基于钆塞酸二钠增强MRI高度异型增生结节分类对2018版LI-RADS诊断效能的影响

The Effect of High-Grade Dysplastic Nodule(HGDN)Categories on Diagnostic Performance of LI-RADS v2018 Criteria based on Gd-EOB-DTPA-Enhanced MRI

目的在肝细胞癌(HCC)高危因素背景下,探讨基于钆塞酸二钠(Gd-EOB-DTPA)增强MRI高度异型增生结节(HGDN)分类处理对2018版肝脏影像报告与数据系统(LI-RADS)诊断效能的影响。方法回顾性分析2015年1月至2021年12月在桂林医学院附属医院放射科,行Gd-EOB-DTPA增强MRI检查并术后经病理诊断的169例HCC高风险患者,共192个肝结节。两名放射科医师独立评估患者的影像征象,根据2018版LI-RADS分类标准将结节分类,统计分析LR-4/5、LR-5的诊断效能。为HGDN增加一个新分类LR-DN,将病理结果为HGDN的结节分类为LR-DN,其他结节分类不变,重新统计LR-4/5、LR-5的诊断效能。结果2018版LI-RADS分类LR-4/5、LR-5的灵敏度、特异度、准确率分别为92.81%、56.41%、85.42%和82.35%、79.49%、81.77%。新增LR-DN分类后,LR-4/5、LR-5的灵敏度无变化,特异度分别为87.18%、92.31%,准确率分别为91.67%、84.38%。结论Gd-EOB-DTPA增强MRI结合2018版LI-RADS对小肝癌(sHCC)的诊断效能较高,新增LR-DN分类可以提高LI-RADS分类对sHCC的诊断效能,同时加强临床对HGDN的分类管理、随访及诊疗策略。

Objective To investigate the effect of HGDN categories on diagnostic performance of LI-RADS v2018 criteria based on Gd-EOB-DTPA-enhanced MRI with high-risk patients for HCC.Methods From January 2015 to December 2021 in the Department of Radiology,Affiliated Hospital of Guilin Medical University,169 patients with high-risk for HCC underwent Gd-EOB-DTPA-enhanced MRI,total of 192 nodules diagnosed by pathological.Two radiologists evaluated the imaging features and classified according to LI-RADS v2018 criteria independently,calculate the diagnostic performance of LR-4/5,LR-5 based on consensus data.LR-DN was added for HGDNs,while other nodules remained unchanged,and the diagnostic performance of LR-4/5,LR-5 was recalculated.Results The sensitivity,specificity,accuracy of LR-4/5,LR-5 using the version of 2018 LI-RADS criteria were 92.81%,56.41%,85.42%and 82.35%,79.49%,81.77%,respectively.After the addition of the LR-DN,the sensitivity of LR-4/5,LR-5 unchanged,with accuracies of 91.67%,84.38%,specificities of 87.18%,92.31%,respectively.Conclusion Gd-EOB-DTPA-enhanced MRI combined with LI-RADS v2018 criteria may improve the sensitivity of LR-4/5,LR-5 for sHCC.The addition of the LR-DN classification may improve the sensitivity and specificity of LR-4/5,LR-5 and may help to strengthen the clinical classification management.