(F-212) Exosomal PD-L1-microRNA Pairs Detected by an Exo-PROS Liquid Biopsy Assay Predict Treatment Responses to Immune Checkpoint Inhibitor

Introduction: Immune checkpoint inhibitors (ICI) based immunotherapies that target PD-1 and PD-L1 have revolutionized treatment for many advanced cancers including lung cancer and melanoma. However, only 20-30% of patients receive durable clinical benefits. Tumor PD-L1 expression is used to select patients for ICI therapy, but it is not a sensitive standalone biomarker in predicting ICI responses. Tissue biopsy is an invasive procedure, which cannot be performed frequently to monitor treatment responses. To overcome these challenges, we developed the innovative exosome protein microRNA one stop (Exo-PROS) assay, which in situ and simultaneously detects two distinct types of exosomal biomarkers to greatly enhance predictive accuracy. In this study, exosomal PD-L1-miR-21 and PD-L1-miR-34a pairs successfully distinguished responders from non-responders to ICI therapy with AUC of 0.92, significantly higher than tumor PD-L1 expression, predictive in only 28.9% of cases.

Materials and Methods: Exo-PROS assay is a surface plasmon resonance (SPR)-based, molecular beacon (MB)-assisted liquid biopsy test (Fig. 1a). When 10uL serum was applied on the biochip, anti-PD-L1 antibodies capture exosomes expressing PD-L1 (i.e. PD-L1+ exosomes), which changed the local reflective index and allowed optical detection of exosomal PD-L1 levels via SPR. Then MB, sensing probes for microRNAs were added to detect miR-21 or miR-34a in PD-L1+ exosomes. The binding between MB and miR-21 or miR-34a separated the fluorescence dye from the quencher, and thus restored the fluorescence signals of MB, which were used to calculate microRNA expression. The levels of exosomal PD-L1-miR-21 and PD-L1-miR-34a pairs were measured in sera from lung cancer and melanoma patients (16 responders and 37 non-responders) at pre-treatment and 3-6 weeks post-treatment time points. The predictive performances (sensitivity, specificity, AUC) of exosomal PD-L1-miRNA pairs were determined by ROC analysis.

Results, Conclusions, and Discussions: At the pre-treatment baseline, responders had significantly lower levels of exosomal PD-L1 but higher levels of miR-21 and miR-34a than non-responders. At 3-6 weeks post-treatment, we observed increased exosomal PD-L1 levels but decreased miR-21 and miR-34a levels in responders compared with non-responders (Fig. 1b, c). When the baseline expression of PD-L1-miR-21/miR-34a pairs was used as the biomarkers, we distinguished responders from non-responders with a sensitivity of 0.7, specificity of 0.75 and AUC of 0.79, much higher than tumor PD-L1 expression (28.9%). When the expression fold changes of exosomal PD-L1-miR-21/miR-34a pairs between 3-6 weeks post-treatment and pretreatment baseline were used as the biomarkers, the predictive performance was greatly improved with a sensitivity of 0.73, specificity of 1.0 and AUC of 0.92 in distinguishing responders from non-responders. These results demonstrated that Exo-PROS assay is a potent companion test to screen patients for ICI therapy.

Acknowledgements (Optional): :

References (Optional): :