1,247 pubs · 114,977 cites · h-index 161
The transcription factor achaete-scute complex homolog 1 (ASCL1) is a lineage oncogene that is central for the growth and survival of small cell lung cancers (SCLC) and neuroendocrine non-small cell lung cancers (NSCLC-NE) that express it. Targeting ASCL1, or its downstream pathways, remains a challenge. However, a potential clue to overcoming this challenage has been information that SCLC and NSCLC-NE that express ASCL1 exhibit extremely low ERK1/2 activity, and efforts to increase ERK1/2 activity lead to inhibition of SCLC growth and surival. Of course, this is in dramatic contrast to the majority of NSCLCs where high activity of the ERK pathway plays a major role in cancer pathogenesis. A major knowledge gap is defining the mechanism(s) underlying the low ERK1/2 activity in SCLC, determining if ERK1/2 activity and ASCL1 function are inter-related, and if manipulating ERK1/2 activity provides a new therapeutic strategy for SCLC. We first found that expression of ERK signaling and ASCL1 have an inverse relationship in NE lung cancers: knocking down ASCL1 in SCLCs and NE-NSCLCs increased active ERK1/2, while inhibition of residual SCLC/NSCLC-NE ERK1/2 activity with a MEK inhibitor increased ASCL1 expression. To determine the effects of ERK activity on expression of other genes, we obtained RNA-seq from ASCL1-expressing lung tumor cells treated with an ERK pathway MEK inhibitor and identified down-regulated genes (such as SPRY4, ETV5, DUSP6, SPRED1) that potentially could influence SCLC/NSCLC-NE tumor cell survival. This led us to discover that genes regulated by MEK inhibition suppress ERK activation and CHIP-seq demonstrated these are bound by ASCL1. In addition, SPRY4, DUSP6, SPRED1 are known suppressors of the ERK1/2 pathway, while ETV5 regulates DUSP6. Survival of NE lung tumors was inhibited by activation of ERK1/2 and a subset of ASCL1-high NE lung tumors expressed DUSP6. Because the dual specificity phosphatase 6 (DUSP6) is an ERK1/2-selective phosphatase that inactivates these kinases and has a pharmacologic inhibitor, we focused mechanistic studies on DUSP6. These studies showed: Inhibition of DUSP6 increased active ERK1/2, which accumulated in the nucleus; pharmacologic and genetic inhibition of DUSP6 affected proliferation and survival of ASCL1-high NE lung cancers; and that knockout of DUSP6 "cured" some SCLCs while in others resistance rapidly developed indicating a bypass mechanism was activated. Thus, our findings fill this knowledge gap and indicate that combined expression of ASCL1, DUSP6 and low phospho-ERK1/2 identify some neuroendocrine lung cancers for which DUSP6 may be a therapeutic target.
Abstract Purpose: We have previously mapped a major susceptibility locus influencing familial lung cancer risk to chromosome 6q23-25. However, the causal gene at this locus remains undetermined. In this study, we further refined this locus to identify a single candidate gene, by fine mapping using microsatellite markers and association studies using high-density single nucleotide polymorphisms (SNP). Experimental Design: Six multigenerational families with five or more affected members were chosen for fine-mapping the 6q linkage region using microsatellite markers. For association mapping, we genotyped 24 6q-linked cases and 72 unrelated noncancer controls from the Genetic Epidemiology of Lung Cancer Consortium resources using the Affymetrix 500K chipset. Significant associations were validated in two independent familial lung cancer populations: 226 familial lung cases and 313 controls from the Genetic Epidemiology of Lung Cancer Consortium, and 154 familial cases and 325 controls from Mayo Clinic. Each familial case was chosen from one high-risk lung cancer family that has three or more affected members. Results: A region-wide scan across 6q23-25 found significant association between lung cancer susceptibility and three single nucleotide polymorphisms in the first intron of the RGS17 gene. This association was further confirmed in two independent familial lung cancer populations. By quantitative real-time PCR analysis of matched tumor and normal human tissues, we found that RGS17 transcript accumulation is highly and consistently increased in sporadic lung cancers. Human lung tumor cell proliferation and tumorigenesis in nude mice are inhibited upon knockdown of RGS17 levels. Conclusion:RGS17 is a major candidate for the familial lung cancer susceptibility locus on chromosome 6q23-25.
