Mutations in TP53 and other heterogeneous genes help to distinguish metastases from new primary malignancies

Background Distinguishing metastases from new primary malignancies or vice versa is important because misclassification can result in inappropriate management. However, for some cases this distinction can be challenging, particularly for squamous cell carcinomas in which the usual surgical pathology approach, predominantly morphology and immunohistochemistry, are frequently non-contributory. We analysed tumor-associated mutations in order to determine whether they could help with this diagnostic dilemma.

Methods Mutations in specific genes were identified with cBioPortal, a large publically available tumor sequence data set. Genes were selected based upon either their high overall prevalence of mutation, or their inclusion in an in-house tumor sequencing set. Tumor types analysed included various common adenocarcinomas, squamous cell carcinomas from multiple sites, urothelial carcinoma and melanoma. Individual mutations and sets of mutations within gene cohorts were compared by their diversity (or heterogeneity) index, prevalence and cumulative prevalence. We demonstrated the utility of this method by performing in-house sequencing of candidate genes in tumors from three patients for which morphology and immunohistochemistry were unable to distinguish between a metastasis and a new primary malignancy.

Results Sequence data from relatively small cohorts of candidate genes readily identified highly diverse, low prevalence mutation profiles in most common malignancies including squamous cell carcinomas. The diversity index predicted the likelihood of an identical mutation profile occurring in an unrelated tumor. High yield gene cohorts could be predicted based on the primary tumor type. Most cohorts included TP53 due to both its high mutation prevalence and high mutation index of diversity. Identical, low prevalence mutations in multiple tumors from patients in the three case studies provided strong diagnostic certainty for metastases rather than new primary malignancies.

Conclusions Most common tumors, including squamous cell carcinomas, have a readily identifiable mutation profiles that occur at a sufficiently low prevalence to effectively barcode or fingerprint the tumor. An identical mutation profile in a primary tumor and a new lesion provides strong evidence for a metastasis and effectively excludes a new primary malignancy, providing diagnostic confidence and aiding clinical management certainty.