Follow-up of Brain Abscess Volume by Magnetic Resonance Imaging: a Pilot Study

Background: The follow-up of brain abscess (BA) by neuro-imaging studies is poorly standardized. We aimed to estimate volume changes as a function of time under appropriate medical treatment in adult patients with bacterial BA.

Methods: We retrospectively reviewed all cases of bacterial BA followed-up in one University Hospital from January 2011 to February 2016. Cases were identified through computerized database. We excluded patients who underwent abscess excision. Only cases for whom magnetic resonance imaging (MRI) studies were available at baseline, and after 4-12 weeks of adequate antibacterial treatment were analysed. The BA volume was automatically calculated after manual delineation in each slice of contrast-enhanced T1-weighted sequence. Evolution of BA volume was defined by the volume difference between the 2 MRI (expressed in %), divided by number of weeks between both MRI.

Results: Of the 60 cases of BA screened, 50 were excluded because of no follow-up MRI during the pre-defined 4-12 weeks window (n=38), parasitic BA (n=6), surgical excision (n=5), and post-biopsy hematoma (n=1). The 10 patients enrolled had a mean age of 47 years, sex ratio was 1/1. BA were due to Mycobacterium tuberculosis (n=2), Fusobacterium sp. (n=2), and Staphylococcus constellatus, Eikenella corrodens, Peptoniphilus harei, Aggregatibacter aphrophilus, Streptococcus constellatus, Porphyromonas gingivalis (one patient each). Patients were treated in agreement with institutional guidelines for neuromeningeal tuberculosis (9 months-regimen) or bacterial BA (i.e. six-weeks combination of ceftriaxone and metronidazole). Initial mean size of BA as evaluated by MRI was 12.2 cm³. Four patients had multiple BA. The mean decrease of BA volume was -10.4 ± 6.9% per week, with striking differences between tuberculosis BA (+0.6 ± 1.1% per week), and non-tuberculous bacterial BA (-13.2 ± 4.3% per week). Mean decrease of BA volume was similar in the 5 patients initially managed with stereotactic aspiration/biopsy (-11.7 ± 3.9%). All patients had favourable outcome.

Conclusion: Under appropriate antibacterial treatment, non-tuberculous BA volume decrease by 10-15% per week, as assessed by MRI, during the first 3 months. These figures may be used for systematic follow-up.