Section 3 -

Early Gastric Cancer Geraint T Williams Pathology Department Wales College of Medicine, Cardiff University

Early gastric cancer (EGC) is defined as carcinoma that is confined to the mucosa and submucosa of the stomach, [1] irrespective of the presence of lymph node metastasis. This definition was conceived by Japanese gastroenterologists in the middle of the last century to identify cases of gastric cancer that were potentially curable – it does not necessarily imply that they all EGCs are 'early' in a chronological sense. As with all malignant tumours, it is now appreciated that there is a great biological variation in growth and progression, and in the stomach there is evidence to suggest that up to a third of EGCs remain at an early stage five years after diagnosis. [2] Increasing use of endoscopy has meant that more than 50% of gastric cancers in Japan, and up to 20% in the West, are diagnosed as EGCs. Radical surgery for such tumours is nearly always curative, but has significant morbidity and mortality, especially in the older population, and this has led to the development in the Far East of less radical but equally effective alternatives [3] that are now becoming increasingly used in the West also. Foremost of these is endoscopic mucosal resection (EMR), which has been demonstrated to be safe and to lead to an excellent outcome, provided that there is careful patient selection and painstaking histopathological evaluation of the resection specimen to identify cases requiring more radical treatment. This presentation will concentrate of histopathological aspects of the diagnosis of EGC, the selection of patients for EMR, the assessment of EMR specimens, and the surveillance of the stomach following an EMR procedure. From a theoretical perspective, the pre-treatment diagnosis of EGC requires its distinction from precancerous lesions (gastric dysplasia) and non-neoplastic lesions on the one hand and carcinoma invading the muscularis propria on the other. Microscopy has a major part to play in the former, but relatively little role in the latter because of the relatively superficial nature of endoscopic biopsies. It is the endoscopic appearances of the lesion itself – especially its size, and whether it is raised, flat, depressed or ulcerated, supplemented by the use of endoscopic ultrasound, that provides the best information here. The histological diagnosis of gastric dysplasia, and its distinction from adenocarcinoma are topics that are well recognised as challenging and controversial. International differences in interpretation and nomenclature between Eastern and Western pathologists have complicated the issue further, [4] and implementation of the Vienna classification, an attempt to resolve this, does not appear to have had the impact that was hoped. [5] The issue is further complicated by the fact that heterogeneity within lesions means that those with histological changes of low or high grade dysplasia in one area may have overt invasive carcinoma in another – a problem that can be partly resolved by multiple sampling. [6]In practice, the unequivocal demonstration of malignancy in what appears to be a superficial neoplastic lesion is of limited practical importance, because most centres would now manage high-grade dysplastic lesions in the same way as EGC and undertake EMR for well demarcated low-grade dysplastic lesions. The distinction between truly neoplastic conditions and reactive non-neoplastic mimics is much more important, [7] and a significant cause of litigation in histopathology. Constant vigilance by the pathologist for the potential for misdiagnosis and careful consideration of the clinical and endoscopic features of the case are the most important guardians from disaster. Immunohisto-chemical studies have relatively little role in the interpretation of atypical lesions in diagnostic practice, apart from possibly the finding of strong diffuse nuclear positivity for p53 (but no less staining), which is supportive evidence of a neoplastic lesion. However, the converse is not true: a significant proportion of neoplastic lesions are p53-negative with conventional technology. Once a histological diagnosis has let to a decision that extirpation of a lesion is necessary, the selection of patients for EMR or more radical surgery is based on the endoscopic and endoscopic ultrasound features and any co-morbidity in the patient. Criteria indicating suitability for EMR are based on studies of factors that predict lymph node metastasis in EGC [8] and vary a little between centres, but the ultimate decision is based on the judgement of the endoscopist in each individual case. It is generally agreed that ulcerating lesions are unsuitable for EMR, and most endoscopists are reluctant to use the procedure for poorly-differentiated and signet ring-cell type cancers. In general, lesions that are amenable to EMR are those that are considered to be intramucosal, are not ulcerated, are well demarcated endoscopically, measure less than 3cm in diameter, are well or moderately differentiated on biopsy, and have no lymph node metastases on endoscopic ultrasound. Using these criteria, less than 0.5% would be expected to have lymph node metastasis. In a large series of 479 lesions selected in this way at the National Cancer Centre in Tokyo, unpredicted submucosal invasion was found on histology in 15%. No deaths occurred and local recurrence developed in 5%. [9] Another review of 1732 cases from 12 