Tumors of the Small Intestine

Published on 20/05/2015 by admin

Filed under Gastroenterology and Hepatology

Last modified 22/04/2025

Print this page

This article have been viewed 3670 times

CHAPTER 121 Tumors of the Small Intestine

Emmy Ludwig, Robert C. Kurtz

EPIDEMIOLOGY

Small intestinal neoplasms are a diverse and rare group of tumors that affect the gastrointestinal tract from the duodenum to the ileum. Although the small intestine represents 75% of the length and 90% of the absorptive area of the gastrointestinal tract,¹ less than 5% of all primary gastrointestinal tract tumors originate in the small intestine.

Many types of tumors arise from the small intestine, and about 65% of them are malignant.² The four most common types of malignancies—adenocarcinomas, carcinoids, lymphomas, and sarcomas (including gastrointestinal stromal tumors or GISTs)—account for 95% of all small intestinal tumors. In most case series of Western patients, adenocarcinoma is the most common small intestinal neoplasm, followed closely by carcinoid tumor.^3,⁴ Adenocarcinoma (Fig. 121-1) most commonly arises in the duodenum, and carcinoid tumor and lymphoma predominate in the jejunum and ileum. Sarcoma distributes equally among all three segments of the small intestine.⁵

Figure 121-1. Endoscopic image of a mucosal neoplasm in the jejunum of a patient with metachronous small intestine adenocarcinomas.

(Courtesy of Mark A. Schattner, MD, New York.)

A review of 144 cases of malignant tumors of the small intestine showed that 64% of patients were men, with a median age of 55.7 years; 47% of the tumors were adenocarcinomas, 28% were carcinoid tumors, 13% were sarcomas, and 12% were lymphomas. Overall five-year survival was 57% and median survival was 53 months. Survival was highest for early-stage tumors and when curative resection was performed successfully.⁶ Survival was not related to location of tumor.

A recent study using the Connecticut Tumor Registry data reviewed the epidemiologic and clinical characteristics of 1260 cases of small intestinal tumors. Of these tumors, 49.8% were found in men; the mean age at presentation was 65.2 years. As for location, 25.4% occurred in the duodenum, 15.3% were in the jejunum, and 29.7% arose from the ileum; the remainder of tumors were either multifocal or their location was unspecified. By type, 33.1% were carcinoids, 30.1% were adenocarcinomas, 16.5% were lymphomas, and 7.1% were GISTs. Virtually all (99%) of the tumors were invasive. Surgery was the primary treatment in 87.7% of patients.⁷

Surveillance Epidemiology and End Results (SEER) data estimate that in the United States in 2008, small intestinal cancers were diagnosed in 3200 men and 2910 women, of whom 1110 people will die of these tumors. From data for 2001 to 2005, the median ages at diagnosis and death were 67 years and 71 years respectively. The overall five-year relative survival rate for 1996 to 2004 from 17 SEER geographic areas was 57.8%; 31% of patients had localized disease at diagnosis, 33% had locally advanced disease, and 29% had metastatic disease at presentation. The corresponding five-year relative survival rates were 77.0% with localized disease, 62.2% with locally advanced disease, 36.0% with metastatic disease, and 40.2% for patients whose stage of disease was not known.⁸

Worldwide, rates of small intestinal tumors vary; rates are higher in North America and Western Europe and lower in Asia. The Maori of New Zealand have an unusually high incidence of small intestinal tumors.⁹ Within the United States, incidence rates appear to be somewhat higher in African Americans and in men. Some have suggested an increasing incidence of small intestinal cancer over time.¹⁰

Carcinoid tumor (see Chapter 31) is the second most commonly diagnosed small intestinal malignancy, with an annual incidence of approximately three cases per million persons in the United States. The average age at diagnosis is 55 years. The most common sites of extranodal lymphomas (Fig. 121-2) are the stomach and small intestine (see Chapter 29). GISTs can be found throughout the small intestine (see Chapter 30).

Figure 121-2. Endoscopic image of a mucosal neoplasm in the duodenum in a patient with anemia and overt gastrointestinal bleeding. Pathology revealed an Epstein-Barr virus and associated B-cell lymphoma.

(Courtesy of Satish Nagula, MD, New York.)

