Practice Patterns and Outcomes of Intravitreal Anti-VEGF Injection for Retinopathy of Prematurity: An International Multicenter Study

Purpose: To report practice patterns of intravitreal injections of anti-VEGF for retinopathy of prematurity (ROP) and outcomes data with a focus on retreatments and complications. Design: Multicenter, international, retrospective, consecutive series. Subjects: Patients with ROP treated with anti-VEGF injections from 2007 to 2021. Methods: Twenfty-three sites (16 United States [US] and 7 non-US) participated. Data collected included demographics, birth characteristics, examination findings, and methods of injections. Comparisons between US and non-US sites were made. Main Outcome Measures: Primary outcomes included number and types of retreatments as well as complications. Secondary outcomes included specifics of the injection protocols, including types of medication, doses, distance from limbus, use of antibiotics, and quadrants where injections were delivered. Results: A total of 1677 eyes of 918 patients (43% female, 57% male) were included. Mean gestational age was 25.7 weeks (range, 21.2–41.5 weeks), and mean birth weight was 787 g (range, 300–2700 g). Overall, a 30-gauge needle was most commonly used (51%), and the quadrant injected was most frequently the inferior-temporal (51.3%). The distance from the limbus ranged from 0.75 to 2 mm, with 1 mm being the most common (65%). Bevacizumab was the most common anti-VEGF (71.4%), with a dose of 0.625 mg in 64% of cases. Overall, 604 (36%) eyes required retreatment. Of those, 79.8% were retreated with laser alone, 10.6% with anti-VEGF injection alone, and 9.6% with combined laser and injection. Complications after anti-VEGF injections occurred in 15 (0.9%) eyes, and no cases of endophthalmitis were reported. Patients in the United States had lower birth weights and gestational ages (665.6 g and 24.5 weeks, respectively) compared with non-US patients (912.7 g and 26.9 weeks, respectively) (P < 0.0001). Retreatment with reinjection and laser was significantly more common in the US compared with the non-US group (8.5% vs. 4.7% [P = 0.0016] and 55% vs. 7.2% [P < 0.001], respectively). There was no difference in the incidence of complications between the 2 geographic subgroups. Conclusions: Anti-VEGF injections for ROP were safe and well tolerated despite a variance in practice patterns. Infants with ROP receiving injections in the US tended to be younger and smaller, and they were treated earlier with more retreatments than non-US neonates with ROP.