Introduction of intermediate serrated polyp through CpG island methylator phenotype and BRAF/KRAS mutation analysis

Abstract

Introduction: Both difficulties in histological distinction and shared molecular features between SSA/Ps and HPs suggest that these two lesions might be of the same spectrum, and it is possible that an intermediate lesion might exist between the two spectral ends of a HP and a SSA/P. We hypothesized that such an intermediate serrated polyp (ISP), lying in the middle of the morphological spectrum between HPs and SSA/Ps, might have an intermediate status between HPs and SSA/Ps in terms of molecular alterations. In the present study, we have analyzed three serrated lesions, a HP, an intermediate serrated polyp, and a SSA/P for BRAF mutation and CpG island methylator phenotype (CIMP) status. Our aim was the identification whether the ISP has an intermediate status of not only morphological features but also of molecular alterations between a HP and a SSA/P. Methods: We retrospectively selected 103 cases of polypectomized hyperplastic polyp (HP) specimens in 2011 at Seoul National University Hospital, Seoul, Korea. Cases were selected on the basis that they were well oriented and had sufficient tissue available for molecular study and immunohistochemical analysis. The samples were divided into three groups: HP, ISP and SSA/P. HPs are subdivided again into microvesicular (MVHP), goblet cell (GCHP) and mucin-poor (MP). DNA methylation analysis, we used real-time PCR (RT-PCR) based MethyLight technology. PCR and sequence analysis of the KRAS codons12 and 13 and BRAF codon 600 were performed. A tissue (2mm in diameter) was obtained from individual paraffin blocks and placed in a new paraffin block using a trephine apparatus (Superbiochips Laboratories, Seoul, Korea). Immunohistochemical staining was performed using the Bond-Max Autostainer (Leica Microsystems, Illinois, USA) (Antibodies: MUC1, MUC2, MUC5AC, MUC6, CDX2, β-catenin, p53, Ki-67, SLIT2, CK7 and CK20 antigens). Results: When we compared the number of methylated genes by histologic subtypes, the number of methylated gene per samples was the highest (1.91) in SSA, followed by ISP (0.92), MVHP (0.25) and GCHP (0.13). There was a general trend toward increased hypermethylation in right-side samples as compared to left-side samples within the same histological category. KRAS mutation was identified in 29.2% of GCHP, 6.3% of MVHP, 3.8% of ISP and 10.7% of SSA. BRAF mutation was identified in 18.8% of MVHP, 38.0% of ISP, 60.0% of SSA, and not observed in GCHP. The mutation status of BRAF and KRAS in GCHP were different from MVHP as well as ISP and SSA. MUC6 was positive in 23.5% of total polyps. MUC6 expression was not observed in GCHP. Differences in MUC6 expression between GCHP and subgroups were observed, while no differences in MUC6 expression between ISPs and SSAs were observed. We found out that CK7 positive was found in MVHP, ISPs and SSAs, and CK7 negative was found in GCHP. We found out that loss of expression of SLIT2 was found SSAs, ISP and MVHP. The characteristic features of GCHPs were different from MVHPs as well as ISPs and SSAs. Discussion: Our findings suggest the presence of an ISP lying in between microvesicular HPs and SSA/Ps in terms of both morphological and molecular aspects. Thus favors the contention that SSA/Ps might arise from microvesicular HPs through ISPs. Combining morphological and molecular analysis, we believe ISPs are the middle step toward to colorectal serrated adenocarcinoma. If a lesion cannot be labeled between MVHPs and SSA/Ps, ISPs can be used for the diagnosis. Also it is important to recognize ISPs as a precursor lesion of colorectal cancer.