Streamlining the CAR T cell therapy workflow: A look into closed and semi-automated processes

Abstract

There is considerable excitement surrounding the treatment potential of T cell immunotherapies. Despite the clinical success of chimeric antigen receptor (CAR) T cells, there remain challenges associated with routinely offering these products as treatment alternatives. These challenges include the costly manufacturing process relying on lengthy and complex open workflows with high manual labor requirements that influence product variability.

This article describes the details of a robust CAR T cell manufacturing workflow that can be adapted for current good manufacturing practices (cGMP) compliance in commercial production of CAR T cells. This semi-automated, closed CAR T process achieves 1 × 10¹⁰ expanded T cells. Enhanced green fluorescent protein (eGFP) transduction efficiency is greater than 80% across an 8-day manufacturing process.

Introduction

In 2023 alone, an estimated 1.9 million new cases of cancer are expected to be diagnosed in the US, and about 600 000 Americans are expected to die of cancer—equivalent to 1670 deaths per day (1). Although chemotherapy combined with drugs continues to be the standard of care, an increasing number of adults and children are refractory to conventional treatment modalities, motivating the “next generation” of approaches for cancer therapy.

Unlike conventional chemotherapy, novel approaches are now being developed based on a deeper understanding of immuno-oncology. Among these approaches, adoptive T cell therapy has gained significant attention for its curative potential. Approval of autologous CAR T cell therapies from Novartis (Kymriah™) and Gilead/Kite (Yescarta™) (2) underscores both the promise and challenges ahead in order to make these therapies available to patients.

There is now a growing need for end-to-end workflow solutions supporting industrial scale-up and scale-out of CAR T cell therapies. To address these challenges, we have investigated individual CAR T unit operations to identify commercially available reagents and modular equipment that drive process closure and automation as a method to improve workflow efficiency and product consistency. The CAR T workflow is shown in Figure 1.

Fig 1. Workflow for the semi-automated, closed CAR T cell manufacturing process. The cell material at the start and end of the process is highlighted in green. The unit operations are shown in white.

We evaluated commercially available T cell enrichment and activation reagents, closed small-scale culture vessel options, alternative solutions to enhance transduction, and the specific timing of process steps. The aim was to develop a modular platform process that is robust and flexible for the varied needs of CAR T developers.

Detailed materials and methods are provided at the end of the article.

Results and discussion

Comparison of thawing and washing methods

Open, manual methods are available for thawing and washing cells from leukopaks. However, these do not comply with evolving GMP requirements and might introduce risks associated with operator variability and contamination. As an alternative, a closed, automated approach can be used to thaw and wash the cells. Such an approach can reduce manual labor requirements and product variability.

Thawing of frozen leukopaks using an automated dry thawing device*, washing using a functionally closed cell processing system† resulted in cell recovery and viability results comparable to a manual method (p ≥ 0.05 by 2-sample t-test, n = 3) (Fig 2A). Cells from the automated method also gave comparable T cell recovery, purity, and viability (p ≥ 0.05 by 2-sample t-test, n = 3) (Fig 2B).

* For upstream thawing we used a proprietary device. An alternative would be to use the VIA Thaw™ instrument for both upstream and downstream processing.
† a functionally closed cell processing device was used in this study. However, the Sepax™ C-Pro instrument is recommended as the equivalent instrument for manufacturing purposes.

Fig 2. Recovery and viability after manual and automated thaw and wash. (A) Viable peripheral blood mononuclear cell (PBMC) recovery and viability immediately after manual or automated thaw and wash; (B) recovery, purity and viability of enriched T cells after EasySep Release Human CD3 Positive Selection. Bars represent the average of the biological triplicates with individual donor values shown by colored dots.

Comparison of T cell enrichment methods

There are several commercially available methods for enriching T cell populations. Two common systems are CD3⁺selection and CD14⁺/CD19⁺ depletion. In order to determine the optimal protocol for enriching T cells, two CD3⁺selection methods and one CD14⁺/CD19⁺ depletion method were evaluated.

As expected, depletion of the CD14⁺ and CD19⁺populations resulted in significantly lower T cell purities than either of the CD3⁺selection methods (p < 0.001 by one-way ANOVA, n = 3) (Fig 3B). T cell enrichment using the EasySep™ Release Human CD3 Positive Selection Kit resulted in significantly higher recoveries than selection with the QuadroMACS™ system (p = 0.01 by one-way ANOVA, n = 3) (Fig 3B). However, there were no statistically significant differences in the viability and 5-day expansion kinetics between any of the enriched T cell populations (p ≥ 0.05 by one-way ANOVA, n=3) (Fig 3C).

