Global Protein Aberration Score (glopas): A Comprehensive Risk Score to Predict Hepatocellular Carcinoma Biology and Estimate Patients’ Survival.

287 Background: Existing hepatocellular carcinoma (HCC) staging systems use a small number of empirically selected clinical/biological variables. We hypothesize that GloPAS (global protein aberration score) will capture a more global measurement of disease biology that can partition HCC into three subsets: (1) essentially normal profiles, (2) extremely aberrant profiles, and (3) slightly aberrant profiles. Methods: We collected plasma samples and clinical data prospectively from 767 HCC patients (pts) and 200 healthy controls, and quantified 317 pts using Myriad RBM CLIA-certified panel, Austin, TX. We applied a deconvolution algorithm, originally developed for determining percent normal contamination for tumor, to quantify the degree of global protein aberration for each pt relative to normal controls. We defined three distinct groups of pts with low ( < 0.3), medium (0.3-0.8), and high ( > 0.8) GloPAS and assessed GloPAS association with overall survival (OS) and other prognostic factors using log-rank tests and compared the prognostic abilities of GloPAS vs. existing systems using concordance index(C-index). We developed a single-sample GloPAS (ssGloPAS) using an algorithm that can be applied to single sample setting. Results: Although determined without using information about OS or any pt-level clinical or demographic factors (i.e. unsupervised), GloPAS showed remarkable prognostic separability (low/med/high GloPAS, with median OS 38.2mo/18.3mo/7.1mo, p < 0.0001). GloPAS prognostic ability was far above any existing HCC staging system (C-index = 0.75 vs. 0.58-0.70 p < 0.0001), demonstrating even more prognostic information than key factors as metastatic status and vascular invasion. The ssGloPAS was able to recapitulate the global signal captured by the GloPAS score with a much smaller subset of 14 proteins. Conclusions: GloPAS significantly improved prediction of OS and prognostic stratification of HCC. After further validation, ssGloPAS could be used to guide therapy decisions and stratify HCC pts in clinical trials. The novel concept underlying GloPAS methodology can be applied to other cancers to build disease-specific prognostic scores.