Your SlideShare is downloading. ×
Acs0508 Tumors Of The Stomach, Duodenum, And Small Bowel 2005
Upcoming SlideShare
Loading in...5

Thanks for flagging this SlideShare!

Oops! An error has occurred.

Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

Acs0508 Tumors Of The Stomach, Duodenum, And Small Bowel 2005


Published on

1 Like
  • Be the first to comment

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

No notes for slide


  • 1. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 1 8 TUMORS OF THE STOMACH, DUODENUM, AND SMALL BOWEL Jeffrey D.Wayne, M.D., F.A.C.S., and Mark S.Talamonti, M.D., F.A.C.S. Gastric Adenocarcinoma reported between cancer of the gastric cardia and infection with The incidence of gastric carcinoma exhibits significant geo- Helicobacter pylori or Epstein-Barr virus.10,11 graphic variability. The disease is most common in Japan and CLASSIFICATION China, and high rates of occurrence have also been reported in Central and South America, Eastern Europe, and parts of the Adenocarcinoma of the stomach may be divided into two his- Middle East.1 In most of the more developed nations, however, tologic subtypes, intestinal and diffuse.12 Each subtype has unique gastric carcinoma is relatively uncommon.The overall incidence of pathologic, epidemiologic, etiologic, and prognostic features. The this condition has decreased in the past few decades, but gastric intestinal (or glandular) subtype usually arises in the distal stom- carcinoma remains the second leading cause of cancer death ach (often after a long precancerous phase), is more common in worldwide. The reported reductions in gastric cancer mortality elderly patients, and has been closely associated with atrophic gas- may be linked to better refrigeration and a concomitant decrease tritis and diets high in nitrates and nitrose compounds.13 The char- in the intake of salted, pickled, smoked, and chemically preserved acteristic histologic finding is cohesive neoplastic cells that form foods; however, this link remains controversial. An inverse associ- glandlike tubular structures.The diffuse subtype occurs more fre- ation with the consumption of fresh fruits and vegetables has also quently in younger patients and has no identifiable precursor le- been noted.2 sion. It may develop in any part of the stomach but shows a pre- Gastric cancer occurs 1.5 to 2.5 times more frequently in males dilection for the cardia. Cell cohesion is absent; thus, individual than in females. It is rarely diagnosed before the age of 40, and its cancer cells infiltrate and thicken the stomach wall without form- incidence peaks in the seventh decade of life. African Americans, ing a discrete ulcer or mass. Hispanic Americans, and Native Americans are two times more In general, the prognosis for the diffuse subtype is worse than likely to have gastric cancer than white Americans are.3 that for the intestinal subtype. Whereas intestinal lesions are seen In the United States in particular, the incidence of stomach more frequently in regions with a high incidence of gastric cancer, cancer has fallen substantially over the past 70 years.4 Whereas this the incidence of diffuse lesions is constant among various popula- disease was once a leading cause of cancer-related death in the tions throughout the world.14 Accordingly, the overall decline in gas- United States, it now ranks 13th among major causes. Unfortun- tric cancer over the past century has been attributed to a decline in ately, the decline in incidence has not translated into an improve- intestinal lesions and to a decline in the incidence of H. pylori ment in the 5-year survival rate.5 Across all races, the 5-year rela- infection (see below). tive survival was 23% for the period extending from 1992 to 1999.3 RISK FACTORS This result is probably related to the advanced stage at which most patients present. A 1995 study from the Commission on Cancer Historical studies of specimens obtained during operation or at of the American College of Surgeons (ACS) found that 66% of autopsy suggest that gastric carcinoma, especially of the intestinal patients with gastric cancer presented with locally advanced or subtype, frequently develops in the presence of chronic atrophic metastatic disease.6 Resection rates ranged from 30% to 50%, and 5- gastritis and associated intestinal metaplasia. It has generally been year survival rates after resection with curative intent were directly assumed that adenocarcinoma of the distal stomach progresses related to stage at presentation. For stage I disease, the survival rate from chronic gastritis to metaplasia through the teratogenic influ- was 43%; for stage II, 37%; for stage III, 18%; and for stage IV, 20%. ence of environmental factors. The most commonly studied envi- Another relevant change in the epidemiology of gastric cancer is ronmental factors are the nitrates and nitrose compounds present a shift in the distribution of primary lesion sites within the stom- in high levels in salted, smoked, or pickled foods consumed in ach. In the first quarter of the 20th century, two thirds of gastric areas where gastric cancer is endemic.15 To date, however, no cancers were located in the antrum and the prepyloric area, and prospective studies have conclusively demonstrated that modern only 10% arose in the cardia or the esophagogastric junction. Since refrigeration practices and the subsequent decline in the salting, the 1970s, however, adenocarcinoma of the proximal stomach has smoking, and pickling of food have been responsible for the rela- become increasingly common. In one study, the incidence of ade- tive decline in intestinal gastric cancer. Furthermore, the intestinal nocarcinoma of the gastric cardia rose from 29.1% to 52.2% in the subtype may arise in the absence of metaplasia. Finally, the emer- period between 1984 and 1993.7 In another, which included 18,365 gence of chronic infection with H. pylori as the dominant risk fac- gastric cancer patients from ACS-approved hospitals, a full 31% of tor for gastric adenocarcinoma has challenged the paradigm of the tumors were found to be in the proximal stomach, compared with atrophic gastritis–intestinal metaplasia–gastric cancer sequence. only 26% in the distal third.8 In the United States, carcinoma of the Epidemiologic studies across various populations worldwide cardia occurs primarily in whites, with a male-to-female ratio of have consistently demonstrated a strong association between H. approximately 2:1. Cancer of the cardia appears to be distinct from pylori infection and gastric cancer.16 Prospective serologic studies adenocarcinoma of the distal esophagus, which frequently arises in have confirmed that persons with evidence of such infection are the setting of Barrett’s esophagus.9 Associations have also been three to six times more likely to have gastric cancer than persons
  • 2. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 2 Table 1 American Joint Committee of Cancer TNM INVESTIGATIVE STUDIES Clinical Classification of Gastric Carcinoma Until comparatively recently, an upper gastrointestinal series was often the first diagnostic test ordered to evaluate symptoms TX Primary tumor cannot be assessed related to the upper GI tract. However, even with double-con- T0 No evidence of primary tumor trast techniques, which allow improved visualization of mucosal Tis Carcinoma in situ: intraepithelial tumor without invasion detail, false negative rates as high as 25% were reported, espe- of lamina propria T1 Tumor invades lamina propria or submucosa cially with small lesions (i.e., 5 to10 mm).24 Accordingly, in most Primary tumor (T) T2 Tumor invades muscularis propria or subserosa large series, fiberoptic endoscopy with biopsy has replaced con- T2a: tumor invades muscularis propria trast radiography as the primary diagnostic technique.25 Upper T2b: tumor invades subserosa GI endoscopy with biopsy has been reported to have a diagnos- T3 Tumor penetrates serosa (visceral peritoneum) without tic accuracy of 95%.20 However, false negatives have been report- invasion of adjacent structures ed, especially in the context of inadequate biopsies.Thus, it is re- T4 Tumor invades adjacent structures commended that at least four biopsy specimens taken from the NX Regional lymph node(s) cannot be assessed region of any atypical findings.26 N0 No regional lymph node metastasis Regional lymph N1 Metastasis in 1–6 regional lymph nodes STAGING nodes (N) N2 Metastasis in 7–15 regional lymph nodes Two major classification systems are available for staging gas- N3 Metastasis in > 15 regional lymph nodes tric cancer. The first is the one used in Japan, where gastric can- MX Distant metastasis cannot be assessed cer is staged according to the general rules for gastric study in Distant metastasis (M) M0 No distant metastasis surgery and pathology published by the Japanese Research Society M1 Distant metastasis for Gastric Cancer (JRSGC).27 This elaborate system focuses on the anatomic involvement of specifically numbered lymph node stations. The second system is the one generally used in Western who are seronegative.17 Still, only a very small fraction of infected countries—namely, the familiar tumor-node-metastasis (TNM) persons have gastric cancer. It has been estimated that more than system developed by the American Joint Committee on Cancer half of the world’s inhabitants may be infected with H. pylori—a (AJCC) and the International Union Against Cancer (UICC) [see number that dwarfs the actual incidence of gastric cancer.What is Tables 1 and 2].28 The AJCC/UICC staging system is based on a clear is that H. pylori infection of the gastric mucosa leads to a state gastric cancer database and classifies lesions according to the of chronic active inflammation that lasts for decades. This inflam- depth to which the primary tumor penetrates the gastric wall, the matory process appears to be modulated by multiple forces, in- extent of lymph node involvement, and the presence or absence cluding genetic and environmental factors.18 Inherited traits may of distant metastases. confer susceptibility or resistance to carcinogenesis. Indeed, first- The primary goal in the evaluation of gastric cancer patients is degree relatives of gastric cancer patients have a two to three times to stratify them into two clinical stage groups: those with loco- higher relative risk of contracting the disease.19 Gastric irritants regional disease (AJCC stages I to III) and those with systemic may act as promoters, and antioxidants may have a protective disease (AJCC stage IV).29 The National Comprehensive Cancer effect (which may be part of the reason for the reduced risk of gas- Network (NCCN) has developed consensus guidelines for the tric cancer associated with diets rich in fruits and vegetables).20 clinical evaluation and staging of patients with possible gastric Unlike intestinal cancers, diffuse cancers appear not to be asso- cancer. These guidelines are accessible to any practitioner via ciated with H. pylori infection. Diffuse adenocarcinoma of the stomach is more common in young patients and has no known precursor lesion.9 The incidence of genetically associated diffuse cancers is estimated to be in the range of 5% to 10%.19 Familial Table 2 American Joint Committee on Cancer cases of diffuse gastric cancer occur at an average age of 38 years Staging System for Gastric Carcinoma and are inherited in an autosomal dominant fashion with 70% penetrance.21 Patients with blood group A have a 16% to 20% Stage T N M increased risk of gastric cancer.22 Stage 0 Tis N0 M0 CLINICAL EVALUATION Stage IA T1 N0 M0 In high-risk areas (e.g., Japan), mass screening programs have been successful in identifying early gastric cancer, which is gener- Stage IB T1 N1 M0 ally amenable to surgical cure.23 In fact, in some Japanese studies, T2a, T2b N0 M0 as many as 40% of newly diagnosed patients had early gastric can- T1 N2 M0 cer. Unfortunately, in Western countries, the disease is almost Stage II T2a, T2b N1 M0 always diagnosed relatively late, when it is locally advanced or T3 N0 M0 metastatic.When it is superficial, gastric cancer typically produces no symptoms. As it progresses, however, a constellation of vague, T2a, T2b N2 M0 Stage IIIA T3 N1 M0 nonspecific symptoms may develop, including anorexia, fatigue, T4 N0 M0 weight loss, and epigastric discomfort. Dysphagia, early satiety, vomiting, and hematemesis also are seen, albeit rarely; when present, Stage IIIB T3 N2 M0 they often indicate advanced disease. Indeed, early gastric cancer T4 N1, N2, N3 M0 has no characteristic physical findings, and many patients are not Stage IV T1, T2, T3 N3 M0 diagnosed until they present with jaundice, ascites, or a palpable Any T Any N M1 mass, all of which signal incurable disease.
