Improving Treatment of Pancreatic Cysts

The majority of patients who develop pancreatic cysts require no treatment due to the benign nature of their cysts. However, a small percentage of cysts will develop into pancreatic cancer if left untreated, and early identification and treatment of pancreatic cancer is a major factor influencing patient outcomes. Current methods at identifying which patients require surgery is ineffective, and many patients with benign cysts undergo unnecessary surgeries that are associated with high morbidity and potential mortality. DNA from pancreatic cyst fluid has been identified as a potential method for classifying pancreatic cysts, but this approach has not been tested in a large, comparative study. CompCyst is a novel tool for assessing risk and guiding treatment selection in patients with pancreatic cysts. CompCyst uses clinical features, imaging characteristics, and genetic and biochemical markers from cyst fluid to stratify patients into one of three groups: benign cyst requiring no treatment, potentially malignant cyst requiring surveillance, and high-grade dysplasia or overt cancer requiring surgical resection.

In a study published in Science Translational Medicine, the CompCyst algorithm was used to evaluate the molecular profiles of 426 patients with pancreatic cysts who had undergone surgical resection. Classification with CompCyst was compared to current clinical standards for cyst classification, with both of these classifications compared to gold-standard histopathological analysis of the resected cyst. CompCyst was more effective than current clinical practice in identifying the optimal treatment approach for patients with pancreatic cysts, correctly classifying cysts and directing treatment for 69% of patients, compared to 56% with current standards. Among patients with benign tumors, CompCyst was able to correctly identify 60.4% of patients, compared to only 18.9% using current clinical standards. Of patients with potentially malignant cysts requiring observation, CompCyst was able to correctly identify 48.6%, while only 34.4% were identified with current standards. CompCyst and current classification systems had similar accuracy in identifying patients with malignant cysts requiring surgery (91% vs 89%).

The investigators concluded that the CompCyst algorithm was more effective than conventional criteria at identifying risk classification in patients with pancreatic cysts. The authors highlighted that use of CompCyst would have allowed 60% of patients to avoid unnecessary surgery and that use of this tool in clinical practice may help reduce the morbidity and costs associated with over treatment of patients with benign or low-risk pancreatic cysts. A larger, prospective study is currently being planned to more fully assess the accuracy of CompCyst in directing treatment for pancreatic cysts.

Read more about this study on Medscape Oncology.

Sci Transl Med. 2019 July 17. [Epub ahead of print.]
CompCyst-pancreatic-cysts