Classification of Microsatellite Alterations Detected in Endoscopic Biopsy Specimens of Gastric Cancers

ABSTRACT

PURPOSE: Individual gastric cancers demonstrate complicated genetic alterations. The PCR-based analysis of polymorphic microsatellite sequences on cancer-related chromosomes has been used to detect chromosomal loss and microsatellite instability. For the purpose of preoperative usage, we analyzed the correspondance rate of the microsatellite genotype between endoscopic biopsy and surgical specimens. MATERIALS AND METHODS: Seventy-three pairs of biopsy and surgical specimens were examined for loss of heterozygosity and microsatellite instability by using 40 microsatellite markers on eight chromosomes. Microsatellite alterations in tumor DNAs were classified into a high-risk group (baseline- level loss of heterozygosity 1 chromosomal loss in diffuse type and high-level loss of heterozygosity 4 or more chromosomal losses) and a low-risk group (microsatellite instability and low-level loss of heterozygosity 2 or 3 chromosomal losses in diffuse type or 1~3 chromosomal losses in intestinal type) based on the extent of chromosomal loss and microsatellite instability. RESULTS: The chromosomal losses of the biopsy and the surgical specimens were found to be different in 21 of the 73 cases, 19 cases of which were categorized into a genotype group of similar extent. In 100 surgical specimens, the high-risk genotype group showed a high incidence of nodal involvement (19 of 23 cases 5 cm) irrespective of tumor size while the incidence of nodal involvement for the low-risk genotype group depended on tumor size (5 of 26 cases 5 cm). Extraserosal invasion was more frequent in large-sized tumor in both the high-risk genotype group (5 cm 23 of 24 cases) and the low-risk genotype group (5 cm 16 of 27 cases). The preoperative prediction of tumor invasion and nodal involvement based on tumor size and genotype corresponded closely to the pathologic tumor stage (ROC area>0.7). CONCLUSION: An endoscopic biopsy specimen of gastric cancer can be used to make a preoperative genetic diagnosis that accurately reflect the genotype of the corresponding surgical specimen.