Abstract
Purpose::
To study the vitreous and serum pharmacokinetics of bevacizumab in human subjects following intravitreal injection.
Methods::
Following institutional review board approval, patients undergoing intravitreal bevacizumab for the following conditions, neovascular age-related macular degeneration and proliferative diabetic retinopathy were recruited. Following obtaining informed consent, 27 serum samples and 18 vitreous samples, at the time of planned vitrectomy, were obtained at time points from 2 to 45 days following an intravitreal injection of 1.25 mg of bevacizumab. A sandwich ELISA using VEGF165 coated to the plate and detection by an anti-human IgG antibody quantified unbound bevacizumab. Bevacizumab bound to VEGF was quantified by coating the plate with an anti-VEGF antibody and detection with an anti-human IgG antibody.
Results::
Peak intravitreal bevacizumab were detected at 2-5 days following injection and ranged from 80 - 170 ug/ml. Disappearance appeared to follow first-order kinetics with a t1/2 of approximately 10 days. No difference in levels of bound or free bevacizumab was seen. Serum levels of bevacizumab paralleled those of the vitreous levels with peak levels seen at 5 days following intravitreal injection and ranged from 20 - 687 ng/ml. However, in only 15% of patients were serum bevacizumab levels detectable at any point. Serum levels fell below the level of detection (15 ng/ml) by 45 days following injection. Differences between bound and free levels of bevacizumab varied between patients and ranged from 2 - 10%.
Conclusions::
Vitreal levels of bevacizumab disappear slowly supporting the clinical impression of longer duration of action of the full-length antibody. Absorption into the serum occurs in some patients with levels attaining pharmacologic levels and disappearance paralleling elimination from the vitreous supporting the concept of systemic monitoring of patients.
Keywords: drug toxicity/drug effects • age-related macular degeneration • diabetic retinopathy