Anti-CD19 chimeric antigen receptor T cell (CAR19-T) immunotherapy has shown curative potential in B cell malignancies. However, clinical remissions in relapsed/refractory CD19+ lymphomas and lymphoproliferative disorders are often short-lived, with therapeutic benefit for less than half of patients, highlighting the need for more effective CAR-based strategies.

iNKT cells are rare but powerful immunoregulatory and cytotoxic T lymphocytes, playing a pivotal anti-tumour role. They are restricted by CD1d, a non-polymorphic, glycolipid-presenting HLA I-like molecule, expressed on malignant CD19+ B cells in mantle cell lymphoma (MCL) and marginal zone lymphoma (MZL) cells, while in up to 50% of patients with chronic lymphocytic leukaemia (CLL) CD1d expression is very low or negative.

I tested the hypotheses that a) bi-specific CAR19-iNKT cells, targeting simultaneously CD19 and CD1d, via the CD19-specific CAR and their natural invariant TCR respectively, would be more effective than CAR19-T cells against CD19+CD1d+ B cell malignancies and b) transcriptional enhancement of CD1d expression would increase the CAR19-iNKT cytotoxic effect, including against CLL cells.

I established and optimized a novel, highly efficient protocol for manufacturing clinical scale CAR19-iNKT cells. In vitro validation demonstrated that CAR19-iNKT cells are CD19-specific, retain their natural CD1d-restricted reactivity and exert additive dual-specific cytotoxicity against CD1d+CD19+ targets. Compared to same-donor CAR19-T, CAR19-iNKT cells display a significantly higher expandability and proliferative potential, they are equally or more effective in killing CD19+CD1d+ lymphoid cell lines and consistently more effective against primary MCL, MZL and CLL cells.

Notably, I found that in CD1dlow/– primary CLL cells, surface CD1d expression can be restored by clinically relevant concentrations of all-trans retinoic acid (ATRA) and, in line with my hypothesis, CAR19-iNKT but not CAR19T cells displayed higher cytotoxic activity against ATRA-treated CLL cells.

Finally, in an NSG xenograft model of lymphoma, CAR19-iNKT cell immunotherapy led to a significantly improved overall survival, with earlier, more profound and sustained complete responses, which resulted in a significantly improved tumour-free survival as well as eradication of CNS lymphoma.

I conclude that CAR19-iNKT are more effective than CAR19-T cells against CD1d+CD19+ B cell malignancies in vitro and in vivo. This, together with the previously demonstrated ability of third donor-derived iNKT cells to protect from acute Graft-versus-Host Disease (aGVHD), raise the prospect of developing a more effective ‘off-the-shelf’ CAR19-iNKT immunotherapy for lymphomas. Furthermore, the finding that ATRA-mediated restoration of CD1d expression enhances the anti-lymphoma effect of CAR19-iNKT immunotherapy against CD1d low/negative tumour cells in vitro suggests that CAR-iNKT cells, in combination with transcriptional/epigenetic modulation of CD1d, may represent a highly efficient platform for CAR-based immunotherapy also for other CD1d-negative disorders.