Query 15-11: Intravitreal bevacizumab for retinal conditions: real-world safety assessment
This project was originated with a query by CADTH and funded in June 2016 for one year. The objective of this study is to examine complications of anti-VEGF injections in four different cohorts: i) neovascular age-related macular degeneration cohort (AMD cohort); ii) diabetic retinopathy cohort; iii) diabetic macular edema cohort; and iv) retinal neovascularisation cohort.
We evaluated the risk associated with current exposure of anti-VEGF agents, examined as both therapeutic class and individual drugs, on primary and secondary outcomes in Marketscan data (2011-2015). We also presented updated results of the AMD cohort for anti-VEGF agents examined as individual drugs, using one additional year of Markescan data (2011-2016) in a report to DSEN in July 2018. We presented these alongside results from one previously published population-based nested case-control study using health care data sets in Ontario (April 2006-March 2011) comparing serious adverse events between bevacizumab and ranibizumab. We noted that bevacizumab was widely used in Ontario prior to public formulary coverage through the Ontario Drug Benefit (ODB) program. The use of bevacizumab for retinal diseases has become negligible in Ontario among patients 65 years of age and older (ODB eligible population) since the initiation of coverage within the ODB formulary for the Health Canada approved drug ranibizumab.
Summary of the results
Our Marketscan results suggested similar risk of primary (ocular) outcomes between bevacizumab or aflibercept versus the comparator ranibizumab.
Similar results were found in our analysis for systemic events. For the analysis with the composite of systemic outcomes using Marketscan cohort the adjusted HR was 1.11 (95% CI: 0.86-1.44) for bevacizumab and 0.86 (95% CI: 0.68-1.09) for aflibercept, both compared to ranibizumab.
Comparing specific systemic outcomes in the Marketscan data and the Ontario study several results appear to have roughly the same risk direction and risk magnitude. In the Ontario study although there was no definite statistical difference, the adjusted estimates did include the possibility that AMI events were slightly more associated with bevacizumab versus ranibizumab (adjusted OR, aOR: 1.23; 95% CI: 0.85-1.77). In the Marketscan cohort, we found even less suggestion of greater association of bevacizumab with AMI (adjusted HR, aHR: 1.17; 95% CI: 0.86-1.57, comparator was ranibizumab). The risk of stroke and CVA were also very similar between the two cohorts (aOR: 1.03; 95% CI: 0.67-1.60, in the Ontario study, and aHR: 1.00; 95% CI: 0.77-1.29, in the Marketscan cohort). For venous thromboembolism and DVT, both estimates showed indeterminant results. In the Ontario study, the aOR was 0.92 comparing bevacizumab to ranibizumab users (95% CI: 0.51-1.69), while in the Marketscan cohort the aHR was 1.38 (95% CI: 0.75-2.53).
We are close to finalizing agreements with Canadian Partnership for Tomorrow Cohorts (CPTP) the Ontario Health Study and the BC Generations Project, as well as with the Canadian Institute for Health Information (CIHI)’s National Prescription Drug Utilization Information System Metadata (NPDUIS). This will provide updated analysis on new agents. We have also initiated an application with CPTP’s Alberta’s Tomorrow Project as well as beginning the process to access administrative health data from Atlantic PATH.
This query has been finalized, a full report was submitted in December 2018. Our real-world population-based samples suggest that different anti-VEGF agents have similar safety profiles regarding systemic events. In the direct comparisons of different anti-VEGF agents in diabetic retinal conditions, we found an overall greater risk of adverse ocular outcomes (aHR: 1.52; 95% CI: 1.34-1.72) for bevacizumab versus ranibizumab. We also found a lower risk of vitreous hemorrhage (aHR: 0.75; 95% CI: 0.60-0.94) with aflibercept versus ranibizumab. However, total adverse ocular outcomes were not clearly different for aflibercept versus ranibizumab. In AMD patients, for aflibercept versus ranibizumab, we found a higher risk of uveitis (aHR: 1.67; 95% CI: 1.02 -2.74), and a lower risk of vitreous hemorrhage (aHR: 0.49; 95% CI: 0.26-0.94). However, total adverse ocular outcomes were not clearly different for aflibercept versus ranibizumab in AMD.
Moura et al. Safety of anti-vascular endothelial growth factor (anti-VEGF) intravitreal injections for the treatment of patients with age-related macular degeneration (AMD). 2017 (ICPE abstract)
Machado et al. Safety of intravitreal injections for the treatment of patients with diabetic retinopathy: a population-based study. 2017 (ICPE abstract)
Full report submitted to DSEN and query submitters.
For more information contact: Jessica (Lifang) Wang. Research Coordinator. [email protected] 514.934.1934 ext.44718