• Juri Katchanov, Eberhard Siebert, Matthias Endres, and Randolf Klingebiel.
• Department of Neurology, Campus Charité Mitte, Charité, Berlin, Germany, juri.katchanov@charite.de.
• Neuroradiology. 2009 Nov 1; 51 (11): 723-9.

IntroductionHere, we analyzed the frequency, morphological pattern, and imaging characteristics of focal lesions as a consequence of community-acquired bacterial meningitis. We hypothesized that diffusion-weighted imaging combined with contrast-enhanced imaging, serial scanning, and multimodal vascular studies would provide further insight into the pathological basis of such parenchymal lesions in bacterial meningitis.MethodsWe reviewed clinical and imaging data (i.e., magnetic resonance tomography, magnetic resonance angiography, computed tomography angiography, digital subtraction angiography) of 68 adult patients admitted to our neurological intensive care unit between March 1998 and February 2009 with the diagnosis of community-acquired bacterial meningitis.ResultsWe identified seven patients with parenchymal lesions. These lesions could be attributed to four morphological patterns: (1) territorial cerebral ischemia, (2) perforating vessels ischemia, (3) ischemia of presumed cardiac origin, and (4) isolated cortical lesions. Whereas the patterns (1) and (2) were associated with vasculopathy of large- and medium-sized vessels (as shown by cerebral vascular imaging), vessel imaging in (3) and (4) did not show abnormal findings.ConclusionOur study implies that parenchymal lesions in acute bacterial meningitis are mainly ischemic and due to involvement of large-, medium-, and small-sized arteries of the brain. Diffusion-weighted imaging combined with conventional, CT-, or MR-based cerebral angiography revealed the underlying pathophysiological mechanisms in the majority of patients. Furthermore, we detected two patients with isolated bilateral cortical involvement and normal vessel imaging. These lesions might represent ischemia due to the involvement of small pial and intracortical arteries.

35 keywords...

hide…