Clinical and management factors linked to good visual outcomes in Acanthamoeba keratitis: A Systematic review and meta-analysis

Abstract

Purpose

Summarize clinical characteristics and management-related factors associated with good visual outcomes in Acanthamoeba keratitis (AK), aiming to evaluate their collective impact on a global scale.

Design

Systematic review and metanalysis.

Methods

The study was registered with PROSPERO (CRD42024538378). Searches were conducted in PubMed, Embase, Virtual Health Library, and MedXRiv on January 13, 2024. Primary studies involving at least 10 eyes diagnosed with AK were included. The risk of bias was assessed using validated tools depending on the study design. Data extraction focused on demographics, clinical features, diagnostic methods, and treatment outcomes. Narrative synthesis, proportion meta-analysis, I² heterogeneity assessment, and Mann-Whitney U test were performed.

Results

Eighty-one studies involving 3,854 individuals (4,033 eyes) were analyzed. Patients who achieved good visual outcomes (final Corrected Distance Visual Acuity [CDVA] of 20/25 or better) were younger (mean age 27.7 vs. 35.8 years, p=0.031). Significant predictors of good visual outcomes included radial keratoneuritis (OR=3.44, 95 % CI: 1.57–7.52) and corneal epitheliopathy alone (OR=2.22, 95 % CI: 1.27–3.89) at presentation. Conversely, the presence of a ring-shaped stromal infiltrate at presentation (OR=0.21, 95 % CI: 0.07–0.58) and the use of corticosteroids before diagnosis (OR=0.18, 95 % CI: 0.09–0.68) were associated with a lower likelihood of achieving good visual outcomes. Post-diagnosis corticosteroid use did not significantly affect outcomes.

Conclusion

Early-stage AK signs, indicated by corneal epitheliopathy and radial keratoneuritis without a ring-shaped stromal infiltrate, are associated with good visual outcomes. Corticosteroid use before diagnosis is detrimental, but post-diagnosis use does not significantly impact outcomes. Younger patients generally have a better prognosis.