DWI in Pediatric Small-Bowel Crohn Disease: Are Apparent Diffusion Coefficients Surrogates for Disease Activity in Patients Receiving Infliximab Therapy?

Abstract

The purpose of this study was to determine prospectively whether bowel wall apparent diffusion coefficient (ADC) measurements can be used to monitor treatment response to infliximab therapy in the setting of pediatric small-bowel Crohn disease. Twenty-eight pediatric subjects with newly diagnosed biopsy-proven Crohn disease of the distal or terminal ileum treated with infliximab were enrolled. Subjects underwent MR enterography at baseline, 1 month after therapy, and 6 months after therapy. Imaging features were documented, including bowel wall ADC and arterial or enteric phase contrast-enhanced signal intensity normalized to that of unenhanced imaging. A linear mixed model assessed the relationship between ADC and time; patient age and sex and azathioprine combination therapy were covariates. The diagnostic performance (with 95% CIs) of an increase in bowel wall ADC of 20% or more for identifying response to infliximab was calculated using a decrease in normalized contrast-enhanced bowel wall signal intensity of 20% or more as the reference standard. Bowel wall ADC increased over time (mean [± SD], 1180 ± 200 × 10-6 mm2/s at baseline, 1420 ± 420 × 10-6 mm2/s at 1 month, and 1450 ± 450 × 10-6 mm2/s at 6 months; p = 0.0003); azathioprine therapy modulated this rate of change (p = 0.003). There was a statistically significant negative correlation between change in ADC and change in normalized contrast-enhanced signal intensity over time (ρ = -0.36; p < 0.001). The diagnostic performance of change in ADC for identifying response to infliximab therapy was sensitivity of 0.58 (95% CI, 0.34-0.80), specificity of 0.52 (95% CI, 0.31-0.72), positive predictive value of 0.48 (95% CI, 0.27-0.69), and negative predictive value of 0.62 (95% CI, 0.38-0.82). Bowel wall ADC increases over time in pediatric subjects receiving infliximab, but the diagnostic performance of ADC is likely insufficient for reliable treatment monitoring.