Systematic Automation of an Autologous Cellular Immunotherapy Biomanufacturing Process 

Systematic Automation of an Autologous Cellular Immunotherapy Biomanufacturing Process 

As more and more approvals of immuno-oncology therapies become reality, how will these therapies be efficiently and sustainably scaled up and commercialised for long-term success?

The first to market will probably be Novartis, who ‘set the stage’ for approval of CTL019 therapeutic with revealing new data from its phase II, ELIANA trial at ASH. The trial found that 82% of patients (41 out of 50 patients) achieved complete remission or complete remission with incomplete blood count recovery at three months after an infusion of CTL019 in young relapsed/refractory patients with B-cell late-stage acute lymphoblastic leukaemia (ALL). The Swiss company is planning to file for approval this spring with the FDA.

Hot on their heels is Kite Pharma, who are starting a rolling submission with the FDA for a BLA of its KTE-C19 med, now known as axicabtagene ciloleucel, as a treatment for patients with relapsed/refractory aggressive B-cell non-Hodgkin lymphoma who are ineligible for autologous stem cell transplant.


Systematic automation of an autologous cellular immunotherapy biomanufacturing process

Current biomanufacturing strategies have limited amenability to commercial-scale biomanufacturing scale-up/scale-out. As such, these processes represent a major barrier to industry development that could be alleviated by innovation in bioprocess automation and downstream processing technologies, which can be achieved using quality by design (QbD), as well as in cell population purification and enrichment.

A robust automation system removes the human element from bioprocessing and enhances consistency for both product and process. By limiting the human-operator interactions with the product stream you are ensuring higher sterility and reducing opportunities for contamination, whilst also lessening the potential for errors and expanding production hours beyond attended operation. Quality and consistency of products will lead to fewer out-of-specification results and product losses. Controlled and reproducible handling, plus reduced labour needs, should also confer savings on the cost-of-goods (CoGs).

Brindley et al. [April 2016] suggested that to successfully automate this complex process, a stepped approach should be used (as shown in Figure 1 below):


Figure 1. A stepped approach to bioprocessing automation

Developers should break down the process into separate, sequential steps that include a single function called a ‘unit operation’. The development team should designate appropriate equipment for those unit operations, with the connections between them determining whether linkages are affordable relative to their benefits or demand unjustified expenditures. Taking into account unit operations, components, connections, and timelines, this will ensure that a company ends up with the most robust and affordable iTx manufacturing system with the greatest equipment use rates.


Opportunities to optimise immunotherapy manufacturing 

Cell purification and enrichment

The protocol of producing CAR T cells requires several, carefully monitored steps. Firstly, using leukapheresis, blood is removed from the patient’s body, and the leukocytes are separated (with the remainder of the blood returning to circulation). After a sufficient number of leukocytes have been harvested, the leukapheresis product is enriched for T cells (Figure 2). This process involves washing the cells out of the leukapheresis buffer, which contains anticoagulants (Levine et al., 2017). 

Figure 2. After a sufficient number of leukocytes have been harvested from the patient’s blood via leukapheresis, the anticoagulants in the leukapheresis buffer are washed out of the product and the cells are concentrated by counterflow centrifugal elutriation, which separates cells by size and density. [Levine et al., 2017]

Table 1. Cell purification methods in widespread use [data sourced from Brindley, et al. 2016].
Cell purification method Properties for separation Limitations
       Physical separation Cell size and density, typically using techniques such as density and counterflow centrifugation

Commonly used for pre-purification and cell concentration but have low specificity because they separate cells based on sedimentation properties, not by characteristic features or expression of specific markers; i.e. centrifugation cannot separate cells with similar sedimentation properties.

Surface-marker-based separation Using capture ligands via techniques such as magnetic cell separation (MACS) or fluorescence-activated cell sorting (FACS).

Currently dominated by MACS approaches but this leaves large numbers of paramagnetic particles bound to separated cells, which is highly undesirable for cell-based therapeutic applications. Minimizing the number of attached or internalized beads represents a formidable obstacle here because most products based on magnetic beads lack the ability to readily release bound cells from capture molecules without altering their viability and phenotype.

Additionally, the resulting isolated cells require a lengthy culture process to reduce magnetic-bead concentrations to meet clinical release criteria for patients receiving cell-based therapeutic infusions.


Purification and enrichment of specific cell subpopulations from heterogeneous starting populations is a critical component of multiple biologics workflows. The key objective is to obtain a highly purified and viable cell population for downstream applications.

Purifying autologous antigen-presenting cells (APCs) from the patient is a labour intensive process and for this reason, standardised, efficient methods have been pioneered. The use of anti-CD3 antibodies alone or in combination with feeder cells and growth factors, such as IL-2, has been the practice to isolate T cells for many years. Life Technologies introduced beads coated with anti-CD3/anti-CD28 monoclonal antibodies or cell-based artificial APCs (aAPCs; derived from the chronic myelogenous leukaemia cell line K562, which can be engineered to express the required costimulatory ligands), to optimise this process, with these agents being easily removed from the culture through magnetic separation.

T cells can grow logarithmically in a perfusion bioreactor for several weeks, in the presence of interleukin-2 and aAPCs. Culture conditions can be further refined to polarise T cells to a specific phenotype (i.e., Th2 or Th17) during expansion and this can result in increased efficacy in the clinic [Levine et al, 2017].

Additionally, Brindley et al. (2016) recommended the use of hydrogel technology to overcome some of the limitations of current cell separation methods. A biocompatible phase-change hydrogel can be easily functionalized with cell-capture agents such as antibodies (coated on different substrates such as paramagnetic beads, polymethyl methacrylate (PMMA) acrylic, polydimethylsiloxane (PDMS) silicone, polystyrene), and it can be dissolved rapidly in a buffer containing a chelating agent. One of these platforms has been designed by Quad Technologies and enables liquid-phase cell separation, overcoming many of the challenges listed in the table above.


