摘要目的:探讨肝癌患者血清微小RNA（miR）-92的表达及与不良预后的关系。方法:回顾性分析2015年5月至2018年5月浙江省湖州市第一人民医院和浙江省湖州市中心医院收治的70例肝癌患者的临床资料，采用实时定量聚合酶链反应法检测miR-92表达水平，并与80例健康体检者比较。绘制受试者工作特征（ROC）曲线，评价血清miR-92诊断肝癌和早期（Ⅰ~Ⅱ期）肝癌的效能；分析血清miR-92表达水平与临床病理特征的关系；多因素Cox回归模型分析影响肝癌患者预后的独立危险因素。结果:肝癌患者血清miR-92表达水平明显高于健康体检者（4.10 ± 1.74比1.88 ± 0.78），差异有统计学意义（ P<0.05）。ROC曲线分析结果显示，血清miR-92诊断肝癌的曲线下面积（AUC）为0.874，最佳截断值为2.43，灵敏度为82.86%，特异度为86.25%；血清miR-92诊断早期肝癌的AUC为0.755，最佳截断值为2.38，灵敏度为68.57%，特异度为84.42%。血清miR-92表达水平与TNM分期和淋巴结转移相关（ P<0.01），而与性别、年龄、肿瘤直径、肝炎病史、肝硬化、分化程度及门静脉癌栓无关（ P>0.05）。血清miR-92高表达肝癌患者5年总生存率明显低于血清miR-92低表达肝癌患者（17.1%比31.5%），差异有统计学意义（ χ2 = 5.561， P<0.05）。多因素Cox回归分析结果显示，miR-92是影响肝癌患者预后的独立危险因素（ HR = 0.282，95% CI 1.179~3.562， P<0.05）。 结论:肝癌患者血清miR-92表达水平明显升高，在病情进展中发挥重要作用；血清miR-92可作为肝癌早期诊断、预后评估的指标。

abstractsObjective:To investigate the expression of serum microRNA (miR)-92 in patients with liver cancer and its relationship with poor prognosis.Methods:The clinical data of 70 patients with liver cancer in the First People′s Hospital of Huzhou City, Zhejiang Province and the Central Hospital of Huzhou City, Zhejiang Province from May 2015 to May 2018 were retrospectively analyzed. The miR-92 expression level was detected by real-time quantitative polymerase chain reaction method and compared with that of 80 healthy subjects. Receiver operating characteristic (ROC) curve was drawn to evaluate the efficacy of serum miR-92 in diagnosing liver cancer and early stage (stage Ⅰ to Ⅱ) liver cancer. The relationship between serum miR-92 expression level and clinicopathological characteristic was analyzed. Independent risk factors affecting the prognosis in patients with liver cancer were analyzed by multivariate Cox regression model.Results:The expression level of serum miR-92 in patients with liver cancer was significantly higher than that in healthy subjects (4.10 ± 1.74 vs 1.88 ± 0.78), and there was statistical difference ( P<0.05). ROC curve analysis result showed that the area under curve (AUC) of serum miR-92 for the diagnosis liver cancer was 0.874, the best cut-off value was 2.43, the sensitivity was82.86%, and the specificity was 86.25%; the AUC of serum miR-92 for the diagnosis early liver cancer was 0.755, the best cutoff value was 2.38, the sensitivity was 68.57%, and the specificity was 84.42%. The serum miR-92 expression level was related to TNM stage and lymph node metastasis ( P<0.01), but not related to gender, age, tumor diameter, history of hepatitis, liver cirrhosis, degree of differentiation and portal vein tumor thrombus ( P>0.05). The 5-year overall survival rate in liver cancer patients with high serum miR-92 expression was significantly lower than that in liver cancer patients with low serum miR-92 expression (17.1% vs. 31.5%), and there was statistical difference ( χ2 = 5.561, P<0.05). Multivariate Cox regression analysis result showed that miR-92 was an independent risk factor affecting the prognosis in patients with liver cancer ( HR = 0.282, 95% CI 1.179 to 3.562, P<0.05). Conclusions:The expression level of serum miR-92 in patients with liver cancer is significantly increased, which plays an important role in the progression of the disease. Serum miR-92 can be used as an indicator for early diagnosis and prognosis evaluation of liver cancer.