival (OS) Time from explore enrollment or randomization until death.

Progression-free survival (PFS) Time from enrollment or randomization until disease progression or death.

Event-free survival (EFS) Time from enrollment or randomization until disease progression, death, or discontinuation of treatment.

European Medicines Agency (EMA)

European authority responsible in the European Union for evaluating marketing authorisations of medicinal products including CAR T cells submitted through the centralized procedure.

Good manufacturing practice (GMP)

Production of medicinal products under defined high-quality standards.

Toxicity associated with CAR T cell therapy

On-target/off-tumor toxicity Side effects caused by killing of healthy tissue by CAR T cells due to target antigen expression outside tumor tissue.

Off-target toxicity Side effects in CAR T cell-treated patients due to cross-reactivity of the engineered antigen tying domain with a non-related surface protein.

Cytokine-release syndrome (CRS) Systemic inflammatory response resulting in non-infective fever with elevated levels of inflammatory cytokines such as interleukin-6 and interferon-γ.

Neurotoxicity Presence of neurocognitive deficits.

Introduction

For many decades, cancer therapy mainly relied on surgery, chemotherapy, and radiotherapy. In latest years, the concept of stimulating the patient’s immune response and the observed durability of responses has established cancer immunotherapies as a novel treatment option for a series of cancer types. One promising treatment is the adoptive transfer of T cells genetically engineered to express a chimeric antigen receptor (CAR) (Fig 1A). Such CAR T cells recognize surface antigens independently from MHC limitation. When targeted to tumor surface antigens, CAR T cells proliferate and kill tumor cells upon antigen contact (Fesnak et al, 2016).

Figure 1. CAR T cell therapy–principle and clinical trial overview

(A) The CAR T cell therapy process. T cells are isolated from blood of the patient or a donor, activated, and then genetically engineered to express the CAR construct (an example shown in gray above the vector particle in violet). After ex vivo expansion of the CAR T cells, they are formulated into the final product. The patient undergoes either a conditional chemotherapy or the CAR T cell product is directly infused. (B) Schematic representation of a T cell receptor (TCR) and four types of chimeric antigen receptors (CARs) being displayed on the surface of a T cell while contacting their antigen (crimson) on the tumor cell. The single-chain variable fragment (scFv) as ligand-binding domain mediating tumor cell recognition in CARs is shown in light blue with the VH and VL domains being connected via a long limber linker and transmembrane domain to intracellular signaling domains. Pro-inflammatory cytokines or co-stimulatory ligands voiced by the CAR T cells are depicted for the four th generation. (C) Overview of so-called brainy CAR T cells products. Pooled CAR T cell products consist of two or more single-targeting CAR T cell types with distinct antigen specificities. Multi-CAR T cells harbor several CAR molecules with different antigen specificities. A tandem CAR T cell voices a CAR construct harboring two ligand-binding domains with different antigen specificities. In a conditional CAR T cell activation and co-stimulation are separated on two CAR constructs recognizing different target antigens. In the split CAR construct the ligand-binding or signaling domain is physically separated permitting managed CAR T cell activation. iCAR T cells additionally express a receptor engineered to recognize an antigen voiced on normal tissue to provide an inhibitory signal in turn. In addition CAR T cells can be tooled with suicide genes or switches (e.g., iCasp9) permitting ablation of CAR T cells. (D) Left, status of published CAR T cell gene therapy trials or trials registered at ClinicalTrials.gov including long-term follow-up studies. The status of one trial is unknown and not listed. The total number of clinical trials (dark blue bars) is compared to published clinical trials (light blue bars). The asterisk indicates zero trials. Right, phases of CAR T cell gene therapy trials. Long-term follow-up studies are not included. For nine trials, the phase classification is unknown. The asterisk indicates zero trials.

