• OBJECTIVES To determine in a prospective randomized clinical trial, the effect of extended lymph node dissection (R3 gastrectomy) versus conventional surgery (R0/1 gastrectomy) on tumor recurrence and survival in Chinese patients with advanced cancer of the stomach (adenocarcinoma invasion beyond submucosa).
• STATISTICAL CONSIDERATIONS The major end points are disease free survival and survival. The log-rank test will be used as analytic method on disease-free survival and overall durvival. The follow-up time after treatment is five years.

Previous experience provides an estimation of a 20% 5-year survival rate with conventional surgery for gastric cancer. Assuming that this new treatment (radical surgery) may increase the 5-year survival rate to 40%, we estimate that 118 evaluable patients are required to detected a significance at a of 0.05 level and power of 0.8 using a one tailed test.

We expect to enter 50 patients per year and finish accrual of patient within 2.5 years.

* TREATMENT PLAN

Patients are randomized intraoperatively into arm I (conventional [R0/R1] gastrectomy) or arm II (extended radical [R3] lymph node dissection).

Arm I: Resection of stomach, greater omentum, lesser omentum, and adjacent organ(s) when the primary tumor invaded directly. A total or distal subtotal gastrectomy is decided by the proximal distance from cardia to the tumor. A margin of 3 cm was required for well-defined (Borrmann I or II) tumors, and of 5 cm for ill-defined (Borrmann III or IV) tumors. When a total gastrectomy is performed, the distal pancreas and spleen should be preserved unless there is direct invasion into these organs. The left gastric artery should not be ligated at its origin but distal to its bifurcation into ascending and descending branches.

Arm II: The procedure are the same as in Arm I, additionally, extended lymph node dissection includes dissection of the N1, N2 and N3 nodes. These groups (N) are comprised of lymph node stations which depended on the location of the primary tumors (Fig. 1).

1. For N2 dissection, the left gastric artery is ligated at its origin to facilitate the dissection of No.7. If a total gastrectomy is indicated, a splenectomy is essential for the dissection of station No. 10, and a distal pancreatectomy for the dissection of station No.11. Combined resection of these adjacent organs is not performed during a distal subtotal gastrectomy.
2. If the primary cancer is located at the middle or lower third of stomach, No.1 (nodes at right cardiac) and 11 (nodes along the splenic artery) are excised routinely for frozen section before gastric resection. If no cancer metastasis is seen in both No.1 and 11 lymph nodes is performed. If only No. 1 lymph nodes is positive for cancer metastasis, total gastrectomy is performed instead of distal subtotal resection. If No. 11 lymph nodes are positive, splenectomy and distal pancreatectomy are performed in addition to the radical resection of the stomach and lymph nodes.
3. For N3 dissection: No. 12 (nodes in the hepatoduodenal ligament), No. 13 (nodes at retropancreatic region) and No. 14 (nodes at the root of mesentery) are dissected.

For patients who had gastric cancer recurrence after gastrectomy, chemotherapy with cisplatin、 5-FU、Leucovorin will be given.

Cisplatin 20mg/m2 5-FU 450 mg/m2 Leucovorin 90 mg/m2

The above three drugs to be given in 500ml of normal saline and infused intravenously simultaneously over 96 hours and repeated every 3 weeks. (21days)

Dose modification： Delay treatment for one week (if WBC<4000/dl immediately before treatment). Reduce dose of subsequent cycles of 5-FU by 20% if WBC nadir <1000/dl.

Delay treatment for one week (if there is oral mucositis or diarrhea immediately before treatment). Reduce dose of subsequent cycles of 5-FU by 10% for grade 2 and 20% for grade 3 or 4 mucositis or diarrhea.

Stop treatment for any grade 4 non-hematological toxicity. Stop cisplatin treatment if serum creatinine > 2mg/dl.

Duration of treatment： Continue treatment for patients until disease progression.

* DATA AND PROTOCOL MANAGEMENT 10.1 Registration, randomization will be performed by the Cancer Clinical Trial Operations Office by calling 789-9046. (47)& 7852459.

Quality control of the data will be performed by the Cancer Clinical Trial Operations Office, Institute of Biomedical Sciences, Academia Sinica.

Data will be evaluated by the statistician and the study chairman before data analyzed

10.2 Protocol Compliance: the attending physician and oncology research nurse must see each patient prior to treatment and at follow up to ensure all investigations, treatment procedures and data records are done and recorded according to protocol.

• Inclusion Criteria:

  1. Patients with histologically proven adenocarcinoma of the stomach.
  2. No clinical early stage stomach cancer (mucosa or submucosal invasion only).
  3. No esophageal invasion.
  4. No distant metastatic disease.
  5. No para-aortic, macrolymphadenopathy
  6. No N3, macrolymphadenopathy
  7. Suitable for gastric surgery with curative intent.
  8. Medically fit for surgery.
  9. No emergency resection required.
  10. No previous gastrectomy.
  11. Not older than 75 years.
  12. Signed informed consent form.
  13. No previous or concomitant other cancer.

• Exclusion Criteria:

  1. Patients is not adenocarcinoma of the stomach.
  2. distant metastatic disease.
  3. previous gastrectomy.
  4. older than 75 years.