PATHOLOGY

Neoplasms can arise from any of a wide range of epithelial and nonepithelial cells of the small intestine. Adenomas and adenocarcinomas arise from glandular mucosa; carcinoid tumors originate from argentaffin cells; lymphomas develop from clonal proliferation of lymphocytes; and GISTs have been identified as originating from the interstitial cells of Cajal, which help control the motility of the intestine and have elements of both smooth muscle and neural differentiation.¹¹ Leiomyomas arise from smooth muscle cells in the muscularis propria or muscularis mucosa, can be intra- or extraluminal, and account for about 40% of benign intestinal neoplasms. Leiomyomas with more than two mitoses per high-power field are reclassified as leiomyosarcomas. Lipomas represent approximately 20% of benign small bowel neoplasms, are most commonly intraluminal, and are found in the ileum; they rarely bleed, but they can obstruct or intussuscept (Fig. 121-3).¹² The remainder of benign neoplasms include desmoid tumors (see Chapter 122) and hemangiomas (see Chapter 36).

Figure 121-3. A, Film from an enteroclysis study that demonstrates a smooth submucosal lesion that was found to be a lipoma (arrow). B, Surgical resection specimen of a lipoma from another patient who presented with intussusception and bleeding.

(Courtesy of Igor Laufer, MD, Philadelphia.)

Small intestinal adenocarcinoma is staged using the TNM system as shown in Tables 121-1 and 121-2. Subtypes of this cancer include adenocarcinoma in situ, mucinous adenocarcinoma, signet cell carcinoma, squamous cell carcinoma, adenosquamous cell carcinoma, small cell carcinoma, medullary carcinoma, and undifferentiated carcinoma.

Table 121-1 TNM Classification

CATEGORY	DESCRIPTION
Tumor (T)
Tx	Primary tumor cannot be assessed
T0	No evidence of primary tumor
Tis	Carcinoma in situ
T1	Tumor invades lamina propria or submucosa
T2	Tumor invades muscularis propria
T3	Tumor invades through muscularis propria into subserosa
T4	Tumor invades through visceral peritoneum or directly invades other organs or structures
Regional Lymph Nodes (N)
Nx	Regional lymph nodes cannot be assessed
N1	No regional lymph node metastases
N2	Regional lymph node metastases
Distant Metastases (M)
Mx	Distant metastases cannot be assessed
M1	No distant metastases
M2	Distant metastases

Table 121-2 Staging of Small Intestinal Adenocarcinoma

Typical patterns of distribution are seen with benign lesions of the small intestine just as with malignant lesions. The most common benign tumor is the adenoma, which can be found throughout the small intestine but is slightly more common in the duodenum and ileum. Adenomas account for about 30% of all benign small intestinal neoplasms. They are more likely to be polypoid in the ileum and sessile and villous in the duodenum. Villous adenomas tend to be somewhat larger and often occur in the second portion of the duodenum; 40% to 45% can have pathologic features of malignancy.

Benign lesions of the small intestine can often be removed by endoscopic means, after being localized using some or all of the methods described later. Techniques for resection and destruction include snare polypectomy, endoscopic mucosal resection (EMR), and argon plasma coagulation. Early duodenal cancers also have been removed using EMR with good results. Because the duodenal wall is thin, the risk of perforation with EMR is high; careful submucosal injection of the duodenum is recommended before removing a duodenal lesion.¹³ Two studies have shown EMR to be safe and effective, with the potential of reducing the need for surgical intervention for these relatively rare conditions.^14,¹⁵

Table 121-3 shows the variety of tumors that can arise in the small intestine.

Table 121-3 Some Benign and Malignant Tumors of the Small Intestine

Non-Epithelial Benign Tumors

Non-Epithelial Malignant Tumors

Fibrosarcoma

Gastrointestinal stromal tumor (GIST)*

Malignant Lymphomas

Burkitt’s lymphoma

Diffuse large B-cell lymphoma

Immunoproliferative small intestinal disease (IPSID)

Mantle cell lymphoma

T-cell lymphoma

Western type B-cell lymphoma (MALT)

* GIST may be a benign or malignant tumor, although all GISTs have malignant potential (see Chapter 29).

