Building a CAR Garage: Preparing for the delivery of commercial CAR T products at Memorial Sloan Kettering Cancer Center.

Abstract

Two commercial Chimeric Antigen Receptor (CAR) T cell therapies for CD19 expressing B cell malignancies, Kymriah and Yescarta, have recently been approved by the FDA. The administration of CAR T cells is a complex endeavor involving cell manufacture, tracking and shipping of apheresis products, and management of novel and severe toxicities. At Memorial Sloan Kettering Cancer Center, we have identified eight essential tasks that define the CAR T cell workflow. In this review, we discuss practical aspects of CAR T cell program development, including clinical, administrative, and regulatory challenges for successful implementation.

The recent FDA approval of Chimeric Antigen Receptor (CAR) T cell therapies is the culmination of decades of scientific and clinical development¹. Modern CARs were built from advances in tumor-specific cell therapy²,³, the development of ‘T-bodies’ by Zelig Eshar⁴,⁵, advances in genetic manipulation of primary human T cells⁶,and the design of a second generation of CARs with significantly enhanced potency^7,8. Next, physician-scientists led a series of single center trials that demonstrated CAR T cell activity in relapsed and aggressive forms of non-Hodgkin’s lymphoma (NHL)^9-11, adult acute lymphoblastic leukemia (ALL)¹²,¹³, pediatric ALL¹⁴,¹⁵, and chronic lymphocytic leukemia (CLL)¹⁶.

We are now faced with administrative and logistic challenge of delivering CAR therapy outside the scope of clinical trials. Attention in the scientific and lay press has led to tremendous excitement and high demand, and meeting expectations will be a challenge. Patients may assume there will be broad access to these products in a similar fashion to other chemotherapeutic agents; however, CAR T cells will only be limited to certain centers that are certified by the manufacturers, and only provided to patients who fulfill strict label requirements. The process of building cell therapy infrastructure, meeting regulatory requirements, and ultimately treating patients is a significant undertaking. This review describes our institutional experience with the development of a CAR T cell program at Memorial Sloan Kettering Cancer Center (MSKCC) in New York City.

CARs in the Clinic

CARs are minimally composed of two domains: 1) an extracellular antibody fragment domain that recognizes tumor-associated antigens, fused via linker and transmembrane domains to 2) an internal activation signaling domain (such as CD3 zeta) that activates T cells (Figure 1). Although CARs can theoretically be designed to target any extracellular antigen, the target antigen CD19, displayed on both normal B cells and in malignant B cell lymphomas and leukemia at high levels³¹, has achieved the most clinical success. To enhance signaling and improve efficacy, second generation CARs^7,8 additionally include a co-stimulatory signaling domain derived from CD28 ^17,18 or 41BB ^18,19. While CD28 containing CARs expand robustly and to rapidly eradicate tumor²⁰, second generation 4-1BB containing CARs are initially fewer in number²¹ but may better withstand T cell exhaustion induced by tonic CAR signaling^22,23. A full assessment of the relative benefits of each approach is beyond the scope of this review, but as described below, both CD28 and 4-1BB containing CAR T cells will enter the clinic as FDA approved cellular therapies.

Figure 1: Antigen Recognition by T cells.

T cell receptors recognize intracellular antigens presented as peptides in the context of major histocompatibility complex (MHC). First generation (1G) CAR T cells target extracellular tumor-associated antigens using an antibody-derived Single Chain Variable Fragment (scFV; blue) fused to an adaptor domain (tan) and internal CD3ζ signaling domain (purple). Second Generation (2G) CAR T cells incorporate a co-stimulatory domain (red) for enhanced T cell function. Co-stimulation is commonly provided through signaling of CD28 or 4-1BB.

Tisagenlecleucel (Kymirah)

In 2012, Novartis and the University of Pennsylvania entered into a global collaboration and licensing agreement which culminated in Novartis’ tisagenlecleucel, tradename Kymriah. The ongoing ELIANA study ( NCT02435849) is the first pediatric global CAR T cell therapy trial²⁴, while the JULIET study ( NCT02445248) is evaluating the efficacy and safety of tisagenlecleucel in adult patients with relapsed/refractory DLBCL²⁵.

