Magnetic resonance imaging in acute meningoencephalitis of viral and unknown origin: frequent findings and prognostic potential

Background

Magnetic resonance imaging (MRI) findings in meningoencephalitis have mainly been described in terms of their diagnostic value rather than their prognostic potential, except for herpes simplex virus (HSV) encephalitis. The aims of our study were to describe frequency and anatomic locations of MRI abnormalities specific to limbic, circadian and motor systems in a cohort of meningoencephalitis patients, as well as to investigate the prognostic value of these MRI findings.

Methods

A secondary, selective analysis of a retrospective database including all meningitis, meningoencephalitis and encephalitis cases treated between 2016 and 2018 in the University hospital of Bern, Switzerland was performed. Patients with meningitis of any cause, bacterial or autoimmune causes of encephalitis were excluded.

Results

MRI scans and clinical data from 129 meningoencephalitis cases found that the most frequent causes were tick-borne encephalitis (TBE, 42%), unknown pathogens (40%), VZV (7%), and HSV1 (5%). At discharge, median modified Rankin Score (mRS) was 3 (interquartile range, IQR, 1), 88% of patients had persisting signs and symptoms. After a median of 17 months, median Glasgow Outcome Score (GOS) was 5 (IQR 1), 39% of patients still had residual signs or symptoms. All patients with HSV, 27% with TBE and 31% of those with meningoencephalitis of unknown etiology had fluid-attenuated inversion recovery (FLAIR) and to a lesser extent diffusion-weighted imaging (DWI) lesions in their initial MRI, with highly overlapping anatomical distribution. In one fifth of TBE patients the limbic system was affected. Worse outcome was associated with presence of DWI and/or FLAIR lesions and lower normalized apparent diffusion coefficient (ADC) signal intensities.

Conclusion

Presence of FLAIR lesions, restricted diffusion as well as the extent of ADC hypointensity in initial MRI are parameters which might be of prognostic value regarding the longterm clinical outcome for patients with meningoencephalitis of viral and of unknown origin. Although not described before, affection of limbic structures by TBE is possible as shown by our results: A substantial proportion of our TBE patients had FLAIR signal abnormalities in these regions.

Can we forecast poor outcome in herpes simplex and varicella zoster encephalitis? A narrative review

Herpes simplex virus (HSV) and varicella zoster virus (VZV) are among the most commonly diagnosed infectious causes of sporadic encephalitis worldwide. Despite treatment, mortality and morbidity rates remain high, especially for HSV encephalitis. This review is intended to provide an overview of the existing scientific literature on this topic from the perspective of a clinician who is confronted with serious decisions about continuation or withdrawal of therapeutic interventions. We performed a literature review searching two databases and included 55 studies in the review. These studies documented or investigated specifically outcome and predictive parameters of outcome of HSV and/or VZV encephalitis. Two reviewers independently screened and reviewed full-text articles meeting the inclusion criteria. Key data were extracted and presented as a narrative summary. Both, HSV and VZV encephalitis have mortality rates between 5 and 20% and complete recovery rates range from 14 to 43% for HSV and 33 to 49% for VZV encephalitis. Prognostic factors for both VZV and HSV encephalitis are older age and comorbidity, as well as severity of disease and extent of magnetic resonance imaging (MRI) lesions on admission, and delay in treatment initiation for HSV encephalitis. Although numerous studies are available, the main limiting factors are the inconsistent patient selection and case definitions as well as the non-standardised outcome measures, which hampers the comparability of the studies. Therefore, larger and standardised observational studies applying validated case definitions and outcome measures including quality of life assessment are required to provide solid evidence to answer the research question.