BACKGROUND. The lymph node (N) classification in the International Union Ag
ainst Cancer (UICC) TNM staging system for gastric adenocarcinoma has been
revised. The new classification is based on the number of positive regional
lymph nodes instead of the anatomic location of the regional lymph node me
tastasis. Both classification systems were compared for prognostic signific
ance.
METHODS. A total of 4362 gastric carcinoma patients who underwent resection
between 1969 and 1990 were analyzed.
RESULTS. Thirteen percent of patients could not be staged according to the
new system. Based on the previous classification, 647 patients were classif
ied as pN1 and 711 patients as pN2. When reclassified, 587 patients remaine
d pN1, 54 patients became pN2, and 6 patients became pN3. Of the 711 pN2 pa
tients, 333 became pN1, 267 remained pN2, and 111 patients became pN3. Both
lymph node classification methods defined groups with widely differing pro
gnoses. The prognoses of patients classified as new pN2 were more homogeneo
us than those of the group classified as old pN2. Survival of new pT4/pN1 p
atients was significantly better than that of other subgroups in Stage IV.
Nine potential prognostic factors, including lymph node metastasis, were st
udied in multivariate analysis. The hazard ratios were 1.38 (1.16-1.64) for
pN1 and 2.55 (2.16-3.01) for pN2, based on the old classification. They we
re 1.51 (1.29-78) for pN1, 3.11 (2.56-3.78) for pN2, and 3.88 (2.98-5.05) f
or pN3, based on the new classification.
CONCLUSIONS. The new N classification is superior as a prognostic factor to
the old N classification, although there is inadequacy in stage grouping.
[See editorial on pages 1763-5, this issue.] Cancer 2000;88:1796-800. (C) 2
000 American Cancer Society.