A Cleveland Clinic case series found that tyrosine kinase inhibitors (TKIs) may act as an additional risk factor for retinal vein occlusion (RVO) in patients with existing vascular comorbidities.
Findings from a recent study published in Eye examined the occurrence of retinal vein occlusion (RVO) in cancer patients undergoing treatment with tyrosine kinase inhibitors (TKIs), identifying 11 patients across a decade-long review who developed the complication while on therapy.
The results point to a probable drug-event link and raise questions about how these patients should be monitored.
Give me some background first.
TKIs have changed the treatment landscape for a range of cancers. They work by blocking specific growth factor pathways that drive tumor growth, and they tend to come with fewer systemic side effects than traditional chemotherapy.
But they’re not without ocular risks. Side effects such as keratitis, macular edema, and periorbital edema have been documented with TKI use.
- And scattered case reports over the past several years have flagged a more concerning complication: RVO.
Some context, please.
RVOs are vascular occlusions of the retinal veins that can lead to macular edema, neovascular glaucoma, proliferative retinopathy, and vitreous hemorrhage. Common risk factors include hypertension, diabetes, hyperlipidemia, and hypercoagulable states.
Prior case reports have documented RVO following use of axitinib, regorafenib, and sorafenib.
However: Multi-patient data has been limited, making it hard to assess whether TKIs are truly contributing or whether these are coincidental events in a high-risk population.
Now, talk about the study.
A team at the Cleveland Clinic Cole Eye Institute ran a retrospective chart review of all patients at their institution who had both a TKI prescription and an RVO diagnosis between January 2014 and January 2024.
The initial search: 49 patients with concurrent RVO and TKI prescriptions.
From there, Investigators excluded patients:
- who never started the TKI
- Whose RVO predated TKI use
- Who discontinued the drug more than 6 months before the RVO
- Without a confirmed RVO diagnosis
- Without complete records
Who was included in the study?
That left 11 patients (12 eyes). Mean age at RVO presentation was 75.9 ± 9.8 years. Eight patients (72.7%) were male.
Risk factor profile: All 11 had pre-existing hypertension. Six (54.5%) had diabetes mellitus. Eight (72.7%) had hyperlipidemia. Five (45.5%) had chronic kidney disease. None smoked at the time of their RVO, though six were former smokers.
Ocular history: Six patients (54.5%) had non-proliferative diabetic retinopathy (NPDR) in both eyes, two (18.2%) had primary open angle glaucoma, and two (18.2%) had hypertensive retinopathy.
Cancer indications: Five patients (45.5%) had renal cell carcinoma, three (27.3%) had chronic myeloid leukemia (CML), two (18.2%) had chronic lymphocytic leukemia (CLL), and one (9.1%) had gastrointestinal stromal tumur (GIST). TKIs used were axitinib (n=5), imatinib (n=3), ibrutinib (n=2), and regorafenib (n=1).
Findings?
RVO developed after a mean of 2.8 ± 2.0 years on TKI therapy (range: 0.8 to 6.5 years).
Of the 12 RVOs, eight were central retinal vein occlusions (CRVOs) and four were branch retinal vein occlusions (BRVOs). Presenting visual acuity (VA) ranged from 20/30 to count fingers.
Treatment: Ten eyes received anti-vascular endothelial growth factor (VEGF) injections, including combinations of bevacizumab, aflibercept, faricimab, and dexamethasone implants. Two eyes went untreated.
Outcomes: At final follow-up, six eyes (50%) showed improvement in VA, three (25%) held steady, and three (25%) worsened. Final acuity ranged from 20/25 to count fingers.
The Naranjo score: The mean Naranjo Adverse Drug Reaction Probability Score came in at 5.2 (range: 3 to 8), placing the association in the “probable” category.
Tell me more.
One case stood out. A 73-year-old female on regorafenib for GIST developed a CRVO in her right eye roughly 1 year after starting the drug. Her vision dropped from 20/20 to count fingers. Imaging showed retinal non-perfusion and intraretinal fluid.
Seven months later, while still on regorafenib (at a reduced dose), she developed a CRVO in the left eye. She was switched to ripretinib and treated with monthly bevacizumab injections and dexamethasone implants in both eyes. As of last follow-up, her vision sat at 20/150 bilaterally.
And why does that matter?
That bilateral presentation earned the highest Naranjo scores in the cohort: 8 for each eye, driven by the adverse event reappearing after continued drug exposure.
Worth noting: Eight of the 11 patients (72.7%) were actually in remission at the time of their RVO, continuing on maintenance TKI therapy. Since malignancy itself can drive hypercoagulability, the fact that most had controlled disease makes the underlying cancer less likely to be the primary contributor.
Limitations?
This was a retrospective, single-center study with no control group. Causation could not be established.
The comorbidity issue: Every patient in the cohort had hypertension. Most had hyperlipidemia or diabetes on top of that. Each of these conditions is independently associated with RVO (the authors cited odds ratios of 3.5 for hypertension, 2.5 for hyperlipidemia, and 1.5 for diabetes), making it difficult to tease apart how much the TKI itself contributed.
On prevalence: The institution identified 11,538 patients prescribed a TKI over the study period, putting the crude observed frequency at roughly 0.1%. But the authors noted they couldn’t determine how many of those patients actually filled or took the medication, so the true rate is likely higher.
The Naranjo scale also has limitations in this context: It was designed for drug reactions broadly, and vascular occlusions are multifactorial events where drug rechallenge is rarely feasible. The authors cautioned that Naranjo scores may overestimate causality here.
Expert opinion?
No outside expert commentary was included in the study.
That said: The authors argued that TKIs targeting the VEGF signalling pathway may increase thrombosis risk by disrupting normal endothelial antiplatelet effects, which are typically mediated by nitric oxide and prostacyclin.
- Even TKIs targeting other pathways could throw off the balance between prothrombotic and antithrombotic factors through disruption of regular cell function and cytokine production.
Their position: Patients starting TKIs, particularly those with pre-existing vascular and ocular risk factors, should be counseled on the warning signs of RVO and know how to access ophthalmology quickly.
Anything else?
Three patients continued their TKIs after the RVO event (with a mean 2.0 ± 1.4 years of additional follow-up) and did not develop RVO in the other eye or see worsening in the affected eye.
Why some patients on TKIs develop the complication while others don’t, even on the same drug at similar doses, remains an open question.
In practice: The authors flagged that a baseline ophthalmic evaluation may be particularly worthwhile for older patients with multiple vascular or ocular risk factors who are starting VEGF-pathway TKI therapy.
For patients already on these drugs, the emphasis should be on prompt reporting of any sudden vision changes.
Take home.
RVO during TKI therapy appears to be rare but carries real consequences for vision.
For clinicians coordinating cancer care: Patients on TKIs with vascular risk factors should know to report sudden vision changes immediately.
As such: Early detection and timely referral to a retina specialist can make the difference between preserved and permanently lost vision. A baseline eye exam before starting TKI therapy is worth considering for high-risk patients.