Project description:Lung cancers are a heterogeneous group of diseases with respect to biology and clinical behavior. So far, diagnosis and classification are based on histological morphology and immunohistological methods for discrimination between two main histologic groups: small cell lung cancer (SCLC) and non-small cell lung cancer which account for 20% and 80% of lung carcinomas, respectively. While SCLCs express properties of neuroendocrine cells, NSCLCs, which are divided into the three major subtypes adenocarcinoma, squamous cell carcinoma and dedifferentiated large cell carcinoma, show different characteristics such as the expression of certain keratins or production of mucin and lack neuroedocrine differentiation. The molecular pathogenesis of lung cancer involves the accumulation of genetic und epigenetic alterations including the activation of proto-oncogenes and inactivation of tumor suppressor genes which are different for lung cancer subgroups. The development of microarray technologies opened up the possibility to quantify the expression of a large number of genes simultaneously in a given sample. There are several recent reports on expression profiling on lung cancers but the analysis interpretation of the results might be difficult because of the heterogeneity of cellular components. A contamination of the tumor sample with normal epithelia, blood vessels, stromal cells, leucocytes and tumor necrosis may confound the true expression profile of the tumor. The use of laser capture microdissection (LCM) greatly improves the sample preparation for microarray expression analysis. Consequently, we used advanced technology including LCM and microarray analysis. In detail, we examined gene expression profiles of tumor cells from 29 previously untreated patients with lung cancer (10 adenocarcinomas (AC), 10 squamous cell carcinomas (SCC), 9 small cell lung cancer (SCLC)) in comparison to normal lung tissue (LT) of 5 control patients without tumor. Bronchoscopical biopsies from the primary lung tumor were taken before treatment. Biopsies were cut into 8Âµm sections and from each section cancer cells were isolated using laser capture microdissection in order to obtain pure samples of tumor cells. Total RNA was extracted, reversely transcribed, in-vitro transcribed, labelled and hybridized to the array. For expression analysis, microarrays covering 8793 defined genes (Human HG Focus Array, Affymetrix) were used. Following quality control, array data were normalized and analysed for significant differences using variance stabilizing transformation (VSN) and significance analysis of microarrays (SAM), respectively. Based on differentially expressed genes cancer samples could be clearly separated from non cancer samples using hierarchical clustering. Comparing AC, SCC and SCLC with normal lung tissue, we found 205, 335 and 404 genes, respectively, that were at least 2-fold differentially expressed with an estimated false discovery rate < 2.6%. Each histological subtype showed a distinct expression profile. Further, using a genetic programming approach we constructed a classificator to discriminate AC, SCC, NT and SCLC. To this end, the 50 genes with the greatest signal-to-noise ratio were selected to train the classificator. By leave-one-out cross validation all 34 samples were correctly classified in this training set. In order to validate the 50-gene-classificator on a test set, further 13 microdissected lung cancer samples were used and correctly classified in concordance to pathologic finding. In conclusion, the different lung cancer subtypes have distinct molecular phenotypes which reflect biological characteristics of the tumor cells and which might be the basis for development of targeted therapy. Moreover, gene expression profiling and genetic programming is a suitable tool for classification and discrimination of different histological subtypes in lung cancer in comparison to normal lung tissue. Experiment Overall Design: Comparison of gene expression profiles of normal lung tissues, adenocarcinomas, squamous cell carcinomas and small cell lung cancers.

Project description:Background Neuroendcrine carcinoma (NEC) of lung consists of small-cell lung carcinoma (SCLC) and large cell neuroendocrine carcinoma (LCNEC). Although long-term outcomes of SCLC patients are usually poor, a few patients achieve long-term survivals. Prognostic factors in SCLC have not been fully elucidated. microRNAs (miRNAs) are known to negatively regulate gene expression and to be relevant to tumorigenesis, tumor classification and prognosis. Methods 63 samples (46 neuroendocrine carcinoma including 35 SCLC and 11 LCNEC, 4 adenocarcinoma, 5 squamous cell carcinoma and 8 normal lung) were obtained through surgical resection. miRNA expression in each sample was comprehensively investigated using miRNA microarray. Results Unsupervised hierarchial clustering classified the all NEC into two subgroups, group 1 and 2. Group 1 was felt into an independent branch that consisted of only NEC, whereas group 2 was included in a branch that contained both NEC and non-NEC. Compared with SCLC of group 1 (SCLC 1), SCLC of group 2 (SCLC 2) was a distinct subgroup with surprisingly good prognosis (0% v 79% survived for 5 years; p=0.007). Serum proGRP level was significantly lower in SCLC 2 than in SCLC 1 despite of similar tumor stages between the two groups. Among 12 cases treated by presurgical chemotherapy, PR:NC ratio was 6:3 in SCLC 1 and 2:1 in SCLC 2, suggesting no apparent difference of chemosensitivity. Histologically, distinction between both groups was difficult. Cox regression analysis demonstrated that three miRNA signature was correlated with survival of SCLC patients. These prognostic miRNAs were differentially expressed between SCLC 1 and 2, which were validated by quantitative real time RT-PCR. Bioinformatics analysis predicted that these miRNAs targeted not only genes involving with tumorigenesis but also genes associated with neuroendocrine function. Conclusion miRNA expression profiling identified a distinct subgroup of SCLC with favorable prognosis. This subgroup might have less neuroendocrine character than usual SCLC. All samples were tissue samples obtained by surgical resection. 35 samples of small-cell lung cancer were investigated. 11 samples of larege cell neuroendocrine cancer, 4 samples of adenocarcinoma, 5 samples of squamous cell carcinoma and 8 samples of normal lung were also included as reference.

