Clinical features and prognostic factors in adults with brain abscess

Abstract

Studies on brain abscess are hampered by single-centre design with limited sample size and incomplete follow-up. Thus, robust analyses on clinical prognostic factors remain scarce. This Danish nationwide, population-based cohort study included clinical details of all adults (≥18 years) diagnosed with brain abscess in the Danish National Patient Registry from 2007 through 2014 and the prospective clinical database of the Danish Study Group of Infections of the Brain covering all Danish departments of infectious diseases from 2015 through 2020. All patients were followed for 6 months after discharge. Prognostic factors for mortality at 6 months after discharge were examined by adjusted modified Poisson regression to compute relative risks with 95% confidence intervals (CI). Among 485 identified cases, the median age was 59 years [interquartile range (IQR 48-67)] and 167 (34%) were female. The incidence of brain abscess increased from 0.4 in 2007 to 0.8 per 100 000 adults in 2020. Immuno-compromise was prevalent in 192/485 (40%) and the clinical presentation was predominated by neurological deficits 396/485 (82%), headache 270/411 (66%), and fever 208/382 (54%). The median time from admission until first brain imaging was 4.8 h (IQR 1.4-27). Underlying conditions included dental infections 91/485 (19%) and ear, nose and throat infections 67/485 (14%), and the most frequent pathogens were oral cavity bacteria (59%), Staphylococcus aureus (6%), and Enterobacteriaceae (3%). Neurosurgical interventions comprised aspiration 356/485 (73%) or excision 7/485 (1%) and was preceded by antibiotics in 377/459 (82%). Fatal outcome increased from 29/485 (6%) at discharge to 56/485 (12%) 6 months thereafter. Adjusted relative risks for mortality at 6 months after discharge was 3.48 (95% CI 1.92-6.34) for intraventricular rupture, 2.84 (95% CI 1.45-5.56) for immunocompromise, 2.18 (95% CI 1.21-3.91) for age >65 years, 1.81 (95% CI 1.00-3.28) for abscess diameter >3 cm, and 0.31 (95% CI 0.16-0.61) for oral cavity bacteria as causative pathogen. Sex, neurosurgical treatment, antibiotics before neurosurgery, and corticosteroids were not associated with mortality. This study suggests that prevention of rupture of brain abscess is crucial. Yet, antibiotics may be withheld until neurosurgery, if planned within a reasonable time period (e.g. 24 h), in some clinically stable patients. Adjunctive corticosteroids for symptomatic perifocal brain oedema was not associated with increased mortality.