PDF file - 1617KB, Figure S1. Malignant Characteristics of Invasive Adenocarcinoma in K-rasLA1 Mice; Figure S2. Radiation Effects on the Incidence of Various Endpoints in K-rasLA1 Mice; Figure S3. Comparative Genomic Analyses and Classifier Isolation from Irradiated Versus Unirradiated Control K-rasLA1 Mice; Figure S4. Comparative Genomic Analysis of Whole Lungs Reveals Unique Gene Classifiers Capable of Specifying Individual Experimental Cohorts; Figure S5. Only "Fractionated" Classifier Demonstrates Clinical Relevance for Lung Cancer Patient Survival; Figure S6. "Fractionated" Classifier Capable of Predicting Overall Survival in Patients with Breast, but not Lung Squamous Cell Cancer; Supplementary Figure S7. Cox Regression Analysis Exposes 6 Genes Within "Fractionated" Classifier Which Retain Predictive Capacity; Table S1. Logistic Regression Analysis of Unirradiated K-rasLA1 Mice for Gender and Strain Effects on Various Endpoints; Table S2. Logistic Regression Analysis for Gender and Strain Effects on Various Endpoints Controlling for Experiment; Table S3. Logistic Regression Analysis of Radiation Effects on the Incidence of Invasive Adenocarcinoma Controlling for Gender and Strain; Table S4. Multivariate Cox Analysis for Gender and Strain Effects on Overall Survival Controlling for Experiment; Table S5. Multivariate Cox Analysis of Radiation Effects on Overall Survival Controlling for Gender and Strain Effects; Table S6. IPA Network Annotations Associated with Corresponding Gene Lists.
PDF file - 1617KB, Figure S1. Malignant Characteristics of Invasive Adenocarcinoma in K-rasLA1 Mice; Figure S2. Radiation Effects on the Incidence of Various Endpoints in K-rasLA1 Mice; Figure S3. Comparative Genomic Analyses and Classifier Isolation from Irradiated Versus Unirradiated Control K-rasLA1 Mice; Figure S4. Comparative Genomic Analysis of Whole Lungs Reveals Unique Gene Classifiers Capable of Specifying Individual Experimental Cohorts; Figure S5. Only "Fractionated" Classifier Demonstrates Clinical Relevance for Lung Cancer Patient Survival; Figure S6. "Fractionated" Classifier Capable of Predicting Overall Survival in Patients with Breast, but not Lung Squamous Cell Cancer; Supplementary Figure S7. Cox Regression Analysis Exposes 6 Genes Within "Fractionated" Classifier Which Retain Predictive Capacity; Table S1. Logistic Regression Analysis of Unirradiated K-rasLA1 Mice for Gender and Strain Effects on Various Endpoints; Table S2. Logistic Regression Analysis for Gender and Strain Effects on Various Endpoints Controlling for Experiment; Table S3. Logistic Regression Analysis of Radiation Effects on the Incidence of Invasive Adenocarcinoma Controlling for Gender and Strain; Table S4. Multivariate Cox Analysis for Gender and Strain Effects on Overall Survival Controlling for Experiment; Table S5. Multivariate Cox Analysis of Radiation Effects on Overall Survival Controlling for Gender and Strain Effects; Table S6. IPA Network Annotations Associated with Corresponding Gene Lists.
