Background: Cystic pancreatic neoplasms may be benign, premalignant, or malignant. These lesions may remain asymptomatic for long periods and can be quite large at the time of presentation.
Methods: A retrospective analysis was used to determine whether preoperative evaluation can predict pathology and determine resectability and outcome.
Results: Over 12 years, 145 cystic pancreatic neoplasms, of which 24 (17%) were larger than 10 cm, were managed at the Johns Hopkins Hospital. Those 24 large tumors included 9 of 73 cystadenomas (12%), 7 of 27 cystadenocarcinomas (26%), 2 of 35 adenocarcinomas producing mucin or associated with a cyst (6%), 5 of 9 Hamoudi tumors (55%), and 1 dermoid cyst. Clinical symptoms, liver function tests, and computed tomographic scans did not distinguish benign from malignant pathology. On 18 angiograms, 2 malignant and 4 benign neoplasms demonstrated encasement or occlusion; however, 3 of these 6 tumors were resectable. Twenty of 22 patients (91%) who were explored underwent resection with no hospital mortality. For the entire series, 5-year survival for those with cystadenomas, cystadenocarcinomas, and cystic adenocarcinomas was 97%, 38%, and 9%, respectively. Three-year survival for those 7 with cystadenocarcinomas larger than 10 cm was 54%, compared with 51% for those 20 with smaller cystadenocarcinomas.
Conclusions: Preoperative evaluation usually does not predict pathology, resectability, or outcome. Moreover, resectability is high and morbidity is low, irrespective of size. Large cystic pancreatic tumors should be explored to determine pathology, attempt resection, and provide an opportunity for long-term survival.