A study of the clinical utility of a 20-minute secretin-stimulated endoscopic pancreas function test and performance according to clinical variables.

ABSTRACT

BACKGROUND AND AIMS: Direct pancreas juice testing of bicarbonate, lipase, or trypsin after stimulation by secretin or cholecystokinin is used to determine exocrine function, a surrogate for diagnosing chronic pancreatitis (CP). Endoscopic pancreas function tests (ePFTs), where a peak bicarbonate concentration (PBC) ≥80 mEq/L in pancreas juice is considered normal, are now used more frequently. In this ePFT, aspirates start 35 minutes after secretin administration because pancreas output peaks 30 minutes after secretagogue administration. The performance of ePFT in a cohort of patients with a presumptive diagnosis of CP referred to a pancreas clinic for consideration of an intervention including total pancreatectomy and islet autotransplantation was studied, compared with EUS, ERCP, histology, and consensus diagnosis. The effect of sedation, narcotic use, aspirate volume, body mass index, age, and proton pump inhibitors (PPIs) on test performance is reported. METHODS: After a test dose, synthetic human secretin was administered intravenously, and 30 minutes later sedation was achieved with midazolam and fentanyl or propofol. A gastroscope was advanced to the major papilla where 4 continuous aspiration samples were performed at 5-minute intervals in sealed bottles. PBC ≥80 mEq/L was normal. RESULTS: Eighty-one patients had ePFTs from August 2010 through October 2015. Twenty-seven patients (33%) were diagnosed with CP. Eighteen of the 27 patients with CP and 1 of the 54 patients without CP had an abnormal ePFT, producing a sensitivity of 66% (95% CI, 46.0-83.5), specificity 98% (95% CI, 90.1-99.9), positive predictive value 94.7% (95% CI, 74-99.9), and negative predictive value 85.5% (95% CI, 74.2-93.1). ERCP and PBC concordance was generally poor, but none of the patients without CP had major EUS changes, and only 3 patients with a PBC <80 mEq/L had a normal EUS. The PBC was affected by narcotics and PPI use. CONCLUSION: A 20-minute ePFT after secretin administration had a marginal sensitivity for diagnosis of CP. The diagnosis of CP should not rely on a single study and certainly not a PFT. The duodenal aspirate volume did not correlate with the PBC, which contrasts with current secretin-enhanced MRCP knowledge; therefore, further studies on this subject are warranted. Neither type of sedation, BMI, nor age affected test performance. Narcotics and PPIs may affect the PBC, so borderline results should be interpreted with caution in these groups.