Previously, there were no validated magnetic resonance imaging (MRI) index for evaluating perianal fistulas in patients with Crohn disease (CD).
In this study, researchers developed and internally validated a new index called MAGNIFI-CD can effectively assess MRI scans of perianal fistulizing CD activity, and offers improved characteristics over previous indices, according to study results published in Gastroenterology.
Researchers evaluated 160 MRI scans collected at baseline and week 24 during a phase 3, randomized, clinical trial of stem-cell therapy to treat perianal fistulizing CD. Four expert radiologists blinded to study time point and clinical characteristics scored all MRI scans for disease activity using index items from previous studies as well as exploratory items.
Final items for the new index were identified using a backward, step-down, model-building approach with P =.05 as the criterion for item retention. Index items that were retained to develop and refine the MAGNIFI-CD MRI index included number of fistula tracts, fistula length, hyperintensity of primary tract on post-contrast T1-weighted images, dominant feature, extension, and inflammatory mass. The optimism-corrected R2 of the model and the original sample were 0.71 and 0.74, respectively, and the calibration slope for the model was 0.98.
Researchers explored the operating properties of MAGNIFI-CD, then compared the new index to other frequently used indices for assessing perianal fistulas, the Van Assche Index (VAI) and the modified VAI (mVAI). Intra-rater reliability was higher for the MAGNIFI-CD at 0.85 (95% confidence interval [CI], 0.77-0.90), with reliability for the VAI and MVAI at 0.81 (95% CI, 0.71-0.86), and 0.81 (95% CI, 0.74-0.86), respectively.
Inter-rater reliability was also higher for MAGNIFI-CD at 0.74 (95% CI, 0.63-0.80), while inter-rater reliability for the VAI and MVAI was 0.68 (95% CI, 0.56-0.77) and 0.67 (95% CI, 0.55-0.75), respectively. Lastly, standardized effect size estimates for the VAI, mVAI, and MAGNIFI-CD were 0.68 (95% CI 0.32-1.03), 0.84 (95% CI 0.48-1.21), and 1.02 (95% CI 0.65-1.39), respectively. These results revealed that the MAGNIFI-CD had improved operating characteristics compared with the original and modified versions of the VAI.
This study had limitations. Patients with simple fistulas, rectovaginal fistulas, and rectal or anal stenosis and/or severe proctitis were excluded from the dataset, limiting generalizability of the index in those patient subgroups. The expertise of the radiologists in pelvic MRI and item scoring conventions may also limit generalization of the index to other settings. The MAGNIFI-CD was also not independently assessed in a second data set. Finally, the operating properties of the MAGNIFI-CD were limited to an intervention of stem cell therapy injection and still need to be assessed in other treatment types.
The study researchers concluded that the MAGNIFI-CD has improved operating characteristics compared to other indices assessing MRI scans of perianal fistula activity in CD, and that the new index may be useful as an outcome measure in clinical trials investigating treatment for perianal fistulizing CD.
Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.
Hindryckx P, Jairath V, Zou G, et al. Development and validation a magnetic resonance index for assessing fistulas in patients with Crohn’s disease [published online July 20, 2019]. Gastroenterology. doi: 10.1053/j.gastro.2019.07.027.