[Intraocular fluids analysis for etiologic diagnosis of presumed infectious uveitis].
PURPOSE: To study the etiologic distribution of patients with presumed infectious uveitis who underwent ocular fluid analysis. PATIENTS AND METHODS: We retrospectively analyzed vitreous and or aqueous humor samples of patients with presumed infectious uveitis, referred to the department of Ophthalmology of the University Hospital of Rouen, France, between January 1997 and June 2004. We excluded patients with postsurgical or endogenous endophthalmitis. We noted clinical features and intraocular sample analysis methods for each pathogen. RESULTS: The study included 42 patients, 24 men and 18 women, aged between 6 and 79 years (mean, 39.5 years). Uveitis was unilateral in 88% of cases and bilateral in 12%. Posterior uveitis was predominant (52%), followed by panuveitis (24%), anterior (14%), and intermediate uveitis (10%). Aqueous humor and vitreous analysis confirmed etiological diagnosis for 13 of 41 patients (31%) and three of six cases (50%), respectively. Inflammation was controlled or stabilized in all cases. The most pejorative visual outcome was observed for candidiasis and viral retinitis. Positivity of intraocular samples was variable, depending on the suspected pathogen, with results similar to those reported in other studies. DISCUSSION: Atypical features in potential infectious uveitis justify ocular paracentesis or vitrectomy, which are more efficient since molecular biological tools have become available. Sensitivity of aqueous humor analysis is high for posterior uveitis and suspicion of viral retinitis. New molecular variants applied to ocular samples will improve the etiological diagnosis of infectious uveitis, particularly for pathogens that are difficult to culture, such as fungi and bacteria.