Villous Adenocarcinoma of the Duodenum Invading the Ampulla of Vater: Case Report and Review of Literature

We report a case of villous adenocarcinoma of duodenum arising from the ampulla of Vater with a review of the literature. Although preoperative endoscopic biopsies were performed, no malignancy was identified. Because of the pathological uncertainty we decided to perform a pylorus-preserving pancreatoduodenectomy. Microscopic examination demonstrated glandular dysplasia with aspects of villous adenoma and well differentiated adenocarcinoma. We conclude that both in malignant cases and in cases with uncertain diagnosis a pylorus-preserving pancreatoduodenectomy is the best surgical treatment because it results in better 5 year survival.


INTRODUCTION
Villous adenomata of the duodenum represent 45% of all benign lesions of the duodenum, 1% of all duodenal tumors and 2.4% of all benign tumors of the small bowel2. This tumor wasfirst described by Perry in 1893 and to date 236 cases have been reported. Recently a high incidence of malignant degeneration has been reported ranging from 35% to 63% 3,4 Other reports have indicated that the incidence of villous adenoma associated with carcinoma in situ is 16O/o and that the ampulla of Vater is involved in 47,8% to 82% of cases 3, 6-11" We report a new case of villous adenoma of the duodenum involving the ampulla of Vater and a re-

CASE REPORT
A. 46 year old male presented with a 4 year history ofrecurrent episodes ofjaundice, nausea and intermittent fever that resolve with antibiotic therapy. An ultrasound scan revealed an enlarged liver with dilatation ofthe gallbladder, pancreatic duct and biliary tree to 2 cm diameter. Endoscopy revealed a tumor invading the second portion of duodenum at the level of the papilla, which on biopsy was a villous adenoma. Endoscopic retrograde cholangio-pancreatography (ERCP) demonstrated a mass invading the duodenum and ampulla of Vater (Fig. 1). A biopsy demonstrated a villous adenoma with moderate dysplasia and positive carcino embryonic antigen (CEA).  Blood analyses were all normal except for an alkaline phosphatase of 373 U/I and a white count of 11,500/mm3.
Hypotonic duodenography with contrast revealed a thickening of the wall at the level of the third part of the duodenum. Abdominal CT scan did not identify any enlarged lymphnodes, hepatic or pancreatic lesions.
On the basis of the pathological uncertainty we decide to perform a pylorus preserving pancreatoduodenectomy (PPD).
Gross analysis of the operative specimen showed a plaque lesion 6x4 cm with villi invading circumferentially the pancreatic duct and the common bile duct at the papilla level for 0.8 cm; the duodenal mucosa appearances were of a pseudopolyp (Fig. 2). Microscopic examination demonstrated a glandular dysplasia with aspects of villous adenoma and well differenciated adenocarcinoma. The tumor was predominantly intramucosal with spread in the muscolaris mucosa. Nodes showed only reactive change (Fig. 3). The postoperative period was unremarkable with a hospital stay of 15 days. Three months postoperatively Tc 99 (technetium) scintigraphy showed a prompt visualization of the liver with late images of the ectatic intrahepathic biliary tree. Gastroduodenal manometry demonstrated Motor Migrating Complex (CMM) in the jejunum with the normal presence of phase I after phase II and a longer phase III of Interdegestive Motor Complex (IMC). At two years follow-up the patient is disease free.

