Endophthalmitis after Open Globe Injuries

Abstract

O PEN GLOBE INJURIES ARE A SIGNIFICANT CAUSE OF both monocular and binocular visual loss. Poor vision from open globe injuries may be the result of numerous factors, including the impact on vital ocular structures, retinal detachment with or without proliferative vitreoretinopathy, ocular toxicity from a retained intraocular foreign body (IOFB), as well as endophthalmitis. Endophthalmitis is a particularly devastating complication. Endophthalmitis has been reported to occur in approximately 4% to 8% of open globe injuries and may be higher at 6.9% to 30% in IOFB injuries. Factors increasing the risk of endophthalmitis include delayed primary closure, presence of IOFB, lens disruption, rural setting, soil-associated injury, and age more than 50 years. Factors associated with better outcomes are no involvement of the crystalline lens and anterior segment IOFB. Factors associated with the worst outcomes are ball bearing (BB) or pellet gun injury and afferent pupillary defect. Endophthalmitis in an eye with an IOFB represents additional challenges in localization and removal of the foreign body (FB). In this issue of THE JOURNAL, Andreoli and associates report a large consecutive case series of open globe injuries treated between 2000 and 2007. The percentage of endophthalmitis cases was 0.4% (2/463) in open globe injuries without an IOFB and 3.2% (3/95) in cases with an IOFB. After a standardized protocol, open globe wounds generally were repaired within 24 hours of the injury and intravitreal antibiotics were not used routinely. IOFBs were removed at the time of primary repair and no prophylactic scleral buckles were placed. Systemic antibiotics, including intravenous vancomycin and ceftazidime, were given initially in these patients. Patients were admitted to the hospital for at least 48 hours to receive treatment with intravenous antibiotics, but were discharged without continuation of systemic antibiotics. In this study, risk factors for endophthalmitis included IOFB (P .03) and primary intraocular lens (IOL) placement (P .05). Some of the current controversies in the management of open globe injuries are the following: 1) time to initial closure and IOFB removal, 2) use of intravitreal antibiotics, 3) selection and duration of systemic antibiotics, and 4) IOL placement in eyes with trauma-induced cataract. Regarding timing of repair, IOFBs traditionally have been considered surgical ophthalmic emergencies, and initiation of surgery within hours has been recommended. Mieler and associates reported a retrospective series of 27 patients with IOFBs treated from 1986 through 1989. Surgery, including primary wound closure and pars plana vitrectomy to remove the IOFB, was performed at an average of 4.5 hours from presentation. Endophthalmitis did not develop despite positive vitreous cultures for bacterial pathogens in 7 of 19 cases evaluated. The authors of this small study recommend prompt surgery to lower endophthalmitis rates and to maximize outcomes. Although prompt surgery for open globe wound repair and IOFB removal is the traditional recommendation, at least 2 recent publications have indicated that immediate removal of IOFBs may not be as important as previously thought. Colyer and associates reported a retrospective study of 79 eyes with IOFBs resulting from combat-associated ocular trauma in Iraq between 2003 and 2005. In this study, primary closure occurred within hours of injury and both topical and systemic antibiotics were initiated early. IOFB removal was not immediate, but occurred between 72 hours and 4 weeks afterward, depending on patient stability. There were no cases of endophthalmitis in this series. Similarly, Ehlers and associates reported a large, retrospective series focusing on the management and outcomes of 96 IOFBs. The overall endophthalmitis rate in this study was 4%, but there was no significant association between time to IOFB removal and visual acuity outcome. Prophylactic intravitreal antibiotics may be considered in selected high-risk patients with open globe injuries. Intravitreal vancomycin (for gram-positive coverage) plus ceftazidime or amikacin (for gram-negative coverage) are common antibiotic choices. In Mieler and associates’ series of 27 patients, intravitreal antibiotics were given in selected cases and endophthalmitis did not develop in any case. Two large series reported that good results can be achieved without routine prophylaxis with intravitreal antibiotics. Both Ehlers and associates and Colyer and associates reported large series of IOFBs in which no intravitreal antibiotics were used and yet rates of progression to endophthalmitis were very low. In a multicenter, randomized clinical trial of 346 eyes involving the use of intravitreal gentamicin and clindamycin vs a balanced salt solution control group, rates of posttraumatic endophthalmitis were See accompanying Article on page 601. Accepted for publication Dec 9, 2008. From the Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami, Miller School of Medicine, Miami, Florida. Inquiries to Harry W. Flynn, Jr, Bascom Palmer Eye Institute, 900 NW 17th Street, Miami, FL 33136; e-mail: hflynn@med.miami.edu