Performance and considerations in the use of diagnostic mutation panels for clonality testing in non-small-cell lung carcinoma.

INTRODUCTION

Next-generation sequencing (NGS) mutation panels are widely implemented and commonly applied to aid clonality classification for non-small-cell lung carcinoma (NSCLC) patients with multiple tumors. Performance of different NGS panels for clonality classification, however, remains unresolved.

METHODS

We assembled 210 primary and metastatic pairs from lung adenocarcinoma (LUAD) and lung squamous-cell carcinomas (LUSC) of the TRACERx421 cohort for which gold standard clonality was confirmed by whole exome sequencing. We used four NGS panels ranging from 12 to 523 genes to determine clonality using the 2024 International Association for the Study of Lung Cancer (IASLC) recommendations.

RESULTS

With an oncogene panel, 30% LUAD and 74% LUSC pairs remained inconclusive with, respectively, 2% and 0% misclassified. Addition of tumor suppressor genes results in 5% LUAD and 5% LUSC inconclusive and, respectively, 2% and 1% misclassified. For large panels, 0%-1% was inconclusive and 1% misclassified for both LUAD and LUSC. Misclassifications occurred due to discordant KRAS mutations in clonal pairs or coincidentally shared PIK3CA mutations in non-clonal pairs.

CONCLUSION

Oncogene panels result in many inconclusive results, most of which can be resolved by adding tumor suppressor genes. Notwithstanding, 1%-2% of patients remain challenging. Large NGS panels detect mutations in more genes than the IASLC recommendations, allowing definitive clonality classification.