ETIOLOGY

Despite similarities in genetic mechanisms underlying adenocarcinoma of the small intestine and colon, the incidence of small intestinal adenocarcinoma is far lower than that of colonic adenocarcinoma. Given the identical genetic makeup and similar tissue anatomy of the small intestine and colon, the marked difference in cancer incidence between the two sites is difficult to explain. Both tumors in both locations likely develop via an adenoma-carcinoma sequence, and mutated K-ras, Ki-ras, TP53, DPC-4 (Smad-4), β-catenin, and mismatch repair genes are implicated in tumor progression (see Chapter 123).¹⁶ Factors that may reduce the incidence of small intestinal adenocarcinoma include higher intraluminal pH, liquidity of small intestinal content, rapid transit time of chyme, faster turnover rate of small intestinal epithelium, lower load of carcinogen-producing bacteria, and increased amounts of immunoglobulin (Ig)A-secreting lymphoid tissue surrounding the small intestinal lumen.⁷

RISK FACTORS

Despite the low incidence of malignant small intestinal tumors, several risk factors have been identified that contribute to their development (Table 121-4). A dietary study found that increased ingestion of refined carbohydrates, sugar, and red meat increased the risk of small intestinal adenocarcinoma, whereas a diet high in fish, fruits, and vegetables was protective. Alcohol consumption did not affect risk.¹⁷ A similar study also found a significantly increased risk with red meat consumption, as well as salt-cured or smoked foods, but did not show fruit and vegetable consumption to be protective.¹⁸ Tobacco use and alcohol consumption did not affect development of small intestinal cancer.¹⁸

Table 121-4 Risk Factors for Small Intestinal Tumors

African American ethnicity

Inherited adenomatous and hamartomatous polyposis syndromes

Familial adenomatous polyposis

Hereditary nonpolyposis colorectal cancer

Peutz-Jeghers syndrome

Other etiologic factors for the development of small intestinal cancer include Crohn’s disease (see Chapter 111), in which the risk for both colon cancer and small intestinal adenocarcinoma is elevated.¹⁹^–²¹ A recent meta-analysis pooling 34 studies of 60,122 patients with Crohn’s disease found a relative risk of 28.4 (95% confidence interval [CI]: 14.46 to 55.66) for the development of small intestinal adenocarcinoma.²² Celiac disease (see Chapter 104) also is thought to be a risk factor for developing non-Hodgkin’s lymphoma (NHL) and small intestinal adenocarcinoma.²³ Another study showed an increase in esophageal cancer, melanoma, and NHL in patients with celiac disease; only the risk of NHL persisted despite patient adherence to a gluten-free diet.²⁴

Familial adenomatous polyposis (FAP) predisposes to small intestinal adenomas with associated malignant potential (see Chapter 122). Most of these FAP-associated lesions are found in the periampullary duodenum, although they can occur throughout the length of the duodenum. More than 90% of patients with FAP have a varied number of duodenal adenomas, but only 5% to 10% develop duodenal cancer.^25,²⁶ Upper gastrointestinal polyps in patients with FAP might follow a different genetic pathway from that of colorectal cancer.²⁷ FAP mice models show overexpression of the β-catenin gene and deregulation of its signaling pathway. In these animals, a diet of chenodeoxycholate results in more significant duodenal polyposis secondary to the overexpression of β-catenin.²⁸ Bile has been proposed as a potential risk factor for small intestinal cancer,^29,³⁰ and there is some suggestion it may be synergistic with the genetic mutations seen in patients with FAP.³¹

Hereditary nonpolyposis colon cancer (HNPCC) patients also are at increased risk for developing small intestinal cancer (see Chapter 122). HNPCC patients can present with a small intestinal adenocarcinoma as their first and only tumor.³² Screening recommendations for patients with these inherited syndromes might change with the advent of newer imaging techniques for the small intestine.^33,³⁴

All adenomas of the small intestine, like adenomas of the colon, should be considered precancerous. Adenomas in both locations appear to progress through an adenoma-to-adenocarcinoma sequence. Malignant features are found in approximately one third of all small intestinal adenomas.³⁵ For reasons that are not clear, adenomas at the ampulla of Vater are typically villous, large, and more likely to be malignant than adenomas found elsewhere in the small intestine.²

Patients with Peutz-Jeghers syndrome develop hamartomatous polyps of the small intestine and colon (see Chapters 22 and 122), and adenocarcinoma can arise within adenomatous foci in the hamartoma, although the overall risk of this is unknown.³⁶ Other hamartomatous polyposis syndromes (see Chapter 122) include juvenile polyposis, which has an increased risk of gastrointestinal tract malignancy, also probably from adenocarcinoma arising from hamartomatous polyps. Other syndromes include Bannayan-Ruvalcaba-Riley syndrome, Cowden’s syndrome, Cronkhite-Canada syndrome, and Devon family syndrome. The gastrointestinal cancer risk in Cowden’s syndrome is well described, the others are less so.³⁷

SMALL INTESTINAL ADENOCARCINOMA

CLINICAL FEATURES

The symptoms and signs of small intestinal adenocarcinoma are directly related to the location of the lesion. As an example, adenocarcinoma of the duodenum rarely manifests with symptoms of intestinal obstruction; instead, patients tend to develop frank or occult gastrointestinal blood loss, abdominal pain, and biliary obstruction. In contrast, patients with adenocarcinoma of the jejunum and ileum have vague nonspecific complaints, including abdominal pain, nausea, vomiting and abdominal distention, coupled with overt or occult bleeding. Those who present with iron-deficiency anemia may have improved survival because of their earlier presentation.

