Accurate Classification of Non-Small Cell Lung Carcinoma Using a Novel MicroRNA-Based Approach
JA Bishop, H Benjamin, H Cholakh, A Chajut, DP Clark, WH Westra. The Johns Hopkins Medical Institutions, Baltimore, MD; Rosetta Genomics Ltd., Rehovot, Israel; Rosetta Genomics Inc., Philadelphia, PA
Background: Advances in targeted lung cancer therapy now demand accurate classification of non-small cell lung cancer (NSCLC). MicroRNAs (miRNAs) are recently discovered short, non-coding genes that play essential roles in tissue differentiation during normal development and tumorigenesis. For example, hsa-miR-205 is a miRNA that is highly expressed in lung squamous cell carcinomas (SqCCs) but not in lung adenocarcinomas (ACs). The differential expression of miRNAs could be exploited to distinguish these tumor types.
Design: 102 resected NSCLCs were classified as SqCC or AC based on their histologic features and immunohistochemical profiles. Corresponding pre-operative biopsies/aspirates that had been originally diagnosed as poorly differentiated NSCLCs were available for 21 cases. A qRT-PCR diagnostic assay which measures the expression level of hsa-miR-205 was used to classify the carcinomas as SqCC or AC based solely on expression levels. The two sets of diagnoses were compared.
Results: Using standard pathologic methods of classification (i.e. microscopy and immunohistochemistry), 52 resected lung carcinomas were classified as SqCCs and 50 as ACs. There was 100% concordance between the diagnoses established by conventional and miRNA-based methods. MiRNA profiling also correctly classified 20 of the 21 preoperative biopsy specimens.
Conclusions: MiRNA profiling is a highly reliable strategy for classifying NSCLCs. Indeed, classification is consistently accurate even in small biopsies/aspirates of poorly differentiated tumors. Confirmation of its reliability across the full range of tumor grades and specimen types represents an important step towards broad application.
Online Posters2ViewTM Session: Pulmonary