Utility of biliary microcrystal analysis in predicting composition of common bile duct stones

The high failure rate of contact litholytic therapy for common bile duct stones with currently available cholesterol solvents has been attributed to the inclusion of patients with pigment stones, as no pretreatment investigation is undertaken to distinguish the two stone types. In 36 patients with common bile duct stones we prospectively evaluated the utility of microscopic examination of bile collected from the biliary tree in predicting stone composition. The bile, obtained by means of either an endoscopically placed nasobiliary catheter (n = 27) or a surgically placed T-tube (n = 9), was subjected to microscopic examination, and findings were compared with the composition of stones retrieved subsequently. On the basis of quantitative infrared spectroscopy, stones were classified as cholesterol (n = 28) or pigment (n = 8) stones. The presence of cholesterol crystals in bile correctly identified 24 of 28 patients with cholesterol stones (sensitivity, 86%; specificity, 100%; positive predictive value, 100%). Calcium bilirubinate granules, when present alone (without cholesterol crystals), correctly identified all eight patients with pigment stones (sensitivity, 100%; specificity, 89%; positive predictive value, 73%). Thus, microscopic examination of bile from the biliary tree, if used as a pretreatment screening test, may help distinguish between patients with cholesterol and pigment stones and thereby considerably improve the results of litholytic therapy of common bile duct stones with currently available solvents.