Most neoplasms of the small intestine are not associated with symptoms and are diagnosed either late in their course or incidentally at laparotomy or autopsy. The general absence of symptoms can be attributed to the distensibility of the small bowel wall and the liquid nature of its luminal contents. If a lesion leads to symptoms, the presentation depends upon the pathology of the neoplasm and its location. At least 50% of benign lesions remain asymptomatic, whereas 70% to 90% of malignant lesions are associated with symptoms. No symptoms or signs are specific for either benign or malignant tumors, however, and even though the duration of symptoms tends to be shorter for patients with a malignancy than for those with benign lesions, months can elapse before the diagnosis is made.

If a lesion becomes large enough, a patient can present with cramping periumbilical pain, bloating, and nausea and vomiting resulting from mechanical small bowel obstruction. Small bowel obstruction is the most common presentation for benign lesions and occurs in as many as 70% of such cases. Obstruction can result from either luminal constriction or intussusception. In fact, a benign small bowel neoplasm is the most common cause of intussusception in adults. Adenomas, adenocarcinomas, and lymphomas tend to grow into the lumen, but they can grow through the submucosa and muscle, often becoming large before a diagnosis is made; these lesions also can cause volvulus. As many as 80% of malignant tumors are associated with abdominal pain, although the pain is not caused by obstruction. Back pain in a patient with a primary malignant small bowel lesion suggests spread to the retroperitoneum. Mechanisms of back pain include hemorrhage into the tumor, invasion of enteric ganglia, ischemia, and involvement of the serosa.

Gastrointestinal bleeding, usually chronic, is the second most common symptom of neoplasms of the small intestine. Bleeding occurs in 20% to 50% of patients with benign lesions (e.g., leiomyomas) and, less commonly, in patients with malignant lesions. Massive hemorrhage is more common with sarcomas (e.g., GISTs) than it is with carcinomas, carcinoid tumors, or lymphomas. Weight loss is rare in patients with benign lesions, but, along with anorexia, it is noted in 50% of patients with malignancies. Intestinal perforation also is rare in patients with benign lesions but occurs in 10% of patients with sarcomas and lymphomas. Periampullary lesions can result in jaundice or pancreatitis. In patients with carcinoid tumors, symptoms of the carcinoid syndrome typically do not develop in the absence of hepatic metastases, and even with hepatic lesions, 28% to 50% of patients with such tumors remain free of the carcinoid syndrome (see Chapter 30).

Physical examination might not reveal a small intestinal malignancy, although up to 25% of such patients have a palpable abdominal mass. Twenty-five percent of patients with malignancy also present with the findings of obstruction: distention, borborygmi, a palpable mass, and diffuse mild to moderate abdominal tenderness. Some patients have a positive fecal occult blood test, although the incidence of this positivity is highly variable and ranges from 10% to 70%. Other patients present with jaundice secondary to either biliary obstruction or hepatic replacement by metastases. Cachexia, hepatomegaly, and ascites may be present in patients with advanced metastatic disease.

Laboratory studies performed in patients presenting with early-stage disease may be unremarkable, which might in part explain why the diagnosis is often delayed. Gastrointestinal blood loss can lead to a microcytic iron-deficiency anemia. The presence of a duodenal lesion can block the biliary system, leading to an increase in serum alkaline phosphatase and bilirubin levels.

Serum carcinoembryonic antigen (CEA) is an unreliable tumor marker in small intestinal adenocarcinoma, much as it is in other gastrointestinal cancers, and is rarely elevated unless hepatic metastases are present.

5-Hydroxyindole acetic acid (5-HIAA) is a useful marker for carcinoid tumors, but because it undergoes first-pass metabolism by the liver, urinary levels of 5-HIAA are not elevated other than in the setting of significant hepatic metastatic disease.

