Incidental cystic neoplasms of pancreas: what is the optimal interval of imaging surveillance?

Am J Gastroenterol. 2008 Jul;103(7):1657-62. doi: 10.1111/j.1572-0241.2008.01893.x. Epub 2008 Jun 28.


Background: The optimal interval of imaging studies for surveillance of incidental pancreatic cystic neoplasms is not known.

Objective: A retrospective analysis of longitudinal medical records of patients with pancreatic cystic neoplasms was performed to examine the natural history of incidentally detected cystic pancreatic neoplasms with respect to the development of significant growth and to identify predictors of such growth.

Results: After excluding patients with small (<10 mm) cysts (N = 144) and inadequate clinical follow-up of less than 6 months (N = 79) and those with a clinical diagnosis of pancreatic pseudocysts, serous cystadenoma, main duct intraductal papillary mucinous neoplasm (N = 29), and neuroendocrine tumor (N = 3), in total, 166 cysts in 150 patients were available for analysis. The working diagnoses on these cysts (based on clinical, radiological features, aspiration cytology, cyst fluid analysis, and surgical pathology data when available) were mucinous cystic neoplasm in 117 and branch-type intraductal papillary mucinous neoplasm in 49. The mean standard error (SE) initial size of these cysts was 2 (0.1) cm. Over a median period of follow-up of 32 (IQR [inter-quartile range] 19-48) months, 89% of all the cysts did not show significant growth during the follow-up. In a multivariate Cox proportional hazards model, the initial size of the cystic lesion was an independent predictor of significant growth during follow-up (relative risk 1.28, 95% confidence interval [CI] 1.08-1.61, P= 0.01); the only other significant variable was the presence of intracystic or mural nodule (relative risk 38.6, 95% CI 2.3-654, P= 0.01).

Conclusion: Most incidentally detected cystic neoplasms of the pancreas did not have significant growth during follow-up. Such growth is unlikely to occur before 2 yr of the baseline evaluation, and we suggest that the optimal imaging interval during follow-up of these patients should be at 2 yr from the baseline evaluation, particularly in cystic lesions 3.0 cm or less in size and without intracystic or mural nodules.

MeSH terms

  • Aged
  • Disease Progression
  • Female
  • Humans
  • Incidental Findings
  • Male
  • Pancreatic Cyst / diagnosis*
  • Pancreatic Cyst / pathology
  • Pancreatic Neoplasms / diagnosis*
  • Pancreatic Neoplasms / pathology
  • Proportional Hazards Models
  • Retrospective Studies
  • Time Factors