Refining Lung Cancer Staging in the Molecular Oncology Era

Genomic Sequencing to Distinguish Separate Primary Tumors From Metastatic Lesions

H. Jack West, MD


May 19, 2015

Identification of Independent Primary Tumors and Intrapulmonary Metastases Using DNA Rearrangements in Non-Small-Cell Lung Cancer

Murphy SJ, Aubry MC, Harris FR, et al
J Clin Oncol. 2014;32:4050-4058

Ambiguity in staging of lung nodules is a common challenge. When a lung cancer has been diagnosed by a biopsy of a primary cancer but another lung nodule is present, it can be challenging to distinguish between the nodule being a metastatic lesion or a new primary lung cancer. But the clinical implications are profound. In the case of metastatic disease, the cornerstone of treatment is systemic, and the goal of treatment is almost invariably palliative. For those with a second primary cancer, local therapies to both lesions, potentially with additional interventions, may well prove curative. Pathologist interpretation of biopsies from both lesions may prove helpful, particularly in the fortunate but uncommon situation that the lesions are of different histologies. Can the genetic pattern of DNA arrangements as assessed by next-generation sequencing (NGS) help discriminate the lineage of these lesions more reliably?

To answer this question, Dr Stephen J. Murphy and colleagues from Mayo Clinic performed genetic testing on a control set of multiple defined conditions, evaluated the patterns that emerged, and then applied them to a study set. The learning set included four cases comparing two blocks from the same tumor, three cases comparing cancers of different histology and therefore presumed separate primary cancers in the same patient, and cases comparing four pairs of tumor tissue from a primary tumor and presumed metastatic lesion from the same patient. The study set consisted of 11 cases in which synchronous lung nodules were detected and compared carefully by two pathologists and also by NGS testing for genetic similarities following microdissection of cancer tissue.

Altogether, this genomic analysis reviewed results from 41 tumors that consisted of 33 adenocarcinomas and 8 squamous cancers. Among these, 2203 unique DNA arrangements were identified, with a wide range in number of breakpoints that ranged from 3 to 276 per tumor, averaging approximately 50.

The learning set demonstrated that among biopsies from the same tumor, there was a high concordance of DNA breakpoints, as expected. Each was unique, with no shared genomic alterations across tumors. In contrast, tumors of different histology from the same patient demonstrated no shared DNA breakpoints, with the exception of a single genomic alteration in one case that was also shared with normal tissue and blood DNA and therefor consistent with a germline mutation. Among paired primary tumors and metastatic lesions, NGS testing revealed significant sharing of genomic alterations for each pair, consistent with their shared lineage.

The study set then compared conclusions from the two pathologists with the NGS panel. Pathologists and the genomic prediction model agreed on 8 of 9 cases in which the pathologists were able to offer a clear conclusion, with two cases inconclusive to both pathologists. Specifically, genomic testing corroborated that the tumors were separate primary tumors in four cases in which the pathologists favored this conclusion, and the genomic prediction model agreed with the pathologists in 4 of 5 cases in which the pathologists favored the conclusion of metastatic disease from the same cancer. In a single case ascribed to metastatic disease by pathologists, there were no shared genomic alterations except for a single DNA breakpoint that was also seen in normal tissue and blood from that case and therefore more consistent with a germline mutation and two separate primary cancers. For the two cases in which the pathologists could not favor either conclusion, NGS testing revealed no shared DNA breakpoints, therefore indicative of separate primary tumors.

While availability and cost of NGS testing are certainly barriers to widespread use, this work provides a very provocative proof of principle that genomic testing can provide insights in ambiguous cases that seem more reliable and conclusive than a pathologist's interpretation. In addition to determining whether two lung lesions represent the same cancer, this work suggests that NGS testing can reliably clarify whether metachronous cancers represent recurrence or a new primary lesion, whether metastatic lesions in a patient with a history of more than one cancer represents one or the other, and can even potentially enlighten us about the biology of entities like multifocal adenocarcinoma in situ, such as whether these lesions represent regional spread of the same cancer throughout the lungs or a proclivity to develop similar but distinct cancers from lung tissue at high risk. With genomic testing becoming more readily available and likely less expensive over time, we can look forward to being better able to refine treatment decisions for our patients as well as gaining a better understanding of the behavior of many cancers.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.