Assessment of consistency of multiple magnetic resonance imaging modalities in measurement of gross tumor volume for nasopharyngeal carcinoma

Abstract

Objective To investigate the consistency of multiple magnetic resonance imaging (MRI) modalities in measurement of gross tumor volume (GTV) for nasopharyngeal carcinoma. Methods A retrospective analysis was performed among 45 patients who were newly diagnosed and pathologically confirmed with nasopharyngeal carcinoma. All patients underwent plain, enhanced, and diffusion-weighted MRI scans. Primary tumor was manually delineated on enhanced axial T1WI, axial T2WI, and apparent diffusion coefficient (ADC) maps. GTV was calculated by the sum of areas. The obtained GTV was divided into T1+ C group, T2WI group, and ADC group. One-way analysis of variance was used to evaluate the differences between the mean values from the three groups. The Bland-Altman method was used to evaluate the consistency of different imaging modalities in GTV measurement using the mean difference between the two groups and 95% distribution of the difference. Results There was no significant difference in mean GTV between the T1+ C group, the T2WI group, and the ADC group (20.8±18.1) vs. (21.5±17.2) vs. (24.4±20.8) cm3(P=0.617). The mean differences and 95% limits of agreement between the T2WI group and the T1+ C group, between the ADC group and the T2WI group, and between the ADC group and the T1+ C group were 0.74, 2.94, and 3.68 cm3 and (-5.97-7.44), (-8.25-13.69), and (-3.34-10.70) cm3, respectively. Conclusions The multiple MRI modalities have poor consistency in measurement of GTV for nasopharyngeal carcinoma. Therefore, they cannot replace each other. Enhanced T1WI combined with T2WI and DWMRI can determine a relatively accurate range of target volume for nasopharyngeal carcinoma. Key words: Multimodality magnetic resonance; Gross tumor volumel; Consistency