Japanese centres found only a single death from gastric cancer on follow-up. [3] These studies concluded that the histopathological indications for further surgery were incomplete excision, submucosal invasion, and lymphovascular invasion. Accordingly, the pathologist who is receipt of an EMR specimen must deal with it meticulously in order to report these three features as unambiguously as possible. This requires careful co-operation between the endoscopist, his assistant, and the pathologist. If it is feasible to resect the lesion intact in its entirety, then the specimen should be carefully pinned out on a cork or wax block, the resection margins carefully marked with indelible markers, and the whole of the specimen embedded for histological examination. [10] If resection has to be undertaken piecemeal, then a reasonably reliable assessment of true excision margins can be made if the pathologist is informed of the exact location of each of the resected pieces. The gradual introduction of EMR into Western centres has demonstrated that lessons learned in the Far East apply equally in the West. Careful assessment of EMR specimens indicates that the preoperative biopsies "under-assess" the degree of neoplasia in about one third of cases, and margin involvement is present in a similar number. [10] Local recurrence occurs in up to 50% of cases with margin involvement, and a clearance of 2mm in the fixed specimen is required to be confident of complete excision. [11] Depending on co-morbidity in the patient, incompletely excised intramucosal lesions can be treated by further endoscopic resection, while those with submucosal or lymphovascular invasion can be offered more radical treatment. Patients who have undergone EMR frequently have further endoscopy within a few days or weeks of EMR, to confirm completeness of excision or to survey any residual neoplasia. The histological interpretation of such biopsies taken can be particularly challenging because of reparative changes following the resection. [12] Many of these changes recapitulate the 'atypical' appearances that are recognised to occur in any regenerating or ulcerating lesion, but lacy architectural change, clear cell degeneration and signet ring cell-like change, can be particularly problematical. [12] The lack of a desmoplastic stromal reaction can be particularly helpful in making the correct diagnosis in these scenarios. Although the great majority of early gastric cancers have an excellent prognosis, there is a minority that do not. Studies in the early 1980s [13]showed that these tumours had certain histological characteristics, namely a raised configuration, an intestinal or differentiated phenotype with an expansive or "penetrating" growth pattern in the submucosa, frequent aneuploidy, and a high frequency of lymphovascular invasion. 25% of such EGCs had liver metastases and the 10-year survival was 65% compared with >90% for the remainder. It may be that these tumours represent a particularly aggressive form of truly "early" gastric neoplasia with more rapid progression to advanced gastric cancer than the more conventional forms of EGC. With the development of more promising chemotherapeutic approaches to gastric carcinoma, consideration of adjuvant therapy may be appropriate for these uncommon aggressive tumours. References Hamilton SR, Aaltonen LA. Pathology and Genetics of Tumours of the Digestive System. World Health Organization Classification of Tumours. Vol 2. IARC Press, Lyon, 2000. Tsukuma H, Oshima A, Narahara H, Morii T. Natural history of early gastric cancer: a non-concurrent, long term, follow up study. Gut 2000; 47: 618-21. Kojima T, Parra-Blanco A, Takahashi H et al. Outcome of endoscopic mucosal resection in early gastric cancer: review of Japanese literature. Gastrointest Endosc 1998; 48: 550-4. Schlemper R J, Itabashi M, Kato Y et al Differences in diagnostic criteria for gastric carcinoma between Japanese and Western pathologists. Lancet 1997; 349: 1725–9. Schlemper RJ, Riddell RH, Kato Y et al The Vienna classification of gastrointestinal epithelial neoplasia. Gut 2000; 47: 251-5. Hull MJ, Mino-Kenudson M, Nishioka NS, et al. Endoscopic mucosal resection: an improved diagnostic procedure for early gastroesophageal epithelial neoplasms. Am J Surg Pathol 2006; 30: 114-8. Isaacson P. Biopsy appearances easily mistaken for malignancy in gastrointestinal endoscopy. Histopathology 1982; 6: 377–89. Kim YK, Joo JK, Ryu SY et al Factors related to lymph node metastasis and surgical strategy used to treat early gastric cancer. World J Gastroenterol 2004; 10: 737-40. Ono H, Kondo H, Gotoda T et al. Endoscopic mucosal resection for treatment of early gastric cancer. Gut 2001; 48: 225-9. Lauwers GY, Ban S, Mino M, et al. Endoscopic mucosal resection for gastric epithelial neoplasms: a study of 39 cases with emphasis on the evaluation of specimens and recommendations for optimal pathologic analysis. Mod Pathol 2004; 17: 2-8. Hamada T, Kondo K, Itagaki Y et al Endoscopic mucosal resection for early gastric cancer. Nippon Rinsho 1996; 54: 1292-7. Mitsuhashi T, Lauwers GY, Ban S et al. Post-gastric endoscopic mucosal resection surveillance biopsies: evaluation of mucosal changes and recognition of potential mimics of residual adenocarcinoma. Am J Surg Pathol 2006; 30: 650-6. Kodama Y, Inokuchi K, Soejima K et al Growth patterns and prognosis in early gastric carcinoma: superficially spreading and penetrating growth types. Cancer 1983; 51: 320–6.