Early Switch to Oral Antibiotics May Work for Brain Abscess, Review Finds

People treated for a brain abscess may not need weeks of uninterrupted intravenous antibiotics, according to a systematic review and meta-analysis published in Drugs & Therapy Perspectives. The analysis, which pooled five retrospective cohort studies covering 488 patients, found that shorter intravenous exposure before an oral switch was associated with higher treatment success than longer intravenous exposure, with an overall pooled odds ratio of 2.69 (95% confidence interval 1.66–4.37; p < 0.001) and no observed statistical heterogeneity between studies (I² = 0%). The authors caution that the pooled estimate reflects a directional association rather than the effect of a standardized intervention, because exposure thresholds and outcome definitions varied across the included studies.

Brain abscess is a focal collection of necrotic brain tissue enclosed by a surrounding membrane. It typically arises as a complication of contiguous infections such as sinusitis or otitis media, from hematogenous spread of distant infections including infective endocarditis or pulmonary infections, or following cranioencephalic trauma and neurosurgical procedures. Standard practice in the United States, based on clinical consensus, involves image-guided aspiration or surgical drainage followed by four to eight weeks of intravenous antibiotics. In contemporary United Kingdom cohorts, many patients historically received at least six weeks of intravenous therapy before oral transition was even considered. That prolonged approach carries real costs: extended hospital stays raise the risk of nosocomial infections and significantly increase healthcare expenditure.

The review, conducted according to PRISMA 2020 guidelines and prospectively registered in PROSPERO (CRD420251060028), searched five electronic databases—PubMed, Embase, Scopus, Web of Science, and the Cochrane Library—up to May 11, 2025. Eligible studies included patients aged 16 years or older hospitalized with brain abscess who received oral antibiotic therapy and reported at least one predefined clinical outcome. Case reports and small descriptive case series were excluded, as were studies without extractable outcome data. From 215 records initially identified, 99 duplicates were removed and screening ultimately yielded five retrospective cohort studies comprising 488 patients. Four of the five provided sufficient stratified data for the meta-analysis of treatment success; the fifth, by Evans and colleagues, was retained in the qualitative synthesis but excluded from pooling because its published report did not allow accurate subgroup classification by timing of switch.

The headline numbers are striking. Overall treatment success across the pooled data was 70.2% (321 of 457 patients). When outcomes were grouped by each study’s own definition of intravenous exposure duration, success reached 83.2% (154 of 185) in the shorter-exposure groups compared with 61.4% (167 of 272) in the longer-exposure groups. Because no uniform protocol existed across cohorts, the reviewers retained each study’s original threshold for ‘early’ switch—examples include less than 10 days, less than 14 days, and less than 3 weeks—to avoid misclassification. Individual study estimates varied: Besnard and colleagues showed the largest effect (OR 3.31, 95% CI 1.54–7.12), Lauda-Maillen and colleagues reported an OR of 3.13 (95% CI 1.19–8.22), while Hettle and colleagues (OR 1.98, 95% CI 0.63–6.20) and Asquier-Khati and colleagues (OR 1.81, 95% CI 0.56–5.91) showed directionally similar but individually non-significant estimates. Besnard and colleagues contributed the greatest weight to the pooled result at 39.98%.

Robustness checks lent support to the association. A leave-one-out sensitivity analysis using a random-effects DerSimonian–Laird model produced pooled odds ratios ranging from 2.34 to 2.91, with all confidence intervals entirely above the null and all p values between 0.000 and 0.008. Heterogeneity testing yielded a Q statistic of 1.08 (df = 3, p = 0.78), T² of 0.00, and I² of 0.0%. Funnel plot inspection suggested a generally symmetrical distribution around the pooled log odds ratio, though the authors emphasize that with only four pooled studies the power to detect asymmetry was limited and small-study effects cannot be excluded.

The timing of the switch itself varied considerably. Median time to oral transition, reported in three studies, ranged from 9 days (Besnard, interquartile range 5–15) to 19 days (Asquier-Khati, interquartile range 12–28); Lauda-Maillen reported a median of 10 days for early switchers and 23 days for late switchers. Notably, oral conversion did not shorten the total antibiotic course: across the three studies reporting total duration, study-level medians ranged from 42 to 70 days, comparable to or longer than the 42–56 days of intravenous therapy traditionally recommended. What the strategy appears to change is where that therapy is delivered. Median hospital stays ranged from 24 days (Lauda-Maillen) to 37 days (Asquier-Khati), with Hettle reporting 29.5 days, and Lauda-Maillen’s early-switch subgroup stayed a median of 17 days versus 29 days for the intravenous group.

