1062 — Adverse Events after Intravitreal Injection of Ranibizumab and Bevacizumab for Age-Related Macular Degeneration
Kaboli PJ, Lund B, and Abramoff MD, Iowa City VA Healthcare System and University of Iowa; Alexander GC, University of Chicago; Cowan C, Washington DC VAMC; Ross JS, New Haven VAMC and Yale University;
Age-related macular degeneration (AMD) is the leading cause of blindness in the US. The principal treatments for AMD with choroidal neovascularization are vascular endothelial growth factor (VEGF) inhibitors. Ranibizumab (Lucentis) has FDA approval for this indication, while a similar preparation, bevacizumab (Avastin), is used off-label. Bevacizumab is proven to be equally effective, but with substantially lower cost (~4% the cost of ranibizumab). Bevacizumab requires pharmacy dose distribution and recent concerns about reported serious side effects have led to a ban on VA use. The objective of this observational cohort study was to evaluate the differential risk of serious adverse events for two equally effective treatments.
VA administrative data files (FY2005-10) were used to identify all patients who received intraocular VEGF inhibitors. Primary outcomes of interest were endophthalmitis within 30 days of injection and stroke within 1 year [i.e., ischemic, hemorrhagic, and transient ischemic attack (TIA)]. Multivariable logistic regression models adjusted for age, gender, injection count, and comorbidity.
A total of 7,477 patients received bevacizumab (N = 4,306) or ranibizumab (N = 3,171) over 5 years with mean age of 73.8 years and 3.1% were women. Endophthalmitis occurred in 0.39% of patients receiving bevacizumab (N = 18) and 0.27% receiving ranibizumab (N = 9); (Adjusted OR = 1.68; 0.74-3.82). Stroke rates at 1 year were 3.2% for bevacizumab (137) and 3.3% for ranibizumab (N = 104); (adjusted OR = 0.91; 0.69-1.19). Similar results were observed with sub-groups of stroke [ischemic = 2.04% vs. 2.11%; OR = 0.97; 0.70-1.33, hemorrhagic = 0.19% vs. 0.22%; OR = 0.77 0.27-2.22, TIA = 0.95% vs. 0.95%; OR = 0.99; 0.61-1.62)]. Two potential infectious clusters occurred at different hospitals, one involved bevacizumab (3 cases in 5 days) and the other ranibizumab (2 cases in 2 days).
Despite recent concern of higher risks of endophthalmitis and stoke after injection of bevacizumab, rates were the same as ranibizumab and similar to prior studies, although potentially underpowered. Infections and other complications are consequence of invasive procedures and should not be abandoned if they occur, but risks evaluated and processes engaged to minimize risk.
A recent VA ban on bevacizumab use due to greater risk than ranibizumab is not supported by these data. Risks, benefits, and cost implications should be further evaluated.