Correlation Between Tumor-Specific Mutated and Methylated DNA in Colorectal Cancer

Abstract

PURPOSE Analysis of circulating tumor DNA (ctDNA) is a potential improvement in precision medicine. In colorectal cancer (CRC), somatic mutations such as RAS and RAF in the blood (mut-ctDNA) are investigated for prognostic and predictive purposes. However, they are only present in approximately 60% of patients. Recently, ctDNA has been detected in patients with RAS/ RAF wild type (WT) by methylated ctDNA (meth-ctDNA). The aim of this study was to compare mutated DNA with methylated DNA in malignant and nonmalignant tissue and plasma from CRC cohorts to establish a universal biomarker for ctDNA in all patients with CRC. MATERIALS AND METHODS Tissue (n = 170) and plasma (n = 147) samples were analyzed for RAS/ RAF mutations and neuropeptide Y methylation by droplet digital polymerase chain reaction. Tissue originated from nonmalignant WT and RAS/ RAF-mutated adenomas, tumor-adjacent colorectal tissue, and WT and RAS/ RAF-mutated tumor tissue. Plasma samples represented healthy donors and localized and metastatic CRCs. RESULTS The level of neuropeptide Y–methylated DNA in the tissue cohorts differed between nonmalignant and malignant/premalignant tissues with minimal overlap. Furthermore, meth-ctDNA was detected in plasma from 100% of patients with metastatic disease, compared with 67% of those with localized disease and 8% of healthy donors. Median fraction of meth-ctDNA in metastatic and localized cancers was 13.25% and 0.04%, respectively. Correlation between mut-ctDNA and meth-ctDNA was high ( r = 0.77 and 0.80 in localized and metastatic settings, respectively). CONCLUSION Mut-ctDNA is interchangeable with meth-ctDNA in patients with CRC. On the basis of our results, meth-ctDNA should be considered a universal biomarker in metastatic CRC, but additional investigations of clinical utility are warranted.

Why the Therapeutic Impact of RAS Mutation Clearance in Plasma ctDNA Deserves to Be Further Explored in Metastatic Colorectal Cancer

“…In this study, NPY was discarded due to positivity in normal mucosa and in whole blood samples (which contain white blood cells as the main contaminant of cfDNA), and WIF1 was found to have limited differential methylation between normal healthy/normal adjacent mucosa and tumor. Other recently published studies found NPY methylation in the plasma of healthy donors (14) and non-cancer patients (15). Therefore, to date, we are far from true agreement on the use of WIF-1/NPY methylation testing to exclude or confirm the presence of ctDNA.…”

Comparison of Mutated KRAS and Methylated HOXA9 Tumor-Specific DNA in Advanced Lung Adenocarcinoma

“…Meth-ctDNA has already been suggested as a universal cancer-specific biomarker in previous studies [ 14 , 17 , 31 ]. The present study finds that meth-HOXA9 is a sensitive and specific biomarker of advanced lung adenocarcinoma in tumor tissue, and it has promising qualities for use in liquid biopsies.…”

“…Genomic DNA was extracted from formalin-fixed, paraffin-embedded (FFPE) lung tissue (tumor or non-tumor), as previously described [ 31 ]. In tumor tissue, areas with a high degree of tumor cells were selected by an experienced pathologist.…”

“…KRAS mutations were analyzed on DNA from the tumor samples as well as on plasma DNA. The ddPCR method was previously reported [ 31 ]. The analysis was performed as described above for meth-HOXA9, except for the bisulfite conversion of the DNA.…”

“…The limit of blank in plasma was only established for the KRAS G12D mutation, which is one of the common mutations. Plasma from healthy donors ( n = 20) was analyzed by ddPCR, and <2 positive droplets were observed in all analyses [ 31 ]. A pool of lymphocyte DNA from healthy donors was analyzed for all KRAS mutations, and <3 mutation-positive droplets were found in all analyses.…”