Detection of KRAS mutations in circulating tumour DNA from plasma and urine of patients with colorectal cancer

Abstract

Background: Circulating tumour DNA (ctDNA) is very useful for purposes of cancer genetics; however, it has some limitations. Recently, ctDNA in body fluids, such as urine, sputum, and pleural effusion, has been investigated. The aim of this study was to evaluate the quantity of ctDNA derived from urine (trans-renal ctDNA) and the accuracy of KRAS mutation detection in relation to disease stage in colorectal cancer.

Methods: Urine, plasma, and tissue samples were collected from consecutively resected colorectal cancer patients. DNA was extracted from each sample and the quantity was determined. From each DNA sample, KRAS mutations were detected using droplet digital PCR.

Results: 200 patients participated and KRAS mutations were detected in 84 patients (42.0%) from tumour tissue. The concentration of trans-renal ctDNA (trtDNA) was significantly lower than that of plasma; however, there was no significant difference between the sensitivity using ctDNA and that using trtDNA (29.8% VS 33.3%, p = 0.62). Concordance between these two tests was only 17.5%. Combination analysis (ctDNA + trtDNA) improved the sensitivity to 53.6%, and sensitivity was significantly higher than that of corresponding single assays (p = 0.003). In early cancer stages, trtDNA had greater sensitivity for detecting KRAS mutations than ctDNA (37.7% vs. 21.3%, p = 0.047). Conversely, it was less useful for advanced cancer stages (21.7% vs. 52.2%, p = 0.07). Notably, KRAS mutations were detected using ctDNA or trtDNA in 12 of 116 (10.3%) patients who had no KRAS mutations in their tissue samples.

Conclusions: trtDNA and ctDNA have equal potential and combination analysis significantly improved the sensitivity.

Keywords: Circulating tumour DNA; Colorectal cancer; Liquid biopsy; Minimal residual disease; Trans-Renal tumour DNA.