Surgical treatment

Actually, surgery is the only reliable possibility of a curative treatment. The aim of surgery is to remove as completely as possible all grossly visible tumor tissue and to obtain histologically free surgical margins. This goal is usually reached in 45% of cases of diagnosed GC in population-based series and up to 5560% of cases in specialized centers During the operative procedure, gastric resection depends on cancer spread, i.e. tumoral infiltration through the gastric wall, tumoral extension to adjacent organs, and lymph node involvement.

Limits of gastric resection

The extension of gastric resection depends on the location of the tumor y Cancer located to the body or the corpus of the stomach requires total gastrectomy. Reconstruction of digestive continuity is then realized by a Roux-en-Y oesojejunostomy. Pouch and Roux-en-Y reconstruction seem to improve postoperative quality of life after total gastrectomy. y Cancer of the antrum (distal third and pylorus), may be managed by sub-total distal gastrectomy. Reconstruction is realized by a method similar to Bilroth I or II procedures. Roux-en-Y gastrojejunostomy has been proposed to avoid bile-reflux in the gastric remnant , but vagotomy is mandatory to prevent anastomotic peptic ulcer, depending on the size of the gastric remnant. This type of reconstruction called total duodenal diversion has been proposed in patients with severe gastrooesophageal reflux disease ; however, it can be indicated for GC after distal gastrectomy, in cases of patients with a hope of long term survival as early gastric cancer (EGC) for example. For a long time, routine radical total gastrectomy was proposed for distal lesions by general authors , particularly from Japan ; it rationale was based on the effect that survival was better after an extensive lymphadenectomy, including pancreatic tail resection; in these conditions, total gastric resection was necessary. Actually, routine total gastrectomy is no more the only recommended treatment for distal lesions as it was demonstrated in a French prospective multicentric controlled study . In this study , there was no significant difference on the 5-year survival rate between total or subtotal distal gastrectomy for distal lesions. However, one important point was the fact that after subtotal distal gastrectomy, free margins of resection should not be less than 56 cm on the stomach and no less than 2 cm on the proximal duodenum. y Cancer of the cardia needs a particular approach. In fact, they are consider as a different clinical entity , while others assimilate them to a lower oesophageal cancer whatever histological differences. As a matter of fact, the limits of the resection depends on the oesophageal extension itself. Despite a lack of controlled study, there is a tendency to achieve total gastrectomy for lesions limited to the cardia and proximal oesophago-gastrectomy by abdominal and right thoracic combined approach (Lewis-Santy procedure) for lesions extended to the lower oesophagus . y Borrmann type 4 infiltrative GC, whatever their topography (partial or total), are usually treated by total gastrectomy because of a frequent wide extension through the gastric wall . The relative incidence of this type of cancer is increasing with time, and wider resection to surrounding organs including extended lymphadenectomy, as proposed by Japanese authors , does not seem to improve prognosis for all of the TNM stages, except probably in stage III , but no controlled trials has been performed in Borrmann type 4 GC. y At last, EGC, whatever its location, requires similar treatment as for other types because of the possibility of wide submucosal extension and lymph-node involvement. Some Japanese authors suggested limited resections for EGC, but these procedures are not currently diffused in Europe.

Lymph-node dissection

Lymph node extension is the most prognostic feature in GC. Therefore the crucial question of the last twenty years was related to the prognostic impact of lymph node dissection. Many studies have been realized successively: retrospective, prospective and finally controlled prospective studies. During the last ten years, four prospective controlled studies compared the type of lymph-node dissection limited (D1) versus extended (D2), in the surgical treatment of GC. The two last controlled studies gave firstly only results about mortality and morbidity of the D1 and D2 procedures , then secondly results about long term survival . Actually, until now, none of these randomized trials has demonstrated the superiority of D2 versus D1 specially in term of 5-year survival (no superiority after D2 dissection); furthermore, they showed an increased incidence of postoperative complications rate after D2 vs. D1 dissection. Highest rates of morbidity and mortality are partly due to anastomotic leakage and consequences of the pancreatic tail resection during D2 dissection. These findings led some authors to conclude that D2 resection without pancreatico-splenectomy, excluding the negative effect on operative mortality, may be a good approach for the lymph-node dissection of GC rather than standard D1. The debate seemed to be definitively closed, but a recent multicentric prospective non controlled study pointed out a significant improvement of 10-year survival rate in TNM stage II cancer after D2 dissection (defined in this study as an extended lymph-node dissection with more than 25 removed nodes). Although this study was multicentric, involving many surgeons, standardization of lymph-node dissection was set and routinely realized after several meetings, as well as the last controlled trials . In addition to these results, extended lymph-node dissection did not increase mortality and morbidity rates. In France, standardized extended lymphadenectomy for gastric cancer is not routinely performed by all the surgical teams, and systematic count of lymph-nodes in the specimen, as proposed in most of actual trials specially by Japanese authors, but also by western ones, is not standardized for all western pathologists. In the near future, the standardized techniques of lymph-node dissection and pathological analysis might be needed for oncological accreditation. As morbidity and mortality are correlated to resection of the spleen and pancreatic tail in controlled studies, the German study suggests that D2 resection should be associated to subtotal distal gastrectomy in stage II and IIIA to prevent the necessity of a resection of the spleen and pancreatic tail. When a total gastrectomy is performed in curative intent, intermediate 1.5 lymph-node dissection between D1 and D2 could be realized, including splenic lymph-node dissection without splenectomy as it was suggested by a Japanese author . Despite the effect of lymph-node dissection on survival has not been proved, it is worth to standardize and familiarize European teams, to precise staging and margin clearance.