CARs are composed of an extracellular strapping domain, a hinge region, a transmembrane domain, and one or more intracellular signaling domains (Fig 1B). Single-chain variable fragments (scFvs) derived from tumor antigen-reactive antibodies are commonly used as extracellular roping domains. All CARs harbor the CD3ζ chain domain as the intracellular signaling domain. Second- or third-generation CARs also contain co-stimulatory domains, like CD28 and/or 4-1BB, improving proliferation, cytokine secretion, resistance to apoptosis, and in vivo persistence. Third-generation CARs exhibit improved effector functions and in vivo persistence as compared to second-generation CARs, whereas fourth-generation CARs, so-called TRUCKs or armored CARs, combine the expression of a second-generation CAR with factors that enhance anti-tumoral activity, such as cytokines, co-stimulatory ligands, or enzymes that degrade the extracellular matrix of solid tumors (Fig 1B; Chmielewski & Abken, 2015). To enhance the safety of CAR T cell therapy, so-called brainy T cells which are either tooled with a suicide gene or include synthetic control devices are under non-clinical and clinical investigation (Fig 1C; Zhang & Xu, 2017).

Thus, CAR T cells are complicated medicinal products with the unique feature of being able to self-amplify and persist in treated patients. Their translation from basic and pre-clinical research to clinical trials therefore poses many challenges that slow down clinical development, while many cancer patients despairingly await novel treatment options. With the aim of identifying the hurdles in clinical translation of this therapeutic concept, we have analyzed all available data from ongoing and finished clinical trials. Based on our analysis, we suggest suggestions to facilitate translation of CAR T cell products especially in Europe.

Finished and ongoing CAR T cell clinical trials

The very first CAR T cell trials initiated about twenty years ago included patients with advanced epithelial ovarian cancer or metastatic renal cell carcinoma and targeted the folate receptor or carbonic anhydrase IX (CAIX), respectively (Kershaw et al, 2006; Lamers et al, 2006). The next two registered clinical trials with published results reported on single patients suffering from neuroblastoma (Dataset EV3) or follicular lymphoma (Dataset EV4) reaching finish response (Park et al, 2007; Till et al, 2008). However, the breakthrough was achieved over the following years with CD19-specific CAR T cells targeting B-cell malignancies. Finish or partial response was reported not only for single individuals but also for the majority of patients in some trials (Dataset EV4). From then on, the number of CAR T cell trials substantially enhanced and now grows exponentially (Fig 2A). In two thousand sixteen alone, sixty two fresh CAR T cell clinical trials have been entered into ClinicalTrials.gov.

Figure Two. CAR T cell trials over time and geographical distribution

Figure Trio. Indication, age, CAR generation, and targeted antigen in clinical trials

Typically, CAR T cells are infused intravenously. However, intra-tumoral (You et al, 2016), intracranial (Brown et al, 2015) or intra-peritoneal injection (Koneru et al, 2015), hepatic artery (Katz et al, 2015), pleural (Petrausch et al, 2012), or transcatheter arterial infusion are being investigated as well (Datasets EV1 and EV2). To increase the tolerability of the treatment and to lower the risk of side effects, the given CAR T cell dose is often split over numerous injections (e.g., three injections each one day apart; Datasets EV3 and EV4). The total treatment dose is in the range of 7.Five × ten 7 –Trio.Four × ten 8 CAR T cells if a immobile dose is applied (Fig 4A). However, the majority of trials use an inter- or intra-patient dose escalation regime (Datasets EV1 and EV2). Dose escalation usually covers 2-log steps beginning somewhere inbetween one × ten 6 and one × ten 9 CAR T cells (Fig 4B). Notably, the total number of infused cells depends on the percentage of CAR-positive T cells within the product, which is very variable, not only inbetween different studies (Fig 4C) but also within single trials (Fig 4D and Datasets EV3 and EV4).

