Dataset Information

Molecular heterogeneity of multinodular hepatocellular carcinoma [gene expression]

ABSTRACT: Genome-wide gene expression profiles of multiple hepatocellular carcinoma tumors and non-tumor liver tissues from the same patients. Introduction: Hepatocellular Carcinoma (HCC) is the second cause of cancer-related death worldwide. Understanding tumor heterogeneity is relevant for the pathogenesis and treatment of this neoplasm. Particularly, whether multinodular tumors result from intrahepatic metastasis (IM or clonal tumors) or de novo cancers (synchronic tumors) have direct implications in treatment decisionmaking for resection and transplantation. We aimed to assess the genomic heterogeneity of multifocal HCC using single-nucleotide polymorphism (SNP) and gene expression analyses. Methods: Among 544 HCC patients consecutively transplanted at Mount Sinai Hospital between 1990- 2007, we selected 18 patients with a total of 42 multinodular tumors (2-3 non-satellite foci HCCs) for this molecular study. Formalin-fixed blocks were collected for each tumor along with corresponding clinico-pathological data. SNP array and gene expression profiling was generated. Clonality was defined by measuring the similarity of genome-wide copy number variation (CNV) profiles between two nodules. Unsupervised hierarchical clustering based on gene expression data was performed to measure genetic proximity between each tumor. Prediction of previously published signatures was performed using Nearest Template Prediction (NTP). Association of clonality with clinic-pathological parameters was assessed. Results: A total of 42 tumors have been analyzed (10 patients- 2 tumors; 8- 3 tumors). Most patients were male (17/18, 95%) with median age of 53-yr-old (range 39–67), HCV (10/18, 56%) or HBV infection (6/18, 33%). Median tumor size was 3 cm (range 1.5-6.5), satellites were present in 3 (17%) and vascular invasion in 11 patients (61%), respectively. CNV profiles predicted clonal tumors in 38% (6/16) and non-clonality in 62% of cases (10/16). CNV profiles of the remaining 2 cases were not informative. Clonal tumors were significantly associated with HCV infected patients (5/6 vs 3/10, p=0.007), whereas all HBV-induced HCC were synchronic tumors (0/6 vs 6/10, p=0.03). Furthermore, clonal tumors were significantly associated with presence of satellites and shorter time-torecurrence. Assessment of genetic proximity based on gene expression revealed that each clonal tumor showed proximity to its paired tumor and clustered around the same node (6/6,100%) as opposed to non-clonal (2/9,22%). When exploring molecular subclasses within clonal tumors (intrahepatic metastases), while half of them retained the molecular fingerprint, the other half switch to more aggressive subclasses. Conversely, all non-clonal tumors within the same patient belonged to distinct molecular subclasses. Conclusion: Multinodular HCCs undergoing transplantation are molecularly heterogeneous. Using CNV profiling we identified clonal multinodular tumors (true intrahepatic metastasis) in 40% of cases and de novo tumors in 60%. Clonal tumors were significantly associated with HCV infection, satellites and recurrence. Genetic proximity was observed in clonal tumors, but molecular subclasses prediction revealed that intrahepatic metastases share more aggressive subclasses in half of cases.

ORGANISM(S): Homo sapiens

PROVIDER: GSE65372 | GEO | 2018/01/28

REPOSITORIES: GEO

ACCESS DATA

Dataset's files

Source:

			Action	DRS
		Other

Items per page:

