Endophthalmitis after Intravitreal Injection of Vascular Endothelial Growth Factor Inhibitors: Management and Visual Outcomes

• Published in: Ophthalmology (Rochester, Minn.) Vol. 125; no. 8; pp. 1279 - 1286
• Main Authors: Xu, Kunyong, Chin, Eric K, Bennett, Steven R, Williams, David F, Ryan, Edwin H, Dev, Sundeep, Mittra, Robert A, Quiram, Polly A, Davies, John B, Parke, 3rd, D Wilkin, Johnson, Jill B, Cantrill, Herbert L, Almeida, David R P
• Format: Journal Article
• Language: English
• Published: United States 01.08.2018
• ISSN: 1549-4713
• DOI: 10.1016/j.ophtha.2018.01.022

We describe the presentation of patients developing endophthalmitis after intravitreal injection with vascular endothelial growth factor (VEGF) inhibitors. Moreover, we evaluate the management by comparing the outcomes of immediate tap and injection of intravitreal antibiotics (TAI) versus initial surgical pars plana vitrectomy (PPV). Finally, we analyze the predictive factors of visual outcomes at 6-month follow-up. Retrospective, single-center, nonrandomized interventional study. Patients developing endophthalmitis after receiving an intravitreal injection of anti-VEGF agent between 2006 and 2016. All patients received a vitreous biopsy sent for cultures before the initiation of treatment: TAI group versus PPV with intravitreal antibiotics (PPV group). Best-corrected visual acuity (BCVA) at 6-month follow-up after treatment for endophthalmitis. A total of 258 357 intravitreal injections occurred over the course of the 10-year period, of which 40 patients (0.016%) had endophthalmitis within 3 weeks after injection. In total, 34 patients (85.0%) had pain and 25 patients (62.5%) had hypopyon on initial examination. Among 24 culture-positive cases, 66.7% of the causative organisms were coagulase-negative Staphylococcus, followed by Streptococcus species (10.0%). The best-corrected visual acuity (BCVA) (logarithm of the minimum angle of resolution [logMAR]) at 6-month follow-up was significantly worse for patients who had a positive culture for Streptococcus species (4.0; standard deviation [SD], 0.8) (approximately light perception) compared with those who had a positive culture for coagulase-negative Staphylococcus (0.4; SD, 0.3) (∼20/50) (P < 0.0001). Compared with the TAI group, a higher proportion of samples were culture-positive in the PPV group (90.9% vs. 48.3%, P = 0.03). There was no statistically significant difference in BCVA at 6-month follow-up between the TAI and PPV groups. Younger age (<85 years) and lower intraocular pressure (IOP) (≤25 mmHg) at presentation were predictive of achieving a BCVA of 20/400 or better at 6-month follow-up after treatment. Initial management (TAI vs. PPV), duration of symptoms, presence of pain, presence of hypopyon, presenting BCVA, and culture status (positive vs. negative) were not found to be predictive of visual outcomes at 6-month follow-up. No significant difference in BCVA at 6-month follow-up was detected between the TAI and PPV groups. Younger age and lower IOP at presentation were associated with better visual outcomes at 6-month follow-up.