A Prospective Validation Study of Lung Cancer Gene Panel Testing Using Cytological Specimens

Abstract

Background: Genetic panel tests require sufficient tissue samples, and therefore, cannot always be performed. Although collecting cytological specimens is easier than tissue collection, there are no validation studies on the diagnostic accuracy of lung cancer gene panel tests using cytology samples. Methods: Using an amplicon-based high-sensitivity next-generation sequencing panel test capable of measuring eight druggable genes, we prospectively enrolled consecutive patients who underwent diagnostic procedures. We evaluated the analysis accuracy rate, nucleic acid yield, and the quality of cytological specimens under brushing, needle aspiration, and pleural effusion. We then compared these specimens with collected tissue samples. Results: In 163 prospectively enrolled cases, nucleic acid extraction and analysis accuracy was 100% in cases diagnosed with adenocarcinoma. Gene mutations were found in 68.7% of cases with 99.5% (95% CI: 98.2–99.9) concordance to companion diagnostics. The median DNA/RNA yield and DNA/RNA integrity number were 475/321 ng and 7.9/5.7, respectively. The correlation coefficient of the gene allele ratio in 64 cases compared with tissue samples was 0.711. Conclusion: The success of gene analysis using cytological specimens was high, and the yield and quality of the extracted nucleic acid were sufficient for panel analysis. Moreover, the allele frequency of gene mutations in cytological specimens showed high correlations with tissue specimens.