Published ahead of print on February 21, 2003, doi:10.1165/rcmb.2002-0183RC
American Journal of Respiratory Cell and Molecular Biology. Vol. 29, pp. 157-162, 2003
© 2003 American Thoracic Society
DOI: 10.1165/rcmb.2002-0183RC
Gene Expression in Lung Adenocarcinomas of Smokers and Nonsmokers
Charles A. Powell*,
Avrum Spira*,
Adnan Derti,
Charles DeLisi,
Gang Liu,
Alain Borczuk,
Steve Busch,
Sudhir Sahasrabudhe,
Yangde Chen,
David Sugarbaker,
Raphael Bueno,
William G. Richards and
Jerome S. Brody
Department of Medicine, Columbia College of Physicians and Surgeons, New York, New York; Pulmonary Center and Department of Medicine, Boston University School of Medicine, and Bioinformatics Program, College of Engineering, Boston University, Boston, Massachusetts; Department of Pathology, Columbia University School of Medicine, New York, New York; Aventis Pharmaceuticals, Bridgewater, New Jersey; and Division of Thoracic Surgery, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts
Address correspondence to: Jerome S. Brody, M.D., Pulmonary Center R3, Boston University Medical Center, 715 Albany St., Boston, MA 02118. E-mail: jbrody{at}bu.edu
 |
Abstract
|
|---|
Adenocarcinoma (AC) has become the most frequent type of lung cancer in men and women, and is the major form of lung cancer in nonsmokers. Our goal in this paper was to determine if AC in smokers and nonsmokers represents the same genetic disease. We compared gene expression profiles in resected samples of nonmalignant lung tissue and tumor tissue in six never-smokers with AC and in six smokers with AC, who were matched for clinical staging and histologic criteria of cell differentiation. Results were analyzed using a variety of bioinformatic tools. Four times as many genes changed expression in the transition from noninvolved lung to tumor in nonsmokers as in smokers, suggesting that AC in nonsmokers evolves locally, whereas AC in smokers evolves in a field of genetically altered tissue. There were some similarities in gene expression in smokers and nonsmokers, but many differences, suggesting different pathways of cell transformation and tumor formation. Gene expression in the noninvolved lungs of smokers differed from that of nonsmokers, and multidimensional scaling showed that noninvolved lungs of smokers groups with tumors rather than noninvolved lungs of nonsmokers. In addition, expression of a number of genes correlated with smoking intensity. Our findings, although limited by small sample size, suggest that additional studies comparing noninvolved to tumor tissue may identify pathogenetic mechanisms and therapeutic targets that differ in AC of smokers and nonsmokers.
Abbreviations: adenocarcinoma, AC • multidimensional scaling, MDS • normal, N • nonsmoker normal, Nns • smoker normal, Ns • real-time polymerase chain reaction, QRT-PCR • tumor, T • transforming growth factor, TGF • nonsmoker tumor, Tns • smoker tumor, Ts
|
Introduction
|
|---|
Lung cancer is the leading cause of cancer deaths in both men and women in the United States (1). Adenocarcinoma (AC) is the most common form of lung cancer and has increased relative to other histologic types of lung cancer over the last several decades. Although cigarette smoking is the principle causal agent of lung cancer, 10–15% of patients with lung cancer have no history of smoking. The majority of these individuals have AC of the lung and are women (Report of the Lung Cancer Progress Review Group, National Cancer Institute, August 2001. http://osp.nci.nih.gov/prg_assess/prg/lungprg/lung_rpt.htm). The changing pattern of lung cancer, and the rising incidence of lung cancer in nonsmokers, prompted us to ask whether AC in nonsmokers is the same genetic disease as in smokers.
DNA microarray technology is capable of producing large gene expression data sets that can provide novel insights into fundamental cancer biology at the molecular level. Microarrays have been applied to a number of different cancers, and several studies comparing different histologic types of lung cancer and lung cancer with different outcomes have appeared recently (2–6). However, none of these studies have explored the specific differences between lung cancer in smokers and nonsmokers, and none have compared nonmalignant to tumor tissue in the same subject. In this study we used gene expression arrays to compare AC with nonmalignant lungs of six smokers and six nonsmokers who were matched for clinical and histologic features of cancer. Although the small numbers in our study limit conclusions regarding specific genes, our findings suggest AC in smokers arises in a field of genetically altered lung tissue, whereas AC in nonsmokers arises locally in relatively normal lung tissue. Although changes in a number of genes are shared in the transition from nonmalignant lung to tumor in smokers and nonsmokers, the majority of genes differ. In addition, there were a number of genes whose expression differed between the nonmalignant lung of smokers compared with nonsmokers, and many of these genes correlated with smoking history. Our studies, although only suggestive because of small sample size, point to the need for further studies of adenocarcinomas of smokers and nonsmokers.
