Gastric Cancer Surgical Practice Guidelines

July 1, 1997

The Society of Surgical Oncology surgical practice guidelines focus on the signs and symptoms of primary cancer, timely evaluation of the symptomatic patient, appropriate preoperative extent of disease evaluation, and role of the surgeon in

The Society of Surgical Oncology surgical practice guidelines focuson the signs and symptoms of primary cancer, timely evaluation of the symptomaticpatient, appropriate preoperative extent of disease evaluation, and roleof the surgeon in diagnosis and treatment. Separate sections on adjuvanttherapy, follow-up programs, or management of recurrent cancer have beenintentionally omitted. Where appropriate, perioperative adjuvant combined-modalitytherapy is discussed under surgical management. Each guideline is presentedin minimal outline form as a delineation of therapeutic options.

Since the development of treatment protocols was not the specific aimof the Society, the extensive development cycle necessary to produce evidence-basedpractice guidelines did not apply. We used the broad clinical experienceresiding in the membership of the Society, under the direction of AlfredM. Cohen, MD, Chief, Colorectal Service, Memorial Sloan-Kettering CancerCenter, to produce guidelines that were not likely to result in significantcontroversy.

Following each guideline is a brief narrative highlighting and expandingon selected sections of the guideline document, with a few relevant references.The current staging system for the site and approximate 5-year survivaldata are also included.

The Society does not suggest that these guidelines replace good medicaljudgment. That always comes first. We do believe that the family physician,as well as the health maintenance organization director, will appreciatethe provision of these guidelines as a reference for better patient care.

Societyof Surgical Oncology Practice Guidelines: Gastric Cancer

Symptoms and Signs

Early-stage disease

  • No symptoms or signs
  • Megaloblastic anemia
  • Dyspepsia or mild epigastric or substernal pain
  • Early satiety

Advanced-stage disease

  • Upper gastrointestinal bleeding
  • Anemia, especially microcytic if chronic
  • Frank upper gastrointestinal bleeding with hematemesis
  • Pain: epigastric "ulcer"-type pain, pain that bores to theback, substernal pain, right or left upper quadrant pain (occasional)
  • Obstruction: dysphagia secondary to gastroesophageal junction tumor,difficulty with digesting solids before liquids, nausea and vomiting withreturn of undigested food, secondary to gastric outlet obstruction
  • Weight loss, malaise, fever
  • Ascites, abdominal or pelvic masses

Evaluation of the Symptomatic Patient

Upper gastrointestinal barium study

  • Appropriate initial assessment

Endoscopy with biopsy

  • Required for diagnosis since a gastric neoplasm may be an adenocarcinoma,a lymphoma, or, rarely, a gastrointestinal stromal tumor (GIST)

Digital rectal examination

  • Extremely important not only for the evaluation of blood in the stoolbut also to rule out the presence of a Blummer's shelf.

Thorough physical examination

  • Evaluate supraclavicular nodes to rule out obvious clinical metastasis.

Appropriate timeliness of surgical referral

  • The evaluation of the patient with gastric carcinoma should proceedwith due diligence and rapidity.
  • Patients with gastric ulcers that are biopsy-negative should be placedon intensive medical therapy and reevaluated with endoscopy in 6 weeks. Ifthe ulcer is healing, another 6 weeks of medical therapy is appropriate.If the ulcer has not healed completely, surgery is indicated, as antralor lesser curvature ulcers may be malignant.

Preoperative Evaluation for Extent of Disease

Physical examination

  • Assess for lymphadenopathy with careful attention to left supraclavicularfossa.
  • Assess for abdominal mass, ascites.
  • Assess for blood in stool and Blummer's shelf by digital examination.
  • Assess for ovarian or peritoneal metastasis on pelvic examination inwomen.

Esophagogastrodenoscopy (EGD) (required)

  • Assess tumor extent.
  • Assess tumor location.
  • Assess degree of obstruction. EGD may also be utilized with dilation,laser, or electrofulguration to temporarily relieve obstruction due togastroesophageal junction tumors
  • Assess for and control hemorrhage.
  • Obtain biopsy, which is essential, as a gastric neoplasm may be anadenocarcinoma, a lymphoma or, rarely, a GIST.

CT scan of lower chest and abdomen/barium study

  • Assess extent of local disease (barium study may help delineate proximalextent of disease, and therefore, facilitate planning of surgery) and extentof metastatic disease.

Standard preoperative tests

  • Standard tests, including an ECG, chemistry profile, and electrolytes,are appropriate as warranted for anesthesia.

