An estimated 10-20% of CAR-T batches fail due to slow growth rates of Tcells during the expansion step of the process. Furthermore, the starting material for every batch differs in many ways including the distribution of phenotypes in the Tcell population, yet the final product specification is fixed.
This project then aims to evaluate how levels of key cytokines effect the growth rate of important Tcell phenotypes, so as to promote a more consistent final product. To accomplish this, a properly scaled-down “bioreactor” must be used.
Screen CART patients’ cells for enhanced growth prior to start of the clinical batch using a transcriptome-driven mathematical model and scaled-down bioreactor.
Reduce number of failed batches of CART cells due to slow growth or incorrect phenotype distribution.
By implementing a magnetically‑actuated microbioreactor platform alongside ANFIS‑based modeling of cytokine effects on key T‑cell phenotypes, an organization will reduce process‑development costs–enabling reliable, high‑oxygen‑transfer scale‑down studies with minimal reagents. This approach will accelerate CAR‑T development timelines through predictive, patient‑specific cytokine set‑point selection and faster iteration cycles and will deliver consistently targeted T‑cell phenotypes validated by RNA‑seq—ensuring robust product quality and GMP‑aligned documentation without compromising clinical or regulatory expectations.
Two technologies were developed through this project: 1). a microbioreactor with mixing capabilities (via magnetically-actuated microposts) 2.) a mathematical modelling approach using neuro-Fuzzy inference systems (ANFIS) to model the effects of three key cytokines (IL2, IL7 and IL15) on four Tcell phenotypes deemed important for CART therapies, with the potential to use the model before treatment to assess optimal Tcell growth conditions for different patients
Oxygen Transfer in the wells with RedBud magnetically-actuated Posts rates were found to be comparable to larger scale bioreactors. Increasing stirrer speed (3500-9000 rpm) and decreasing fill volume (150-350 µL) in these mixed wells led to increasing Oxygen Transfer (kLa: 4-88 h-1)
Improved growth of CD4+ Human Tcells and a significant increase in Chinese Hamster Ovary growth was observed in the Redbud mixed wells (580,000 cells/mL) verse unmixed control wells (420,000 cells/mL); t(12.814) = 8.3678, p=
The CD4+ and CD8+ naïve Human Tcell phenotypes had higher growth rates than the memory phenotypes, with the high IL-2 and high IL-7 conditions eliciting high growth for both. CD4+ memory T cells grew faster than the CD8+ memory phenotype, with the moderate IL-2 and IL-15 condition being the best condition.
The growth study was analyzed using machine-learning techniques, where adaptive neuro-Fuzzy inference systems (ANFIS) effectively modelled the growth of each T cell phenotype as a function of levels of supplemented γ-chain cytokines. The same approach was also used to model differentiation of the naïve T cell phenotypes cells to memory T cell phenotype. This modelling approach can help CART providers ensure optimal Tcell growth rates for each patient
Using RNA-Sequencing techniques, we studied the cells’ gene regulation through differential gene expression analysis; specifically comparing the phenotypes with one another. For example, the CD4+ naïve and memory T cell phenotypes differentially express genes related to response to type I interferon and CD8+ naïve and memory T cell phenotypes express genes related to stem cell maintenance and epithelial cell differentiation. This study drives understanding of how the phenotypes respond differently to cytokines, and how to best cultivate each phenotype for ex vivo growth.
Coppola, C., Hopkins, B., Huhn, S., Du, Z., Huang, Z., & Kelly, W. J. (2020). Investigation of the Impact from IL-2, IL-7, and IL-15 on the Growth and Signaling of Activated CD4(+) T Cells. International Journal of Molecular Sciences, 21(21), 7814. https://doi.org/10.3390/ijms21217814
Fisher, J. T., Gurney, T. O., Mason, B. M., Fisher, J. K., & Kelly, W. J. (2021). Mixing and oxygen transfer characteristics of a microplate bioreactor with surface-attached microposts. Biotechnology Journal, 16(5). https://doi.org/10.1002/biot.202000257
Hopkins, B., Fisher, J., Chang, M., Tang, X., Du, Z., Kelly, W. J., & Huang, Z. (2022). An In-Vitro Study of the Expansion and Transcriptomics of CD(4+) and CD(8+) Naive and Memory T Cells Stimulated by IL-2, IL-7 and IL-15. Cells, 11(10). https://doi.org/10.3390/cells11101701
Login to the NIIMBL member portal to access additional project information, including presentations, progress updates, reports, and more.
Not yet a member? Learn more about which level of NIIMBL membership is right for you and your organization.
Villanova University
Merck Sharp & Dohme LLC
Redbud Labs Inc
