Classification of Non-small Cell Lung Carcinoma in Transthoracic Needle Specimens Using MicroRNA Expression Profiling

Abstract

Emerging targeted lung cancer therapies require the accurate morphologic subclassification of non-small cell lung cancer (NSCLC), even in scant and distorted specimens obtained by transthoracic needle aspiration (TTNA). MicroRNAs (miRNAs) are small noncoding genes recently reported as useful in differentiating squamous cell carcinoma (SCC) from adenocarcinoma (AD) in resected tumor specimens. We investigated their ability to do so in TTNA specimens. Smears, immunocytochemistry slides, and corresponding cell blocks of 31 NSCLC TTNA specimens were retrieved and classified as AD or SCC based on their cytologic features and immunocytochemical profiles. Data on EGFR and K-RAS mutational status were available for all cases of AD. We quantified the hsa-let-7 family and hsa-miR-205 by quantitative reverse transcription-polymerase chain reaction and compared the miRNA expression levels in AD and SCC using Student t test. Eighteen cases were classified as AD and 13 as SCC by light microscopy and immunocytochemistry. miRNA expression profiles demonstrated considerable, statistically significant differences between AD and SCC, showing an upregulation of hsa-let-7a, hsa-let-7b, hsa-let-7c,hsa-let-7f, hsa-let-7g, hsa-let-7i, and hsa-miR-98 and a downregulation of hsa-miR-205 in AD specimens (all P < .05; t test). Profiling the hsa-let-7 family and hsa-miR-205 is a promising method for differentiating AD from SCC, even in such small specimens as transthoracic aspirates. Subject to the validation of these findings in further, larger studies, this could prove to be a reliable, standardizable tool for the subclassification of NSCLC.