PURPOSE: To study the prevalence, pathogenicity, and virulence of Propionibacterium acnes keratitis. METHODS: All cases of infectious keratitis submitted to the microbiologic laboratory of the New York Eye and Ear Infirmary between January 1, 2003, and April 6, 2006, were reviewed. Those cases in which P. acnes was recovered from culture were collected, and the medical records studied in depth. RESULTS: Of 1555 cultures submitted to the microbiology laboratory, 1329 (85.5%) were positive for growth. One hundred twenty four (9.3%) of the 1329 cases yielded P. acnes in at least 1 culture medium. Seventy eight (62.9%) of 124 cases had not been pretreated with antibiotics before culture, and 66.7% of the nonpretreated ulcers were monomicrobial (P. acnes only). Fifty one (65.4%) of 78 cases of the nonpretreated corneal ulcers presented with a cellular reaction in anterior chamber, 12 (15.4%) with a hypopyon (6 were monomicrobial), 21 (26.9%) had stromal thinning (12 of which were monomicrobial), and 2 (2.6%) progressed to perforation (both polymicrobial). Corneal ulcers associated with P. acnes tended to be small (66.7%) and were widely distributed: central (n = 17, 21.8%), paracentral (n = 44, 56.4%), and peripheral (n = 17, 21.8%). The most common risk factors were contact lens wear and previous history of ocular surgery. Three of the 78 nonpretreated patients were unresponsive to medical treatment and required surgery for keratitis. CONCLUSION: This study provides evidence that P. acnes is a frequent cause of bacterial keratitis and may cause significant morbidity.
PURPOSE: To study the prevalence, pathogenicity, and virulence of Propionibacterium acnes keratitis. METHODS: All cases of infectious keratitis submitted to the microbiologic laboratory of the New York Eye and Ear Infirmary between January 1, 2003, and April 6, 2006, were reviewed. Those cases in which P. acnes was recovered from culture were collected, and the medical records studied in depth. RESULTS: Of 1555 cultures submitted to the microbiology laboratory, 1329 (85.5%) were positive for growth. One hundred twenty four (9.3%) of the 1329 cases yielded P. acnes in at least 1 culture medium. Seventy eight (62.9%) of 124 cases had not been pretreated with antibiotics before culture, and 66.7% of the nonpretreated ulcers were monomicrobial (P. acnes only). Fifty one (65.4%) of 78 cases of the nonpretreated corneal ulcers presented with a cellular reaction in anterior chamber, 12 (15.4%) with a hypopyon (6 were monomicrobial), 21 (26.9%) had stromal thinning (12 of which were monomicrobial), and 2 (2.6%) progressed to perforation (both polymicrobial). Corneal ulcers associated with P. acnes tended to be small (66.7%) and were widely distributed: central (n = 17, 21.8%), paracentral (n = 44, 56.4%), and peripheral (n = 17, 21.8%). The most common risk factors were contact lens wear and previous history of ocular surgery. Three of the 78 nonpretreated patients were unresponsive to medical treatment and required surgery for keratitis. CONCLUSION: This study provides evidence that P. acnes is a frequent cause of bacterial keratitis and may cause significant morbidity.