DIAGNOSIS

Plain Films

Plain abdominal films rarely are useful for localization of small intestinal tumors, but they may be helpful clinically in the setting of a suspected intestinal obstruction or perforation. In 50% to 80% of patients with small intestinal neoplasms, an upper gastrointestinal series with a small intestinal follow-through (UGIS-SBFT) reveals an abnormality ³⁸ and specifically demonstrates a tumor in 30% to 44% of cases (Fig. 121-4); this figure increases to 90% if an enteroclysis study (small bowel enema) is performed (see Fig. 121-3).³⁹ A barium enema can reveal lesions in the terminal ileum if there is sufficient reflux of barium through the ileocecal valve.

Figure 121-4. Film from a small bowel follow-through examination demonstrating an apple-core appearance caused by a metastatic lesion to the small intestine from a scirrhous gastric cancer.

(Courtesy of Igor Laufer, MD, Philadelphia, Pa.)

Video Capsule Endoscopy

Video capsule endoscopy (VCE) uses a capsule that contains a radiotransmitter to generate photographic images of the small intestine. The patient swallows the capsule, which is propelled by peristalsis through the small intestine. The capsule generates two images per second and has a battery life of eight hours. Images are captured by a digital recorder worn by the patient, and the images are downloaded in a computer to be reviewed later by the physician.^40,⁴¹ This technology is useful for detecting tumors of the small intestine, especially when they are beyond the reach of push enteroscopy (Fig. 121-5). In 2002, a prospective trial compared traditional barium UGIS-SBFT with VCE in 22 patients with suspected small intestinal abnormalities. SBFT was found to be diagnostic in four (20%) patients; VCE was diagnostic in nine (45%) patients and suspicious in eight (45%).⁴² VCE was well tolerated. A much larger study published in 2006 looked at 562 patients who underwent VCE from August 2001 to November 2003 for a variety of indications; 79% (443/562) had evidence of occult gastrointestinal bleeding. A diagnosis was made by VCE in 277 patients (49.3%). Of the 562 patients studied, 50 (8.9%) were found to have a small intestinal tumor, 48% of which were malignant.⁴³

Figure 121-5. Capsule endoscopy view of an ulcerated mass in a patient who presented with gastrointestinal bleeding. Four ulcerated, bleeding masses were found throughout the small bowel; these were confirmed at surgery and found to be sarcomas.

(Courtesy of Ann Marie Joyce, MD, Philadelphia, Pa.)

VCE also has been used to locate the primary small intestinal lesion in patients with metastatic neuroendocrine tumors. In a study performed in 2006, 20 such patients were evaluated by CT scan, enteroclysis, nuclear imaging (¹¹¹In pentetreotide scintigraphy) and VCE.⁴⁴ Nuclear imaging revealed an abnormality in the abdominal region in 13 patients but could not localize the abnormality further. In contrast, VCE found nine small intestinal tumors, of which five were carcinoid tumors, for a diagnostic yield similar to that of nuclear imaging; VCE findings were felt to be more clinically helpful and relevant.

Other studies have shown the superiority of VCE in detecting small intestinal cancer over enteroclysis, push enteroscopy, colonoscopy, and CT scans. VCE was instrumental in diagnosis or management in 55% to 77% of patients evaluated.⁴⁵^–⁴⁸ A study reported in 2006 by Bailey and colleagues looked at the diagnosis and outcome of small intestinal tumors found by VCE.⁴⁹ In their study, 416 VCEs were performed in Australia between May 2001 and May 2004, identifying 27 tumors in 26 patients. Twenty-one of the 26 evaluations were for occult gastrointestinal bleeding. Twenty-three patients had both colonoscopy and upper endoscopy, which were unrevealing,²³ and had prior radiologic or nuclear evaluation (SBFT, enteroclysis, abdominal CT, tagged RBC study); eight (34.7%) of these patients had only a possible finding. VCE revealed nine benign lesions, eight of which were resected. Four were hamartomatous polyps.¹⁷ Thirteen patients were found to have small intestinal malignancies. Adenocarcinoma was identified in five, carcinoid tumor in six, and GIST in two. Most importantly, 23 of 26 patients underwent surgery, and in 12 of these 23 patients, surgery was regarded as curative, including six patients with malignant small intestinal tumors. None of the patients who underwent a curative resection developed evidence of recurrent disease at a follow-up interval that ranged from 26 to 51 months.