Other outcomes were broadly reassuring but heterogeneous. Crude overall mortality across all five studies was 9.4% (46 of 488), reported at follow-up points ranging from one month to one year, which limits direct comparison; one-year mortality reached 18.1% in one cohort and 13.3% in another. Recurrence, formally reported in three studies, was low, with a crude overall proportion of 2.5% (10 of 399). Adverse events were systematically reported in only two studies, and both identified reversible thiamphenicol-induced cytopenia as the dominant toxicity—Besnard and colleagues documented at least one adverse effect in 76.3% of their cohort (145 of 190), while Asquier-Khati reported adverse drug reactions leading to antibiotic discontinuation in 25.0% (27 of 108). The reviewers stress that these events reflect specific drug toxicity rather than the oral-switch strategy itself, since they occur regardless of route of administration.

Significant caveats temper the findings. All included studies were retrospective, and risk of bias assessed with the ROBINS-I V2 tool was judged serious in three studies and moderate in two. Confounding was the most prominent problem: allocation to early switch was typically determined by clinician discretion rather than predefined criteria, baseline severity differed between comparison groups, and in some cohorts exposure classification depended on surviving long enough to reach the switch threshold, raising the possibility of immortal time bias. Patients selected for oral transition were generally more clinically stable, so the observed advantage of shorter intravenous exposure may partly reflect case selection rather than the treatment strategy itself. Treatment success was also defined inconsistently—ranging from favorable Glasgow Outcome Scale scores at three months to absence of a composite failure endpoint—and the reviewers did not contact study authors for unpublished subgroup data.

The findings arrive amid conflicting guidance. European guidelines from ESCMID in 2024 advise against routine oral switch due to insufficient evidence, whereas the British Society for Antimicrobial Chemotherapy suggested as early as 2000 that transition could be considered after one to two weeks of intravenous therapy in stable patients. The review’s authors suggest that a transition within the second week may be safe and effective in carefully selected patients—those who have undergone adequate surgical drainage, are clinically stable with resolution or improvement of fever, show normalization or significant decline in inflammatory markers such as CRP, and receive oral agents with high bioavailability and adequate blood–brain barrier penetration—while noting these criteria were not uniformly defined across studies and their individual contributions remain unquantified. The question may soon be settled prospectively: the ongoing ORAL trial (NCT04140903), an international open-label randomized non-inferiority trial planning to enroll 450 adults across multiple countries, is testing whether oral therapy after at least two weeks of intravenous treatment matches six to eight weeks of standard intravenous therapy, with a six-month composite primary endpoint. The stakes extend beyond high-income settings; in low- and middle-income countries, where nine of ten people cannot access basic surgical care and antimicrobial stewardship programs face limited pharmacy and microbiology capacity, reducing unnecessary intravenous days could meaningfully ease the burden of managing this serious infection.

Subject of Research: Effectiveness of intravenous-to-oral antibiotic switch strategies for brain abscess treatment

Article Title: Effectiveness of oral antibiotic therapy for brain abscess: A systematic review and meta-analysis of intravenous-to-oral switch strategies

Article References: Jimenez, I., Valladares, C., Núñez Vásquez, C. E., Sanchez Montenegro, A., Chia, J. E., Alderete Q, S. B., Quispe Juan de Dios, J., & Nina Yupanqui, E. (2026). Effectiveness of oral antibiotic therapy for brain abscess: A systematic review and meta-analysis of intravenous-to-oral switch strategies. Drugs & Therapy Perspectives. https://doi.org/10.1007/s40267-026-01262-9

Image Credits: AI Generated

DOI: 10.1007/s40267-026-01262-9

Keywords: brain abscess, oral antibiotics, intravenous-to-oral switch, systematic review, meta-analysis, antimicrobial stewardship, antibiotic duration, treatment success, retrospective cohort, hospital length of stay, ORAL trial, antimicrobial resistance