Rectal tumour volume (GTV) delineation using T2-weighted and diffusion-weighted MRI: Implications for radiotherapy planning

Abstract

PurposeTo compare the rectal tumour gross target volume (GTV) delineated on T2 weighted (T2W MRI) and diffusion weighted MRI (DWI) images by two different observers and to assess if agreement is improved by DWI. Material and methods27 consecutive patients (15 male, range 27.1–88.8 years, mean 66.9 years) underwent 1.5T MRI prior to chemoradiation (45Gy in 25 fractions; oral capecitabine 850mg/m2), including axial T2W MRI (TR=6600ms, TE=90ms) and DWI (TR=3000ms, TE=77ms, b=0, 100, 800s/mm2). 3D tumour volume (cm3) was measured by volume of interest (VOI) analysis by two independent readers for the T2W MRI and b800 DWI axial images, and the T2W MRI and DWI volumes compared using Mann–Whitney test. Observer agreement was assessed using Bland–Altman statistics. Significance was at 5%. ResultsArtefacts precluded DWI analysis in 1 patient. In the remaining 26 patients evaluated, median (range) T2W MRI MRI and DWI (b=800s/mm2) 3D GTVin cm3 were 33.97 (4.44–199.8) and 31.38 (2.43–228), respectively, for Reader One and 43.78 (7.57–267.7) and 42.45 (3.68–251) for Reader Two. T2W MRI GTVs were slightly larger but not statistically different from DWI volumes: p=0.52 Reader One; p=0.92 Reader Two. Interobserver mean difference (95% limits of agreement) for T2W MRI and DWI GTVs were −9.84 (−54.96 to +35.28) cm3 and −14.79 (−54.01 to +24.43) cm3 respectively. ConclusionSmaller DWI volumes may result from better tumour conspicuity but overall observer agreement is not improved by DWI.