Tisagenlecleucel utilizes a murine single chain variable fragment to recognize human CD19, linked to a CD8a spacer domain (Table 1). Like the University of Pennsylvania construct on which it is based, it is a second-generation CAR that utilizes a 4-1BB costimulatory domain, which has important implications for persistence and activity, as discussed previously.

Table 1:

Characteristics of Approved CAR T Cell Products

Company	Gilead (KITE)	Novartis
Product Name	Axicabtagene Ciloleucel	Tisagenlecleucel
Academic License	NCI	University of Pennsylvania
Binding Domain (All Murine ScFv)	FMC63	FMC63
Indications	DLBCL, TFL PMBCL	Pediatric ALL <26 yrs old
Spacer Domain	CD28	CD8α
Transmembrane Domain	CD28	CD8α
Stimulatory Domain	CD28-CD3ζ	4-1BB-CD3ζ
Starting Cell Population Selection	None	None
Final CD4/CD8 ratio	Variable	Variable
Ablation Technology	None	None
Viral Vector	Gamma retrovirus	Lentivirus
Outcomes
CR Rate	NHL 54%	ALL 83%
Grade 3-5 Cytokine Release	30%	30%
Grade 3-5 Neurotoxicity	30%	30%
Average LOS on Study	14 days	NS
Projected Demand in the United States	4000 to 6000 patients	400-600 patients

Interim analysis of the ELIANA trial for relapsed/refractory pediatric ALL published in June 2017²⁴ showed that 83% of patients achieved CR/CRi, with 79% survival at 12 months. CAR T cells and concomitant B-cell aplasia in peripheral blood were observed for more than one year in some patients. Approximately 8% of patients suffered manufacturing failures, and 10% were not infused due to death or adverse events.

Kymriah (Tisagenlecleucel) was recently FDA approved and carries an indication for patients up to 25 years of age with B-cell precursor ALL that is refractory or in second or later relapse. Notably, ELIANA eligibility criteria required CD19 positivity confirmed within 3 months of enrollment, excluded patients with active CNS disease or Burkitt’s lymphoma/leukemia, and required a 6-month delay from transplant.

Axicabtagene Ciloleucel (Yescarta)

Kite pharmaceuticals (Santa Monica, CA), which was acquired by Gilead, developed CAR technology licensed technology from the National Cancer Institute (NCI). Axicabtagene ciloleucel, brand name Yescarta, is being evaluated in the ZUMA-1 trial ( NCT02614066) for refractory aggressive non-Hodgkin lymphoma²⁶ (Table 1). Like Kymriah, it utilizes a murine CD19 binding domain, but unlike Kymriah it contains both a CD28 spacer and a CD28 co-stimulatory signaling domain.

Response durability in NHL is lower than that seen in ALL²⁶; of 111 eligible patients, ORR was 82% and CR 59%; at a median of 15.4 months, 47% of patients were still responding and 40% in CR. Overall survival at 18 months was 52%.

Patients enrolled in the trial were >18 years old and had relapsed/refractory diffuse large B cell lymphoma (DLBCL), primary mediastinal B cell lymphoma (PMBCL) or transformed follicular lymphoma (TFL). Results from the NCI suggest that this construct can be effective in ‘double-hit’ lymphomas²⁷, suggesting CAR T cells have a role in treating aggressive disease. However, as in ALL, the patient population most likely to benefit from CAR T cell administration, and the optimal timing of therapy, are not yet known.

Building a CAR T Cell Program

The process of CAR T cell administration involves lymphocyte collection, genetic manipulation, and reinfusion. There are complex logistics involved in collecting cells in the apheresis center, shipping to the manufacturer for production, and coordinating receipt of the product.