Abstract Steroid receptor coactivator-3 (SRC-3) is a histone acetyltransferase and nuclear hormone receptor coactivator, located on 20q12, which is amplified in several epithelial cancers and well studied in breast cancer. However, its possible role in lung cancer pathogenesis is unknown. We found SRC-3 to be overexpressed in 27% of non–small cell lung cancer (NSCLC) patients (n = 311) by immunohistochemistry, which correlated with poor disease-free (P = 0.0015) and overall (P = 0.0008) survival. Twenty-seven percent of NSCLCs exhibited SRC-3 gene amplification, and we found that lung cancer cell lines expressed higher levels of SRC-3 than did immortalized human bronchial epithelial cells (HBEC), which in turn expressed higher levels of SRC-3 than did cultured primary human HBECs. Small interfering RNA–mediated downregulation of SRC-3 in high-expressing, but not in low-expressing, lung cancer cells significantly inhibited tumor cell growth and induced apoptosis. Finally, we found that SRC-3 expression is inversely correlated with gefitinib sensitivity and that SRC-3 knockdown results in epidermal growth factor receptor tyrosine kinase inhibitor–resistant lung cancers becoming more sensitive to gefitinib. Taken together, these data suggest that SRC-3 may be an important oncogene and therapeutic target for lung cancer. Cancer Res; 70(16); 6477–85.
The human adrenal cortex is composed of distinct zones that are the main source of steroid hormone production. The mechanism of adrenocortical cell differentiation into several functionally organized populations with distinctive identities remains poorly understood. Human adrenal disease has been difficult to study, in part due to the absence of cultured cell lines that faithfully represent adrenal cell precursors in the early stages of transformation. Here, Human Adrenocortical Adenoma (HAA1) cell line derived from a patient's macronodular adrenocortical hyperplasia and was treated with histone deacetylase inhibitors (HDACis) and gene expression was examined. We describe a patient-derived HAA1 cell line derived from the zona reticularis, the innermost zone of the adrenal cortex. The HAA1 cell line is unique in its ability to exit a latent state and respond with steroidogenic gene expression upon treatment with histone deacetylase inhibitors. The gene expression pattern of differentiated HAA1 cells partially recreates the roster of genes in the adrenal layer that they have been derived from. Gene ontology analysis of whole genome RNA-seq corroborated increased expression of steroidogenic genes upon HDAC inhibition. Surprisingly, HDACi treatment induced broad activation of the Tumor Necrosis Factor (TNF) alpha pathway. This novel cell line we developed will hopefully be instrumental in understanding the molecular and biochemical mechanisms controlling adrenocortical differentiation and steroidogenesis.
BACKGROUND: Approximately 20% of lung adenocarcinoma (LUAD) is negative for the lineage-specific oncogene Thyroid transcription factor 1 (TTF-1) and exhibits worse clinical outcome with a low frequency of actionable genomic alterations. To identify molecular features associated with TTF-1-negative LUAD, we compared the transcriptomic and proteomic profiles of LUAD cell lines. SRGN , a chondroitin sulfate proteoglycan Serglycin, was identified as a markedly overexpressed gene in TTF-1-negative LUAD. We therefore investigated the roles and regulation of SRGN in TTF-1-negative LUAD. METHODS: Proteomic and metabolomic analyses of 41 LUAD cell lines were done using mass spectrometry. The function of SRGN was investigated in 3 TTF-1-negative and 4 TTF-1-positive LUAD cell lines and in a syngeneic mouse model (n = 5 to 8 mice per group). Expression of SRGN was evaluated in 94 and 105 surgically resected LUAD tumor specimens using immunohistochemistry. All statistical tests were 2-sided. RESULTS: SRGN was markedly overexpressed at mRNA and protein levels in TTF-1-negative LUAD cell lines (P < .001 for both mRNA and protein levels). Expression of SRGN in LUAD tumor tissue was associated with poor outcome (hazard ratio = 4.22, 95% confidence interval = 1.12 to 15.86, likelihood ratio test, P = .03), and with higher expression of Programmed cell death 1 ligand 1 (PD-L1) in tumor cells and higher infiltration of Programmed cell death protein 1-positive lymphocytes. SRGN regulated expression of PD-L1 as well as proinflammatory cytokines, including Interleukin-6, Interleukin-8, and C-X-C motif chemokine 1 in LUAD cell lines; increased migratory and invasive properties of LUAD cells and fibroblasts; and enhanced angiogenesis. SRGN was induced by DNA demethylation resulting from Nicotinamide N-methyltransferase-mediated impairment of methionine metabolism. CONCLUSIONS: Our findings suggest that SRGN plays a pivotal role in tumor-stromal interaction and reprogramming into an aggressive and immunosuppressive tumor microenvironment in TTF-1-negative LUAD.
