Early gastric cancer.

Although common in Japan, early gastric cancer (EGC = gastric adenocarcinoma confined to the mucosa and submucosa of the stomach, with or without regional lymph node metastases) is thought to be an infrequent occurrence in the United States. However, a review of all "curative" resections for carcinoma of the gastric body and antrum at the University of Virginia between 1974 and 1982 revealed EGC in five of 31 patients (16%). The purpose of the present study was to compare EGC to more advanced gastric cancer (ADV; n = 26) to determine whether any presenting historical, laboratory, x-ray, or endoscopic features distinguished the two groups before surgery and to ascertain whether postoperative survival in the United States mimicked the Japanese experience. All surviving patients were contacted, all charts were abstracted, all pathologic specimens were reexamined, and all radiographs were reviewed blindly by an experienced radiologist. Statistical evaluation was accomplished using Kaplan-Meier plots, chi square analysis, and unpaired "t" tests, as appropriate. At presentation, patients with EGC were younger (44 +/- 6 vs. 67 +/- 2 years, p less than 0.01) with higher admission albumin levels (4.1 +/- 0.2 vs. 3.7 +/- 0.1 mgm/dl, p less than 0.01). Although not significantly different, admission hemoglobin tended to be higher (41 +/- 2 vs. 35 +/- 2%), the incidence of weight loss tended to be less (40 vs. 65%), duration of symptoms tended to be longer (21 +/- 11 vs. 8 +/- 3 months), and tumor diameter tended to be smaller (1.7 +/- 0.6 vs. 5.8 +/- 0.7 cm) in EGC. No differences were apparent with respect to endoscopic or radiographic appearance, tumor location (greater than 70% antrum), presence of regional lymph node metastases (EGC = 2/5; ADV = 20/26), or type of resection (subtotal gastrectomy in 4/5 EGC, in 19/26 ADV). On median 5-year follow-up, however, survival with EGC has been 100%. In contrast, the Kaplan-Meier estimate of 5-year survival in ADV is 15% (42% with muscularis invasion, 0% with serosal invasion, 12% with extra-gastric spread; p less than 0.01 vs. EGC). One suture line recurrence in EGC was successfully treated by re-resection. No ADV patient with recurrence survives (p less than 0.01). Thus, EGC behaves similarly in the United States and Japan; for example, prognosis is excellent even in the presence of lymph node metastases. Inability to distinguish EGC from ADV before surgery justifies an aggressive surgical approach to all patients with resectable gastric neoplasms.