Quantitative Predictive Cellular Analytics Driving Improvements in Advanced Therapy Development and Manufacturing

Abstract

Background & Aim

Advanced therapy medicinal products (ATMPs) such as chimeric antigen receptor (CAR) T-cell therapies have shown great success, but challenges remain. These include variability in patient and donor cell quality and complex manufacturing processes, highlighting the need for robust analytical methods to ensure quality and consistency while increasing access. Process automation offers benefits, but fit-for-purpose cell-based analytics are crucial for providing predictive and prescriptive insights from R&D through manufacturing and quality control (QC). Connecting analytics throughout the development and manufacturing process can reduce the need for comparability studies, lower costs, and improve product quality. This study provides data supporting Laser Force Cytology (LFC) as a comprehensive cell-based analytical tool for use throughout the entire CAR T-cell development process, including donor screening, process development, monitoring, and QC.

Methodology

LumaCyte's Radiance® instrument uses LFC, combining optical and hydrodynamic force measurements to analyze single cells without the need for antibodies or fluorescent labels. To demonstrate the technology, naive, mock transduced, and CD19+ CAR T cells were manufactured from healthy human donors. Transduced cells were incubated with NALM6 cells for a 24h period at different effector:target cell ratios. After the 24h incubation period, co-cultures and monocultures were analyzed using LFC and in parallel monitored for luciferase activity, which served as an orthogonal measurement of the NALM6 cell killing.

Results

Co-culture assay results demonstrated a strong correlation between the killing efficacy measured using the label-free LFC data with the orthogonal luciferase assay. After establishing the co-culture killing capacity of each manufactured CAR T cell lot, donors were classified as either high or low-killing efficiency. Using the LFC analysis of the PBMC starting materials from each donor, the classification as either high or low-killing efficiency could be predicted.

Conclusion

LFC precisely measures subtle cellular changes, providing near real-time data that enables donor and process prediction to speed process development and ultimately improve manufacturing and clinical outcomes. Our data demonstrate how this novel technology improves CGT development and production assays, compress development time, and reduces costs for these life-saving treatments.