Search

The results of bile salt treatment in patients with radiolucent stones and a functioning gall bladder have been poor. In 42 of these patients awaiting cholecystectomy we determined the value of duodenal bile examination in predicting gall stone composition, and thus identifying those less likely to respond to bile salt therapy. Based on chemical analysis and scanning electron microscopy, 28 of 42 (67%) gall stones retrieved at surgery were potentially insoluble. Microscopic examination of duodenal bile correctly identified 21 (75%) of them: it predicted all four (100%) pigment stones, three of six (50%) calcium carbonate containing cholesterol stones, and 14 of 18 (78%) cholesterol stones with pigment shells. It was nearly as reliable as microscopic examination of bile aspirated directly from the gall bladder during surgery (21 (75%) v 23 (82%); p = NS). Furthermore, the presence of cholesterol crystals in duodenal bile was a more sensitive indicator than chemical detection of supersaturation (34 of 38 (89%) v 25 of 35 (71%); p < 0.05) for prediction of cholesterol gall stones. Microscopic examination of duodenal bile, if used as a screening test, could help to exclude potential non-responders and thereby improve considerably the results of oral bile salt treatment for gall stone dissolution.

Thirty-nine patients with cholelithiasis were prospectively studied to evaluate the qualitative and quantitative differences between duodenal bile and gallbladder bile. Duodenal bile obtained before cholecystectomy by nasoduodenal intubation and ceruletide injection was qualitatively similar to gallbladder bile obtained during surgery. Microscopic cholesterol crystals as an indicator of cholesterol gallstones (n = 35) could be detected in 31 (89%) and 35 (100%; p = NS), respectively. Moreover, there was no difference in the molar percentage of three biliary lipids and the mean cholesterol saturation index (1.54 +/- 0.72 and 1.74 +/- 0.42; p = NS) of the two sources of bile. Duodenal bile was, however, dilute as compared with gallbladder bile, as evidenced by lower cholesterol crystal counts (167 +/- 247 versus 705 +/- 978; p < 0.01), lower total lipid concentration (5.8 +/- 2.7 versus 11.1 +/- 5.6 g/dl; p < 0.001), and lower concentrations (in mmol/l) of the three bile lipids--that is, total bile acids, phospholipids and cholesterol (p < 0.001). Good concentrated bile (total lipid concentration > or = 5 g/dl) could be obtained in 74% of duodenal bile samples, compared with 90% of gallbladder bile (p = NS). Our study shows that, although duodenal bile is dilute as compared with gallbladder bile, it is qualitatively similar to gallbladder bile and, because of the ease and safety of its collection, can be used to study serial alterations in biliary composition in individual subjects.