Floaters: Drugs That Cause or Treat Them

What Floaters Are and Why They Form

Floaters appear as dark spots, threads, or cobweb-like shapes drifting across the visual field. They are shadows cast on the retina by clumps of collagen fibers within the vitreous gel.

The vitreous body is 99% water and 1% structural macromolecules, primarily type II collagen and hyaluronic acid. With aging, hyaluronic acid concentration decreases regionally, allowing collagen fibrils to aggregate into visible opacities. This process, called vitreous syneresis, accelerates after age 50 and is nearly universal by the eighth decade 1. A posterior vitreous detachment (PVD) occurs in approximately 65% of people over age 65, and the detached vitreous cortex often produces a prominent ring-shaped floater known as a Weiss ring 2.

Myopia (nearsightedness) of -3 diopters or greater accelerates vitreous liquefaction by roughly a decade. Prior intraocular surgery, ocular trauma, and diabetes-related vitreous hemorrhage also contribute. The distinction between benign age-related floaters and those signaling a retinal tear is clinical and time-sensitive: a sudden onset of multiple new floaters with photopsia (light flashes) or a curtain-like visual field defect requires same-day dilated fundus examination 3.

Drugs That Can Cause or Worsen Floaters

Several medication classes carry documented associations with new or worsening floaters. The mechanisms range from crystalline vitreous deposits to posterior subcapsular cataract formation that mimics floater symptoms.

Corticosteroids (systemic and intravitreal). Long-term oral prednisone at doses above 10 mg/day and intravitreal triamcinolone acetonide injections produce posterior subcapsular cataracts in 25%, 40% of patients within 1 to 2 years 4. These lens opacities scatter light in patterns patients describe as floaters. Intravitreal steroid implants (dexamethasone, Ozurdex) also cause transient vitreous condensation around the dissolving pellet.

Fluoroquinolone antibiotics. A pharmacovigilance analysis of the FDA Adverse Event Reporting System (FAERS) identified elevated reporting odds for "vitreous floaters" with ciprofloxacin and moxifloxacin 5. The proposed mechanism involves fluoroquinolone-induced collagen cross-linking disruption in the vitreous, analogous to the tendon collagen damage well-documented with this drug class. The European Medicines Agency updated fluoroquinolone labeling in 2018 to include musculoskeletal and connective-tissue warnings that extend to ocular collagen.

Tamoxifen. At cumulative doses exceeding 100 g, tamoxifen produces refractile crystalline deposits in the inner retinal layers and occasionally in the vitreous 6. Patients report bilateral floaters; the deposits are visible on optical coherence tomography (OCT). Dose reduction or switching to an aromatase inhibitor may stabilize symptoms.

Rifabutin. Used in Mycobacterium avium complex prophylaxis for immunocompromised patients, rifabutin at doses above 600 mg/day causes anterior uveitis with vitreous cells mimicking floaters in up to 8% of patients 7. The mechanism is a direct drug-induced hypersensitivity uveitis.

Other implicated agents. Case reports link isotretinoin, bisphosphonates (zoledronic acid), and checkpoint inhibitors (nivolumab, pembrolizumab) to vitreous inflammation or deposits presenting as floaters, though large-scale data remain sparse.

Pharmacologic Treatments: What Exists and What Does Not

No drug has received FDA approval specifically for vitreous floaters. This absence reflects the challenge of dissolving or dispersing collagen aggregates inside a gel-filled compartment without damaging adjacent retinal tissue.

Ocriplasmin (Jetrea). This recombinant truncated form of human plasmin received FDA approval in 2012 for symptomatic vitreomacular adhesion (VMA). In the MIVI-TRUST trials (N=652 combined), a single 0.125 mg intravitreal injection resolved VMA in 26.5% of treated eyes versus 10.1% with placebo at day 28 8. While not approved for floaters per se, some retina specialists have explored ocriplasmin off-label for Weiss ring floaters associated with partial PVD, hypothesizing that completing the vitreous separation might reposition the opacity away from the visual axis. Published data on this off-label use remain limited to small retrospective series.

Hyaluronidase (Vitrase). Ovine-derived hyaluronidase was studied in the CLEARIT-1 trial for vitreous hemorrhage clearance. Though it accelerated hemorrhage resolution, it showed no clinically meaningful effect on collagen-based floaters, and the program was discontinued for this indication.

Topical NSAIDs and antioxidants. Some practitioners prescribe topical nepafenac or oral antioxidant supplements (lutein, zeaxanthin, vitamin C) based on the theory that reducing oxidative damage slows collagen cross-linking. No randomized controlled trial supports floater reduction from these interventions. The American Academy of Ophthalmology (AAO) Preferred Practice Pattern for posterior vitreous detachment does not recommend any supplement for floater management 9.

YAG Laser Vitreolysis

Nd:YAG laser vitreolysis uses focused 532 nm or 1064 nm laser pulses to vaporize vitreous opacities into gas microbubbles that dissipate over days. The procedure is performed in-office under topical anesthesia.

A prospective comparative trial by Shah and Heier (N=52 eyes) randomized patients with symptomatic Weiss ring floaters to YAG vitreolysis versus sham. At 6 months, 54% of treated eyes reported significant or complete improvement versus 9% of sham eyes (P<0.001) 10. Larger retrospective series by Delaney et al. report symptom improvement in 77%, 93% of patients with well-defined Weiss ring opacities.

Candidacy matters. Floaters that are dense, well-circumscribed, and located at least 3 mm from both the retina and the lens respond best. Diffuse cloud-like opacities scattered throughout the vitreous gel are poor candidates. Risks include transient IOP elevation (seen in 5%, 8% of cases), lens pitting if treatment is too anterior, and retinal damage if pulses are misdirected posteriorly.

