Encephalitis in the Immunocompromised: Clinical Presentation, Diagnostic Findings, Treatment, and Comparison of Infectious and Autoimmune Etiologies
Anna Kolchinski1, Margaret Li2, Ralph Habis1, John Probasco3, Rodrigo Hasbun4, Arun Venkatesan3
1Johns Hopkins University School of Medicine, 2Neuroscience, Johns Hopkins University, 3Neurology, The Johns Hopkins Hospital, 4Neurology, University of Texas Houston
Objective:
To characterize differences in clinical presentation and diagnostic studies between immunocompromised and immunocompetent encephalitis patients.
Background:
There is limited data on how immunocompromised patients with encephalitis differ from the general encephalitis population. The aim of this study is to characterize such differences in all-cause encephalitis, and in the infectious and autoimmune subtypes.
Design/Methods:
This is a retrospective study of 647 encephalitis patient records from two academic medical centers, dichotomized into immunocompromised (HIV, current immunosuppressive medication use, recent chemotherapy, or organ transplant) and immunocompetent groups. Key inclusion criteria included a confirmed encephalitis diagnosis by standardized case definitions. Clinical characteristics considered included comorbidities, presentation, CSF and imaging results.
Results:
This study included 144 immunocompromised and 503 immunocompetent patients. Immunocompromised patients were more likely to have varicella zoster virus (14% vs 5%, p<.001), cytomegalovirus (4% vs .4%, p<.001), or cryptococcal encephalitis (6% vs 0.2%, p<.001) and less likely to be diagnosed with anti-NMDAR (2% vs 10.8%, p<.001) and seronegative autoimmune encephalitis (4% vs 11.9%, p <.01) than the immunocompetent. Among those with infectious encephalitis, immunocompromised patients were more likely to present with focal neurological deficits (46% vs 33%, p=0.049) and confusion 88% vs 74%, p = 0.008), to have hypoglycorrhachia (42% vs 18%, p<0.001) and high CSF neutrophil count (57/mm3 vs 27/mm3, p=0.038), and to have their encephalitis be treated with vancomycin (67/85 or 79% vs. 117/180 or 65%, p= 0.023), rituximab (4/85 or 5% vs 0%, p =.003) or corticosteroids (50/65 or 59% vs 64/178 or 36%, p< .001) as compared to the immunocompetent group. Nine immunocompromised individuals developed autoimmune encephalitis, 5 of whom did not have identifiable predisposing factors.
Conclusions:
Immunocompromised patients with encephalitis are more likely to have infectious encephalitis but can present with autoimmune encephalitis. Our results highlight the need for a broad differential of infectious and autoimmune encephalitis etiologies in the immunocompromised.