Anti-4-1BBxPDL1 Bispecific Antibody Reinvigorates Tumor-Specific Exhausted CD8<SUP>+</SUP> T Cells and Enhances the Efficacy of Anti-PD1 Blockade

Abstract

Purpose: To overcome the limited efficacy of immune checkpoint blockade, there is a need to find novel cancer immunotherapeutic strategies for the optimal treatment of cancer. The novel anti-4-1BBxPDL1 bispecific antibody-ABL503 (also known as TJ-L14B)-was designed to simultaneously target PDL1 and 4-1BB and demonstrated strong antitumor T-cell responses without considerable toxicity. In this study, we investigated the mechanisms by which the combination of ABL503 and anti-PD1 blockade affected the reinvigoration of exhausted tumor-infiltrating CD8(+) T cells (CD8(+) TIL) and antitumor efficacy. Experimental Design: Single-cell suspensions of hepatocellular carcinoma and ovarian cancer tissues from treatment-naive patients were used for immunophenotyping of CD8(+) TILs and in vitro functional assays. Humanized hPD1/hPDL1/h4-1BB triple-knock-in mice were used to evaluate the effects of ABL503 and anti-PD1 blockade in vivo. Results: We observed that ABL503 successfully restored the functions of 4-1BB(+) exhausted CD8(+) TILs, which were enriched for tumor-specific T cells but unresponsive to anti-PD1 blockade. Importantly, compared with anti-PD1 blockade alone, the combination of ABL503 and anti-PD1 blockade further enhanced the functional restoration of human CD8(+) TILs in vitro. Consistently, the combination of ABL503 with anti-PD1 in vivo significantly alleviated tumor growth and induced enhanced infiltration and activation of CD8(+) TILs. Conclusions: ABL503, a PDL1 and 4-1BB dual-targeting bispecific antibody, elicits pronounced additive tumor growth inhibition, with increased infiltration and functionality of exhausted CD8(+) T cells, which in turn enhances the anticancer effects of anti-PD1 blockade. These promising findings suggest that ABL503 (TJ-L14B) in combination with PD1 inhibitors will likely further enhance therapeutic benefit in clinical trials. See related commentary by Molero-Glez et al., p. 3971