          |
 |
OMIG Main Page | 2005 Abstracts | < Previous | Next > Microbiology of Endophthalmitis from India DW Stroman1, JG Bartell1, JJ Dajcs1, L Therese2 and HN Madhavan2 1Alcon Research, Ltd.,Ft. Worth TX, 2Research Vision Foundation, Chennai, India Purpose: Characterize the etiology of infectious endophthalmitis in Indian patients. Methods: Aqueous humor (AH) or vitreous (VIT) samples were collected from 45 patients with clinical signs of endophthalmitis and 6 patients without signs of infection. DNA was extracted and rRNA gene amplified using pan-bacterial or pan-fungal primers. Because many samples were polymicrobic, the mixed PCR amplicons were separated by denaturing HPLC and subsequently sequenced to identify microorganisms to the species level. Results: Endophthalmitis: Bacteria were cultured from 3 of 29 AH samples, fungi from 8 of 29. From 29 AH extracts, 18 had bacterial sequences, 9 had fungal sequences, and 2 had both bacteria and fungal. Species identification from 29 AH extracts revealed 41 bacteria and 11 fungi. 39% (16/41) of bacteria were Gram-positive, 61% were Gram-negative. 3 of 11 fungi were yeasts. Bacteria were cultured from 1 of 16 VIT samples, fungi from 3 of 16. From the 16 VIT extracts, 7 had bacterial sequences, 4 had fungal sequences, and 1 had both bacteria and fungal. Species identification from the 16 VIT extracts revealed 18 bacteria and 5 fungi. 6% (1/16) of the bacteria were Gram-positive, 94% were Gram-negative, 2 of 5 fungi were yeasts. Non-endophthalmitis: No bacteria or fungi were cultured or detected from the 1 AH and 5 VIT samples from patients without clinical signs of endophthalmitis. Conclusions: Species-level identification revealed that bacteria were detected in 70% of AH extracts versus 10% based upon culturing. Similarly, bacteria were identified in 50% of the VIT samples versus 6% based upon culturing. In AH, 40 % of bacteria were Gram- positive, whereas in VIT, less than 10% were Gram-positive. Bacterial endophthalmitis rates based upon culturing significantly underestimate the actual bacterial infection rate. Disclosure code: E OMIG Main Page | 2005 Abstracts | < Previous | Next >
|
|