|
Materials and Methods
|
|---|
Patients and Tissue Specimens
We collected primary lung adenocarcinomas and histologically nonmalignant lung tissue from patients undergoing lung cancer resection at Brigham and Women's Hospital (Boston, MA). Samples were snap-frozen in liquid nitrogen and stored at -140°C. We chose six cases of AC in never-smokers from the database, and six cases of adenocarcinoma from cigarette smokers, selected by matching for the following criteria in a descending order of priority: (i) cell type; (ii) histologic stage of differentiation; (iii) clinical stage; and (iv) patient age. Each sample was accompanied by an adjacent section for histologic confirmation. A pathologist determined the proportion of tumor cells and emphysema in each section. Nonmalignant samples contained no recognizable tumor cells; tumor cells made up on average 60% of cells in tumor tissues. (There were no significant differences in amount of emphysema between nonmalignant samples.) The study was approved by the Human Studies Committees of Brigham and Women's Hospital and Boston University Medical Center.
Microarray Data Acquisition and Normalization
RNA was extracted using RNAeasy kits (Qiagen, Valencia, CA). The HuGene FL array (Affymetrix, Santa Clara, CA), which contains  6,800 human genes, was used to generate gene expression profiles for the 24 tissue samples. Six to eight micrograms of total RNA from each of the 24 samples was prepared, hybridized onto the HuGene FL array, and scanned following the Affymetrix protocol as described (7). A single weighted mean expression level for each gene along with a detection P value (which indicates whether the transcript was reliably detected) was derived using Microarray Suite 5.0 software (Affymetrix).
To filter out arrays of poor quality, we assessed several quality control measures and excluded from further analysis two arrays that differed significantly from the others. We scaled the data from each array to normalize the results for inter-array comparisons. We filtered out genes whose median detection P value was not less than 0.05 in at least one of the four comparative groups, leaving 2,047 genes for which data was available in 22 samples. Many of the genes traditionally linked with cancer (i.e., p53, retinoblastoma etc.) were not found on the array or were filtered (see www.netaffx.com for list of all genes on the HU6800 array).
Statistical Analysis
Following data normalization and filtering, we used several methods to generate lists of genes with different levels of expression among four comparative groups of samples. We performed a paired t test to identify genes with statistically significant differences in expression levels between three groups of matched experiments (each patient matched to themselves): smoker tumor (Ts) versus smoker normal (Ns), nonsmoker tumor (Tns) versus nonsmoker normal (Nns), and all tumor (T) versus all normal (N). An unpaired t test was used to identify genes with differences in expression levels between Ns versus Nns tissue. We selected a P value threshold of 0.01 for statistical significance. Due to the presence of multiple comparisons, there was the potential problem of finding genes differentially expressed between two groups when no difference actually exists (8). Current methods for adjustments for multiple comparisons break down quickly, as they are too conservative (9). In addition, they assume independence of the different tests, which is unlikely to hold true in the microarray setting, where multiple genes are co-regulated (10). We first compared the number of genes at our P value threshold with the expected number of false positives under the null hypothesis to estimate the overabundance of information in the analyzed dataset (11). In addition, we employed a permutation test to assess the significance of our P value threshold for any given gene's comparison between two groups (9).
We also performed a permutation-based neighborhood analysis to select the top 50 genes that distinguish the comparative groups (12). The permutation-test neighborhood analysis was used to select marker genes for each of the classes and to assess their statistical significance with respect to a reference empirical distribution obtained by permuting the phenotype class labels (12). The neighborhood analysis using the t test distance function and 1,000 permutations was performed with Genecluster 2.0b software (http://www.genome.wi.mit.edu/cancer/software/software.html).
Following the application of a gene variation filter (see www.bubiopulmatics.org for details), hierarchical clustering of the genes and samples using a Pearson correlation (uncentered) similarity metric and average linkage clustering was performed using CLUSTER and TREVIEW software programs obtained at http://rana.lbl.gov/EisenSoftware.htm. Multidimensional scaling (MDS) of samples according to the expression of the top 30 paired t test genes was performed using the Partek 5.0 software (www.partek.com). In addition, MDS was performed on all tumor samples according to the expression of top 20 t test genes that distinguished Ns versus Nns. We also calculated a correlation coefficient (r value ranging from –1 to 1) for each gene, to quantify how well a given gene's expression level in smoker normal lung correlates with pack-years of smoking (see www.bubiopulmatics.org for details).
