Cranial magnetic resonance imaging in multiple sclerosis--indicated only in exceptional cases?

Clinical examination, spinal fluid laboratory data, evoked potentials (EP), cranial computed tomography (CT), and magnetic resonance imaging (MRI) were compared for their value in diagnosing multiple sclerosis (MS) in 51 patients aged 18-60 yrs. MRI was superior to CT imaging: supratentorial lesions were found in 90% of the patients with confirmed MS. These lesions are not specific, however, as vascular processes, vitamin B12 deficiency or chronic encephalitis, can show similar MRI-patterns. The detection of infratentorial lesions in symptomatic patients was poor (8% compared to 64% pathological EP-findings). Thirty-eight patients (= 75%) could be classified correctly as "confirmed MS" according to Bauer-criteria without any CT or MR imaging. In the remaining 13 patients presenting classificatory problems due to normal CSF, first manifestation or possible alternate processes, MRI helped for the final diagnosis in only 6 cases. The value of MRI in diagnosing MS seems to lie in its superiority to CT in excluding other pathological processes or for research rather than in its value for the confirmation of the diagnosis. In a minority of cases, it can detect supratentorial lesions in patients with pure spinal symptoms, normal CSF, or first manifestations.