BACKGROUND: Distinguishing dysplasia from hepatocellular carcinoma (HCC) by fine-needle aspiration (FNA) cytology is difficult. The aim of this study was to diagnose HCC and the distinction of liver cell dysplasia from HCC with different grades by interpreting and scoring the cyto-morphological features. METHODS: Eighty-three cirrhotic patients undertook a sonography-guided FNA and subsequent needle biopsy for the tumor. HCC was confirmed in 68 cases and cirrhosis with dysplasia in 15 cases by pathology and follow-up for longer than 2 years. Eighteen cytological features were scored as degree of one, two or three according to their presence or prominence. RESULTS: Two cases of well-differentiated HCC were diagnosed as negative for HCC initially. The sensitivity, specificity, false positive, false negative and accuracy were 97%, 100%, 0%, 3% and 97.6% for FNA cytology in the diagnosis of HCC, respectively. The score of dysplasia was 20.8 +/- 1.3 (mean +/- SD) and lower than 26.2 +/- 3.4 in Edmondson's grade I HCC (P < 0.01), 28.9 +/- 2.9 in grade II HCC (P < 0.01), and 34.9 +/- 4.3 in grade III/IV HCC (P < 0.01). The score was also significantly lower in grade II HCC than in grade III/IV HCC (P < 0.01). CONCLUSIONS: FNA yielded a high accuracy in the distinction of dysplasia from HCC with different grades. There is a good correlation in cyto-morphological scores of liver cell dysplasia and HCC with different grades. Dysplasia displayed the lowest score and the score increased in order from dysplasia to grade III/IV HCC.
BACKGROUND: Distinguishing dysplasia from hepatocellular carcinoma (HCC) by fine-needle aspiration (FNA) cytology is difficult. The aim of this study was to diagnose HCC and the distinction of liver cell dysplasia from HCC with different grades by interpreting and scoring the cyto-morphological features. METHODS: Eighty-three cirrhoticpatients undertook a sonography-guided FNA and subsequent needle biopsy for the tumor. HCC was confirmed in 68 cases and cirrhosis with dysplasia in 15 cases by pathology and follow-up for longer than 2 years. Eighteen cytological features were scored as degree of one, two or three according to their presence or prominence. RESULTS: Two cases of well-differentiated HCC were diagnosed as negative for HCC initially. The sensitivity, specificity, false positive, false negative and accuracy were 97%, 100%, 0%, 3% and 97.6% for FNA cytology in the diagnosis of HCC, respectively. The score of dysplasia was 20.8 +/- 1.3 (mean +/- SD) and lower than 26.2 +/- 3.4 in Edmondson's grade I HCC (P < 0.01), 28.9 +/- 2.9 in grade II HCC (P < 0.01), and 34.9 +/- 4.3 in grade III/IV HCC (P < 0.01). The score was also significantly lower in grade II HCC than in grade III/IV HCC (P < 0.01). CONCLUSIONS: FNA yielded a high accuracy in the distinction of dysplasia from HCC with different grades. There is a good correlation in cyto-morphological scores of liver cell dysplasia and HCC with different grades. Dysplasia displayed the lowest score and the score increased in order from dysplasia to grade III/IV HCC.