Project description:Background Neuroendcrine carcinoma (NEC) of lung consists of small-cell lung carcinoma (SCLC) and large cell neuroendocrine carcinoma (LCNEC). Although long-term outcomes of SCLC patients are usually poor, a few patients achieve long-term survivals. Prognostic factors in SCLC have not been fully elucidated. microRNAs (miRNAs) are known to negatively regulate gene expression and to be relevant to tumorigenesis, tumor classification and prognosis. Methods 63 samples (46 neuroendocrine carcinoma including 35 SCLC and 11 LCNEC, 4 adenocarcinoma, 5 squamous cell carcinoma and 8 normal lung) were obtained through surgical resection. miRNA expression in each sample was comprehensively investigated using miRNA microarray. Results Unsupervised hierarchial clustering classified the all NEC into two subgroups, group 1 and 2. Group 1 was felt into an independent branch that consisted of only NEC, whereas group 2 was included in a branch that contained both NEC and non-NEC. Compared with SCLC of group 1 (SCLC 1), SCLC of group 2 (SCLC 2) was a distinct subgroup with surprisingly good prognosis (0% v 79% survived for 5 years; p=0.007). Serum proGRP level was significantly lower in SCLC 2 than in SCLC 1 despite of similar tumor stages between the two groups. Among 12 cases treated by presurgical chemotherapy, PR:NC ratio was 6:3 in SCLC 1 and 2:1 in SCLC 2, suggesting no apparent difference of chemosensitivity. Histologically, distinction between both groups was difficult. Cox regression analysis demonstrated that three miRNA signature was correlated with survival of SCLC patients. These prognostic miRNAs were differentially expressed between SCLC 1 and 2, which were validated by quantitative real time RT-PCR. Bioinformatics analysis predicted that these miRNAs targeted not only genes involving with tumorigenesis but also genes associated with neuroendocrine function. Conclusion miRNA expression profiling identified a distinct subgroup of SCLC with favorable prognosis. This subgroup might have less neuroendocrine character than usual SCLC.

