Novel quantitative immunohistochemical analysis for evaluating PD-L1 expression with phosphor-integrated dots for predicting the efficacy of patients with cancer treated with immune checkpoint inhibitors.

ABSTRACT

Introduction: Programmed cell death ligand 1 (PD-L1) expression in tumor tissues is measured as a predictor of the therapeutic efficacy of immune checkpoint inhibitors (ICIs) in many cancer types. PD-L1 expression is evaluated by immunohistochemical staining using 3,3´-diaminobenzidine (DAB) chronogenesis (IHC-DAB); however, quantitative and reproducibility issues remain. We focused on a highly sensitive quantitative immunohistochemical method using phosphor-integrated dots (PIDs), which are fluorescent nanoparticles, and evaluated PD-L1 expression between the PID method and conventional DAB method. Methods: In total, 155 patients with metastatic or recurrent cancer treated with ICIs were enrolled from four university hospitals. Tumor tissue specimens collected before treatment were subjected to immunohistochemical staining with both the PID and conventional DAB methods to evaluate PD-L1 protein expression. Results: PD-L1 expression assessed using the PID and DAB methods was positively correlated. We quantified PD-L1 expression using the PID method and calculated PD-L1 PID scores. The PID score was significantly higher in the responder group than in the non-responder group. Survival analysis demonstrated that PD-L1 expression evaluated using the IHC-DAB method was not associated with progression-free survival (PFS) or overall survival (OS). Yet, PFS and OS were strikingly prolonged in the high PD-L1 PID score group. Conclusion: Quantification of PD-L1 expression as a PID score was more effective in predicting the treatment efficacy and prognosis of patients with cancer treated with ICIs. The quantitative evaluation of PD-L1 expression using the PID method is a novel strategy for protein detection. It is highly significant that the PID method was able to identify a group of patients with a favorable prognosis who could not be identified by the conventional DAB method.