MRI-based neck imaging reporting and data systems in diagnosis of extracavity local recurrent nasopharyngeal carcinoma and posttreatment changes / 中华放射学杂志

ABSTRACT

Objective:To evaluate the diagnostic value of neck imaging reporting and data systems (NI-RADS) based on MRI in extracavity local recurrent nasopharyngeal carcinoma (ELRNPC) and posttreatment changes (PTC).Methods:From April 2015 to September 2020, 33 cases of ELRNPC and 37 cases of PTC confirmed by pathology or follow-up were retrospectively enrolled at Zhongshan City People′s Hospital. Two radiologists independently evaluated the conventional MRI findings using NI-RADS criteria in the first step, then evaluated with conventional MRI and diffusion weighted imaging (DWI) sequences in the second step. All images were re-evaluated by one senior radiologist in the same steps after three months. Inter- and intra-reader agreements were assessed with Cohen′s Kappa test. Receiver operating characteristic curves were generated to assess the diagnostic values of NI-RADS categories between ELRNPC and PTC. The area under the curve (AUC) was compared by Delong test.Results:Inter- and intra-reader agreements of Kappa value were 0.742 and 0.909 for conventional MRI and 0.807 and 0.934 for conventional MRI with DWI. In the differential diagnosis of ELRNPC and PTC, the AUC, sensitivity, and specificity of NI-RADS categories based on conventional MRI were 0.932 (95%CI 0.846-0.978), 87.9% (95%CI 71.8%-96.6%), 94.6% (95%CI 81.8%-99.3%), and of NI-RADS based on conventional MRI with DWI were 0.991 (95%CI 0.933-1.000), 93.9% (95%CI 79.8%-99.3%), 97.3% (95%CI 85.8%-99.9%), respectively. There was a statistical difference between the AUCs of the two categories ( Z=2.20, P=0.028). Conclusions:For both the NI-RADS based on MRI with or without DWI, the differential diagnostic value of ELRNPC and PTC is excellent, while the consistency and diagnostic performance are more substantial when combined with DWI.