Get Off on the Right Foot: How to Plan an Efficient Leukocytapheresis to Collect T Cells for CAR T-Cell Manufacturing

Abstract

Background

The major drug regulatory agencies have approved chimeric antigen receptor (CAR) T cells for the treatment of some B-cell lymphoproliferative diseases. Their use is expanding, and new indications will be approved. Efficient mononuclear cell collection by apheresis providing enough T cells is a critical step in further CAR T-cell manufacturing process. It is important that apheresis units are prepared for the collection of the required T cells for manufacturing with the highest efficiency and safety for the patient.

Summary

Several series have studied different characteristics that could influence the collection efficiency of T cells for CAR T-cell manufacturing. Also, an effort has been made to identify predictors of the total number of target cells collected. Despite these publications and the large number of ongoing clinical trials, consensus protocols in apheresis are scarce.

Key Messages

The aim of this review was to summarize the set of measures described to optimize apheresis and ensure patient safety. Moreover, we also propose, in a practical approach, a way to apply this knowledge to the daily routine in the apheresis unit.

Introduction

Adoptive cell therapy is a type of immunotherapy in which immune effector cells, such as T or NK cells, are given to the patient to help the body fight diseases. Sometimes, T cells are modified to express high-affinity natural T-cell receptors or antibody-like receptors to selected antigens using gene transfer technology. The latter are synthetic proteins consisting of a fusion of different proteins known as chimeric receptors. Therefore, second-generation chimeric antigen receptor (CAR) T cells are T cells expressing a CAR, a molecule with an extracellular domain composed of a single-chain variable fragment of immunoglobulin, which recognizes the “tumor” antigen, an intracellular domain composed of the intracellular domain of the zeta chain of the CD3 component of the T-cell receptor and a costimulatory domain (CD28 or 41BB), and a transmembrane region that links the two extracellular and intracellular domains through the cell membrane [1].

Autologous CAR T-cell therapies have been shown to be effective in the treatment of various recurrent or refractory hematologic malignancies, such as multiple myeloma (MM), non-Hodgkin lymphoma (NHL), acute lymphoblastic leukemia (ALL), and chronic lymphocytic leukemia (CLL). Additionally, other CAR T cells designed against antigen targets for other malignancies are currently being investigated in preclinical and early clinical studies. This increase in therapeutic indications has meant that a high number of hospital centers were authorized to apply these therapeutic strategies. Therefore, apheresis units must be prepared because the collection of T cells from patient's peripheral blood using apheresis devices for autologous production of CAR T cells is the first step in the manufacturing process. This first step is critical because high-quality starting material impacts the CAR T-cell manufacturing process [2, 3].

At this point, two issues deserve to be pointed out. First, from a regulatory point of view, collection of cells is usually performed in hospitals or blood banks where the obtained product follows Recommendation No. R (95) 15 of the Committee of Ministers to member states on the preparation, use, and quality assurance of blood components [4]. However, when the product is obtained, the hospital or blood bank acts as a service provider to industry where CAR T-cell therapies fall under the advanced therapy medicinal products framework [5]. In addition, hospital exemption is a regulation created for exempting those advanced therapy medicinal products manufactured in hospitals or universities where the medicine is prescribed for individual patients under the care of a medical practitioner [6]. Finally, the FACT-JACIE accreditation system covers a wide range of important aspects that can be of use for centers that aim to be accredited to provide CAR T-cell therapy [7]. Second, there are several challenges when performing apheresis to collect lymphocytes in patients for CAR T-cell manufacturing. Although physicians working in an apheresis unit are used to collecting lymphocytes from non-mobilized healthy donors in order to treat patients who relapsed after allogeneic hematopoietic cell transplantation [8], several caveats must be taken into consideration when collecting T cells from non-mobilized patients: difficulties in establishing the buffy coat (BC) due to low white blood cell count in the peripheral blood count (common in patients with refractory NHL and ALL), the clinical status of the patients, and treatments used before the collection [3, 9].