Several organizations provide standards and regulations for this process. The Federal Drug Administration (FDA) approves and regulates products, manages adverse event reporting, and oversee a Risk Evaluation and Mitigation (REMS) process in partnership with the manufacturers. The Foundation for the Accreditation of Cellular Therapy (FACT) provides standards for handling, processing, and tracking of cell therapy products^28-30. The Center for International Blood and Marrow Transplant Research functions as a data repository and conduit between centers and the FDA. Finally, manufacturers themselves inspect apheresis and storage facilities, authorize individual centers to provide therapeutics, and provide both the products themselves and mechanisms for product shipping and tracking.

At MSKCC, we identified eight essential tasks that define the CAR T cell workflow and are required to successfully implement the delivery of commercial CAR T cell products in the clinic. We developed this workflow by leveraging the experience of both our investigational CAR T cell team (Cellular Therapeutic Center) and our adult and pediatric hematopoietic stem cell transplantation (HSCT) programs. The overall development was guided by a task force headed by a project managers and lead physicians. The task force included physician representative from leukemia, lymphoma, and bone marrow transplantation, nursing leadership, laboratory medicine, apheresis and transfusion medicine, legal services, financial services, social work and hospital administration. The eight tasks together with the personnel necessary to implement them are depicted in Figure 2.

Figure 2:

The eight essential steps required to establish a CAR T cell program and the personnel required to implement them. Key regulatory agencies involved in each step are denoted by a bullet.

Task 1: Patient Intake

Institutions must build an effective administrative infrastructure to facilitate patient referral, selection, and evaluation. This includes administrative staff to field calls and intake, nursing staff for triage and coordination, and financial and insurance advisers to ensure eligibility. While common to all medical enterprise, intake and pre-screening are particularly important for CAR T cell, as demand is likely to be high, eligibility is limited, and the availability of CAR T cells will at least initially be limited to a small number of centers.

Task 2: CAR T Cell Consultation Service:

We next developed a CAR T cell consultation service to handle patient selection and treatment. The consultation service consists of physicians, nurses, and social workers with expertise in CAR T cell therapy. Specialized physicians must be trained and registered with both the institution and manufacturers as CAR certified MDs who are authorized to prescribe CAR products and manage CAR patients.

A key task of the consultation service includes assessing patients for therapy. Appropriate candidates must have disease that will benefit from CAR T cell treatment, fit the labeling indication for the product, be well enough to receive therapy, have a caregiver, be able to stay within 2 hours of the treating institution for one month, and have insurance coverage or other financing for treatment. Disease burden at time of evaluation is a critical consideration, as patients with low burden of disease experience fewer treatment related toxicities and appear to derive more benefit from treatment¹³.

If investigational cell therapy products are available, patients must be triaged between the commercially available products and clinical trials. In some cases, this will be straightforward and based on product labeling; for example, a 27-year-old with ALL is not eligible for Kymriah, and an adult with Mantle Cell Lymphoma is not eligible for Yescarta. At other times, patients will be eligible for commercial CAR products, standard of care treatment, and clinical trials. Therefore, it is important to create a process by which patients and physicians can decide on the appropriate therapy.

Of particular importance is that patients continue to receive therapy that is most appropriate for their disease; thus, for example, patients with chemosensitive relapsed DLBCL should be treated with an autologous HSCT after high dose BEAM until the appropriate randomized trials of CAR T vs autologous HSCT have been performed. At our institution, candidates are reviewed at a weekly conference which brings together CAR certified MDs and disease-focused specialists.

Task 3: CAR T Cell Collection, Ordering, Shipping and Receiving

CAR T cell therapy requires an apheresis center and cell storage facility where T cells can be harvested, maintained, and easily transferred. This requires training of apheresis personnel, cell therapy technologists, administrative personnel, and nursing coordinators to process CAR products. Such centers and the products they generate are regulated by and must be registered with the FDA.

Despite concerns regarding leukopenia in these heavily pre-treated patients, minimal required collections can be successful in more than 90% of patients, with optimal yields in approximately 70% ³¹. There are, however, minimal requirements for initiation of collection; the Novartis product, for example, requires absolute lymphocyte counts > 500 mcL/L or absolute CD3+ T cell counts > 150 mcL/L.