Confirmatory Array Studies
To confirm genes differentially expressed, we analyzed a second set of unmatched samples, one each of Ts, Ns, Tns, and Nns, using an Affymetrix U95A array that included most of the genes present on the HU6800 array. Sample collection, microarray data acquisition, and data normalization were performed as described above. We performed multidimensional scaling of these four new samples with the initial samples for three comparisons (T versus N, Ts versus Ns, and Tns versus Nns), using the top 30 genes from the paired t test comparisons.
Real-time polymerase chain reaction (QRT-PCR) was used to confirm the differential expression of a select number of genes that changed in transition from normal lung to tumor among nonsmokers. Primer sequences were designed with Primer Express software (Applied Biosystems, Foster City, CA). Forty cycles of amplification, data acquisition, and data analysis were performed in an ABI Prism 7,700 Sequence Detector (Applied Biosystems).
In addition, we compared our results with data generated by Bhattacharjee and coworkers (2), to confirm our findings among a larger number of lung adenocarcinoma samples. Microarray Suite 5.0 software (Affymetrix) was used to analyze Affymetrix U95A Genechip image files from the 139 adenocarcinoma and 17 normal lung samples from that study. Given that the smoking status of all patients (including all normal samples) in that study was unavailable, we limited our comparison to the genes that distinguish all tumors from all normal lung samples regardless of smoking status. Hierarchical clustering of all samples from the Bhattacharjee dataset was performed using genes identified as differentialy expressed (P < 0.01) between T versus N in our study.
Supplemental Information
Additional information on the statistical methods, microarray protocol, expression levels for all genes in all samples (stored in a relational database), and data analysis is available at www.bubiopulmatics.org.
|
Results
|
|---|
Table 1 presents demographic data for matched pairs of smokers and nonsmokers. Figure 1 shows that four times as many genes differed at the P < 0.01 level in Tns versus Nns than in Ts versus Ns. The strength of this type of analysis lies in the fact that each subject's own tissue served as the control for tumor gene expression. The small sample size in our study limits the conclusions one can draw about specific gene changes because of the false discovery rate inherent in the multiple comparisons. However, the overabundance of information in the analyzed dataset demonstrates that more than four times as many differences in gene expression occurred than would be expected as a result of multiple comparisons for Tns versus Nns, and twice as many changes for Ts versus Ns (Figure 1). Although a number of changes were shared in smokers and nonsmokers, there were many differences in the Ts and Tns transition.
View larger version (114K):
[in this window]
[in a new window]
|
Figure 1. Number of genes that differ between groups by t test. (A) Total number of genes that differ at the P < 0.01 level between T and N, Tns and Nns, and Ts and Ns. Bar for each group includes genes that decreased (bottom) and genes that increased (top). (B) Overabundance graph: a comparison of the number of genes at various P value thresholds with the expected number of false positives under the null hypothesis reveals that 20 differences would be expected at a P < 0.01 and 102 at a P < 0.05. The differences for Tns versus Nns were more than four times that expected, and for Ts versus Ns were twice that expected.
|
|
Table 2 shows selected changes that occurred in Tns/Nns but not in Ts/Ns, changes that occurred in both Tns/Nns and Ts/Ns, and changes in Tns/Nns that also appeared in Ns/Nns comparisons. A complete list of genes that changed in each setting can be found on the website (www.bubiopulmatics.org). Tns was characterized by decreased expression of two transforming growth factor (TGF)-ß receptors and of several components of TGF-ß signaling such as dematopoetin and endoglin, and downstream TGF-ß genes such as connective tissue growth factor. These microarray findings were validated by QRT-PCR, which revealed downregualtion of TGF-ßR2, TGF-ßR3, endoglin, and CTGF in transition from normal lung to adenocarcinoma in nonsmokers (see www.bubiopulmatics.org for QRT-PCR data). TGF-ß has been shown to be a growth suppression gene in a number of epithelial cells. Tns was also characterized by increased expression of several potential oncogenes and decreased expression of several putative tumor suppressor genes, e.g., FEZ1 and BTG1. In addition, a number of cell–cell genes and cell–matrix genes decreased and matrix invasion genes increased, suggesting movement of tumor cells through matrix in Tns. Table 2 shows some of the genes that were shared in the t test comparison between Tns/Nns and Ns/Nns. Several of the genes that decreased (e.g., Id1, Id3, and mig2) are putative tumor suppressor genes, and one, DTR or heparin-binding epidermal growth factor–like gene, is a known oncogene. Many of these genes were also identified in the neighborhood analysis (see below).
