DNA image analysis combined with routine cytology improves diagnostic sensitivity of common bile duct brushing

S. Krishnamurthy et al., DNA image analysis combined with routine cytology improves diagnostic sensitivity of common bile duct brushing, CANC CYTOP, 93(3), 2001, pp. 229-235
Citations number
Categorie Soggetti
Oncology,"Onconogenesis & Cancer Research
Journal title
ISSN journal
0008-543X → ACNP
Year of publication
229 - 235
SICI code
BACKGROUND. Cytologic evaluation of common bile duct brushings has a low se nsitivity for diagnosing malignancy because of scant cellularity, poor cell ular preservation, or sampling errors occur. The aim of this study was to e valuate whether cytology combined with image analysis improves the diagnost ic accuracy of bile duct brushing in comparison with cytology alone. METHODS. Forty-nine specimens of bile duct brushings obtained from 45 patie nts during endoscopic retrograde cholangiopancreatography were evaluated us ing cytology and image analysis. Specimens were classified as negative, aty pical, suspicious, or malignant by using cytologic evaluation. DNA histogra ms were classified as diploid (D), broad diploid (BD), aneuploid (A), or te traploid (T). Degree of hyperploidy (DH), representing cells with a DNA con tent > 5C was evaluated using a cutoff value of greater than or equal to 1% . Final diagnosis of cancer was based on tissue specimens that were obtaine d by fine-needle aspiration or surgical biopsy and clinical follow-up. RESULTS. Thirty-four patients ultimately proved to have a malignancy. Cytol ogy revealed 19 negative cases, 15 atypical cases, 9 suspicious cases, and 6 malignant cases. Together, suspicious and malignant cytology cases yielde d a sensitivity of 44% and a specificity of 100% for a cytologic diagnosis of cancer. The DNA histogram pattern was D in 24 cases, ED in 9 cases, and A in 16 cases. ED and A patterns were significantly associated with maligna ncy (P < 0.001). A DH > 1% was noted in 22 cases. DH alone had a sensitivit y of 62% and a specificity of 91% and was significantly associated with mal ignancy (P < 0.004). Atypical cytology alone had a false-negative rate of 2 9%, but in combination with a DH <greater than or equal to> 1%, the false-n egative rate decreased to 7%. Additionally, when the authors combined atypi cal, suspicious, and malignant cytology with a DH > 1%, the diagnostic sens itivity increased to 88%, but the specificity decreased to 73%. CONCLUSIONS. Combined cytology and image analysis of bile duct brushing inc reased diagnostic sensitivity compared with cytology alone. The findings su ggest that image analysis may help select patients having atypical cytology who should undergo a more rigorous evaluation for malignancy. A larger pro spective study of the usefulness of combined cytology and image analysis of bile duct brushing is warranted. Cancer (Cancer Cytopathol) 2001;93:229-23 5. (C) 2001 American Cancer Society.