EUS-FNA bests CT/MRI in diagnosing certain malignant pancreatic lesions
Click Here to Manage Email Alerts
Endoscopic ultrasound with fine needle aspiration has “incremental value” over imaging techniques like computer tomography and MRI for identifying malignant branch-duct intraductal papillary mucinous neoplasms, according to retrospective study data.
“Our study further endorses the practice of incorporating [endoscopic ultrasound (EUS)] in the management of most (branch-duct intraductal papillary mucinous neoplasms (BD-IPMNs)],” which is currently recommended for some pancreatic cysts by International Consensus Guidelines and AGA guidelines, the researchers wrote. “Its ability to detect mural nodules missed by CT/MRI highlights the limitation of CT/MRI in predicting malignancy in some BD-IPMNs.”
To evaluate the performance of EUS-FNA for identifying malignant BD-IPMNs, compared with the predictive ability of imaging-defined high-risk stigmata and worrisome features as recommended by ICG 2012, researchers performed a retrospective cohort study of 364 patients with BD-IPMNs diagnosed at Indiana University Hospital from 2001 through 2013. They also evaluated for recurrence and long-term outcomes in patients who underwent surgery (n = 135) or imaging surveillance (n = 229).
All patients whose BD-IPMNs were surgically resected underwent CT and/or MRI and 78% underwent EUS-FNA. High-risk stigmata and worrisome features found on CT or MRI were comparable between BD-IPMNs that were determined by pathology to be benign (n = 117) or malignant (n = 18). However, main pancreatic duct dilation (5-9 mm) was more frequently associated with malignancy (P = .01)
In comparison, EUS-FNA features suspicious for malignancy were identified in malignant lesions more frequently. These included definite mural nodules (33% sensitivity, 94% specificity and 86% accuracy), main pancreatic duct dilation features suspicious for involvement (42%, 91% and 83%) and suspicious/positive malignant cytology (33%, 91% and 82%). CT/MRI missed the mural nodules identified by EUS in malignant lesions 28% of the time.
Finally, the researchers found that patients with malignant lesions had an increased risk for IPMN recurrence during a mean follow-up period of 131 months (P = .01), and some patients had benign IPMN recurrence up to 8 years after surgery.
“The high specificity and accuracy of EUS features of malignancy we report herein strongly position EUS-FNA as the optimum tool for diagnosing malignant BD-IPMNs, particularly in patients without [worrisome features] and with smaller cysts,” the researchers concluded.
These findings raise questions about current guideline recommendations on the management of BD-IPMNs, according to a related editorial written by Omer Basar, MD, and William R. Brugge, MD, both from the Pancreas Biliary Center, Gastrointestinal Unit at Massachusetts General Hospital.
ICG 2012, also known as the Fukuoka consensus guidelines, recommend EUS in patients with a history of clinical pancreatitis and worrisome features, while the AGA guidelines recommend EUS when there are two or more high-risk features, “so the question of when EUS-FNA should be performed is controversial,” they wrote.
“Furthermore, in low risk patients, the AGA guidelines suggest surveillance with cross-sectional imaging,” and discontinuation of imaging is recommended if there are no signs of significant change in size or morphology over 5 years. “There is concern that these guidelines may interfere with the detection of early malignancy,” they added.
Finally, the AGA guideline recommendations for postsurgical surveillance, which “are opposed to monitoring for resected IPMNs with a positive margin (except for high-grade dysplasia),” may increase the risk for “missing some malignant cysts in the initial evaluation and recurrent IPMNs after surgical resection,” they concluded. – by Adam Leitenberger
Disclosures: The researchers and editorial authors report no relevant financial disclosures.