Although VCE is a useful technology for detecting small intestinal tumors, it is not without its limitations. As yet, no tissue for definitive diagnosis can be obtained by VCE. Tumors that can be seen on endoscopic push enteroscopy can be missed by VCE because views are often rapidly fleeting.⁵⁰ Push enteroscopy, however, is limited by its ability to proceed only to about 50 cm beyond the ligament of Treitz.⁵¹

Double-Balloon Endoscopy

The development of double-balloon endoscopy (DBE) has allowed evaluation of the entire small intestine, much like VCE, but with the important added ability to biopsy tissue and potentially to perform interventional maneuvers, such as coagulation of bleeding lesions and polypectomy. Yamamoto and colleagues first described DBE in 2001 ⁵² and elaborated on the clinical experience and outcome using it in 2004.⁵³ In this report, 66 of 123 patients were evaluated for occult gastrointestinal bleeding; 10 patients who presented with occult bleeding had polyps or tumors identified by the procedure.⁵¹ Another study of DBE demonstrated similar findings in investigating 64 patients with occult gastrointestinal bleeding: 13% had evidence of tumors or polyps.⁵⁴ Other studies have shown the efficacy and utility of DBE for the investigation of the small intestine.^55,⁵⁶ A case series showed that some small intestinal lesions missed on VCE are seen on DBE, suggesting that the two technologies are complementary in the evaluation and management of patients with small intestinal diseases.⁵⁷

Computed Tomography

Computed tomography (CT) in conjunction with newer modalities has been used with success to better image the small intestine. A study by Pfannenburg and colleagues showed that anatomic functional image fusion using combined transmission and emission tomography (SPET/CT) had a diagnostic accuracy of 99% in classifying neuroendocrine lesions and improved tumor localization and characterization in patients with neuroendocrine tumors.⁵⁸

The advent of multidetector CT enteroclysis (MDCT-E) has enabled demonstration of various findings including wall thickening, hypervascularity, and mural and extramural abnormalities.⁵⁹ MDCT-E requires nasojejunal intubation to enable rapid delivery of an enteral contrast medium to distend the small intestinal loops homogenously; water has been found to be the most useful neutral contrast medium.⁶⁰ Intravenous contrast also is necessary. MDCT-E has been shown to be a useful technology for detecting small intestinal abnormalities (Fig. 121-6).

Figure 121-6. A, An image from a CT enteroclysis study that revealed abnormal focal wall enhancement and multiple discrete masses in the ileum. B, Surgical resection specimen from the same lesion in A, which was found to be a carcinoid tumor.

(A, Courtesy of Ilan Weisberg, MD, New York. B, Courtesy of William R. Jarnagin, MD, New York.)

A prospective study reported by Boudiaf and colleagues in 2004, looked at 107 consecutive patients with suspected small intestinal pathology.⁶¹ Results were compared with the results of endoscopy, enteroscopy, VCE, pathology findings, and clinical follow-up. MDCT-E revealed 21 patients with small intestinal mass lesions, nine patients with active Crohn’s disease, two patients with small intestinal tuberculosis, four patients with small intestinal lymphoma complicating celiac disease, and 12 patients with low-grade small intestinal obstruction.⁵⁸ Sensitivity and specificity of MDCT-E were 100% and 95%, respectively, compared with the appropriate gold standard (endoscopy, enteroscopy, VCE, or pathology).⁶¹

Another study looked MDCT-E in symptomatic patients with suspected small intestinal neoplasm; researchers used endoscopic, pathology, and clinical follow-up findings as reference standards.⁶² After having normal endoscopy and colonoscopy, 219 patients underwent MDCT-E. Fifty-five patients had positive findings on MDCT-E, for an overall sensitivity and specificity of 84.7% and 96.9%, respectively. Pathologic evaluation revealed 50 patients with small intestinal tumors, including 19 with carcinoid, seven with adenocarcinoma, five with lymphoma, nine with jejunal adenoma, five with GIST, two with ectopic pancreas, two with angiomatous masses, and one with a metastatic lesion. Five patients had false-positive evaluations.⁶²

Endoscopic Ultrasonography

Endoscopic ultrasonography (EUS) is potentially helpful in determining the origin of intestinal tumors and whether they are benign or malignant. To date, however, the role of EUS in the small intestine has mainly been in the evaluation of pancreatic and ampullary tumors,⁶³ although its role is sure to evolve with more experience (Fig. 121-7).

Figure 121-7. A, Endoscopic ultrasonography in two patients with a gastrointestinal stromal tumor (GIST). A, A benign GIST appears as a small (<3 cm), homogeneous, hypoechoic lesion arising, most commonly, within and contained within the muscularis propria layer. B, A malignant GIST appears as large (>3 cm), inhomogeneous lesions arising within the muscularis propria and with possible extension into other wall layers and surrounding structures.