We also used more robust methods designed to show relations between subject groups. However, these methods do not allow comparisons between tumor and nontumor sample from the same subject. Unsupervised hierarchical clustering (Figure 2) separated tumor and nontumor samples, and grouped 4/5 noninvolved smokers tissue together. However, Ts and Tns were not completely separated by this method. Neighborhood analysis, an algorithm that selects those genes that best distinguish classes and quantifies their ability to distinguish classes (as compared with what would be expected by chance), showed clear differences between T and N, Tns and Nns, and Ts and Ns (Figure 3); the mean t score, a measure of difference in gene expression between groups, was greater than that expected by chance (5% permutation) in all three groups. A complete list of the top 50 genes in each category can be found on the website. Sixteen of the 50 top genes changed in a similar fashion in the transition from noninvolved lung to tumor in smokers and nonsmokers. These included decreased expression of cell–cell and cell–matrix-associated genes such as CDH5, A2M, and GPC3; the latter two have been shown to bind and inactivate growth factors. There was altered expression of hypoxia/stress-related genes such as GAPDH and PKM2 (increased) and GPX and Cox7A (decreased), and changes in cancer-related genes such as RAGE, Cav1, EMP2, and KAL1, all decreased. In addition, two endothelial genes, PCAM and TEK, decreased, and  1Col increased, suggesting a change in lung interstitial tissue components. Differential expression of RAGE and GPC3 was confirmed by QRT-PCR (see www.bubiopulmatics.org for data). Differences between noninvolved lung and tumor in smokers and nonsmokers included, in smokers, several myc-associated genes (APEX and PHB); and ras-associated genes (GFR, GADD45, and PCTK1); and in nonsmokers, expression of several cell and matrix adhesion molecules (e.g., TIMP3 and MDK) and several genes associated with poor prognosis in other tumors (e.g., CEACAM1, LGAL3BP, and MUC1).
View larger version (21K):
[in this window]
[in a new window]
|
Figure 2. Dendrogram of unsupervised hierarchical clustering. Unsupervised clustering of all samples according to the expression of all filtered genes. T and N samples separated into two groups. Within N samples, four of the five Ns clustered together. Within the T group, four of six Tns clustered together.
|
|
View larger version (57K):
[in this window]
[in a new window]
|
Figure 3. Permutation-based neighborhood analysis. Top 50 genes (by t score) differentially expressed between three comparative groups. Genes in red were increased, and genes in blue decreased compared with the mean expression level (see legend). On top are individual samples; on side are genes (list of genes on website). Genes that distinguish between T and N, Tns and Nns, and Ts and Ns are fairly consistent within each class. However, the top genes in the Tns versus Nns neighborhood analysis provide a stronger and more consistent distinction between classes as compared with genes differentially expressed in Ts versus Ns. This is confirmed by the higher t scores, which represent the mean difference between groups for each of the 50 genes divided by the sum of their standard variations, for Tns versus Nns. Class labels were permuted 1,000 times, and the 5% permutation score represents the t score that would be expected to occur by chance in less than 5% of permutations if there were no true difference between the classes (see website for further detail).
|
|
Although Ns versus Nns were also separated by the neighborhood analysis method, the differences were not as dramatic as the first three groups, indicated by the mix of red and blue in Figure 4A and by the 5% permutation score exceeding the t score (Figure 4). The Ns group contained individuals who smoked between 15 and 150 pack-years. If gene expression was related to amount of smoking, one might expect highly variant patterns of gene expression in Ns. Indeed, 166 genes had a correlation coefficient of > 0.8 when plotted against pack-years of smoking. Genes that correlated with pack-years (see Figure 4 and website) included a number of immunomodulatory genes (e.g., butyrophilin; r = 0.95), apoptosis-related genes (SLC25A; r = -0.92), and signaling genes (zynin; r = 0.91, Pax6; r = 0.99), as well as oxidative stress genes (mitochondrial NADH; r = 0.93). In addition, hnRNP, a gene that has been studied as an early diagnostic marker of lung cancer in sputum and in alveolar lavage (13), was positively correlated with pack-years of smoking.
