1 Department of Radiology, the Affiliated Cancer Hospital of Zhengzhou University, Henan Cancer Hospital, 127 Dongming Rd, Zhengzhou 45008, China.
2 Department of Imaging, Dana-Farber Cancer Institute, Boston, MA.
AJR Am J Roentgenol. 2018 Mar;210(3):677-684. doi: 10.2214/AJR.17.18104. Epub 2018 Jan 11.
The objective of our study was to retrospectively evaluate the efficacy of combined analysis of T2-weighted imaging and DWI in the diagnosis of parametrial invasion (PMI) in cervical carcinoma.
The clinical records of 192 patients with cervical carcinoma who met the study requirements were reviewed for this retrospective study. The signal intensities of suspicious PMI tissue were assessed on T2-weighted images, DW images, and apparent diffusion coefficient maps independently by two experienced radiologists. The radiologist observers predicted the presence of PMI by scoring T2-weighted imaging alone and then by scoring T2-weighted imaging and DWI combined. The results were compared with histopathologic findings.
Histopathologic findings revealed PMI in 24 of 192 study subjects. In positively predicting the presence of PMI, T2-weighted imaging and DWI combined scored significantly better than T2-weighted imaging alone, as proven by high sensitivity (T2-weighted imaging alone vs T2-weighted imaging and DWI combined: observer 1, 75.0% vs 83.3% [p = 0.477]; observer 2, 66.7% vs 91.7% [p < 0.05]), high specificity (T2-weighted imaging alone vs T2-weighted imaging and DWI combined: observer 1, 84.5% vs 98.8% [p < 0.001]; observer 2, 85.7% vs 98.8% [p < 0.001]), and high accuracy (T2-weighted imaging alone vs T2-weighted imaging and DWI combined: observer 1, 83.3% vs 96.9% [p < 0.001]; observer 2, 83.3% vs 97.9% [p < 0.001]). The area under the ROC curve was also significantly higher for T2-weighted imaging and DWI combined (observer 1, 0.911; observer 2, 0.952) than for T2-weighted imaging alone (observer 1, 0.798; observer 2, 0.762). Although the interobserver agreement was good for T2-weighted imaging (κ = 0.695) and excellent for T2-weighted imaging and DWI combined (κ = 0.753), the improvement failed to achieve statistical significance (p = 0.28).
Combined analysis of T2-weighted imaging and DWI enhances the accuracy of diagnosing PMI in patients with cervical cancer compared with T2-weighted imaging alone.
本研究旨在回顾性评估 T2 加权成像与 DWI 联合分析在宫颈癌旁肌层浸润（PMI）诊断中的效能。
本回顾性研究纳入了 192 名符合研究要求的宫颈癌患者的临床资料。由 2 位有经验的放射科医生分别对可疑 PMI 组织的 T2 加权图像、DWI 图像和表观扩散系数图上的信号强度进行评估。放射科医生通过单独评分 T2 加权成像，然后通过联合评分 T2 加权成像和 DWI 来预测 PMI 的存在。并将结果与组织病理学发现进行比较。
组织病理学发现 192 例研究对象中有 24 例存在 PMI。在阳性预测 PMI 存在方面，T2 加权成像和 DWI 联合评分明显优于 T2 加权成像单独评分，证明具有较高的敏感性（观察者 1：T2 加权成像单独评分 vs T2 加权成像和 DWI 联合评分，75.0% vs 83.3%[p=0.477]；观察者 2：66.7% vs 91.7%[p<0.05]）、较高的特异性（观察者 1：T2 加权成像单独评分 vs T2 加权成像和 DWI 联合评分，84.5% vs 98.8%[p<0.001]；观察者 2：85.7% vs 98.8%[p<0.001]）和较高的准确性（观察者 1：T2 加权成像单独评分 vs T2 加权成像和 DWI 联合评分，83.3% vs 96.9%[p<0.001]；观察者 2：83.3% vs 97.9%[p<0.001]）。T2 加权成像和 DWI 联合评分的 ROC 曲线下面积也明显高于 T2 加权成像单独评分（观察者 1：0.911；观察者 2：0.952）。尽管 T2 加权成像的观察者间一致性良好（κ=0.695），T2 加权成像和 DWI 联合的观察者间一致性极好（κ=0.753），但差异无统计学意义（p=0.28）。
与 T2 加权成像单独分析相比，T2 加权成像与 DWI 联合分析可提高宫颈癌旁肌层浸润诊断的准确性。