  • 3. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 3 a b Figure 1 Shown are CT scans of a patient with a T3 carcinoma involving the posterior wall of the gastric antrum, taken with the patient supine (a) and prone (b). By placing the patient in the prone position and dis- tending the stomach with water, better definition of the extent of the tumor and clearer delineation of the inter- face between the stomach and the pancreas is achieved. the Internet ( CT with intravenous contrast plus appropriate gastric distention gastric.pdf) and are updated annually. Multidisciplinary evalua- with 600 to 800 ml of water (a negative contrast agent), have tion is recommended for all patients. A careful history is obtained allowed modest improvements in overall staging with CT [see and a thorough physical examination performed, with special Figure 1]. Nevertheless, CT is still limited in its ability to evaluate attention paid to comorbid conditions that might preclude opera- peritoneal disease and liver metastases smaller than 5 mm.31 tive intervention. Initial laboratory studies include a complete Given the limitations of CT, we believe that in the absence of blood cell count with a platelet count; determination of serum elec- obvious metastatic disease, locoregional staging with endoscop- trolyte, blood urea nitrogen, creatinine, and glucose concentrations; ic ultrasonography (EUS) is vital for accurately assessing tumor and a liver function panel. Chest radiography is performed, along penetration through the gastric wall (T stage) and ascertaining with computed tomography of the abdomen and pelvis. whether regional nodes (N stage) or even mediastinal or para-aor- Whereas CT is invaluable for detecting ascites, bulky adenopa- tic lymph nodes may be involved (which would be considered M1 thy, and significant visceral metastases, its overall accuracy in stag- disease) [see Figure 2]. EUS is unique among imaging modalities ing tumors is modest: only 70% for advanced lesions and 44% for in its ability to image the gastric wall as a five-layer structure, with early lesions.30 CT assesses lymph node involvement primarily on each layer correlating with an actual histologic layer.32 The overall the basis of node size. Thus, its sensitivity for N1 and N2 disease accuracy of EUS in determining the extent of infiltration ranges is low, ranging from 24% to 43%; however, its specificity is high, from 67% to 92%.33 EUS features that suggest lymph node approaching 100%. Technical advances, such as spiral (helical) metastasis include a rounded shape, hypoechoic patterns, and a a b Figure 2 (a) Shown is an EUS image of a T3 gastric neoplasm. (b) EUS reveals the presence of suspicious perigastric (N1) nodes, later confirmed as malignant at operation.
  • 4. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 4 Patient has biopsy-proven gastric adenocarcinoma Perform helical CT of abdomen and pelvis with oral and I.V. contrast. Perform EUS. Tumor is resectable Tumor is not resectable Assess severity of symptoms. Evidence of widely metastatic disease or the presence of locally advanced unresectable disease is a contraindication to resection. Provide supportive care. Consider palliative chemotherapy, endoluminal stenting, or entry into advanced disease clinical trials. Symptoms are severe Symptoms are mild or absent (e.g., bleeding or obstruction) Stage tumor with laparoscopy. Perform exploratory laparotomy. Resect tumor. Laparoscopic staging reveals Laparoscopic staging no evidence of M1 disease reveals M1 disease Exploration reveals M1 disease Exploration reveals no evidence of M1 disease Resect tumor. Provide supportive care. Resection is palliative in intent. Consider neoadjuvant therapy Consider palliative chemotherapy, Resection is curative in intent. in a clinical trial setting. endoluminal stenting, or entry into advanced disease clinical trials. Potentially curative resection involves • Clear surgical margins • Complete nodal dissection • Adjuvant 5-FU–based chemoradiation therapy Tumor is in distal stomach Tumor is in fundus or proximal stomach Tumor is in distal esophagus, esophagogastric junction, or cardia Perform subtotal gastrectomy with Perform total gastrectomy with D2 dissection D2 dissection and Billroth II reconstruction. and esophagojejunal reconstruction. Perform transthoracic or transhiatal esophagogastrectomy with D2 dissection. To watch for disease recurrence after resection, obtain complete history and physical examination every 4 mo for 1 yr, then every 6 mo for 1 yr, then yearly thereafter. Order CBC and comprehensive chemistry panel. If new symptoms arise, consider diagnostic imaging (e.g., CT or endoscopy). Figure 3 Algorithm illustrates workup and treatment of patient with gastric carcinoma. size larger than 1 cm. In one study comparing preoperative find- ume ascites. If cytologic study of the ascitic fluid so obtained con- ings from EUS with pathologic findings at operation, EUS was firms the presence of malignant cells, the patient is considered to 100% sensitive for N0 disease and 66.7% sensitive for N1 dis- have metastatic disease and therefore is not eligible for curative- ease.34 EUS also allows identification and aspiration of small-vol- intent surgery. For all of these reasons, EUS is now widely accept-
  • 5. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 5 ed as superior to conventional CT in the regional staging of gas- an R0 resection.41 Given the propensity of tumor for submucosal tric cancer.9 spreading, many authors consider proximal margins of 5 to 6 cm, with routine frozen-section analysis, to be optimal.42,43 Role of Laparoscopy In an effort to lower the positive margin rate, some surgeons The ultimate goal of any staging evaluation is to ensure that have proposed that total gastrectomy be considered the operation patients with metastatic disease are not treated with nontherapeu- of choice for all operable gastric cancers. This approach, original- tic laparotomy or other local therapies (e.g., radiation therapy), ly based on historical data from single institutions, has been test- which are generally ineffective against advanced disease. Even ed in three clinical trials. In the first trial, elective total gastrecto- small-volume metastatic disease identified on the surface of the my was compared with subtotal gastrectomy as curative-intent liver or the peritoneum at laparotomy is associated with poor sur- therapy for adenocarcinoma of the antrum.44 Elective total gas- vival: in one study, patients with such disease had a life expectan- trectomy did not increase mortality, but it also did not improve 5- cy of only 6 to 9 months.35 In these situations, there is little to be year survival (which was 48% in both treatment arms). In the sec- gained from attempts at palliative resection. ond trial, patients with antral cancer were randomly assigned to Staging laparoscopy [see 5:20 Gastric and Duodenal Dis.] has undergo either subtotal gastrectomy or total gastrectomy with proved to be highly relevant to the evaluation of patients with gas- extended lymph node dissection (ELND) and en bloc distal pan- tric cancer. In a study from the Memorial Sloan-Kettering Cancer createctomy and splenectomy.45 Total gastrectomy was associated Center (MSKCC), the investigators performed laparoscopic explor- with increased operative time, greater transfusion requirements, ation on 110 of 111 patients with newly diagnosed gastric cancer.36 and longer hospital stay; however, median survival was signifi- Of these 110 patients, 94% were accurately staged, with a sensitiv- cantly better in the subtotal gastrectomy group (1,511 days versus ity of 84% and a specificity of 100%, and 37% were found to have 922 days). In the third trial, the investigators concluded that sub- subclinical metastatic disease. Hospital stay was substantially total gastrectomy should be the procedure of choice for cancer of shorter in the 24 patients who underwent diagnostic laparoscopy the distal half of the stomach, provided that an adequate negative with biopsy only (average, 1.4 days) than in comparable patients who proximal margin could be achieved.46 This conclusion was based underwent exploratory laparotomy without resection (average, 6.5 on their finding that 5-year survival probabilities were essentially days). Finally, at the time the data were reported, none of the equivalent in the two groups studied (65.3% in the subtotal gas- patients who underwent laparoscopy had required palliative sur- trectomy group versus 62.4% in the total gastrectomy group). gery. Subsequent single-institution series confirmed the utility of staging laparoscopy, reporting accuracy rates ranging from 95% to Options for proximal gastric cancer As noted (see above), 97% and occult M1 disease rates approaching 30%.37,38 Taken as a adenocarcinoma of the gastric cardia and the esophagogastric whole, the data, though derived from relatively small single-institu- junction appears to be clinically distinct from adenocarcinoma of tion experiences, are compelling, and they have led the NCCN to the distal stomach,47 and its incidence is currently escalating across encourage laparoscopic staging strongly, either before or at the all races and age groups. Accordingly, it is imperative that sur- time of the planned resection.39 geons understand the surgical options for treatment of proximal gastric cancer.48 MANAGEMENT For tumors originating from the distal esophagus, esophagec- tomy—either transhiatal esophagectomy with a cervical anasto- Surgical Therapy mosis or transthoracic (Ivor-Lewis) esophagectomy with a tho- Surgical resection [see 5:20 Gastric and Duodenal Dis.] remains racic anastomosis—is clearly the procedure of choice [see 4:7 Open the only potentially curative therapy for localized gastric cancer Esophageal Procedures]. For tumors of the cardia, it has been sug- [see Figure 3]. Cure requires removal of all gross and microscopic gested that esophagogastrectomy might offer a survival advantage disease. More specifically, a margin-negative (R0) resection entails over total gastrectomy with an esophagojejunal anastomosis. This wide local excision of the primary tumor with en bloc removal of suggestion was evaluated in a study of 1,002 patients with adeno- all associated lymphatic