Duravyu

Vorolanib intravitreal insert (DURAVYU, EYP-1901) is an investigational product and is not approved by the FDA. Two identical phase 3 non-inferiority trials compared vorolanib intravitreal insert 2.7 mg every 6 months with aflibercept 2 mg every 8 weeks: LUGANO enrolled 432 participants and LUCIA enrolled 475 participants. On August 17, 2026, EyePoint reported that LUGANO did not meet its primary endpoint (mean change in best-corrected visual acuity, blended weeks 52 and 56) in the full dataset. In an ad hoc analysis that excluded 9 of 211 participants (4%) in the vorolanib insert arm who lost 15 or more letters from causes the sponsor attributed to conditions other than wet AMD, the sponsor reported non-inferiority with a nominal p-value of 0.0096. LUCIA topline data are expected in the fourth quarter of 2026. The sponsor plans a pre-NDA meeting in the fourth quarter of 2026 and an NDA submission in the first half of 2027, provided that LUCIA results are positive. Two phase 3 trials in diabetic macular edema (COMO and CAPRI) are fully enrolled, with topline data expected in the fourth quarter of 2027.

In the phase 2 DAVIO 2 trial, a single vorolanib insert dose (2,060 µg or 3,090 µg) was given at week 8. Mean plasma vorolanib concentrations were 0.000 pg/mL at week 8 (the visit at which the insert was administered), 23.206 pg/mL (2,060 µg; 46 participants) and 27.121 pg/mL (3,090 µg; 47 participants) at week 20, and 7.733 pg/mL (45 participants) and 10.392 pg/mL (50 participants) at week 56. Mean aqueous humor concentrations at week 20 were 17.426 ng/mL (43 participants) and 25.066 ng/mL (49 participants). Plasma concentrations were reported in picograms per milliliter and aqueous humor concentrations in nanograms per milliliter. Mean plasma concentrations of the main metabolite did not exceed 2.46 pg/mL at any sampled visit.

Through week 56 of LUGANO, the ocular adverse events in the study eye with the largest absolute differences between vorolanib intravitreal insert 2.7 mg (211 participants) and aflibercept 2 mg (221 participants) were conjunctival haemorrhage (11.4% and 5.0%), vitreous floaters (9.0% and 3.6%), and neovascular age-related macular degeneration (7.1% and 1.4%). Intraocular inflammation (0.5% in each arm) and retinal detachment (0.5% in each arm) were infrequent, and the sponsor reported no cases of insert migration into the anterior chamber, retinal vasculitis, or severe intraocular inflammation. Nine participants (4%) in the vorolanib insert arm lost 15 or more letters from causes attributed by the sponsor to geographic atrophy (6), glaucoma (2), or retinal detachment (1); no participant in the aflibercept arm had such a loss. These data are preliminary and pending final analysis.

In LUGANO and LUCIA, all participants received scheduled aflibercept 2 mg from day 1 through week 8. Participants randomized to vorolanib intravitreal insert received 2.7 mg (two inserts in a single injection) at week 8 and were redosed every 6 months (every 24 weeks), with the second dose at week 32 and the third planned dose at week 56, for a total of four doses over the 2-year trials. Supplemental aflibercept was permitted in both arms when protocol criteria were met.

In the phase 3 program, vorolanib intravitreal insert 2.7 mg is given by in-office intravitreal injection every 6 months after aflibercept loading, and it requires no cold storage. Approved labeling for aflibercept 2 mg recommends monthly injections for 3 months followed by injections every 8 weeks. Aflibercept 8 mg is given monthly for 3 doses followed by every 8 to 16 weeks, with extension to every 20 weeks considered after 1 year of successful response. Faricimab is given monthly for 4 doses followed by one of three regimens with intervals of up to 16 weeks. Ranibizumab 0.5 mg is given monthly, with less frequent regimens described as not as effective. The ranibizumab port delivery system is refilled every 24 weeks but requires surgical implantation in an operating room and carries a boxed warning for endophthalmitis. The sustained-release TKI axitinib intravitreal hydrogel (AXPAXLI) is investigational; its sponsor plans an NDA submission in the fourth quarter of 2026. No head-to-head trials of vorolanib intravitreal insert against aflibercept 8 mg, faricimab, ranibizumab, the port delivery system, or axitinib intravitreal hydrogel were identified; the phase 3 comparator was aflibercept 2 mg every 8 weeks.