Figure Four. CAR T cell dose and percentages of CAR-positive cells within CAR T cell products

(A, B) CAR T cell dose indicated in the probe description of published CAR T cell gene therapy trials or trials registered at ClinicalTrials.gov. The CAR T cell dose is normalized to seventy five kg or 1.72 m two per dose. CAR T cells are either administered as a immobile dose (A) or in a dose escalation regimen (B). Each dot represents a single trial. (C, D) The reported number of CAR-positive cells as given in Datasets EV3 and EV4 in column «%CAR + cells (median; range)» was used to identify the median amount of CAR-positive T cells in the various cell products within one clinical trial (C) and the range of variability inbetween cell products within one clinical trial (lowest (min) and highest (max) percentage CAR-positive T cells per trial) (D).

Clinical benefit for many cancer patients

The most famous case of a patient who benefitted from CAR T cell therapy is most likely that of Emily Whitehead, a child suffering from recurrent acute lymphoblastic leukemia (ALL), who has now reached five years of cancer-free survival (http://emilywhitehead.com). Emily was part of the NCT01626495 trial, and thus, one of the more than three hundred patients with hematological malignancies in addition to the about one hundred fifty patients with solid tumors for which published data are available (Datasets EV3 and EV4).

CAR T cell therapy shows up to be especially active against B-cell malignancies. This is due to the tumor cell selective and homogenous expression of CD19 or CD20 as well as the lighter access for CAR T cells. Taking into account the dismal prognosis for late-stage patients, even when balanced against the observed toxicities (see below), the clinical response observed in CAR T cell trials treating CD19-positive malignancies is substantial. From the two hundred forty three patients (199 adult, forty four pediatric) treated with CD19-CAR T cells, objective response has been observed for more than 60% while only 20% did not react (Fig 5A, Dataset EV4). Of note, in trials including pediatric and adult patients, the clinical outcome appeared to be independent from age (Cruz et al, 2013; Maude et al, 2014; Lee et al, 2015; Zhang et al, 2016). Whether this holds true for overall survival will be established once long-term monitoring data become available.

Figure Five. Clinical outcome

In five trials, more than 85% of treated patients reached accomplish response (CR) as best clinical outcome (NCT00968760, NCT01865617, NCT01815749, NCT01626495, NCT01044069) (for details, see Dataset EV4). In these trials, the time points for evaluation range from four weeks (Turtle et al, 2016) to thirty months (Kebriaei et al, 2016). They included patients suffering from ALL or non-Hodgkin lymphoma (NHL) with different grades of detectable disease in bone marrow (BM), extramedullary sites, or cerebrospinal fluid. Patients with minimal residual disease (MRD) or morphologic disease, defined by the percentage of blasts in BM, were also included (Dataset EV4). Interestingly, there shows up to be no major difference in remission rates inbetween patients with morphologic disease and patients with MRD (Davila et al, 2014; Maude et al, 2014; Turtle et al, 2016; Park et al, 2017). However, patients with MRD lived significantly longer (18-month follow-up) (Park et al, 2017). Thus, low disease cargo seems to improve the durability of CAR T cell therapy, at least in ALL.

Overall, there is a tendency for CD19-CAR T cells to be most effective in patients suffering from ALL, slightly less so for NHL, and the least for chronic lymphocytic leukemia (CLL), suggesting an influence of the disease type on efficacy. Albeit preliminary, the results of CAR T cells in NHL show up promising. A mixed patient population included nine high-grade chemorefractory NHL cases, of which four achieved finish remission and extra two partial remission (Kochenderfer et al, 2015). The initial result of a subsequent trial sponsored by Kite Pharma with the same CAR T cell product confirmed these results, with four out of seven patients treated achieving accomplish remission, which is continuing after twelve months (Locke et al, 2017). Considering that the prognosis of high-grade chemorefractory NHL is particularly dismal with a median survival of few weeks, these results prompted filing of the Kite Pharma CAR T cell product with the Food and Drug Administration (FDA). For other tumor antigens of hematological malignancies, clinical benefit is less pronounced (Fig 5A, Dataset EV4), albeit final results from these mostly ongoing trials are still to come.