vessels and any local or regional extension carcinoma of the esophagogastric junction.49 The investigators of disease.The downside of surgical resection as a sole modality of divided tumors into three types on the basis of the location of the therapy is that it is associated with a high rate of relapse. Conse- tumor center—cancers of the distal esophagus (type I), cancers of quently, several areas of surgical treatment of stomach cancer the cardia (type II), and cancers of the subcardial fundus (type remain subject to controversy. In particular, the extent of gastric re- III)—and analyzed the demographic and long-term survival data. section, the extent of lymph node dissection, the optimal approach Operative mortality proved to be higher with esophagogastrecto- to proximal stomach lesions, and the role of splenectomy and my than with extended total gastrectomy. Furthermore, R0 resec- adjacent organ resection continue to generate significant debate. tion and lymph node status were found to be the dominant prog- nostic factors influencing survival. Finally, in patients with type II Extent of gastric resection R0 resection (i.e., resection of all lesions, the pattern of lymphatic spread was primarily to paracar- gross disease with microscopically negative margins) has been dial, lesser curvature, and left gastric node groups. These data, shown to have a clear impact on overall survival after potentially taken together, led the authors to conclude that total gastrectomy curative surgery. In the German Gastric Cancer Study, a prospec- is preferable to esophagogastrectomy in this setting if a margin- tive multicenter observational trial, the calculated 10-year survival negative resection can be achieved. rate in the entire population was 26.3%, compared with 36.1% in An alternative approach to treating proximal gastric cancer is to patients who underwent an R0 resection.40 In a large multi-institu- perform a proximal subtotal gastrectomy. To date, no prospective tional adjuvant therapy trial, 19% of patients underwent an R1 studies have compared this method with total gastrectomy or resection (i.e., had resection-line involvement); only 9% of patients transhiatal esophagogastrectomy for esophagogastric junction tu- with stage I, II, or III disease and resection-line involvement sur- mors, but surgeons from MSKCC have published their retrospec- vived beyond 5 years, compared with 27% of those who underwent tive experience with 98 patients who underwent either total gas-
  • 6. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 6 trectomy or proximal subtotal gastrectomy for proximal gastric ed metastases underwent either total gastrectomy or, if 5 cm prox- cancer over a 10-year period.50 There were no significant differ- imal margins could be obtained, distal gastrectomy. In this study, ences between the groups with respect to morbidity, mortality, or a D2 lymph node dissection entailed distal pancreatectomy and 5-year survival. It remains to be seen whether such excellent splenectomy. Both morbidity and mortality were significantly high- results can be achieved at other centers. er in the D2 group than in the D1 group, and D2 dissection con- Thus, the evidence at present does not support routine perfor- ferred no demonstrable survival advantage at a median follow-up mance of total gastrectomy for lesions of the distal fundus or of 72 months. antrum, provided that histologically negative margins are achiev- In a trial from the Medical Research Council in the United able without compromise of the gastric inlet. Our current practice Kingdom, 400 patients with stage I to IIB disease were randomly is to perform a subtotal gastrectomy with Billroth II reconstruc- assigned to undergo either a D1 or a D2 lymph node dissection.58 tion for tumors of the distal stomach, a total gastrectomy with There was no significant difference in overall 5-year survival Roux-en-Y esophagojejunostomy for most cancers of the fundus between the two arms, but multivariate analysis demonstrated that and the proximal stomach [see 5:20 Gastric and Duodenal Dis.], clinical stages II and III, advanced age, male sex, and removal of and either a transthoracic esophagogastrectomy or a transhiatal the pancreas and the spleen were independently associated with esophagogastrectomy with gastric interposition for tumors of the poor outcome. The authors concluded that the classic Japanese esophagogastric junction and the cardia [see 4:7 Open Esophageal D2 dissection offered no survival advantage over D1 dissection. Procedures]. However, they hypothesized that D2 dissection with preservation of the distal pancreas and the spleen might lead to decreased mor- Extent of lymph node dissection Over the past decade, few bidity and mortality within the extended resection group and thus topics in the surgical literature have generated more debate than potentially to superior outcomes. the optimal extent of regional lymphadenectomy for gastric can- Further support for this hypothesis was provided by two non- cer. In Japan, where radical surgery for gastric cancer is now uni- randomized trials from specialized centers. The Italian Gastric versally accepted, the JRSGC has codified the extent of lymphat- Cancer Study Group (IGCSG) completed a phase II multicenter ic dissection according to the level of nodes dissected.51 A D1 trial designed to evaluate the safety and efficacy of pancreas-pre- lymph node dissection involves resection of the perigastric lymph serving D2 lymph node dissection.59 Quality control measures nodes along the greater and lesser curvature of the stomach. A D0 included supervision by a surgeon who had studied the technique dissection is anything less than a D1 dissection. A D2 dissection of D2 lymph node dissection at the National Cancer Center entails resection of the D1 nodes along with nodes along the com- Hospital in Tokyo. At a median follow-up time of 4.38 years, the mon hepatic artery, the left gastric artery, the celiac axis, and the overall morbidity rate for D2 dissection in the 191 patients enrolled splenic artery. A D3 lymph node dissection adds resection of was 20.9%, and the in-hospital mortality was 3.1%. The 5-year nodes in the hepatoduodenal ligament and the root of the mesen- survival rate for eligible patients was 55%. In a prospective series tery. Finally, a D4 resection calls for a D3 dissection plus resection of 125 patients undergoing standardized D2 lymph node dissec- of the retroperitoneal para-aortic and paracolic lymph nodes.52 tion at a single Western center, the investigators reported a mor- The JRSGC defines a curative operation as a gastric resection that tality of 1.37% and an overall morbidity of 33.5%.60 As in the includes lymph nodes one level beyond the level of pathologic IGCSG study, distal pancreatectomy was avoided in all cases, nodal involvement.Thus, in Japan, a D2 lymph node dissection is except when direct extension was suspected on the basis of macro- considered the standard resection for even relatively early cancers, scopic findings (5.5% of cases). Overall 5- and 10-year survival and numerous studies have cited the benefits of D3 and even D4 rates for this highly selected cohort were 52.3% and 40%, respec- lymphadenectomy for advanced carcinoma.53-55 tively. These studies suggest D2 lymph node dissection may be Western surgeons have been reluctant to embrace radical lym- safely performed in Western centers, when accompanied by care- phadenectomy, arguing that it has yet to demonstrate an unequiv- ful selection of patients, strict standardization of technique, and a ocal survival advantage in any prospective, randomized trial from strategy of pancreatic preservation. a Western institution or cooperative group. Detractors further Current AJCC guidelines state that pathologic examination of argue that the survival advantage associated with more radical at least 15 lymph nodes is required for adequate staging.28 In an procedures simply reflects stage migration, a higher incidence of effort to confirm the benefit of this staging system, investigators early gastric cancers, and differences in tumor biology and body from MSKCC reviewed their experience with 1,038 patients who habitus between Japanese and Western populations, and they point underwent R0 resection for gastric cancer.61 The location of posi- to the increases in operating time and morbidity that often accom- tive lymph nodes (within 3 cm of the primary tumor versus more pany extended gastric resections. One retrospective review of the than 3 cm away) did not significantly affect median survival; how- tumor registries of over 2,000 hospitals in the United States found ever, the number of positive lymph nodes had a profound effect on that D2 lymph node dissection had no survival advantage over D1 survival. Furthermore, in cases in which at least 15 nodes were lymph node dissection in terms of either the median survival time examined (27% of the total), the median survival for patients with or the 5-year survival rate.56 N1 (metastasis in one to six regional lymph nodes), N2 (metasta- Two prospective trials from Western Europe examined this issue sis in seven to 15 regional lymph nodes), and N3 disease (metas- further in an effort to evaluate the safety and efficacy of ELND. In tasis in more than 15 regional lymph nodes) was significantly the Dutch Gastric Cancer Group trial, 711 patients were ran- longer than the median survival reported in cases in which 14 or domly assigned to undergo either D1 or D2 lymphadenectomy as fewer nodes were resected with the specimens. These findings are part of a potentially curative gastrectomy for biopsy-proven ade- consistent with published data from our own institution (North- nocarcinoma.57 This trial was unique in its use of extensive quali- western University Feinberg School of Medicine), which indicate ty control measures, which included instruction and operative that the number of positive lymph nodes is a highly significant pre- supervision by an expert gastric cancer surgeon from Japan (who dictor of survival.62 In our series of 110 patients, those with N2 or also assisted with the processing and pathologic examination of N3 disease (seven or more positive lymph nodes) had a median the surgical specimens). Patients without evidence of disseminat- disease-free survival (DFS) of 17.6 months, whereas those with
