CpG island methylator phenotype (CIMP) of colorectal cancer is best characterised by quantitative DNA methylation analysis and prospective cohort studies

Background: The concept of CpG island methylator phenotype ( CIMP) is not universally accepted. Even if specific clinicopathological features have been associated with CIMP, investigators often failed to demonstrate a bimodal distribution of the number of methylated markers, which would suggest CIMP as a distinct subtype of colorectal cancer. Previous studies primarily used methylation specific polymerase chain reaction which might detect biologically insignificant low levels of methylation. Aim: To demonstrate a distinct genetic profile of CIMP colorectal cancer using quantitative DNA methylation analysis that can distinguish high from low levels of DNA methylation. Materials and methods: We developed quantitative real time polymerase chain reaction (MethyLight) assays and measured DNA methylation (percentage of methylated reference) of five carefully selected loci (promoters of CACNA1G, CDKN2A (p16), CRABP1, MLH1, and NEUROG1) in 460 colorectal cancers from large prospective cohorts. Results: There was a clear bimodal distribution of 80 microsatellite instability-high (MSI-H) tumours according to the number of methylated promoters, with no tumours showing 3/5 methylated loci. Thus we defined CIMP as having >= 4/5 methylated loci, and 17% (78) of the 460 tumours were classified as CIMP. CIMP was significantly associated with female sex, MSI, BRAF mutations, and wild-type KRAS. Both CIMP MSI-H tumours and CIMP microsatellite stable (MSS) tumours showed much higher frequencies of BRAF mutations (63% and 54%) than non-CIMP counterparts (non-CIMP MSI-H (0%, p < 10(-5)) and non-CIMP MSS tumours (6.6%, p < 10(-4)), respectively). Conclusion: CIMP is best characterised by quantitative DNA methylation analysis. CIMP is a distinct epigenotype of colorectal cancer and may be less frequent than previously reported.