No Clear Benefit of Vitrectomy for Post-Injection Endophthalmitis in Age-Related Macular Degeneration

Abstract

Purpose : With the increasing use of anti-Vascular Endothelial Growth Factor (anti-VEGF) injections for conditions such as age-related macular degeneration (AMD), post-injection endophthalmitis and its treatment is a growing concern. No clear guidelines exist as to the role of vitrectomy versus intravitreal injection of antibiotics in this setting. A review of a single center's management and outcomes of post-injection endophthalmitis was performed.

Methods : A retrospective chart review was performed at a single institution, analyzing billing data from 1/1/2007 to 8/17/2016 and correlating it to clinic charts. Patient demographic data, visual acuities, infection source, treatment type and timing, as well as AMD treatment frequency was recorded for visits up to one year before and after onset of infection.

Results : Of the 559 unique patients identified with endophthalmitis, 134 were as a result of injections performed at the practice during this time frame, with an overall incidence ranging between 0.1% to 0.15%. Of these injections, 77 were anti-VEGF treatments given for AMD in a population of mainly pseudophakic (73%) women (62%).

On average, patients lost 0.48 ± 1.0 logMAR acuity 3 months after infection and 0.53 ± 1.0 logMAR at 12 months or last observation. Eyes that underwent vitrectomy in addition to antibiotic injections 32 (41%) had significantly more vision loss at time of last observation (0.89 logMAR) compared to eyes treated with injection alone (0.29 logMAR; p = 0.04). Even vitrectomies performed within a week of antibiotic injection had a trend towards more vision loss at last observation (p = .17). Analysis of Variance demonstrated worse visual outcomes with worse vision at presentation (p = 0.01) and better outcomes with earlier initiation of steroid treatment (p = 0.04).

AMD anti-VEGF treatment frequency decreased from 0.74 ± 0.23 injections per month before the onset of endophthalmitis to 0.37 ± 0.34 injections per month afterwards (p < .0000000001). Multivariable linear regression of post-infection anti-VEGF treatment rates correlated with pre-infection rates and visual acuity 3 months after infections (p = 0.0000004, R² = .35).

Conclusions : Though the rate of post-injection endophthalmitis is low, it leads to significant vision loss. No clear benefit of vitrectomy over antibiotic injection alone was found, although early initiation of steroid therapy was beneficial.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.