Abstract
617
Objectives: While the immune system has the capacity to inhibit tumor growth, tumor cells can evade death through the expression of PD-L1 (B7-H1), which exerts an immunosuppressive effect. PD-L1, upregulated through a variety of mechanisms, binds its cognate receptor, PD-1 on T-cells, thus inactivating the cytotoxic effect of T-cells. PD-L1 expression levels, assessed by immunohistochemistry (IHC), have been positively correlated with patient response. Immunotherapies that block immune checkpoint inhibition have had stunning success in some but not all patients, putatively due to inaccurate assessment of PD-L1 levels. Therefore, non- invasive imaging of PD-L1 that avoids sampling errors could serve as a predictive biomarker for immunotherapy. To clarify the role of imaging PD-L1 in preclinical models, we investigated changes of PD-L1 density on irradiated and control (non-irradiated) tumors with PET/CT imaging, flow cytometry (FC), and IHC
Methods: A syngeneic model of head and neck squamous cell carcinoma (MEER, murine HPV positive HNSCC) was used as a model in immuno-competent mice with irradiation of the neck tumor (2GyX10) where the flank tumor served as a non-irradiated control. Zr-89-DFO-anti- mouse-PD-L1 (clone 10F.9G2) was administered after irradiation, followed by PET imaging and a biodistribution. A second cohort of mice with the same irradiation protocol was sacrificed for FC analysis and IHC of the dissected tumors.
Results: AT 96 h post injection, FC analysis showed significant radiation induced upregulation of PD-L1 in the irradiated tumors (neck) relative to the non-irradiated tumors (flank, ~3.5 fold). Moreover, the upregulation was only observed on tumor infiltrating CD45+ cells and not on CD45- tumor cells. The net increase of PD-L1 expression was due to the influx of CD45+ tumors cells which increased from ~20% to ~40% in the irradiated tumor. The biodistribution (n=6/group) showed uptake in PD-L1-enriched spleen (%ID/g ± SD, 32 ± 11) and thymus (%ID/g, 16 ± 7) with statistically significant differences between irradiated tumor (neck, %ID/g ± SD, 19±5.4%), and the non-irradiated control tumor (flank, %ID/g ± SD, 6.8±1.2%, P < 0.05, paired t-test) which were readily visualized by pre-clinical PET/CT. The tumor/blood ratios were relatively high and statistically significant (neck, 19±0.5, flank, 11 ±0.80, P < 0.05). Quantitative IHC analysis of the dissected tumors did not show an increase of the irradiated tumor relative to the non-irradiated tumor.
Conclusion: These results demonstrate that the89Zr-labeled anti-mouse Mab is sensitive to changes in PD-L1 densities after irradiation and localizes to tissues known to be enriched in PD- L1. The increases in PD-L1 expression are on the infiltrating myeloid cells and not on CD45- cells and the results from flow cytometry correlated well with the imaging results. The IHC data did not agree with biodistribution, imaging and flow cytometry results, possibly due to heterogeneous PD-L1 expression within the tumor, which further emphasizes the value of PET imaging in assessing PD-L1 status of the entire tumor. Research Support: This work was supported by National Institute of Health grants R01 CA206517, DE019727, P50 CA097190, T32 CA060397 (RLF), the University of Pittsburgh Cancer Institute award P30 CA047904 (RLF), and 1R21EB023364 (WBE).