Purpose: Fungal ocular infections cause significant ocular morbidity, particularly when diagnosis and treatment are delayed. Accurate morphologic identification of Fusarium spp. beyond the genus is time-consuming and insensitive. It was the purpose of this study to examine the usefulness of the nuclear ribosomal RNA (rRNA) internal transcribed spacer regions (ITS1 and -2) to detect and differentiate Fusarium spp. responsible for ocular infections.
Methods: Fifty-eight archived isolates from ocular sources of 52 patients diagnosed with Fusarium keratitis at the Bascom Palmer Eye Institute (Miami, FL) from April 2000 to May 2007 were analyzed. The archived samples, which were initially classified according to morphologic characteristics, were analyzed by DNA sequence data generated from the ITS regions of the rRNA genes.
Results: Fifteen distinct sequences were identified among the 58 isolates. Sequence analysis identified the isolates as Fusarium solani (75%), F. oxysporum (16%), F. incarnatum-equiseti (5%), F. dimerum (2%), and one Fusarium sp. (2%) that was not classified within any species complex. Species identification based on sequence data correlated well with the morphologic classification when performed by a mycology reference laboratory, but a higher rate of mismatch was observed based on identification by a nonreference laboratory.
Conclusions: Most of the isolates of Fusarium ocular infections belong to the F. solani or F. oxysporum species complexes. Morphologic classification at the species level yielded inconsistent results at a general microbiology laboratory. In contrast, the sequence variation within the ITS region allowed reliable and faster discrimination of the isolates at both the genus and species level.