基于增强MRI影像学征象和血清学指标预测肝细胞癌微血管侵犯的价值

The value of predicting microvascular invasion in hepatocellular carcinoma based on enhanced MRI and serological indices

目的:探究肝细胞癌(HCC)增强MRI影像学征象和血清学指标与微血管侵犯(MVI)的关系,旨在术前预测MVI,为评估患者预后和指导个性化治疗决策提供参考。方法:回顾性纳入84例经手术病理证实的HCC患者。采用“7点”基线取材法获取病理组织,通过免疫组织化学方法检测MVI,将患者划分为MVI阳性组和MVI阴性组。两位放射科医师回顾患者术前MRI增强图像,独立评估MRI影像学特征。采用单因素分析和多因素逻辑回归分析探究HCC的临床病理学特征、血清学指标和MRI影像学特征与MVI关系。结果:84例患者中,34例(40.5%)为MVI阳性,50例(59.5%)为MVI阴性。MVI阳性患者的Ki-67标记指数更高(P=0.027),甲胎蛋白(AFP)水平更高(P=0.001),肿瘤长轴直径更大(P<0.001),短轴直径更大(P<0.001),更易出现动脉期瘤内强化小动脉(P=0.013)、瘤周强化(P=0.002)、瘤周无或不完整强化包膜(P=0.039)、延迟期不光滑的肿瘤边界(P=0.001)和微静脉侵犯双特征预测因子(TTPVI)(P<0.001)。两位放射科医师评估MRI影像学特征具有良好的观察者间一致性。多因素逻辑回归分析显示,AFP、肿瘤长轴直径和TTPVI是与MVI独立相关的因素。联合三个特征诊断MVI的曲线下面积达到0.864(95%CI:0.767~0.881),敏感度为75.8%,特异度为96.0%,阳性预测值为92.6%,阴性预测值为85.7%。结论:AFP、肿瘤长轴直径和TTPVI是预测MVI的独立因素,联合患者血清学特征和MRI肿瘤影像学特征对HCC患者的MVI状态具有良好的预测价值,有潜力为HCC治疗决策提供参考依据。

Objective:To investigate the relationships of serological indicators and contrast-enhanced MRI features with microvascular invasion(MVI)in hepatocellular carcinoma(HCC),with the aim of preoperative prediction of MVI and providing reference for evaluating patient prognosis and guiding personalized treatment.Methods:Eighty-four patients with pathologically-confirmed HCC were retrospectively enrolled.Pathological tissues were obtained using the"7-point"baseline sampling method.The patients were divided into MVI positive and negative groups.Two radiologists retrospectively analyzed preoperative MRI and independently evaluated imaging features.Univariate analysis and multivariate logistic regression analysis were used to investigate the relationships of clinicopathologic characteristics,serological indicators and MRI features with MVI.Results:Of the 84 patients,34(40.5%)were MVI-positive and 50(59.5%)were MVI-negative.MVI positive patients had a higher Ki-67 labelling index(P=0.027),a higher level of AFP(P=0.001),a larger tumour diameter in the long axis(P<0.001),a larger diameter in the short axis(P<0.001),and a greater tendency to develop arterial stage intratumoural enhancement of small arteries(P=0.013),peritumoural enhancement(P=0.002),no or incomplete peritumoural enhancement of the peritumoural envelope(P=0.039),delayed stage non-smooth tumour borders(P=0.001)and two-trait predictor of venous invasion(TTPVI)(P<0.001).There was good inter-observer agreement between the two radiologists assessing MRI imaging features.Multivariate logistic regression analysis revealed that MVI was independently associated with AFP,tumor size and TTPVI in patients with HCC.The area under the curve for combining the three features to diagnose MVI reached 0.864(95%CI:0.767~0.881),with a sensitivity of 75.8%,a specificity of 96.0%,a positive predictive value of 92.6%,and a negative predictive value of 85.7%.Conclusion:AFP,tumor size and TTPVI are independent factors for predicting MVI.The proposed combined model achieved a promising performance for noninvasively predicting MVI in HCC,with the potential to provide a reference basis for decision-making on HCC treatment.