  • 7. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 7 Table 3 Survival after Curative Resection of Gastric Cancer According to AJCC Lymph Node Status Roder et al1 Kodera et al105 Karpeh et al61 Lymph Node Status Cases (% of 5-Year Cases (% of 5-Year Cases (% of 5-Year Total) Survival (%) Total) Survival (%) Total) Survival (%) N1 (1–6 positive nodes) 258 (54) 45 306 (62) 70 392 (62) 38 N2 (7–15 positive nodes) 137 (29) 30 94 (19) 39 178 (28) 12 N3 (> 15 positive nodes) 82 (17) 10 93 (19) 24 65 (10) 5 N0 or N1 disease (six or fewer positive nodes) had a median DFS routine splenectomy does not increase survival, and it should be of 44 months. Data from other centers support this view as well reserved for situations in which the gastric tumor directly invades [see Table 3]. the splenic hilum or there is evidence of gross nodal metastases It is our current practice to perform a D2 lymph node dissec- along the splenic artery. tion, with resection of all perigastric lymph nodes along the greater and lesser curvatures of the stomach, as well as those along Nonsurgical Therapy the common hepatic artery, the left gastric artery, the celiac axis, Adjuvant therapy As noted (see above), the majority of and the splenic artery [see 5:20 Gastric and Duodenal Dis.]. We patients who present with gastric carcinoma and undergo poten- make every attempt to preserve the tail of the pancreas and spleen, tially curative surgical treatment will experience locoregional fail- with multivisceral resection reserved for cases of overt direct ex- ure, distant metastasis, or both and will succumb to their disease. tension of malignant disease in the absence of disseminated me- Accordingly, numerous adjuvant approaches—including chemo- tastasis.This strategy should provide adequate staging in terms of therapy, radiotherapy, chemoradiation, immunochemotherapy, and the AJCC guidelines, minimize morbidity, and possibly confer a intraperitoneal chemotherapy—have been tried in gastric cancer survival advantage on certain patient subgroups, as suggested by patients with the aim of improving overall survival and DFS. The the results of the trials mentioned. results, for the most part, have been disappointing. Results from prospective, randomized, controlled trials of adju- Role of splenectomy Routine splenectomy has been pro- vant radiation therapy in this setting have failed to establish a sur- posed as a means of facilitating clearance of metastatic nodes vival benefit. In a multi-institutional trial from 1994, patients were along the splenic artery and in the splenic hilum, but there is little randomly assigned to undergo surgery alone, surgery plus adju- evidence to support this practice in the treatment of proximal gas- vant radiation, or surgery plus adjuvant multiagent chemothera- tric cancers. Indeed, numerous studies have documented the dele- py.66 There was no significant benefit to either adjuvant regimen: terious effect of splenectomy when it is performed as part of an overall 5-year survival was 20% for surgery alone, compared with extended gastric resection. 12% for surgery plus radiation therapy and 19% for surgery plus In a retrospective study of 392 patients who underwent curative chemotherapy. gastrectomy at a high-volume cancer center, the impact of splen- Results from trials of chemotherapy alone have been equally un- ectomy on survival and postoperative morbidity was evaluated.63 satisfactory. Because of the established inefficacy of single-agent Splenectomy was not predictive of death on multivariate analysis, 5-fluorouracil (5-FU) therapy, combination chemotherapy regi- but complications were far more frequent in patients who under- mens have been employed. Such regimens have included nitro- went splenectomy as part of surgical treatment than in those who sourea compounds, mitomycin-C, anthracyclines, and members did not (45% versus 21%). Specifically, the incidence of infectious of the cisplatin family.23 In a meta-analysis of 13 trials comparing complications was far higher in the splenectomy group than in the adjuvant chemotherapy with observation in non-Asian countries, nonsplenectomy group (75% versus 47%). the odds ratio for death in the treated group was 0.8, correspond- In a review of data from an American College of Surgeons ing to a relative risk of 0.94.67 This result did not, however, reflect Pattern of Care Study, the investigators reported that the opera- a statistically significant improvement. Most oncologists have now tive mortality was 9.8% in patients who underwent splenectomy abandoned the use of chemotherapy by itself in the adjuvant setting. during gastric resection, compared with 8.6% for those who did In an effort to derive greater therapeutic benefit than can be not.64 More significantly, the 5-year observed survival rate was achieved with either radiation therapy or chemotherapy alone, 20.9% in the splenectomy group, compared with 31% in the non- combinations of the two have been used in the adjuvant setting. In splenectomy group. Intergroup Trial 0116, 556 patients who had undergone R0 resec- In a randomized, prospective trial, early and late results of total tion of adenocarcinoma of the stomach or the esophagogastric gastrectomy alone were compared with those of total gastrectomy junction were randomly assigned to treatment with either surgery plus splenectomy in patients being treated for cancers of the upper alone or surgery plus postoperative chemoradiotherapy.68 Patients third of the stomach.65 All patients underwent a D2 lymph node with tumors ranging from stage IB to stage IVM0 were included; dissection. The operative mortalities and the 5-year survival rates the majority had T3 tumors and node-positive disease.The thera- were similar in the two groups, but the splenectomy group had peutic regimen consisted of 5-FU and leucovorin administered con- more infectious complications. Specifically, the splenectomy group comitantly with 45 Gy of external-beam irradiation over a period had higher incidences of pulmonary complications, postoperative of 5 weeks. Median overall survival in the surgery-only group was fever higher than 38° C (100° F), and subphrenic abscess forma- 27 months, compared with 36 months in the surgery-chemoradi- tion. We agree with the conclusions of the authors of this study: ation group. In addition, the 3-year survival rate was 41% in the
  • 8. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 8 surgery-only group, compared with 50% in the surgery-chemora- gical resection with D2 lymphadenectomy; four patients (17%) diation group.The hazard ratio for death in the surgery-only group had progressive disease and did not undergo resection. Morbidi- as compared with the surgery-chemoradiation group was 1.35. ty and death rates were acceptable (32% and 5%, respectively), In the United States, the results of Intergroup Trial 0116 have and 11% of patients exhibited complete pathologic responses. led to the acceptance of chemoradiotherapy as standard adjuvant Overall, 63% of patients showed pathologic evidence of a signif- therapy for patients who have undergone curative-intent resection icant treatment effect. of gastric cancer. Nonetheless, numerous criticisms of this trial Newer neoadjuvant treatment strategies employ multiagent in- have been expressed. Specifically, a review of the operative and duction chemotherapy followed by chemoradiotherapy and planned pathology reports of 453 of the patients revealed a lack of surgical gastric resection in patients with locally advanced but potentially standardization.69 When the extent of lymphadenectomy was cat- resectable gastric cancer.73 egorized, the majority (54.2%) of the patients were found to have FOLLOW-UP AND MANAGEMENT OF RECURRENT DISEASE undergone a D0 dissection; 38.1% underwent a D1 dissection, and only 7.5% underwent a D2 or D3 dissection. These findings Even after gross resection of all disease with microscopically suggest that the main effect of the chemoradiation therapy may negative margins (R0 resection), recurrence of gastric carcinoma have been simply to compensate for inadequate surgery.This sug- is common. Adenocarcinoma of the stomach may spread through gestion is supported by the observation that the number of patients direct extension, via lymphatic channels to regional and distant with local and regional recurrences was higher in the surgery-only lymph nodes, or via the bloodstream to distant sites. Further- group (178 versus 101), whereas the number of patients with dis- more, once tumors have penetrated the serosa (T3), peritoneal tant failure was slightly higher in the adjuvant-therapy arm (40 metastasis becomes a possibility. Through autopsy series and versus 32). Furthermore, when the Maruyama Index of Un- clinical studies, certain definite patterns of locoregional failure resected Disease (a computer model developed for accurate pre- and distant metastasis have been established. Locoregional diction of nodal station involvement in gastric cancer) was applied recurrences are common in the gastric bed and the adjacent to the 556 patients eligible for the Intergroup Trial, the median lymph nodes. Clinical and reoperative evaluation have docu- Maruyama Index was 70.70 This value was far above the level con- mented recurrent disease at the anastomosis, in the retroperi- sidered to represent optimal surgical therapy (i.e., Maruyama toneum, or in the regional lymph nodes in 3% to 69% of Index < 5) and led the authors to conclude that the vast majority patients; the incidence of recurrence may vary, depending on of patients in the trial had been surgically undertreated. whether the patients had received adjuvant therapy.23 One autopsy Currently, physicians, especially in Europe, generally eschew series documented a locoregional recurrence rate of 94% in pa- adjuvant therapy after R0 resection of gastric cancer, except under tients treated with surgery alone. The peritoneum is ultimately the auspices of a clinical trial.9 The Radiation Therapy Oncology involved in 17% to 50% of all patients. The most common sites Group has initiated a phase II trial of adjuvant chemoradiothera- of visceral metastases are the liver and the lungs. py using 45 Gy of external