View larger version (67K):
[in this window]
[in a new window]
|
Figure 4. Neighborhood analysis and correlation with smoking history in Ns. Neighborhood analysis (on left) reveals that the top 50 genes (by t score) separating Ns and Nns show considerable variability within each class, as indicated by the overlap of red and blue expression values and by the mean t score being less than 5% permutation score (more than 5% of permutations would be expected to have a higher t score by chance). Pearson correlation for pack-years and gene expression is shown, on the right, for select genes with R value > 0.8 or < -0.8. Genes positively correlated with pack-years include, top to bottom, BRD2, DXS59928, TGM1, PAX6, and BTN2A2. Genes negatively correlated with pack-years, from top to bottom, are SFTPB, COX4I1, CAPZA1, FXR1, RAF1, NFI/B, and RAP1A.
|
|
Multidimensional scaling using the top 30 genes from the paired t tests effectively distinguished between T and N, Ts and Ns, and Tns and Nns (Figure 5). New unpaired samples were assigned to the correct categories. In addition, MDS and hierarchcal clustering of all tumor samples according to the expression of the top 20 t test genes that distinguish Ns versus Nns revealed clustering of the tumor samples closer to the Ns samples (Figures 6A and 6B). Hierarchical clustering of all adenocarcinoma and normal lung tissue samples from the study by Bhattacharjee and colleagues, using the genes that distinguished all T versus N in Table 1, resulted in all but one of the normal lung samples clustering separately from the tumors (see supplementary figure at www.bubiopulmatics.org).
View larger version (30K):
[in this window]
[in a new window]
|
Figure 5. Multidimensional Scaling (MDS) with addition of four new samples. MDS plot of samples in three-dimensional space according to the expression of top 30 genes from the paired t test comparisons. The four new samples (circled) group with their appropriate class from the initial sample set. Ns, normal smoker sample; Nns, normal nonsmoker sample; Ts, tumor smoker sample; Tns, Tumor nonsmoker sample.
|
|
View larger version (108K):
[in this window]
[in a new window]
|
Figure 6. "Normal" lungs of smokers group with tumor lungs. Hierarchical clustering and MDS plot using top 20 t test genes that distinguish Ns from Nns show that Ns samples group closer to tumor samples than do Nns samples, consistent with the concept that Ns have premalignant changes associated with a smoking-induced "field defect". (A) In cluster figure, green represents low level of expression, and red high level of expression. Green bar above cluster identifies Ns, and red bar identifies Nns samples. (B) MDS in two-dimensional space shows that Ns group with all tumors rather than with Nns lungs.
|
|
|
Discussion
|
|---|
Access to nonmalignant lung tissue and lung tissue with AC from the same individual in nonsmokers and smokers who were matched for tumor cell type, histologic state of differentiation, clinical stage, and age, has provided a unique opportunity to determine whether AC of the lung in smokers and nonsmokers evolve by similar mechanisms. This feature of our study design enabled us to use each subject's noninvolved lung tissue as a control, and thus enhance our ability to detect gene changes that occur in transition from normal lung tissue to AC of the lung. Four times as many genes changed in the transition between nonmalignant lung and tumor in nonsmokers as in smokers, suggesting, because the tumors were matched for clinical and histologic stage, that the noncancerous lungs of smokers already had many alterations in gene expression. This is confirmed by comparison of noncancerous lungs of smokers and nonsmokers, and is consistent with the concept that smoking causes widespread field defects in lungs and airways of smokers, and that tumors in smokers arise within a field of genetically abnormal cells (14, 15). In contrast, AC in nonsmokers likely arises in a field of relatively normal cells, as might be the case with prior infection, such as seen in "scar carcinomas."
The asymmetry between the number of genes measured and the number of samples limits the analysis of specific gene changes identified by paired t tests; therefore, we used alternative analytic approaches that compared groups of individuals but eliminated the comparison of normal and tumor tissue in the same individual. Several different analytic methods confirmed the conclusion that the transition between N and T differed in smokers and nonsmokers, and that the noncancerous lungs of smokers differed from those of nonsmokers. This conclusion was confirmed by multidimensional scaling, which assigned both original samples and several new samples not used to establish the original groupings, to the correct class in multidimensional space.
Although these methods identified genes that changed in a similar fashion and thus were characteristic of AC in smokers and nonsmokers, a number of genes changed only in Ts or in Tns. Tns is characterized by decreased expression of many genes associated with TGF-ß signaling and by changes in cell and matrix genes, suggesting major changes in cell–cell and cell–matrix interactions. In addition, there were differences in gene expression between Ns and Nns, and expression levels of many genes correlated positively or negatively with the intensity or duration of smoking. Many of these latter genes likely represent premalignant changes in the noncancerous lungs of smokers. This is supported by the fact that tumor samples group together with Ns samples according to the expression of the top 20 genes that distinguish Ns versus Nns (Figure 5B). However, it is not known whether these changes are unique to smokers who develop lung cancer or are present in all smokers. Similarly, it is not know whether any of these changes are associated with the presence of concomitant COPD, although only one of the 166 genes whose expression correlates with pack-years of smoking in this study also correlated with pack-years of smoking in a study of lung tissue gene expression in COPD that is underway in our laboratory (data not published).
