Background: Pancreatic ductal adenocarcinoma (PDAC) has a poor prognosis. Variant profiling is crucial for developing personalized treatment and elucidating the etiology of this disease.
Methods: Patients with PDAC undergoing surgery from 2007 to 2012 (n 73) were followed from diag-nosis until death or the end of the study. We applied an Anchored Multiplex PCR (AMP™)-based next-generation sequencing (NGS) method to a panel of 65 selected genes and assessed analytical performance by sequencing a quantitative multiplex DNA reference standard. In clinical PDAC samples, detection of low-level KRAS (Kirsten rat sarcoma viral oncogene homolog) mutations was val-idated by allele-specific PCR and digital PCR. We com-pared overall survival of patients according to KRAS mu-tation status by log-rank test and applied logistic regression to evaluate the association between smoking and tumor variant types.
Results: The AMP-based NGS method could detect variants with allele frequencies as low as 1% given sufficient sequencing depth (>1500x). Low-frequency KRAS G12 mutations (allele frequency 1%–5%) were all confirmed by allele-specific PCR and digital PCR. The most prevalent genetic alterations were in KRAS (78% of patients), TP53 (tumor protein p53) (25%), and SMAD4 (SMAD family member 4) (8%). Overall sur-vival in T3-stage PDAC patients differed among KRAS mutation subtypes (P=0.019). Transversion variants were more common in ever-smokers than in never-smokers (odds ratio 5.7; 95% CI 1.2–27.8).
Conclusions: The AMP-based NGS method is applicable for profiling tumor variants. Using this approach, we demonstrated that in PDAC patients, KRAS mutant sub-type G12V is associated with poorer survival, and that transversion variants are more common among smokers.