beam radiation with cisplatin and pax- In view of the high recurrence rates, all patients who have itaxel, with or without 5-FU. If promising results are found, a undergone resection should be seen for routine surveillance exam- phase III trial will follow. It is to be hoped that ongoing trials will inations. Currently, the NCCN recommends that a complete his- shed further light on this complex management issue. tory and physical examination be conducted every 4 months for 1 year, then every 6 months for 1 year, and then yearly thereafter.39 Neoadjuvant therapy As a response to the disappointing A complete blood count, serum electrolyte concentrations, and results of adjuvant therapy and the inability of many patients to liver function studies should also be considered. Imaging studies regain adequate performance status after radical gastric surgery, (e.g., CT and endoscopy) are ordered as indicated, usually in neoadjuvant therapy protocols have been proposed.71 The theo- response to new symptoms. In addition, long-term vitamin B12 retical benefits of a neoadjuvant treatment strategy include treat- supplementation should be initiated for patients who have under- ment-induced tumor downstaging, which may enhance resectabil- gone a proximal or subtotal gastrectomy. ity, and early administration of systemic therapy, which allows almost all patients to receive and complete the prescribed treat- ment. Furthermore, because treatment is administered when mea- Other Gastric Malignancies surable disease is present, response to therapy may be assessed and GASTRIC LYMPHOMA continued only in patients who are likely to benefit. Finally, patients who are found to have rapidly progressive disease during preoper- Gastric lymphoma is the second most common malignancy of ative chemotherapy may be spared having to undergo a nonthera- the stomach, accounting for 2% to 9% of gastric tumors in the peutic gastrectomy.72 United States. Lymphomas of the stomach are of the non-Hodgkin In a report of three phase II trials from the M. D. Anderson type.The stomach is the most common site of extranodal involve- Cancer Center, encompassing 83 patients who received neoadjuvant ment of non-Hodgkin lymphoma (NHL) and accounts for nearly chemotherapy before planned surgical resection, clinical response 50% of all such cases.74 rates ranged from 24% to 38%, with three patients (4%) exhibit- ing a complete pathologic response.72 Sixty-one patients (73%) Clinical Evaluation were able to undergo a curative-intent resection, and the response The presenting symptoms of gastric lymphoma, like those of to chemotherapy was the only significant predictor of survival on gastric adenocarcinoma, are nonspecific and include loss of ap- multivariate analysis. petite, weight loss, vomiting, and bleeding. Overt clinical symp- Preoperative chemoradiation therapy has also been shown to toms (e.g., fever and night sweats) are relatively rare: in one mul- be feasible in phase II trials. In a 2001 trial that included 23 ticenter trial concerned with primary gastric lymphoma, they patients, 96% of the study population received combined-modal- occurred in fewer than 12% of patients enrolled.75 Risk factors for ity therapy.71 Nineteen patients (83%) were able to undergo sur- gastric lymphoma include H. pylori infection, immunosuppression
  • 9. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 9 after solid-organ transplantation, celiac disease, inflammatory bowel bleeding or perforation. In the German Multicenter Study Group disease, and HIV infection.76 trial, 185 patients with stage I or II gastric lymphoma were treat- ed either with gastrectomy followed by radiation or (in the case of Investigative Studies high-grade lesions) chemotherapy plus radiation or with chemo- The diagnosis of gastric lymphoma is most frequently estab- therapy and radiotherapy alone.75 There was no significant differ- lished by means of endoscopy with biopsy. Staging studies in- ence in survival between the group receiving surgical treatment clude a comprehensive blood count, a lactate dehydrogenase and the group receiving nonoperative therapy: overall 5-year sur- (LDH) level, and a comprehensive chemistry panel; CT of the vival rates were 82.5% and 84%, respectively. There were no per- chest, the abdomen, and the pelvis; and, often, a bone marrow forations, and there was only one hemorrhage (in a patient treat- biopsy. All pathology slides should be reviewed by an experienced ed with chemotherapy alone). Similarly, in a single-institution, hematopathologist.77 prospective, randomized trial comparing chemotherapy alone with chemotherapy plus surgery for stage I and II lymphoma, Staging and Prognosis there were no instances of perforation and only three instances of Numerous staging systems have been employed to stage NHL GI bleeding in the chemotherapy group, compared with two of the GI tract. Of these, the one most commonly applied is a bleeding episodes in the surgery plus chemotherapy group.79 modification of the Ann Arbor staging system for lymphoma.76 Currently, patients with early-stage high-grade gastric lym- For surgeons, the most important determination is often whether phomas are treated primarily with chemotherapy or radiation the NHL (1) is confined to the stomach and the perigastric nodes therapy; only rarely do they require surgical intervention for com- (stage I and II disease), (2) involves other intra-abdominal nodes plications encountered during therapy. Patients with locally ad- and organs (stage III), or (3) extends outside the abdomen vanced (stage III) or disseminated (stage IV) gastric lymphoma are (stage IV).78 clearly best treated with chemotherapy, with or without radiation. Occasionally, surgery is indicated in such patients to treat residual Management disease confined to the stomach or to palliate bleeding or obstruc- Over the past decade, the management of patients with gastric tion that does not resolve with nonoperative therapy. Primary sur- lymphoma has undergone significant changes. Generally, there has gical therapy is to be avoided in these patients because of the sig- been a shift away from surgical management, even in relatively nificant risk of complications and the delay in initiating systemic localized cases (stages I and II).79 This shift is the result not only of therapy. the documented success of chemotherapy alone for more ad- GASTROINTESTINAL STROMAL TUMOR vanced cases (stages III and IV) but also of a better understanding of the etiology of gastric lymphoma.80 Approximately 45% of all Gastrointestinal stromal tumor (GIST), though relatively rare gastric lymphomas are low-grade mucosa-associated lymphoid tis- in absolute terms, is the most common sarcoma of the GI tract,85 sue (MALT) lymphomas.75 The gastric mucosa is normally devoid with approximately 6,000 cases reported each year in the United of lymphoid tissue. It is hypothesized that MALT develops in the States alone. The stomach is the most common site of involve- stomach in response to chronic H. pylori infection.81 ment, accounting for 60% to 70% of cases86; the small intestine (25%), the rectum (5%), the esophagus (2%), and a variety of Nonsurgical therapy Low-grade MALT lymphoma usually other locations account for the remainder. On the basis of their presents as stage I or II disease and has an indolent course. Since appearance on light microscopy, GISTs were once thought to be 1993, when regression of low-grade MALT lymphoma after erad- of smooth muscle origin, and most were classified as leiomyosar- ication of H. pylori was first reported, numerous trials have docu- comas.87 Thus, extended gastric resection, often including con- mented the efficacy of anti–H. pylori therapy, with complete remis- tiguous organs, was advised. Recurrence developed after R0 resec- sion rates ranging from 50% to 100%.79 In the German MALT tion in approximately 50% of cases.88 With the advent of immuno- Lymphoma Study, the complete remission rate was 81%; 9% of histochemistry and electron microscopy, it became clear that patients exhibited partial responses, and 10% showed no response.82 GIST has both smooth muscle and neural elements, and the cell Low-grade lymphomas that are more advanced or do not regress of origin is now believed to be the interstitial cell of Cajal, an with antibiotic therapy may be treated with H. pylori eradication intestinal pacemaker cell.89 The diagnosis of GIST is secured by and radiation (with or without chemotherapy).83 For localized per- immunohistochemical staining for the tyrosine kinase receptor sistent disease, modest doses of radiation, on the order of 30 Gy, KIT (CD117), which highlights the presence of interstitial cells of may be employed.When chemotherapy is required, multiagent re- Cajal. More than 95% of GISTs exhibit unequivocal staining for gimens, such as cyclophosphamide-vincristine-prednisolone (COP), KIT.86 Approximately two thirds of GISTs also express CD34. are often used. Histologically, these tumors may exhibit a spindle cell pattern, an Approximately 55% of gastric lymphomas are high-grade le- epithelioid pattern, or a mixed subtype. sions, which can occur with or without a low-grade MALT com- ponent.75 These lymphomas are treated with chemotherapy and Clinical Evaluation radiation therapy according to the extent of the disease.The cyclo- The median age of incidence is 63 years, and tumors are gen- phosphamide-doxorubicin-vincristine-prednisolone (CHOP) regi- erally between 0.5 and 44 cm in diameter at the time of diagnosis men is the one most frequently employed. In some studies, the (median diameter, 6 cm).86 Mass-related symptoms (e.g., abdom- anti-CD20 monoclonal antibody rituximab has been either added inal pain, bloating, and early satiety) may be present. Melena or to standard therapy or used alone, with encouraging results.84 anemia from overlying mucosal ulceration may be present as well. A small subset of patients have peritonitis as a consequence of Surgical therapy Surgical resection, once thought to be tumor rupture and subsequent hemorrhage. Finally, many GISTs essential for the diagnosis, staging, and treatment of early-stage are discovered incidentally during operation, abdominal imaging, gastric lymphoma, now is used mainly in patients who experience or endoscopy.