                              (A)

                              (B)

                              (C)

Fig 3. Comparison of enrichment methods. (A) FlowJo™ representation of gated cells before and after EasySep CD3⁺ selection; (B) recovery, purity, and viability of T cells after either EasySep CD3⁺ selection, QuadroMACS CD3⁺ selection, or QuadroMACS CD14⁺/CD19⁺ depletion; (C) day 5 viability and cumulative fold expansion relative to the EasySep CD3⁺ enriched T cell growth of the respective donor. Bars represent the average of the biological triplicates with individual donor values shown by colored dots.

Comparison of T cell transduction methods

In order to optimize transduction, the timepoint to perform transduction after the initial activation of the T cells was evaluated. Earlier timepoints were prioritized to maintain as short a process duration as possible. Several methods for transducing T cells are currently in use. However, a number of these rely on open and manual processing in plates or flasks, which are unlikely to be amenable to GMP requirements.

As an alternative, we tested the automated, functionally closed SpinOculation C-Pro application on the Sepax™ C-Pro instrument to transduce the T cells with lentivirus (LV)-eGFP.

Figure 4 shows results from the determination of optimal day for transduction. T cells enriched through CD3⁺selection had higher transduction efficiencies than the depletion method when transduced on day 0 (p = 0.000 by one-way ANOVA, n = 3) and day 1 (p < 0.001 by one-way ANOVA, n = 3). However, no differences were seen on cells transduced 2 days after activation (p ≥ 0.05 by one-way ANOVA, n = 3). Overall, the highest transduction efficiency was observed by transducing CD3+ selected cells 1 day after activation (p < 0.02 by one-way ANOVA, n = 3).

Fig 4. Determination of optimal day for transduction. Percentage of eGFP-positive T cells 72 h after transduction on various days after activation. Bars represent the average of the biological triplicates with individual donor values shown by colored dots.

As described in the methods, various manual transduction approaches currently used by numerous groups (T-flasks, shake flasks, and RetroNectin™) were compared to the SpinOculation C-Pro application on the Sepax™ C-Pro instrument. The automated, closed transduction method with SpinOculation C-Pro on the Sepax™ C-Pro resulted in both comparable viability (Fig 5A) and transduction efficiency at the end of culture (Fig 5B).

Fig 5. Comparison of transduction methods. (A) Viability of cells using various transduction methods after activation phase of culture (3 days or 4 days after transduction); (B) percentage of eGFP-positive T cells at the end of culture (7 or 8 days after transduction). Bars represent the average of technical triplicates that were run for each of two biological donors.

Comparison of T cell expansion methods

There are several open methods for culturing T cells at small scale prior to entering a bioreactor, with culture in T-flasks being the more common method. There are also closed methods such as gas permeable bags and G-Rex™, but in our previous evaluations, these were not able to reach the same growth kinetics observed in T-flasks (data not published). We investigated Erlenmeyer shake flasks as a closed culturing alternative that could match the fold expansion of T cells inoculated in T-flasks.

No statistically significant differences were found in the T cell viability or expansion (p ≥ 0.05 by 2-sample t-test, n = 4) (Fig 6A). Phenotypic comparison of cells at the end of culture are shown in Figure 6B. No statistically significant differences were found in T cell purity or CD4⁺ and CD8⁺ subpopulations (p ≥ 0.05 by 2-sample t-test, n = 4).

Fig 6. Comparison of small-scale culturing methods. (A) Day 5 viability and cumulative fold expansion relative to the T cell growth in T-flask of the respective donor; (B) phenotypic analysis of T cells and subpopulations CD4⁺ and CD8⁺ cells by flow cytometry at the end of culture. Bars represent the average of four biological replicates with individual values shown by colored dots.

Large-scale transduced T cell production

The automated thawing and washing of leukopaks were implemented along with T cell enrichment using the EasySep Release Human CD3 Positive Selection Kit, SpinOculation C-Pro application on the Sepax™ C-Pro instrument, and small-scale activation and expansion in shake flasks and Xuri™ Cell Expansion System W25 (Fig 7).

