PURPOSE: Crohn's disease has been associated with restricted diffusion in diseased small bowel segments on diffusion-weighted imaging (DWI). However, data addressing longitudinal changes in DWI findings and their potential clinical ramifications in the pediatric population are lacking. The purpose of this study was to follow DWI changes in diseased small bowel segments between serial magnetic resonance enterography (MRE) examinations, and to correlate these changes with other imaging parameters and clinical status. METHODS: This retrospective study evaluated patients less than 21 years of age undergoing serial MRE examinations including DWI for Crohn's disease involving the small bowel. All patients carried a diagnosis of Crohn's disease established by pathology or corroborative clinical and imaging findings. Longitudinal changes in mean apparent diffusion coefficient (ADC) values within the wall of affected small bowel lesions were recorded and normalized to both unaffected bowel and skeletal muscle. ADC changes were correlated with qualitative imaging phenotype, as reflected by a defined set of non-DWI imaging parameters, as well as with clinical disease activity. RESULTS: Seventeen lesions were evaluated longitudinally, distributed among 13 patients (9 boys and 4 girls, mean age at baseline 16.6 years), each of whom had two sequential MRE examinations. Lesions demonstrating a fibrostenotic imaging phenotype at follow-up MRE had a significantly lower change in mean ADC value between examinations than lesions that did not have a fibrostenotic imaging phenotype (p = 0.0005), an effect that persisted when ADC values were normalized to unaffected bowel and skeletal muscle. Across all studies, lesions with a fibrostenotic imaging phenotype had lower ADC values than those with an inflammatory imaging phenotype, which were in turn lower than those with a normal imaging phenotype (p = 0.0001). CONCLUSION: Patterns of longitudinal DWI changes in Crohn's disease may differ among small bowel lesions depending upon their specific natural histories. These findings may assist in the evaluation of the ADC value as a potential imaging surrogate when evaluating lesion status, particularly as it pertains to relative contributions of inflammation and fibrosis.
PURPOSE:Crohn's disease has been associated with restricted diffusion in diseased small bowel segments on diffusion-weighted imaging (DWI). However, data addressing longitudinal changes in DWI findings and their potential clinical ramifications in the pediatric population are lacking. The purpose of this study was to follow DWI changes in diseased small bowel segments between serial magnetic resonance enterography (MRE) examinations, and to correlate these changes with other imaging parameters and clinical status. METHODS: This retrospective study evaluated patients less than 21 years of age undergoing serial MRE examinations including DWI for Crohn's disease involving the small bowel. All patients carried a diagnosis of Crohn's disease established by pathology or corroborative clinical and imaging findings. Longitudinal changes in mean apparent diffusion coefficient (ADC) values within the wall of affected small bowel lesions were recorded and normalized to both unaffected bowel and skeletal muscle. ADC changes were correlated with qualitative imaging phenotype, as reflected by a defined set of non-DWI imaging parameters, as well as with clinical disease activity. RESULTS: Seventeen lesions were evaluated longitudinally, distributed among 13 patients (9 boys and 4 girls, mean age at baseline 16.6 years), each of whom had two sequential MRE examinations. Lesions demonstrating a fibrostenotic imaging phenotype at follow-up MRE had a significantly lower change in mean ADC value between examinations than lesions that did not have a fibrostenotic imaging phenotype (p = 0.0005), an effect that persisted when ADC values were normalized to unaffected bowel and skeletal muscle. Across all studies, lesions with a fibrostenotic imaging phenotype had lower ADC values than those with an inflammatory imaging phenotype, which were in turn lower than those with a normal imaging phenotype (p = 0.0001). CONCLUSION: Patterns of longitudinal DWI changes in Crohn's disease may differ among small bowel lesions depending upon their specific natural histories. These findings may assist in the evaluation of the ADC value as a potential imaging surrogate when evaluating lesion status, particularly as it pertains to relative contributions of inflammation and fibrosis.