Endoscopy 2020; 52(S 01): S297
DOI: 10.1055/s-0040-1704947
ESGE Days 2020 ePoster presentations
Colon and rectum 09:00–17:00 Thursday, April 23, 2020 ePoster area
© Georg Thieme Verlag KG Stuttgart · New York

DIAGNOSTIC YIELD OF BRUSH CYTOLOGY AT ERCP FOR BILE DUCT STRICTURES

G Demetriou
1   University Hospital of Heraklion, Gastroenterology, Heraklion, Greece
,
E Lappa
2   University Hospital of Heraklion, Cytology, Heraklion, Greece
,
A Veniamin
1   University Hospital of Heraklion, Gastroenterology, Heraklion, Greece
,
A Machaira
1   University Hospital of Heraklion, Gastroenterology, Heraklion, Greece
,
V Valatas
1   University Hospital of Heraklion, Gastroenterology, Heraklion, Greece
,
A Kalogeraki
2   University Hospital of Heraklion, Cytology, Heraklion, Greece
,
D Tamiolakis
2   University Hospital of Heraklion, Cytology, Heraklion, Greece
,
E Kalaitzakis
1   University Hospital of Heraklion, Gastroenterology, Heraklion, Greece
› Author Affiliations
Further Information

Publication History

Publication Date:
23 April 2020 (online)

 
 

    Aims Commonly brush cytology samples are obtained from bile duct strictures during ERCP. According to a recent meta-analysis, their sensitivity to exclude cancer ranges 40%-50% with a specificity of 98%-100% (Gastrointest Endosc 2015). Our aim was to perform a clinical audit evaluating the diagnostic utility of brush cytology samples from patients undergoing ERCP due to bile duct strictures in our institution.

    Methods All patients who underwent ERCP in 2002–2018 due to obstructive jaundice and were found to have a bile duct stricture from which brush cytology samples were obtained were retrospectively enrolled. Cytology results were reviewed and classified as a.non-diagnostic, b.negative for malignancy, c.atypia, d.suspicious for malignancy, or e.positive for malignancy. Diagnosis of cancer after cyto- or histopathology obtained in other procedures or clinical follow-up after a 12-month period served as the diagnostic gold standard.

    Results A total of 84 specimens were obtained from 79 patients (median age 72, 56% male). Overall, 37 samples(44%) were negative for malignancy, 15(18%) showed atypia, 9(11%) were suspicious and 21(25%) were positive for malignancy. Two samples were non-diagnostic(2%). A final diagnosis of pancreaticobiliary cancer was put in 53(67%) (gold standard). Considering suspicious specimens as negative for malignancy, the sensitivity and specificity of brush cytology was 39%(95%CI, 25%-53%) and 94% (95%CI, 80%-99%), respectively. Considering suspicious specimens as positive, sensitivity was 55%(95%CI, 37%-73%) and specificity 90%(95%CI, 82%-98%).

    Conclusions The observed sensitivity(39%-55%) and specificity(90-94%) of brush cytology from bile duct strictures is in line with the current literature. The relatively low specificity may be attributed to specimen-processing pitfalls, or errors upon cytopathology review. To increase the diagnostic yield of brush cytology, we proposed the use of processing methods such as Thinprep, including a more detailed patient history with relevant imaging findings in the cytopathology referral, and using international nomenclature in the cytopathology report.


    #