Abstract

Approximately 40% of colon cancer harbor Kirsten rat sarcoma viral oncogene ( KRAS ) mutations, but the prognostic value of KRAS mutations in colon cancer is still controversial. We enrolled 412 colon adenocarcinoma (COAD) patients with KRAS mutations, 644 COAD patients with KRAS wild-type and 357 COAD patients lacking information on KRAS status from five independent cohorts. A random forest model was developed to estimate the KRAS status. The prognostic signature was established using least absolute shrinkage and selection operator-Cox regression and evaluated by Kaplan-Meier survival analysis, multivariate-Cox analysis, receiver operating characteristic curve and nomogram. The expression data of KRAS -mutant COAD cell lines from the Cancer Cell Line Encyclopedia database and the corresponding drug sensitivity data from the Genomics of Drug Sensitivity in Cancer database were used for potential target and agent exploration. We established a 36-gene prognostic signature classifying the KRAS -mutant COAD as high and low risk. High risk patients had inferior prognoses compared to those with low risk, while the signature failed to distinguish the prognosis of COAD with KRAS wild-type. The risk score was the independent prognostic factor for KRAS -mutant COAD and we further fabricated the nomograms with good predictive efficiency. Moreover, we suggested FMNL1 as a potential drug target and three drugs as potential therapeutic agents for KRAS -mutant COAD with high risk. We established a precise 36-gene prognostic signature with great performance in prognosis prediction of KRAS -mutant COAD providing a new strategy for personalized prognosis management and precision treatment for KRAS -mutant COAD.

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