September 29, 1995
Clinical Alert to Ophthalmologists
This letter is to inform you of findings from the Endophthalmitis VitrectomyStudy (EVS) that warrant your attention prior to publication of the results.This randomized, controlled clinical trial is supported by the National EyeInstitute (NEI), part of the U.S. National Institutes of Health, and was conductedat 27 clinical centers nationwide.
The purpose of the EVS was:
- To determine the role of immediate pars plana vitrectomy in the managementof endophthalmitis occurring after cataract surgery or secondary intraocularlens implantation.
- To determine the role of intravenous antibiotics in the management ofendophthalmitis.
EVS patients had clinical evidence of bacterial endophthalmitis within sixweeks of cataract surgery or secondary lens implantation. The involved eyehad 1) either hypopyon or enough clouding of anterior chamber or vitreousmedia to obscure clear visualization of second-order retinal arterioles; 2)a cornea and anterior chamber in the involved eye clear enough to visualizesome part of the iris; and, 3) a cornea clear enough to allow the possibilityof pars plana vitrectomy. To enter the study, the patient had to have visualacuity worse than 20/50, but at least light perception.
The study received and analyzed outcome data from 420 patients who were randomlyassigned to either immediate pars plana vitrectomy (VIT) or to vitreous tapor biopsy (TAP). They also were randomly assigned to either intravenous antibiotictreatment or to no intravenous treatment. All patients received intravitrealantibiotics.
A final outcome evaluation of visual acuity was assessed at 9-12 months andthe results indicated that:
- There was no difference in final visual acuity or media clarity whetheror not systemic antibiotics were employed.
- Patients who presented with hand motions or better vision did equallywell whether they had immediate vitrectomy or immediate tap/biopsy.
- Patients who presented with vision of light perception only had muchbetter visual results with immediate pars plana vitrectomy than with tap/biopsy.With vitrectomy, these patients were three times more likely to achieve20/40 final vision (33% compared to 11%), twice as likely to achieve 20/100final vision (56% compared to 30%), and less than one half as likely toincur severe visual loss of less than 5/200 (20% compared to 47%).
Based on the above results, we conclude that routine immediate vitrectomywas of no additional benefit in patients who met EVS entry criteria and presentedwith hand motions or better vision. However, immediate vitrectomy was of substantialbenefit for those who presented with light perception only vision. In addition,systemic antibiotics provided no additional advantage to final visual outcome.
A scientific paper was submitted to the Archives of Ophthalmology forexpedited review. The paper has been reviewed and revised and is expectedto be published shortly. In addition, a symposium presenting the EVS resultswill occur at the annual meeting of the American Academy of Ophthalmologyin Atlanta on Thursday, November 2, 1995, at 10:00 a.m.
Carl Kupfer, M.D.
National Eye Institute
Bernard H. Doft, M.D.
Study Chair, EVS
Sheryl Kelsey, Ph.D.
Director, EVS Coordinating Center
Department of Epidemiology
University of Pittsburgh