Predictors of unprovoked seizures in surgically treated pyogenic brain abscess: Does perioperative adjunctive use of steroids has any protective effect?

OBJECTIVE: Though outcome of brain abscess has been improved due to stereotactic drainage and appropriate antibiotic treatment, late unprovoked seizure often occurs after the abscess is resolved. The purpose of this study is to reveal the factors related to the late unprovoked seizure and effect of steroid on the prevention of seizure.

PATIENTS AND METHODS: From January 2002 to August 2016, 119 patients with supratentorial brain abscesses were retrospectively analyzed. Initial volume of abscess, surgical methods, use of antiepileptic drugs (AEDs), and seizure free survival according to the use of steroid were compared between seizure free and late unprovoked seizure patients groups. Factors related to the late unprovoked seizure were evaluated by multivariate logistic regression model.

RESULTS: All patients underwent surgery, which were either by burrhole aspiration or craniotomy. 22/119 patients (18.5%) had late unprovoked seizure. Initial abscess volume was significantly larger in the group of late unprovoked seizure (28.31 ± 22.68 cm3 vs. 17.03 ± 14.53 cm3 , p = 0.015). The mean time to the late unprovoked seizure was 487.7 ± 446 days (range, 11-1369 days). Steroid was used to relieve perilesional edema in 35 patients. Proportion of late onset seizure was not different between the group of steroid non-use group and steroid use (17.1% vs. 19%, p = 0.52) and mean seizure free periods of steroid non-use group and steroid use group were 40.75 ± 9.23 months and 48.85 ± 8.50 months, respectively (P = 0.89, by log rank test). Initial presentation of seizure at the diagnosis of brain abscess and initial volume larger than 20.89 cm3 were risk factors for late unprovoked seizure with odds ratio of 4.1 (95% C.I 1.44-11.69) and 3.08 (95% C.I 1.19-7.96), respectively.

CONCLUSION: Late unprovoked seizure in patients with brain abscess was affected by initial presentation of seizure and initial volume of the abscess whereas methods of surgical intervention and steroid use had no effect on the occurrence of late unprovoked seizure.