Abstract

Injection of glycolipid KRN7000 (α-galactosylceramide) into mice induces activation of a specific subpopulation of T cells called invariant NKT (iNKT) cells which are important for tumor cell eradication. NKT cell activation is antigen-specific and dependent on presentation of agonist by the non-polymorphic class I-like molecule CD1d on antigen-presenting cells. In contrast to mouse, injection of KRN7000 into cancer patients did not induce any clinical responses in a published phase I study. This lack of response in part reflects the very low frequency of iNKT (Vα24/Vβ11 T cell receptor) cells in humans and especially in some cancer patients. Therefore, therapies which can increase iNKT cell numbers in patients may enhance the chances of tumor cell killing. Adoptive transfer of in vitro-expanded iNKT cells in combination with KRN7000 may be one therapeutic option. Our group and others have shown that CD1d-positive tumor cell lines are very sensitive to killing by human NKT cells. In contrast, killing of CD1d-negative tumor cells is not directly mediated by NKT cells but can be influenced through cytokines secreted by NKT cells. Based on these results and the reduced frequency of iNKT cells in patients, we hypothesized that NKT cell adoptive therapy could be most beneficial for treatment of patients with CD1d-positive cancers. Previous reports have shown that multiple myeloma (MM), acute myeloid leukemia (AML)-M4, and T-acute lymphoblastic leukemia (ALL) tumor cells express CD1d and may be susceptible to NKT cell killing in vitro. In this proof-of-concept study, we analyzed iNKT cell frequency, expansion, and function from six patients with hematopoietic malignancies (MM and AML). Although the frequency of iNKT cells in patients was extremely low (<0.01% in 5 of 6 patients), in vitro culture of PBMCs with KRN7000 and rhIL-2 induced tremendous NKT cell expansion. After the initial 12–14 day culture, iNKT cells expanded from 300-fold to greater than 2000-fold in 5 of 6 patient samples. Purified iNKT cell lines were generated from 5 of 6 patients and then restimulated every 8–12 days. Within one month, it was possible to obtain more than 100 million NKT cells (4 of 6 donors) from a starting culture of 10 million PBMCs (∼100–6,000 NKT cells). The sorted and expanded iNKT cell lines were then tested for cytokine secretion and cytotoxic activity. Upon activation, patient iNKT cell lines produced relatively normal levels of IFN-γ, IL-4, IL-5, and GM-CSF but slightly reduced amounts of IL-2 and TNFα compared to normal donor iNKT lines. Importantly, patient iNKT cell lines showed potent and dose-dependent killing activity against B, T and monocytic tumor cell lines in a 4h chromium release assay. The cytotoxic activity was both CD1d-restricted and antigen-specific. Near maximal killing of tumor cells was seen at an effector to target cell ratio of 5:1, but killing was seen even at 1:1 (E:T). Three of four patient iNKT cell lines showed strong killing of Jurkat (57–66% killing), MOLT-4 (36–57% killing), CEM (30–46%killing), U-937 (47–49% killing), and CD1d-transfected B cell (54–78% killing) targets at a 5:1 NKT:tumor cell ratio. Based on this data and results of the preliminary toxicology study in non-human primate, we believe that adoptive transfer of expanded iNKT cells plus KRN7000 is safe and may be particularly effective against CD1d-positive tumor cells in vivo.

Author notes

Disclosure: No relevant conflicts of interest to declare.