Clinical-scale, modular manufacturing of tumor-reactive TILs using a closed and automated culture system

Abstract

Recent studies have revealed the potential of tumor-infiltrating lymphocytes (TILs) to treat solid tumors effectively and safely. However, the translation of TIL therapy for patients is still hampered by non-standardized and laborious manufacturing procedures that are expensive and produce highly variable cellular products. To address these limitations, the CliniMACS Prodigy^® Tumor Reactive T cell (TRT) Process has been developed. The TRT Process allows the automated isolation, transduction, and expansion of tumor-reactive T cells in a clinically compliant and closed system under GMP conditions. The TRT Process can generate tumor-reactive T cells using several methodologies which reflect clinically relevant applications. It can manage an automated Rapid Expansion Protocol (REP) using GMP-compliant reagents to generate a TIL cell product from solid tumors, including melanoma. Additionally, the TRT Process automates the closed selection of CD137-expressing TILs directly from tumor digest followed by the direct expansion of selected cells. Enriched CD137⁺ TILs could be robustly expanded even when as few as 1x10⁴ TILs were used to seed the REP phase. These data provide proof-of-concept for the isolation and expansion of tumor-reactive T cells from tumor digest in a closed, automated manner in the CliniMACS Prodigy, allowing for an efficient, simple, and reproducible manufacturing of TIL products. The direct selection of CD137⁺ TILs from tumor digest removes the need for the pre-REP phase, selects for therapeutically relevant cells, and can dramatically shorten the manufacturing time compared to conventional methods.

Keywords: CD137; CliniMACS Prodigy; GMP compliant cell manufacturing; REP; TILs; automation; tumor reactive T cells.

Figure 1

The Tumor Reactive T cell (TRT) Process using the CliniMACS Prodigy. (A) Schematic overview of the manufacturing process. Automated expansion of starting cell numbers of up to 2x10⁶ yTILs for 12 – 15 days. Cell counts, T cell expansion, viability and cell subsets of expanded (B) melanoma yTILs (n=9) and (C) yTILs from other solid tumor entities (n=7) were determined using flow cytometry analysis. Line marks the mean. Student's paired t-test was used to calculate significance. All p values <0.05 were considered statistically significant and are indicated as *p<0.05.

Figure 2

Differentiation status and specificity of TILs expanded with the TRT Process were determined using flow cytometry analysis. (A) Proportion of naïve or stem cell-like memory (T_N/SCM), central memory (T_CM), effector memory (T_EM) and effector (T_EFF) T cells (left graph) and expression of exhaustion markers TIM-3 (n=8), LAG-3 (n=7), PD-1 (n=8) and TIM3⁺/LAG3⁺/PD1⁺ (n=5) (right graph) among viable CD3⁺ T cells in the final product. Line marks the median (n=8). (B) Representative data of IFN-γ, TNF-α, CD107a and CD137 expression by TILs after an 18 hour overnight co-culture with autologous tumor cells. (C) Selected expanded TILs were re-stimulated with autologous or allogeneic tumor cells overnight for 18 hours and specific response was assessed by intracellular IFN-γ staining (n=4, mean with SD). Student's paired t-test was used to calculate significance. All p values <0.05 were considered statistically significant and are indicated as *p<0.05 and **p<0.01.

Figure 3

Automated stimulation, enrichment and expansion of virus-specific T cells from LP. All parameters were determined using flow cytometry analysis. (A) Performance of the separation. Frequency of CD137⁺ T cells was assessed in the starting material and in the positive fraction after enrichment. Recovery of CD137⁺ T cells, viability and cellular composition in the positive fraction was evaluated after enrichment (n=12). (B) Automated expansion of VSTs. T cell expansion kinetics, viability and the cellular composition of expanded VSTs (n=14). (C) Differentiation and exhaustion status among viable CD3⁺ T cells at the end of the expansion culture (n=14). For A-C, line marks the mean. (D) IFN-γ production of expanded VSTs and after re-stimulation with virus-specific PepTivator Peptide Pools (n=14, mean with SD). Student's paired t-test was used to calculate significance. All p values <0.05 were considered statistically significant and are indicated as **p<0.01 and ****p<0.0001.

Figure 4

Automated manufacturing of tumor-reactive T cells from melanoma samples (n=3). All parameters were determined using flow cytometry analysis. (A) Representative flow cytometry analysis of the CD137 expression in the starting material, in the negative and in the positive fraction after enrichment. (B) Performance of the separation. Frequency of CD137⁺ T cells was assessed in the starting and in the positive fraction after enrichment. Recovery of CD137⁺ T cells, viability and cellular composition in the positive fraction after enrichment. (C) Automated expansion of TRTs. T cell expansion, viability and the cellular composition of expanded TRTs at the end of the culture. For B+C, line marks the mean. (D) Differentiation and exhaustion status among viable CD3⁺ T cells at the end of the expansion culture. Line marks the median. Student's paired t-test was used to calculate significance. All p values <0.05 were considered statistically significant and are indicated as *p<0.05 and **p<0.01.

Figure 5

Tumor specificity of CD137-enriched and expanded TRTs as determined by flow cytometry analysis. (A) Reactivity of CD137-enriched TRTs (n=3, mean with SD) and (B) CD137-depleted TILs against the autologous tumor cell line. (C) Reactivity of CD137-enriched TRTs and CD137-depleted TILs from one selected donor against autologous and allogeneic tumor cells (technical triplicates, mean with SD). Student's paired t-test was used to calculate significance. All p values <0.05 were considered statistically significant.