The authors report the results of a retrospective multicentre study conducted at 17 retina referral centres to investigate the microbial spectrum of endophthalmitis after anti-VEGF injections. A total of 83 cases of culture-positive endophthalmitis (58.4%) were included. The most frequent pathogens were coagulase-negative staphylococci (59%), viridans streptococci (15%) and staphylococcus aureus (12%). Non-staphylococcal and non-streptococcal pathogens accounted for 12% of all endophthalmitis cases in this study. In all cases, pre-injection betadine sterilisation of the fornix was performed. The number of endopththalmitis cases in the operation room was four and in the office was 37. Onset of endophthalmitis from injection (n=77), days 4.0±3.6. Data about post-treatment antibiotic eye drops were provided for 56 cases (64.3%). In 34 cases from four different centres (Melbourne n=15, Tel Aviv n=10, Liverpool n=8, Sydney n=1), post-injection antibiotic therapy after the last injection was denied and, also, in 22 cases from six different centres. Regimens of antibiotic therapy differed among the centres. Used drugs were moxifloxacin (40.9%; nine cases: Buenos Aires n=8; Perugia n=1), azithromycin (36.4%; eight cases: Dijon n=8), tobramycin (13.6%, three cases: Bangkok n=2, Chiang Mai n=1), and ofloxacin (9.1%, two cases: Dijon n=1, Coimbra n=1). VA after one month was 0.92±0.77 log MAR in non-streptococcal cases, versus 2.04±1.09 log MAR in streptococcus-associated cases; VA prior to endophthalmitis being predictive for visual outcome after one month. Initial treatment was by tap and intravitreal antibiotic injection in 49 patients (59%). Twenty-one of these patients (43%) did not require any further surgical treatment, 12 patients (25%) underwent at least one further tap and inject procedure, and in 16 patients (33%) secondary pars plana vitrectomy (PPV) was performed. Thirty-four patients (41%) were treated by primary PPV. Eighteen of these patients (53%) did not require any additional surgical procedure, 12 patients (35%) underwent a second PPV and in three patients (9%) a secondary tap and inject procedure was performed. In one patient (3%) secondary enucleation was conducted. All patients received at least one intravitreal antibiotic injection. The mean number of intravitreal antibiotic injections was 1.8±0.8 (range: one to four injections). The intravitreal antibiotic therapies used were: vancomycin plus ceftazidime in 72 patients (87%), vancomycin plus ceftazidime and dexamethasone in 10 patients (12%), and vancomycin plus amikacin in one patient (1%). Sixty-two patients (75%) received additional systemic antibiotic therapy. The most frequently used systemic antibiotics were vancomycin plus ceftazidime intravenously (32%), ciprofloxacin orally (27%), moxifloxacin orally (16%), and imipenem plus cilastatin intravenously combined with oral levofloxacin (16%). The use of postoperative antibiotics and performance of injections in an operating room setting significantly reduced the rate of streptococcus-induced endophthalmitis cases (p=0.01) for both. Limitations: Retrospective, small sample size, lack of preoperative OCT scans, and information whether ranibizumab was used in a prefilled syringe or from a vial was not provided. Some geographic areas were under-represented or not represented.

Endophthalmitis after intravitreal anti-VEGF injection: an international multicenter study.
Busch C, Iglicki M, Okada M, et al.
Share This
Sofia Rokerya

MBBS MRCOphth FRCSI, King's College University Hospital, UK.

View Full Profile