OBJECTIVES: Biliary brushing during endoscopic retrograde cholangiography (ERC) may allow the diagnosis of cancer, but this technique is limited by a low (18-57%) sensitivity. We compared the cellular and diagnostic yields of a new sampling method (using stricture dilation and a grasping basket) with those of brushing.
METHODS: Sixty consecutive patients with a suspected malignant obstruction of the common bile duct were included in a consecutive, nonrandomized order. They underwent sampling during ERC using the new method (study patients, N = 30) or a brush (control patients, N = 30). A final diagnosis of cancer was made in 22/30 (73%) study patients and 21/30 (70%) control patients. At the end of the study period, all smears were interpreted at cytopathological examination in a blind and random fashion for cellularity (insufficient, low, moderate, or high) and diagnosis (normal, atypical considered reactive, highly atypical suspicious for cancer, or malignant). "Highly atypical" and "malignant" diagnoses were considered as indicative of cancer.
RESULTS: Compared with biliary brushing, the new method provided a high cellular yield in more cases (19/30 vs 10/30, P= 0.020), a higher sensitivity for the detection of cancer (19/22 [86%] vs 11/21 [52%], P= 0.015), and a higher diagnostic accuracy (27/30 [90%] vs 20/30 [67%], P= 0.028). Specificity for the detection of cancer was 100%. The 30-d complication rate observed with the new method was 3%.
CONCLUSIONS: Significantly higher cellular and diagnostic yields can be obtained at ERC in patients with suspected malignant biliary stricture by combining stricture dilation and a grasping basket compared with brushing.