PTU-024 Pancreatic cysts – can investigations be safely rationalised?

Introduction Pancreatic cysts are a frequent incidental finding on cross-sectional imaging of the abdomen. We examined the decision making and outcome of patients with a pancreatic cyst(s) discussed at the Royal Derby Hospital HPB cancer MDT and compared practice against the 2015 American Gastroenterology Association (AGA) and 2017 International Association of Pancreatology (IAP) guidelines on the management of pancreatic cysts. Methods A search of HPB MDT meeting reports, from January 2016 to October 2017 (n=1144) identified 88 patient (51 female) reports relating to the first discussion of a pancreatic cyst. Electronic medical records were examined to collect data pertaining to subsequent investigations and outcome. Details of medical comorbidities were used to calculate a Charleson comorbidity index. Results The median age was 72 years (range 32–87) and the median estimated 10 year survival based on the Charleson Comorbidity Index was 53% (range 0%–98%). 86% of pancreatic cysts were judged to be an incidental finding. The median cyst diameter was 19.5 mm (range 4–110 mm). 43/88 (49%) patients proceeded to endoscopic ultrasound (EUS), with 33 having a fine needle aspiration (FNA). 4/88 (5%) patients had probably malignant (C4) or malignant (C5) cytology. All 4 patients had “high risk stigmata” on their initial CT/MRI. The final outcome for most patients was no further intervention (56%) or follow-up imaging (36%), with 5 (6%) patients offered surgery. Applying the 2017 IAP management algorithm, 13 (15%) patients had “high-risk stigmata” on CT/MRI and except where their performance status or co-morbidity precluded further investigation/treatment (5), were recommended for surgery (1) or EUS (7). Of the remaining 75 patients, 45 (60%) had no worrisome features on CT/MRI and so would not have required EUS. 21/45 (47%) of these patients in our practice underwent EUS, but none demonstrated definite mural nodules, main duct involvement or suspicious/positive cytology. The 2015 AGA management algorithm could only be applied to those patients who had undergone initial radiological assessment with MRI (n=11). None of these patients had two positive features on MRI, indicating a need for EUS. 6/11 (55%) patients did, however, have an EUS, with none identifying positive features or concerning cytology. Conclusions These findings suggest that a significant proportion (24%) of patients with pancreatic cysts underwent unnecessary EUS. Application of international guidelines can reduce the number of patients who require an endoscopic ultrasound.