P985: cll-1 is an adoptive immunotherapy target in juvenile myelomonocytic leukemia

Topic: 15. Myeloproliferative neoplasms - Biology & Translational Research Background: Juvenile myelomonocytic leukemia (JMML) is a myeloproliferative/myelodysplastic malignancy of young children that is treated with hematopoietic stem cell transplantation (HSCT). However, only half of all patients will experience long-term survival after HSCT. Adoptive immunotherapy with chimeric antigen receptor (CAR)-T cells has been shown to be effective in advanced lymphoid malignancies but has not been explored extensively in JMML. Aims: We hypothesized that identification of cell surface targets on JMML cells would be amenable to targeting with CAR-T cells even in patients who relapsed after HSCT. We therefore sought to identify potential CAR-T cell targets and test them in vitro using primary samples and in vivo using patient derived xenograft (PDX) models. Methods: Using bulk RNA sequencing (RNAseq) of peripheral blood (PB) and bone marrow (BM) cells from 88 patients and comparing it to RNASeq from healthy donors, we identified surface antigens highly expressed on JMML that were minimally expressed on healthy tissue. Due to potential differences in transcriptome-proteome correlation, we validated the transcriptomic results by flow cytometry. CARs were cloned by Gibson Assembly. Jurkat and healthy donor CAR-T cells were generated by lentiviral transduction and used for subsequent experiments including luciferase-based cytotoxicity assays. Results: After performing hierarchical clustering of surface protein encoding genes, we identified that JMML, AML and ALL patients display distinct clustering (Figure 1A) which therefore warrants JMML-specific studies. Our selection pipeline (Figure 1B) revealed that CLEC12A (which encodes for CLL-1) was highly expressed in JMML but not in pediatric healthy cells or in healthy tissue as per the GTEx database. Flow cytometry confirmed higher mean fluorescence intensity and CLL-1+ percentages in JMML cells compared to normal PB and BM cells (Figure 1C). CAR-T cells containing a CD8α hinge, CD8α transmembrane and 4-1BB costimulatory domain were most effective against CLL-1 expressing U937 cells and more persistent in vitro than three other constructs with varying hinge, transmembrane and costimulatory domains. CLL-1 and CD33 CAR-T cells (used as positive control known to be effective in myeloid malignancies) effectively killed CD33+, CD11b+, CLL-1+ and CD14+ JMML cells in vitro (Figure 1D). PDX studies comparing CLL-1, CD33, empty CAR-T cells and vehicle are ongoing. Summary/Conclusion: RNAseq and flow cytometric validation demonstrate that CLL-1 is broadly expressed on JMML cells. In vitro data indicate that anti-CLL-1 CAR-T cells can effectively kill JMML cells. In vivo studies in PDXs are ongoing and will investigate the efficacy of this therapy in JMML. CLL-1 CAR-T cells could potentially be used as salvage therapy in those who relapse after transplant.Keywords: JMML, Cancer immunotherapy, CAR-T, Adoptive immunotherapy