术前超声微血流分级预测肝细胞癌微血管侵犯的价值

The value of preoperative ultrasonic microflow grading in predicting microvascular invasion of hepatocellular carcinoma

目的探讨术前超声微血流分级预测肝细胞癌(HCC)患者微血管侵犯(MVI)的价值。方法2018年8月至2021年8月鹤煤总医院收治的HCC患者138例,根据手术病理结果将患者分为有MVI组60例和无MVI组78例。术前采用免疫荧光法检测血清甲胎蛋白(AFP)水平,同时行超声微血流成像及造影记录肿瘤大小、微血流分级、微血流分布、肿瘤边缘形态、瘤周强化等影像学特征。分析MVI发生的独立预测因素,探讨血清AFP水平联合影像学特征对MVI的预测效能。结果有MVI组血清AFP水平≥400μg/L的患者17例(28.3%),无MVI组为10例(12.8%)(P<0.05)。有MVI组肿瘤>5cm、微血流分级为Ⅲ级、肿瘤边缘不光滑和瘤周强化的占比分别为58.3%(36/60)、40.0%(24/60)、83.3%(50/60)和51.7%(31/60),高于无MVI组的34.6%(27/78)、20.5%(16/78)、51.3%(40/78)和17.9%(14/78),差异均有统计学意义(P<0.05)。血清AFP水平、肿瘤大小、微血流分级、肿瘤边缘不光滑和瘤周强化是HCC患者MVI的独立预测因素(P<0.05)。血清AFP水平联合影像学特征(包括肿瘤大小、微血流分级、肿瘤边缘不光滑和瘤周强化)预测MVI的曲线下面积(AUC)、敏感度、特异度和准确度分别为0.954、93.3%、97.4%和95.7%。结论除血清AFP水平、肿瘤边缘不光滑、肿瘤大小、瘤周强化外,术前超声微血流分级也是HCC患者MVI的独立预测因素,联合上述指标可提高预测效能。

Objective To investigate the value of preoperative ultrasonic microflow grading in predicting microvascular invasion(MVI)in patients with hepatocellular carcinoma(HCC).Methods A total of 138 HCC patients were admitted to our hospital between August 2018 and August 2021.They were divided into MVI group(60 cases)and MVI group(78 cases)according to the pathological results.Serum alpha-fetoprotein(AFP)level was detected by immunofluorescence method before surgery.At the same time,ultrasound microflow imaging and contrast imaging were performed to record the tumor size,microflow grade,microflow distribution,tumor edge morphology,peritumor enhancement and other imaging characteristics.The independent predictive factors of MVI were analyzed,and the predictive efficacy of serum AFP level combined with imaging characteristics of MVI was investigated.Results The proportion of serum AFP level≥400μg/L in MVI group was 28.3%,which was significantly higher than that of 12.8%in MVI group(P<0.05);The proportions of tumor size>5 cm,gradeⅢmicroflow grade,unsmooth tumor margin and peritumoral enhancement in MVI group were 58.3%,40.0%,83.3%and 51.7%,respectively,which were significantly higher than those of 34.6%,20.5%,51.3%and 17.9%in MVI group(P<0.05).Serum AFP level,tumor size,micro-blood flow grade,unsmooth tumor margin and peritumoral enhancement were independent predictors of MVI in HCC patients(P<0.05).The area under curve(AUC),sensitivity,specificity and accuracy of serum AFP level combined with imaging features(including tumor size,microflow grading,tumor margin unsmoothness and peritumoral enhancement)in predicting MVI were 0.954,93.3%,97.4%and 95.7%,respectively.Conclusion In addition to serum AFP level,unsmooth tumor margin,tumor size and peri-tumor strength,preoperative ultrasonic microflow grade was also an independent predictor of MVI in HCC patients.The prediction efficiency could be further improved by a combination of the above indicators.