Editors' ChoiceCancer

The clone wars

See allHide authors and affiliations

Science Translational Medicine  15 Feb 2017:
Vol. 9, Issue 377, eaam6061
DOI: 10.1126/scitranslmed.aam6061

Abstract

Genome sequencing of patients with pancreatic adenocarcinoma reveals high concordance of driver alterations across metastases.

Pancreatic cancer commonly metastasizes to distant organs and responds poorly to systemic anticancer therapies. Understanding how tumors spread and the role of intraindividual tumor heterogeneity can provide insights into the biology of metastasis and identify new therapeutic approaches. Makohon-Moore et al. examined 26 sites of metastasis in four patients with untreated metastatic pancreatic adenocarcinoma by whole-genome sequencing and compared these sites with multiple regions of the primary tumor. Although there were differences between all tumor sites in individual patients, mutations involving reported “driver” genes (including KRAS, TP53, SMAD4, CDKN2A, ARID1A, ATM) were remarkably uniform. Genetic “driver” mutations provide a selective growth advantage to tumor cells, whereas “passenger” mutations do not directly affect tumor formation or clonal expansion. Genomic copy number alterations involving shared driver gene regions also tended to be greater and more focal compared with private passenger alterations.

To elucidate patterns of evolution and quantify the degree of genetic heterogeneity in patients, the investigators applied targeted sequencing and mathematical evolutionary modeling to calculate the evolutionary distance (“divergence”) and similarity between tumor sites. Overall, there was a high level of similarity involving both driver and passenger genes among the founder cells that seeded metastatic lesions in individual patients. Phylogenetic analyses revealed that multiple subclones within a primary tumor are capable of metastatic spread, giving rise to highly related yet independent metastases. There was no evidence of metastasis-to-metastasis seeding. Overall, these data suggest that pancreatic cancer undergoes at least one major episode of selection that affects the majority of cells within the primary tumor and results in genomic changes that are maintained and perhaps even responsible for distant spread.

Results from this study suggest that targeting shared drivers in pancreatic adenocarcinoma may be an effective therapeutic approach, though unfortunately most involve tumor suppressors and other difficult to target mutations. Nonetheless, it does not appear as if clonal heterogeneity in itself can explain the lethal pancreatic cancer phenotype, paving the way for future studies aimed toward understanding how epigenetic and microenvironment changes—as well as immune, host, and other factors—might cooperate with these genomic changes to shape clonal architecture and drive tumor progression and ultimately death.

Highlighted Article

View Abstract

Navigate This Article