Identification of Microvascular Invasion Biomarkers in Hepatocellular Carcinomas by MALDI Imaging Mass Spectrometry
Nicolas Pote, Theodore Alexandrov, Samira Laouirem, Jacques Belghiti, Pierre Bedossa, Valerie Paradis. Beaujon Hospital, Assistance Publique-Hôpitaux de Paris, Clichy, France; INSERM U773, Beaujon Hospital, Clichy, France; University of Bremen, Bremen, Germany; University Denis Diderot, Paris, France
Background: Microvascular invasion (mVI), a major predictive factor of post-operative tumoral recurrence in patients with hepatocellular carcinoma (HCC), is only detectable after surgery on histological examination of the surgical specimen. So far, there is no reliable biomarker to predict the presence of mVI prior to surgical procedures. MALDI Imaging Mass Spectrometry (IMS) represents a new analytical tool to directly provide the relative abundance and spatial distribution of the whole proteins expressed in a tissue section. The aim of this study was to compare, using MALDI IMS, the tissue proteome of HCC without and with mVI in order to identify biomarkers of mVI.
Design: A total of 56 HCC from various etiologies obtained from surgical specimen, for which clinicopathological data and frozen samples were available, were retrospectively collected. All patients were cirrhotics, and none of them received any neo-adjuvant treatment. After histological examination, two groups of tumors were defined (26 HCC without mVI; 30 HCC with mVI). Cryosectioning was done to yield tissue sections analysed by a pathologist to determine tissue morphology and mirror sections for MALDI IMS. A statistical comparative analysis, using a cross classification model, was then performed in order to identify protein peaks differentially expressed between both groups.
Results: 24 discriminant protein peaks were differentially expressed between both groups, of which 11 were discriminant in more than 30/56 cross classifications. 2 peaks increased in HCC without mVI, whereas 22 increased in HCC with mVI. The latter, which could be used as tissue mVI biomarkers, are under selection for protein characterization. Tissue distribution analysis of the two most discriminant peaks (m/z 10,042 and 8,904) showed that m/z 10,042 was preferentially expressed in tumor cells, whereas m/z 8,904 was overexpressed in the stroma. Interestingly, we have previously described m/z 8,904 as a serum biomarker of HCC (Paradis et al., Hepatology 2005).
Conclusions: These results highlight the potential of MALDI IMS to uncover new biomarkers in liver carcinogenesis, and to allow their tissue localization. The identification of mVI biomarkers would be helpful in the therapeutic strategy of patients with HCC.
Category: Special Category - Pan-genomic/Pan-proteomic approaches to Cancer
Tuesday, March 20, 2012 9:30 AM
Poster Session III # 272, Tuesday Morning