RESUMEN
BACKGROUND/AIMS: Brush cytology during ERCP has been reported to have a low sensitivity. A new device, Howell biliary system (Wilson-Cook), enables targeted biopsies for histopathologic assessment. The aim was to compare histopathology with brush cytology. METHODOLOGY: Brush cytology followed by biopsies obtained by the Howell device was taken consecutively from bile duct strictures. Coded slides were scored by 3 pathologists and 2 cytologists in a 3-graded scale; 2: benign; 3: suspicious of malignancy; 4: malignant. The clinical outcome including autopsy served as the gold standard for the definite diagnoses. RESULTS: Twenty-one malignant and 6 benign strictures were evaluated. The histopathology revealed 11 out of the 21 malignant as certain or suspected malignant (score > or = 3) (sensitivity: 0.52). The cytology scored 17 out of 21 > or = 3 (sensitivity: 0.80). The in pair kappa values for the 3 pathologists were: (0.37; 0.26; 0.41) vs. 0.56 for the 2 cytologists. Among the evaluable strictures the pathologists scorings were; (median: 3.0, SD: 0.72) for the malignant and (median: 2.3, SD: 0.98) for the benign (p = 0.27) and the cytology scorings were; (median: 3.5, SD: 0.73) for the malignant and (median: 2.7, SD: 0.65) for the benign (p = 0.09). CONCLUSIONS: Brush cytology has a higher accuracy than the targeted biopsies and should be used in combination with other methods to reach the correct diagnosis.