Both Novartis and Kite have an extensive contracting process to establish technical and quality agreements for apheresis and cell processing facilities. The goal of these agreements is to ensure a high-quality cell product is delivered reliably to the manufacturer. Manufacturers work with the institution to define processes for tracking and verifying product identity, manipulating and storing products, and training personnel. Importantly, cell manipulation and freezing protocols must be reconciled with technical standards established by each company without significantly disrupting existing practice. This is a thorough and time-consuming process which involves on-site inspections by the manufacturer and, in certain cases, mock shipments.

Kymriah, but not Yescarta, can be collected, frozen, and stored for later use, a strategy Novartis has termed “at risk collection.” The acceptable duration of storage is validated up to 9 months per the manufacturer. Similarly, the optimal timing for collection is not known, although theoretically early collection before patients are heavily pre-treated and leukopenic may be beneficial.

Once collected, apheresis products are shipped to a commercial facility where T cells are isolated, activated, genetically modified with a CAR-encoding vector, and expanded prior to cryopreservation. Our own institutional process of clinical CAR- T manufacturing is reviewed in Wang et al³². After processing, they are shipped back to the originating institution. One critical area of focus, and a key focus of FACT standards, is ensuring chain of custody of the cell product. Staff must be trained on the software programs developed by each product manufacturer which coordinates shipping and ensures product identity.

Task 4: Bridging Strategy

Average turn-around time for initial apheresis to return of cells to the clinical site is 17 days for Yescarta³³, with some products requiring 2-4 weeks before minimum cell numbers are obtained; slow expansion, insurance authorization and treatment planning can extend the bridging period to 1-2 months. While time-to-infusion has been drastically reduced, many patients have aggressive malignancies which must be carefully managed during insurance authorization, apheresis and manufacture. In selecting patients for CAR T cell therapy, clinicians must evaluate the pace of disease progression, the potential for near-term complications, and the expected date of infusion.

This process is managed in collaboration between the disease-specific primary oncologist and the CAR T cell consultation service, with nursing and administrative staff in support. Alternative therapies for patients who suffer manufacturing failures or disease complications should be considered; for example, in ZUMA-1, 101 of 111 patients ultimately received therapy²⁶, with one manufacturing failure, one rapid disease progression, and several interval adverse events and other medical complications.

Two to seven days prior to infusion, patients receive a lymphodepleting conditioning regimen to enhance CAR T cell function and engraftment^34,35. This requires the development of order sets specific to the conditioning regimen used for each commercial product.

While there are slight differences in the conditioning regimens use for different products and trials, investigators have coalesced around low-dose cyclophosphamide and fludarabine, as utilized in JULIET and ZUMA-1. The optimal conditioning regimen has not been rigorously defined, but experiences at the NCI¹⁰ and Fred Hutchinson Cancer Center³⁶ are illustrative, with more aggressively lymphodepleting regimens associated with both increased expansion and toxicity.

Task 5: CAR T-Cell Infusion

Infusion requires the development of order sets, nursing standard of practice protocols, and documentation and verification procedures for administration. As part of FACT accreditation, electronic cell therapy orders must be developed for collection and thaw/infusion products and must be recorded in chemotherapy administration records. Therefore, pharmacy may track and register cell products and a work-flow for pharmacy involvement in product administration should be developed.

Consents must be developed for cell collection and infusion. Although specific consents are not mandated by regulatory agencies, the complex process and significant toxicity suggests a thorough patient consent process should be a central part of CAR T cell administration.

Task 6, Post Infusion Care (Day 0 to 30):

After infusion, patients are at risk for CAR T cell related adverse effects. The Cytokine Release Syndrome (CRS)^12,37,38 occurs on a spectrum from low grade fever, which may not require any treatment, to fulminant multi-organ failure, which requires intensive care and aggressive immunosuppression. In some studies, approximately 10-15% of CAR T cell patients have required vasopressor support for hypotension and/or mechanical ventilation^36,39. The mainstays of treatment for severe CRS include the anti-IL6R antibody Tocilizumab, high-dose steroids, and supportive care delivered in an intensive care unit⁴⁰.

