Abstract
Purpose: :
To assess the utility of polymerase chain reaction (PCR) of intraocular fluid in the diagnosis and treatment of uveitis patients.
Methods: :
Uveitis patients seen at the National Eye Institute between July 2007 and November 2011 with idiopathic uveitis who had aqueous or vitreous tap for PCR were identified through the microbiology department records. Patient records were reviewed for clinical characteristics associated with positive results and change in diagnosis and treatment based on the PCR results.
Results: :
Of the 102 patients identified, 52 patients had sufficient data. Twenty six patients were diagnosed as non-infectious idiopathic uveitis and PCR was negative in all. In 26 patients in whom infectious uveitis was suspected, 16 (61.5%) had a positive result for at least one diagnostic assay. Of these 16 patients, 9 (56.3%) were positive for CMV, 2 (12.5%) for HSV, 2 (12.5%) for VZV, 2 (12.5%) for Toxoplasmosis, and 1 (6.2%) for EBV. Of those with a positive result, 14 (87.5%) were immunocompromised. Change in initial diagnosis based on PCR results occurred in 7 of the 52 (13.5%) patients. While all patients suspected of viral retinitis on clinical exam had positive PCR confirmation, only 2 of the 4 patients (3 aqueous and 1 vitreous sample) (50%) with a clinical diagnosis of Toxoplasmosis, had positive PCR results (1 aqueous and 1 vitreous sample). Change in treatment based on PCR results occurred in 9 of the 52 patients (17.3%). Of these, 4 (44.4%) did not have complete change in choice of treatment, but rather allowed for more specific coverage. Eight patients with granulomatous panuveitis and a positive PPD in the absence of active systemic tuberculosis (TB) had mycobacterial PCR, of which all were negative. Serial quantitative PCR performed in 8 patients helped guide antiviral treatment. The predictive value of positive and negative PCR testing in this group was 100% and 72%, respectively; sensitivity and specificity were 62% and 100%, respectively. Importantly, there were no serious complications from the intraocular fluid sampling.
Conclusions: :
PCR from the aqueous or vitreous has a high positive predictive value and specificity and is a useful auxiliary test for confirming specific diagnoses in infectious uveitis allowing for more definitive treatment. PCR is most useful in immunocompromised patients with suspicion for viral retinitis; it is less useful in patients with suspected ocular toxoplasmosis and ocular tuberculosis.
Keywords: clinical laboratory testing