We retrospectively examined 177 lesions, comprising 51 and 126 CRAs and CDAs, respectively, which were obtained from patients who underwent surgical or endoscopic resection for GC at Iwate Medical University Hospital (Iwate, Japan) between 2010 and 2018. We defined CRA as an adenocarcinoma having branching or anastomosing glands resembling the shapes of the letters W, H, Y or X composed of neoplastic epithelium with low-grade nuclear atypia, as previously proposed [5, 7]. Meanwhile, CDA was defined as an adenocarcinoma that is characterized by tumors having tubular or papillary formations (i.e., CDA corresponds to differentiated-type cancers). The resected specimens were fixed in 10% buffered formalin. After paraffin embedding, representative 3 µm-thick sections were stained with hematoxylin and eosin (H&E) for immunohistochemistry analysis.
The sections were examined by at least two pathologists (T.S. and Y.F.). Histological classification of the tumors in this study was evaluated according to the Japanese classification of gastric carcinoma (the 15th edition)  and diagnosis of pTis (dysplasia) and pT1a (intramucosal adenocarcinoma) was made based on the previously described criteria .
The study was approved by the Ethical Research Committee of Iwate Medical University (H29-78 and HGH29-17).
Immunohistochemical analysis of Muc2 (Ccp58, dilution 1:100; Leica Biosystems, Nußloch, Germany), Muc5AC (CLH2, dilution 1:100; Leica Biosystems), Muc6 glycoprotein (CLH5, dilution 1:100; Leica Biosystems), CD10 (56C6, ready to use; Agilent Technologies, California, USA), CDX-2 (DAK-CDX2, ready to use; Agilent Technologies), p53 (Dako), β-catenin (β-Catenin-1, ready to use; Agilent Technologies) and MLH-1 (ES05, ready to use; Agilent Technologies) was conducted on 3 µm-thick representative paraffin sections. The DAKO EnVision+ system (dextran polymers conjugated with horseradish peroxidase; DAKO, Copenhagen, Denmark) was used to examine immunohistochemical expression of these markers, as previously described .
Evaluation of cancer cell phenotypes
Cancer cell phenotypes were subclassified into four groups: gastric type (positive for Muc5AC and/or Muc6, and negative for Muc2 and CD10), intestinal type (negative for Muc5AC and Muc6, and positive for Muc2 and/or CD10), mixed type (positive for Muc5AC and/or Muc6, and also positive for Muc2 and/or CD10) and unclassified type (negative for all markers), according to a previous report .
Assessment of immunohistochemical expression
To standardize evaluations, we used the following criteria to analyze immunohistochemical staining of mucin markers (MIUC2, MUC5AC, and MUC6), CD10, β-catenin, CDX2, and p53. The staining intensity scores were divided into four categories: no staining, weak/equivocal staining, moderate staining, and strong staining. Moderate or strong staining was considered to be positive expression. The percentage of cells with positive expression was scored as follows: 0: 0–10% cells; 1: 10% to < 30% cells; 2: 30% to < 60% cells; 3: 60% to < 100% cells; and 4: 100% cells. In this study, a score greater than 1 indicated positive expression of the markers in the lesions. Finally MLH-1 expression in <5% of the tumor cell population was defined as loss of MLH-1 expression.
Tissue dissection and DNA extraction
DNA was extracted from manually micro-dissected paraffin-embedded tissue sliced into 10-μm thick sections (Figure 1A and B) in which >60% of cells were identified as tumor cells using TaKaRa DEXPAT (TAKARA Bio Inc., Japan) according to the manufacturer’s instructions.
TP53 gene and direct sequence
Single strand conformation polymorphism (SSCP) analysis was performed as previously described, with some modifications . Briefly, the PCR products (2 μl) were mixed with 10 μl gel loading solution (9.5% deionized formamide, 20 mM EDTA-Na, 0.05% xylene cyanol and bromophenol blue), denatured at 95°C for 5 minutes and then kept on ice until loading. Non-denaturing 7.5% polyacrylamide gels were used for electrophoresis, which was carried out at 260 to 300 V for 3 to 12 hours at 22°C using a temperature controller (Resolmax, ATTO Co., Tokyo). The gels were visualized by silver staining and photographed.
Sequencing was performed twice on original PCR products of TP53 exons 5-8 for all SSCP-positive samples to confirm the TP53 mutation status of the samples using a direct sequence method. The results of the first and second sequencing runs were identical. PCR products were recovered from 3% agarose gels and the eluted DNA fragment was precipitated with ethanol before direct sequencing. Sequence primers were the same as those used for PCR. Direct sequencing was performed using fluorescent-labeled dideoxynucleotide triphosphates for automated DNA sequence analysis (Applied Biosystems 373A sequencer; Applied Biosystems, USA, CA).
