Endophthalmitis after intravitreal injections in France
Reviewed by Efrosini Papagiannuli

In this nationwide, retrospective, multicentre case series, the authors looked at the incidence and characteristics of endophthalmitis after intravitreal injections of anti-VEGFs and steroids, given for a variety of posterior segment conditions (macular oedema secondary to diabetes or retinal vein occlusion, neovascular macular degeneration, degenerative myopia and miscellaneous cases). They wished to describe the clinical and bacteriological features, management and outcome (visual acuity) of these eyes. A total of 316,576 intravitreal injections from 25 centres were identified that underwent injections from January 2008 to June 2013. The intravitreal drugs used consisted of ranibizumab, bevacizumab, triamcinolone acetonide and dexamethasone implant. Case note review detected 65 cases of presumed endophthalmitis (0.021% incidence), with a median time to presentation of four days post-injection. Presumed endophthalmitis was defined as any acute intraocular inflammation presenting within four weeks of intravitreal injection and requiring intravitreal antibiotics (vancomycin with ceftazidime). The commonest symptom was loss of vision and an offending germ was detected in nearly 45%. The most frequent pathogen was coagulase negative Staphylococcus (78.3%).There was a significant association (p=0.001) with an increased incidence of endophthalmitis if a disposable conjunctival mould assist device or prophylaxis with antibiotic / antiseptic were used. Sixty-five percent of patients post episode had worse visual acuity at three months follow-up and one patient developed phthisis. In 61.7% of patients intravitreal injections were restarted after resolution of the endophthalmitis at a median time of 113 days. The authors conclude that the incidence of endophthalmitis after intravitreals is low, but associated with a poor visual prognosis. The strength of this study lies in its large sample size, but they acknowledge the various limitations, primarily its retrospective nature, the possibility of missing data or underreported endophthalmitis cases, the heterogenicity of the management in different French centres (relating to the use of corticosteroids and initial vitrectomy) and the fact that systemic antibiotics were part of the management, which is current practice in France but not supported in the literature. Furthermore they point out that because of variability in the level of asepsis and antisepsis from country to country (use of drapes, face masks, surgical hat, sterile gloves and gowns), any generalisability of their results to other countries ought to be very careful. They recommend further studies to define specific guidelines for endophthalmitis management.

Endophthalmitis after intravitreal injections: incidence, management and visual outcome.
Dossarps D, Bron AM, Koehrer P, et al.
AMERICAN JOURNAL OF OPHTHALMOLOGY
Epub ahead of print.