Diagnostic value of computerized nuclear morphometry for the prediction of malignancy in liver fine needle aspiration cytology.

ABSTRACT

OBJECTIVE: To evaluate the diagnostic value of nuclear parameters using computerized morphometry in the differential diagnosis of well-differentiated hepatocellular carcinoma (WD-HCC) and regenerative atypia in liver fine needle aspiration cytology (FNAC). STUDY DESIGN: All 78 cases, including 30 negative, 30 positive and 18 suspicious liver FNA cytology smears were examined with regard to nuclear area, nuclear perimeter, largest to smallest diameter ratio of the nuclei (L/S ratio) and coefficient of variation of the nuclear area (NACV). The diagnostic value was assessed using receiver operating characteristics (ROC) curves. Cutoff points were determined by the area under the ROC curve (AUC), which provides the highest combined sensitivity and specificity. RESULTS: The nuclear area and perimeter in liver FNA were more helpful than L/S ratio and NACV in discriminating between hepatocytes with regenerative atypia and WD-HCC. Cutoff points of 56.67 for nuclear area and 27.16 for perimeter were suggested to help shift the suspicious diagnoses to WD-HCC. CONCLUSION: When faced with difficult diagnoses on liver FNA, we suggest application of computerized morphometry may be a helpful ancillary diagnostic tool in the further classification of such cases.