Believe it or not, although obtaining T cells from patients is the first step in this complex process, little attention has been paid to this issue, and no guidelines or recommendations have been made until recently [10, 11, 12, 13, 14, 15]. In fact, a detailed leukocytapheresis collection manual, with all requirements of the collection process, is not available in some protocols [16]. Table 1 presents data regarding different requirements for cell collection as starting material for CAR T-cell therapy [16]. The aim of the present review was to give readers with a list of criteria that are necessary to include in a collection manual before starting leukocytapheresis as well as a list of considerations for collection parameters in order to customize the apheresis procedures to adult donors and to collect cells in an efficient and safe way.

Table 1.

Cell collection requirements in different leukocytapheresis collection manuals for CART cells

Product	Axicabtagene ciloleucel	Brexucabtagene autoleucel	Tisagenlecleucel	Lisocabtagene maraleucel	Idecabtagene vicleucel	Ciltacabtagene autoleucel
Registered name	Yescarta®	Tecartus®	Kymriah®	Breyanzi®	Abecma®	Carvykti®
Manufacturer	Gilead	Gilead	Novartis	Juno-Celgene-BMS	Bluebird bio-Celgene-BMS	Legend Therapeutics-Janssen
Targeted antigen	CD19	CD19	CD19	CD19	BCMA	BCMA
Approval year	2017	2017	2020	2021	2021	2022
Cell dose target to collect	5–10 × 109 MNCsa	5–10 × 109 MNCs	1–4 × 109 CD3+ cells ≥2 × 109 TNCsb ≥3% CD3+ of TNCs	Not stated. A collection bag with 450 mL is required.	Not stated	Not stated
Blood volume to process	12–15 L	12–15 L	6–10 L	7 L if lymphocytes in PBc are ≥1,000/µL 12 L if lymphocytes in PB are <1,000/µL	Not stated	Not stated

^aMNCs, mononuclear cells.

^bTNCs, total nucleated cells.

^cPB, peripheral blood.

Factors Affecting Mononuclear Cell Yield

As stated before, detailed information about how to proceed to collect cells as a raw material for CAR T-cell manufacturing is not available in some protocols [16]. At least, the desired number of cells to be collected must be specified in the protocol [17]. However, some protocols specify a range of whole blood to be processed. In our opinion, this lack of information is the reason that described approaches in apheresis units suggest variability [15]. For some reason or another, the main goal of performing an apheresis procedure is because the collection of a minimum number of a target cell is desired. Therefore, this desired number of cells must be stated in the protocol in order to plan an efficient collection of adequate cellular products for CAR T-cell manufacturing. If the collection target is specified, either a minimum or an optimal cell dose, the apheresis unit team will be able to determine collection feasibility with some basic calculations.

However, it is important to note that with current apheresis devices, it is possible to pull in peripheral blood and to separate it by density centrifugation or by elutriation into its major components (plasma, BC, and red blood cells [RBCs]). In this case, the apheresis device is used to selectively separate and siphon off the BC and return back the plasma and the RBCs [3]. Because of limitations of either density centrifugation or elutriation technology, the collected BC contains not only the desired target cells but also other mononuclear cells (MNCs) as well as platelets and some RBCs. In summary, the purity of the collected product is limited [3, 11, 18].

Blood Cell Count in Peripheral Blood before Starting Cell Collection

It seems logical that the greater the number of cells in peripheral blood, the greater the number of collected cells [3, 19]. This direct relationship has been corroborated in studies with healthy donors [20]. More importantly, this correlation has also been proven in series of patients diagnosed with NHL, ALL, MM, CLL, and solid neoplasms [12, 21, 22, 23, 24, 25, 26]. While the power of the correlation between CD3+ lymphocyte count in peripheral blood and the final target cell yield was high (correlation coefficient r = 0.57–0.86) [21, 25, 26], there is no established minimum peripheral blood lymphocyte count before starting the collection of T cells [13]. Published series have described successful collections even with cell counts as low as 160–330 lymphocytes/µL [13, 21, 22, 25, 27, 28, 29]. Although these data demonstrate that with low pre-apheresis cell counts, collection targets can be obtained, it is also true that low pre-apheresis cell counts have been associated with higher collection failure rates [21, 24, 25, 28].

