A randomized crossover study of secretin-stimulated endoscopic and dreiling tube pancreatic function test methods in healthy subjects

OBJECTIVES: We have developed an endoscopic method of secretin endoscopic pancreatic function testing (ePFT) to simplify duodenal fluid collection. Validation of the ePFT requires a direct comparison to the traditional PFT using a Dreiling tube (DT). Our aim was to compare bicarbonate concentrations [HCO₃^-] obtained by the ePFT and DT methods in healthy subjects (HS). METHODS: HS were randomized to either DT or ePFT, then crossed over to the other test after a minimum 1-wk washout. An age/weight-based sedation bolus was used for each test. DT protocol: Endoscopic placement of a DT was confirmed by fluoroscopy. After a baseline 15-min collection and administration of IV synthetic secretin, fluid was continuously collected in 15-min aliquots for an hour. ePFT protocol: Endoscopy was performed using a 6-mm endoscope. After gastric aspiration and discard and IV secretin, duodenal aspirates were obtained every 15-min for an hour. Fluid specimens were auto-analyzed for [HCO₃^-]. RESULTS: Twelve HS were enrolled (6F, mean age 37 yr). The difference in [HCO₃^-] between the two methods was not significant at the 0-, 30-, 45-, or 60-min collections. An excellent correlation in peak [HCO₃^-] was observed (R²= 0.84, p < 0.001). Using a peak [HCO₃ ^-] cutpoint 80 mEq/L, there was 100% agreement between the methods; using cutpoint 90 mEq/L, there was 83% agreement. CONCLUSIONS: The accuracy of the ePFT is similar to DT: There were minimal differences in [HCO ₃^-] at each of the timed collections and at peak. There is an excellent correlation in peak [HCO₃^-] and high level of diagnostic agreement between the tests.