Diffusion-weighted magnetic resonance for prediction of response after neoadjuvant chemoradiation therapy for locally advanced rectal cancer: preliminary results of a monoinstitutional prospective study.

PURPOSE: To evaluate diffusion-weighted imaging (DWI) for assessment of treatment response in locally advanced rectal cancer (LARC) 8 weeks after neoadjuvant chemoradiotherapy (CRT).

METHODS AND MATERIALS: A total of 28 patients with LARC underwent magnetic resonance imaging (MRI) prior to and 8 weeks after CRT. Tumor volume (TV) was calculated on T2-weighted MRI scans as well as the apparent diffusion coefficient (ADC) was calculated using Echo-planar DWI-sequences. All data were correlated to surgical results and histopathologic tumor regression grade (TRG), according to Mandard's classification. Post-treatment difference ADC (%ΔADC) and TV (%ΔTV) changes at 8 weeks were compared complete response (CR; TRG1) and non-complete response tumors (non-CR; TRG2-5).

RESULTS: The mean % ADC increase of CR group was significantly higher compared to non-CR group (77.2 ± 54.63% vs. 36.0 ± 29.44%; p = 0.05). Conversely, the mean % TV reduction did not significantly differ in CR group from non-CR group (73.7% vs. 63.77%; p = 0.21). Accordingly, the diagnostic accuracy of the mean % ADC increase to discriminate CR from non-CR group was significantly higher than that of the mean % TV reduction (0.913 vs. 0.658; p = 0.022). No correlation was found between mean % TV reduction and TRG (rho = 0.22; p = 0.3037), whereas a negative correlation between mean % ADC increase and TRG was recorded (r = -0.69; p = 0.006).

CONCLUSION: The mean % ADC increase appears to be a reliable tool to differentiate CR from non-CR after CRT in patients with LARC.