Total Blood Volume Processed

Total blood volume (TBV) is an important parameter to take into account because it is something that we can manipulate in order to customize the MNC collection procedure. It also seems logical that the greater the TBV processed, the greater the number of collected target cells. However, we must adjust the TBV processed to the minimum for collecting the required target cells for several reasons. First, the higher the TBV processed, the longer the apheresis procedure. Second, the longer the apheresis procedure, the higher the risk for donors to experience adverse events, particularly toxicity due to citrate anticoagulation [30]. The TBV processed in published series on apheresis for CAR T-cell manufacturing ranges from a median of 4.4–11.6 L [12, 21, 22, 24, 25, 26, 27, 28, 29]. We will show how to adjust the TBV processed by doing simple calculations.

Collection Efficiency

Cell collection efficiency (CE), often calculated as a quality metric, refers to how many cells of interest are actually collected in the product bag relative to the number of processed cells that pass through the apheresis device [31]. Therefore, the formula used to calculate CE is a ratio where the numerator is the yield of cells collected in the bag and the denominator represents the number of cells processed throughout the apheresis procedure. Because CE is expressed in %, the ratio is multiplied by 100, and CE = 0% means that no cells were collected during the procedure, while CE = 100% means that all processed cells were collected.

The numerator of this formula can easily be obtained by performing a cell count with a representative sample of the collected bag. In contrast, in order to calculate the denominator, it is necessary to know the cell count in peripheral blood of the donor as well as the whole blood processed (note that the volume of citrate must not be added). If we use only the cell count in peripheral blood obtained just before starting the apheresis procedure, we can calculate the CE2(%). However, if we use the cell count in peripheral blood obtained just before starting and just after finishing the apheresis procedure, we can calculate the CE1(%) [32].

By doing this additional work of taking into account not only the pre-apheresis peripheral blood count but also the post-apheresis cell count, we are calculating a more accurate value for CE. CE1 is more accurate than CE2 because CE1 formula describes the mean of cells present during the apheresis procedure. In other words, CE1 formula reflects an approximation to the real conditions during the apheresis procedure [33]. If we only used the pre-apheresis cell count, we would be assuming that the cell count was constant throughout the entire apheresis procedure. In contrast, using pre-apheresis and post-apheresis cell counts, we are taking into account if the cell count is increasing or decreasing throughout the apheresis procedure. Obviously, if we had obtained more cell counts at different time points during the apheresis procedure, we would have done a more realistic approach to the real conditions during the apheresis procedure. However, because of practical reasons as well as economic constraints, it is common to perform just a pre-apheresis cell count. We recommend, however, obtaining pre-apheresis and post-apheresis cell counts in order to calculate the CE1 of the apheresis procedure [34].

High variability in cell CE is reported in published studies. In some studies, authors reported mean CE2 of CD3+ cells of 18–68% [21, 23, 24, 25, 26, 29]. Other authors reported CE1 or CE2 of lymphocytes between 43 and 83% [22, 28]. Nevertheless, the vast majority of studies describe an average CE of lymphocytes/CD3+ cells of around 55% [21, 22, 24, 25, 26, 27, 29].