  • 10. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 10 Investigative Studies GASTRIC CARCINOID When a GIST is suspected, abdominal and pelvic imaging with Gastric carcinoid tumors are rare, accounting for fewer than either CT or MRI is indicated. Chest imaging is performed as 11% to 30% of all GI carcinoids and fewer than 1% of all gastric well. Endoscopy, with or without EUS, may occasionally help with tumors.92 The median age at diagnosis is 62, and tumors are surgical planning, but because of the infrequency of mucosal in- equally distributed between men and women. volvement, it is rarely diagnostic.90 Surgical consultation should be Clinical Evaluation and Investigative Studies obtained to determine whether the lesion can be resected. If the tumor is resectable, biopsy should not be performed, because of Gastric carcinoid tumors are often discovered during endo- the risk of tumor rupture and intra-abdominal dissemination. Biop- scopic examination of patients experiencing chronic abdominal sy may be required, however, if the patient has widespread disease pain; patients may also complain of vomiting and diarrhea. These or may be enrolling in a trial of neoadjuvant therapy. In such cases, tumors are rarely associated with symptoms of the carcinoid syn- biopsy may be performed percutaneously or at the time of EUS. drome. Diagnosis is usually confirmed by endoscopic biopsy, and EUS is helpful in determining the extent of gastric wall penetra- Staging and Prognosis tion and the degree of regional lymph node involvement. Although the majority of gastric GISTs have a benign course, a Gastric carcinoid tumors have been divided into three types, wide spectrum of biologic behavior has been observed. Of the primarily on the basis of their association (or lack thereof) with prognostic factors examined to date, tumor size and mitotic rate hypergastrinemia.Type I tumors are associated with chronic atro- appear to be the most valuable. If the tumor is less than 2 cm in phic gastritis, are generally small (< 1 cm), and are often multiple diameter and the mitotic count is lower than five per high-power and polypoid. They grow slowly and only rarely metastasize to field (HPF), the risk of an aggressive disease course is considered regional nodes or distant sites. Type II tumors are associated with to be very low. Conversely, if the tumor is larger than 10 cm, if the the Zollinger-Ellison syndrome and multiple endocrine neoplasia mitotic count is higher than 10/HPF, or if the tumor is larger than type I (MEN I) and, like type I tumors, are usually small and mul- 5 cm with a mitotic count higher than 5/HPF, the risk of aggres- tiple. They also grow slowly, but they are more likely to metasta- sive clinical behavior is considered to be high. For all other tumors, size than type I gastric carcinoids are. Type III (sporadic) gastric the risk of aggressive disease is considered to be intermediate.86 carcinoid tumors are the most biologically aggressive type. They are often large (> 1 cm) at the time of diagnosis and are not asso- Management ciated with hypergastrinemia. Type III lesions frequently metasta- Surgical therapy The role of surgery in the treatment of a size to regional nodes (54%) or the liver (24%).92 GIST is to resect the tumor with grossly negative margins and an Management intact pseudocapsule. Lymph node involvement is rare with For patients with small, solitary type I tumors, endoscopic GISTs, and thus, no effort is made to perform ELND.The tumor polypectomy [see 5:18 Gastrointestinal Endoscopy] or open resection must be handled with care to prevent intra-abdominal rupture. via gastrotomy (local excision) [see 5:20 Gastric and Duodenal Dis.] Formal gastric resection is rarely required: as a rule, it is indicated is the procedure of choice. For patients with multiple or recurrent only for lesions in close proximity to the pylorus or the esopha- tumors, antrectomy [see 5:20 Gastric and Duodenal Dis.] is indicat- gogastric junction. ed to remove the source of the hypergastrinemia. For patients with type II lesions, treatment is similar to that for patients with type I Nonsurgical therapy If the tumor has metastasized or has lesions, with the extent of gastric resection determined by the size advanced locally to the point where surgical therapy would result and number of lesions. For patients with type III lesions, however, in excessive morbidity, the patient is treated with the tyrosine kinase either distal or total gastrectomy with ELND is required.93 All inhibitor imatinib mesylate. Imatinib is a selective inhibitor of a patients undergoing a less than total gastrectomy should be fol- family of protein kinases that includes the KIT -receptor tyrosine lowed with serial endoscopy at regular intervals.94 kinase, which is expressed in the majority of GISTs. Originally indicated for the treatment of chronic myelocytic leukemia, ima- tinib was approved for the treatment of KIT -positive GIST in Table 4 American Joint Committee of Cancer TNM 2002, when phase II clinical trials documented sustained objective Clinical Classification of Small Bowel Carcinoma responses in a majority of patients with advanced unresectable or metastatic GIST.91 Patients with borderline resectable lesions should TX Primary tumor cannot be assessed be treated with imatinib until they exhibit a maximal response as T0 No evidence of primary tumor documented by CT and positron emission tomography (PET); TisCarcinoma in situ surgery may then be undertaken to resect any residual foci of dis- T1 Tumor invades lamina propria or submucosa ease. Similarly, whereas patients with metastatic disease are unlike- T2 Tumor invades muscularis propria ly to manifest a complete response to imatinib therapy, they should Primary tumor (T) T3 Tumor penetrates < 2 cm into subserosa or into nonperitonealized perimuscular tissue (mesentery for be periodically reevaluated and considered for resection should sur- jejunum or ileum, retroperitoneum for duodenum) gical treatment become technically feasible.90 T4 Tumor penetrates visceral peritoneum or directly After an R0 resection of a GIST, no adjuvant therapy is indi- invades > 2 cm into adjacent structures cated unless the patient is participating in a clinical trial. The NX Regional lymph node(s) cannot be assessed American College of Surgeons Oncology Group is currently con- Regional lymph N0 No regional lymph node metastasis ducting two trials of imatinib in the postoperative setting. A phase nodes (N) N1 Regional lymph node metastasis II trial (Z9000) of imatinib, 400 mg/day, for patients with high-risk MX Distant metastasis cannot be assessed GIST, has reached accrual, and a phase III trial (Z9001) compar- Distant metastasis (M) M0 No distant metastasis ing 1 year of imatinib, 400 mg/day, with placebo in patients with M1 Distant metastasis intermediate-risk GIST is currently under way.
  • 11. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 11 Table 5 American Joint Committee on Cancer nosis is often made by means of esophagogastroduodenoscopy Staging System for Small Bowel Carcinoma (EGD). Lesions within the first 100 cm of the small bowel may be evaluated with push enteroscopy.When the adenocarcinoma is sit- uated elsewhere in the small bowel, it is localized with small bowel Stage T N M radiographs. Some authors consider enteroclysis to be superior to the more commonly used small bowel follow-through in this set- Stage 0 Tis N0 M0 ting, in that enteroclysis is better able to demonstrate fine mucosal Stage I T1, T2 N0 M0 detail.99 In experienced hands, enteroclysis may therefore be more Stage II T3, T4 N0 M0 sensitive.100 Some lesions are identified when CT or MRI is per- formed to evaluate complaints of abdominal pain. Furthermore, Stage III Any T N1 M0 abdominal imaging may yield complementary staging information Stage IV Any T Any N M1 (e.g., the presence of regional adenopathy or metastatic disease). One promising new method for the identification of small bowel tumors is wireless capsule endoscopy.101 This minimally invasive technique Small Bowel Malignancies may be particularly useful in identifying small lesions in the distal Malignant tumors of the small intestine are rare, accounting for jejunum and ileum that cannot be identified radiographically. fewer than 5% of all GI tract malignancies. In the United States, MANAGEMENT only a few thousand new cases of small bowel cancer are report- ed each year.95 The majority of small bowel malignancies are ade- Aggressive surgical resection remains the cornerstone of thera- nocarcinomas, lymphomas, or carcinoid tumors,96 though GISTs py for adenocarcinoma of the small intestine.102 For periampullary are being noted with increasing frequency in the small intestine. lesions, pancreaticoduodenectomy is typically required to achieve a Treatment of lymphomas, carcinoid tumors, and GISTs in the margin-negative resection. For lesions in the distal duodenum, a small bowel is nearly identical to treatment of the same lesions in segmental sleeve resection with a duodenojejunostomy is appro- the stomach [see Other Gastric Malignancies, above] and thus will priate. For lesions in the jejunum or the ileum, segmental resection not