Conventional Magnetic Resonance Sequences
Conventional magnetic resonance (MR) sequences used in routine practice include T1-weighted image (T1-W), T2-weighted image (T2-W) and fluid-attenuated inversion recovery pulse image (FLAIR) to identify abnormal areas involved in viral encephalitis. Abnormal hyperintensities on limbic and hypothalamic areas by T2-W and FLAIR sequences of MRI are evident in patients with HSE. These abnormal intensity areas appeared hypointense on T1-W images. Furthermore, these abnormal intensities disappeared following the extensive treatment with acyclovir. At this point, it should be mentioned that previous studies have revealed that FLAIR imaging is more sensitive than T2-W spin-echo sequences in detecting encephalitic lesions in the acute phase of the illness.[10,11] In typical cases, MRI examination by T2-W and FLAIR imaging can detect the lesions of viral infection within the first 48 h. At present, however, we do not have definite evidence for the correlation between brain MRI findings and the extent of the abnormalities in cerebrospinal fluid and electroencephalograpy examinations.
Diffusion-weighted imaging (DWI) is now increasingly used in various diseases involving the brain and spine. It has been mostly used to make a diagnosis of fresh cerebral ischemia. Hyperintensity areas on DWI appear within 1 h after the onset of brain infarction, by which we can differentiate the fresh lesions from old infarction. Recently, however, MR imaging involving DWI in other conditions such as intracranial infections and tumors has been explored, although its application on infectious diseases is limited and most of these studies are confined to brain abscesses. Since then, DWI sequences have been reported to be very useful in detecting lesions in toxoplasmosis and Rasmussen encephalitis. The increased sensitivity of DWI sequences with regard to viral encephalitis has been also shown in several studies.[17,18] As shown in Figure 1, our previous patient with HSE showed remarkable hyperintensity lesions in the temporal lobe on DWI sequnces, while the T2-W image did not show clear hyperintense lesions in the same brain region. The MR appearance using DWI is closely related to pathologic changes that occur following viral invasion. In the acute stage, there are areas of congestion, lymphocytic perivascular cuffing and pathological thrombus formation. These areas might be responsible for the cytotoxic edema that leads to restricted diffusion and low apparent diffusion coefficient (ADC). The involved areas on DWI sequences show hyperintensity lesions. This might be due to the presence of cytotoxic edema primarily in the affected gray matter neurons. In the late acute and early subacute stages, the components of vasculitis and perivascular cuffing diminish. Therefore, the proportion of diffusion restriction decreases and ADC starts to increase. This stage of the infection is also known to be accompanied by the vasogenic and/or interstitial collection of fluid that accounts for the lesions becoming hyperintense on T2-W sequences. A pattern of cytotoxic edema might suggest fulminant necrotizing changes and a poor prognosis for patients, whereas a pattern of vasogenic edema represents mild pathological changes and a better prognosis. In fact, Prakash et al. reported that there was a significant direct correlation of ADC values with disease duration in patients with Japanese encephalitis. Furthermore, initial DWI and ADC map imaging could predict outcome in some patients. Thus, we should consider that the addition of DWI sequence to the routine imaging protocol in viral encephalitis is mandatory, as it detects early encephalitic lesions more effectively and depicts the border of infected areas more precisely. In fact, a previous study has shown that in adults, DWI appears to be more sensitive than T2-W and FLAIR imaging in the early detection of the cytotoxic edema, which also show as reduced ADC. At this point, however, we should mention that the above notion does not always hold true for all patients with viral encephalitis. Our recent case of VZV encephalitis demonstrated that FLAIR sequences were more sensitive in detecting lesions than DWI sequences (Figure 2).
MRI of a 71-year-old female patient with herpes simplex encephalitis who presented with fever and impaired consciousness.
Real-time PCR of her cerebrospinal fluid revealed 320,000 copies/ml of HSV-1. (A) DWI sequences showed hyperintensity in the temporal lobe. (B) T2-W fast spin echo axial image showed a faint hyperintensity in the same region. (C) ADC map showed hyperintensity in the same region. Arrows indicate the hyperintensity lesions.
ADC: Apparent diffusion coefficient; DWI: Diffusion-weighted imaging; T2-W: T2-weighted.
Encephalitis caused by varicella zoster virus in a 56-year-old male patient who presented fever and consciousness disturbance.
(A) Axial DWI sequences showed fainter and fewer hyperintensity lesions in the brain stem and temporal lobe cortex and subcortical white matter. (B) Axial FLAIR pulse image showed a clearer and larger number of hyperintensity areas than those of (A). (C) ADC map showed increased value on some lesions detected by FLAIR imaging. Arrows indicate the hyperintensity lesions.
ADC: Apparent diffusion coefficient; DWI: Diffusion-weighted imaging; FLAIR: Fluid-attenuated inversion recovery.
Future Virology. 2012;7(9):901-909. © 2012 Future Medicine Ltd.