Cardiorespiratory assessment

  • In selected patients

Role of the Surgeon in Initial Management

Evaluation of the symptomatic patient

  • The surgeon should be involved in the evaluation of the symptomaticpatient, especially as it relates to the timing of an operation.
  • Patients with gastric outlet obstruction may benefit from a short periodof nasogastric intubation to decompress the stomach. Fluid and electrolytesshould be corrected during this period of nasogas- tric suction.
  • Similarly, the surgeon may also perform the upper endoscopy to definethe extent of disease. Clearly, the surgeon should be involved at an earlystage in the evaluation of the extent of the patient's disease and theassessment of significant comorbid disease.

Diagnostic procedures

  • Endoscopy usually provides the biopsy that confirms the diagnosis.Similarly, enlarged left supraclavicular nodes are amenable to biopsy.

Surgical considerations
The options for the management of gastric cancer are diverse. The surgicalprocedures that may be performed are:

  • Diagnostic laparoscopy (to evaluate stage of disease and to providea histologic diagnosis)
  • Palliative distal gastrectomy
  • Palliative proximal gastrectomy
  • Palliative total gastrectomy
  • Curative distal gastrectomy
  • Curative proximal gastrectomy
  • Curative total gastrectomy
  • The role of routine D2 nodal dissection remains controversial.

These guidelines are copyrighted by the Society of Surgical Oncology(SSO). All rights reserved. These guidelines may not be reproduced in anyform without the express written permission of SSO. Requests for reprintsshould be sent to: James R. Slawny, Executive Director, Society of SurgicalOncology, 85 W Algonquin Road, Arlington Heights, IL 60005.

Gastric adenocarcinoma has declined in frequency by more than 40% overthe last 3 decades but is still the eighth most common cause of cancer-relateddeaths in the United States.[1] It remains a difficult cancer to treat,with overall 5-year survival rates of 5% to 15% in the United States, largelydue to the advanced stage of disease at the time of diagnosis.[2] The incidenceof proximal gastric carcinoma is also increasing in western Europe andthe United States.[3,4]

Adjuvant therapy has not improved survival in patients who have undergonea potentially curative resection in eight of nine prospective, randomizedtrials,[5-13] and the only trial that demonstrated a significant positiveeffect[8] has not been confirmed by subsequent trials. In a current adjuvanttherapy trial, the addition of chemoradiation to potentially curative surgeryis being compared to surgery alone. However, this trial has not yet beencompleted.

The lack of efficacy of current multimodality treatment may be due toboth imprecise staging of gastric cancer and regimens that are not sufficientlyactive. To demonstrate a survival benefit for the use of multimodalitytherapy requires either a minor treatment effect of combination therapyin a disease that is uniformly fatal, such as pancreatic cancer, or a majoreffect in a disease that is slightly less aggressive (eg, testicular cancer[14]).Thus, determination of the effect of adjuvant therapy in gastric cancerawaits improvements in both staging and combination therapy.


The TNM classification[15] is used to stage gastric carcinoma. Thisstaging system (Table 1) is based largelyon the recommendations of the Japanese Surgical Society,[16] which identifieddifferent combinations of pathologic T- and N-stages that have similarclinical outcomes. The N1 lymph nodes are those within 3 cm of the primarytumor, while the N2 nodes are those that are more than 3 cm from the primarytumor and include the celiac, splenic, common hepatic, and left gastricnodes. The N3 and N4 nodes include the periportal, retropancreatic, andpara-aortic nodes.

Five-year survival data from Maruyama[17] demonstrate the interactionbetween T- and N-staging to produce substages (Table1). In this study, all patients had N2 nodes removed, so that patientswhose disease was classified as either N0 or N1 were not likely to haveundocumented involvement of N2 nodes. These data also indicated that involvementof N3 or N4 nodes has the same impact on survival as does the presenceof distant metastases.

These findings may explain why the survival of American and Europeanpatients appears to be worse than that of Japanese patients.[2] However,survival of Western patients with early gastric cancer may be similar tothat of Japanese patients if they undergo removal of nodes to ensure adequatestaging.

Node Dissections

Node dissections are termed D1, D2, or D3 depending on whether onlythe N1 nodes, the N1 and N2 nodes, or the N1, N2, and N3 nodes are removed,respectively. In patients with stage IB gastric cancer, 5-year disease-freesurvival rates have ranged from 0% to 85%.[3,18-21] The worst survivalwas observed in the series that performed predominantly D1 node dissections.When D2 dissections were performed routinely, survival was as high as 85%.[19]

Similarly, survival rates in patients with stage II or III disease treatedwith D2 dissections by Gall and Hermanek[22] were 70% and 30%, respectively.These survival rates compare favorably with the 66% and 48% to 56% ratesobserved by Abe et al[23] for the same disease stages. Rohde et al[18]also achieved similar results for stage IIIA and IIIB disease in a seriesin which there was a high frequency of D2 node dissections.

