Using Genomics to Differentiate Multiple Primaries From Metastatic Lung Cancer.

INTRODUCTION:Genomic technologies present a promising mechanism of resolving the clinical dilemma of distinguishing independent primary tumors from intrapulmonary metastases in NSCLC. We evaluated the utility of discordant mapping somatic junctions from chromosomal rearrangements in diagnosing metastatic disease compared to the current standard histologic review. MATERIAL AND METHODS:Mate-pair sequencing was performed on DNA extracted from 76 distinct tumors from 37 cases of multiple lung cancers. Discordant mapping junctions and chromosomal copy levels were assessed for each tumor. Blood-derived DNA was available on 22 of these cases for germline assessments. A lung cancer next-generation sequencing panel was additionally performed on tumor pairs from 17 patients. RESULTS:Whereas mate-pair sequencing was able to classify lineage in all tumor pairs, histologic review appeared to misclassify lineage in 9 of 33 (27%) same-histology tumor pair comparisons. Based on disagreement between the reviewing pathologists, histopathologic lineage was classified as indeterminate in seven cases. In two cases where pathologists agreed on a metastatic call, no shared junctions were found suggesting independent primaries. Although germline junctions passing algorithmic filters were common, on average less than three were present and all had predictable structures of small focal rearrangements or transposons. Evaluation of shared chromosomal copy changes and driver mutations through a lung cancer next-generation sequencing panel, while informative, were nondefinitive in calling lineage in all cases. CONCLUSIONS:The highly unique nature and prevalence of chromosomal rearrangement in lung cancers provide a useful and definitive technique for calling lineage in multifocal lung cancer.