Endobiliary brush biopsy: Intra- and interobserver variation in cytological evaluation of brushings from bile duct...

Obtaining cytological specimens by wire-guided endobiliary brushing at the time of endoscopic retrograde cholangiopancreatography (ERCP) is a convenient way to reach a diagnosis. Sensitivity for malignant disease is generally around 50% and specificity around 100%. The present study was designed to assess the reproducibility of the cytological examination.

Samples were obtained from 55 consecutive patients with biliary duct strictures that eventually turned out to be caused by malignant disease in 41 patients (73%). The cytology specimens were evaluated twice in different random order with an interval of at least 4 months by two pathologists blinded to the final diagnoses. Suitability for diagnosis (suitable, suboptimal or unsuitable) and cytologic diagnosis (benign, atypical, suspicious for malignancy and malignant cells) were registered. Kappa analysis was applied.

Regarding suitability for diagnosis, the two observers reproduced their findings in 84% (kappa 0.76) and 87% (0.59) (p>0.05), while the interobserver variations on the two occasions were 76% and 78% (kappa 0.49 and 0.58, respectively) (p>0.05). Five percent of samples were rated as inadequate for diagnosis in at least one reading; 18% were suboptimal and 62-82% were suitable. The observers reproduced their cytological diagnosis in 77% and 71% (weighted kappa 0.83 and 0.75) (p>0.05), and agreed on the cytological diagnosis in 65% and 73% (weighted kappa 0.65 and 0.75) (p>0.05).

The intra- and interobserver agreement in cytological evaluation of endobiliary brushings from bile duct strictures is generally good. The rates of inadequate and suboptimal samples can probably be reduced by modifications of the brush design.