MSCT3期增强扫描诊断肾透明细胞癌术前T分期的准确性与限度

The accuracy and limitations of preoperative T staging with three-phase contrast-enhanced multi-slice CT for clear cell renal cell carcinoma

目的 :探讨MSCT 3期增强扫描诊断肾透明细胞癌(clear cell renal cell carcinoma,cc RCC)术前T分期的准确性及其影响因素。方法:回顾性分析经手术病理证实的110例cc RCC患者术前MSCT 3期增强扫描资料,依据2010版TNM分期标准判断术前T分期并评价其诊断效能,分析与T分期相关的肾包膜、肾静脉和(或)下腔静脉、肾周和(或)肾窦脂肪、肾周筋膜受侵情况。结果:1CT术前T分期与病理分期有较好的一致性(K=0.811);2T3a期准确性(86.4%)、特异性(91.1%)和阴性预测值(83.6%)最低,而高估率(8.9%)最高;T4期敏感性(66.7%)和阳性预测值(66.7%)最低,而低估率(33.3%)最高;3共15例评估错误(低估66.7%)均与T3a期认定相关;肾窦脂肪、肾脏脂肪与肾包膜受侵的CT诊断漏诊率分别为57.1%、25.0%和15.8%。结论:MSCT 3期增强扫描对cc RCC的T分期诊断符合率较高,但T3a期的诊断价值相对较低。

Objective：To evaluate the diagnostic accuracy and limitations of multi-slice computed tomography（MSCT） for preoperative T-staging of clear cell renal cell carcinoma（cc RCC）. Methods：In this retrospective study,a total of 110 consecutive patients with cc RCC confirmed by histopathological results were assessed for the preoperative T-staging value using triphasic MDCT scan. And whether renal capsule,renal vein,inferior vena cava,renal sinus fat,perirenal fat,and perirenal fascia invaded by tumor was analyzed. Results ： 1 There was good consistency for T staging of cc RCC between the preoperative CT scan and pathological staging（Kappa=0.811）. 2The accuracy（86.4%）,the specificity（91.1%） and the negative predictive value（83.6%） of T3 astage were lowest;meanwhile the overestimate（8.9%） of T3 astage was highest. The sensitivity（66.7%） and the positive predictive value（66.7%） of T4 stage were lowest;meanwhile the undervalued rate（33.3%） of T4 stage was highest. 3A total of 15 patients were diagnosed with wrong T staging,which were all related to the T3 astaging and the undervalued rate was 66.7%. The missed rate of CT in the diagnosis of renal sinus fat,renal fat and renal capsule was 15.8%,25% and 57.1%respectively. Conclusion：Three-phase contrast-enhanced multi-slice CT scanning is an accurate method for preoperative T staging of RCC. But the diagnosis of T3 astage is still subject to certain restrictions.