476 AUTHORS' REPLY The alterations of four genes (APC, K-ras, p53, and DCC) which are characteristic of the colorectal adenoma-carcinoma sequence were studied in gastric adenomas and adenocarcinomas. In gastric carcinomas, the rarity of K-ras mutations and the frequency of pS3 mutations and DCC loss of heterozygosity (LOH) have also been described by other investigators with incidences similar to ours.24 Frequent mutations of APC are reportedly restricted to certain histological types: very well differentiated adenocarcinoma and signet-ring cell c a r ~ i n o m a Well . ~ to moderately differentiated adenocarcinomas, which constitute the great majority of differentiated adenocarcinomas of the stomach, did not carry APC m u t a t i o i i ~ . ~The .~ difference in the incidences of APC mutations in adenomas and adenocarcinomas of the stomach reported by Tahara and by ourselves might indeed have resulted from differences in the histological criteria used, as they suggest. Methodological underestimation, however, is unlikely because we were able to detect frequent APC mutations in colorectal carcinomas by the same SSCP procedure.' Furthermore, other lines of evidence have suggested that APC is not the true target of Sq-LOH in other than colorectal carcinoma^.^.^ In gastric adenoma, APC mutations occurred at an incidence of 20 per cent (6/30), but alterations of K-ras, pS3, and DCC were absent or rare.' In addition, our recent microsatellite and allelotype analyses have revealed no evidence for the adenoma-carcinoma sequence in gastric carcinogenesis, because the incidence of LOH did not exceed 10 per cent at any of the 39 microsatellite loci covering each autosomal chromosome arm, including those at which differentiated adenocarcinomas exhibited frequent LOH.lO." Tohdo et a1.I2 reported that 30 per cent (3/10) of gastric adenomas carried pS3 mutations. However, two of the three were silent mutations which did not contribute to carcinogenesis. If gastric adenoma progresses to carcinoma, why do few silent mutations occur in gastric carcinoma? A case of well differentiated adenocarcinoma arising from tubular adenoma is cited to emphasize the existence of the adenoma-carcinoma sequence in the stomach. We have experience of the co-existence of a.denoma (, 1996 by John Wiley & Sons, Ltd. and adenocarcinoma in the stomach, but in our experience these tumours usually develop apart from each other. In addition the so-called 'carcinoma in adenoma', which is very common in the colon, is quite rare in the stomach. For these reasons, we reaffirm our conclusion that the adenoma-carcinoma sequence is not the major pathway of carcinogenesis in the stomach. G. TAMURA, C. MAESAWA AND R. SATODATE Department of' Pathology Iwate Medical University School of Medicine Uchimuru 19-1, 020 Morioka Japan REFERENCES 1. Maesawa C, Tamura G, Suzuki Y, ef a[. The sequential accumulation of 2. 3. 4. 5. 6. 7. X. 9. 10. 11. 12. genetic alterations characteristic of the colorectal adenomaxarcinoma sequence does not occui- between gastric adenoma and adenocarcinoma. J Pulhol 1995; 176: 249-258. Koshiba M, Ogawa 0, Habuchi T, ef a/. Infrequent ras mutation in human stomach cancers. Jpn J Cuncer Res 1993; 8 4 163-167. Uchino S, Noguchi M, Ochiai A, Saito T, Kobayashi M, Hirohashi S. p53 mutation in gastric cancer: a genetic model for carcinogenesis is common to gastric and colorectal cancer. Int J Cuncer 1993; 5 4 759-764. Uchino S, Tsuda H, Noguchi M, eta/. Frequent loss of heterozygosity at the DCC locus in gastric cancer. C a n ~ Res r 1992, 52: 3099-3102. Nakatsuru S, Yanagisawa A, Ichii S, C I a/. Somatic mutation of the AI'C gene in gastric cancer: frequent mutations in very well diffcrcntiated adenocarcinoma and signet-ring cell carcinoma. Hum M u / Genet 1992; 1: 559-563. Horii A, Nakatsuru S, Miyoshi Y, et a/. The APC gene, responsible for familial adenomatous polyposis, is mutated in human gastric cancer Cuncrr Re.( 1992: 52: 3231-3233. Ogasawara S, Maesawa C, Tamura G, Satodate R . Lack of mutations of the adenomatons polyposis coli gene in oesophageal and gastric carcinomas. I'irthoiw Arch 1994; 424: 607-61 1. Ogasawara S, Tamura G, Maesawa C, ef al. Common delcted region on the long arm of chromosome 5 in esophageal carcinoma. Gu.sfroenterology(in press). Tamura G, Ogasawara S, Nishiruka S, rt u/. Two distinct regions of deletion on the long arm of chromosome 5 in dilkrentiated adenocarcinomiis of the stomach. Cancer Re.r (in press). Tamura G, Sakata K, Maesawa C, ef ul. Microsatellite altcrations in ndenoma and differentiated adenocarcinoma of the stomach. S a n w r Rt,s 1995; 5 5 1933-1936. Tamura G, Sakata K, Nishiznka S, et u/. Allelotype of adenoma and differentiated adenocarcinoma of the stomach. (Submitted for publication.) Tohdo H, Yokozaki H, Haruma K , Kajiyama G, Tahara E. p53 gene mutations in gastric adenomas. Virchows Arch B [ C d Pu/hol/ 1993; 63: 191-195.