本研究利用肝癌組織及正常組織間基因表現程度的差異，建立一個能區辨此兩種組織的診斷指標。從文獻中挑選30個肝癌的相關基因，應用即時反轉錄聚合?連鎖反應，測量這些基因於細胞株Hep-G2, Huh7及4位HBV-positive HCCs組織的表現量。從30個基因中挑選出9個表現程度較顯著的基因。然後測定此9個基因在20 HBV-positive HCCs癌組織及正常組織的表現量。將這些基因表現量進行單變數分析，並計算ROC曲線下面積，並以上述分析結果作為變數選擇條件。結果我們從這9個基因中選出5個基因進行多變量分析；亦即利用條件式邏輯斯迴歸(conditional logistic regression)，對基因表現量做組合，以得到一綜合性診斷指標。若以指標分數65當作診斷切點，則可以得到敏感度為95%及特異度為80%之回溯性診斷切點。本研究建立的診斷指標，能區辦分化良好(well-differentiated)的癌組織及正常組織之不同，有助於提供依據，作為初發階段之肝癌病人不易區辨或確診的情形下，另一診斷方法的探討。; The aim of this study was to develop and validate a molecular index for the diagnosis of hepatocellular carcinoma (HCC) based on genes whose specificity and level of expression are the most discriminating for HCC. The level of expression of 30 genes was assessed by a real-time reverse transcription-polymerase chain reaction approach at Hep-G2, Huh7 and 4 HBV-positive HCCs. The most informative genes were selected as diagnostic indices. As a result, 9 out of the 30 genes were differentially expressed in HCC. Moreover, 5 out of these 9 genes were eventually selected according to their performance and univariate analysis results. A multivariate analysis was then used to obtain an expression for the diagnostic index by conditional logistic regression. Using a cut-off score of 65, sensitivity and specificity are 95% and 80%, respectively, in 20 HBV-positive HCCs. It may offer a basis for future studies on the differentiation of ambiguous cases with early-stage HCCs.
