Mucinous Tumors of the Exocrine Pancreas: Intraductal Mucin-Hypersecreting Neoplasms

Mucinous Tumors of the Exocrine Pancreas: Intraductal Mucin-Hypersecreting Neoplasms

Intraductal mucin-hypersecreting neoplasms, also known as mucinous ductal ectasia (MDE), represent a form of pancreatic exocrine tumor previously confused with cystic neoplasms. This clinical entity, first described in 1982 by Ohhashi and colleagues, [3] was referred to as "mucin-producing intraductal tumor." The initial report described four patients with pancreatic carcinoma associated with unique clinical features distinct from the typical pancreatic ductal adenocarcinoma. Intraductal excretion of mucin resulted in diffuse dilatation of the main pancreatic duct and extrusion of mucus through a patulous ampullary orifice. The variety of terms subsequently used to describe MDE (Table 2) stems from a lack of understanding of this clinicopathologic entity. It seems likely, however, that various authors have focused on different stages in an adenoma-carcinoma sequence when choosing a descriptive diagnosis. [4] A recent study by Rivera et al [5] showed that intraductal papillary neoplasms with or without MDE represent a spectrum of main-duct papillary tumors ranging from adenoma to carcinoma. The authors therefore suggested the term intraductal papillary mucinous tumors (IPMTs) of the pancreas to encompass both tumors.
Little information is available on the epidemiology and etiology of IPMTs because they occur rarely and have been only recently recognized. The study by Rivera et al [5] suggested that cigarette smoking is a risk factor for IPMTs, which was also noted in a series based on all pancreatic cancers. [6] Data from animal models suggest that nitrosamines play a significant role in the etiology of both IPMTs and solid ductal adenocarcinomas. [7] IPMTs arise from the epithelial lining of the pancreatic ducts and progress along the duct system. Approximately 75% of the IPMTs arise from the main pancreatic duct within the head of the pancreas. Some IPMTs involve the ampulla of Vater and/or the minor papilla from where they protrude into the duodenum. The dilated duct segment contains a grossly visible papillary/villous tumor and/or viscous mucin. IPMTs may be benign, borderline, or malignant according to the differentiation of the proliferating columnar epithelial cells.
IPMTs occur mostly in men in the sixth to seventh decade (range = 40 to 85 years). The most common presenting symptom is abdominal discomfort associated with mild elevations of pancreatic serum enzymes mimicking chronic or relapsing pancreatitis. The history of pancreatitis or recurrent pancreatitis varies from 29% [5] to 80% [8] in the literature. These symptoms are more commonly observed in patients with mucinhypersecreting tumors and are most likely due to intermittent obstruction of the main pancreatic duct by plugs of viscous mucin or intraluminal tumor. Back pain (25%), jaundice (25%), weight loss (42%), steator-rhea (37.5%), and diabetes (37.5%) are among various presentations reported in one series. [5] In a recent report from our institution describing outcomes of 25 patients with IPMTs, [9] 52% of patients had acute relapsing pancreatitis with elevated pancreatic enzymes, 32% had pain with normal enzymes, 12% had weight loss with no pain, and 4% had incidental findings. Therefore, due to the indolent nature of these tumors and the limited awareness of this lesion among the medical community, a delay in diagnosis ranging from months to years is not uncommon.
Conventional imaging studies such as abdominal CT scanning and transabdominal ultrasonography reveal nonspecific changes that are indistinguishable from those of chronic pancreatitis, pancreatic pseudo-cysts, or nonmucinous producing cystic neoplasms. These studies typically demonstrate well-defined unilocular or multilocular cystic pancreatic masses and/or dilatation of the main pancreatic duct. The term cystic duct here would be a misnomer since these cystic structures represent dilated ductular structures. The cystic changes of IPMTs may occasionally be so extensive as that they are confused with a mucinous cystic neoplasm. An important distinction between the two lesions is that, in IPMTs, the "cysts" communicate with the duct system, whereas MCNs generally do not show any connection with the duct system.
Endoscopy and endoscopic retrograde pancreatography (ERP) are essential to clarify the diagnosis. The typical ERP reveals extreme dilatation along the entire length of the main pancreatic duct, with associated side branch ectasia, in the absence of a demonstrable pancreatic duct stricture that would explain the duct dilation (Fig 1). The cystically dilated portions of the duct are filled with mucus, and the dilation is the result of mucus obstruction. In a report by Cunningham and colleagues, [9] ERCP findings were patulous papilla (80%), visible mucin (84%), ductal filling defects (88%), and dilated pancreatic duct (100%), with whole pancreatic duct dilated in 64% and focal segmental dilation in 36%. Tissue specimens (aspiration cytology, brush cytology, or pancreatic biopsy) obtained endoscopically have a high false-negative rate but are helpful when positive.