Small cell lung cancer (SCLC) is classified as a high-grade neuroendocrine (NE) tumor, but a subset of SCLC has been termed "variant" due to the loss of NE characteristics. In this study, we computed NE scores for patient-derived SCLC cell lines and xenografts, as well as human tumors. We aligned NE properties with transcription factor-defined molecular subtypes. Then we investigated the different immune phenotypes associated with high and low NE scores. We found repression of immune response genes as a shared feature between classic SCLC and pulmonary neuroendocrine cells of the healthy lung. With loss of NE fate, variant SCLC tumors regain cell-autonomous immune gene expression and exhibit higher tumor-immune interactions. Pan-cancer analysis revealed this NE lineage-specific immune phenotype in other cancers. Additionally, we observed MHC I re-expression in SCLC upon development of chemoresistance. These findings may help guide the design of treatment regimens in SCLC.
MYC stimulates both metabolism and protein synthesis, but how cells coordinate these complementary programs is unknown. Previous work reported that, in a subset of small-cell lung cancer (SCLC) cell lines, MYC activates guanosine triphosphate (GTP) synthesis and results in sensitivity to inhibitors of the GTP synthesis enzyme inosine monophosphate dehydrogenase (IMPDH). Here, we demonstrated that primary MYChi human SCLC tumors also contained abundant guanosine nucleotides. We also found that elevated MYC in SCLCs with acquired chemoresistance rendered these otherwise recalcitrant tumors dependent on IMPDH. Unexpectedly, our data indicated that IMPDH linked the metabolic and protein synthesis outputs of oncogenic MYC. Coexpression analysis placed IMPDH within the MYC-driven ribosome program, and GTP depletion prevented RNA polymerase I (Pol I) from localizing to ribosomal DNA. Furthermore, the GTPases GPN1 and GPN3 were upregulated by MYC and directed Pol I to ribosomal DNA. Constitutively GTP-bound GPN1/3 mutants mitigated the effect of GTP depletion on Pol I, protecting chemoresistant SCLC cells from IMPDH inhibition. GTP therefore functioned as a metabolic gate tethering MYC-dependent ribosome biogenesis to nucleotide sufficiency through GPN1 and GPN3. IMPDH dependence is a targetable vulnerability in chemoresistant MYChi SCLC.
Small cell lung cancer (SCLC) is classified as a high-grade neuroendocrine (NE) tumor, but a subset of SCLC has been termed “variant” due to the loss of NE characteristics. In this study, we computed NE scores for patient-derived SCLC cell lines and xenografts, as well as human tumors. We aligned NE properties with transcription factor-defined molecular subtypes. Then we investigated the different immune phenotypes associated with high and low NE scores. We found repression of immune response genes as a shared feature between classic SCLC and pulmonary neuroendocrine cells of the healthy lung. With loss of NE fate, variant SCLC tumors regain cell-autonomous immune gene expression and exhibit higher tumor-immune interactions. Pan-cancer analysis revealed this NE lineage-specific immune phenotype in other cancers. Additionally, we observed MHC I re-expression in SCLC upon development of chemoresistance. These findings provide a new framework to guide design of treatment regimens in SCLC.
