Project description:Exposure to cigarette smoke creates a field of injury throughout the entire respiratory tract inducing genomic alterations that lead to an “at-risk” airway where and lung cancers develop. Lung squamous cell carcinoma (SCC) arises in the epithelial layer of the bronchial airways and is often preceded by the development of premalignant lesions(PMLs). The presence of high-grade persistent or progressive PMLs is a marker of increased lung cancer risk both at the lesion site elsewhere in the lung. Effective tools to identify and treat premalignant lesions at highest risk of progression to invasive carcinoma are lacking. We profiled via RNA-Seq airway brushing and biopsies (divided into two cohorts: discovery and validation) obtained from high-risk smokers undergoing lung cancer screening via serial auto-fluorescence bronchoscopy procedures. We identified four distinct molecular subtypes (Proliferative, Inflammatory, Secretory, and Normal) in the bronchial biopsies that correspond to a spectrum of biological and morphological alterations. The Proliferative subtype was enriched with dysplastic PMLs from current smokers that exhibit up-regulation of KRT5 and KI67 as well as metabolic and cell cycle pathways, and down-regulation of cilium-associated processes and FOXJ1 expression. Molecular subtype classification in the validation cohort biopsies replicated these significant clinical and biological associations. Airway brushes from normal fluorescing areas of the large airway classified as the Proliferative subtype specifically predicts the presence of biopsies of this same subtype. Within the Proliferative subtype, genes associated with interferon signaling and antigen processing and presentation were observed to be down-regulated among dysplastic biopsies that persistent or progress in the future. Innate and adaptive immune cells were computationally predicted to be depleted in these biopsies and this was confirmed via immunofluorescence staining of adjacent biopsies. These findings provide a proof of concept that molecular biomarkers of endobronchial biopsies can enhance histopathological grading and that immunoprevention strategies are an important future direction in intercepting the progression of PMLs to lung cancer.

Project description:While lung cancer is the leading cause of cancer death in the US, we have a limited understanding of the earliest molecular events preceding the onset of disease. Prior work has demonstrated that cigarette smoke creates a molecular “field of injury” throughout the airway epithelium and that there are distinct alterations in the airway transcriptome among smokers who have lung cancer. Molecular characterization of this airway “field of injury” in current and former smokers with premalignant lesions (PMLs) could provide novel insights into the earliest molecular events associated with lung carcinogenesis and identify relatively accessible biomarkers to guide lung cancer detection and early intervention. Using mRNA sequencing (mRNA-Seq), we profiled 82 cytologically normal bronchial airway epithelial cells collected during autofluorescence bronchoscopy from high-risk smokers with and without bronchial PMLs, 75 of which were used in downstream analyses. We identified 280 genes differentially expressed in the “field of injury” between subjects with (n=50) and without (n=25) PMLs (FDR<0.002), 81 of which were up-regulated in subjects with PMLs. Oxidative phosphorylation (OXPHOS), the electron transport chain (ETC), and mitochondrial protein transport pathways were strongly enriched among these up-regulated genes (FDR<0.05). We next demonstrated that OXPHOS activation is shared between the “field” and the PMLs with increased oxygen consumption and increased staining for mitochondrial markers in biopsies of PMLs from patients as well as an animal model of lung squamous cell carcinoma (SCC) premalignancy. The 280-gene signature also has a significant concordant relationship to gene expression changes identified in PMLs adjacent to lung SCC tumors, in lung SCC tumors, and in the cytologically normal airway of individuals with lung cancer (FDR<0.05). These findings suggest that these expression changes are reflective of early cancer-associated changes occurring throughout the respiratory tract, and that pathways such as OXPHOS may be targets for chemoprevention. We subsequently developed an airway gene expression biomarker that predicts the presence of PMLs (AUC=0.92, n=17 samples in test set) and show that changes in the biomarker score are associated with progression and regression of PMLs in an independent cohort (AUC=0.75, n=51 samples). The biomarker results indicate that molecular alterations in the field of injury are dynamic with progression or regression of PMLs, suggesting that these changes may be leveraged to stratify high-risk smokers with progressive disease into early intervention trials and monitor disease progression or recurrence.

Project description:Individuals who present with premalignant endobronchial lesions are considered at high risk of lung cancer. Nonetheless, premalignant lesions behave erratically and only a minority progresses towards lung cancer. Therefore, biomarkers need to be discovered that can aid in assessing an individual’s risk for subsequent cancer to better tailor treatment choices and avoid unnecessary follow-up procedures. We recently proposed a classifier of DNA copy number alterations (CNAs) at 3p26.3-p11.1, 3q26.2-29, and 6p25.3-24.3 as risk predictor for endobronchial cancer. The current study was set out to validate the classifier among an independent series of premalignant endobronchial lesions with various histological grades. A series of 36 endobronchial premalignant lesions (8 squamous metaplasia, and 28 various grades of dysplasia) identified during autofluorescence bronchoscopy of 12 case subjects who had carcinoma in situ or carcinoma (≥CIS) during follow-up bronchoscopy at the initial site and 24 control subjects who remained cancer-free, was subjected to array Comparative Genomic Hybridization (arrayCGH). DNA copy number profiles were related to lesion outcome. Prediction accuracy of the previously defined molecular classifier to predict endobronchial cancer in this series was determined. Unsupervised hierarchical clustering analysis revealed a significant association between cluster assignment and lesion outcome (p< 0.001), independent of histological grade, with quiescent profiles in controls (24/24) and aberrant profiles in the majority of cases (9/12). Our pre-defined classifier demonstrated 92% accuracy for predicting cancer outcome in the current sample series. Our validated classifier holds great promise for stratification of patients with premalignant endobronchial lesions for risk of subsequent cancer. Fresh frozen specimens of 36 premalignant endobronchial biopsies. Test samples were compared to an external pool of normal male/female reference DNA.

Dataset Information

Molecular subtyping reveals immune alterations associated with progression of bronchial premalignant lesions

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets