Pancreatic cystic neoplasia

All too often a pancreatic tumour turns out to be the common type of ductal adenocarcinoma, with its strong potential for early invasion and metastasis and the consequent infrequency of cure. The incidence of neuro-endocrine tumours of the pancreas roughly reflects the proportion of endocrine tissue within the gland (less than 5%). Separation between benign and malignant variants of neuroendocrine tissue is influenced by the size of the tumour. In ‘borderline’ cases the distinction may be made as easily by the operating surgeon who finds evidence of nodal or hepatic metastases as by the histopathologist looking down the microscope. It may be best to consider all neuroendocrine tumours as potentially malignant, though in the case of insulinoma the risk is less than 10%. En passant, I wonder whether any other tumour in the body that is as small as the average insulinoma or gastrinoma can cause such devastating systemic effects.

The only other pancreatic tumour that occurs with any degree of frequency is a cystic neoplasm. These interesting lesions embrace a number of different subtypes, three of which formed the subject of reports in the last issue of this journal, while a fourth was described in a paper published in 2000. Some ‘ordinary’ ductal carcinomas can present a cystic appearance on imaging because of central necrosis, and from time to time a neuroendocrine tumour will also undergo cystic degeneration. These curiosities apart, there are four main variants of pancreatic cystic neoplasm. Of these, the most innocuous is serous cystadenoma, though even here malignant transformation is a remote possibility. This tumour is also known as microcystic adenoma, and its honeycomb appearance on CT may allow quite a confident preoperative diagnosis. Indeed, since these lesions often arise in elderly women, a typical appearance on imaging may deter the surgeon from undertaking a major pancreatic resection.

Mucinous cystic tumours are much less innocent than their serous counterparts. They show a gradation from a truly benign neoplasm (cystadenoma) through a localised malignancy (cystadenocarcinoma) to an advanced, invasive cancer. Other things being equal, therefore, mucinous tumours should be resected whatever their site of origin within the gland. Fortunately, in this regard, they tend to affect the middle-aged more often than the elderly. Papillary-Cystic neoplasm has almost as many synonyms as case reports. It is both rare and virtually confined to young women. As the name suggests, solid and cystic components are combined within the tumour, which should be regarded as one of low-grade malignancy. The fourth type of cystic neoplasm is now termed intraductal papillary mucinous tumour or IPMT. The alternative label – mucinous ductal ectasia – draws attention to its main characteristic, namely a grossly dilated pancreatic duct that is full of mucin. Extrusion of mucin through the papilla at duodenoscopy is the crucial clue that the patient (usually a man) may have not just florid chronic pancreatitis but an underlying neoplasm with a dinstinct potential to undergo malignant change.

What can be learned about these pancreatic ‘fascinomas’ from the four papers that we have recently published? Fischer and colleagues from Edinburgh, UK, describe a 44-year-old woman with a mucinous cystadenoma of the pancreas who presented with acute pancreatitis ¹. She had a sudden onset of pain and vomiting with epigastric tenderness and a raised serum amylase. Imaging revealed a mature cyst in the tail of pancreas with peripancreatic inflammation and (remarkably) concomitant infarction of the spleen. The authors were rightly suspicious about the nature of this lesion and proceeded to remove the distal pancreas and spleen. Within the wall of the cyst there was a small focus of mucinous cystadenoma with ovarian-type stroma. This case is rather disquieting. In trying to distinguish a cystic neoplasm from an inflammatory psuedocyst, one usually relies on the absence of an antecedent history of acute and chronic pancreatitis (as well as the contour of the cyst on scanning). Thus the recent acute pancreatitis should really have ruled out the diagnosis of cystic neoplasm in this patient. The authors speculate that obstruction of the pancreatic duct by the small tumour may have triggered the attack of pancreatitis; alternatively this was a truly idiopathic attack and the cyst was an incidental (if fortuitous) discovery. Of course it has long been known that acute pancreatitis is occasionally caused by carcinoma of the ampulla, and I have encountered the same association with endocrine tumours.

Chedid and his associates from Porto Alegre, Brazil report another case of papillary cystic neoplasm in a young woman ². Notable features were her relative youth (18 years) and the fact that the site of origin within the pancreatic head necessitated a Whipple's operation. They suggest a hormonal influence to explain the particular propensity for this tumour to affect women of reproductive age. The histogenesis of papillary cystic neoplasm is obscure; both acinar, ductal and endocrine cells have been considered as the cell of origin, together with an undifferentiated progenitor cell. The rarity of these tumours does not facilitate detailed studies!

From Aarhus, Denmark, comes a report of six cases of IPMT described by Bendix Holme and associates ³. All six patients had diffuse dilatation of the main pancreatic duct, and all six were treated by total pancreatectomy. A seventh patient had liver metastases and was not operated; he lived another 17 months. In contradistinction to the other types of cystic neoplasm of the pancreas, each case of IPMT affected a male patient (median age 68 years). For myself, I have seen two new cases of IPMT in the last year, one in a man of 65 years and one in a woman of 81 years. In each case mucin was seen coming through the papilla at endoscopy, though the younger patient also had a localised cyst ulcerating into the duodenum which could have been a mucinous cystadenocarcinoma or even a chronic pseduocyst. Proximal pancreatoduodenectomy was converted into a total pancreatic resection when frozen-section histology of the duct at the resection margin showed invasive IPMT. This experience inclines me to accept the suggestion from the Danish group that total pancreatectomy is the treatment of choice for this condition. They point out that IPMT patients usually have symptoms of chronic pancreatitis, which helps to justify such radical treatment.

To complete the survey of cystic neoplasms I will refer to a paper by Formentini and colleagues from Ulm, Germany that we published in early 2000 ⁴. They described a series of 25 patients with serous cystic neoplasms of the pancreas who had undergone operation over a 12-year period. Three of these tumours were serous cystadenocarcinomas, while another two benign serous cystadenomas were associated with ductal cancers. This experience is a timely warning that serous tumours are not always innocent lesions.

It is my belief that the increasing use of ultrasound and CT scanning to pursue unexplained abdominal pain will increase the yield of patients presenting with resectable cystic neoplasms of the pancreas. I have lately seen two such lesions that were detected on routine antenatal scans in pregnant women. Congenital pancreatic cyst might be a reasonable differential diagnosis in such a case, but this condition seems even rarer, at least as an isolated lesion. The most important conclusion to be drawn from this crop of cases is as follows: Beware the diagnosis of pseudocyst in a patient without either a documented attack of acute pancreatitis or clear-cut aetiological factors to support a diagnosis of pancreatitis. The outcome is bound to be unsatisfactory when the surgeon assumes that a cystic neoplasm in or adjacent to the pancreas is inflammatory in origin and fails to carry out any (or adequate) biopsy of its wall ⁵. Under these circumstances an internal drainage procedure is doomed to failure whether stomach or jejunum is used or, as I have seen on two occasions, stomach followed by jejunum at successive operations.