Wet (neovascular) age-related macular degeneration (AMD) is the late-stage form of AMD in which abnormal blood vessels (macular neovascularization, historically called choroidal neovascularization) grow beneath or into the macula. The American Academy of Ophthalmology (AAO) estimates that about 80% of people with AMD have the non-neovascular form, while the neovascular form accounts for most of the severe vision loss. Unmodifiable risk factors are increasing age, Northern European ancestry, and genetic factors; cigarette smoking is the principal modifiable risk factor. In the Global Burden of Disease Study 2021, tobacco accounted for 9.96% of age-standardized disability-adjusted life-years attributed to AMD.

In a Bayesian meta-analysis of cohort studies in populations of European ancestry, the annual incidence of neovascular AMD in White US adults aged 50 years or older was 1.8 per 1,000, and the incidence of any late AMD was 3.5 per 1,000 (about 293,000 new cases per year). A US cost-benefit analysis estimated 168,400 new-onset neovascular AMD cases in 2018. In the Beaver Dam Eye Study (3,917 participants), the 15-year cumulative incidence of exudative AMD was 4.4% in participants aged 75 years or older at baseline and 0.4% in those aged 43 to 54 years.

In a Bayesian meta-regression for 2019 using CDC Vision and Eye Health Surveillance System data, an estimated 1.49 million US adults aged 40 years or older (crude prevalence 0.94%) had late-stage AMD (choroidal neovascularization and/or geographic atrophy) and 18.34 million had early-stage AMD. The 2004 Eye Diseases Prevalence Research Group estimate was 1.75 million US adults with neovascular AMD and/or geographic atrophy (1.47%). No recent US estimate specific to the neovascular subtype was retrieved. Globally, the pooled prevalence of late AMD in people aged 45 to 85 years was 0.37% and of any AMD 8.69%, with 196 million people projected to have AMD in 2020 and 288 million in 2040.

In a meta-analysis of 53 studies including 4,362 participants with untreated neovascular AMD, mean visual acuity loss was 1 line at 3 months, 2.7 lines at 12 months, and 4 lines at 24 months; severe vision loss (more than 6 lines) occurred in 21.3% at 6 months and 41.9% by 3 years, and the proportion with visual acuity worse than 20/200 rose from 19.7% at baseline to 75.7% by 3 years. In the sham arm of the MARINA trial, 62.2% of participants lost fewer than 15 letters at 12 months and mean visual acuity decreased by 10.4 letters. The AAO attributes nearly 90% of AMD-related severe visual acuity loss (20/200 or worse) to the neovascular form. The AAO cites Salisbury Eye Study data linking visual acuity loss to activities of daily living and mortality risk, and a US cost-utility model estimated that 11 years of anti-VEGF therapy averts a 1.0 year-of-life loss.

The AAO Preferred Practice Pattern describes diagnosis based on history (metamorphopsia, decreased vision, scotoma), examination including stereoscopic biomicroscopy of the macula, and imaging. Optical coherence tomography (OCT) is described as important for detecting subretinal and intraretinal fluid and for monitoring response to therapy; fluorescein angiography is indicated to detect and characterize macular neovascularization, and OCT angiography had a pooled sensitivity of 0.87 and specificity of 0.97 for detecting macular neovascularization in a meta-analysis cited by the guideline.

In the CATT trial, 480 of 1,059 eyes (45.3%) receiving anti-VEGF therapy developed scar by 2 years (fibrotic scar 24.7%, nonfibrotic scar 20.6%). A 2026 review reports that 20% to 67% of treated participants develop macular fibrosis within 2 to 7 years. In a US TriNetX cohort of 92,849 people with neovascular AMD, 21.24% had a diagnosis of visual impairment, the highest proportion among the single ocular diseases analyzed. Globally, 1.85 million people were blind and 6.23 million had moderate or severe vision impairment due to AMD in 2020. In a French nationwide cohort of 1,662 people with incident treated neovascular AMD, the condition was associated with a 19% increase in the risk of disability in activities of daily living and of mood disorders.

