Treatment of pre-B cell acute lymphoblastic leukemia (ALL) using chimeric antigen receptor expressing T cells (CART) targeting CD19 have demonstrated impressive clinical results in children and young adults with up to 70-90% complete remission rate in multiple clinical trials. However, about 30% of patients relapse due to loss of the targeted epitope on CD19 or CART failure. Our CD22-targeted CAR trial has generated promising results in relapsed/refractory ALL, including CD19 antigen negative ALL, but relapse associated with decreased CD22 site density has occurred. Thus, developing strategies to prevent relapses due to changes in antigen expression have the potential to increase the likelihood of durable remissions. In addition, dual targeting of both CD19 and CD22 on pre-B ALL may be synergistic compared to targeting a single antigen, a potential approach to improve efficacy in patients with heterogeneous expression of CD19 and CD22 on leukemic blasts. We describe the systematic development and comparison of the structure and therapeutic function of three different types (over 15 different constructs) of novel CARs targeting both CD19 and CD22: (1) Bivalent Tandem CAR, (2) Bivalent Loop CAR, and (3) Bicistronic CAR. These dual CARs were assembled using CD19- and CD22-binding single chain fragment variable (scFv) regions derived from clinically validated single antigen targeted CARs. They are structurally different in design: both tandem and loop CARs have the CD19 and CD22 scFv covalently linked in the same CAR in different orders, whereas, bicistronic CARs have 2 complete CAR constructs connected with a cleavable linker. The surface expression on the transduced T cell of the CD19/CD22 dual CARs was detected with CD22 Fc and anti-idiotype of CD19 and compared to single CD19 or CD22 CARs. Activities of dual CARs to either CD19 or CD22 were evaluated in vitro with cytotoxicity assays or killing assays against K562 cells expressing either CD19 or CD22 or both antigens and also tested against a leukemia CD19+/CD22+ cell line, NALM6, and NALM6 with CRISPER/CAS9 knockout of CD19 or CD22 or both antigens. Therapeutic function of the top candidates of the dual CARs was then validated in vivo against these NALM6 leukemia lines. Some of these dual CARs were also further tested against patient-derived xenografts. Finally, we tested the dual targeting CARs in an artificial relapse model in which mice were co-injected with a mix of CD19 knockout and CD22 knockout NALM6 leukemia lines. From these studies, we established that the order of the scFv, size of the linker, type of leader sequence, and co-stimulatory domain in the CAR constructs all impact the efficacy of the dual targeting CARs. Tandem, Loop, and Bicistronic CARs all demonstrate some levels of in vitro and in vivo activities, but the bicistronic CAR was most effective at clearing leukemia and preventing relapse. In the CD19+/CD22+ NALM6 model, bicistronic CAR treated mice remain disease free while CD19 CAR or CD22 CAR treated mice already died or relapsed on day 27. In the relapse model, as expected, CD19 or CD22 single CAR T cell treatment resulted in progression of the corresponding antigen-negative NALM6. Treatment with dual targeted bicistronic CARs resulted in clearance of both CD19 and CD22 negative ALL with durable remission. In summary, we described novel CD19/CD22 dual targeting CARs with robust pre-clinical activity against pre-B cell ALL, and validated this approach in the prevention of resistance to single-antigen targeted CARs in preclinical models.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution