Cardiovascular disease, mortality and treatment in retinal vein occlusion: a nationwide registry-based, and a clinical study

Background and aim
Retinal vein occlusion (RVO) is a common retinal vascular disease affecting millions of people worldwide. The occlusion causes compromised retinal blood drainage in the upstream area of the affected vein, often resulting in macular edema and decreased visual acuity.
Previously, treatment options were limited, and many patients only received treatment for complications such as retinal neovascularization and neovascular glaucoma. In 2011, intravitreal treatment with vascular endothelial growth factor (VEGF) inhibitors was approved for use in the treatment of RVO by the European Medicines Agency and quickly became the treatment of choice. The efficacy on visual acuity outcomes was convincing, and systemic safety evaluated in clinical trials was highly acceptable.
RVO has been associated with cardiovascular diseases (CVD), both as a predisposing risk factor and as a subsequent event. While previous studies have found correlations between RVO and subsequent risk of CVD and mortality, these studies were conducted before the introduction of VEGF inhibitors, where only a selected portion of all RVO patients were referred for treatment. Thus, a contemporary analysis would be needed to clarify these risks in a cohort reflecting todays RVO patients, in order to guide clinicians in handling and advising the patients.
Furthermore, intravitreally delivered VEGF inhibitors enter the systemic circulation, and controversy exist on the cardiovascular safety of the treatment. Danish registries might supply novel insight into this area, since the clinical studies on angiostatic agents were not powered to evaluate rare events and often excluded patients with high a priori risk of CVD and mortality, and since no larger population-based studies have evaluated the safety of these agents in RVO patients.
Irrespective of any unknown risks of the treatment, it is expensive for society and inconvenient for the patients, having to receive many intraocular injections with the burden of hospital visits and risk of iatrogenic complications. In the interest of lowering anti-VEGF treatment need, macular laser could potentially play a role, having been the first line treatment for patients with branch retinal vein occlusion (BRVO) for decades. The novel navigated laser delivery system allows for preplanning treatment, targeting specific retinal pathology and applying the laser spots with a safe and precise semi-automated system, which could prove useful in combination treatment in RVO patients. In addition, exploring novel biomarkers of this multifactorial disease would be a first step in advancement towards stratified or individualized patient treatment.
Thus, the aim of the PhD study was to provide a contemporary analysis on Danish nationwide registry data of the risks of cardiovascular diseases and mortality associated with: 1) RVO and 2) Anti-VEGF treatment of RVO (Study A).
Further, we aimed to: 3) evaluate the prospects of navigated central retinal laser in lowering the need for anti-VEGF treatment in patients with treatment-naïve BRVO and macular edema, and: 4) evaluate the value of novel biomarkers in predicting the re-treatment need as well as structural and functional outcomes in these patients (Study B).
Methods
Study A: The study was a prospective cohort study utilizing data from the Danish nationwide registries from 1998 to 2018. We evaluated the associations between RVO, CVD and all-cause mortality including subgroup analyses before and after the introduction of intravitreal therapy in 2011. Furthermore, we explored what effect VEGF inhibitory treatment had on cardiovascular events and mortality in the cohort of RVO patients registered from 2012 to 2018. Cox proportional hazards models were fitted, adjusting for possible confounding factors, and reported as hazard ratios (HR) with 95% confidence intervals (CI).
Study B: The study was a randomized, controlled clinical trial, including patients with treatment-naïve BRVO and macular edema. All patients received three loading doses of 2.0 mg aflibercept, after which one group received central retinal navigated laser (Group A) and one did not (Group B). All patients were followed up until month 12, receiving further aflibercept injections according to a pre-defined pro re nata (as-needed) regimen. We evaluated patients monthly with optical coherence tomography (OCT) and best corrected visual acuity. At baseline, patients were further evaluated with OCT angiography, retinal oximetry and microperimetry, the latter of which was repeated at month 12. Primary endpoint was number of patients needing anti-VEGF re-treatment after loading phase.
Main results and conclusion
We found that RVO patients had increased risk of overall, ischemic and non-ischemic CVD (HR 1.13 95% CI 1.09 to 1.17, HR 1.16 95% CI 1.11 to 1.21 and HR 1.23 95% CI 1.15 to 1.33), without marked difference between BRVO and central RVO patients, but with higher risks in a subgroup of patients registered after 2011. We found an increased risk of all-cause mortality in CRVO patients registered prior to 2011, but not in those registered after 2011. When evaluating the association between VEGF inhibitory treatment and CVD and mortality, we found an increased risk of intracranial hemorrhage (HR 1.66, 95% CI 1.02 to 2.71), but no other increased risks of other CVDs or all-cause mortality.
In BRVO patients with macular edema, navigated central laser as a supplement, did not lower re-treatment need (percentage of patients needing re-treatment after month three was 71% (95% CI 49 to 87%) and 80% (95% CI 56;93%) in groups A and B, respectively, p=0.72). Groups did not differ in number of injections after loading phase (1 (0;3) vs. 2 (1;3), p=0.43), change in best corrected visual acuity (+12.8 ± 9.4 vs. +15.1 ± 9.6 letters, p=0.48), change in central retinal thickness (-195 (-276;-145) vs. -181 (-263;-157) µm, p=0.81) or change in retinal sensitivity (+3.3 ± 2.3 vs. +4.1 ± 3.5 dB, p=0.67). Retinal arteriolar oxygen saturation tended to associate with no need for re-treatment after loading phase (OR 0.80 (95% CI 0.64 to 1.00), adjusted AUC 0.92, p=0.054), but none of the evaluated biomarkers had statistically significant predictive value on either re-treatment need, structural or functional outcome.
In conclusion, RVO patients had increased risk of cardiovascular diseases that was higher in patients registered after 2011, while CRVO patients after 2011 were found to have lower all-cause mortality than CRVO patients prior to 2011. Overall, we found treatment with VEGF inhibitors to be safe in regards to cardiovascular diseases and mortality, but an increased risk of intracranial hemorrhage should to be verified and quantified more precisely in larger population-based cohorts.
Supplementary treatment with navigated central laser had no significant effect in reducing treatment need in BRVO patients, but we found no detrimental effect of the laser evaluated by microperimetry. Supplementary laser in selected patients, due to e.g. non-response to VEGF inhibition or low compliance was not evaluated in the current study, and might be indicated. While none of the evaluated biomarkers proved significant in predicting re-treatment need, functional or structural outcomes, retinal arteriolar oxygen saturation tended to, and should be considered in future studies.