An optimized bicistronic chimeric antigen receptor against GPC2 or CD276 overcomes heterogeneous expression in neuroblastoma

Chimeric antigen receptor (CAR) T cell therapies targeting single antigens have performed poorly in clinical trials for solid tumors due to heterogenous expression of tumor-associated antigens (TAAs), limited T cell persistence, and T cell exhaustion. Here, we aimed to identify optimal CARs against glypican 2 (GPC2) or CD276 (B7-H3), which were highly but heterogeneously expressed in neuroblastoma (NB), a lethal extracranial solid tumor of childhood. First, we examined CAR T cell expansion in the presence of targets by digital droplet PCR. Next, using pooled competitive optimization of CAR by cellular indexing of transcriptomes and epitopes by sequencing (CITE-Seq), termed P-COCC, we simultaneously analyzed protein and transcriptome expression of CAR T cells to identify high-activity CARs. Finally, we performed cytotoxicity assays to identify the most effective CAR against each target and combined the CARs into a bicistronic "OR" CAR (BiCisCAR). BiCisCAR T cells effectively eliminated tumor cells expressing GPC2 or CD276. Furthermore, the BiCisCAR T cells demonstrated prolonged persistence and resistance to exhaustion when compared with CARs targeting a single antigen. This study illustrated that targeting multiple TAAs with BiCisCAR may overcome heterogenous expression of target antigens in solid tumors and identified a potent, clinically relevant CAR against NB. Moreover, our multimodal approach integrating competitive expansion, P-COCC, and cytotoxicity assays is an effective strategy to identify potent CARs among a pool of candidates.