早發與晚發B型肝炎相關肝細胞癌臨床與表觀基因表現之比較

Abstract

背景：B型肝癌病毒帶原者發生早發型與晚發型肝細胞癌間可能有不同的致癌機制，DNA甲基化已被認為和各種內源性與外源性因素，包括各種制癌物質暴露與發炎免疫反應相關。根據先前針對早發型肝細胞癌，周邊血液白血球DNA使用全基因體晶片找出的候選甲基化標記，本篇研究目的為：(a)比較早、晚發肝細胞癌在臨床、候選標記基因甲基化程度的差異，(b)使用早發型肝細胞癌病例-手足配對驗證先前所找出的候選標記，(c)利用早發型肝細胞癌病例-手足配對之DNA甲基化程度，建立肝細胞癌診斷模型，同時(d)評估該診斷模型是否可應用於晚發族群。 材料和方法：由多家醫學中心研究，蒐集了早發型肝細胞癌病例(n=48)-手足(n=59)配對與晚發型肝細胞癌病例(n=48)-手足(n=48)配對的臨床訊息與周邊白血球DNA，使用焦磷酸定序技術，於早發型肝細胞癌病例-手足配對，驗證所選出的20組基因標記，且建構鑑別模型，並使用AUC來評估所建立的鑑別模型之準確性。 結果：早發與晚發型肝細胞癌病患的臨床訊息在本篇研究中並無顯著差異，早發與晚發型族群間的甲基化程度有差異。早發型肝細胞癌病例-手足配對從20個甲基化探針中，成功驗證了7個探針。本篇研究以5個甲基化位點建構鑑別模型Ccon，在早發型病例-手足比較中，Ccon的AUC值為0.797(95% CI 0.705-0.890)，其中依據Youden index選擇最好的切點在8.7371，此切點的敏感度與特異度分別為79.2%與79.7%，在晚發型病例-手足比較中，Ccon的AUC值為0.716(95% CI 0.612-0.819)，其中依據Youden index選擇的切點在9.5474，此切點的敏感度與特異度分別為62.5%與83.3%。 結論：早發型與晚發型肝細胞癌的甲基化程度，會因為發展成癌症的原因不一樣而有所不同。Ccon可以有效的於年輕族群與年長族群進行肝細胞癌的檢測。

Background & Aims: There may be different mechanisms causing early- and late-onset hepatocellular carcinoma (HCC) in hepatitis B carriers. DNA methylation has been associated with a diversity of endogenous and exogenous factors, including various carcinogen exposures and inflammation/immune response. Based on candidate methylation markers indentified from a previous genome-wide scan of early-onset HCC on peripheral leukocyte DNA, the purpose of this study was four-fold: (a) to compare clinical and methylation status of candidate markers between early-and late-onset HCC; (b)to conduct a replication–study using an early-onset HCC-sibling case-control matched set; (c) to construct a DNA methylation classifier for diagnosis of early-onset HCC; as well as (d) to evaluate whether the constructed DNA methylation classifier for early-onset HCC could also be applicable in late-onset HCC. Method: Clinical data and leukocyte DNA sample were assessed in 48 early-onset HCC (vs. 59 siblings) and 48 late-onset HCC (vs. 48 siblings) case-sibling matched sets obtained from a multicenter collaboration study. We typed a total 20 candidate methylation probes and neighboring CpG sites by pyrosequencing technology. We examined whether methylation levels at identified CpG sites showed an association with under the receiver operating characteristic curve (AUC) to evaluate the discriminatory accuracy of the constructed DNA methylation classifier. Results: Clinical status were no significantly different between early- and late-onset HCC in this study. Methylation levels were different between early- and late-onset case-sibling set. We identified association between methylation in 7 probes of 20 candidate probs and HCC with early-onset HCC-sibling case-control matched set. We identified a methylation classifier (Ccon) containing five CpG sites that could detect hepatocellular carcinoma. In early-onset case-sibling study, Ccon had an AUC 0.797(95% CI 0.705-0.890) to discriminate individuals with hepatocellular carcinoma with a sensitivity of 79.2% and specificity of 79.7%. The AUC for Ccon in late-onset case-sibling study was 0.716(95% CI 0.612-0.819), sensitivity and specificity were 62.5% and 83.3%, respectively. Conclusion: There were different mechanisms causing methylation levels changed between early- and late-onset HCC in hepatitis B carriers. The methylation classifier Ccon could be valuable to detect early-onset and late-onset hepatocellular carcinoma.