In addition to CRS, a syndrome of neurologic toxicity has been described^41,42. Severe neurologic toxicities include encephalopathy, nerve palsies, and seizures, and have resulted in patient deaths^43,44; these are often preceded by more subtle deficits in complex commands such as hand-writing, word-finding, and attention⁴⁰. Information on the appropriate clinical management of CRS and other CAR T toxicities is both extensively reviewed elsewhere^40,45 and is available on the package insert of approved therapies.

Subspecialists in Neurology and Intensive Care must be educated about CAR specific toxicities. We have developed management guidelines to harmonize treatment of CRS and neurotoxicity across departments and between commercial products and clinical trials. However, different products and disease indications have different rates and patterns of toxicity, and clinical judgement must be used in applying guidelines. Tisagenlecleucel induced severe CRS (Grade 3 or 4 by the UPenn toxicity criteria) in 48% of pediatric and young adult ALL patients in ELIANA²⁴ and 28% of adult DLBCL patients in JULIET²⁵, while ZUMA-1 reported a 13% rate (by NCI Common Terminology Criteria for Adverse Events) for axicabtagene ciloleucel in adult DLBCL.

CRS correlates with disease burden in pediatric trials⁴⁶. Similarly, in adult ALL trials, severe CRS has only been observed in patients with morphologic evidence of disease at time of T cell infusion (> 5% bone marrow blasts) ^13,47. Therefore, patients with a low burden of disease but a high risk of relapse may be optimal candidates for CAR therapy.

Nurses and ancillary medical personnel must also be educated about the recognition of CAR T cell specific care and toxicities, including CRS and neurotoxicity. Nursing units with experience in managing leukemia, lymphoma, and transplant patients are also well suited to handling more common issues including infection⁴⁸ and transfusion. All physicians, mid-level providers, pharmacists and nurses who will interact with CAR T cell patients must undergo FDA mandated training as part of a Risk Evaluation Mitigation Strategy (REMS). This includes training which is provided by manufacturers on CAR-specific toxicity management with a knowledge assessment and the ability to audit these assessments if requested by the FDA as part of an adverse event assessment.

By FDA mandate, the inpatient pharmacy will need to stock at least two doses of Tocilizumab for each patient prior to CAR T cell administration, and ordering protocols and order sets will need to be developed to assure timely administration of both anti-IL6 therapy as well as corticosteroids and supportive medications as needed for CRS.

Each institution must undertake a thoughtful assessment of anticipated patient volumes, which depend on referral rates, relapse and treatment patterns, average length of stay, and relapse and complication rates. Although neither of the two products currently approved require an up-front hospitalization, both carry a significant risk of neutropenic fever, CRS, and neurotoxicity, resulting in an average a length-of-stay of 10-14 days. The need for inpatient admission, intensive care, and subspecialty care (such as neurology consultation) needs to be anticipated and requires planning by multiple departments.

Outpatient administration and management of patients undergoing CAR T cell therapy is feasible but requires well-staffed, dedicated outpatient units like those used to monitor outpatient autologous HCT recipients. Older patients, patients with bulky disease and patients with more comorbidities are likely to be at higher risk for complications and thus closer monitoring will be reasonable. Since a large proportion of patients will be coming from a significant distance, provisions must be made for lodging, transportation and living expenses. Patients and caregivers must be educated on CAR T toxicities and are given a manufacturer-provided wallet card describing symptoms that warrant immediate evaluation, such as fevers which can herald the onset of CRS or infection.

Task 7, Post-Infusion Care (Day 30 Onwards):

Documentation and hand-off procedures must be established to transition successfully treated patients to the outpatient setting. Order sets, standards of practice, and documentation must be developed for this purpose. The role of CAR T cell consultation and disease-specific groups must be delineated in patient monitoring and follow up. As discussed below, patients must be assessed for disease response at one month.

