ABSTRACT
We introduce a new AI-ready computational pathology dataset containing restained and co-registered digitized images from eight head-and-neck squamous cell carcinoma patients. Specifically, the same tumor sections were stained with the expensive multiplex immunofluorescence (mIF) assay first and then restained with cheaper multiplex immunohistochemistry (mIHC). This is a first public dataset that demonstrates the equivalence of these two staining methods which in turn allows several use cases; due to the equivalence, our cheaper mIHC staining protocol can offset the need for expensive mIF staining/scanning which requires highly-skilled lab technicians. As opposed to subjective and error-prone immune cell annotations from individual pathologists (disagreement > 50%) to drive SOTA deep learning approaches, this dataset provides objective immune and tumor cell annotations via mIF/mIHC restaining for more reproducible and accurate characterization of tumor immune microenvironment (e.g. for immunotherapy). We demonstrate the effectiveness of this dataset in three use cases: (1) IHC quantification of CD3/CD8 tumor-infiltrating lymphocytes via style transfer, (2) virtual translation of cheap mIHC stains to more expensive mIF stains, and (3) virtual tumor/immune cellular phenotyping on standard hematoxylin images. The dataset is available at https://github.com/nadeemlab/DeepLIIF.
ABSTRACT
We introduce a new AI-ready computational pathology dataset containing restained and co-registered digitized images from eight head-and-neck squamous cell carcinoma patients. Specifically, the same tumor sections were stained with the expensive multiplex immunofluorescence (mIF) assay first and then restained with cheaper multiplex immunohistochemistry (mIHC). This is a first public dataset that demonstrates the equivalence of these two staining methods which in turn allows several use cases; due to the equivalence, our cheaper mIHC staining protocol can offset the need for expensive mIF staining/scanning which requires highly-skilled lab technicians. As opposed to subjective and error-prone immune cell annotations from individual pathologists (disagreement > 50%) to drive SOTA deep learning approaches, this dataset provides objective immune and tumor cell annotations via mIF/mIHC restaining for more reproducible and accurate characterization of tumor immune microenvironment (e.g. for immunotherapy). We demonstrate the effectiveness of this dataset in three use cases: (1) IHC quantification of CD3/CD8 tumor-infiltrating lymphocytes via style transfer, (2) virtual translation of cheap mIHC stains to more expensive mIF stains, and (3) virtual tumor/immune cellular phenotyping on standard hematoxylin images. The dataset is available at \url{https://github.com/nadeemlab/DeepLIIF}.
ABSTRACT
Head and neck squamous cell carcinoma (HNSCC) is an aggressive epithelial malignancy characterized by frequent mutations and metastasis. Long noncoding RNAs (lncRNAs) have been implicated in tumorigenesis and serve as novel prognostic biomarkers in different cancers. To enhance our understanding of lncRNAs that may have biological significance in HNSCC and may serve as prognostic biomarkers, we globally profiled lncRNAs in HNSCC by analyzing the RNA-seq data from The Atlas of Noncoding RNAs in Cancer (TANRIC) database. Of 3576 lncRNAs, we identified 926 (higher-688, lower-238) lncRNAs with a 2-fold abundance difference among the forty HNSCC and paired adjacent normal tissue. We investigated differential abundance of lncRNAs based on TP53 mutation and p16 status. We found 133 lncRNAs to have differential abundance by 2-fold among the mutant vs wild-type TP53 samples, whereas among p16-negative vs positive samples, we identified 710 lncRNAs with the same criteria. Meanwhile, analysis of the 15 most abundant lncRNAs in the tumor samples identified five lncRNAs whose higher abundance was associated with poor overall patient survival. Among these five, higher abundance of LINC00460 associated with poor patient survival in an independent cohort of 82 HNSCC patients. To further evaluate the potential function of LINC00460, we performed lncRNA-mRNAs co-expression analysis and found that higher abundance of LINC00460 associated with cancer-related biological pathways including EMT and other inflammatory response pathways. In summary, we report LINC00460 is more abundant in tumors compared to adjacent normal tissue and that it may serve as a potential prognostic biomarker in HNSCC.
ABSTRACT
The molecular determinants of lung cancer risk remain largely unknown. Airway epithelial cells are prone to assault by risk factors and are considered to be the primary cell type involved in the field of cancerization. To investigate risk-associated changes in the bronchial epithelium proteome that may offer new insights into the molecular pathogenesis of lung cancer, proteins were identified in the airway epithelial cells of bronchial brushing specimens from risk-stratified individuals by shotgun proteomics. Differential expression of selected proteins was validated by parallel reaction monitoring mass spectrometry in an independent set of individual bronchial brushings. We identified 2,869 proteins, of which 312 proteins demonstrated a trend in expression. Pathway analysis revealed enrichment of carbohydrate metabolic enzymes in high-risk individuals. Glucose consumption and lactate production were increased in human bronchial epithelial BEAS2B cells treated with cigarette smoke condensate for 7 months. Increased lipid biosynthetic capacity and net reductive carboxylation were revealed by metabolic flux analyses of [U-13C5] glutamine in this in vitro model, suggesting profound metabolic reprogramming in the airway epithelium of high-risk individuals. These results provide a rationale for the development of potentially new chemopreventive strategies and selection of patients for surveillance programs.