KRAS and BRAF genes
PCR-pyrosequencing using a PyroMark Q24 instrument (Qiagen, Venlo, the Netherlands) was performed for KRAS (exon 2) and BRAF (exon 15; codon 600) using a previously reported method . Briefly, the polymerase chain reaction (PCR) product (25 μL) was bound to streptavidin Sepharose HP (GE Healthcare, Brøndby, Denmark), purified, washed, denatured in 0.2 M NaOH and washed again. Before pyrosequencing, 0.3 μM sequencing primer was annealed to the purified single-stranded PCR product by heating to 80 °C for 2 min.
Analysis of Allelic Imbalance (AI)
AI analysis was performed using a PCR-microsatellite assay (GeneAmp PCR System 9600; Perkin-Elmer, CA, USA) according to previously reported procedures [10-11]. AIs on chromosomes 1p, 3p, 4p, 5q, 8p, 9p, 13q, 17p (TP53), 18q and 22q were examined in paired cancer and normal DNA samples using 22 highly pleomorphic microsatellite markers (D1S228, D1S548, D3S1234, D3S2402, D4S1601, D4S2639, D5S582, D5S107, D5S299, D8S201, D8S513, D8S532, D9S171, D9S1118, D13S162, TP53, D18S34, D18S487, D22S274, D22S1140, D22S1168). These markers have frequently been used in studies of GCs [12-13]. In addition, a variable number of tandem repeat polymorphisms at the DCC locus were tested.
PCR reactions were performed using a thermal cycler (GeneAmp PCR System 9600, Perkin-Elmer, CA, USA) with 50-100 ng genomic DNA as a template, 25 pM of each primer, 0.2 mM deoxynucleotide triphosphate (dNTP), 1x reaction buffer containing 1.5 mM MgCl2, and 1.5 U Taq polymerase (Boehringer Mannheim Co., Germany) in a final reaction volume of 25 μl. Samples were processed for 25 to 30 cycles, with each cycle consisting of 30 s at 94°C, 1 min at 55 to 58°C, and 2 min at 72°C, followed by a final 10 min extension at 72°C. For quantitative detection of the allelic loss at each locus, PCR-LOH (loss of heterozygosity) analysis was performed as described previously. A 1μl aliquot of the PCR product was added to 3 μl formamide and 0.5 μl of TAMRA 500 size standard (Applied Biosystems, CA) and was loaded onto a 6% polyacrylamide-8 M urea gel, and run for 2-6 hours in a 373A Automated Sequencer (Applied Biosystems, CA, USA) at a constant power of 30 W.
Peaks generated from normal DNA samples were used to determine homozygous (1 peak) or heterozygous (2 peaks). The allelic ratio was calculated as previously described . A cancer was considered to have AI when the allele peak ratio was <0.60, representing an allelic signal reduction of at least 40%. MSI at a given locus was not evaluated. The data were collected and analyzed using GeneMapper software v. 4.0 (Applied Biosystems, CA, USA).
Analysis of Microsatellite Instability (MSI)
The PCR-based assay for evaluation of MSI was described previously [10-11]. Two adenine mononucleotide repeats (BAT25 and BAT26) and three dinucleotide repeats (D2S123, D5S346 and D17S250) were used to determine the presence of tumor MSI . Tumors were considered positive for MSI when abnormally-sized peaks in the tumor sample relative to the paired normal sample were detected for at least two of the five markers.
DNA Methylation Analysis
The DNA methylation status was examined by PCR analysis of bisulfite-modified genomic DNA (EpiTect Bisulfite Kit; Qiagen) using pyrosequencing for quantitative methylation analysis (PyroMark Q24; Qiagen NV). The primers were designed using the PyroMark Assay Design Software package (Qiagen NV). We quantified DNA methylation in 6 specific promoters described by Yagi et al . High methylation epigenotype (HME) tumors were defined as those having at least 2 methylated markers in the first marker panel (RUNX3, MINT31 and LOX). The remaining tumors were screened using a second marker panel (NEUROG1, ELMO1 and THBD); intermediate methylation epigenotype (IME) tumors were defined as those having at least 2 methylated markers. The other tumors were designated as having a low methylation epigenotype (LME). Methylation of MLH-1 was also quantified. The cut-off value was 30% according to a previously described method using six specific promoters .
Statistical analyses were performed using the statistical computing software R version 3.3.2 (R Foundation for Statistical Computing, Vienna, Austria). To determine significant differences in age and tumor size, the Wilcoxon signed-rank test of variance was used. The Fisher’s exact test was used to compare other clinicopathological factors, immunohistochemical studies and molecular analysis results. Multiple-comparison analysis was carried out using the Bonferroni correction. P values <0.05 were considered to indicate statistical significance.