Time trade-off utility values in participants with AMD decline with visual acuity in the better-seeing eye, from 0.89 at 20/20 to 20/25 to 0.40 at counting fingers to light perception (80 participants); participants in the lowest visual acuity group were willing to trade 60% of remaining lifetime for perfect vision. National Eye Institute Visual Function Questionnaire-25 mental health subscale scores were lowest in participants with visual acuity of 20/200 or worse (median 43.8). Among 130 participants in a Norwegian survey, 33.1% reported stress in the week before an injection and 54.6% described the injection as uncomfortable.

In 2024, Medicare Part B spending was $2.37 billion for aflibercept 2 mg (Eylea), $1.92 billion for faricimab (Vabysmo), $0.67 billion for aflibercept 8 mg (Eylea HD), and $0.13 billion for ranibizumab (Lucentis); these totals include all approved indications (for example, diabetic macular edema and retinal vein occlusion), not neovascular AMD alone. In 2020, aflibercept was the second-highest Medicare Part B drug by total spending (more than $3 billion). In a US commercial claims cohort of 6,076 participants with neovascular AMD, annual neovascular AMD-related outpatient visit costs were $8,658 for active choroidal neovascularization compared with $2,406 for inactive disease. In a US survey of 200 participants, mean annual societal ophthalmic costs were $39,910 with visual loss and $6,116 with good vision, rising to $82,984 at 20/800 or worse. Each treatment visit required a mean of 90 minutes in clinic, and patients reported almost 12 hours per visit including travel and recovery. In the IRIS Registry, 11.6% of 156,327 treatment-naive participants were lost to follow-up within 12 months.

Caregivers of people receiving anti-VEGF therapy carry a documented time and cost burden. In a UK survey of 250 patient-caregiver pairs, 70.8% of caregivers had attended 10 or more appointments in the previous year, 24.8% required time off work, and caregivers reported loss of income as a consequence. In a US time-and-motion study, caregivers took time away from work (22%) and personal activities (28%) to provide transportation. In a nine-country survey, 56.7% of 910 participants were usually taken to appointments by a caregiver, and a caregiver being unable to take the patient was the most common reason for a missed appointment (25.9%). In a US societal cost survey, direct nonmedical costs, primarily caregiver costs, made up 67.1% of societal ophthalmic costs in participants with visual loss.

Diagnosis and retreatment decisions rely on retinal imaging, principally OCT (CPT 92134), with fluorescein angiography when indicated. In a US Medicare analysis covering 2008 to 2015, reimbursement for OCT imaging used to guide anti-VEGF therapy for neovascular AMD totaled an estimated $0.8 billion, while OCT-guided personalized dosing was estimated to have saved Medicare $10.3 billion in anti-VEGF drug and injection costs, for a net government saving of $9.0 billion. The estimated annual beneficiary copayment for OCT imaging was $6, compared with $242 for aflibercept on a personalized injection schedule. No current Medicare fee schedule amount for OCT or fluorescein angiography was retrieved from a primary CMS source.

Current treatment of neovascular AMD relies on repeated intravitreal injections of anti-VEGF agents given monthly, every 8 weeks, or at extended intervals of up to 16 weeks under treat-and-extend or personalized-interval regimens, with continuous office monitoring. Real-world patients receive fewer injections than trial participants and have smaller visual gains: in 49,485 treatment-naive United States eyes, the mean gain at 1 year was 0.95 letter after a mean of 7.3 injections; in the eight-country AURA study (2,227 patients), mean gains were +2.4 letters at year 1 and +0.6 letters at year 2 with 5.0 and 2.2 injections; and in LUMINOUS (3,379 treatment-naive patients with 1-year data), the mean gain was 3.1 letters with 5.0 injections. In one United States retina practice, 2,003 of 9,007 patients (22.2%) were lost to follow-up for 12 months or longer after an injection. In CATT, vision gains from the first 2 years were not maintained at 5 years. Agent-specific safety limitations include intraocular inflammation with retinal vasculitis or occlusion after brolucizumab (definite or probable intraocular inflammation in 4.6% of brolucizumab-treated eyes in HAWK and HARRIER) and endophthalmitis in 11 of 555 participants (2.0%) who received the port delivery system with ranibizumab, which carries a boxed warning and was voluntarily recalled in 2022 before reintroduction in 2024. Payer prior authorization requests for anti-VEGF agents were approved in 96.2% of cases but delayed care by more than 24 hours in 59.6% of approvals.

