Small but powerful: developing nanobody-based B7-H3 CAR T-cells for the treatment of paediatric brain tumours
Central nervous system tumours are the leading cause of mortality in paediatric cancer patients. Despite intensive treatment regimens, overall survival rates remain low: there is an urgent need for effective, targeted therapy. Following success against haematological malignancies, chimeric antigen receptor (CAR) T-cell therapy offers a promising alternative approach. In previous studies, B7-H3 was reported as a target of interest for paediatric brain tumours. Thus far, clinical efficacy and persistence of current B7-H3-directed CAR T-cells remains limited. We hypothesise that replacing the scFv-based antigen binding domain (ABD) of CAR T-cells with a nanobody (VHH) will improve efficacy. The inherent features of VHH make them a viable alternative; amongst others, they are small yet stable, have low immunogenicity, and do not aggregate, therefore limiting tonic signalling and premature exhaustion.
In this study, we compared five novel B7-H3-targeting VHH to an existing scFv. All ABDs were cloned into a second generation CAR construct containing 4-1BB and CD3z signalling domains. A cell-line-based NFAT-based reporter system was used to select the highest-performing VHH-based constructs. These were transduced into primary T cells and used in short-term cytotoxicity assays against DIPG and medulloblastoma cell lines. Killing capacity and cytokine secretion were measured using a luciferase reporter assay and LegendPLEX™, respectively. Finally, a combination of live microscopy, spectral flow cytometry, and LegendPLEX™ was used to compare persistence, phenotype, cytokine profile, and capacity of each CAR T-cell to respond to repeated stimulation with tumour cells over a period of 21 days.
Of the five VHH tested, two showed a high activation capacity in preliminary activation assays, similar to that of the control scFv. In 24 hour co-culture assays, one VHH showed a comparable killing capacity compared to the scFv-CAR, with significantly improved cytotoxicity against one cell line. In addition, the cytokine profile of these VHH CAR T-cells was more favourable following co-culture, indicating activation, proliferation, and cytotoxicity. Moreover, in long-term rechallenge assays, we observed a clear capacity of these VHH CAR T-cells to respond to repeated stimulation with tumour cells, extending beyond that of the scFv-based CAR T-cells.
Our results show that VHH-based CAR T-cells provide a promising alternative to scFv-based CAR T-cells. In both short- and long-term efficacy tests modelling cytotoxicity and persistence, we show a VHH-based CAR T-cell which exceeds the performance of an existing scFv-based CAR-T cell. By taking activation profiles, phenotype, and longevity into account, our approach promotes selection of the best performing candidate. Our results, alongside the inherent characteristics of a VHH-based construct, provide evidence that these VHH CAR T-cells show a high potential for clinical success and thus improved treatments for patients with central nervous system tumours.