124I-Labeled Monoclonal Antibody and Fragment for the Noninvasive Evaluation of Tumor PD-L1 Expression In Vivo.

ABSTRACT

Lung cancer is a highly heterogeneous cancer and is divided broadly into small and nonsmall cell lung cancer (SCLC or NSCLC). In all NSCLC patients, it is estimated that 50%-60% are programmed cell death ligand 1 (PD-L1) positive, and anti-PD-1/PD-L1 therapies have shown their clinical application prospects in advanced NSCLC. To avoid unnecessary adverse effects and provide anti-PD-1/PD-L1 therapy to the most appropriate patient population, the PD-L1 expression in patients preparing for treatment must be evaluated accurately and in real time. In this study, we noninvasively evaluate the PD-L1 expression in an NSCLC xenograft using 124I-labeled F(ab')2 fragments of durvalumab (Durva) and compared it with the 124I-labeled intact antibody in terms of the biodistribution and dosimetry. The aim is to develop a nuclide labeled molecular probe with better performance for PD-L1 immunoPET imaging. After cleaving using IdeS protease, the F(ab')2 fragments of Durva were labeled with 124I. The radioligand showed a high radiochemical purity (>96%) and outstanding stability. Western blot, quantitative real-time polymerase chain reaction, and flow cytometry were performed on the two selected NSCLC cell lines to measure the in vitro PD-L1 expression. The H460 cells showed a much higher PD-L1 expression than the A549 cells, both at the protein level and the mRNA level. In the following cell binding experiment and binding specificity assay, the labeled radioligand showed good affinity to high PD-L1 expression cells and could be blocked with excess unlabeled intact Durva. The results of the biodistribution and the positron emission tomography (PET) image showed that the peak tumor uptake of 124I-Durva-F(ab')2 was close to 124I-Durva, but much earlier (5.29 ± 0.42% ID/g for 124I-Durva-F(ab')2 at 12 h vs 5.18 ± 0.73% ID/g for 124I-Durva at 48 h). Compared with 124I-Durva, an accelerated blood clearance was observed for 124I-Durva-F(ab')2. The faster blood clearance allowed for a higher tumor-to-background ratio, which was reflected on the image in contrast. The H460 tumors showed excellent contrast as early as 4 h after injection with 124I-Durva-F(ab')2, and for 124I-Durva, the xenograft could not be distinguished clearly until 24 h after injection. Interestingly, 124I-Durva-F(ab')2 showed lower accumulations compared to other metal isotopes labeled PD-L1 antibodies in bone, liver, spleen etc., which will be beneficial for metastasis detection. Another benefit of accelerated blood clearance was a reduction in the radiation dose. According to the results of the OLINDA/EXM, the effective dose for the total body of 124I-Durva was 4.25-times greater than that of 124I-Durva-F(ab')2 (186 µSv/MBq vs 43.8 µSv/MBq). All of these data indicated that 124I-Durva-F(ab')2 is a promising immunoPET tracer for evaluating the in vivo PD-L1 levels in an NSCLC model and is expected to be successful in future clinical application.