Characteristics, management and outcome of Herpes Simplex and Varicella-Zoster virus encephalitis: a multicentre prospective cohort study

Objective To characterize differences between Herpes Simplex Virus and Varicella-Zoster Virus encephalitis (HSVE and VZVE) and other aetiologies of infectious encephalitis (IE), and to investigate the impact of time-to-aciclovir (ACV) start, ACV dose and duration on outcome. Methods We compared 132 HSVE, 65 VZVE, and 297 others IE enrolled in a prospective cohort (ENCEIF). We estimated associations between time-to-ACV start, dose or duration and outcome through adjusted odds-ratio (aOR) using logistic regression analysis. Results Prevalence of immunodepression differed among aetiologies: 15/65 (23%) for VZVE, 13/132 (10%) for HSVE and 30/297 (10%) for others IE (P<0.05), as was presence of seizure at admission: 27/132 (20%) for HSVE, 4/65 (6%) for VZVE and 43/297 (14%) for others IE (P<0.05). Poor outcome at hospital discharge (Glasgow outcome scale ≤3) differed among the three groups: 40/127 (31%) for HSVE, 12/65 (18%) for VZVE and 38/290 (13%) for other IE (P<0.05). Time-to-ACV start was associated with outcome in HSVE (aOR 3.61 [1.25 – 10.40]), but not in VZVE (aOR 0.84 [0.18 – 3.85]). Increased ACV dose was not associated with outcome among HSVE (aOR 1.25 [0.44 – 3.64]) nor VZVE (aOR 1.16 [0.24 – 5.73]). Conclusion HSVE and VZVE are distinct in clinical presentation, outcome, and prognostic factors. The impact of early ACV initiation was more apparent for HSVE than for VZVE, however, this could be due to VZVE smaller sample size and lower outcome rate leading to low statistical power, or due to potential distinct IE pathophysiology. To characterize differences between Herpes Simplex Virus and Varicella-Zoster Virus encephalitis (HSVE and VZVE) and other aetiologies of infectious encephalitis (IE), and to investigate the impact of time-to-aciclovir (ACV) start, ACV dose and duration on outcome. We compared 132 HSVE, 65 VZVE, and 297 others IE enrolled in a prospective cohort (ENCEIF). We estimated associations between time-to-ACV start, dose or duration and outcome through adjusted odds-ratio (aOR) using logistic regression analysis. Prevalence of immunodepression differed among aetiologies: 15/65 (23%) for VZVE, 13/132 (10%) for HSVE and 30/297 (10%) for others IE (P<0.05), as was presence of seizure at admission: 27/132 (20%) for HSVE, 4/65 (6%) for VZVE and 43/297 (14%) for others IE (P<0.05). Poor outcome at hospital discharge (Glasgow outcome scale ≤3) differed among the three groups: 40/127 (31%) for HSVE, 12/65 (18%) for VZVE and 38/290 (13%) for other IE (P<0.05). Time-to-ACV start was associated with outcome in HSVE (aOR 3.61 [1.25 – 10.40]), but not in VZVE (aOR 0.84 [0.18 – 3.85]). Increased ACV dose was not associated with outcome among HSVE (aOR 1.25 [0.44 – 3.64]) nor VZVE (aOR 1.16 [0.24 – 5.73]). HSVE and VZVE are distinct in clinical presentation, outcome, and prognostic factors. The impact of early ACV initiation was more apparent for HSVE than for VZVE, however, this could be due to VZVE smaller sample size and lower outcome rate leading to low statistical power, or due to potential distinct IE pathophysiology.