Project description:Lung cancers are a heterogeneous group of diseases with respect to biology and clinical behavior. So far, diagnosis and classification are based on histological morphology and immunohistological methods for discrimination between two main histologic groups: small cell lung cancer (SCLC) and non-small cell lung cancer which account for 20% and 80% of lung carcinomas, respectively. While SCLCs express properties of neuroendocrine cells, NSCLCs, which are divided into the three major subtypes adenocarcinoma, squamous cell carcinoma and dedifferentiated large cell carcinoma, show different characteristics such as the expression of certain keratins or production of mucin and lack neuroedocrine differentiation. The molecular pathogenesis of lung cancer involves the accumulation of genetic und epigenetic alterations including the activation of proto-oncogenes and inactivation of tumor suppressor genes which are different for lung cancer subgroups. The development of microarray technologies opened up the possibility to quantify the expression of a large number of genes simultaneously in a given sample. There are several recent reports on expression profiling on lung cancers but the analysis interpretation of the results might be difficult because of the heterogeneity of cellular components. A contamination of the tumor sample with normal epithelia, blood vessels, stromal cells, leucocytes and tumor necrosis may confound the true expression profile of the tumor. The use of laser capture microdissection (LCM) greatly improves the sample preparation for microarray expression analysis. Consequently, we used advanced technology including LCM and microarray analysis. In detail, we examined gene expression profiles of tumor cells from 29 previously untreated patients with lung cancer (10 adenocarcinomas (AC), 10 squamous cell carcinomas (SCC), 9 small cell lung cancer (SCLC)) in comparison to normal lung tissue (LT) of 5 control patients without tumor. Bronchoscopical biopsies from the primary lung tumor were taken before treatment. Biopsies were cut into 8µm sections and from each section cancer cells were isolated using laser capture microdissection in order to obtain pure samples of tumor cells. Total RNA was extracted, reversely transcribed, in-vitro transcribed, labelled and hybridized to the array. For expression analysis, microarrays covering 8793 defined genes (Human HG Focus Array, Affymetrix) were used. Following quality control, array data were normalized and analysed for significant differences using variance stabilizing transformation (VSN) and significance analysis of microarrays (SAM), respectively. Based on differentially expressed genes cancer samples could be clearly separated from non cancer samples using hierarchical clustering. Comparing AC, SCC and SCLC with normal lung tissue, we found 205, 335 and 404 genes, respectively, that were at least 2-fold differentially expressed with an estimated false discovery rate < 2.6%. Each histological subtype showed a distinct expression profile. Further, using a genetic programming approach we constructed a classificator to discriminate AC, SCC, NT and SCLC. To this end, the 50 genes with the greatest signal-to-noise ratio were selected to train the classificator. By leave-one-out cross validation all 34 samples were correctly classified in this training set. In order to validate the 50-gene-classificator on a test set, further 13 microdissected lung cancer samples were used and correctly classified in concordance to pathologic finding. In conclusion, the different lung cancer subtypes have distinct molecular phenotypes which reflect biological characteristics of the tumor cells and which might be the basis for development of targeted therapy. Moreover, gene expression profiling and genetic programming is a suitable tool for classification and discrimination of different histological subtypes in lung cancer in comparison to normal lung tissue. Keywords: ordered

Project description:Lung cancers are a heterogeneous group of diseases with respect to biology and clinical behavior. Currently, diagnosis and classification are based on histological morphology and immunohistological methods for discrimination between two main histologic groups: small cell lung cancer (SCLC) and non-small cell lung cancer which account for 20% and 80% of lung carcinomas, respectively. NSCLCs, which are divided into the three major subtypes adenocarcinoma, squamous cell carcinoma and dedifferentiated large cell carcinoma, show different characteristics such as the expression of certain keratins or production of mucin and lack of neuroedocrine differentiation. The molecular pathogenesis of lung cancer involves the accumulation of genetic und epigenetic alterations including the activation of proto-oncogenes and inactivation of tumor suppressor genes which are different for lung cancer subgroups. The development of microarray technologies opened up the possibility to quantify the expression of a large number of genes simultaneously in a given sample. There are several recent reports on expression profiling on lung cancers but the analysis interpretation of the results might be difficult because of the heterogeneity of cellular components. The methods used for sample selection and processing can have a strong influence on the expression values obtained through microarray profiling. Laser capture microdissection (LCM) provides higher specificity in the selection of target cells compared to traditional bulk tissue selection methods, but at an increased processing cost. Here we describe the use of an expression microarray study on NSCLC samples and surrounding tissue, comparing macroscopic lung tumor and tissue samples (“grind and bind”), versus tumor and alveolar compartment cells laser capture microdissected (LCM) from the same macroscopic lung samples. In this study, a set of 31 pairs and one non-paired sample of macroscopic tumor and non-tumor samples (10 pairs and 1 non-paired sample squamous-cell carcinoma, 19 pairs and one non-paired samples adenocarcinoma, 2 pairs adeno-squamous-cell carcinoma) was selected for bulk/macro sampling. Of these 31 pairs and 2 non-paired samples, 16 pairs plus 15 non paired samples were reanalyzed using laser capture microdissection (LCM) for sampling the cells (7 pairs and 3 non-paired samples squamous-cell carcinoma, 8 pairs and 11 non-paired samples adeno carcinomas, 1 pair and 1 non paired sample Adeno-squamous-cell carcinoma). For macroscopic samples, 50 to 80 µg of tissue was used to isolate total RNA. Gene expression profile was determined using Affymetrix Human Genome Gene 1.0 ST genechip. For the LCM samples, from representative slides histologically confirmed and mapped by a pathologist, approximately 1000 cells/sample were collected by LCM;. cDNA was amplified using Nugen WT-Ovation One-Direct amplification system. Here we describe the use of an expression microarray study on NSCLC samples and surrounding tissue, comparing macroscopic lung tumor and tissue samples (“grind and bind”), versus tumor and alveolar compartment cells laser capture microdissected (LCM) from the same macroscopic lung samples.

Dataset Information

The miRNA expression profiles in three subtypes of lung carcinomas

Dataset's files

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets