Contrast-Enhanced Brain MRI for Ambulatory Evaluation of Headache: Comprehensive or Excessive?

Background and purpose: Headache is one of the most common indications for brain MR imaging. Our purpose was to assess the necessity and yield of contrast-enhanced brain MRI protocols relative to noncontrast protocols for the outpatient evaluation of headache in adults.

Materials and methods: We retrospectively reviewed 874 consecutive adult outpatient brain contrast-enhanced MRIs for the indication of headache. Examination impressions were reviewed for any findings contributing to headache and stratified into 18 categories. Incidental findings unrelated to headache but mentioned in the report impressions were recorded. A subset of 100 contrast-relevant cases were selected for independent blinded review by 2 neuroradiologists first without contrast and then after a 3-week washout period, with and without contrast. Reviewers were asked to classify cases as positive/negative and to describe relevant findings. Specific diagnoses were compared between reviews and readers on a case-by-case basis. Intrareader agreement was calculated between noncontrast and postcontrast reviews.

Results: Eighteen percent of examinations had findings potentially related to headaches. Two percent of cases were deemed contrast relevant. Incidental findings were reported in 4% of cases. On the blinded double review, the percentage of positive cases remained essentially unchanged, decreasing by 1.5% with contrast-enhanced images available. All intra- and extra-axial masses associated with parenchymal edema were detected without contrast. Detection rate of small, incidental meningiomas doubled on postcontrast sequences.

Conclusions: Routine acquisition of contrast-enhanced MRIs for outpatients with an indication of headache had a low yield. Noncontrast examinations demonstrated similar sensitivity for actionable intracranial pathology, while contrast-enhanced MRIs doubled the detection rate of incidental meningiomas. A more practical initial approach may be starting with noncontrast examinations with callbacks or additional imaging for a small percentage of patients.