Vorolanib intravitreal insert is investigational and has not been approved by the FDA. EyePoint anticipates a New Drug Application in the first half of 2027, contingent on LUCIA topline results expected in the fourth quarter of 2026. LUGANO did not meet its primary best-corrected visual acuity noninferiority endpoint in the full dataset, so the approved population and label are not yet defined. In the phase 3 trials, the insert (2.7 mg, delivered as 2 inserts in a single injection) was given every 6 months starting at month 2, after aflibercept loading doses, with supplemental aflibercept allowed by prespecified disease-activity criteria. The trials enrolled both treatment-naive participants (75.4% of the LUGANO insert arm) and previously treated participants, who needed at least 2 anti-VEGF injections in the prior 6 months. The phase 2 DAVIO 2 trial enrolled only previously treated participants. Based on these designs, the anticipated use is maintenance therapy after anti-VEGF induction, with rescue anti-VEGF injections as needed, in treatment-naive or previously treated patients. The sponsor expects physicians to incorporate the product into existing treatment approaches. Administration is an office-based intravitreal injection, the same care setting as current anti-VEGF therapy. An independent 2025 review stated that larger controlled trials are needed to define the clinical role of sustained-release tyrosine kinase inhibitors.

Formal subgroup results from LUGANO have not been published; EyePoint stated that subgroup analyses would be presented at retina meetings beginning with the Retina Society meeting of September 23 to 26, 2026. In the topline data, 9 of 211 participants (4%) in the vorolanib insert arm lost 15 or more letters from causes the sponsor attributed to conditions other than wet AMD: geographic atrophy (6), glaucoma (2), and retinal detachment (1). No aflibercept participants lost vision from such causes. Baseline history of glaucoma in the study eye was 11.4% in the insert arm and 5.9% in the aflibercept arm. Geographic atrophy lesions were closer to the fovea at baseline in the insert arm (825 versus 1,009 microns). Excluding these 9 participants was an ad hoc analysis. The prespecified supplement-free subgroup (113 participants, 54%) met noninferiority (nominal p = 0.0035). The EyePoint chief executive officer stated that visual acuity results were similar in treatment-naive and treatment-experienced participants, but no numerical subgroup results have been released. In DAVIO 2, which enrolled only previously treated participants, EyePoint reported that visual and anatomic outcomes were not meaningfully influenced by baseline visual acuity, duration of wet AMD, or prior treatment burden. These subgroup findings are sponsor-reported, partly post hoc, and not peer reviewed.

In the phase 3 LUGANO trial (432 participants; vorolanib insert 2.7 mg every 6 months versus aflibercept 2 mg every 8 weeks), the vorolanib insert did not achieve noninferiority for mean change in best-corrected visual acuity at weeks 52 and 56 in the full dataset. An ad hoc analysis excluding 9 of 211 insert-arm participants met the noninferiority criterion (nominal p = 0.0096). Relative to on-label aflibercept, the insert reduced treatment burden by 42% (nominal p < 0.0001), about 2 fewer injections through week 56. Of insert-treated participants, 76% were supplement-free through week 32, 54% through week 56, and 79% received zero or one supplemental injection through week 56. The mean difference in central subfield thickness versus aflibercept was 4 microns at week 56. Rates of cataract, elevated intraocular pressure, and intraocular inflammation did not differ between arms. In the phase 2 DAVIO 2 trial of previously treated participants, the 2 mg and 3 mg doses differed from aflibercept by -0.3 and -0.4 letters, respectively, at weeks 28 and 32, and 65% and 64% of eyes were supplement-free through 6 months. All data are topline or preliminary sponsor reports. LUCIA results are pending.