Lentiviral Transduction Of Notch Ligand Expanded Cord Blood HSC’s To Express a CD19-Specific CAR Generates NK and Myeloid Progeny With Antitumor Activity

Over the past decade, significant progress has been made using chimeric antigen receptors (CAR) to redirect the specificity of T cells against an array of cell surface tumor antigens. Using this same technology, we hypothesized that CD34+ hematopoietic stem and progenitor cells (HSPC) derived from cord blood (CB) could be transduced with lentivirus to efficiently express a CAR specific for CD19 that would enable targeted lysis through its continued expression in differentiated progeny effector cells. As previously reported, the Notch-mediated ex vivo expansion of CB HSPC is a clinically validated cell therapy product that is well tolerated, can be given off the shelf without HLA matching, and provides transient myeloid engraftment in both the HCT and intensive chemotherapy setting. CB-derived CD34+ cells are selected using the Miltenyi CliniMACS or AutoMACS platform and plated in StemSpan serum free expansion media supplemented with IL-6, IL-3, TPO, Flt3L, and SCF and in the presence of immobilized Notch ligand and retronectin. The cells are cultured for 14-16 days resulting in >150 fold expansion of CD34+ cells on average and a final heterogeneous cell product consisting of 14.5% (range 6.2-26) CD34+, 20.5% (6-36) CD15+, 11.3 (1.8-23) CD14+, 5.4%(2.2-13.6) CD56+CD16+, and 0.1(0-0.2)CD20+ which is cryopreserved for future clinical use. Importantly, no T cells are derived from this culture system. Off the shelf expanded units have been infused into 45 patients and no serious adverse events have been noted except for one allergic reaction attributed to DMSO. Additionally, there has been no persistent engraftment beyond day 180 in the HCT setting and 14 days post infusion in the chemotherapy setting.

To engineer anti-CD19 activity into this product, cells are transduced at an MOI of 3 on day 3 of culture using a self inactivating lentivirus that encodes for a second generation CD19 CAR with 41BB costimulation as well as an EGFR tag which lacks an intracellular signaling domain and serves as a marker of transduction and a selection marker. Transduction efficiency ranges from 20-40% and equally transduces the CD34+ and CD34- populations. Copy number analysis demonstrates between 1-4 copies/cell. Expression of the transgene does not affect the final culture phenotype at 14 days and transgene expression is seen in all cell subsets and is stable over the culture period. Due to concerns of expression of the CAR on HSC with potential signaling capacity, irradiated LCL was added on day 7 of culture at a 1:1 ratio to provide antigen stimulation and did not negatively affect the growth of the cultures, nor the expression of the transgene. Functional NK cells were derived in vitro from the end culture population by an additional week of culture with RPMI supplemented with L-glutamine, human serum, IL-2 and IL-15. The CD19-CAR expressing NK cells had enhanced cytotoxic activity in a CRA against LCL compared with non-transduced NK cells (50 v 30%) whereas both killed K562 targets equally (75 v 80%). Ongoing experiments are being conducted to further evaluate the expression, signaling capacity and function of the CAR on each of the cell compartments.

The NOG mouse model was used for in vivo functional assays to look at effects of CD19 CAR expression on human repopulating ability and to assess anti-CD19 activity. NOG mice underwent sub-lethal irradiation followed by tail vein injection of the expanded progeny of 20,000 selected CD34+ CB cells. There was no effect on early engraftment of transduced cells, however there was no detectable CD19 engraftment in the mice transplanted with CD19 CAR expressing cells compared with untransduced cells, in which the CD19 population was >20% of the engrafted cells, indicating anti-CD19 activity. NK cell populations were increased using NS0-IL15 secreting cells, irradiated and injected subcutaneously three times per week starting at week 3 to provide enhanced effector function. Current studies are ongoing to evaluate the anti-CD19 activity using the Raji CD19+ tumor model in the NOG mice.

These initial results are promising as a way to engineer a graft versus leukemia effect into cord blood transplant. Furthermore, transduction of a CD19 CAR into our existing universal donor expanded CB cell therapy product allows for infusion of an anti-CD19 cell product to be given immediately following identification of a patient with clinical need for therapy, for example one in relapse or with persistent MRD.