No abstract

Background Seizures can be part of the clinical presentation of central nervous system (CNS) infections. We describe patients suspected of a neurological infection who present with a seizure and study diagnostic accuracy of clinical and laboratory features predictive of CNS infection in this population. Methods We analyzed all consecutive patients presenting with a seizure from two prospective Dutch cohort studies, in which patients were included who underwent cerebrospinal fluid (CSF) examination because of the suspicion of a CNS infection. Results Of 900 episodes of suspected CNS infection, 124 (14%) presented with a seizure. The median age in these 124 episodes was 60 years (IQR 45–71) and 53% of patients was female. CSF examination showed a leukocyte count ≥ 5/mm3 in 41% of episodes. A CNS infection was diagnosed in 27 of 124 episodes (22%), a CNS inflammatory disorder in 8 (6%) episodes, a systemic infection in 10 (8%), other neurological disease in 77 (62%) and in 2 (2%) episodes another systemic disease was diagnosed. Diagnostic accuracy of clinical and laboratory characteristics for the diagnosis of CNS infection in this population was low. CSF leukocyte count was the best predictor for CNS infection in patients with suspected CNS infection presenting with a seizure (area under the curve 0.94, [95% CI 0.88 – 1.00]). Conclusions Clinical and laboratory features fail to distinguish CNS infections from other causes of seizures in patients with a suspected CNS infection. CSF leukocyte count is the best predictor for the diagnosis of CNS infection in this population.

Background Vilobelimab, a complement 5a (C5a)-specific monoclonal antibody, reduced mortality in critically ill COVID-19 patients in a phase 3 multicentre, randomized, double-blind, placebo-controlled study. As part of the study, vilobelimab concentrations and C5a levels as well as antidrug antibodies (ADAs) to vilobelimab were analysed. Results From Oct 1, 2020 to Oct 4, 2021, 368 invasively mechanically ventilated COVID-19 patients were randomized: 177 patients were randomly assigned to receive vilobelimab while 191 patients received placebo. Pharmacokinetic sampling was only performed at sites in Western Europe. Blood samples for vilobelimab measurements were available for 93 of 177 (53%) patients in the vilobelimab group and 99 of 191 (52%) patients in the placebo group. On day 8, after three infusions, mean vilobelimab (trough) concentrations ranged from 21,799.3 to 302,972.1 ng/mL (geometric mean 137,881.3 ng/mL). Blood samples for C5a measurements were available for 94 of 177 (53%) patients in the vilobelimab group and 99 of 191 (52%) patients in the placebo group. At screening, C5a levels were highly elevated and comparable between groups. In the vilobelimab group, median C5a levels were 118.3 ng/mL [IQR 71.2–168.2 ng/mL] and in the placebo group, median C5a levels were 104.6 ng/mL [IQR 77.5–156.6 ng/mL]. By day 8, median C5a levels were reduced by 87% in the vilobelimab group (median 14.5 ng/mL [IQR 9.5–21.0 ng/mL], p < 0.001) versus an 11% increase in the placebo group (median 119.2 ng/mL [IQR 85.9–152.1 ng/mL]). Beyond day 8, though plasma sampling was sparse, C5a levels did not reach screening levels in the vilobelimab group while C5a levels remained elevated in the placebo group. Treatment-emergent ADAs were observed in one patient in the vilobelimab group at hospital discharge on day 40 and in one patient in the placebo group at hospital discharge on day 25. Conclusions This analysis shows that vilobelimab efficiently inhibits C5a in critically ill COVID-19 patients. There was no evidence of immunogenicity associated with vilobelimab treatment. Trialregistration ClinicalTrials.gov, NCT04333420. Registered 3 April 2020, https://clinicaltrials.gov/ct2/show/NCT04333420