The high remission rates obtained with CD19-CAR T cells in B-ALL compare favorably to standard chemotherapy as well as to recently approved antibody-based therapies, such as blinatumomab, a bispecific T cell engager (BiTE) directed against CD3 and CD19. Blinatumomab induced accomplish remission in 42.9% of patients with Philadelphia chromosome-negative relapsed or refractory ALL, with a median relapse-free survival of Five.9 months according to the European public assessment report (EPAR; EMA/CHMP/469312/2015). The single-arm phase II probe was the basis for conditional marketing authorization in the EU. The favorable outcome and superior efficacy of blinatumomab (34% accomplish remission rate at twelve weeks, median OS 7.7 months) over standard chemotherapy (16% CR rate, Four.0-month median OS) have been confirmed in a randomized managed trial (Kantarjian et al, 2017). Direct comparison of accomplish remission rates inbetween blinatumomab and CAR T cells in B-ALL is misleading, however, due to differences in time points of outcome assessment (12 weeks in the case of blinatumomab and twenty eight days in the case of CAR T cells). Yet, the reported durable remissions and event-free survival rates following CAR T cell administration are particularly promising, especially since these patient populations included cases refractory to blinatumomab (Maude et al, 2014). Very recently, Novartis announced that the corresponding CAR T cell product CTL019 (tisagenlecleucel) received recommendation for approval by the FDA Oncologic Drugs Advisory Committee for the treatment of relapsed or refractory pediatric and youthful adult patients with B-cell ALL (www.novartis.com/news/media-releases/novartis-car-t-cell-therapy-ctl019-unanimously-10-0-recommended-approval-fda). The final decision about marketing approval will be taken by the FDA and is expected within the next months.

As can be lightly verified in the examples listed above, CAR T cell trials differ in many parameters including disease entity, disease cargo, CAR construct design, production and amplification of CAR T cells, patient pre-conditioning and administered doses, to name just a few. Given this complexity, the identification of the most relevant parameters for a positive clinical outcome is the prime concentrate of ongoing research. In those CD19-CAR T cell trials which resulted in substantial benefit for the patients, second-generation CARs were usually used (Dataset EV4). However, many other CD19-CAR T cell trials that also used second-generation CARs showcased a less promising outcome. Obviously, other still unknown parameters impair a positive outcome.

Lymphodepletion was demonstrated to be beneficial for enhanced in vivo CAR T cell expansion and persistence (Dai et al, 2015; Turtle et al, 2016), while IL-2 co-administration was not recommended (Zhang et al, 2015). Another significant factor influencing in vivo expansion is the CAR T cell phenotype. Products containing higher amounts of CAR T cells with a central (CD62L + ) and/or stem cell memory phenotype (CD45RA + ) displayed enhanced in vivo expansion (Xu et al, 2014). Notably, this phenotype can be preserved during manufacturing by using IL-7 and IL-15 instead of IL-2 for cultivation (Casucci et al, 2013). Such CAR T cells may become less fatigued after repeated antigen-specific stimulation (Xu et al, 2014). In a similar direction, administration of a defined CD4:CD8 CAR T cell composition (ratio 1:1) demonstrated for the very first time a correlation inbetween cell dose and the time point of peak CAR T cell expansion (Turtle et al, 2016). Interestingly, the absolute number of CD8 + CAR T cells was higher than that of CD4 + CAR T cells at the peak of expansion, despite CD4 + and CD8 + CAR T cells having been infused in a 1:1 ratio. Also, the used co-stimulatory domain influences CAR T cell activity. The CD28 domain confers higher anti-tumoral activity (Zhao et al, 2015), whereas longer persistence of CAR T cells is observed for the 4-1BB domain [<1 month for CD28 (Brentjens et al, 2013) versus up to four years for 4-1BB (Porter et al, 2015)].

