A phase 3 multicenter study to evaluate the accuracy and clinical usefulness of the high-sensitivity next-generation sequencing panel using cytological specimens for lung cancer (cPANEL)

<bold>Background:</bold> The diagnosis of lung cancer can be delayed due to difficulties in obtaining tissue samples by biopsy. Instead, cytology or liquid samples may be used as an alternative. To date, there have been no prospective multicenter studies to estimate whether cytology specimens can replace tissue samples using lung cancer gene panel analysis. <bold>Methods:</bold> We prospectively validate an amplicon-based high-sensitivity next generation sequencing panel test using cytological specimens in a multicenter phase 3 study. The primary endpoint was to confirm gene analysis success rate superiority using cytological specimens against the assumption of a 90% success rate for conventional panel testing using tissue specimens. <bold>Results:</bold> Full analysis set of 248 cases were included. The success rate for gene panel analysis using cytology specimens was 98.4% (95%CI: 95.9-99.6%) with a positive concordance rate of 97.4% (95%CI 91.0-99.7%) by other companion diagnostic kit. Among 150 cases of lung adenocarcinoma, driver gene mutation was detected in 93 cases (62%). The median value for nucleic acid yield and quality (DNA/RNA integrated number) of cytology specimens was 546.0/426.5 ng and 9.2/4.7 for DNA/RNA, respectively. For mutant cases, Pearson correlation coefficient of variant allele frequency between tissue formalin-fixed and paraffin-embedded (FFPE) sample and cytology specimens was 0.815. <bold>Conclusion:</bold> The success rate was significantly higher for gene analysis using cytological specimens. Furthermore, due to the sufficient nucleic acid yield, high quality, and high correlation of mutant alleles frequency found in comparison with FFPE, cytological specimens are considered well-suited for panel testing as tissue substitutes.