be covered further in this chapter. Our focus here is on the may be performed with a wide mesenteric resection to encompass presentation, diagnosis, and treatment of adenocarcinoma of the potentially involved regional lymph nodes. Contiguous organs are small bowel. Like gastric adenocarcinoma, small bowel adenocar- resected en bloc as necessary.98 cinoma is usually staged according to the AJCC/UICC TNM Because the presenting signs and symptoms are often vague and classification system [see Tables 4 and 5]. nonspecific, diagnosis is often delayed. In one series, only 6 (11%) of the 53 patients were suspected of having a small bowel tumor at CLINICAL EVALUATION admission.102 In a retrospective review of patients with small bowel Between 46% and 55% of small bowel adenocarcinomas occur tumors treated at our institution, the mean duration of symptoms in the duodenum.96,97 Patients frequently present with nausea, before surgical management was 110 months, and more than 50% vomiting, abdominal pain, weight loss, and GI bleeding98; occa- of the patients were found to have stage III or IV disease.98 sionally, they present with iron deficiency anemia or a positive fecal The 5-year survival rate continues to be low (24% to occult blood test result. In rare cases, small bowel obstruction, 37%).98,103,104 Significant predictors of good overall survival in- often with the tumor serving as a lead point for intussusception, is clude complete (R0) resection and low AJCC tumor stage.98,103,104 the first manifestation of the disease.97 The available evidence indicates that all patients with small bowel neoplasms should be offered an oncologically sound surgical resec- INVESTIGATIVE STUDIES tion. In one series, curative (R0) resection was accomplished in When an adenocarcinoma is located in the duodenum, the diag- 51% of cases.103 References 1. Roder DM: The epidemiology of gastric cancer. the changing pattern of esophago-gastric cancer: classification. Acta Pathol Micrbiol Scand 64:31, Gastric Cancer 5(1):5, 2002 data from a UK regional cancer registry. Cancer 1965 2. Neugut AI, Hayek M, Howe G: Epidemiology of Causes Control 12:943, 2001 13. Ogimoto I, Shibata A, Fukuda K: World Cancer gastric cancer. Semin Oncol 23:281, 1996 8. Wanebo HJ, Kennedy BJ, Chmiel J, et al: Cancer Research Fund/American Institute of Cancer 3. Jemal A,Tiwari RC, Murray T, et al: Cancer Sta- of the stomach: a patient care study by the Amer- Research 1997 recommendations: applicability tistics. CA Cancer J Clin 54:8, 2004 ican College of Surgeons. Ann Surg 218:579, 1993 to digestive tract cancer in Japan. Cancer Causes 9. Hohenberger P, Gretschel S: Gastric cancer. Control 11(1):9, 2000 4. Alberts SR, Cervantes A, van de Velde CJ: Gas- tric cancer: epidemiology, pathology and treat- Lancet 362:305, 2003 14. Plummer M, Franceschi S, Munoz N: Epide- ment. Ann Oncol 14(2):ii, 2003 10. Corvalan A, Koriyama C, Akiba S, et al: Epstein- miology of gastric cancer. IARC Sci Publ Barr virus in gastric carcinoma is associated with 157:311, 2004 5. Cancer Statistics Working Group. United States Cancer Statistics: 1999–2001 Incidence and Mor- location in the cardia and with a diffuse histol- 15. Hansson LE, Nyren O, Bergstrom R, et al: tality Web-based Report Version. Department of ogy: a study in one area of Chile. Int J Cancer Nutrients and gastric cancer risk: a population- Health and Human Services, Centers for Dis- 94:527, 2001 based cases-control study in Sweden. Int J Can- ease Control and Prevention, and National Can- 11. Fukayama M, Chong JM, Uozaki H: Pathology cer 57:638, 1994 cer Institute, Atlanta, 2004 and molecular pathology of Epstein-Barr virus- 16. Gastric cancer and Helicobacter pylori: a com- associated gastric carcinoma. Curr Top Micro- bined analysis of 12 case control studies nested 6. Lawrence W Jr, Menck HR, Steele GD Jr, et al: biol Immunol 258:91, 2001 within prospective cohorts. Helicobacter and Can- The National Cancer Data Base report on gas- 12. Lauren P: The two histological main types of cer Collaborative Group. Gut 49:347, 2001 tric cancer. Cancer 75:1734, 1995 gastric carcinoma: diffuse and so-called intesti- 17. Correa P: Bacterial infections as a cause of can- 7. Wayman J, Forman D, Griffin SM: Monitoring nal-type carcinoma: an attempt at a histological cer. JNCI 95(7):E3, 2003
  • 12. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 12 18. Peek RM Jr, Blaser MJ: Helicobacter pylori and results of the German Gastric Cancer Study. node staging in gastric cancer: is location more gastrointestinal tract adenocarcinomas. Nat Rev Ann Surg 228:449, 1998 important than number? An analysis of 1,038 Cancer 2:28, 2002 41. Hallissey MT, Jewkes AJ, Dunn JA, et al: Resec- patients. Ann Surg 232:362, 2000 19. La Vecchia C, Negri E, Franceschi S, et al: Fami- tion-line involvement in gastric cancer: a contin- 62. Talamonti MS, Kim SP, Yao KA, et al: Surgical ly history and the risk of stomach and colorectal uing problem. Br J Surg 80:1418, 1993 outcomes of patients with gastric carcinoma: the cancer. Cancer 70:50, 1992 42. Jakl RJ, Miholic J, Koller R, et al: Prognostic fac- importance of primary tumor location and 20. Fuchs CS, Mayer RJ: Gastric carcinoma. N Engl tors in adenocarcinoma of the cardia. Am J Surg microvessel invasion. Surgery 134:720, 2003 J Med 333:32, 1995 169:316, 1995 63. Brady MS, Rogatko A, Dent LL, et al: Effect of 21. Lin J, Beer DG: Molecular biology of upper gas- 43. Kooby DA, Coit DG: Controversies in the surgi- splenectomy on morbidity and survival following trointestinal malignancies. Sem Oncol 31:476, cal management of gastric cancer. J National curative gastrectomy for carcinoma. Arch Surg 2004 Comprehensive Cancer Network 1(1):115, 2003 126:359, 1991 22. Ebert MP, Malfertheiner P: Review article: 44. Gouzi JL, Huguier M, Fagniez PL, et al: Total 64. Wanebo HJ, Kennedy BJ, Winchester DP, et al: pathogenesis of sporadic and familial gastric can- versus subtotal gastrectomy for adenocarcinoma Role of splenectomy in gastric cancer surgery: cer—implications for clinical management and can- of the gastric antrum: a French prospective con- adverse effect of elective splenectomy on long- cer prevention. Aliment Pharmacol Ther 16:1059, trolled study. Ann Surg 209:162, 1989 term survival. J Am Coll Surg 185:177, 1997 2002 45. Robertson CS, Chung SC, Woods SD, et al: A 65. Csendes A, Burdiles P, Rojas J, et al: A prospec- 23. Karpeh MS, Kelsen DP, Tepper JE: Cancer of prospective randomized trial comparing R1 sub- tive randomized study comparing D2 total gas- the stomach. Cancer Principles and Practice of total gastrectomy with R3 total gastrectomy for trectomy versus D2 total gastrectomy plus Oncology, 6th ed. DeVita VT Jr, Hellman S, antral cancer. Ann Surg 220:176, 1994 splenectomy in 187 patients with gastric carcino- Rosenberg S, Eds. Lippincott Williams & ma. Surgery 131:401, 2002 46. Bozzetti F, Marubini E, Bonfanti G, et al: Sub- Wilkins, Philadelphia, 2001, p 1092 66. Hallissey MT, Dunn JA,Ward LC, et al: The sec- total versus total gastrectomy for gastric cancer: 24. Oohara T, Aono G, Ukawa S, et al: Clinical diag- ond British stomach cancer group trial of adju- five-year survival rates in a multicenter random- nosis of minute gastric cancer less than 5 mm in vant radiotherapy or chemotherapy in respectable ized Italian trial. Italian Gastrointestinal Tumor diameter. Cancer 53:162, 1984 gastric cancer: five-year follow-up. Lancet 343: Study Group. Ann Surg 230:170, 1999 1309, 1994 25. Grise K, McFadden D: Gastric cancer: three de- 47. Spechler SJ: The role of gastric carditis in meta- cades of surgical management. Am J Surg 64: 930, 67. Earle CC, Maroun JA: Adjuvant chemotherapy plasia and neoplasia at the gastroesophageal 1998 after curative resection for gastric cancer in non- junction. Gastroenterology 117:218, 1999 Asian patients: revisiting a meta-analysis of ran- 26. Yalmarthi S, Witherspoon P, McCole D, et al: 48. Devesa SS, Blot WJ, Fraumeni JF Jr: Changing domized trials. Eur J Cancer 35:1059, 1999 Missed diagnosis in patients with upper gastroin- patterns in the incidence of esophageal and gas- testinal cancers. Endoscopy 36:874, 2004 68. MacDonald JS, Smalley SR, Benedetti J, et al: tric carcinoma in the United States. Cancer 83: Chemoradiotherapy after surgery compared with 27. Nio Y, Tsubono M, Kawabata K, et al: Compar- 2049, 1998 surgery alone for adenocarcinoma of the stom- ison of survival curves of gastric cancer patients 49. Rudiger Siewert J, Feith M, Werner M, et al: ach or gastroesophageal junction. N Engl J Med after surgery according to the UICC stage classi- Adenocarcinoma of the esophagogastric junc- 345:725, 2001 fication and the general rules for gastric cancer tion: results of surgical therapy based on anatom- study by the Japanese Research Society for gas- 69. Estes NC, MacDonald JS, Touijer K, et al: In- ical/topographic classification in 1,002 consecu- tric cancer. Ann Surg 218:47, 1993 adequate documentation and resection for gas- tive patients. Ann Surg 232:353, 2000 tric cancer in the United States: a preliminary 28. American Joint Committee on Cancer: Stomach. 