152 Immuno-oncology protein biomarker panels provide greater biological insights and improved patient stratification

<h3>Background</h3> Immunotherapeutic approaches have achieved durable treatment responses in certain cancer patients and raise hope for further advancements. However, the lack of responses and acquired drug resistance remain major clinical challenges. An integrated longitudinal analyses of circulatory plasma proteins, combined with TME transcriptomics, showed prognostic significance and utility of plasma proteomics in biomarker discovery for cancer immunotherapy. Yet, successful implementation of blood biomarker profiling relies on technologies that allow scalable multiplexing while delivering consistent, high data quality. <h3>Methods</h3> Olink’s proximity extension assay (PEA), established on the biding of two antibodies labeled with complementary oligonucleotides to the target protein and a genomic readout, enables high specificity and sensitivity. PEA therefore ensures scalable multiplexing, allowing analyses from thousands to dozens of proteins simultaneously with uncompromised data quality. Here we used the Olink Flex library, composed of a wide range of pre-validated protein biomarkers to rapidly build custom immuno-oncology panels, which were then applied for analyzing target proteins in the plasma of cancer patients receiving immune checkpoint blockade (ICB). <h3>Results</h3> Our data, derived from the large-scale proteomic analyses of circulatory proteins measured in the plasma of melanoma, lung and other solid tumors, identified several proteins associated with ICB resistance such as MIA, LIF, ST2, IL6 and IL8. Increased plasma levels of proteins encoded by interferon-stimulated genes (such as CCL2, CCL7, IL6, IL15) were shown to correlate with poor treatment response in melanoma and renal cell carcinoma, confirming transcriptomics data and highlighting the context dependent role of IFN signaling in tumor immunity. Furthermore, proteomic profiling of ICB-treated melanoma patients revealed that an early increase in circulating IFNγ and relative levels of CXCL9/10/11 in the plasma may predict immune related adverse events. The levels of these proteins were further evaluated and used to establish absolute threshold and compose Immuno-oncology protein biomarker panels. <h3>Conclusions</h3> Next generation proteomics has been extensively used to elucidate mechanisms of treatment resistance and predict patient responses to immunotherapy through deep immunoprofiling of circulatory plasma proteins. The ability to validate discovered protein biomarkers using single scalable technology such as PEA will accelerate translational biomarkers discovery to improve clinical management and development of new therapies.