(A and B, Courtesy of Gregory Ginsberg, MD, Philadelphia, Pa.)

Angiography

Angiography is of limited value except to detect tumors with a significant vascular component and for localizing a site of severe bleeding. A technetium-labeled red blood cell scan (tagged RBC scan) is more sensitive than angiography to determine if a patient has active gastrointestinal bleeding and can be used to localize the lesion, but it cannot definitively diagnose the cause.

Future Prospects

It is likely that more comparison studies will be reported over the next few years to better elucidate the optimal algorithm for the diagnostic evaluation of patients with suspected small intestinal lesions. All of these new technologies—VCE, EUS, DBE, and MDCT-E—are likely to continue to play complementary roles in evaluating the small intestine.

TREATMENT

Benign Neoplasms

Therapy of benign neoplasms of the small intestine usually consists of resection, either via an endoscope or at laparotomy, as determined by the size, growth pattern, and location of the lesion. Most adenomas may be adequately removed at surgery by segmental resection, although if there is any evidence of malignant degeneration, the margins of resection should be extended and draining lymph nodes should be excised. Pedunculated duodenal adenomas may be removed endoscopically by snare cautery, whereas sessile duodenal adenomas are often managed by endoscopic mucosal resection.

If an adenoma involves the ampulla, it may be possible to remove it endoscopically or by sphincterotomy with biliary and/or pancreatic stenting. Larger periampullary or ampullary lesions, which are more likely to harbor malignancy, may require pancreaticoduodenectomy (Whipple’s procedure), the decision for which is facilitated by histopathology of frozen sections. Patients must undergo periodic surveillance following removal of an adenoma, especially if it is a villous adenoma, because there is a significant chance of recurrence; however, survival is excellent for these patients.

Malignant Neoplasms

Surgery remains the only known curative treatment for small intestinal adenocarcinoma. Curative surgical resection is successful in 40% to 65% of patients.^6,⁶⁴ Surgical treatment for adenocarcinoma of the duodenum, because of anatomic location, is a pancreaticoduodenectomy (Whipple’s procedure). In patients with resectable adenocarcinoma of the jejunum, removal of the tumor with wide margins and adjacent lymph nodes is performed. Terminal ileal adenocarcinoma is managed similar to cecal cancer, with a terminal ileal resection, right hemicolectomy, and lymph node resection.

Dabaja and colleagues presented a large retrospective review of 217 patients with small intestinal adenocarcinoma in 2004.⁶⁵ Surgery with curative intent was performed in 146 patients (67%). Adjuvant chemotherapy was administered in 59 patients (27%) and palliative chemotherapy in 48 patients (22%). The median overall survival time was 20 months. The five-year overall survival rate was 26%. Cancer-directed surgery, early-stage disease, and lymph node involvement ratio were significantly associated with overall survival by univariate analysis. Cancer-directed surgery and lymph node involvement ratio were independent predictors of overall survival in a multivariate analysis (adjusted rate ratio, 0.14; 95% CI: 0.04 to 0.46; P = 0.001). In a study reported from British Columbia, 47 patients with small bowel adenocarcinoma were identified between 1990 and 2000; 21 of the 47 were given chemotherapy, with the majority receiving 5-fluorouracil (5-FU)-based therapy. Of 19 patients treated with curative intent, the median overall survival was 38.6 months. Sixteen of 37 patients with advanced disease received chemotherapy. The median overall survival for those who received palliative chemotherapy was 15.6 months, compared with 7.7 months for those who did not.⁶⁶

There have been no large prospective studies of chemotherapy for adenocarcinoma of the small intestine because this tumor is uncommon. Most of the chemotherapy data, therefore, come from retrospective analyses, case reports, and case series. One such large retrospective review of continuous infusion 5-FU in advanced small intestinal adenocarcinoma showed an overall response rate of 37.5% and a median survival of 13 months.⁶⁷ In another retrospective study, combination chemotherapy with 5-FU plus cisplatin, carboplatin, or oxaliplatin in advanced small intestinal adenocarcinoma had an overall response rate of 21% and an overall survival of 14 months. Second-line chemotherapy using 5-FU plus irinotecan has shown promise, but use of continuous infusion 5-FU or 5-FU with cisplatin is not effective.⁶⁸