Simian virus 40 (SV40) is a DNA tumor virus capable of infecting and transforming human mesothelial (HM) cells in vitro. Hamsters injected intracardially to expose most tissue types to SV40 preferentially develop mesotheliomas. In humans, asbestos is the main cause of mesothelioma, and asbestos and SV40 are co-carcinogens in transforming HM cells in tissue culture and in causing mesothelioma in hamsters. Laser microdissection experiments conducted in the laboratory of Adi Gazdar demonstrated that SV40 was present specifically in the malignant mesothelioma cells and not in nearby stromal cells. Further experiments demonstrated that SV40 remains episomal in HM cells and astrocytes because of the production of a long antisense RNA that represses viral capsid protein production. Thus, the potent SV40 oncoprotein, T-antigen (Tag), is expressed, but because the capsid proteins are not produced, the cells are not lysed and, instead, become transformed. Together this evidence suggests that SV40 may contribute to the development of mesotheliomas in humans. However, epidemiological evidence to support this hypothesis is lacking. This chapter also summarizes the introduction of SV40, a monkey virus, into the human population as an unrecognized contaminant of early poliovaccines. In addition to mesotheliomas, SV40 now is linked with brain cancers, osteosarcomas, and lymphomas in humans. Explanations are provided for the apparent geographic variations in SV40 prevalence and for controversies about the role of SV40 in human cancer.
Tremendous amount of whole-genome sequencing data have been provided by large consortium projects such as TCGA (The Cancer Genome Atlas), COSMIC and so on, which creates incredible opportunities for functional gene research and cancer associated mechanism uncovering. While the existing web servers are valuable and widely used, many whole genome analysis functions urgently needed by experimental biologists are still not adequately addressed. A cloud-based platform, named CG (ClickGene), therefore, was developed for DIY analyzing of user's private in-house data or public genome data without any requirement of software installation or system configuration. CG platform provides key interactive and customized functions including Bee-swarm plot, linear regression analyses, Mountain plot, Directional Manhattan plot, Deflection plot and Volcano plot. Using these tools, global profiling or individual gene distributions for expression and copy number variation (CNV) analyses can be generated by only mouse button clicking. The easy accessibility of such comprehensive pan-cancer genome analysis greatly facilitates data mining in wide research areas, such as therapeutic discovery process. Therefore, it fills in the gaps between big cancer genomics data and the delivery of integrated knowledge to end-users, thus helping unleash the value of the current data resources. More importantly, unlike other R-based web platforms, Dubbo, a cloud distributed service governance framework for 'big data' stream global transferring, was used to develop CG platform. After being developed, CG is run on an independent cloud-server, which ensures its steady global accessibility. More than 2 years running history of CG proved that advanced plots for hundreds of whole-genome data can be created through it within seconds by end-users anytime and anywhere. CG is available at http://www.clickgenome.org/.
BACKGROUND: Small cell lung cancer (SCLC) is usually diagnosed in the advanced stage. It has a very poor prognosis, with no advancements in therapy in the last few decades. A recent phase 1 clinical study, using an antibody-drug conjugate directed against DLL3, showed promising results. A prerequisite for this therapy is an immunohistochemical test for DLL3 expression. The antibody used in the clinical trial was bound to a specific platform, which is not available in all pathology laboratories. In this study, the expression of DLL3 was analyzed using different DLL3 antibodies in high-grade neuroendocrine tumors of the lung and cell cultures. Additionally, correlation of DLL3 expression with Rb1 loss and TP53 mutation was evaluated. METHODS: The study cohort consisted of surgically resected cases, 24 SCLC and 29 large cell neuroendocrine carcinoma (LCNEC), from which tissue microarrays (TMAs) were constructed. The validation cohort included 46 SCLC samples, mostly small biopsies. Additionally, well-characterized SCLC cell lines were used. Immunohistochemical analysis was performed using four different DLL3 antibodies, as well as TP53 and Rb1 antibodies. Expression was evaluated microscopically and manually scored. RESULTS: The comparison of all DLL3 antibodies showed poor results for the overall agreement, as well as positive and negative agreement. Differences were observed regardless of the applied cut-off values and the tumor type. The antibody used in the clinical trial was the only which always positively stained the tumor cells obtained from cell cultures with known DLL3 expression and was negative on cells that did not express DLL3. There was no correlation between p53 and DLL3 expression in SCLC and LCNEC. RB1 loss in SCLC showed statistical significant correlation with the DLL3 positivity (p = 0.037), while no correlation was found in LCNEC. CONCLUSION: The DLL3 antibody used in the clinical trial demonstrated superiority in the detection of DLL3 expression. Cell cultures, which can be used for DLL3 antibodies as positive and negative probes, were established. Evidence of DLL3 expression in high proportions of patients with LCNEC might provide basis for studies of new therapy options in this group of patients.
