Abstract 1808: Neuroblastoma cell lines established from progressive disease that exhibit partial or multi drug resistance are highly sensitive to chimeric receptor scFv(ch14.18)-zeta mediated NK cell killing

Introduction: Neuroblastoma (NB) is a neuroectodermal tumor of childhood characterized by a poor prognosis especially for high-risk patients. To specifically direct the cytotoxic abilities of NK cells towards NB cells expressing the glycolipid antigen ganglioside GD2, the human NK cell line NK-92 was genetically engineered to express a chimeric receptor, consisting of a GD2-specific ch14.18scFv-antibody fragment and the signal transducing zeta-chain of the CD3 complex (NK-92-scFv(ch14.18)-zeta). Recently, we showed that NK-92-scFv(ch14.18)-zeta (NK-92tr) are able to efficiently lyse GD2 expressing NB cells. Furthermore, plate-bound GD2 is sufficient to induce the activation of NK-92tr, indicated by increased granzymeB and perforin production. Both effects can be abrogated by blocking the chimeric receptor with Id14G2A-17-9. The treatment of high risk NB patients is still challenging, as the majority of these patients develops progressive disease after positive initial responses. These relapsed tumors often show resistance towards one or multiple drugs (MDR). Therefore we tested the hypothesis whether NK-92tr are able to efficiently lyse GD2+ tumor cell lines established from progressive disease that exhibit partial or MDR. Aims and Methods: We employed NB cell line pairs from five patients in Cr51 release cytotoxicity assays. Each pair consists of a cell line established at time point of diagnosis (SK-N-BE(1), SMS-KAN, SMS-KCN, CHLA-15, CHLA-122) and another cell line established from progressive disease (SK-N-BE(2), SMS-KANR, SMS-KCNR, CHLA-20, CHLA-136). Two of the relapse cell lines exhibit MDR (SK-N-BE(2), CHLA-20), the others partial drug resistance. Flow cytometry was utilized to analyze GD2 expression on all cell line pairs to analyze whether GD2 expression levels changed during progressive disease after initial chemotherapy. Further, we inhibited ganglioside synthesis in relapse cell lines by using the glucosylceramide synthase inhibitor PPPP to abrogate NK-92tr mediated cell killing. Results and Conclusions: The comparison of cell lines established at different time points from the same patients revealed for all five patients that the relapse cell lines show significantly higher sensitivity towards NK-92tr mediated lysis. This effect correlates with a significantly increased GD2 surface expression on relapse cell lines and can be abrogated by blocking ganglioside synthesis. Our results indicate that relapse cell lines which exhibit partial or MDR