50. Harrison LE, Karpeh MS, Brennan MF: Total report. Am Surg 64:680, 1998 AJCC Cancer Staging Manual, 6th ed. Greene gastrectomy is not necessary for proximal gastric F, Page DL, Fleming ID, et al, Eds. Springer- 70. Hundahl SA, MacDonald JS, Bendetti J, et al: cancer. Surgery 123:127, 1998 Verlag, New York, 2002, p 99 Surgical treatment variation in a prospective, 51. Kajitani T: The general rules for the gastric can- randomized trial of chemoradiotherapy in gastric 29. Abdalla EK, Pisters PWT: Staging and preoper- cer study in surgery and pathology. Japan J Surg cancer: the effect of undertreatment. Ann Surg ative evaluation of upper gastrointestinal malig- 2:127, 1981 Oncol 9:278, 2002 nancies. Semin Oncol 31:513, 2004 52. Japanese Gastric Cancer Association: Japanese 71. Lowy AM, Feig BW, Janjan N, et al: A pilot study 30. Takao M, Fukuda T, Iwanaga S, et al: Gastric Classification of Gastric Adenocarcinoma—2nd of preoperative chemoradiotherapy for respec- cancer: evaluation of triphasic spiral CT and English edition. Gastric Cancer 1:10, 1998 table gastric cancer. Ann Surg Oncol 8:519, 2001 radiologic-pathologic correlation. J Comput 53. Maeta M, Yamashiro H, Saito H, et al: A pros- 72. Lowy AM, Mansfield PF, Leach SD, et al: Res- Assist Tomogr 22:288, 1998 pective pilot study of extended (D3) and super- ponse to neoadjuvant chemotherapy best predicts 31. Davies J, Chalmers AG, Sue-Ling HM, et al: extended para-aortic lymphadenectomty (D4) in survival after curative resection of gastric cancer. Spiral computed tomography and operative stag- patients with T3 or T4 gastric cancer managed by Ann Surg 229:303, 1999 ing of gastric carcinoma: a comparison with his- total gastrectomy. Surgery 125:325, 1999 topathological staging. Gut 41:314, 1997 73. Yao JC, Mansfield PF, Pisters PWT, et al: Com- 54. Baba M, Hokita S, Natsugoe S, et al: Paraaortic bined-modality therapy for gastric cancer. Sem 32. Pollack BJ, Chak A, Sivak MV Jr: Endoscopic lymphadenectomy in patients with advanced car- Surg Oncol 21:223, 2003 ultrasonography. Semin Oncol 23:336, 1996 cinoma of the upper-third of the stomach. 74. Gurney KA, Cartwright RA, Gilman EA: Des- 33. Messmann H, Schlottmann K: Role of endos- Hepatogastroenterology 47:893, 2000 criptive epidemiology of gastrointestinal non- copy in the staging of esophageal and gastric can- 55. Isozaki H, Okajima K, Fujii K, et al: Effective- Hodgkin’s lymphoma in a population-based reg- cer. Semin Surg Oncol 20:70, 2001 ness of paraaortic lymph node dissection for ad- istry. Br J Cancer 79:1929, 1999 34. De Manzoni G, Di Leo M, Bonfiglio P, et al: Ex- vanced gastric cancer. Hepatogastroenterology 75. Koch P, del Valle F, Berdel WE, et al: Primary perience of endoscopic ultrasound in staging 46:549, 2000 gastrointestinal non-Hodgkin’s lymphoma: II. adenocarcinoma of the cardia. Eur J Surg Oncol 56. Wanebo HJ, Kennedy BJ, Winchester DP, et al: Combined surgical and conservative or conserv- 25:595, 1999 Gastric carcinoma: does lymph node dissection ative management only in localized gastric lym- 35. Macdonald JS, Gohmann JJ: Chemotherapy of alter survival? J Am Coll Surg 183:616, 1996 phoma—results of the prospective German advanced gastric cancer: present status, future 57. Bonenkamp JJ, Hermans J, Sasako M, et al: Multicenter Study GIT NHL 01/92. J Clin Oncol prospects. Semin Oncol 15:42, 1988 Extended lymph node dissection for gastric can- 19:3874, 2001 36. Burke EC, Karpeh MS Jr, Conlon KC, et al: Lap- cer. N Engl J Med 340:908, 1999 76. Crump M, Gospodarowicz M, Shepherd FA: aroscopy in the management of gastric adenocar- 58. Cuscheri A, Weeden S, Fielding J, et al: Patient Lymphoma of the gastrointestinal tract. Semin cinoma. Ann Surg 225:262, 1997 survival after D1 and D2 resections for gastric Oncol 26:324, 1999 37. Feussner H, Omote K, Fink U, et al: Prethera- cancer: long term results of the MMC random- 77. Non-Hodgkin’s Lymphoma, Practice Guidelines peutic laparoscopic staging in advanced gastric ized surgical trial. Br J Cancer 70:1522, 1999 in Oncology, v.1.2005. National Comprehensive carcinoma. Endoscopy 31:342, 1999 59. Degiuli M, Sasako M, Ponyi A, et al: Survival Cancer Center Network results of a multicentre phase II study to evaluate 38. Charukhchyan SA, Lucas GW: Laparoscopy and lesser sac endoscopy in gastric carcinoma oper- D2 gastrectomy for gastric cancer. Br J Cancer 78. Talamonti MS: Gastric cancer. Cancer Surgery ability assessment. Am Surg 64:160, 1998 90:1727, 2004 for the General Surgeon. Winchester D, Jones 60. Roukos DH, Lorenz M, Encke A: Evidence of RS, Murphy GP, Eds. Lippincott Williams & 39. Gastric Cancer, Practice Guidelines in Oncolo- Wilkins, Philadelphia, 1998 gy, v.1.2005. National Comprehensive Cancer survival benefit of extended (D2) lymphadenec- Center Network tomy in Western patients with gastric cancer based 79. Yoon SS, Coit DG, Portlock CS, et al: The on a new concept: a prospective long-term fol- diminishing role of surgery in the treatment of 40. Siewert JR, Bottcher K, Stein HJ, et al: Relevant low-up study. Surgery 123:573, 1998 gastric lymphoma. Ann Surg 240:28, 2004 prognostic factors in gastric cancer: ten-year 61. Karpeh MS, Leon L, Klimstra D, et al: Lymph 80. Parsonnet J, Hansen S, Rodriguez L, et al: Heli-
  • 13. © 2005 WebMD, Inc. All rights reserved. ACS Surgery: Principles and Practice 5 GASTROINTESTINAL TRACT AND ABDOMEN 8 Tumors of the Stomach, Duodenum, and Small Bowel — 13 cobacter pylori infection and gastric lymphoma. N cal management and staging. Ann Surg 215: 97. Torres M, Matta E, Chinea B, et al: Malignant Engl J Med 330:1267, 1994 68, 1992 tumors of the small intestine. J Clin Gastroenterol 81. Isaacson PG: Recent developments in our un- 88. Conlon KC, Casper ES, Brennan MF: Primary 37:372, 2003 derstanding of gastric lymphomas. Am J Surg gastrointestinal sarcomas: analysis of prognostic 98. Talamonti MS, Goetz LH, Rao S, et al: Primary Pathol 20:S1, 1996 variables. Ann Surg Oncol 2:26, 1995 cancers of the small bowel. Arch Surg 137:564, 2002 82. Stolte M, Bayerdorffer E, Morgner A, et al: Heli- 89. Corless CL, Fletcher JA, Heinrich MC: Biology cobacter and gastric MALT lymphoma. Gut of gastrointestinal stromal tumors. J Clin Oncol 99. Zuckerman GR, Prakash C, Askin MP, et al: AGA 50:III19, 2002 22:3813, 2004 technical review on the evaluation and manage- ment of occult and obscure gastrointestinal 83. Schechter NR, Yahalom J: Low-grade MALT 90. Soft Tissue Sarcoma, Practice Guidelines in On- bleeding. Gastroenterology 118:201, 2000 lymphoma of the stomach: a review of treatment cology, v.1.2005. National Comprehensive Can- cer Center Network 100. Lewis BS: Small intestinal bleeding. Gastroen- options. Int J Radiat Oncol Biol Phys 46:1093, terol Clin North Am 29:67, 2000 2000 91. Demetri GD, von Mehren M, Blanke CD, et al: 101. de Mascarenhas-Saravina MN, de Silva Araujo 84. Martinelli G, Laszlo D, Ferreri AJ, et al: Clinical Lopes LM: Small bowel tumors diagnosed by activity of rituximab in gastric marginal zone Efficacy and safety of imatinib mesylate in ad- vanced gastrointestinal stromal tumors. N Engl J wireless capsule endoscopy: report of five cases. non-Hodgkin’s lymphoma resistant to or not eli- Endoscopy 35:865, 2003 gible for anti–Helicobacter pylori therapy. J Clin Med 347:472, 2002 Oncol 23:1979, 2005 92. Gilligan CJ, Lawton GP, Tang LH, et al: Gastric 102. Lambert P, Minghini A, Pincus W, et al: Treat- carcinoid tumors: the biology and therapy of an ment and prognosis of primary malignant small 85. Nilsson B, Bumming P, Meis-Kindblom JM, et bowel tumors. Am Surg 62:709, 1996 enigmatic and controversial lesion. Am J Gastro- al: Gastrointestinal stromal tumors: the inci- enterol 90:338, 1995 103. Ito H, Perez A, Brooks DC, et al: Surgical treat- dence, prevalence, clinical course, and prognos- 93. Schindl M, Kaserer K, Niederle B: Treatment of ment of small bowel cancer: a 20-year single insti- tication in the preimatinib mesylate era—a pop- gastric neuroendocrine tumors: the necessity of a tution experience. J Gastrointest Surg 7:925, 2003 ulation-based study in western Sweden. Cancer 103:821, 2005 type-adapted treatment. Arch Surg 136:49, 2001 104. Dabaja BS, Suki D, Pro B, et al: Adenocarcin- oma of the small bowel: presentation, prognostic 86. Miettinen M, Sobin LH, Lasota J: Gastrointes- 94. Modlin IM, Cornelius E, Lawton GP: Use of an factors, and outcome of 217 patients. Cancer tinal stromal tumors of the stomach: a clinico- isotopic somatostatin receptor probe to image 101:518, 2004 pathologic, immunohistochemical, and molecu- gut endocrine tumors. Arch Surg 130:367, 1995 105. Kodera Y, Yamamura Y, Shimizu Y, et al: Lymph lar genetic study of 1765 cases with long-term 95. Jemal A, Murray T, Ward E, et al: Cancer Statis- node status assessment for gastric carcinoma: is follow-up. Am J Surg Pathol 29:52, 2005 tics, 2005. CA Cancer J Clin 55:10, 2005 the number of metastatic lymph nodes really 87. Ng EH, Pollock RE, Munsell MF, et al: Prog- 96. Frost DB, Mercado PD, Tyrell JS: Small bowel practical as a parameter for N category in the nostic factors influencing survival in gastroin- cancer: a 30-year review. Ann Surg Oncol 1:290, TNM Classification? Tumor Node Metastasis. testinal leiomyosarcomas: implications for surgi- 1994 J Surg Oncol 69:15, 1998