The diagnostic yield of endoscopic ultrasound (EUS) in non-jaundiced patients with unexplained biliary dilation has been reported to be 6-33%. In our experience, patients with unexplained biliary dilation who use chronic narcotic analgesics might have an even lower diagnostic yield at EUS. However, this correlation has not been well-described as yet.
We reviewed electronic medical records of patients referred to our center for EUS for the primary indication of unexplained biliary dilation between July 2007 and June 2017. Our database search yielded 767 potential subjects. Exclusion criteria included pancreaticobiliary masses, stones, or calcific chronic pancreatitis seen on prior imaging, a total serum bilirubin >3 mg/dl, or alkaline phosphatase (AP) > 1.5x the upper limit of normal. Data were also separately abstracted on patients with liver tests exceeding the above criteria, yet who had no other exclusion criteria.
41 patients met all inclusion and exclusion criteria. The mean age was 62.4 ± 14.5 years, 37 (90.2%) were female, the mean bile duct diameter at EUS was 12.2 ± 3.8 mm, total bilirubin was 0.6 ± 0.3 mg/dL and AP was 112 ± 62.9 U/L. Most patients (33/41, 80.5%) were symptomatic, reporting abdominal pain or weight loss. A majority (29/41, 70.7%) of patients used oral narcotics. EUS demonstrated choledocholithiasis in 3/29 patients (10.3%) using narcotics and 1/12 (8.3%) not using narcotics. No masses or other relevant pathology was observed. This patient cohort was then combined with 39 additional patients with unexplained biliary dilation who had abnormal liver biochemical tests. Multivariable logistic regression analysis revealed that patients with abnormal liver biochemical tests were 3 times more likely to have positive EUS finding (OR 2.9, CI [1.5,5.7], P = 0.002). Patients using chronic narcotic analgesics were less likely to have positive EUS finding (OR 0.3, CI [0.1, 1.1], P = 0.06).
A majority of patients referred for EUS to investigate biliary dilation without biochemical cholestasis were women using chronic narcotic analgesics. The diagnostic yield of EUS in this population was low (9.8%) and no neoplasia was found. In contrast, the yield of EUS in patients with biliary dilation and elevated liver tests was strikingly higher at 85%. Use of chronic narcotic analgesics appeared to predict a reassuring EUS examination.