Palliative surgical treatment

The best palliation in GC whenever possible is still surgical resection. In fact, morbidity and mortality of palliative surgery without resection (laparotomy alone or by-pass procedures) is extremely high and should be avoided. Perhaps a better pre-operative evaluation by CT scan (heliscan) and/or laparoscopic staging indicated in selected patients could decrease the number of explanatory laparotomy in unresectable GC . Although, 25% of the patients with diagnosed GC can benefit from palliative procedures . There are two different types of the palliative treatment of GC: resection of the tumor and surgical by-pass procedures without resection. Actual pre-operative investigations can not always predict the type of operative procedure as exactly as during operative exploration. Laparoscopic staging could be indicated in these conditions . Mostly, in many cases, the possibility of tumoral resection appears to surgeons as a perioperative finding, and peroperative manual exploration may find hepatic metastasis, wide or localized peritoneal implants. in these conditions, palliative surgery depends on local anatomy and preoperative clinical symptoms. A bleeding tumor is more to be resected than an obstructive one for which a by-pass might be recommended. In a general manner, oncologic rules of resections must respect the followings: little free margin on surrounding organs, inutility of lymph node dissections, unless it is required to obtain a free margin. There is a lower mortality and morbidity in palliative

resections rather than in by-pass without resections. However, by-pass procedures can still be indicated when resection risk appears to be too high (morbidity and mortality) and/or in case of biliary and/or digestive obstruction. Then, a gastroenterostomy and/or a biliary diversion may be realized.

Non operative treatment

Non surgical treatment represents 30% of diagnosed GC and is indicated in case of diffuse hepatic, peritoneal and/or extra-abdominal metastasis without obstructive symptoms, sub-clavicular lymph nodes and/or the presence of severe physiological disorders and/or undernutrition. Treatment in these cases is difficult and varies from abstention to endoscopic desobstruction (endoscopic laser therapy, argon beam, heater probe coagulation therapy, endoscopic stent). Chemotherapy can be indicated only in phase II protocols.

http://www.ncbi.nlm.nih.gov/books/NBK6969/

Treatment Treatment depends upon the clinical stage at presentation and the comorbid disease of the patient. For patients who are medically fit and who have potentially resectable disease, surgery should be performed with an effort to achieve an R0 resection that is defined as a curative en bloc resection with negative proximal, distal, and radial margins (i.e. no obvious residual tumor). The surgeon should attempt to achieve > 5 cm proximal and distal margins. For distal tumors, this usually requires a distal gastrectomy. For proximal tumors, a total gastrectomy or proximal gastrectomy is acceptable. For gastric cancers that approach the gastroesophageal junction, the proximal margin should be at least 6 cm; this requirement usually mandates a formal esophagogastric resection with thoracotomy. An extended (D2) lymph node dissection (removing an average of 25-30 lymph nodes) can provide more complete staging than a limited (D1) lymph node dissection (removing an average of 15 lymph nodes), but there are no randomized Western trials that demonstrate a survival advantage associated with a D2 dissection yet. A large multi-center, randomized Dutch trial demonstrated that a D2 dissection was associated with increased morbidity and mortality but not with increased survival. A smaller, controlled British trial confirmed the findings of the Dutch study. Splenectomy should be avoided unless the spleen is involved by tumor and at least 15 lymph nodes should be assessed to properly stage the lymph nodes. To ensure that 15 lymph nodes are removed usually requires removal of the lesser omentum, greater omentum, common hepatic arterial lymph nodes, and the left gastric lymph nodes to the celiac axis. In most cases, this lymph node dissection approximates a D2 dissection with the important exception that the splenic hilar lymph nodes are not removed (which is an important component of most D2 dissections) to avoid removal of the spleen. After curative resection, adjuvant combination chemotherapy and radiotherapy should be offered patients with node positive disease or tumors that penetrate the full thickness of the muscularis propria (T2b). This standard was set with the results from a randomized, controlled clinical trial published in 20017. This multi-center trial (INT-0116) treated patients with Stages Ib through IV M0 with one cycle of 5-FU and Leucovorin followed by 4.5 Gy external beam radiotherapy with 5-FU and Leucovorin, followed by two cycles of 5-FU and Leucovorin at monthly intervals. Between 1991 and 1998, 603 patients entered the study. Nodal metastases were noted in 85%. Improved survival was noted in the treatment arm, and will be discussed below. For patients who are medically fit with locally advanced unresectable disease, combination chemotherapy and radiotherapy can be used. For patients who have comorbid disease that makes the risk of resection greater than the benefit, combination chemotherapy and radiation, chemotherapy alone, or no treatment (best supportive care) are options. For patients with metastatic disease, chemotherapy should be considered. For patients who present with bleeding and are not candidates for resection from the standpoint of either medical condition or metastases, radiation therapy can be helpful for palliation.

http://www.ssat.com/cgi-bin/guidelines_SurgicalTreatmentGastricCancer_EN.cgi