Abstract
Introduction: Programmed-death 1 (PD-1) is expressed by T-cells and is a major co-inhibitory immune checkpoint. Its ligand PD-L1 is expressed by tumor cells and myeloid phagocytic cells. By upregulating PD-L1, tumors are capable of escaping immune attack. Clinical trials with anti-PD-1/PD-L1 immune checkpoint inhibitors have shown impressive and durable responses. However, a large number of non-responding patients is unnecessarily exposed to ineffective, expensive treatment, and its associated side effects. To use these drugs more efficiently, there is an urgent need for a predictive biomarker. Currently, PD-1 and PD-L1 is measured immunohistochemically on a tumor biopsy. However, PD-1 and PD-L1 demonstrate considerable intra- and intertumoral heterogeneity. Moreover, their expression is dynamic and may change because of tumor progression and previous treatment. Therefore, the aim of this study is to develop an imaging technique to non-invasively assess PD-L1 expression and PD-1+ tumor infiltrating T-lymphocytes (TILs). This technique can be used to longitudinally monitor PD-1/PD-L1 of whole tumor lesions and their metastases.
Material and Methods: Anti-murine PD-1 and PD-L1 antibodies were radiolabeled with In-111. The in vitro binding characteristics were assessed using EL-4 (PD-1+) and Renca (PD-L1+) cells. Subsequently, the optimal antibody dose (1 - 1,000 μg) and time point (4, 24, 48, 72, 168 h) for imaging were assessed by ex vivo biodistribution and microSPECT/CT imaging studies, in mice with syngeneic s.c. implanted tumors with known variable expression of PD-L1; Renca, 4T1, CT26, or LLC1 tumors. Imaging findings were validated with immunohistochemistry for the expression of PD-L1, CD8, and PD-1.
Results: In-111-labeled antibodies specifically bound to PD-1 and PD-L1 expressing cell lines in vitro. The optimal antibody dose to target PD-1 was 3 μg. MicroSPECT/CT showed heterogeneous uptake of the radiolabeled anti-PD-1 antibody in 4T1 tumors. Ex vivo biodistribution studies showed a tumor uptake of 14 ± 4%ID/g. Other organs that showed uptake of the anti-PD-1 antibody were thymus (10 ± 2%ID/g), spleen (10 ± 3%ID/g), and lymph nodes (14 ± 5%ID/g). The optimal antibody dose to target PD-L1-expressing Renca tumors was 30 μg and highest tumor-normal tissue contrast was obtained 24 to 72 h post injection. Tumor uptake in Renca, 4T1, CT26, and LLC1 was 15 ± 5%ID/g, 16 ± 6%ID/g, 11 ± 6%ID/g, and 6 ± 3%ID/g, respectively, and correlated with PD-L1 expression on immunehistochemistry. Enhanced uptake was also observed in spleen (17 ± 2%ID/g), brown fat (19 ± 2%ID/g), duodenum (10 ± 2%ID/g), and lymph nodes. Immunohistochemistry confirmed PD-L1 expression in these organs.
Conclusion: These studies demonstrate that PD-1 and PD-L1 can be imaged with microSPECT/CT. These techniques can potentially be used to non-invasively select patients that are most likely to respond to immune checkpoint inhibitor therapy and can monitor PD-1/PD-L1 prior to and during conventional anti-cancer treatment and disease progression.
Citation Format: Sandra Heskamp, Janneke D.m. Molkenboer-Kuenen, Erik H. Aarntzen, Otto C. Boerman. Non-invasive imaging of the PD-1/PD-L1 pathway in syngeneic murine tumor models [abstract]. In: Proceedings of the Second CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; 2016 Sept 25-28; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2016;4(11 Suppl):Abstract nr B077.