The literature on LECs of the pancreas is limited to reports of single or small numbers of cases. Adsay et al. [2
] reported a clinicopathologic analysis of 12 patients with LEC. In this report, the epidemiology of this disease was described as follows. LECs of the pancreas are very rare, constituting approximately 0.5% of pancreatic cysts. They are seen in middle-aged patients (mean age 55 years, range 35–82 years), predominantly but not exclusively in men (M/F = 4/1). The most common symptom at presentation is abdominal pain. Other complaints at presentation include nausea, vomiting, anorexia, weight loss, back pain, fatigue, fever and chills. Many cases are diagnosed during a work-up for other diseases. In the present case, the patient had upper abdominal discomfort. Upper gastrointestinal endoscopy showed no abnormal findings such as gastritis or duodenitis. We considered this symptom to be caused by LEC.
The traditional markers, such as CEA, CA19-9, carbohydrate antigen-125, cancer-related antigen 72-4, pS2 and mucin-like carcinoma-associated antigen, as well as fluid viscosity would be expected to be significantly lower in LECs than in mucinous neoplasms [3
]; however, the limited experience with LECs does not confirm this prediction. Some cases had high CA19-9 levels in the cyst but not in the serum [6
]. Since the excretory ducts of the normal pancreatic tissue and some of the epithelial cells lining the cyst were immunoreactive for CA19-9, it can be concluded that CA19-9 in the cyst contents is probably produced by cells derived from the exocrine pancreas [7
]. This is an interesting finding, considering that squamous metaplasia in pancreatic ductal epithelium is often immunohistochemically negative for these markers [2
]. In the present case, many of the columnar epithelial cells and goblet cells located on the surface of the squamous epithelium lining the cyst were immunoreactive for CA19-9. On the other hand, the squamous epithelial cells were all negative for CA19-9.
LECs are often round and have a well-defined wall that is sharply demarcated from the pancreas and surrounding adipose tissue [2
]. The lesion often protrudes from the pancreatic parenchyma, and many cases appear to be peripancreatic rather than intrapancreatic [8
], as in the present case. The mean size of LECs is 4.7 cm (range 1.2–17 cm), and they are either unilocular (40%) or multilocular (60%) [9
]. In the present case, the lesion was a 6.5 cm, multilocular cyst. The main radiographic finding that may help distinguish LECs is that they are sharply demarcated from the pancreatic tissue [8
]. Cyst contents, which are mostly composed of keratin, have a different radiographic density than the mucin of mucinous cystic neoplasms; however, they may be difficult to distinguish from those of pseudocysts. This is true for the macroscopic appearance as well. The cyst contents display a ‘cheesy’ or ‘caseous’ appearance characteristic of keratinaceous debris or may be clear and serous in some cases [2
]. In the present case, the lesion seemed to be a simple cyst on CT, but EUS showed a mosaic appearance inside the cyst. The cyst contents of the resected specimen seemed cheesy macroscopically.
Microscopically, LECs are characterized by cysts lined by stratified squamous epithelium and immediately adjacent, dense subepithelial lymphoid tissue that contains lymphoid follicles [2
]. In some areas, the lining may appear more transitional and in others, flat, cuboidal and focally denuded [2
]. The origin of pancreatic LECs is uncertain, with several possibilities suggested, including origin from epithelial inclusions in peripancreatic lymph nodes (although such inclusions have only exceptionally been demonstrated and the lymphoid elements of most LECs do not have the features of a residual lymph node), squamous metaplasia of the pancreatic ductal epithelium, derivation as a form of monodermal teratoma, and displacement of a branchial cleft cyst [9
]. Another intriguing suggestion is that there is an interaction between a subset of lymphocytes and the pancreatic ductal epithelium that results in squamous differentiation and proliferation, with cyst formation initially due to secretory activity of the lining cells [2
For many purely cystic lesions, cytologic differentiation can be difficult because of acellular aspirates [10
]. However, for LECs, at least 20 cases have been reported in whom the diagnosis was made based on CT or EUS-FNA [11
]. Cytology classically shows abundant anucleate squamous cells, multinucleated giant cells, mature lymphocytes on a background of keratinaceous debris, and a lack of neoplastic cells [11
]. Interpretation of LECs by EUS-FNA is complicated by frequent contamination of the aspirate by tissues acquired by the needle during the procedure, such as mucinous and glandular epithelium from intestinal sources, making cystic neoplasm difficult to rule out [10
]. Very often, LECs can even contain a thick milky, creamy or frothy aspirate, further confusing the diagnosis [12
]. We performed EUS-FNA and obtained characteristic chylaceous fluid containing abundant anucleate squamous cells with few atypical cells. Those atypical cells are considered to be activated macrophages retrospectively. Mature or immature lymphocytes were not observed in our specimen. Chylaceous fluid with numerous anucleate squamous cells from pancreatic cyst strongly suggest LECs from our case. Chemical analysis of aspirated cyst fluid has proven to be useful in the differential diagnosis of pancreatic cysts in general [2
]. However, since this disease was not suspected, a chemical analysis was not done, and a diagnosis of a pancreatic LEC could not been made preoperatively.
No recurrences or progression into lymphoma or carcinoma have been documented in the cases of LECs for whom follow-up information was available. Whatever the mechanism of formation, pancreatic LECs are benign lesions that are cured by conservative resection. Therefore, if the tumor can be diagnosed preoperatively, the option of ‘wait and watch’ may be clinically acceptable [2
]. However, in most cases, the possibility of another type of pancreatic cystic neoplasm is difficult to rule out with the current investigative methods [2
]. The difficulty in preoperative diagnosis still leads many LECs to surgical resection. Also, because LECs are often well delineated from the pancreatic parenchyma, they may seem amenable to cystectomy. However, before such an operation can be attempted, the diagnosis needs to be confirmed, because this would be a suboptimal therapy for most other macrocystic neoplasms.