Abstract 4818: Validation and clinical utility of a human hepatocellular carcinoma metastasis signature as prognostic classifier

Abstract

Abstract Hepatocellular carcinoma (HCC) is the most frequent malignant tumor in the liver and the third leading cause of cancer death worldwide. The high mortality of HCC is mainly caused by intra-hepatic metastasis or the high rate of recurrence after surgery. Despite potentially curative surgical resection, eligibility is scarce and patients suffer from a high recurrence rate. Recently, we developed a tumor-based 153-gene signature that is predictive of HCC portal venous metastasis and survival in 40 HCC patients (Ye et al, Nat Med, 9: 416-23, 2003). This signature was based on cDNA microarray profiling of 20 well defined HCC cases, in which ten developed tumor thrombi in the major branch of portal vein and ten had no metastasis at one year follow-up. In this study, we sought to validate this signature for its utility in predicting HCC survival using two independent cohorts. The first cohort consisted of 247 predominantly HBV+ HCC cases analyzed by the Affymetrix U133A 2.0 platform. The second cohort included 139 cases with mixed etiology analyzed by the Human Array-Ready Oligo Set (version 2) array platform. We employed a survival risk prediction algorithm with 10-fold cross validation and 1000 random permutation to test whether the 153-gene signature is associated with survival. We found that this gene signature was able to predict prognosis in both independent HCC cohorts. Importantly, the risk prediction was independent of the microarray platform and etiology. In addition, survival prediction was successful in patients with small (<5 cm in diameter) and/or solitary tumor and the signature performed well on patients particularly with early recurrence (<2 years). Thus, the 153-gene signature was validated in three independent cohorts to be related to HCC patient survival. Furthermore, the risk prediction performed independently of clinical characteristics and microarray platform. We recommend that this signture may serve as a molecular diagnostic test useful to assess the risk that an HCC patient will recur within 2 years after resection, particularly for those with tumors <5cm and in a solitary presentation. Supported by NIH intramural research funds Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 4818.