The present figure for accuracy of ERCP-based brush cytology or forceps biopsy for tissue diagnosis is relatively low, usually not exceeding 70%.
In comparison the reported accuracy rates of EUS-guided FNA of pancreatobiliary masses are higher, at 80%.
In a prospective study, Dr Werner and colleagues compared these two modalities used in diagnosis of indeterminate biliary strictures and pancreatic tumors.
50 consecutive patients (29 men, 21 women; mean age 62.1 years) with obstructive jaundice in whom a tissue diagnosis was required were recruited to take part in the study.
During ERCP, intraductal specimens were obtained with a forceps and with two different types of brush (conventional and spiral suction) in random order.
During EUS, only visible mass lesions or localized bile duct wall thickening were aspirated (22-gauge needle), with at least two passes yielding material sufficient for assessment.
A cytopathologist was not present in the procedure room to evaluate specimen adequacy.
The reference methods used were surgery, other biopsy results, follow-up until death, or the conclusion of the study (mean follow-up 20 months).
The final diagnoses were malignancy, 28 (16 pancreatic, 12 biliary), and benign biliary stricture, 22.
Sensitivity and specificity for ERCP-guided biopsy were 36% and 100%, respectively; while for ERCP-guided cytology (when using conventional and spiral suction brushes) they were 46% and 100%, respectively.
For EUS-guided FNA, sensitivity and specificity were 43% and 100%, respectively.
When the researchers considered punctured lesions alone, the sensitivity of EUS-guided FNA was 75%.
They found that, in general, sensitivity was better for ERCP-based techniques in the subgroup biliary tumor (ERCP 75% vs. EUS 25%), whereas EUS-guided biopsy was superior for pancreatic mass (EUS 60% vs. ERCP 38%).
Commenting on their findings, Dr Werner said, "For biliary strictures, combined ERCP- and EUS-guided tissue acquisition seems to be the best approach to tissue diagnosis."
He added, "From a clinical standpoint, it appears reasonable, when a tissue diagnosis is required, to start with ERCP if biliary malignancy is suspected and with EUS when a pancreatic tumor is thought to be the cause of a biliary stricture."