Objective Interferon-β (IFN-β) reduces relapse rates in multiple sclerosis; however clinical response varies between subjects. Can MRI be reliably used to identify poor responders early?
Methods Data from studies of INF-β where early MRI data was given and subjects were classified as good or poor responders using clinical outcomes were used, and overall ORs for poor treatment response according to MRI findings calculated.
Results 1388 papers were screened. seven papers and one abstract met the inclusion criteria, however two were later excluded as insufficiently detailed MRI data was given. The overall OR of poor treatment response in subjects with ≥2 new T2-weighted lesions at 1 year was 9.86 (95% CI 2.33 to 41.70). The OR of poor treatment response in those with ≥1 gad enhancing lesion was 3.34 (95% CI 1.36 to 8.22) and in those with ≥2 gad enhancing lesions 5.46 (95% CI 2.48 to 28.03). Between-study heterogeneity was lower when looking at gad-enhancing lesions. In those with ≥1 new T2 lesion the overall OR for relapses or disease progression was not significantly increased at 2.69 (95% CI 0.72 to 10.04).
Conclusions MRI after starting treatment provides important information about the likelihood of treatment failure. The presence of gad-enhancing lesions seems to provide more reliable information than new T2 weighted lesions. These findings should be incorporated into clinical practice, and second line therapy should be considered in patients with new MRI changes.