Thus, it is possible, with careful staging, to achieve 5-year survivalrates similar to those reported by the Japanese. However, as Shiu et al[24]demonstrated, the node dissection must be one level greater than the levelof pathologically involved nodes in order to improve outcome. This suggeststhat micrometastatic disease may be left behind in second-echelon nodesif a D1 dissection is performed, or that metastases can occur in second-levelnodes without involving the first-level nodes. Either situation will obfuscatedifferences in survival that may be observed in adjuvant trials becausethe histologic status of the regional lymph nodes will not be accuratelydefined.

Finally, recent data suggest that routine performance of a D2 node dissectionmay not improve overall survival of gastric cancer patients but may improvestage-specific survival.[25,26] Previous attempts to perform randomizedprospective trials of node dissections have failed to identify a significantsurvival benefit for wider nodal dissections. For example, a study by Robertsonet al[27], in which patients were randomized between perigastric D1 lymphnode dissection and D3 node dissection, showed no survival benefit forextended lymphadenectomy. In fact, patients in the D1 node dissection grouphad a significantly improved overall survival compared with patients inthe D3 node dissection group. Moreover, extended lymphadenectomy increasedthe morbidity of surgical resection but did not significantly increaseperioperative mortality.

Another randomized prospective trial by Dent et al from South Africa[28]demonstrated that D2 dissection resulted in a much longer operating timeand greater morbidity without improving overall survival. However, thenumber of patients randomized represented only a small subset of the overallstudy. This study demonstrates the problem with this type of surgical trial;namely, the failure to perform a wider specified node dissection in patientswho were supposed to undergo such a dissection. This issue was also addressedin a study by Bunt et al from Leiden.[29] In a very carefully performedprospective trial, these surgeons showed that while stage-specific survivalis improved in patients who undergo D2 node dissections, this is due tothe Will Rogers' effect (stage migration). The overall survival of patientswho had a D2 node dissection was no better than the overall survival ofpatients who underwent the less extensive node dissection.


The data described above suggest that the current management of gastriccarcinoma requires a surgical resection. Identification of the number oflymph nodes that may be involved with cancer is an important prognosticator.Resection of involved lymph nodes is likely to decrease the morbidity oflocoregional recurrence. However, this has not been adequately addressedin the studies performed to date. Nonetheless, accurate staging is an essentialgoal of the primary therapy of gastric carcinoma.

The limits of resection for a gastric cancer are outlined in the guidelines. Administration of adjuvant therapy, such as chemotherapy or radiationtherapy, remains controversial and is the subject of ongoing clinical research.Until a prospective, randomized trial demonstrates a significant improvementin survival, there remains little or no basis for the addition of adjuvanttherapy outside the context of a clinical trial.


1. Parker SL, Tong T, Bolden S, et al: Cancer statistics, 1997. CA CancerJ Clin 47:5-27, 1997.

2. Alexander HR, Kelsen DG, Tepper JC: Cancer of the stomach, in DeVitaVT Jr, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice ofOncology, 5th ed, pp 1021-1054. Philadelphia, Lippincott, 1977.

3. Myers WC, Damiano RJ, Rotolo FS, et al: Adenocarcinoma of the stomach:Changing patterns over the last 4 decades. Ann Surg 205:1-8, 1987.

4. Salvon-Harman JC, Cady B, Nikulasson S, et al: Shifting proportionsof gastric adenocarcinomas. Arch Surg 129(4):381-8; 1994.

5. Dixon WJ, Longmire WP Jr, Holden WD: Use of triethylenethiophosphormiadeas an adjuvant to the surgical treatment of gastric and colorectal carcinoma.Ann Surg 172:26-39, 1971.

6. Serlin O, Wolkoff JS, Amadeo JM, Keehn RJ. Use of 5-fluorodeoxyuridine(FUDR) as an adjuvant to the surgical management of carcinoma of the stomach.Cancer 24:223-228, 1969.

7. The Veterans Administration Surgical Oncology Group: Efficacy ofprolonged intermittent therapy with combined 5-FU and methyl-CCNU followingresection for gastric carcinoma. Cancer 52:1105-1112, 1985.