The spatial organization of different types of cells in tumor tissues reveals important information about the tumor microenvironment (TME). In order to facilitate the study of cellular spatial organization and interactions, we developed a comprehensive nuclei segmentation and classification tool to characterize the TME from standard Hematoxylin and Eosin (H&E)-stained pathology images. This tool can computationally “stain” different types of cell nuclei in H&E pathology images to facilitate pathologists in analyzing the TME. A Mask Regional-Convolutional Neural Network (Mask-RCNN) model was developed to segment the nuclei of tumor, stromal, lymphocyte, macrophage, karyorrhexis and red blood cells in lung adenocarcinoma (ADC). Using this tool, we identified and classified cell nuclei and extracted 48 cell spatial organization-related features that characterize the TME. Using these features, we developed a prognostic model from the National Lung Screening Trial dataset, and independently validated the model in The Cancer Genome Atlas (TCGA) lung ADC dataset, in which the predicted high-risk group showed significantly worse survival than the low-risk group (pv= 0.001), with a hazard ratio of 2.23 [1.37-3.65] after adjusting for clinical variables. Furthermore, the image-derived TME features were significantly correlated with the gene expression of biological pathways. For example, transcription activation of both the T-cell receptor (TCR) and Programmed cell death protein 1 (PD1) pathways was positively correlated with the density of detected lymphocytes in tumor tissues, while expression of the extracellular matrix organization pathway was positively correlated with the density of stromal cells. This study developed a deep learning-based analysis tool to dissect the TME from tumor tissue images. Using this tool, we demonstrated that the spatial organization of different cell types is predictive of patient survival and associated with the gene expression of biological pathways. Although developed from the pathology images of lung ADC, this model can be adapted into other types of cancers.
We constructed a lung cancer-specific database housing expression data and clinical data from over 6700 patients in 56 studies. Expression data from 23 genome-wide platforms were carefully processed and quality controlled, whereas clinical data were standardized and rigorously curated. Empowered by this lung cancer database, we created an open access web resource-the Lung Cancer Explorer (LCE), which enables researchers and clinicians to explore these data and perform analyses. Users can perform meta-analyses on LCE to gain a quick overview of the results on tumor vs non-malignant tissue (normal) differential gene expression and expression-survival association. Individual dataset-based survival analysis, comparative analysis, and correlation analysis are also provided with flexible options to allow for customized analyses from the user.
CREBBP, encoding an acetyltransferase, is among the most frequently mutated genes in small cell lung cancer (SCLC), a deadly neuroendocrine tumor type. We report acceleration of SCLC upon Crebbp inactivation in an autochthonous mouse model. Extending these observations beyond the lung, broad Crebbp deletion in mouse neuroendocrine cells cooperated with Rb1/Trp53 loss to promote neuroendocrine thyroid and pituitary carcinomas. Gene expression analyses showed that Crebbp loss results in reduced expression of tight junction and cell adhesion genes, including Cdh1, across neuroendocrine tumor types, whereas suppression of Cdh1 promoted transformation in SCLC. CDH1 and other adhesion genes exhibited reduced histone acetylation with Crebbp inactivation. Treatment with the histone deacetylase (HDAC) inhibitor Pracinostat increased histone acetylation and restored CDH1 expression. In addition, a subset of Rb1/Trp53/Crebbp-deficient SCLC exhibited exceptional responses to Pracinostat in vivo. Thus, CREBBP acts as a potent tumor suppressor in SCLC, and inactivation of CREBBP enhances responses to a targeted therapy. Significance: Our findings demonstrate that CREBBP loss in SCLC reduces histone acetylation and transcription of cellular adhesion genes, while driving tumorigenesis. These effects can be partially restored by HDAC inhibition, which exhibited enhanced effectiveness in Crebbp-deleted tumors. These data provide a rationale for selectively treating CREBBP-mutant SCLC with HDAC inhibitors. Cancer Discov; 8(11); 1422–37.