This study, although preliminary, suggests that AC of smokers and of nonsmokers have different etiologies, and involve different pathways of cell transformation. This implies that optimal approaches to treatment might differ in AC of smokers and nonsmokers. Our study also suggests that a number of premalignant changes occur in the noninvolved lungs of smokers; these changes in gene expression might represent targets for preventative therapy. The possibilities raised in this article suggest that a larger study comparing both noncancerous and tumor tissue in smokers and nonsmokers, and examining gene expression samples in tissue from smokers with and without cancer, would be of considerable clinical importance.
|
Acknowledgments
|
|---|
This work was supported in part by the Doris Duke Charitable Foundation Clinical Scientist Development Award (A.S.).
|
Footnotes
|
|---|
* These authors contributed equally to the work presented in this article. 
Received in original form September 6, 2002
Received in final form February 14, 2003
|
References
|
|---|
- Greenlee, R. T., M. B. Hill-Harmon, T. Murray, and M. Thun. 2001. Cancer statistics, 2001. CA Cancer J. Clin. 51:15–36.[Abstract/Free Full Text]
- Bhattacharjee, A., W. G. Richards, J. Staunton, C. Li, S. Monti, P. Vasa, C. Ladd, J. Beheshti, R. Bueno, M. Gillette, M. Loda, G. Weber, E. J. Mark, E. S. Lander, W. Wong, B. E. Johnson, T. R. Golub, D. J. Sugarbaker, and M. Meyerson. 2001. Classification of human lung carcinomas by mRNA expression profiling reveals distinct adenocarcinoma subclasses. Proc. Natl. Acad. Sci. USA 98:13790–13795.[Abstract/Free Full Text]
- Garber, M. E., O. G. Troyanskaya, K. Schluens, S. Petersen, Z. Thaesler, M. Pacyna-Gengelbach, R. M. van de Rijn, G. D. Rosen, C. M. Perou, R. I. Whyte, R. B. Altman, P. O. Brown, D. Botstein, and I. Petersen. 2001. Diversity of gene expression in adenocarcinoma of the lung. Proc. Natl. Acad. Sci. USA 98:13784–13789.[Abstract/Free Full Text]
- Giordano, T. J., K. A. Shedden, D. R. Schwartz, R. Kuick, J. M. Taylor, N. Lee, D. E. Misek, J. K. Greenson, S. L. Kardia, D. G. Beer, G. Rennert, K. R. Cho, S. B. Gruber, E. R. Fearon, and S. Hanash. 2001. Organ-specific molecular classification of primary lung, colon, and ovarian adenocarcinomas using gene expression profiles. Am. J. Pathol. 159:1231–1238.[Abstract/Free Full Text]
- Beer, D. G., S. L. Kardia, C. C. Huang, T. J. Giordano, A. M. Levin, D. E. Misek, L. Lin, G. Chen, T. G. Gharib, D. G. Thomas, M. L. Lizyness, R. Kuick, S. Hayasaka, J. M. Taylor, M. D. Iannettoni, M. B. Orringer, and S. Hanash. 2002. Gene-expression profiles predict survival of patients with lung adenocarcinoma. Nat. Med. 8:816–824.[Medline]
- Anbazhagan, R., T. Tihan, D. M. Bornman, J. C. Johnston, J. H. Saltz, A. Weigering, S. Piantadosi, and E. Gabrielson. 1999. Classification of small cell lung cancer and pulmonary carcinoid by gene expression profiles. Cancer Res. 59:5119–5122.[Abstract/Free Full Text]
- Tamayo, P., D. Slonim, J. Mesirov, Q. Zhu, S. Kitareewan, E. Dmitrovsky, E. S. Lander, and T. R. Golub. 1999. Interpreting patterns of gene expression with self-organizing maps: methods and application to hematopoietic differentiation. Proc. Natl. Acad. Sci. USA 96:2907–2912.[Abstract/Free Full Text]
- Benjamini, Y., and Y. Hochberg. 1995. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. (Ser. B) 57:289–300.
