Background: Surgical resection for intraductal papillary mucinous neoplasm (IPMN) of the pancreas has increased over the last decade. While IPMN with main duct communication are generally recommended for resection, indications for resection of side-branch IPMN (SDIPMN) have been less clear. We reviewed our single institutional experience with SDIPMN and indications for resection.
Methods: Patients who underwent resection for IPMN were identified from a prospectively maintained IRB-approved database. Patients with main pancreatic duct communication were excluded. Outcome, clinical and pathologic characteristics were correlated with endoscopic ultrasound (EUS) findings.
Results: From 2000 to 2010, 105 patients who underwent preoperative EUS evaluation and resection for SDIPMN were identified. The mean age was within the sixth decade of life, and there was a slight female predominance (55 vs. 45 %). The most common presenting symptom was abdominal pain (N = 47, 45 %), followed by jaundice (N = 24, 23 %) and weight loss (N = 24, 23 %). Only ten patients (10 %) were asymptomatic at presentation; seven (70 %) had suspicious features on EUS. Of the total cohort, few patients had intracystic septations (N = 27, 26 %) or presence of mural nodules (N = 2, 2 %) on EUS. Of 39 patients who had invasive pancreatic ductal adenocarcinoma (PDAC) on final pathology, EUS-fine needle aspiration (EUS-FNA) demonstrated malignancy in only 21 (54 %). An additional seven (18 %) had EUS-FNA findings of atypia or concern for mucinous neoplasm. EUS evaluation of cyst size was correlated with final pathology. Of 70 patients with EUS cyst size <3 cm, 12 (17 %) had a preoperative EUS diagnosis of malignancy. Final pathology revealed 24 (34 %) to have PDAC: 1 of 7 (14 %) patients with cyst size <1 cm, 2 of 19 (11 %) with cyst size 1-2 cm, and 21of 44 (48 %) with cyst size 2-3 cm. Fifteen of 35 (43 %) patients with cyst size >3 cm had PDAC on final pathology. Of the patients with cyst size <3 cm, 16 (23 %) had high-grade dysplasia on final pathology: 3 of 7 (43 %) with cyst size <1 cm, 3 of 19 (16 %) with cyst size 1-2 cm, and 10 of 44 (23 %) with cyst size 2-3 cm. Seven of 35 (20 %) patients with cyst size >3 cm had high-grade dysplasia on final pathology. Although overall survival (OS) at 48 months stratified by EUS cyst size did not significantly differ between groups, patients with PDAC on final pathology had significantly worse OS compared to noninvasive pathology. A total of eight patients (8 %) developed recurrent disease, all of whom had PDAC on final pathology.
Conclusion: EUS is a helpful modality for the diagnostic evaluation of SDIPMN. Considering the high incidence of malignancy as well as high-grade dysplasia in SDIPMN greater than 2 cm, EUS features should be used in conjunction with other clinical criteria to guide management decisions. Patients with SDIPMN greater than 2 cm that do not undergo surgical resection may benefit from more intensive surveillance.