From bench to IND: Validation of TSLPR CAR T-cells reveal discrepancies in detection modalities and inform dose selection for a Phase 1 clinical trial in relapsed/refractory CRLF2-rearranged B-ALL patients. Free

Introduction: Chimeric antigen receptor (CAR) T-cell therapies targeting CD19 and CD22 have achieved remarkable success in patients with relapsed/refractory B-cell acute lymphoblastic leukemia (ALL). Nonetheless, post-CAR T-cell relapse remains common, including for the cytokine receptor-like factor 2-rearranged (CRLF2-r) B-ALL subtype characterized by overexpression of the thymic stromal lymphopoietin receptor (TSLPR). CRLF2 rearrangements comprise ~50% of high-risk Philadelphia chromosome (Ph)-like ALL with enrichment in adolescents/young adults and ~50% of Down syndrome (DS)-associated ALL (PMID 38657263) who have inferior outcomes with conventional chemotherapy. TSLPR thus represents a compelling alternative immunotherapeutic target for these patients, and our groups previously demonstrated potent activity of TSLPR CAR T-cells (TSLPRCART) in preclinical Ph-like and DS-ALL models (PMID 26041741, 39681640). To prepare for phase 1 clinical evaluation, we conducted preclinical efficacy, safety, and product characterization studies and addressed key issues raised during Investigational New Drug (IND) review.

Methods: We conducted numerous studies to support clinical translation of TSLPRCART. A clinical-grade TSLPR CAR vector incorporating a truncated EGFR (tEGFR) safety tag was validated using a Prodigy-based closed-system manufacturing process. In vitro assays evaluated activation, cytotoxicity, and cytokine secretion of TSLPRCAR/tEGFR T cells (TSLPRCART/tEGFR) against TSLPR⁺ ALL lines with CD19CAR T cells as a killing comparator. In vivo efficacy and persistence were assessed in ALL cell line or patient-derived xenograft (PDX) models established in NSG mice.

To evaluate safety, we mapped the epitope recognized by the parental 3G11 hybridoma and performed off-tumor toxicity assays using primary human cells from various normal tissues. Cytokine-independent growth was assessed in long-term culture without exogenous cytokines. In vivo tracking of CAR and tEGFR expression was performed to assess differences in TSLPRCART/tEGFR and CD19CART/tEGFR persistence.

Results: Clinically-manufactured TSLPRCART/tEGFR expanded efficiently with high CAR (31–53%) and tEGFR detection (54–78%) across six qualification runs. However, flow cytometric TSLPR-Fc staining dropped markedly between day 6 and 9 (18.7–75% reduction), while tEGFR levels remained stable. Further studies revealed CAR internalization, and we observed equivalent anti-ALL cytotoxicity of sorted TSLPR-Fc⁻/tEGFR⁺T-cells, prompting TSLPRCART dosing based on tEGFR⁺ cells.

TSLPRCART/tEGFR showed robust cytotoxicity and cytokine secretion against TSLPR⁺ B-ALL lines, comparable to CD19CART positive controls. In vivo, they demonstrated potent anti-leukemic activity and persistence in TSLPR⁺ ALL cell line and PDX models. High levels of tEGFR⁺ T cells remained >1 month after leukemia clearance. As CD19CAR T cells lacking tEGFR were undetectable at late time points, we generated CD19CART/tEGFR for comparison. For both constructs, tEGFR⁺ were detected at significantly higher levels than CAR+ cells in BM and spleen (day 64), highlighting discrepancies between detection methods that initially complicated safety assessments.

Mapping of the TSLPR epitope to a linear region in the D2 domain (aa118–127) revealed minimal homology to known proteins. Co-culture assays with primary human cells showed no off-target activation or cytokine release.

Long-term culture without exogenous cytokines showed transiently higher viability and activation in TSLPRCART/tEGFR versus bicistronic CD19/CD22CART, but these differences resolved by day 28 (with no viable cells), mitigating concerns of cytokine-independent growth or transformation.

Conclusions: This preclinical study outlines regulatory and technical challenges encountered during development of TSLPRCART immunotherapy for a planned first-in-human phase 1 trial. Our findings underscore the critical importance of CAR detection methodology and functional validation when establishing release criteria for clinical-grade products. Lessons from the IND process, particularly regarding surrogate markers, functional heterogeneity, and product tracking, have informed the design and launch of an upcoming clinical trial. This work highlights the broader translational principle that regulatory readiness hinges not only on preclinical efficacy, but also on rigorous product characterization and alignment of analytical assays with functional performance.