The use of CAR T cells in R/R B-ALL raises the question if subsequent allogeneic stem cell transplant is needed for durable remission. The University of Pennsylvania experience with adult and pediatric ALL showed long-term persistence of CAR cells, and only 9% patients went on to receive all-HSCT¹⁵. Similarly, the recently published MSKCC experience did not demonstrate a survival difference among patients with MRD negative complete response after CAR therapy who underwent transplant compared to those who did not¹³. Prospective randomized control trials comparing a “bridge-to-transplant” compared to CAR-only approach are not available but will be quite informative as this question is key to the successful use of this technology.

When patients do relapse, it is often due to CD19 undetectable disease with detectable CAR T cells in the peripheral blood⁴⁶. Lymphoid cells can also express CD19 isoforms that do not contain CAR binding domains⁴⁹. In either case, re-treatment with CD19 CAR would not have clinical activity; however, alternative targets such as CD20⁵⁰ and CD22⁵¹,⁵² are in development.

Task 8, Financing, Regulatory and Reporting Requirements:

Administrative personnel will need to be educated on the billing and reimbursement protocols for CAR therapy. Novartis has announced a unique payment paradigm for Kymriah, which carries a nearly half-million price tag contingent upon response at one month⁵³. CRs are thankfully quite high, however, suggesting hospitals will face the full costs for at least 4/5^th of patients. Yescarta is priced at $375,000 per patient. Notably, these costs account only for the therapy itself, and do not account for hospitalization and numerous ancillary processes required for CAR T cell therapy (“parts, not labor”). Extrapolating from experience with bone marrow transplantation, management of patient hospitalization and complications is likely to contribute at least another $200,000-$500,000 depending on complication rates ⁵⁴, which brings the total price tag close to one million dollars. Private payors will likely announce strict, label-specific requirements, while public payors will need to create new Diagnosis Related Groups and other funding mechanisms to accommodate cost. Regardless, institutions must be prepared to face significant downside financial risk.

From a regulatory perspective, gene therapy that is deemed “high risk,” including the use of integrating vectors, carries an FDA requirement to track patients for up to 15 years after administration. This requirement stems from potential oncogenic transformation that has been seen in several patients infused with genetically modified hematopoietic stem cells used for the treatment of primary immune deficiency⁵⁵. However, oncogenic transformation in mature T cells has not been observed^56,57.

Data management personnel are needed for obligatory data reporting on outcomes, particularly severe adverse events. FACT recommends that elements of CIBMTR cellular therapy data forms, including patient demographics and outcomes, are captured at regular follow-up intervals. Manufacturers have established reporting requirements for CAR-related toxicities: all incidents of CRS and neurotoxicity must be reported, and manufacturers also encourage reporting of severe and unusual toxicities. Events may be reported directly to the FDA via MedWatch or to the manufacturers.

Conclusion

Built on decades of scientific advancement and clinical development, CAR T cells are a new treatment paradigm for previously intractable malignancies. While CAR T cells were initially developed in a few academic centers, they have been approved at a time where no center has treated more than 400 patients. Each institution that initiates a CAR therapy program will face both clinical and administrative challenges, and the next decade will no doubt bring significant knowledge of the optimal use and management of these powerful tools. Despite significant financial, regulatory, and administrative barriers, we must not lose sight of the tremendous benefit that CAR T cells, skillfully deployed and thoughtfully applied, can bring to our patients.

Highlights:

• Commercial Chimeric Antigen Receptor T cell products are efficacious, but the delivery of cell products requires significant planning and coordination

• A consultation service composed CAR-certified physicians identifies and manages patients

• Apheresis requires significant coordination with manufacturers, including site inspection and sample tracking

• Toxicity management requires REMs training for all providers, and participation of ICU and neurology specialists

• Regulatory standards are provided by the FDA, CIBMTR, FACT and product manufacturers