2017/04/20 by Alexey Bersenev
Engineering · Medicine · #Biomedical and Engineering Education #CAR-T cell therapy research
paper · pdf · doi:10.1111/trf.14110
openalex publication_date 2017/04/20 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/30
Cell therapy is entering an exciting period as multiple cellular immunotherapies are rapidly progressing through different stages of clinical development. Publication of the first clinical cases, describing complete remission of chronic lymphocytic leukemia (CLL) after administration of chimeric antigen receptor (CAR)-modified T cells by a team of researchers from the University of Pennsylvania1 was a huge step forward in the field of cell therapy. This achievement was followed by a commercial licensing arrangement with Novartis in 2012.2 Since that time, CARs have become the most frequent genetic modification of therapeutic immune cells. More than 200 CAR cell clinical trials have been launched worldwide in the past 5 years. Promising results from multicenter international clinical trials in pediatric acute lymphoblastic leukemia (ALL) and non-Hodgkin's lymphoma (NHL) were reported by Novartis3 and Kite Pharma4 at the Annual American Society of Hematology (ASH) meeting in December 2016. Based on these activities, the field is anticipating Food and Drug Administration regulatory approvals of commercial CAR cell–based therapeutics sometime this year. Rapid development of CAR cell therapies has also revealed multiple challenging problems. In addition to important clinical challenges (i.e., relapse rates, cytokine release syndrome, and neurotoxicity), there are many issues related to collection of starting material, manufacturing of a consistent product, and delivery of centrally prepared cellular products to distant clinical sites. This editorial will focus on collection and related manufacturing challenges that should be of particular concern to the TRANSFUSION readership. Nearly all CAR cell trials currently utilize collection of mononuclear apheresis products as the starting material for manufacture. The goal of mononuclear cell (MNC) apheresis collection for manufacturing of a T-cell product is to reduce unwanted cell populations (such as red blood cells, platelets, and granulocytes). Because almost all CAR cell clinical trials currently involve autologous cells, the quality of the apheresis collection is a critical source of variability in T-cell product manufacturing. Unfortunately, enrichment for MNCs does not remove circulating malignant cells, so that underlying disease can significantly affect the quality of collected products. In some cases, collection of an insufficient number of benign T cells can lead to failure to meet the targeted therapeutic dose. In CLL, for example, the MNC collection may consist of more than 90% to 95% leukemic blasts, creating challenges for manufacturing CAR-T cell products. Reported CAR-T cell manufacturing failure rates (mostly related to inability to meet targeted dose) range from as low as 1% to 5%3, 4 to 7% in CLL5 and 14% in lymphomas.6 It is very important to acknowledge and understand patient-to-patient variability of apheresis collections as the starting material for CAR-cell product manufacturing. Each patient has a unique disease history, including stage, type, duration, pretreatment issues, comorbidities, and status of their immune system. Other factors that can contribute to variability may include differences in apheresis machines, operators, and institutional policies for scheduling and conducting procedures. Center-to-center standardization of collection procedures (devices and software protocols) as well as protocols for postcollection manufacturing can be implemented to decrease such variabilities, but will not completely prevent them. In this issue of TRANSFUSION, the Department of Transfusion Medicine at the NIH Clinical Center (Allen et al.7) investigated how the degree of variability in starting apheresis material affects the feasibility of manufacturing an adequate number of CAR-T cells. The study analyzes autologous apheresis collections from 71 patients, enrolled in three different CAR-T cell therapy clinical trials, including those with hematologic malignancies and solid tumors. The authors show that even when the apheresis collections exceeded the goal of 2 × 109 CD3+ cells, 5% of CAR-T cell products did not meet the targeted therapeutic dose (0.3 × 106-3.0 × 106/kg). If fewer than 2 × 109 CD3+ cells were collected, 31% of CAR-T products did not meet the therapeutic goal. Low T-cell number in the starting material could be detrimental to manufacture because specific seeding concentrations are required for successful expansion of T cells. Possible loss of T cells during the multiple initial processing steps before cell culture may also explain the failure to achieve the desired targets. Another important finding reported in