The AAO has not issued a formal position statement endorsing YAG vitreolysis, citing insufficient long-term safety data. The procedure remains off-label for the Nd:YAG platforms currently marketed in the United States.

Pars Plana Vitrectomy

Pars plana vitrectomy (PPV) is the definitive treatment for visually significant floaters. By removing the vitreous gel entirely and replacing it with balanced salt solution, the procedure eliminates the substrate for floater formation.

A prospective study by de Nie et al. (N=116 eyes) demonstrated a patient satisfaction rate of 96% at 6 months post-vitrectomy for primary floater indication 11. The FLIES randomized trial (N=64) compared vitrectomy against watchful waiting and found the surgical group showed statistically and clinically significant improvement in NEI VFQ-25 scores (mean difference 12.4 points, P<0.001) 12.

Risks of PPV include:

• Cataract progression (affecting nearly all phakic patients within 2 years)
• Retinal detachment (1%, 3% lifetime risk post-vitrectomy)
• Endophthalmitis (approximately 0.03%, 0.05%)
• Persistent elevated IOP

Given these risks, ophthalmologists reserve vitrectomy for patients whose floaters produce measurable contrast sensitivity loss or whose quality-of-life impact is documented over at least 6 months of observation.

When Floaters Signal an Emergency

A single new floater noticed casually in bright light is rarely dangerous. A sudden shower of floaters is different.

Retinal tears occur in 8%, 15% of patients presenting with acute symptomatic PVD 13. If untreated, roughly 33%, 50% of retinal tears progress to rhegmatogenous retinal detachment. The clinical triad demanding urgent evaluation: sudden onset of multiple floaters, photopsia (flashing lights), and peripheral visual field loss.

Any patient on anticoagulants (warfarin, apixaban, rivaroxaban) who develops acute floaters should be evaluated for vitreous hemorrhage, which may present identically to benign floaters but carries different management implications. Diabetic patients with proliferative retinopathy face elevated vitreous hemorrhage risk; new floaters in this population warrant dilated exam within 24 hours.

How Floaters Are Diagnosed

The primary diagnostic tool is a dilated fundoscopic examination using slit-lamp biomicroscopy with a condensing lens (78D or 90D). This allows direct visualization of vitreous opacities and assessment of the peripheral retina for tears or detachment.

B-scan ultrasonography is used when media opacity (dense vitreous hemorrhage or cataract) prevents adequate visualization of the posterior segment. It confirms PVD, identifies retinal detachment, and can estimate floater density.

Optical coherence tomography (OCT) documents vitreomacular interface status and identifies tractional components that might influence treatment decisions. Enhanced-depth imaging OCT can visualize premacular floaters but does not reliably image mid-vitreous opacities.

Dynamic light scattering (quantitative ultrasound) is an investigational technique being studied to objectively measure floater burden. The AAO notes that subjective symptom severity correlates poorly with clinical appearance, making objective measurement tools a research priority 9.

Lifestyle and Conservative Management

For patients with mild, stable floaters who do not meet surgical criteria, several strategies reduce symptom perception:

Neuroadaptation occurs in most patients over 3 to 6 months. The brain learns to suppress awareness of stable floaters through the same mechanisms that filter out other constant visual stimuli (such as the nose in the visual field). Reassurance and education about this expected adaptation constitute first-line management per AAO guidelines.

Wearing sunglasses outdoors reduces contrast between floaters and bright backgrounds, making them less noticeable. Reading with a slightly dimmer background light (warm-toned, indirect) similarly decreases perceived floater prominence.

Adequate hydration and avoidance of excessive caffeine have been suggested anecdotally but lack controlled evidence. No dietary intervention has demonstrated floater reduction in published trials.

The Pipeline: Investigational Agents

Several pharmacologic approaches are in preclinical or early clinical development:

Microplasmin variants. Next-generation vitreolytic enzymes with improved specificity for vitreous collagen over retinal laminin are in preclinical testing. The goal is an intravitreal injection that liquefies vitreous collagen aggregates without affecting the internal limiting membrane.

Collagenase formulations. Purified bacterial collagenase (Clostridium histolyticum-derived) has been studied in animal models for vitreous liquefaction. Dose-finding remains challenging due to narrow therapeutic windows between floater dissolution and retinal toxicity.

Pharmacologic vitreolysis cocktails. Combination protocols using tissue plasminogen activator (tPA) plus SF6 gas to induce complete PVD have been reported in small case series for vitreomacular traction. Whether inducing complete PVD pharmacologically could reposition troublesome floaters away from the visual axis is under investigation.

As of May 2026, none of these agents have progressed beyond Phase I/II for a floater-specific indication. The relatively low morbidity of floaters compared to the risks of intravitreal injection makes regulatory approval for this indication particularly challenging.

Drug Interactions and Floater-Relevant Monitoring

Patients on medications associated with floaters benefit from periodic ophthalmologic screening:

• Chronic corticosteroid users (greater than 3 months): annual dilated exam and IOP measurement
• Tamoxifen users: baseline and annual OCT per American Society of Clinical Oncology recommendations 14
• Rifabutin: symptom-triggered evaluation; routine screening is not cost-effective given the self-limited nature of drug-induced uveitis after dose adjustment
• Checkpoint inhibitor recipients: ophthalmology referral for any new visual symptom given the broad spectrum of immune-related ocular adverse events

Patients taking GLP-1 receptor agonists (semaglutide, tirzepatide) should note that while diabetic retinopathy progression was observed in the SUSTAIN-6 trial (HR 1.76 to 95% CI 1.11, 2.78) 15, this was linked to rapid glycemic improvement rather than a direct drug effect on vitreous collagen. New floaters in a GLP-1-treated diabetic patient should prompt vitreous hemorrhage evaluation rather than drug discontinuation.