8. The Gastrointestinal Tumor Study Group: Controlled trial of adjuvantchemotherapy following curative resection for gastric cancer. Cancer 49:1116-1122,1982.

9. The Eastern Cooperative Oncology Group: Postoperative adjuvant 5-fluorouracilplus methyl-CCNU therapy for gastric cancer patients. Cancer 55:1868-1873,1985.

10. Schein P, Coombes R, Chilvers C: A controlled trial of FAM chemotherapyas adjuvant treatment for resected gastric carcinoma. Proc Am Soc ClinOncol 5:79, 1986.

11. Krook JE, O'Connell MJ, Wiend HS: Surgical adjuvant therapy of gastriccancer with doxorubicin and 5-fluorouracil: A joint Mayo/North CentralCancer Treatment Group Study. Proc Am Soc Clin Oncol 7:93, 1988.

12. Allum WH, Hallissey MT, Kelly KA: Adjuvant chemotherapy in operablegastric cancer: 5 year follow-up of first British Stomach Cancer Grouptrial. Lancet 1:571-574, 1989.

13. Bleiberg H, Goffin JC, Dalesio O, et al: Adjuvant radiotherapy andchemotherapy in resectable gastric cancer: A randomized trial of the gastro-intestinaltract cancer cooperative group of the EORTC. Eur J Surg Oncol 15:535-543,1989.

14. Williams SD, Birch R, Einhorn LH, et al: Treatment of disseminatedgerm-cell tumors with cisplatin, bleomycin, and either vinblastine or etoposide.N Engl J Med 316:1435-1440, 1987.

15. Behrs OH, Henson DE, Hutter RVP, et al (eds): Manual for Stagingof Cancer 4th ed, p 65. Chicago, American Joint Committee on Cancer, 1992.

16. Japanese Research Society for Gastric Cancer: The general rulesfor the gastric cancer study in surgery and pathology: Part I. Clinicalclassification. Part II. Histological classification of gastric cancer.Jpn J Surg 11:127-145, 1981.

17. Maruyama K: Surgical Treatment and End Results of Gastric Cancer,p 29. Tokyo, Japan, Department of Surgery, National Cancer Center, 1985.

18. Rohde H, Bauer P, Stutzer H, et al: Proximal compared with distaladenocarcinoma of the stomach: Differences and consequences: German GastricCancer TNM Study Group. Br J Surg 78:1242-1248, 1991.

19. Lawrence M, Shiu MH: Early gastric cancer: Twenty-eight-year experience.Ann Surg 213:327-334, 1991.

20. Bernard A, Obadia JF, Arnould H, et al: Facteurs influencant lasurvie apres resection d'un cancer du cardia: Comparaison de l'oesogastrectomietotale et de l'oesogastrectomie polaire superieure. Ann Chir 44:459-463,1990.

21. Ponsioen CY, Welvaart K, Hermans J: Significance of tumor invasionand lymph node involvement on the prognosis and selection for surgery ofadenocarcinoma of the cardia. Eur J Surg Oncol 15:301-306, 1989.

22. Gall FP, Hermanek P: New aspects in the surgical treatment of gastriccarcinoma--a comparative study of 1636 patients operated on between 1969and 1982. Eur J Surg Oncol 11:219-225, 1985.

23. Abe S, Shiraishi M, Nagaoka S, et al: Serosal invasion as the singleprognostic in- dicator in stage IIIA (T3N1M0) gastric cancer. Surgery 109:582-588,1991.

24. Shiu MH, Moore E, Sanders M, et al: Influence of the extent of resectionon survival after curative treatment of gastric carcinoma: A retrospectivemultivariate analysis. Arch Surg 122:1347-1351, 1987.

25. Bunt AMG, Hermans J, Smit VTHBM, et al: Surgical/pathologic-stagemigration confound comparisons of gastric cancer survival rates betweenJapan and western countries. J Clin Oncol 13:19-25, 1995.

26. Jessup JM: Is bigger better (editorial)? J Clin Oncol 13(1):5-7;1995.

27. Robertson CS, Chung SC, Woods SD, et al: A prospective randomizedtrial comparing R1 subtotal gastrectomy with R3 total gastrectomy for antralcancer. Ann Surg 220:176-182, 1994.

28. Dent DM, Madden MV, Price SK: Randomized comparison of R1 and R2gastrectomy for gastric carcinoma. Br J Surg 75:110-112, 1988.

29. Bunt AM, Hermans J, Smit VT, et al: Surgical-pathologic stage migrationconfounds comparisons of gastric cancer survival rates between Japan andwestern countries. J Clin Oncol 13:19-25, 1995.