PD-L1 Checkpoint Blockade Augments Anti-Tumor Immune Response of GD2 or HER2-Bsab Ex Vivo Armed T-Cells (EVAT) Therapy in Osteosarcoma

Background

ExVivo Armed T-cells (EVAT) carrying zeptomoles (10^-21M) of T-cell engaging GD2-bispecific antibody (GD2-EVAT) or HER2-bispecific antibodies (HER2-EVAT) have potent anti-tumor activity against GD2(+) and/or HER2(+) solid tumors. Strategies to further optimize this approach are highly relevant. PD-1 is a key immune checkpoint receptor expressed mainly by activated T-cells and mediates immune suppression by binding to its ligands PD-L1 or PD-L2. Upregulation of PD-L1 has been found in many cancers including osteosarcoma and associated with aggressive disease and poor outcome. While the use of immune checkpoint inhibitors (ICIs) seems logical, the ideal timing when combined with T-cell engaging bispecific antibody (T-BsAb) or EVAT has yet to be defined. Here, we described the effects of anti-PD-1 or anti-PD-L1 antibodies on GD2-EVAT or HER2-EVAT therapy and explored the impact of its timing in the treatment of osteosarcoma which is GD2(+), HER2(+) and PD-L1(+).

Methods

GD2-BsAb and HER-BsAb were built using the IgG(L)-scFv format (Can Immunol Res, 3:266, 2015, Oncoimmunology, PMID:28405494). T-cells from healthy volunteer donors were isolated, and cultured ex vivo in the presence of CD3/CD28 beads plus 30 IU/mL of interleukin 2 (IL-2). Between day 7 and day 14, activated T-cells (ATCs) were harvested and armed for 20 minutes at room temperature with GD2-BsAb or HER2-BsAb. In vivo anti-tumor activity against GD2(+), HER2(+), and PD-L1(+) osteosarcoma cell line xenografts was tested in BALB-Rag2^-/-IL-2R-γc-KO mice. Anti-human PD-1 antibody (pembrolizumab, anti-PD-1) or anti-human PD-L1 antibody (atezolizumab, anti-PD-L1) were tested for synergy with GD2-EVAT or HER2-EVAT therapy.

Results

The PD-1 expression increased among T-cells that circulated in the blood, that infiltrated the spleen or the tumor after EVAT therapy. While anti-PD-L1 combination therapy with GD2-EVAT or HER2-EVAT improved anti-tumor response against osteosarcoma (P=0.0123 and P=0.0004), anti-PD-1 did not (all P>0.05). The addition of anti-PD-L1 significantly increased T-cell survival in blood and T-cell infiltration of tumor when compared to GD2-EVAT or HER2-EVAT alone (all P<0.0001). Treatment of GD2-EVAT or anti-PD-L1 plus GD2-EVAT downregulated GD2 expression on tumors, but anti-PD-1 plus GD2-EVAT did not. For the next step we tested the impact of different combination schedules of ICIs on GD2-EVAT therapy. Concurrent anti-PD-1 (6 doses along with GD2-EVAT therapy) interfered with GD2-EVAT, while sequential anti-PD-1 (6 doses after GD2-EVAT) did not make a significant effect (P>0.05). On the other hand, while the concurrent use of anti-PD-L1 did not show benefit on GD2-EVAT, sequentially administered anti-PD-L1 produced a significant improvement in tumor control when compared to anti-PD-L1 or GD2-EVAT alone (P=0.002 and P=0.018). When anti-PD-L1 treatment was extended (12 doses after GD2-EVAT), the anti-tumor effect was most pronounced compared to GD2-EVAT alone (P <0.0001), which translated into improved survival (P=0.0057). These in vivo anti-tumor responses were associated with increased CD8(+) tumor infiltrating lymphocytes (TILs) of tumor.

Conclusion

In the arming platform, large numbers of target-specific T-cells can be generated, and this EVAT therapy is a highly effective cellular treatment with high potency in preclinical models. In addition, the advantage of ex vivo cytokine release following T-cell arming and activation could reduce or avoid life threatening cytokine storm if such activation was to proceed in vivo. Adoptive T-cell therapy induced immune response upregulates the inhibitory immune checkpoint PD-1/PD-L1 pathway, and combination treatment with anti-PD-L1 antibody, especially when combined as sequential therapy and continuously treated, significantly improved anti-tumor effect of EVAT, partly through increase in CD8(+) TILs infiltration.