Severe side effects and toxicities

While CAR T cell therapy has shown exceptional clinical benefit, it is sometimes associated with a multiplicity of toxicities that can be life-threatening (see Datasets EV3 and EV4 for published adverse events). Several death cases were reported, especially in the last year. These were due to neurotoxicity caused by cerebral edemas in the CD19-CAR trials sponsored by Juno Therapeutics. After the very first reported deaths, the trial was interrupted and the conditioning regimen was switched from cyclophosphamide and fludarabine to cyclophosphamide alone. However, soon after reinitiation, two more fatal cases excluded the conditioning regimen as causative. These fatal outcomes are on the one mitt surprising when considering that other ongoing CD19-CAR T cell trials did so far not report cumulating fatal cases of cerebral edemas (Table 1) (DeFrancesco, 2017). On the other mitt, there had been at least one fatal outcome of neurotoxicity in another CD19-CAR T trial (NCT01865617) one hundred twenty two days after CAR T cell infusion (Turtle et al, 2016). Furthermore, reversible symptoms of neurotoxicity including confusion, delirium, expressive aphasia, encephalopathy, and seizures were reported in several other studies (Brentjens et al, 2011; Maude et al, 2014; Kochenderfer et al, 2015; Lee et al, 2015; Turtle et al, 2016). In some patients, CD19-CAR T cells have been found in cerebrospinal fluid (CSF; Brentjens et al, 2011; Lee et al, 2015). Whether neurological toxicities are solely restricted to CD19-specific CAR T cells or generally associated with CAR T cell therapy remains to be elucidated. Indeed, the potential causes for the occurrence of neurotoxicity are under debate. The postulated pathophysiological mechanisms include cytokine diffusion and/or translocation of activated CAR T cell across the blood–brain barrier.

Clinical development of CAR T cells–challenges and opportunities in translating innovative treatment concepts

Jessica Hartmann ,

Martina Schüßler-Lenz,

Attilio Bondanza,

Christian J Buchholz

Abstract

Introduction

Figure 1. CAR T cell therapy–principle and clinical trial overview

Ended and ongoing CAR T cell clinical trials

Figure Two. CAR T cell trials over time and geographical distribution

Figure Three. Indication, age, CAR generation, and targeted antigen in clinical trials

Figure Four. CAR T cell dose and percentages of CAR-positive cells within CAR T cell products

Clinical benefit for many cancer patients

Figure Five. Clinical outcome

Severe side effects and toxicities

Clinical development of CAR T cells–challenges and opportunities in translating innovative treatment concepts

Jessica Hartmann ,

Martina Schüßler-Lenz,

Attilio Bondanza,

Christian J Buchholz

Abstract

Introduction

Figure 1. CAR T cell therapy–principle and clinical trial overview

Ended and ongoing CAR T cell clinical trials

Figure Two. CAR T cell trials over time and geographical distribution

Figure Three. Indication, age, CAR generation, and targeted antigen in clinical trials

Figure Four. CAR T cell dose and percentages of CAR-positive cells within CAR T cell products

Clinical benefit for many cancer patients

Figure Five. Clinical outcome

Severe side effects and toxicities

Clinical development of CAR T cells–challenges and opportunities in translating innovative treatment concepts

Jessica Hartmann ,

Martina Schüßler-Lenz,

Attilio Bondanza,

Christian J Buchholz

Abstract

Introduction

Figure 1. CAR T cell therapy–principle and clinical trial overview

Finished and ongoing CAR T cell clinical trials

Figure Two. CAR T cell trials over time and geographical distribution

Figure Trio. Indication, age, CAR generation, and targeted antigen in clinical trials

Figure Four. CAR T cell dose and percentages of CAR-positive cells within CAR T cell products

Clinical benefit for many cancer patients

Figure Five. Clinical outcome

Severe side effects and toxicities

Related movie:

No comments

Leave a Reply Cancel reply