(Enlarge Image)
Figure 1.
Endoscopic retrograde pancreatography demonstrates a diffusely dilated main pancreatic duct with filling defects and dilated irregular side branches consistent with mucinous ductal ectasia (MDE).
[ CLOSE WINDOW ]
Figure 1.
Endoscopic retrograde pancreatography demonstrates a diffusely dilated main pancreatic duct with filling defects and dilated irregular side branches consistent with mucinous ductal ectasia (MDE).
Pancreatoscopy may provide valuable information for the differential diagnosis of amorphous filling defects in the main pancreatic duct and may help to determine the extent of the pathology and the resection margin. Pancreatoscopy can be used to easily differentiate between an intraductal mass and mucin. Also, biopsies can be taken under direct vision from a suspicious mucosal lesion using the "mother-baby scope." [10] Although a diagnosis of malignancy can be obtained preoperatively, this usually does not substitute for surgical exploration and resection of the diseased segment based on frozen section results. In a series reporting on the results of 24 patients, [5] 20 (83%) were resectable. There were 9 pancreatoduodenectomies, 4 distal pancreatectomies, and 7 total pancreatectomies. The operation chosen is usually tailored to conform to the apparent distribution of the intraluminal tumor and not to the distribution of mucin. The inability to achieve histologically negative margins often leads to a total pancreatectomy. At final pathology, invasive adenocarcinoma was seen in 46% of the patients, carcinoma in situ was seen in an additional 24%, and low-grade dysplasia in 12% of the remaining patients. Despite the 88% prevalence of cancer in this series, the overall survival at a mean follow-up of 21 months was 81%. There were no recur-rences or new papillary growths in the pancreatic remnants. Four of the 24 patients had unresectable disease due to liver metastasis. The serum CA 19-9 value was helpful in predicting unresectability. Three of the 4 unresectable patients in this series had values higher than 2000 U/mL.
Cellier and colleagues [11] found invasive carcinoma in 43% of their patients and carcinoma in situ in 21%, with significantly poorer prognosis for patients with invasive carcinoma. The overall actuarial 3-year survival for patients with IPMTs was 83% in their series, whereas the overall recurrence-free 3-year survival was only 21% among patients with invasive disease. A review by Shyr et al [12] found a cumulative 5-year survival of 92% for 51 cases. In a study by Loftus and colleagues [13] involving 15 patients, 3 of 4 patients with malignant IPMTs died of metastatic disease, and 2 of 11 patients with benign disease died although with no evidence of tumor spread.
MRCP has emerged as a new noninvasive technology to assess both pancreatic and bile ducts without the use of contrast material or endoscopy. A study by Koito and colleagues [14] compared MRCP to ERCP in the diagnosis of mucin-hypersecreting tumors of the pancreas. MRCP showed cystic dilated branches of the pancreas better than ERCP. The numbers of nodules or septa that were detectable by MRCP and ERCP were similar. The authors concluded that MRCP appears to be more effective than ERCP for the detection of intraductal cystic lesions of mucin-producing pancreatic tumors. Because some of these tumors are not malignant and do not require surgery, MRCP offers a less invasive alternative to ERCP for follow-up studies. In the study by Yamaguchi et al, [15] the details of cystic lesions of the branch ducts (eg, mural nodule, communication with main pancreatic duct) were more clearly demonstrated by ERCP than by MRCP. ERCP was clear-ly superior in demonstration of normal main pancreatic duct and its branches. These two studies suggest that both ERCP and MRCP are complementary studies that are helpful in the diagnosis and follow-up of patients with IPMTs.
Cellier and colleagues [11] reported on the role of EUS in assessing tumor extension and invasion in patients with IPMTs (Fig 2). They demonstrated that the sensitivity and specificity of EUS for invasive malignancy were 78% and 75%, respectively. The overall accuracy of EUS in assessing tumor invasiveness was 76%. The accuracy for EUS using the GF-UM-3 and GF-UM-20 was 82% and 70%, respectively. This compares with an accuracy of 76% and 79% for CT and ERCP, respectively. The accuracy of EUS in assessing tumor extension was 62% overall and varied from 63% to 60% when using the GFUM-3 and GF-UM-20 instrument systems (Olympus Optical Co, Tokyo, Japan), respectively. Endoluminal endosonography using the UM-3R ultrasonic probe is a useful adjunct in the diagnosis of malignancy in patients with IPMTs. Inui and colleagues [16] studied 9 patients with IPMTs. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value for differentiation of malignant and benign design were 85.7%, 88.9%, 90.5%, 60.0%, and 96.9%, respectively. None of the patients developed acute pancreatitis, but high serum amylase levels were detected in 10% of the patients.

(Enlarge Image)
Figure 2.
EUS revealed a markedly dilated pancreatic duct with no mural vegetations. CBD = common bile duct, PD = pancreatic duct.