Impact of qualitative Diffusion weighted imaging and apparent diffusion coefficient (DWI/ADC) on target volume...

Magnetic resonance imaging (MRI) based IGABT in cervical cancers relies on differential target volume delineation on T2-weighted imaging. Despite decades of experience, there are certain inconsistencies in delineation, with the gross tumor volume (GTV) demonstrating the least concordance. Analyzing current imaging methodologies, in the context of functional imaging, can possibly conceptualize GTV delineation.

MRI pelvis datasets of 20 cervical cancer patients at pre-external beam radiotherapy (EBRT) and pre-brachytherapy were selected. The GTV was delineated on the T2 sequence alone and on the fused T2/DWI/ADC sequences. Quantitative and spatial differences between the two delineation methods were analyzed.

At pre-EBRT, the median absolute difference between GTV_T2 and GTV_T2/DWI was 2 cm³ (-1.72; 8.27, p = 0.117), with strong positive correlation (ρ = 0.943) and moderate to good spatial agreement (Volume ratio: 0.90, DICE Similarity Coefficient (DSC): 0.75, Jaccard Index (JI): 0.61). All lateral tumor extent measurements on one side were higher on T2 imaging, with a maximum of 2.63 cm (1.83;2.87) versus 2.05 cm (1.78;2.39) (p < 0.0001). At pre-brachytherapy, the median difference was 0.30 cm³ (-0.07; 0.77, p = 0.017), with moderately strong correlation (ρ = 0.707) and low to moderate spatial agreement (Volume ratio: 0.79, DSC: 0.33, JI: 0.19). The lateral tumor extent measurement at one of the levels was significantly higher on T2 imaging, being 0.95 cm (0.60;1.78) versus 0.81 cm (0.00;1.39) (p = 0.033).

The two volumes showed strong correlation; however, small yet statistically significant differences were observed at the pre-brachytherapy, with lesser spatial agreement. T2-only imaging demonstrated higher lateral tumor extent at both time points. Overall, these findings indicate discernible differences with functional imaging.