Fig 7. Process workflow. A diagram of the steps and systems used to prepare a clinically relevant dose of CAR T cells from frozen leukapheresis units. Most steps were functionally closed and automated, and process is compatible with GMP guidelines. SF = shake flask.

From three large-scale production runs, we were able to get over 75% cell recovery from the thaw and wash steps using the setup described, with greater than 90% viability (Fig 8A). Populations of enriched T cells were obtained by magnetic separation using the EasySep system, which increased T cell purity starting from 53% to 62% in the initial input population to above 95% in the output, regardless of different starting populations of cells (Fig 8B).

Fig 8. Purity and recovery of T cells after enrichment. (A) Viable PBMC recovery and viability after washing step; (B) Recovery, purity, and viability of enriched T cells after EasySep Release Human CD3 Positive Selection. Bars represent the average of the biological triplicates with individual donor values shown by colored dots.

After isolation, cells were activated, transduced and expanded in shake flasks for 4 days before being transferred into the Xuri™ Cell Expansion System W25. After 4 days, over 68% of the T cell population was transduced (Fig 9A). T cells were expanded in the Xuri™ Cell Expansion System W25 by over 78-fold from the starting cell population (Fig 9B). Transduction was maintained above 84% throughout the culture (Fig 9B) resulting in over 1 × 10¹⁰ transduced T cells on day 8, on average (Fig 9C).

Fig 9. Culture growth and transduction efficiency. (A) Cumulative growth and viability of T cells. Lines represent the expansion profile and triangles represent the viability during the CAR T process; (B) percentage of transduced GFP⁺ T cells after activation in shake flasks and after culture in the Xuri™ Cell Expansion System W25. Bars represent the average of the biological triplicates with individual donor values shown by colored dots; (C) cumulative GFP⁺ T cells on day 8 for donors 1, 2, and 3.

After culture, expanded transduced T cells were harvested using the FlexCell application on the Sefia™ cell processing instrument prior to cryopreservation. After volume reduction, wash, and cryoformulation, total live cell recovery was over 95% without a decrease in viability (Fig 10A). Cryogenic bags of cells were cryopreserved using the VIA Freeze™ Quad controlled ate freezer before being transferred to liquid nitrogen storage.

Figure 10B shows the viability and recovery post-thaw using VIA Thaw™ instrument after manual washing, centrifugation, and buffer exchange to evaluate the cells after cryopreservation (as a surrogate to the product ultimately infused into the patient). As shown, over 70% of the cells were recovered with above 97% viability. The results also show that transduction was maintained above 80% through the cryopreservation process.

Fig 10. Recovery and viability of T cells. (A) Recovery and viability of cultured T cells, from donor 1 and 2 after harvest using FlexCell application on the Sefia instrument; (B) post-thaw recovery and viability of transduced T cells cryopreserved after cryoformulation using FlexCell application on Sefia instrument.

Conclusions

The workflow used in this study is suitable for industrialized, autologous CAR T cell production, because it incorporates several automated and functionally closed unit operations. This large-scale transduced T cell production provided clinically relevant doses (1 × 10¹⁰ cells per donor) with high transduction efficiency (> 80% eGFP+). Phenotypically, the final product demonstrated characteristics of an activated T cell population.

Future work could focus on fine-tuning each step to reduce costs while maintaining a robust, scalable process. For example, bioprocess parameters such as agitation and rocking speed, as well as perfusion strategy, could be further improved by reducing the amount of media exchange and optimizing rocking speed and angle.

Overall, the methodology presented here is compatible with industrial workflows for CAR T manufacturing. This article shows the feasibility of a semi-automated T cell expansion and closure of key unit operations. In particular, the small-scale expansion in shake flasks demonstrates the possibility of an alternative to conventional static culture in gas-permeable bags or closed flasks.

Acknowledgements

All work was performed in collaboration with CCRM through funding from FedDev Ontario and Cytiva at the Centre for Advanced Therapeutic Cell Technologies (CATCT), Toronto, Ontario, Canada. The reporting and interpretation of the research findings are the responsibility of the author(s).

References

1. American Cancer Society. Cancer Facts & Figures 2023| American Cancer Society. www.cancer.org. Published 2023. https://www.cancer.org/research/cancer-facts-statistics/all-cancer-facts-figures/2023-cancer-facts-figures.html.
2. Approved Cellular and Gene Therapy Products, United States Food and Drug Administration. Accessed 22 May 2019.

Learn how we can help you secure and standardize your manufacturing process with closed, automated systems.

Ordering information