Several authors studied different variables that could influence CE in non-mobilized patients. Neither gender nor weight influenced CE [12, 21, 22, 29, 31]. Age was negatively correlated with leukocyte CE1 in a published series of adult patients [22, 31]. However, in a series including young patients and children, this correlation with age was not observed [12, 23, 29]. According to diagnosis, no differences were found among patients diagnosed with ALL, MM, CLL, NHL, solid tumor, and healthy donors [12, 28, 29]. One study, however, identified lower lymphocyte CE1 in patients with ALL when compared with patients with CLL and NHL [22]. History of bone marrow transplantation and time between the last treatment and apheresis procedure were analyzed without finding any association with CE [22, 23]. This last information should be taken with caution because patients included in the study had completed pharmacological washout times. No differences were identified in terms of vascular access [12, 22], except in one study that described a higher CE1 of MNCs in patients when using peripheral veins in comparison to the central line [35]. Finally, different variables related to the used device have been evaluated. In healthy non-mobilized donors, Amicus device demonstrated higher CE1 of CD3+ cells and lower CE1 of platelets than Optia device using the cMNC protocol [35, 36]. In the case of Optia, cMNC protocol showed higher CE1 of platelets than MNC protocol [37]. In a pediatric series, Optia MNC protocol showed higher CE2 of MNCs than COBE Spectra device [27]. In contrast, other authors did not find a relationship between device types (Optia cMNC, Optia MNC, and Spectra) and CE of MNCs or CD3+ cells [22, 24, 37, 38].

Patient Safety

Patients with an indication for CAR T-cell therapy have heavily been multitreated or have recently received antitumor treatment. Moreover, patients have a high risk of infection, as well as they are very sensitive to biochemical alterations, and peripheral venous access is difficult. In summary, we must consider them as fragile patients [39], which is a factor that may increase the opportunity for adverse events during the apheresis procedure. Different published series have reported minor complication rates of up to 9.8%, such as dizziness, hypotension, paresthesia, fever, pain, or agitation [21, 28]. Although severe complications are extremely rare, rates of 0–1.5% have also been reported [21, 22, 23, 24, 26, 28].

In our opinion, the low complication rates were due to an accurate patient selection and good apheresis team practices [40]. People involved in care of patients receiving apheresis procedures may anticipate potential adverse events, and they are prepared to respond to unexpected events. The use of eligibility clinical and laboratory criteria is helpful for this purpose [9, 28]. A Francophone de Greffe de Moelle et de Thérapie Cellulaire Society consensus guide recommends the use of a checklist that includes a performance status of ECOG <2, platelet count >30 × 10⁹/L (adults), hematocrit >24% (adults), and neutrophils >1 × 10⁹/L [41]. In our apheresis unit, we routinely use further preventive measures not only for collection of MNCs in this setting but also in all other apheresis procedures [34, 42]. These preventive measures in current use in our apheresis unit are a careful review of current medications, such as angiotensin-converting enzyme inhibitors [43], β-blockers, and anticoagulant medications [44] as well as prophylactic administration of magnesium sulfate and calcium chloride during the apheresis procedure [30].

How Do We Implement This Knowledge in Clinical Practice?

Apheresis units must be prepared to assume the workload of increasing cellular therapy protocols in daily practice. Although the majority of cellular therapy products are in development, they are increasing with rapid growth trajectories. Therefore, it is important to be efficient in order to collect enough cells for further cell processing. Some authors have suggested interesting target cell yield prediction models, such as calculation of CE2 together with modifying factors [29] and real-time monitoring of collected CD3+ cells [25]. These models, however, have several problems. As the authors pointed out, they lack external validation, which implies a high risk of over-fitting, and the sample size used by subgroups for model calculation was small. Estimating the volume to be processed based on the calculation of the CE is usually the most widely used and, therefore, the most validated approximation [24, 26, 28, 30]. We detail, in the following two sections, current workload in our institution when a patient is admitted to collect MNCs as a starting material for CAR T-cell manufacturing [16].

Pre-Apheresis Visit

The pre-apheresis visit, carried out by an apheresis doctor and an apheresis nurse, will allow us to prevent possible problems and to coordinate with the medical team to determine the date of apheresis. This point is crucial to reduce as much as possible the time between decision of CAR T-cell therapy and its final administration. During the pre-apheresis visit, we, as other authors do, recommend a checklist [16].

