Hierarchical clustering analysis identifies metastatic colorectal cancers patients with more aggressive phenotype

摘要: // Giuseppina Opinto 1, * , Nicola Silvestris 2, Matteo Centonze 1 Giusi Graziano 3 Rosamaria Pinto 4 Livia Fucci 5 Giovanni Simone and Anita Mangia Functional Biomorphology Laboratory, IRCCS-Istituto Tumori, Bari 70124, Italy 2 Medical Oncology, Scientific Direction, Molecular Genetics Pathology Department, These authors have contributed equally to this work Correspondence to: Mangia, email: a.mangia@oncologico.bari.it Keywords: metastatic colorectal cancer, unsupervised hierarchical clustering analyses, immunohistochemistry, TMA, biomarker expression Received: July 27, 2017     Accepted: August 17, Published: September 23, 2017 ABSTRACT A large percentage of cancer (mCRC) patients presents metastasis at the time diagnosis. In last years, great efforts been made in treatment these with identification different phenotypes playing a key role definition new systemic therapies. Unsupervised analysis (HCA) was performed considering clinicopathological characteristics 51 mCRCs. Using immunohistochemistry on tissue microarrays, we assessed β-catenin, NHERF1, RASSF1A, TWIST1, HIF-1α proteins tumors paired liver metastases. We also analyzed RASSF1A methylation status samples same patients. HCA distinguished Group characterized by features. higher number positive lymph nodes ( p =0.0139), poorly differentiated grade <0.0001) high extent tumor spread =0.0053) showing more aggressive phenotype compared 2. both Groups, found common “basal” condition level nuclear TWIST1 <0.0001 ) cytoplasmic β-catenin than Furthermore, hypermethylation overexpression p= 0.0354) metastases tumors. conclusion, identifies mCRC phenotype. Moroever, our results support important contribution progression disease oncogenic evidences, should provide relevant information concerning biology and, as consequence, potential therapeutic approaches.