1 - 1 of 1

Similar Datasets

Project description:Genome-wide DNA copy number profiles of multiple hepatocellular carcinoma tumors and non-tumor liver tissues from the same patients. Introduction: Hepatocellular Carcinoma (HCC) is the second cause of cancer-related death worldwide. Understanding tumor heterogeneity is relevant for the pathogenesis and treatment of this neoplasm. Particularly, whether multinodular tumors result from intrahepatic metastasis (IM or clonal tumors) or de novo cancers (synchronic tumors) have direct implications in treatment decisionmaking for resection and transplantation. We aimed to assess the genomic heterogeneity of multifocal HCC using single-nucleotide polymorphism (SNP) and gene expression analyses. Methods: Among 544 HCC patients consecutively transplanted at Mount Sinai Hospital between 1990- 2007, we selected 18 patients with a total of 42 multinodular tumors (2-3 non-satellite foci HCCs) for this molecular study. Formalin-fixed blocks were collected for each tumor along with corresponding clinico-pathological data. SNP array and gene expression profiling was generated. Clonality was defined by measuring the similarity of genome-wide copy number variation (CNV) profiles between two nodules. Unsupervised hierarchical clustering based on gene expression data was performed to measure genetic proximity between each tumor. Prediction of previously published signatures was performed using Nearest Template Prediction (NTP). Association of clonality with clinic-pathological parameters was assessed. Results: A total of 42 tumors have been analyzed (10 patients- 2 tumors; 8- 3 tumors). Most patients were male (17/18, 95%) with median age of 53-yr-old (range 39–67), HCV (10/18, 56%) or HBV infection (6/18, 33%). Median tumor size was 3 cm (range 1.5-6.5), satellites were present in 3 (17%) and vascular invasion in 11 patients (61%), respectively. CNV profiles predicted clonal tumors in 38% (6/16) and non-clonality in 62% of cases (10/16). CNV profiles of the remaining 2 cases were not informative. Clonal tumors were significantly associated with HCV infected patients (5/6 vs 3/10, p=0.007), whereas all HBV-induced HCC were synchronic tumors (0/6 vs 6/10, p=0.03). Furthermore, clonal tumors were significantly associated with presence of satellites and shorter time-torecurrence. Assessment of genetic proximity based on gene expression revealed that each clonal tumor showed proximity to its paired tumor and clustered around the same node (6/6,100%) as opposed to non-clonal (2/9,22%). When exploring molecular subclasses within clonal tumors (intrahepatic metastases), while half of them retained the molecular fingerprint, the other half switch to more aggressive subclasses. Conversely, all non-clonal tumors within the same patient belonged to distinct molecular subclasses. Conclusion: Multinodular HCCs undergoing transplantation are molecularly heterogeneous. Using CNV profiling we identified clonal multinodular tumors (true intrahepatic metastasis) in 40% of cases and de novo tumors in 60%. Clonal tumors were significantly associated with HCV infection, satellites and recurrence. Genetic proximity was observed in clonal tumors, but molecular subclasses prediction revealed that intrahepatic metastases share more aggressive subclasses in half of cases.

Project description:Genomic analysis of many cancers has led to the identification of novel targets and the development of personalized, targeted therapies. Unfortunately, in the majority of rare tumors, this type of analysis can be particularly challenging. Large series of specimens for analysis are simply not available, allowing recurring patterns to remain hidden. Clinical specimens typically contain variable degrees of non-tumor cells that can mask a potentially critical genomic signature, leaving important clinically relevant events undetected. When analysis is limited to a smaller number of specimens, the effects of heterogeneity within each sample is magnified. In light of these challenges, we used DNA content based flow cytometry to isolate clonal tumor populations from a series of rare cancers for genomic analysis: intrahepatic cholangiocarcinoma, anal carcinoma, adrenal leiomyosarcoma, and pancreatic neuroendocrine tumors. These purified clonal populations are then subject to high definition measurement of copy number aberrations by objectively measuring the height and boundaries of amplicons and by discriminating homozygous from partial deletions. Ranking of these events by copy number facilitates the identification of highly selected aberrations. This approach can garner useful information from a single biopsy. In the cases we describe, several potential therapeutic targets were identified and genomic aberrations correlated with the phenotypic behavior. We propose that clonal genomic analysis can generate unique hypotheses and guide the development of clinical advances for these and other rare tumors. We applied DNA content based flow sorting to isolate the nuclei of clonal populations from tumor biopsies. We coupled this strategy with oligonucleotide array CGH (aCGH) thereby obtaining high definition genomic profiles of clonal populations from different rare tumors including pancreatic neuroendocrine cancers, adrenal leiomyosarcoma, anal carcinoma, and cholangiocarcinoma.

Dataset Information

Molecular heterogeneity of multinodular hepatocellular carcinoma [gene expression]

Dataset's files

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets