Intraocular Inflammation Following Intravitreal Injection of Anti-VEGF Medications for Neovascular Age-Related Macular Degeneration

Intravitreal injection of anti-vascular endothelial growth factor (VEGF) agents has become routine in the management of neovascular age-related macular degeneration (nAMD), diabetic macular edema, and macular edema due to retinal vein occlusion. Ranibizumab, a recombinant high-affinity humanized murine IgG1 monoclonal antibody fragment (Fab) against VEGF-A, was approved by the United States Food and Drug Administration (FDA) in 2006 for the treatment of nAMD. Aflibercept, a recombinant fusion protein with VEGF receptors (VEGFR)-1 & -2 fused to the Fc fragment of human IgG1, was approved by the FDA in 2011 for the treatment of nAMD. Bevacizumab, a humanized murine full-length (Fab and Fc portions) IgG1 monoclonal antibody against VEGF-A originally designed for the treatment of colorectal cancer, is often used off-label for the treatment of nAMD. One of the most feared complications of intravitreal injections regardless of the drug used is endophthalmitis, which may be infectious or non-infectious in etiology. Non-infectious, or sterile, intraocular inflammation has been reported in eyes that have received all of the aforementioned anti-VEGF agents.

Souied and colleagues retrospectively analyzed a large number of physician-level claims over a period of 18 months for intravitreal injections of either ranibizumab (n = 253,647) or aflibercept (n = 179,147) administered for nAMD.¹ They report a significantly higher rate of severe ocular inflammation, determined by claims for International Classification of Disease 9th Revision Clinical Modification (ICD-9) codes for endophthalmitis, following injection of aflibercept (1.06 per 1000 injections) compared to ranibizumab (0.64 per 1000 injections). Most cases were not associated with claims for intravitreal antibiotics (79% for ranibizumab, 71.4% for aflibercept). Information regarding vitreous biopsies and cultures, as well as corticosteroid use, were not available.

The rates of post-injection endophthalmitis from large prospective randomized clinical trials range from 0.02–0.10% of injections,^2–4 which is consistent with those reported by Souied and colleagues. Two recent meta-analyses including 350,535 and 445,503 injections reported rates of post-injection endophthalmitis of 0.056% and 0.046%, respectively.^5,6 One of these studies reported a significantly higher rate of 0.058% for bevacizumab compared to 0.029% for ranibizumab with the difference attributable to significantly more cases of culture-negative endophthalmitis in the bevacizumab group.⁶ However, the difference in endophthalmitis rates between bevacizumab and ranibizumab was not found when analyzing only data from prospective studies.

The majority of severe intraocular inflammation cases reported by Souied and colleagues were not associated with claims for intravitreal antibiotics and assumed to be non-infectious in etiology. The authors hypothesize that the difference in severe ocular inflammation rates may be attributable to aflibercept containing an Fc antibody portion, which is lacking in the ranibizumab molecule. The above-mentioned results of higher rates of culture-negative endophthalmitis following bevacizumab, which also contains an Fc portion, compared to ranibizumab are compatible with this hypothesis. Furthermore, the American Society of Retina Specialists (ASRS) Therapeutic Surveillance Committee (TSC) recently described 56 unsolicited voluntarily reported cases of sterile intraocular inflammation following aflibercept intravitreal injection.⁷ The median time to onset of inflammation following injection was 2 days, compared to 1 day in the study by Souied and colleagues, the median time to resolution was 28 days, and the final vision was overall unchanged, although some patients developed significant vision loss. Thus, non-infectious endophthalmitis appears to be a real complication of intravitreal aflibercept injection that occurs quickly following injection with outcomes seemingly favorable compared to infectious endophthalmitis.

Souied and colleagues provide interesting post-approval data on the use and potential complications of current anti-VEGF therapy in nAMD. It is important to note that while a statistically significant difference in rates of severe intraocular inflammation following injection of ranibizumab versus aflibercept was found, the rates were very low for both medications (≤0.1%) and within the range of previously reported rates for intravitreal anti-VEGF agents. It remains to be determined whether this difference is clinically significant.