3B-01 - Risk of secondary thromboembolism event following intravitreal injections of anti-VEGF agents: A target-trial emulation study

Background: Intravitreal anti-vascular endothelial growth factors (anti-VEGF) agents have been the cornerstone of treatment for macular diseases driven by angiogenesis. However, the risk of secondary thromboembolic events (TEs) was not fully evaluated.

Objectives: To evaluate the risk of secondary TEs following intravitreal injection of anti-VEGF agents.

Methods: We emulated a target trial using data from the National Health Insurance Research Database between 2009-2020. Patients with a history of TEs (such as myocardial infarction, stroke, pulmonary embolism, or deep vein thrombosis), followed by an eligible diagnosis for intravitreal anti-VEGF treatment, were included. The eligible diagnosis for anti-VEGF agents includes retinal vascular occlusion, neovascular age-related macular degeneration, diabetic macular edema, and myopic choroidal neovascularization and the date of first eligible diagnosis was defined as index date. Patients were simultaneously assigned to both treatment and non-treatment arms by creating cloned copies of the study participants. A 180-day grace period was implemented to allow for the reimbursement application process and examinations before the initiation of anti-VEGF agents. The study outcomes included secondary arterial thromboembolism (ATE) and venous thromboembolism (VTE). We implemented an inverse probability of censoring weight based on a set of covariates, and a pooled logistic regression model was used to estimate the risk between treatment and non-treatment group. Subgroup analyses were conducted based on the time interval between the first TEs and the index date.

Results: We identified a total of 224,572 patients with eligible diagnosis of anti-VEGF agents with concomitant history of TEs. Among the study participants, 13,955 received anti-VEGF agents during the follow-up period. Receiving anti-VEGF agents within grace period was not associated with increased risk of secondary ATE (odds ratios: 0.99; 95% confidence intervals 0.96-1.04) and secondary VTE (1.00; 0.65-3.76). When the population was stratified by the time interval between the TEs and the index date, the use of anti-VEGF agents was associated with a higher risk of secondary ATE with time interval less than 360 days. However, patients who received anti-VEGF agents 360 days after the TEs did not show a higher risk of secondary ATE and VTE. The odds ratios based on time since the TEs decreased from 1.86 for less than 90 days to 1.61 for less than 180 days and 1.30 for less than 360 days.

Conclusions: Our study suggested that the risk of secondary TEs following anti-VEGF agents varied based on elapsed time since the patient's earlier thromboembolic event. Administration of anti-VEGF agents after 360 days of occurrence of TEs appears to be safe.