Abstract PR06: PET imaging of biomarkers in immuno-oncology for response prediction and treatment monitoring

Abstract

Background: Cancer immuno-therapy has been established as a powerful way of treating several types of cancer. Expression of biomarkers on tumor and associated immune cells in the tumor microenvironment (TME) seems to reflect clinical outcome. The objective of the current work was to develop a toolbox for non-invasive PET imaging of PD-L1 expression and imaging of CD4+ and CD8a+ tumor-infiltrating lymphocytes for response prediction and treatment monitoring of immune stimulating therapies. Methods: PET radiotracers were generated from intact antibodies (anti-PD-L1, clone 6E11) or antibody fragments (R-anti-M-CD8a+ or R-anti-M-CD4+) and labeled with 89Zr or 64Cu. Syngeneic mouse models were selected based on their respective TME as analyzed by flow cytometry and immunohistochemistry (IHC). PET/CT imaging with 64Cu-NOTA-CD8a-F(ab)’2 in tumor-bearing CT26 mice after treatment with external radiation therapy (XRT) and anti-CTLA-4 was performed for treatment monitoring. Prediction of response to therapy with XRT and anti-PD-L1 was evaluated with 89Zr-DFO-6E11 in 4T1 and CT26 syngeneic mouse models. In vivo imaging data was compared with IHC and flow cytometry. Finally, murine CD3+ T-cells were directly labeled with 89Zr and evaluated in vivo in immunocompetent- and incompetent mice. The labeling efficiency and viability of the cells were evaluated ex vivo, and in vivo cell tracking was performed by longitudinal PET/CT imaging after injection of labeled T-cells in naive immune-competent mice. Results: Optimized protocols for labeling of anti-PD-L1 and fragments directed at CD8a+ and CD4+ with 89Zr or 64Cuwere developed with high radiochemical yield and RCP of >99%. PET/CT imaging with all tracers was performed to compare tumor, spleen and lymph node uptake. Blocking of endogenous CD8a+ and PD-L1 was successfully reflected by PET imaging with 89Zr-DFO-CD8a-F(ab)’2 and 89Zr-DFO-6E11, respectively. Fractionated XRT induced infiltration of CD8a+ cells in spleen and CT26 tumors which was detected by in vivo89Zr-DFO-CD8a-F(ab)’2 PET imaging and ex vivo by IHC analysis. Increased tumor uptake of 64Cu-NOTA-CD8a-F(ab)’2 was seen in CT26 tumor bearing mice treated with fractionated XRT and anti-CTLA-4, and tumor growth was effectively inhibited in treatment groups with high uptake. Imaging with 89Zr-labeled anti-PD-L1 in the CT26 tumor model was able to predict subsequent response to PD-L1 therapy. Finally, longitudinal PET/CT imaging was able to track CD3+ T-cells directly labeled with 89Zr for up to 7 days following intra venous administration in immunocompetent mice. Conclusion: The developed radiotracers enable whole-body visualization of CD8a+ and CD4+ T-cells and PD-L1 expression in various syngeneic mouse models. These radiotracers can be used as tools to predict and monitor immune stimulating therapies in preclinical drug development. Furthermore, PET imaging of direct labeled T-cells is an attractive tool for investigating the distribution of cell-based therapies. Citation Format: Carsten H. Nielsen, Lotte K. Kristensen, Camilla Christensen, Andreas Kjaer. PET imaging of biomarkers in immuno-oncology for response prediction and treatment monitoring [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr PR06. doi:10.1158/1535-7163.TARG-19-PR06