Imaging experimental cerebral malaria in vivo: significant role of ischemic brain edema

Abstract

The first in vivo magnetic resonance study of experimental cerebral malaria is presented. Cerebral involvement is a lethal complication of malaria. To explore the brain of susceptible mice infected with Plasmodium berghei ANKA, multimodal magnetic resonance techniques were applied (imaging, diffusion, perfusion, angiography, spectroscopy). They reveal vascular damage including blood-brain barrier disruption and hemorrhages attributable to inflammatory processes. We provide the first in vivo demonstration for blood-brain barrier breakdown in cerebral malaria. Major edema formation as well as reduced brain perfusion was detected and is accompanied by an ischemic metabolic profile with reduction of high-energy phosphates and elevated brain lactate. In addition, angiography supplies compelling evidence for major hemodynamics dysfunction. Actually, edema further worsens ischemia by compressing cerebral arteries, which subsequently leads to a collapse of the blood flow that ultimately represents the cause of death. These findings demonstrate the coexistence of inflammatory and ischemic lesions and prove the preponderant role of edema in the fatal outcome of experimental cerebral malaria. They improve our understanding of the pathogenesis of cerebral malaria and may provide the necessary noninvasive surrogate markers for quantitative monitoring of treatment.

Figure 1.

Assessment of brain swelling in mice with CM. A, Typical axial multislice T₂-weighted MR images from an uninfected mouse and a mouse with CM (c, cortex; cc, corpus callosum; hi, hippocampus; s, striatum, v, ventricles). Note the areas of hyperintense signal in the striatum and corpus callosum (H). B, Total cerebral, ventricular, and lesion volumes from uninfected mice (n = 8), mice with mild CM (n = 8), and mice with severe CM (n = 5). C, Mid-sagittal T₂-weighted MR images from an uninfected mouse and a mouse with CM. 1, Line going from the pituitary gland to the Sylvius aqueduct; 2, median line crossing the medial cerebellar nucleus; 3, median line stemming from the cerebellar obex (b, brainstem; cb, cerebellum; pg, pituary gland). D, Measurement of distances 1-3 from the mid-sagittal T₂-weighted MR images of uninfected mice (n = 8) and mice with CM (n = 7). ^*p < 0.05; ^**p < 0.01; ^***p < 0.005; ^****p < 0.001. Scale bars: A, C, 1 mm.

Figure 2.

ADC maps reveal cytotoxic edema in mice with CM. Representative ADC maps from uninfected mice and mice with CM are shown. Scale bar, 1 mm.

Figure 3.

Evidence of BBB lesions and vascular damage in mice with CM. A, Typical axial multislice T₁-weighted MR images from an uninfected mouse and two mice with CM before and after the intravenous injection of Gd-DTPA. Arrows indicate areas of contrast enhancement showing major ventriculitis in mice with CM and signal enhancement in lesions in mice with severe CM. B, Typical axial multislice T₂^*-weighted MR images from an uninfected mouse and a mouse with CM showing hemorrhages (arrows). Scale bars, 1 mm.

Figure 4.

Brain perfusion maps reveal decreased CBF, and angiograms show major artery compression in mice with CM. A, Representative perfusion maps from uninfected mice and mice with CM. B, Coronal, axial, and sagittal maximum intensity projections of a three-dimensional time-of-flight angiogram of an uninfected mouse, a mouse with mild CM, and a mouse with severe CM. Orientations: d, dorsal; v, ventral; r, rostral; c, caudal. 1, Anterior cerebral artery; 2, ophthalmic artery; 3, middle cerebral artery; 4, palatine artery; 5, posterior cerebral artery; 6, basilar artery; 7, Willis circle; 8, carotid artery. Scale bars, 1 mm.

Figure 5.

In vivo cerebral ¹H-MR and ³¹P-MR spectra show an ischemic metabolic profile. A, Representative spectra from an uninfected mouse and from a mouse with CM (TE, 16 and 135 ms). B, Original and deconvoluted ³¹P-MR spectra from an uninfected mouse and a mouse with CM.