Differential gene-expression profiles associated with gastric adenoma

Gastric adenomas are considered to be precancerous lesions, but are clinically heterogeneous, since some may progress to adenocarcinoma, whereas others persist unchanged for long periods (Kamiya et al, 1982; Kolodziejczyk et al, 1994; Orlowska et al, 1995). Identification of adenoma cases with a progressive nature is important since intervention (e.g. endoscopic mucosal resection) is mandatory. Tumour size is a prognostic indicator, but exceptional cases are frequently observed. Histological grading of adenomas as per the Vienna classification (Schlemper et al, 2000) has been used to assess the potential for progression. However, exceptional cases are frequent, since 80% of high-grade adenomas progress to adenocarcinomas, whereas 15% of low-grade adenomas progress to high-grade adenomas or adenocarcinomas (Lauwers and Riddell, 1999). Histological diagnosis of biopsy specimens cannot definitively identify adenomas with aggressive potential because sampling errors may contribute to the underestimation of tumour grade or depth of invasion. Thus, an additional prognostic indicator that is independent of conventional clinicopathological findings (e.g. molecular markers) is essential. Recent comprehensive analyses of gene expression, such as a microarray analysis, identified relevant genes whose expression profiles appeared to be linked to tumour stage, histological grade, susceptibility to chemotherapy, clinical aggressiveness or prognosis (Golub et al, 1999; Alizadeh et al, 2000; Perou et al, 2000; Dhanasekaran et al, 2001; Sorlie et al, 2001; Shipp et al, 2002; van’t Veer et al, 2002). Studies on gastric adenocarcinomas revealed several gene-expression profiles that are linked to lymph node metastasis (Hasegawa et al, 2002; Hippo et al, 2002). Using a similar approach, it may be possible to develop an improved classification scheme for gastric tumours that is capable of distinguishing subgroups of adenomas with progressive natures. Such expression profiles have not been applied to gastric adenomas. In the present study, suppression subtractive hybridization (SSH) analysis (Diatchenko et al, 1996; von Stein et al, 1997) was used to identify genes relevant to gastric adenomas. Their expression profiles were subsequently assessed in order to identify different progressive potentials of gastric adenomas in comparison to adenocarcinoma.