Loading
Running six published subtype classifiers over 574 tumours, the authors found they agreed in 88% of cases, and the 12% they disagreed on were genuine hybrids with intermediate survival and intermediate KRAS allele imbalance.
Sequencing data for 574 pancreatic ductal adenocarcinomas from prospective trials and public databases were classified by six published strategies (Moffitt regression and clustering, Collisson, Bailey, Karasinska). Basal-like and classical calls were concordant in 88% and survival differed by subtype. Twelve percent of tumours were discordant and showed intermediate survival in univariate and multivariate analyses, hybrid expression profiles with classical and basal-like genes co-expressed, and intermediate mutant KRAS allelic imbalance (P < 0.001). Nearly 1 in 6 patients had tumours that failed to fall reliably into either subtype on the two regression tools aimed at clinical practice.
A subtype report should carry a confidence, and a trial that dichotomises subtype will misplace about one patient in eight.
Shares Clinical Cancer Research, RNA sequencing & expression profiling, Pancreatic ductal adenocarcinoma.
Shares RNA sequencing & expression profiling, KRAS, Pancreatic ductal adenocarcinoma.
Shares BC Cancer, Clinical Cancer Research, RNA sequencing & expression profiling, KRAS.
Shares KRAS, Pancreatic ductal adenocarcinoma.
Shares Clinical Cancer Research, KRAS, Pancreatic ductal adenocarcinoma.
Shares Clinical Cancer Research, Pancreatic ductal adenocarcinoma.
Shares Clinical Cancer Research, RNA sequencing & expression profiling, Pancreatic ductal adenocarcinoma.
Shares Clinical Cancer Research, RNA sequencing & expression profiling, Pancreatic ductal adenocarcinoma.