REVIEW OF THE LITERATURE
To December 1991 two hundred-thirty-six cases of villous adenoma of the duodenum have been described.
In 148 patients (63%) the tumor involved the ampulla of Vater and are the subject of the study. We excluded 28 patients because lack of information7, having 120 cases in the study group.
We analyzed several factors including: age, sex, benign or malign pathology, and mean survival after diagnosis.
Statistical analysis was performed using the student t-test for unpaired data using a Mystat statistical package running on Apple computer, p < 0,05 was considered significant.
Age and sex information was available in 83 patients. Age ranged between 11 and 80 years with a mean of 60.4; mean age of patient with benign tumors was 60. 8  The histological type of tumour was evaluated in 120 patients, of these 61 were benign and 59 had signs of malignancy.
PPD has been performed in 50 patients. Of these 41 were for malignant tumors and 9 for benign (adenoma with or without cellular atypia). One patient with a benign tumour died post-operatively giving a overall mortality of 1,8%.
Transduodenal or endoscopic excision was performed in 13   Our experience shows that the mean age of affected patients in both groups was 60 years; it is also interesting that females have a higher incidence of benign tumors (male:female 0.7) and males have a higher incidence of malignant tumours (male:female 1,47).
The association between carcinoma and villous adenoma of the duodenum has been reported by several authors14-17; this is demonstrated by the coesistence of adenoma, residual adenomatosis and microadenoma in 91,4% out of 58 cases ofinvasivecarcinoma of ampulla of Vater 15 and of residual adenomatosis in 81,8% of ampullary carcinoma17.
The histological classification often used is that proposed by Komorowski  Diameter, histological type, absence or presence of involved lymphnodes and pancreatic invasion do not have any effect on survival16'17. Kutin states that the presence of a villous adenoma larger than 4 cm can be a sign of degeneration1.
The symptomatology is never obvious. The diagnosis is often made by chance during radiography, endoscopy, duodenal operation for other diseases TM or autopsy5. The most frequent symptoms are: jaundice (69%), abdominal pain (40%) anaemia or melena (19%), weigh loss (18%) and fever (18%). The passing of mucous with consequent loss of electrolytes, described in villous adenoma of colon and rectum, are rarely associated with duodenal adenomas.
Endoscopy is the best procedure to identified villous adenoma because of the opportunity for biopsy but unfortunately the occurrence of false negative is of the order of 33% because of the small dimension of the biopsies; this problem is often encountered during intraoperative biopsiess. Endoscopy can also be used in particular cases to decompress the biliary tree by sphincterotomy or by positioning a bile drainage catheter to avoid infection and complications such as haemorrhahage and postoperative renal failure19. Ultrasonography rarely has been usefuP even if a "pseudokidney" has been identified2. Recently the utilization of echoendoscopy permits more accurate identification of a villous adenoma with malignant degeneration invading submucosal layers21. Hypotonic duodenography, can help to identify the lesion that occurs in the second portion of the duodenum. CT scan of the upper abdomen is useful to show any sign of local spread and nodal involvement.
An open problem is the treatment that can be surgical (PPD or transduodenal excision), endoscopical or in the inoperable cases a palliative by pass.
Chappuis states that endoscopic excision has to be performed in all cases except those with invasive carcinoma. Evans 22 believes that endoscopic exision should be confined to a polypoid lesion. Some authors think that local excision should be reserved for villous adenoma with carcinoma in situ5'12, others believe this should be reserved for tumours of the I, II and IV duodenal portion4'22.
Surgical or endoscopic excision produces an incidence of recurrence of 28% with a disease free mean of 12 months6, also if the tumor involves the lumen of the biliary or pancreatic tree than a higher incidence of haemorrhahage, perforation or recurrence will be found22. Our study shows a post-operative recurrence rate for benign villous adenoma of 12,5%. Unfortunately the recurrence is not always benign. One patient presented with malignant degeneration in the recurrence.
Because ofthe possibility of a false negative biopsy in carcinoma, we think that PPD should also be performed in those patients with uncertain preoperative diagnosis; PPD is also indicated for recurrent tumours with possible degeneration. We also believe that a surgeon must be able intraoperatively to change, on the basis of the local situation (dimension, nodes involvement, and histologic responce), the surgical excision to the more complex PPD. Duodeno-pancreatectomy is associated with a overall mortality of 2%.
Our review shows that the mean survival after PPD differ statisticaly compared with local excision; in particular the 2 year survival is higher in patients who underwent PPD (52%) versus 37% in patients who only had excision of the tumour. The 5 years survival is 22% in the PPD group versus 12% in the excision group. These results in our opinion, suggest that PPD is a better operation for these tumours and is associated with a better 5 years survival.
Surgical by-pass has the lowest mean survival because this procedure is always performed for advanced tumour.
CONCLUSION Diagnosis is often difficult on endoscopy even when examination is associated with a biopsy and echoendoscopy.
In obvious benign cases the choice is between the endoscopic excision (polypoid and small tumors) or trans-duodenal excision.
In malignant cases pylorus preserving pancreatoduodenectomy is the best surgical treatment because it results in better 5 year survival.
If the diagnosis is uncertain than a PPD is to indicated because the possibility of a occult carcinoma.