Intraocular pressure elevation after intravitreal injection (IVI) of anti-VEGF agents (anti-vascular endothelial growth factor drugs) has two patterns: acute intraocular pressure spike (immediately to tens of minutes after injection) and sustained intraocular pressure elevation (weeks to months or longer).
In 2004, pegaptanib (Macugen) was approved as the first anti-VEGF drug for neovascular age-related macular degeneration (nAMD). Since then, many drugs have been approved for intravitreal injection.
| Drug Name | Brand Name | Main Indications |
|---|---|---|
| Bevacizumab | Avastin | Off-label use (0.05 mL / 1.25 mg) |
| Ranibizumab | Lucentis | Age-related macular degeneration, diabetic macular edema (DME), retinal vein occlusion |
| Aflibercept | Eylea | Age-related macular degeneration, diabetic macular edema, retinal vein occlusion |
| Faricimab | Vabysmo | Age-related macular degeneration, diabetic macular edema |
Treatment is planned in two phases: an induction phase and a maintenance phase. The maintenance phase includes fixed dosing, pro re nata (PRN) dosing, and the treat-and-extend regimen.
Most side effects of intravitreal anti-VEGF injections are related to the injection procedure itself, including subconjunctival hemorrhage, eye pain, corneal epithelial damage, and increased intraocular pressure. These are often transient. Serious complications include endophthalmitis, lens damage, and retinal detachment. Systemically, arterial thromboembolic events require the most attention; the incidence of stroke and myocardial infarction with ranibizumab and aflibercept has been reported as 0.6–3% in phase III trials both in Japan and abroad.
Sustained intraocular pressure elevation is a rare complication1), but multiple large-scale studies have shown an association with anti-VEGF therapy. All anti-VEGF drugs carry a theoretical risk of increased intraocular pressure1). For aflibercept 8 mg, intraocular pressure elevation has been reported as an adverse event in 3% or more of cases3).
A transient increase in intraocular pressure immediately after injection occurs in nearly all cases, but it usually returns to baseline within one hour. Sustained intraocular pressure elevation occurs in 2.6–12% of cases and does not happen in everyone. For details, see the “Clinical Findings” section.
The following symptoms may occur during a rapid increase in intraocular pressure:
After injection, check vision. If there is no light perception, there is a high possibility of interruption of ocular blood flow due to high intraocular pressure. Immediately perform anterior chamber paracentesis to sufficiently lower the intraocular pressure.
The intraocular pressure spike after injection is a physical consequence of the volume expansion within the eye.
| Time Point | Change in Intraocular Pressure |
|---|---|
| Before injection | Average ≤18 mmHg |
| 1 minute after | Increases to 28.3–55.2 mmHg |
| 10–15 minutes later | Decreases to 22.8–25.8 mmHg |
| 30 minutes later | Decreases to 17.6–24.5 mmHg |
| 1 hour later | Returns to baseline |
A meta-analysis of 46 reports (2872 eyes) also reported a mean intraocular pressure increase of +23.41 mmHg immediately after injection, +2.51 mmHg at 30 minutes, and -0.63 mmHg at 1 day.
The incidence of long-term sustained intraocular pressure elevation varies by study, but is reported to be 2.6–12%. Representative studies are as follows:
The definition of sustained elevation is not standardized across studies, but “an increase of ≥6 mmHg from baseline and ≥21 mmHg on two consecutive visits” is commonly used as a general criterion.
Two studies have shown that ranibizumab is associated with greater intraocular pressure elevation than aflibercept. Compared with intravitreal steroids, anti-VEGF drugs have a lower risk of intraocular pressure elevation. The rate of intraocular pressure elevation with steroids (e.g., dexamethasone implant) reaches 18% at 1 year 2).
There have been case reports of acute angle-closure glaucoma after intravitreal injection. Rapid intraocular pressure elevation due to increased posterior chamber volume may lead to shallowing of the anterior chamber. It is recommended to assess the risk of angle closure, including gonioscopy, before starting intravitreal injections.
They can be used, but caution is needed. Pre-existing glaucoma is a risk factor for sustained intraocular pressure elevation, and recovery from post-injection pressure spikes may take longer. Treatment should be performed with enhanced intraocular pressure monitoring and preventive measures. For details, see the section on “Management and Prevention”.
If sustained intraocular pressure elevation is observed, evaluate glaucomatous changes with the following tests.
Post-injection intraocular pressure elevation must be differentiated from the following conditions.
In patients at risk for large IOP fluctuations, such as those with advanced glaucoma, the following preventive measures are effective:
The presence of vitreous reflux after injection is associated with lower IOP spikes. Spikes also tend to be reduced in pseudophakic eyes or eyes with a history of glaucoma surgery.
When there are coexisting events such as cataract surgery or vitrectomy, the interpretation of mean intraocular pressure changes depends on the underlying disease, pre-existing glaucoma, and number of injections.
Using glaucoma eye drops before injection or performing anterior chamber paracentesis can reduce acute intraocular pressure spikes. Slow injection of the drug and compression at the injection site are also important. For sustained intraocular pressure elevation, extending the injection interval or adding intraocular pressure-lowering medications may be considered.
The immediate intraocular pressure rise after injection is due to the rapid increase in intraocular volume caused by injecting the drug (usually 0.05 mL) into the vitreous cavity. This is a physical and mechanical consequence, occurring because aqueous humor outflow cannot keep up.
Outflow obstruction
Particulate obstruction: Protein aggregates from the drug package or silicone oil microdroplets from the syringe obstruct the trabecular meshwork.
Reduced outflow facility: A significant decrease in outflow facility measured by Schiøtz tonometry has been confirmed in patients who received 20 or more injections.
Inflammation and cell damage
Direct effect on trabecular meshwork cells: Bevacizumab 4 mg/mL has been shown to delay the metabolism and replication of trabecular meshwork cells in vitro.
Trabeculitis: Inflammatory reactions to monomeric antibodies or protein aggregates can cause trabeculitis with impaired aqueous humor outflow.
Inhibition of nitric oxide synthase (NOS) by anti-VEGF drugs has also been proposed as a mechanism. Decreased NO alters potassium and calcium ion movement in trabecular meshwork cells, changing cell contractility and reducing aqueous humor outflow through intercellular spaces. The effect of NO on smooth muscle has also been linked to systemic hypertension in patients receiving anti-VEGF therapy.
The association between repeated anti-VEGF injections and the onset or progression of glaucoma has been investigated in several large-scale studies.
Cui et al. (2019) analyzed 17,113 eyes and found that patients who received 14 or more injections over 2 years or 20 or more over 3 years had higher odds of initiating intraocular pressure-lowering therapy or receiving a new diagnosis of glaucoma4).
Eadie et al. (2017) conducted a big data analysis in British Columbia and reported that the risk ratio for glaucoma surgery in patients receiving bevacizumab injections for age-related macular degeneration was 2.48 times that of the control group. Seven or more injections were associated with further increased risk5).
It has also been reported that glaucoma patients take longer to recover from acute pressure spikes after injection. Compared to eyes without glaucoma, eyes with glaucoma have a lower rate of reaching below 30 mmHg within 15 minutes after injection.
Findings on the effect of injections on RNFL are mixed.
A meta-analysis by de Vries et al. (4 studies) concluded that RNFL significantly decreased by -3.34 μm at 1 year. However, many individual studies have not shown a clear association between injections and RNFL thinning.
RNFL thickness is influenced not only by injections but also by the underlying retinal disease (e.g., diabetic macular edema, retinal vein occlusion), making it difficult to interpret as a single factor.
