Overview

The first case of duodenal adenocarcinoma was described by Hamburger in 1746 (Hamburger, 1775). In the United States, approximately 5000 patients are diagnosed with small bowel cancer every year, and 1000 of them have duodenal adenocarcinoma (Howe et al, 1999). Duodenal adenocarcinomas are rare tumors that account for approximately 0.3% of all gastrointestinal (GI) malignancies (Spira et al, 1977). Despite the duodenum representing only about 10% of the length of the small bowel, duodenal adenocarcinomas comprise 25% to 45% of small bowel tumors (Alwmark et al, 1980; Sohn et al, 1998); up to half occur near the ampulla of Vater, which explains why some studies group them with other periampullary cancers (Joesting et al, 1981). The incidence of small bowel cancers is estimated to be 1.4 per 100,000, with duodenal adenocarcinoma in approximately 0.17 per 100,000 (DiSario et al, 1994). Other malignant tumors such as carcinoid, gastrointestinal (GI) stromal tumor, lymphoma, and sarcoma can also occur in the duodenum but are far less common there than in the rest of the small intestine.

Presentation

Duodenal adenocarcinoma has a slight male predominance. The median age at presentation is in the sixth and seventh decades of life (DiSario et al, 1994; Rose et al, 1996). Approximately half of the patients with duodenal adenocarcinoma are initially seen with symptoms of abdominal pain or gastric outlet obstruction: nausea, vomiting, early satiety, and anorexia. Duodenal cancers often present as circumferential napkin-ring type lesions (Fig. 59.1). However, patients can also be seen with bleeding, melena, or jaundice. Bauer and colleagues (1994) noted a 21% incidence of a palpable mass on physical exam. A small proportion of patients are asymptomatic; the lesion is identified during the workup of an unrelated medical problem or as a result of screening for familiar adenomatous polyposis (FAP).

FIGURE 59.1 A, Macroscopic photograph of a resected duodenal adenocarcinoma. B, Microscopic image of duodenal adenocarcinoma.

Evaluation

The evaluation of any duodenal mass must assess for possible metastases and for resectability of the lesion (see Chapters 16 and 17). Depending on the patient’s presenting symptoms, the initial study that identifies the lesion will vary (see Fig. 59.1). Radiologic findings highly suggestive of malignancy include central necrosis or an intraluminal exophytic mass or intramural mass with ulceration (Fig. 59.2). Patients will need either a computed tomography scan or magnetic resonance imaging to assess resectability and rule out metastatic disease. For the tumor to be considered potentially resectable, there should be no evidence of major vascular encasement, distant lymphadenopathy, or distant metastases (e.g., liver, peritoneum, lung). If patients undergo upper GI endoscopy, the lesion can be biopsied at that time. Additionally, if a segmental resection is being considered, an upper GI series is helpful, and the relationship of the tumor to the ampulla should be evaluated by endoscopy.

FIGURE 59.2 A, Barium swallow demonstrating narrowing of the duodenum due to duodenal adenocarcinoma. B, Abdominal computed tomographic image demonstrating a duodenal mass in the second portion of the duodenum.

Risk Factors

The biology of duodenal adenocarcinoma is poorly understood. An adenoma-to-carcinoma sequence similar to that seen in colon cancer has been suggested. In various series of endoscopically resected adenomas, the incidence of malignancy varied from 10% to 42%, with villous adenomas being at higher risk for malignant transformation (Seifert et al, 1992; Witteman et al, 1993). Carcinoma in situ can be adequately treated with a local submucosal excision; however, it is generally impossible to know the maximal depth of invasion of the entire lesion before resection, and a more extensive procedure should be performed for any patient with invasive carcinoma (Bjork et al, 1990).

Patients with FAP are at significant risk for duodenal or ampullary adenomas, with an incidence as high as 88% (Church et al, 1992), and they have an increased relative risk of duodenal cancer of 330 compared with the general population (Offerhaus et al, 1992). Groves and colleagues (2002) prospectively followed 114 patients with FAP for 10 years, beginning at a median age of 42 years. Patients underwent upper endoscopy three times a year with a side-viewing duodenoscope. During this time three patients developed duodenal adenocarcinoma and three developed ampullary cancers. This occurred at a median of 6 years’ follow-up and at a median age of 68 years. All patients were classified according to the Spigelman classification: 3 were stage 0, 15 were stage I, 44 were stage II, 41 were stage III, and 11 were stage IV. Stage progression was seen in 0, 3, 7, 5, and 4 patients, respectively. Of the patients who developed invasive carcinoma, 4 were classified as Spigelman IV, but 2 were classified as Spigelman II and III, underscoring the need for surveillance in this population. Soravia and colleagues (1997) prospectively screened 74 patients over 5 years, during which time 11 patients (15%) developed advanced polyposis, and one patient developed duodenal cancer. The patient with adenocarcinoma presented at age 66 years, and the median age for duodenal polyposis for the group was 69 years (range, 38 to 66 years). All the patients in the study who underwent local resection of duodenal polyps developed recurrent polyps at sites of previous local resection. Duodenal and ampullary carcinomas are the second most common cause of death after colon cancer in patients with FAP, underscoring the need for aggressive surveillance.

Gardner syndrome is another condition associated with a high risk for duodenal adenocarcinoma. Patients present with GI polyposis, osteomas, and soft tissue tumors. The risk of malignant degeneration in duodenal polyps is between 3% and 12%, requiring these patients to also undergo an aggressive surveillance program (Sinha & Williamson, 1988). Patients with Peutz-Jehgers syndrome are also at risk, but it is lower than that seen in patients with Gardner syndrome or FAP.

The pathogenesis of duodenal adenocarcinoma is poorly understood. It appears to have a biology similar to, yet distinct from, colon cancer and other pancreatohepatobiliary tumors. Oncogenes and tumor suppressor genes that appear to contribute to the malignant progression include KRAS, TP53, ERBB2, and TGFBRII (Achille et al, 1998; Younes et al, 1997; Zhu et al, 1996). In patients with FAP, the normal duodenal mucosa and duodenal polyps have a higher COX-2 expression than patients with sporadic duodenal adenomas (Brosens et al, 2005). This increased expression may explain the poorer response to inhibitory COX-2 chemoprevention in these lesions.

Therapy