Vascular access is an important aspect to take into account in this pre-apheresis visit because the collection of enough MNCs for cellular therapy protocols can be achieved with one procedure [45]. According to different published series, the number of studies with a preference for the peripheral venous access [12, 22, 23, 24, 35] is greater than for the placement of a central line [21, 25, 29]. Our recommendation is the use of the peripheral access as much as possible because there are no differences in the CE, its placement does not depend on other services, and it is safer. To achieve this goal, implementation of guided ultrasound and nursing expertise in the apheresis units is critical [46].

During Apheresis Procedure

The day of MNC collection arrives and we have to design the apheresis procedure. First, we have to select the apheresis platform to use. The most common apheresis device is the Spectra Optia Terumo BCT (Lakewood, CO, USA), but Amicus (Fresenius Kabi, Bad Homburg, Germany) is also available in some apheresis units [16, 35]. Second, in our experience, anticoagulation with citrate is efficacious and safe. Third, a revision of laboratory parameters including a complete blood count, acid-base balance, and biochemistry parameters (Na, K, Ca, Mg). We also perform a CD3+ cell count with flow cytometry. To estimate the volume of whole blood to process to achieve the collection of the desired amount of cells, the CE2 formula can be used.

Therefore, it means that we have to know the CE2 of our previous apheresis procedures. We recommend calculating the CE2, or even better the CE1, of every apheresis collection in order to use it in future collections [9]. By doing this exercise, we can customize the apheresis procedure for our patients, and we can adjust the whole blood volume to process. In our apheresis units, we use the CE1 because we perform pre-apheresis and post-apheresis cell counts [35]. In our hands, after analyzing 1,071 apheresis procedures in 249 non-mobilized donors, the median CE1(%) of lymphocytes, monocytes, and MNCs were 39.7% (IQR: 19.9–59.0), 52.3% (IQR: 30.1–80.9), and 45.8 (IQR: 28.6–64.7), respectively [35]. More recently, we implemented the CD3+ cell count using flow cytometry from a blood sample obtained just before starting and just after finishing the apheresis procedure as well as the CD3+ cell count in the collected bag. With this approach, we could calculate the CE1 to use in future collection (a manuscript is under revision). These data help us calculate the amount of whole blood to process. Imagine that our donor has 1.0 × 10⁹/L MNCs, and the goal of the apheresis procedure is to collect 7 × 10⁹ MNCs.

In this hypothetical case, the amount of whole blood to process is 15.2 L. However, if the donor had had 5.5 × 10⁹/L MNCs, the amount of whole blood to process would have been 2.8 L.

Conclusions

People working in apheresis units are used to collecting hematopoietic cells from healthy mobilized donors in the past. However, today, the number of collections of cells from non-mobilized patients as a raw material for CAR T-cell manufacturing is increasingly challenging. The challenges arise not only from limitations of the current technology and the device used but also from patient characteristics as well as patient variables, such as low number of white blood cells in peripheral blood, the use of concomitant medication, whole blood processed, and CE. An evaluation of patients performed before performing the apheresis procedure is mandatory to check not only eligibility and suitability but also to assess venous access and to review laboratory tests. The knowledge of all these critical steps in advance helps us get off on the right foot because we can customize the apheresis procedure to minimize these limitations, and we can plan an efficient leukocytapheresis procedure to collect cells for CAR T-cell manufacturing.

Conflict of Interest Statement

Juan A. Piñeyora has no conflicts of interest to declare. Joan Cid received research funding from Cerus, Kawasumi Laboratories, and Sanofi; he also received speaker or advisory fees from Cerus, Fresenius Kabi, Grifols, Macopharma, Pharm-Olam, Sanofi, and Terumo Blood and Cell Technologies. Miquel Lozano has received research support in name of the Clinic Research Foundation from Terumo BCT, Sanofi, and Macopharma. He has received speaker or advisory fees from Grifols, Terumo BCT, and Fresenius Kabi.

Funding Sources

No funding was received for the preparation of this manuscript.

Author Contributions

JAP and JC performed the research and wrote the first draft of the manuscript; JC designed the research study; and JC and ML reviewed and edited the manuscript.

Funding Statement

No funding was received for the preparation of this manuscript.