I n patients with gastric carcinoma, anatomic extent is the strongest predictor of outcome. Following resection for cure (residual tumor classification: R0, no residual tumor), local and lymphatic spread are the most important prognostic factors.
The classification of the local extent (T classification) of gastric carcinoma has been identical in the International Union Against Cancer/American Joint Committee on Cancer (UICC/AJCC) system for more than 20 years, and is also identical in the Japanese classification. In contrast, the classification of regional lymph node (RLN) spread (N classification) is different and presents problems.
Since the first edition of TNM, the classification of regional lymph nodes for gastric carcinoma was based on the localization of involved RLN. The UICC/AJCC classification 1,2 provided relatively simply defined categories, namely, involvement of perigastric lymph nodes near the primary tumor or of more distant lymph nodes, including those at the trunks of major arteries. In contrast, in Japan a complex and sophisticated system was developed. 3 Thus, two different N classifications were used, also resulting in different stage groupings. Consequently, international comparisons of treatment results with consideration of different anatomic extents of disease were not possible.
Since the late 1970s, the prognostic significance of the number of lymph nodes involved has been investigated (for an overview, see Roder et al. 4 ). These studies, however, showed marked variation in the cutoff points. In 1993, the UICC 5 published a proposal for testing a subdivision of lymph node positive cases (N1, N2) according to the number of lymph nodes involved into 3 subcategories: a, 1-3 involved; b, 4 -6 involved; c, more than 6 involved. The testing of this "telescopic ramification" in Germany led to the proposal of a number-based, instead of localization-based, classification of lymphatic spread. It was presented for the first time at the Joint Panel Discussion "Towards common language for research and treatment of gastric cancer," held on 1995, 29, March prior to the 1st International Cancer Congress in Kyoto, Japan. 6 It seemed a way to achieve a uniform classification system acceptable to Western countries as well as Japan. At this time, it was not clear which cutoff points should be adopted,