- MacDonald, T. J., K. M. Brown, B. LaFleur, K. Peterson, C. Lawlor, Y. Chen, R. J. Packer, P. Cogen, and D. A. Stephan. 2001. Expression profiling of medulloblastoma: PDGFRA and the RAS/MAPK pathway as therapeutic targets for metastatic disease. Nat. Genet. 29:143–152.[CrossRef][Medline]
- Tusher, V. G., R. Tibshirani, and G. Chu. 2001. Significance analysis of microarrays applied to the ionizing radiation response. Proc. Natl. Acad. Sci. USA 98:5116–5121.[Abstract/Free Full Text]
- Kaminski, N., and N. Friedman. 2002. Practical approaches to analyzing results of microarray experiments. Am. J. Respir. Cell Mol. Biol. 27:125–132.[Abstract/Free Full Text]
- Pomeroy, S. L., P. Tamayo, M. Gaasenbeek, L. M. Sturla, M. Angelo, M. E. McLaughlin, J. Y. Kim, L. C. Goumnerova, P. M. Black, C. Lau, J. C. Allen, D. Zagzag, J. M. Olson, T. Curran, C. Wetmore, J. A. Biegel, T. Poggio, S. Mukherjee, R. Rifkin, A. Califano, G. Stolovitzky, D. N. Louis, J. P. Mesirov, E. S. Lander, and T. R. Golub. 2002. Prediction of central nervous system embryonal tumour outcome based on gene expression. Nature 415:436–442.[CrossRef][Medline]
- Wistuba, I. I., S. Lam, C. Behrens, A. K. Virmani, K. M. Fong, J. LeRiche, J. M. Samet, S. Srivastava, J. D. Minna, and A. F. Gazdar. 1997. Molecular damage in the bronchial epithelium of current and former smokers. J. Natl. Cancer Inst. 89:1366–1373.[Abstract/Free Full Text]
- Powell, C. A., S. Klares, G. O'Connor, and J. S. Brody. 1999. Loss of heterozygosity in epithelial cells obtained by bronchial brushing: clinical utility in lung cancer. Clin. Cancer Res. 5:2025–2034.[Abstract/Free Full Text]
- Sueoka, E., N. Sueoka, Y. Goto, S. Matsuyama, H. Nishimura, M. Sato, S. Fujimura, H. Chiba, and H. Fujiki. 2001. Heterogeneous nuclear ribonucleoprotein B1 as early cancer biomarker for occult cancer of human lungs and bronchial dysplasia. Cancer Res. 61:1896–1902.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
E. C. Finger, R. S. Turley, M. Dong, T. How, T. A. Fields, and G. C. Blobe
T{beta}RIII suppresses non-small cell lung cancer invasiveness and tumorigenicity
Carcinogenesis,
March 1, 2008;
29(3):
528 - 535.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
E. Zeni, L. Mazzetti, D. Miotto, N. Lo Cascio, P. Maestrelli, P. Querzoli, M. Pedriali, E. De Rosa, L. M. Fabbri, C. E. Mapp, et al.
Macrophage expression of interleukin-10 is a prognostic factor in nonsmall cell lung cancer
Eur. Respir. J.,
October 1, 2007;
30(4):
627 - 632.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C.-K. Toh and W.-T. Lim
Lung cancer in never-smokers
J. Clin. Pathol.,
April 1, 2007;
60(4):
337 - 340.
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. Subramanian and R. Govindan
Lung Cancer in Never Smokers: A Review
J. Clin. Oncol.,
February 10, 2007;
25(5):
561 - 570.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. C. Borczuk and C. A. Powell
Expression Profiling and Lung Cancer Development
Proceedings of the ATS,
January 1, 2007;
4(1):
127 - 132.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
K. M. Lonergan, R. Chari, R. J. deLeeuw, A. Shadeo, B. Chi, M.-S. Tsao, S. Jones, M. Marra, V. Ling, R. Ng, et al.
Identification of Novel Lung Genes in Bronchial Epithelium by Serial Analysis of Gene Expression
Am. J. Respir. Cell Mol. Biol.,
December 1, 2006;
35(6):
651 - 661.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
L. Guo, Y. Ma, R. Ward, V. Castranova, X. Shi, and Y. Qian
Constructing Molecular Classifiers for the Accurate Prognosis of Lung Adenocarcinoma.
Clin. Cancer Res.,
June 1, 2006;
12(11):
3344 - 3354.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C.-K. Toh, F. Gao, W.-T. Lim, S.-S. Leong, K.-W. Fong, S.-P. Yap, A. A.L. Hsu, P. Eng, H.-N. Koong, A. Thirugnanam, et al.
Never-Smokers With Lung Cancer: Epidemiologic Evidence of a Distinct Disease Entity
J. Clin. Oncol.,
May 20, 2006;
24(15):
2245 - 2251.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. Shao, J. Berenguer, A. C. Borczuk, C. A. Powell, T. K. Hei, and Y. Zhao
Epigenetic Inactivation of Betaig-h3 Gene in Human Cancer Cells.
