Department of Radiology, University of California Davis Medical Center, Sacramento, CA 95817, USA.
Cardiovasc Intervent Radiol. 2011 Apr;34(2):306-18. doi: 10.1007/s00270-010-9938-3. Epub 2010 Aug 4.
The primary purpose of this study was to demonstrate intraobserver/interobserver reproducibility for novel semiautomated measurements of hepatic volume used for Yttrium-90 dose calculations as well as whole-liver and necrotic-liver (hypodense/nonenhancing) tumor volume after radioembolization. The secondary aim was to provide initial comparisons of tumor volumetric measurements with linear measurements, as defined by Response Evaluation Criteria in Solid Tumors criteria, and survival outcomes.
Between 2006 and 2009, 23 consecutive radioembolization procedures were performed for 14 cases of hepatocellular carcinoma and 9 cases of hepatic metastases. Baseline and follow-up computed tomography obtained 1 month after treatment were retrospectively analyzed. Three observers measured liver, whole-tumor, and tumor-necrosis volumes twice using semiautomated software.
Good intraobserver/interobserver reproducibility was demonstrated (intraclass correlation [ICC] > 0.9) for tumor and liver volumes. Semiautomated measurements of liver volumes were statistically similar to those obtained with manual tracing (ICC = 0.868), but they required significantly less time to perform (p < 0.0001, ICC = 0.088). There was a positive association between change in linear tumor measurements and whole-tumor volume (p < 0.0001). However, linear measurements did not correlate with volume of necrosis (p > 0.05). Dose, change in tumor diameters, tumor volume, and necrotic volume did not correlate with survival (p > 0.05 in all instances). However, Kaplan-Meier curves suggest that a >10% increase in necrotic volume correlated with survival (p = 0.0472).
Semiautomated volumetric analysis of liver, whole-tumor, and tumor-necrosis volume can be performed with good intraobserver/interobserver reproducibility. In this small retrospective study, measurements of tumor necrosis were suggested to correlate with survival.
本研究的主要目的是展示用于钇-90 剂量计算的新型半自动肝体积测量的观察者内/间可重复性,以及放射性栓塞后全肝和坏死性(低密/无增强)肿瘤体积。次要目的是提供肿瘤体积测量与实体瘤反应评价标准定义的线性测量的初步比较,并提供生存结果。
在 2006 年至 2009 年间,对 14 例肝细胞癌和 9 例肝转移患者进行了 23 例连续的放射性栓塞治疗。回顾性分析治疗后 1 个月获得的基线和随访 CT。三名观察者使用半自动软件两次测量肝、全肿瘤和肿瘤坏死体积。
肿瘤和肝体积的观察者内/间可重复性良好(组内相关系数[ICC]>0.9)。肝体积的半自动测量与手动追踪测量统计学上相似(ICC=0.868),但执行时间明显更短(p<0.0001,ICC=0.088)。线性肿瘤测量的变化与全肿瘤体积呈正相关(p<0.0001)。然而,线性测量与坏死体积不相关(p>0.05)。剂量、肿瘤直径变化、肿瘤体积和坏死体积与生存无关(所有情况下 p>0.05)。然而,Kaplan-Meier 曲线表明,坏死体积增加>10%与生存相关(p=0.0472)。
肝、全肿瘤和肿瘤坏死体积的半自动容积分析可获得良好的观察者内/间可重复性。在这项小型回顾性研究中,肿瘤坏死的测量被认为与生存相关。