A cystic lesion of the pancreas is a radiographic finding that has a broad histologic differential. This differential includes nonneoplastic pseudocysts, benign neoplastic cysts, premalignant cysts, and cystic lesions with invasive carcinoma. The widespread use of high-quality CT scanning in recent years has led to an increasing number of patients in whom identification of small asymptomatic cysts is made. The management of these patients is controversial. This controversy exists primarily because of our limited ability to determine the histologic diagnosis without resection.
The potential benefit of resecting any given cyst must be weighed against the potential substantial morbidity and the measurable mortality associated with resection, particularly in those cases where more complex surgical procedures are required. In the large number of patients currently being identified with lesions less than 3 cm in size, resection would result in a mortality higher than the rate of malignancy. The associated risks notwithstanding, many surgeons have recommended routine resection; others have advocated a more selective approach. Research efforts at many institutions are under way to better define criteria that can be used in a widespread selective approach to these patients.
Approximately seven years ago, we initiated several research efforts that we felt would be important in defining the natural history of pancreatic cystic lesions and would improve our ability to obtain diagnosis without resection. Since then, researchers from Memorial Sloan Kettering Cancer Center and other institutions have reported results of studies using a selective approach to resection in these patients.
All patients evaluated by our surgeons and gastroenterologists for a cystic lesion of the pancreas are enrolled in our pancreatic cyst registry. This registry currently contains data on more than 2,000 patients. Results of our recent review of this registry, published in the Journal of the American College of Surgeons, support the use of this selective approach.(1)
We evaluated 1,424 patients enrolled between 1995 and 2010; 1,141 of the patients (80 percent) were followed radiographically for more than six months. Initial management (within six months of first assessment) was operative in 422 patients (37 percent) and nonoperative in 719 patients (63 percent). Patients followed radiographically were more likely to have cysts that were asymptomatic (72 percent versus 49 percent, p < 0.001), smaller (1.5 cm versus 3 cm, p < 0.001), without solid component (94 percent versus 68 percent, p < 0.001), and without main pancreatic duct dilation (88 percent versus 61 percent, p < 0.001). Changes prompting subsequent operative treatment occurred in 47 patients (6.5 percent), with adenocarcinoma identified in 8 (17 percent) and pancreatic endocrine neoplasm in 4 (8.5 percent). Thus, of the 719 patients initially managed nonoperatively, invasive malignancy was identified in 12 (1.7 percent), with adenocarcinoma seen in 1 percent.
Most studies that have advocated a selective approach to resection for pancreatic cystic lesions have reported the radiographic characteristics of a solid component, cyst size, and symptoms to be associated with treatment recommendations. At Memorial Sloan Kettering, we think that radiographic follow-up is warranted in any patient where the presumed risk of malignancy is less than the risk of mortality from resection (no solid component, smaller than 3 cm, asymptomatic, no main pancreatic duct dilation). Specifically, the majority of patients with incidentally discovered cysts smaller than 3 cm in diameter and without a solid component can be safely followed radiographically. In a young patient with a small mucinous tumor, the additional factors for consideration are the likelihood of progression to malignancy and, in certain circumstances, the patient’s family history for pancreatic malignancy. No data are available for the former.
One of the very interesting areas of current research is into the identification of markers that will allow us to differentiate histopathologic subtypes of cysts and to identify dysplasia in patients who present with intraductal papillary mucinous neoplasms (IPMNs), cystic lesions produced from the secretion of a thick fluid called mucin from abnormal cells lining the pancreatic ductal system. In IPMNs, the cells continue to secrete fluid, which builds up in the pancreatic ductal system and creates a cyst detectable on CT scan. IPMN lesions have the ability to progress to invasive cancer; however, the pace and frequency of progression among patients is not yet fully understood.
Histopathologic analysis of resected IPMN specimens has revealed that these cysts contain a spectrum of dysplasia, ranging from low-grade dysplasia to high-grade dysplasia to invasive pancreatic cancer. Also, most IPMNs show variable amounts of dysplasia throughout the same cyst. Because of the latter characteristic, preoperative sampling of IPMN cyst fluid with fine-needle aspiration (FNA) cytology has not accurately reflected the degree of dysplasia in the cyst as a whole. We have performed multiple studies on the cyst fluid of patients with pancreatic cysts and have identified several interesting markers that we believe will be able to improve our ability to select patients for resection.
In a recent study performed by our group, we hypothesized that dysplasia in IPMN invokes an immunogenic/proinflammatory microenvironment that could be quantified in the cyst fluid and allow for the identification of high-grade dysplasia or carcinoma prior to operation.(2) Between 2005 and 2009, 147 patients underwent pancreatic resection for IPMN at Memorial Sloan Kettering. Pancreatic cyst fluid aspirates were collected at resection from 40 of these patients and all samples were included in this study. Cytokine expression was determined for the samples collected at resection, and correlative studies on preoperative cyst fluid carcinoembryonic antigen and FNA cytology were performed on a subset of these patients. We found levels of interleukin-1β (IL-1β) to be significantly elevated in the cyst fluid of patients with high-grade dysplasia and absent in the fluid of patients with low-grade dysplasia. Cyst fluid carcinoembryonic antigen levels alone, or in combination with IL-1β, did not identify high-risk lesions in the subset, whereas IL-1β levels alone did confidently identify high-risk lesions (p = 0.0007). Validation studies are ongoing, and we hope in the future that markers such as IL-1β will allow us to more appropriately identify patients who need resection and prevent surgery in those who can be safely monitored radiographically.