Cancer Res.,
May 1, 2006;
66(9):
4566 - 4573.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. Spina, L. Ferri, A. Cesario, and P. Granone
Chronic obstructive pulmonary disease as a prognostic factor in non-small cell lung cancer
Eur. J. Cardiothorac. Surg.,
September 1, 2005;
28(3):
505 - 506.
[Full Text]
[PDF]
|
|
|
|
|
|
|
W. R. Fields, R. M. Leonard, P. S. Odom, B. K. Nordskog, M. W. Ogden, and D. J. Doolittle
Gene Expression in Normal Human Bronchial Epithelial (NHBE) Cells Following In Vitro Exposure to Cigarette Smoke Condensate
Toxicol. Sci.,
July 1, 2005;
86(1):
84 - 91.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Meyerson and D. Carbone
Genomic and Proteomic Profiling of Lung Cancers: Lung Cancer Classification in the Age of Targeted Therapy
J. Clin. Oncol.,
May 10, 2005;
23(14):
3219 - 3226.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. A. Granville and P. A. Dennis
An Overview of Lung Cancer Genomics and Proteomics
Am. J. Respir. Cell Mol. Biol.,
March 1, 2005;
32(3):
169 - 176.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
V. Shah, S. Sridhar, J. Beane, J. S. Brody, and A. Spira
SIEGE: Smoking Induced Epithelial Gene Expression Database
Nucleic Acids Res.,
January 1, 2005;
33(suppl_1):
D573 - D579.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
L. T. Nordquist, G. R. Simon, A. Cantor, W. M. Alberts, and G. Bepler
Improved Survival in Never-Smokers vs Current Smokers With Primary Adenocarcinoma of the Lung
Chest,
August 1, 2004;
126(2):
347 - 351.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. Kathuria, Y. X. Cao, M. I. Ramirez, and M. C. Williams
Transcription of the Caveolin-1 Gene Is Differentially Regulated in Lung Type I Epithelial and Endothelial Cell Lines: A ROLE FOR ETS PROTEINS IN EPITHELIAL CELL EXPRESSION
J. Biol. Chem.,
July 16, 2004;
279(29):
30028 - 30036.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. C. Borczuk, L. Shah, G. D. N. Pearson, K. L. Walter, L. Wang, J. H. M. Austin, R. A. Friedman, and C. A. Powell
Molecular Signatures in Biopsy Specimens of Lung Cancer
Am. J. Respir. Crit. Care Med.,
July 15, 2004;
170(2):
167 - 174.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. Spira, J. Beane, V. Shah, G. Liu, F. Schembri, X. Yang, J. Palma, and J. S. Brody
Effects of cigarette smoke on the human airway epithelial cell transcriptome
PNAS,
July 6, 2004;
101(27):
10143 - 10148.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. R. Rhodes, J. Yu, K. Shanker, N. Deshpande, R. Varambally, D. Ghosh, T. Barrette, A. Pandey, and A. M. Chinnaiyan
Large-scale meta-analysis of cancer microarray data identifies common transcriptional profiles of neoplastic transformation and progression
PNAS,
June 22, 2004;
101(25):
9309 - 9314.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. Spira, F. Schembri, J. Beane, V. Shah, G. Liu, and J. S. Brody
Impact of Cigarette Smoke on the Normal Airway Transcriptome
Chest,
May 1, 2004;
125(5_suppl):
115S - 115S.
[Full Text]
[PDF]
|
|
|
|
|
|
|
N. Kaminski and M. Krupsky
Gene Expression Patterns, Prognostic and Diagnostic Markers, and Lung Cancer Biology
Chest,
May 1, 2004;
125(5_suppl):
111S - 115S.
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. Kim, G.-L. Xu, A. C. Borczuk, S. Busch, J. Filmus, M. Capurro, J. S. Brody, J. Lange, J. M. D'Armiento, P. B. Rothman, et al.
The Heparan Sulfate Proteoglycan GPC3 Is a Potential Lung Tumor Suppressor
Am. J. Respir. Cell Mol. Biol.,
December 1, 2003;
29(6):
694 - 701.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. C. Borczuk, L. Gorenstein, K. L. Walter, A. A. Assaad, L. Wang, and C. A. Powell
Non-Small-Cell Lung Cancer Molecular Signatures Recapitulate Lung Developmental Pathways
Am. J. Pathol.,
November 1, 2003;
163(5):
1949 - 1960.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
W. N. Rom and K.-M. Tchou-Wong
Functional Genomics in Lung Cancer and Biomarker Detection
Am. J. Respir. Cell Mol. Biol.,
August 1, 2003;
29(2):
153 - 156.
[Full Text]
[PDF]
|
|
|
Copyright © 2003 American Thoracic Society.
|
Top
Abstract
Introduction