Author ORCID Identifier

0000-0001-8189-2132

Date of Graduation

12-2017

Document Type

Thesis (MS)

Program Affiliation

Biomedical Sciences

Degree Name

Masters of Science (MS)

Advisor/Committee Chair

Scott Kopetz

Committee Member

Jeffrey Morris

Committee Member

Russell Broaddus

Committee Member

Nicholas Navin

Committee Member

Michael Overman

Abstract

Introduction: Mutations in KRAS/NRAS (RAS) predict a lack of benefit from anti-EGFR agents in metastatic colorectal cancer (mCRC). As next generation sequencing (NGS) has advanced, we are discovering atypical and low allele frequency mutations. We aimed to evaluate how NGS can optimally define RAS mutant CRC and the role of relative mutant allele frequency (rMAF) as a biomarker.

Methods: Using institutional and public cohorts of mCRC patients with NGS results, we described the prevalence and clinical impact of atypical (not in current guidelines) and low rMAF RAS mutations (RAS MAF by the MAF of the mutated gene with the highest allele frequency to normalize for tumor content. Functional annotation of 113 RAS mutations was performed and functionality of mutations was compared to rMAF.

Results: RAS mutations were noted in 4244/8609 patients (49.3%), with atypical mutations in 1.3% of patients. The most prevalent atypical mutations were KRAS Q22K (0.2%), KRAS D33E (0.1%) and KRAS T50I (0.1%). Of 113 functionally characterized RAS mutations, all 23 non-activating mutations were atypical, while every guideline cited mutation was activating. Atypical variants (HR 2.45, P=0.0092) and those that resulted in MAPK activity greater than KRAS exon 2 (HR 1.40, P=0.028) had a worse OS. A RAS rMAF >50% was associated with worse OS than rMAFRAS mutation was associated with a worse OS than wild-type patients. The rMAF of any mutated gene was also associated with functional significance in a clinically annotated database, yet the magnitude of difference in rMAF was not sufficient to warrant clinical utility in tissue cohorts. However, a cfDNA cohort did show striking results demonstrating rMAF was associated with a variants functional characterization.

Conclusions: Through a comprehensive atlas of RAS functional characterization, we show that several atypical variants appear clinically relevant. Although rMAF was not useful in characterizing variants as damaging, our findings that RAS rMAF is associated with prognosis suggests allele frequency may be useful information in standard clinical reports.

Keywords

colorectal cancer, biomarker, next generation sequencing, allele frequency, mutation, KRAS, NRAS, cfDNA

Share

COinS