Abstract

Precision medicine depends on the accurate identification of actionable mutations in a tumor sample. It is unknown how heterogeneous the distribution of such mutations can be in a tumor. Morphological (i.e. histopathological) heterogeneity is well described in lung adenocarcinoma and has been specifically recognized in the most recent official clinico-pathological classification. The most predominant subtype present is now used to classify each lung adenocarcinoma. No molecular profile exists to explain the intratumoral differences in lung adenocarcinoma morphology, despite the consistently observed association between specific predominant subtypes and poorer survival. Given a recent proposal stratifying lung adenocarcinoma into subtypes of differing metastatic potential, we questioned the assumption that major mutations are present uniformly throughout tumors; especially those showing discrete different subtypes. We selected formalin-fixed paraffin embedded lung adenocarcinoma specimens that showed discrete areas of different subtypes, extracted subtype DNA samples from those areas and screened for mutations in hotspot regions of the EGFR, KRAS and BRAF genes using high resolution melting. Sanger sequencing was used to confirm all identified mutations. Chromogenic in situ hybridization (CISH) was used to identify mutant allele specific imbalances in tumors with EGFR mutations. Interestingly, we found that KRAS and BRAF mutations could be confined to morphological domains of higher grade. On the other hand, EGFR mutations were found through all histological subtypes in each tumor consistent with the driver status of this mutation. Intratumoral heterogeneity has major implications for tumorigenesis, chemoresistance and the role of histopathology in molecular screening for precision medicine. This study not only confirms that intratumoral mutational heterogeneity does occur, but also that it is associated with morphologically distinct regions in some tumors. From a practical perspective, small biopsies may not adequately represent a tumor's full mutational profile, particularly for later arising but prognostically important mutations such as those in the KRAS and BRAF genes.

Highlights

  • Tumor heterogeneity has different meanings in different clinical disciplines

  • In 24 of the tumors we identified mutations in one or more of the above genes; one of these being a case of pleomorphic carcinoma (50% giant cell/spindle cell, 30% adenocarcinoma, 20% squamous cell carcinoma) with an EGFR mutation and another a pleomorphic carcinoma comprising solid, micropapillary and giant cell components with heterozygous allelic KRAS mutations of codon 12

  • Intratumoral mutational heterogeneity has been reported in renal cell carcinoma [11], colorectal carcinoma [12] and breast carcinoma [13]

Read more

Summary

Introduction

Tumor heterogeneity has different meanings in different clinical disciplines. To the anatomical pathologist, it is related to observable histologic differences in regions of a tumor; to the molecular pathologist, it relates to mutational variation between cells of the same tumor; and to the medical oncologist, it is related to therapeutic response. The existence of intratumoral mutational heterogeneity in lung adenocarcinoma has been debated since activating EGFR mutations were found to predict response to EGFR tyrosine kinase domain inhibitors. It is clinically evidenced after the selective pressure of effective chemotherapy or molecular targeted therapy when tumors later progress. The conclusions of Yatabe et al were based on multiple random samples from five EGFR mutant lung adenocarcinomas without reference to areas of variation in intratumoral morphology Their hypothesis was that previously reported cases of intratumoral mutational heterogeneity were false negative results associated with low tumor purity, insensitive direct sequencing techniques and pseudo-heterogeneity caused by variation of mutant allele copy number across the tumor [5]. A subsequent Korean study has examined EGFR mutations in different subtypes of ten adenocarcinomas and found no evidence of heterogeneity in the two to three selectively dissected tissues from each tumor [6]

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call