BACKGROUND: While tobacco exposure is the cause of the vast majority of lung cancers, an important percentage arise in lifetime never smokers. Documenting the precise extent of tobacco induced molecular changes may be of importance. Also, the contribution of environmental tobacco smoke (ETS) is difficult to assess. METHODS: We developed and validated a quantitative method to assess the extent of tobacco related molecular damage by combing the most characteristic changes associated with tobacco smoke, the tumor mutation burden (TMB) and type of molecular changes present in lung cancers. Using maximum entropy (MaxEnt) as a classifier, we developed a F score. F score values >0 were considered to show evidence of tobacco related molecular damage, while values ≤0 were considered to lack evidence of tobacco related molecular damage. Compared to the stated patient tobacco exposure histories, the F scores had sensitivity, specificity and accuracy values of 85-87%. Using this method, we analyzed public data sets of lung adenocarcinoma (LUAD), lung squamous cell (LUSC) and small cell lung cancer (SCLC). RESULTS: Less than 10% of LUSCs and SCLCs had negative F scores, while 27% to 35% of LUADs had positive scores. The F score showed a highly significant downward trend when LUADs were subdivided into the following categories: ever, reformed ≤15 years, reformed >15 years and never smokers. Most of the examined bronchial carcinoids (a lung cancer type not associated with smoke exposure) had negative F scores. In addition, most LUADs with EGFR mutations had negative F scores, while almost all with KRAS mutations had positive scores. CONCLUSIONS: We have established and validated a quantitative assay that will be of use in assessing the presence and degree of smoke associated molecular damage in lung cancers arising in ever and never smokers.
Pathology images capture tumor histomorphological details in high resolution. However, manual detection and characterization of tumor regions in pathology images is labor intensive and subjective. Using a deep convolutional neural network (CNN), we developed an automated tumor region recognition system for lung cancer pathology images. From the identified tumor regions, we extracted 22 well-defined shape and boundary features and found that 15 of them were significantly associated with patient survival outcome in lung adenocarcinoma patients from the National Lung Screening Trial. A tumor region shape-based prognostic model was developed and validated in an independent patient cohort (n = 389). The predicted high-risk group had significantly worse survival than the low-risk group (p value = 0.0029). Predicted risk group serves as an independent prognostic factor (high-risk vs. low-risk, hazard ratio = 2.25, 95% CI 1.34-3.77, p value = 0.0022) after adjusting for age, gender, smoking status, and stage. This study provides new insights into the relationship between tumor shape and patient prognosis.
Pathology slide images capture tumor histomorphological details in high resolution. However, manual detection and characterization of tumor regions in pathology slides is labor intensive and subjective. Using a deep convolutional neural network (CNN), we developed an automated tumor region recognition system for lung cancer pathology slides. From the identified regions, we extracted 22 well-defined tumor shape features and found that 15 of them were significantly associated with patient survival outcome in lung adenocarcinoma patients from the National Lung Screening Trial. A tumor shape-based prognostic model was developed and validated in an independent patient cohort (n=389). The predicted high-risk group had significantly worse survival than the low-risk group (p value = 0.0029). Predicted risk group serves as an independent prognostic factor (high-risk vs. low-risk, hazard ratio = 2.25, 95% CI 1.34-3.77, p value = 0.0022) after adjusting for age, gender, smoking status, and stage. This study provides new insights into the relationship between tumor shape and patient prognosis.
We constructed a lung cancer-specific database housing expression data and clinical data from over 6,700 patients in 56 studies. Expression data from 23 “whole-genome” based platforms were carefully processed and quality controlled, whereas clinical data were standardized and rigorously curated. Empowered by this lung cancer database, we created an open access web resource – the Lung Cancer Explorer (LCE), which enables researchers and clinicians to explore these data and perform analyses. Users can perform meta-analyses on LCE to gain a quick overview of the results on tumor vs normal differential gene expression and expression-survival association. Individual dataset-based survival analysis, comparative analysis, and correlation analysis are also provided with flexible options to allow for customized analyses from the user.