Background:Colorectal cancer (CRC) is a heterogeneous disease in terms of molecular carcinogenic pathways. Based on recent findings regarding the multiple serrated neoplasia pathway, we revised an eight-marker panel for a new CIMP classification system.Methods:1370 patients who received surgical resection for CRCs were classified into three CIMP subtypes (CIMP-N: 0–4 methylated markers, CIMP-P1: 5–6 methylated markers and CIMP-P2: 7–8 methylated markers). Our findings were validated in a separate set of high-risk stage II or stage III CRCs receiving adjuvant fluoropyrimidine plus oxaliplatin (n=950).Results:A total of 1287/62/21 CRCs cases were classified as CIMP-N/CIMP-P1/CIMP-P2, respectively. CIMP-N showed male predominance, distal location, lower T, N category and devoid of BRAF mutation, microsatellite instability (MSI) and MLH1 methylation. CIMP-P1 showed female predominance, proximal location, advanced TNM stage, mild decrease of CK20 and CDX2 expression, mild increase of CK7 expression, BRAF mutation, MSI and MLH1 methylation. CIMP-P2 showed older age, female predominance, proximal location, advanced T category, markedly reduced CK20 and CDX2 expression, rare KRAS mutation, high frequency of CK7 expression, BRAF mutation, MSI and MLH1 methylation. CIMP-N showed better 5-year cancer-specific survival (CSS; HR=0.47; 95% CI: 0.28–0.78) in discovery set and better 5-year relapse-free survival (RFS; HR=0.50; 95% CI: 0.29–0.88) in validation set compared with CIMP-P1. CIMP-P2 showed marginally better 5-year CSS (HR=0.28, 95% CI: 0.07–1.22) in discovery set and marginally better 5-year RFS (HR=0.21, 95% CI: 0.05–0.92) in validation set compared with CIMP-P1.Conclusions:CIMP subtypes classified using our revised system showed different clinical outcomes, demonstrating the heterogeneity of multiple serrated precursors of CIMP-positive CRCs.

Our results indicate that the density of intratumoural macrophages is a useful prognostic indicator for further stratifying T cell populations in CRC.

BackgroundFormalin-fixed, paraffin-embedded (FFPE) tissues are important resources for profiling DNA methylation changes and for studying a variety of diseases. However, formalin fixation introduces inter-strand crosslinking, which might cause incomplete bisulfite conversion of unmethylated cytosines, which might lead to falsely elevated measurements of methylation levels in pyrosequencing assays. Long interspersed nucleotide element-1 (LINE-1) is a major constituent of repetitive transposable DNA elements, and its methylation is referred to correlates with global DNA methylation. To identify whether formalin fixation might impact the measured values of methylation in LINE-1 repetitive elements and whether prolonged heat-induced denaturation of DNA might reduce the artificial increases in measured values caused by formalin fixation, we analyzed paired fresh-frozen (FF) and FFPE xenograft tissue samples for their methylation levels in LINE-1 using a pyrosequencing assay. To further confirm the effect of a heating step in the measurement of LINE-1 or single gene methylation levels, we analyzed FFPE tissue samples of gastric cancer and colorectal cancer for their methylation status in LINE-1 and eight single genes, respectively.ResultsFormalin fixation led to an increase in the measured values of LINE-1 methylation regardless of the duration of fixation. Prolonged heating of the DNA at 95 °C for 30 min before bisulfite conversion was found (1) to decrease the discrepancy in the measured values between the paired FF and FFPE tissue samples, (2) to decrease the standard deviation of the measured value of LINE-1 methylation levels in FFPE tissue samples of gastric cancer, and (3) to improve the performance in the measurement of single gene methylation levels in FFPE tissue samples of colorectal cancer.ConclusionsFormalin fixation leads to artificial increases in the measured values of LINE-1 methylation, and the application of prolonged heating of DNA samples decreases the discrepancy in the measured values of LINE-1 methylation between paired FF and FFPE tissue samples. The application of prolonged heating of DNA samples improves bisulfite conversion-based measurement of LINE-1 or single gene methylation levels in FFPE tissue samples.Electronic supplementary materialThe online version of this article (doi:10.1186/s13148-016-0308-0) contains supplementary material, which is available to authorized users.

Purpose: Despite the well-known prognostic value of the tumorimmune microenvironment (TIME) in colorectal cancers, objective and readily applicable methods for quantifying tumor-infiltrating lymphocytes (TIL) and the tumor-stroma ratio (TSR) are not yet available.Experimental Design:We established an open-source softwarebased analytic pipeline for quantifying TILs and the TSR from whole-slide images obtained after CD3 and CD8 IHC staining. Using a random forest classifier, the method separately quantified intraepithelial TILs (iTIL) and stromal TILs (sTIL). We applied this method to discovery and validation cohorts of 578 and 283 stage III or high-risk stage II colorectal cancers patients, respectively, who were subjected to curative surgical resection and oxlaliplatin-based adjuvant chemotherapy.Results: Automatic quantification of iTILs and sTILs showed a moderate concordance with that obtained after visual inspection by a pathologist. The K-means-based consensus clustering of 197 TIME parameters that showed robustness against interobserver variations caused colorectal cancers to be grouped into five distinctive subgroups, reminiscent of those for consensus molecular subtypes (CMS1-4 and mixed/intermediate group). In accordance with the original CMS report, the CMS4-like subgroup (cluster 4) was significantly associated with a worse 5-year relapse-free survival and proved to be an independent prognostic factor. The clinicopathologic and prognostic features of the TIME subgroups have been validated in an independent validation cohort.Conclusions: Machine-learning-based image analysis can be useful for extracting quantitative information about the TIME, using whole-slide histopathologic images. This information can classify colorectal cancers into clinicopathologically relevant subgroups without performing a molecular analysis of the tumors.

Background:The association between the CpG island methylator phenotype (CIMP) and clinical outcomes in metastatic colorectal cancer remains unclear. We investigated the prognostic impact of CIMP in patients with metastatic colorectal cancer treated with systemic chemotherapy.Methods:Eight CIMP-specific promoters (CACNA1G, IGF2, NEUROG1, RUNX3, SOCS1, CDKN2A, CRABP1, and MLH1) were examined. The CIMP status was determined by the number of methylated promoters as high (⩾5), low (1–4), and negative (0).Results:A total of 153 patients were included (men/women, 103/50; median age, 61 years; range, 22–80 years). The CIMP status was negative/low/high in 77/ 69/7 patients, respectively. Overall survival (OS) was significantly different among the three CIMP groups, with median values of 35.7, 22.2, and 9.77 months for the negative, low, and high groups, respectively (P<0.001). For patients treated with fluoropyrimidine and oxaliplatin first-line chemotherapy (N=128), OS and progression-free survival (PFS) were significantly different among the three CIMP groups; the median OS was 37.9, 23.8, and 6.77 months for the negative, low, and high groups, respectively (P<0.001), while the median PFS was 9.97, 7.87, and 1.83 months, respectively (P=0.002). Response rates were marginally different among the three CIMP groups (53.4% vs 45.1% vs 16.7%, respectively; P=0.107). For patients treated with fluoropyrimidine and irinotecan second-line chemotherapy (N=86), only OS showed a difference according to the CIMP status, with median values of 20.4, 13.4, and 2.90 months for the negative, low, and high groups, respectively (P<0.001).Conclusions:The CIMP status is a negative prognostic factor for patients with metastatic colorectal cancer treated with chemotherapy.