the study of Allen and colleagues7 is that the concentration of circulating T cells can predict the T-cell yield collected by apheresis. A similar approach has been successfully used in clinical hematopoietic stem cell transplantation field for prediction of CD34+ cell yield in apheresis after mobilization. For consistency and predictability of collection goal and manufacturing, investigators may need to set a T-cell threshold (measured as CD3+ cell number by flow cytometry or, as a surrogate marker, the absolute lymphocyte count [ALC]). Some centers now are starting to include an ALC threshold as part of the inclusion criteria for enrollment into their CAR-T cell trials. For example, Seattle Children's Hospital requires an ALC of at least 100 cells/µL for inclusion in the NCT02028455 trial. However, it is important to understand that setting a CD3+ cell collection goal could differ from center to center, because there are multiple approaches to CAR-T product manufacturing and different potential target therapeutic doses. One limitation of the study of Allen and colleagues7 is that the collections were performed exclusively using the COBE Spectra, and the manufacturer will discontinue support of these devices at the end of this year. Therefore, it will be important to validate the published data on other apheresis devices. Overall, the study by Allen and colleagues sets an example of how a cellular immunotherapy center can evaluate their collection procedures and its potential impact on the subsequent manufacture of T-cell products. Importantly, the authors give practical recommendations for 1) using preapheresis circulating CD3+ cells to predict collection outcome, 2) setting CD3+ cell collection targets, and 3) the volume of blood to process. Besides T-cell quantity, another essential characteristic of collected apheresis product is cell composition. Contaminating cells strongly impact the manufacturing process by: 1) creating additional difficulties for T-cell purification and other processing steps, 2) binding to viral vectors, and 3) inhibiting T-cell growth. For example, myeloid cell contamination of apheresis products can inhibit CAR-T cell expansion,8 and monocyte depletion can help to achieve targeted CAR-T cell doses from apheresis products that may have previously led to manufacturing failures. Thus, development of optimized processing pathways, depending on the cellular composition of incoming apheresis collections, may significantly improve success rates for CAR-T cell manufacture. Few research groups have attempted to correlate the cellular composition of apheresis products and the subsequent success of manufacturing or experimental treatment with CAR-T cell products. Researchers from the Children's Hospital of Philadelphia showed9 that the presence of “early lineage” T cells (such as stem cell memory and naïve T cells) at collection directly correlates with success of T-cell expansion ex vivo. Patients with lower “early lineage” T cells in the blood (and thus in the apheresis products collected) were less likely to achieve preset goals for initial T-cell expansion. Cells collected from 24% of patients (20/83) did not reach this “test expansion” threshold and these patients were excluded from the CD19-CAR-T cell pediatric B-ALL trial. Another group from the University of Pennsylvania recently demonstrated10 a significant correlation between frequency of the CD27+CD45RO– subset in the CD8+ T-cells and durable complete responses in CLL patients treated with CD19-CAR T cells. Intriguingly, researchers from the National Cancer Institute and Kite Pharma did not find a correlation11 between ALC before collection, T-cell subset composition in the apheresis product, and CAR-T cell product manufacturability with clinical responses in NHL patients. These findings highlight the potential impact of apheresis collection on the success of CAR-T cell product manufacturing. We should acknowledge that variability and poor predictability of the quality of the apheresis product used as incoming raw material for CAR-T cell production is a major external factor, which could lead to manufacturing failure. A few recommendations for optimization of apheresis collections and mitigation of related CAR-T cell therapy manufacturing failures are offered. Researchers working in the field of CAR-T cells should collect and share data on the correlation of apheresis collections, manufacturability, and clinical outcomes via meeting abstracts and publications. Technology for developing driverless cars is well on its way to becoming a reality. For CAR-T cell technology, however, hands-on driver input is still required and there is a long and somewhat bumpy road ahead. The author has disclosed no conflicts of interest. Alexey Bersenev, MD, PhD e-mail: [email protected] Advanced Cell Therapy Laboratory Yale-New Haven Hospital Yale University New Haven, CT