Other regimens with activity in gastric and colon cancers have been tried with varying benefit, including 5-FU alone or in combination with doxorubicin, mitomycin-C, cisplatin, and carmustine. A phase II study of 5-FU, doxorubicin, and mitomycin-C in 38 patients with unresectable, biopsy-proven adenocarcinoma of the small intestine showed a response rate of 18.4%, which was no better than that of previous studies of 5-FU-containing regimens.⁶⁹ In a small case series of three patients who had not responded to 5-FU, irinotecan was used as salvage therapy and appeared to have some efficacy, with marked improvement in symptoms, in two of the three patients.⁷⁰ Another study reported on six patients with carcinomatosis from small intestinal adenocarcinoma. The authors demonstrated that patients treated with cytoreductive surgery followed by intraperitoneal hyperthermic chemotherapy with mitomycin C given intraoperatively had a median survival of 45.1 months compared with various historical controls of 3.1 months and 12 months.⁷¹ A retrospective review of chemotherapy in 113 patients with advanced small bowel adenocarcinoma looked at newer agents compared with 5-FU-based regimens. Forty-four patients received palliative chemotherapy, with an overall response rate of 36% during a first- or second-line regimen (9% complete responses and 27% partial responses). Palliative chemotherapy predicted for overall survival in a multivariate analysis (hazard ratio, 0.47; P = 0.035).⁷²

A large multicenter prospective study likely will be required to determine the best adjuvant and palliative regimens for treating small intestinal adenocarcinoma. Radiation therapy has no proven role in treatment of small intestinal adenocarcinoma and can result in significant morbidity from radiation enteritis.

PROGNOSIS

Long-term survival generally is poor for patients with adenocarcinoma of the small intestine. Curative surgical resection is the mainstay of treatment and improves the five-year survival from 0% after palliative treatment to 54%. The absence of involved lymph nodes at the time of surgical resection improves the five-year survival further up to about 63%. Anemia at the time of initial presentation dramatically improves survival, likely because it leads to evaluation at an earlier stage.⁷³

Several studies have looked at prognostic indicators for survival with small intestinal adenocarcinoma. Brucher and colleagues looked a series of 94 patients with primary tumors of the small intestine, 62 of which were malignant; 22 of these were adenocarcinomas. They found a five-year overall survival of 45%.⁷⁴ Univariate analysis showed degree of resection (R status) and advancing tumor stage; the presence of lymph node metastases, distant metastases, carcinomatosis, and vascular invasion were poor prognostic indicators.

Dabaja and colleagues also looked at survival in 217 cases of small intestinal adenocarcinoma.⁶⁵ The median overall survival of these patients was 20 months (95% CI: 16 to 24 months). The five-year overall survival of all patients was 26%. Outcome, as expected, was closely related to stage of disease, with stage IV patients having the worst five-year overall survival (5%) and median survival period of 11 months. The five-year overall survival was significantly shorter for patients with positive lymph nodes compared with patients with negative lymph nodes (32% vs. 52%; median survival, 22 months vs. 78 months; P = 0.0039). The five-year overall survival rate was significantly higher for patients who underwent cancer-directed surgery than for patients who did not, and it was particularly high for those who underwent pancreaticoduodenectomy compared with other types of surgical resection (82 months vs. 29 months; P < 0.0001). Adjuvant chemotherapy did not appear to significantly affect overall survival when given after resection, but chemotherapy for patients with unresectable (stage IV) disease did appear to afford some benefit (12 months vs. 2 months, P < 0.02).

Wu and colleagues retrospectively reviewed prognostic factors in 80 patients with primary small intestinal adenocarcinoma; 60 of 80 underwent surgical management, and 45 of 60 had a curative resection. The five-year survival for patients undergoing resection was 27%; overall survival was 17.5%. In multivariate analysis, earlier tumor stage (I and II) and curative resection were two independent factors with a positive impact on survival. Conversely, the presence of lymph node metastases was an independent variable for poor disease-free survival in patients who underwent a curative resection.⁷⁴

OTHER MALIGNANT SMALL INTESTINAL NEOPLASMS

For diagnosis, management, and prognosis of lymphomas, GISTs, and carcinoids please refer to Chapters 29, 30, and 31, respectively.

Metastatic tumors are more commonly found in the small intestine than are primary tumors. Melanoma is the most common cancer to spread to the small intestine (Fig. 121-8). Patients with metastatic melanoma can present with occult or overt gastrointestinal bleeding. Surgical intervention may be performed as a useful palliation. Other cancers such as breast, lung, colon, stomach, and ovary can directly invade or metastasize to the small intestine (see Fig. 121-9). In these situations, surgery can be useful for palliation, and systemic chemotherapy can help control disease progression. Prognosis remains poor, however.