Introduction/Purpose

Crohn’s disease (CD) is a chronic inflammatory bowel disease characterized by segmental and transmural bowel wall inflammation that often progresses to fibrosis with intestinal strictures and obstruction. Differentiating the fibrostenotic from the inflammatory component of strictures is clinically important; however, it is still insufficiently evaluated by using clinical response, conventional magnetic resonance imaging (MRI), or computed tomography. Recently, magnetization transfer MRI (MTI) has shown potential in detecting bowel fibrosis in CD. The objective of this study was to assess the efficacy of MTI for differentiating fibrotic from inflammatory stricture and for grading of bowel fibrosis in patients with CD, compared with surgical histopathology as the reference standard.

Methods

MTI data of 12 patients (4 men, 8 women; mean age: 32.9 ± 9.2 years) with known CD were acquired using a 3T MR system (MAGNETOM Trio, Siemens Healthcare, Erlangen, Germany), and were analyzed before elective surgery. MTI was performed using two breath-held axial 2D gradient-echo sequences with and without application of a 1.2 kHz off-resonant pre-saturation-pulse. Bowel wall magnetization transfer ratio (MTR) with normalization to skeletal muscle was calculated, and the region-by-region correlation between MTR and histologic degree of fibrosis / inflammation, as well as amount of type I collagen and vessel density from the surgical specimen, was performed.

Results

Normalized MTRs strongly correlated with histologic degree of fibrosis (r=0.711) and moderately correlated with type I collagen scores (r=0.362), but did not correlate with inflammation scores (r=0.177) or vessel densities (r=0.184). Normalized MTRs of 59 examined fibrotic bowel walls were significantly higher compared to 61 non-fibrotic bowel walls (P<0.001). Significant differences in normalized MTRs were also shown among mildly, moderately and severely fibrotic walls (all P<0.05). Receiver operating characteristics (ROC) curve analysis showed high accuracy of normalized MTRs with an area under ROC curve of 0.920 (P<0.001) for differentiating fibrotic and non-fibrotic bowel walls. Using normalized MTR of 0.62 as a cutoff value, the diagnostic sensitivity and specificity for differentiating between fibrotic and non-fibrotic bowel walls were 88.10% and 86.90%, respectively.

Discussion

In our study, the normalized MTR increased with the severity of bowel fibrosis. The ability of MTI to detect and further stratify bowel fibrosis regarding disease severity could assist in surveillance of the natural progression of CD. Meanwhile, MTI was sensitive to the changes in bowel fibrosis but not to that of inflammation. In our study, this feature demonstrates the reliability of MTI for detecting the presence and severity of bowel fibrosis. Moreover, we found a fair correlation between normalized MTR and type I collagen, another hallmark feature of intestinal fibrosis, which has previously been speculated to be nonlinear.

Conclusions

MTI can accurately detect and distinguish varying degrees of bowel fibrosis with or without coexisting inflammation, which can be considered as a potential tool for treatment planning and surveillance of CD progression.

Acknowledgements

No acknowledgement found.

References

No reference found.