A pathological comparative study of diffusion-weighted imaging and computed tomography in determination of lesion length for esophageal carcinoma / 中华放射肿瘤学杂志

ABSTRACT

Objective To investigate the accuracy of measurement of lesion length by computed tomography (CT) scan and diffusion?weighted imaging (DWI) for esophageal carcinoma, and to provide an optimized imaging method as a reference for target delineation in esophageal carcinoma. Methods Thirty?five patients with thoracic esophageal carcinoma from 2012 to 2013 were prospectively enrolled as subjects. All patients underwent examinations of esophageal endoscopy, CT scan of the thorax and abdomen, and DWI before radical surgery. Lesion lengths were measured by the above methods and compared with the real length of pathological specimen resected at surgery. The consistency between the lesion length measured by each imaging method and the pathological standard were evaluated using the intraclass correlation coefficient (ICC) and the Bland?Altman method. Results Four patients had no hyperintense signal on DWI, resulting in a false?negative rate of 11% in all patients and a false?negative rate of 44%(4/ 9) in patients with stage T1 esophageal carcinoma. Thirty?one patients had lesion lengths measured by DWI for analysis. The tumor length of pathological specimen after surgery was 4?? 58 cm, while the tumor lengths measured by endoscopy, CT scan, and DWI with b?values of 600, 800, and 1000 s/ mm2 were 4?? 56, 5?? 58, 4?? 41, 3?? 99, and 3?? 83 cm, respectively. The ICC values were 0?? 703, 0?? 764, 0?? 946, 0?? 890, and 0?? 882, respectively, with P value of 0?? 000 for all. According to the results of the Bland?Altman method, the highest degree of consistency was achieved between the tumor lengths measured by endoscopy and DWI with a b?value of 600 s/ mm2 and the pathological standard. Conclusions The esophageal tumor lengths measured by DWI are close to the real tumor length of the pathological specimen, in which the lesion length measured by DWI with a b?value of 600 s/ mm2 is most reliable. However, the value of DWI in the early diagnosis of esophageal carcinoma is limited.