Activation process

The next phase involves T cells being incubated with the viral vector encoding the CAR, and, after several days, the vector is washed out of the culture by dilution and/or medium exchange.The viral vector uses viral machinery to attach to the patient cells, and, upon entry into the cells, the vector introduces genetic material in the form of RNA. For example, the delivery mechanism via gammaretrovirus (part of the Retroviridae viral family, along with Lentiviruses) is shown below:

Figure 3. Gammaretroviral vectors. A). Genomic structure of MLV-derived ƴ-retroviral vectors. Essential genes gag, pol, and env are removed from the viral backbone and provided in trans for viral production. Transgene encoding CAR is introduced in place of the viral genes. A packaging cell line is transfected with the vector carrying the CAR transgene, packaging and env helper plasmids. If desired, the selective antibiotic pressure is utilized to select for plasmid integration and generate stable virus-producing lines for large-scale production. B). Retroviral particles are collected from the cell culture supernatant and used to transduce stimulated T cells (OKT3/CD28 blasts). After genomic integration, the CAR is stably expressed on the surface of the T cells. (Morgan & Boyerinas, 2016)

As the diagram describes above the genetic material encoding CAR is reverse-transcribed by RNA and permanently integrated into the genome of the patient cells; therefore, CAR expression is maintained as the cells divide and are grown to large numbers in the bioreactor. In addition to gammaretroviral vectors, lentiviral vectors are commonly used in clinical trials of CAR T cell therapies, including Novartis’ CTL019. Other methods of gene transfer, including the Sleeping Beauty transposon system or mRNA transfection.

Bioreactor culture systems are designed to provide the optimal gas exchange requirements and culture mixing necessary to grow large numbers of cells for clinical use. The WAVE Bioreactor (now known as the Xuri; GE Healthcare Life Sciences) utilises a rocking platform in order to expand the CD19-targeted CAR T cell therapy CTL019 whilst the culture system G-Rex (Wilson Wolf), which has the ability to expand cells from low seeding densities.

Figure 4. T Cell Culture and Transduction (Morgan & Boyerinas, 2016)

The big drawback of these culture systems, however, is that the flask must be opened during cell inoculation. One system that is automating the cell preparation, enrichment, activation, transduction, expansion, final formulation and sampling, is the Miltenyi Biotec’s CliniMACS Prodigy (Morgan & Boyerinas, 2016); offered as an ‘all-in-one solution for cell processing in a closed GMP-compliant system’.

Figure 5: Miltenyi Biotec’s CliniMACS Prodigy

A feasibility study of the CliniMACS Prodigy was conducted by Mock et al. in August 2016, with the protocol as follows:

‘Using a closed, single-use tubing set, mononuclear cells were processed from fresh or frozen leukapheresis harvests collected from healthy volunteered donors. Cells were then phenotyped and subjected to automated processing and activation using TransAct, a polymeric nanomatrix activation reagent incorporating CD3/CD28-specific antibodies. Cells were then transduced and expanded in the CentriCult-Unit of the tubing set, under stabilized culture conditions with automated feeding and media exchange.’

The process had been continuously monitored to determine the kinetics of expansion, transduction efficiency, and phenotype of the engineered cells and was considered comparable to existing procedures with yield sufficient for clinical dosing.


Scalability: towards cost-effective commercial scale

Finally, the recent successes of adoptive T-cell immunotherapy for the treatment of hematologic malignancies have highlighted the need for manufacturing processes that are robust and scalable for product commercialisation; particularly as these medicines become widely adopted.

Organising the production of a few dozen cellular products a year can be arranged in a straightforward manner, with limited infrastructure and personnel. However, when implementing manufacturing processes for phase III clinical trials/commercial use, new sets of challenges will arise to produce and distribute, the hundreds or even thousands of cell therapeutic doses per year that are required.

Kaiser, A.D et al. (2015) proposed with their Cancer Gene Therapy paper – ‘Towards a commercial process for the manufacture of genetically modified T cells for therapy’ – different models of cell manufacturing centres shown in Figure 6:

Figure 6. Different models of cell manufacturing centres

A solution that could be adapted from other automated industries is the production line, where a specific product moves from one station to the next.  The patients' cells would enter the ‘processing station’ where a skilled operator would have the task to document and prepare (e.g., perform washes, density gradient separation, subset isolation) enriched T cells, as well as activate them (i.e., the addition of stimulatory reagent). The cells would then move into a (physically) separated space to be transduced (i.e. addition of viral vector). Cells could then be placed in an adjacent suite organised to accommodate the expansion of the cells either using individual stations or modular spaces to accommodate the chosen expansion method.

Alternatively, to achieve scalable manufacturing the industry could take advantages of devices such as CliniMACs Prodigy and develop a unit-based production method i.e. one device is dedicated to the production of one patient product at a time. 



al, B. e. (2016, April 18). Automation of CAR-T Cell Adoptive Immunotherapy Bioprocessing: Technology Opportunities to Debottleneck Manufacturing. Retrieved from Bioprocess International:

Kaiser, A. e. (2015). Towards a commercial process for the manufacture of genetically modified T cells for therapy. Cancer Gene Therapy, 22, 72-78.

Levine, B. e. (2017). Global Manufacturing of CAR T cell therapy. Molecular Therapy Methods & Clinical Development, 92-101.

Mock, U. e. (2015). Automated lentiviral transduction of T cells with Cars using the Clinimacs Prodigy. Blood, 126:2043.

Morgan, R. &. (2016). Genetic Modification of T Cells. Biomedicines.