Apparent diffusion coefficient measurement covering complete tumor area better predicts rectal cancer response to neoadjuvant chemoradiotherapy

Abstract

Aim: To determine the impact of two apparent diffusion coefficient (ADC) measurement techniques on diffusionweighted magnetic resonance images (DW MRI) on the assessment of rectal cancer response to neoadjuvant chemoradiotherapy (CRT). Methods: ADC values were measured prospectively with two different techniques - the first, which measures ADCs in the most cellular tumor parts, and the second, which measures the entire tumor area, in 58 patients with locally advanced rectal cancer on pre-CRT and post-CRT image sets. Areas under the receiver operating characteristic curves (AUCs) and parameters of diagnostic accuracy were calculated for pre- and post-CRT ADC values and numeric and percent ADC change for each technique to determine their performance in tumor response evaluation using histopathological tumor-regression grade as the reference standard. Results: The second technique yielded higher AUCs (0.935 vs 0.704, P < 0.001), percent-change (0.828 vs 0.636, P < 0.001), and numeric-change (0.866 vs 0.653, P < 0.001) than the first technique for post-CRT ADC. Accuracies for post-CRT ADC assessment were 62% for the first and 88% for the second technique (cut-off values: 0.98 and 1.29 × 10-3 mm2/s, respectively) and for ADC change assessment, both numeric and percent, 59% and 74%, respectively (cut-off values: increase of 0.18 and 0.28 × 10-3 mm2/s; increase of 24% and 37%, respectively). Conclusions: The type of measurement technique significantly affected ADC results. ADC measurements covering a larger area better predicted tumor response to therapy. Post-CRT ADCs, regardless of the measurement technique, and numeric ADC change measured in the whole tumor volume accurately identified non-complete responders. Post-CRT ADCs measured in the entire tumor area yielded the highest accuracy level in tumor response evaluation.