Evicenter
MarkdownMedical informationP&T meetings
Vonaprument (ANX007)

Geographic atrophy secondary to age-related macular degeneration

Launch
Not announced
Last updated
167 sources

Executive summary

Vonaprument efficacy and safety

4 evidence topics · 29 sources

Summary: ARCHER II pivotal phase 3 trial (design and status)
  • Design: ARCHER II is a phase 3, multicenter, randomized, parallel-group, double-masked, sham-controlled trial. It enrolled 659 participants, randomized 2:1 to vonaprument 5.0 mg per eye or a sham procedure, each given monthly for 24 months, with treatment masked through 24 months. The phase 3 dose of 5 mg is administered in 25 microliters; the phase 2 ARCHER trial used 5 mg in 100 microliters. The registry lists 132 locations, and participants were enrolled from North America (55%), Europe (43%), and Australia and New Zealand (2%).
  • Eligibility and baseline: participants were at least 50 years of age, with a GA lesion area of 2.5 to 17.5 mm² and BCVA of 45 to 83 ETDRS letters (20/25 to 20/100 Snellen equivalent). The sponsor states that individuals with baseline BCVA below 45 letters were not enrolled “due to very low incidence of BCVA 15-letter loss in ARCHER”. In pooled baseline data, mean age was 78.4 years, mean BCVA was 65.8 letters, and 54.8% of participants had subfoveal lesions.
  • Endpoints: the two primary endpoints are the proportion of participants with BCVA loss of 15 or more letters from baseline at any two consecutive visits through Month 15 and through Month 24. The Month 24 endpoint was added in August 2026, when all eligible participants had received at least 12 months of masked treatment, with alpha allocated between Month 15 (the majority) and Month 24. Secondary endpoints include time to confirmed BCVA ≥15-letter loss, the proportion with LLVA ≥15-letter loss, change from baseline in the area of total ellipsoid zone loss, and the incidence and severity of ocular and systemic treatment-emergent adverse events and adverse events of special interest. The design announced in 2023 planned approximately 400 participants randomized 1:1, with the primary endpoint assessed through 12 months.
  • Regulatory basis: the sponsor reports alignment with the FDA and the EMA on BCVA ≥15-letter loss as the primary outcome measure, and states that the FDA “has not required Annexon to study the slowing of lesion growth”. The single study will be analyzed for the EU filing and, for the U.S. filing, as two pre-defined sub-studies of about 330 participants each, “in accordance with the FDA’s two-trial recommendation”.
  • Status: as of the sponsor’s third-quarter 2026 results release, ARCHER II has reported no efficacy or safety results. All eligible participants had received at least 15 months of treatment, and retention was 90%. The sponsor expects an independent Data Monitoring Committee to assess the Month 15 primary endpoint in the fourth quarter of 2026, sub-study results in the first quarter of 2027 if the endpoint is met, and study completion with the Month 24 analyses in the third quarter of 2027. The Form 10-Q states that “there is no assurance that either timepoint will show a statistically significant treatment effect”. After Month 24, participants may receive monthly vonaprument in an open-label extension study.
Summary: ARCHER phase 2 efficacy

ARCHER was a randomized, double-masked, sham-controlled phase 2 trial that enrolled 270 participants with GA, randomized to intravitreal vonaprument 5 mg monthly (n=89), 5 mg every other month (n=92), or sham monthly or every other month (pooled n=89) for 12 months, followed by a 6-month off-treatment period. The results are available from the registry record and from sponsor press releases, filings, and congress presentations; an independent systematic review published in 2026 states that “no peer-reviewed publications reporting their clinical results are available to date” and excluded them from its evidence synthesis.

  • Primary endpoint (not met): the rate of change (slope) in GA lesion area by fundus autofluorescence from baseline to month 12 did not reach statistical significance. In the registry analysis, which included non-subfoveal participants only, the slope was 1.9902 mm²/year with monthly dosing (38 participants), 2.1575 mm²/year with every-other-month dosing (43), 2.0837 mm²/year for the pooled vonaprument groups (81), and 1.9019 mm²/year with sham (45). The sponsor reported reductions in lesion growth of 6.2% with monthly dosing (p=0.526), 1.3% with every-other-month dosing (p=0.896), and 3.7% for the pooled groups (p=0.673).
  • Confirmed BCVA ≥15-letter loss (prespecified secondary analysis): the registry reports 4 events among 89 participants with monthly dosing, 8 among 92 with every-other-month dosing, and 15 among 89 with sham, defining confirmed loss as loss “observed for 2 consecutive visits”. The 2023 topline release used a definition that also counted “a loss at the end of the 12-month treatment period” and reported 5 of 89 participants (5.6%), 10 of 92 (10.9%), and 19 of 89 (21.3%), with p=0.0021 for monthly and p=0.055 for every-other-month dosing. The 2026 company presentation reports two definitions: loss confirmed “through month 12 or at last study visit” in 5 of 89 (5.6%), 9 of 92 (9.8%), and 19 of 89 (21.3%), with nominal p=0.0021 and p=0.032; and loss with the “month 12 event confirmed at month 15” in 4 of 89 (4.5%), 8 of 92 (8.7%), and 15 of 89 (16.9%), with nominal p=0.0076 and p=0.0994. The counts under the second definition equal the registry counts. The presentation labels these p-values as nominal, from a chi-square test in the ITT population, and the Form 10-K describes the monthly result as “nominally statistically significant”.
  • Time to confirmed BCVA ≥15-letter loss (Kaplan-Meier analysis, nominal log-rank p-values): the 2026 presentation reports a hazard ratio of 0.272 (CI 0.090 to 0.819; nominal p=0.0119) with monthly dosing and 0.504 (CI 0.214 to 1.189; nominal p=0.1098) with every-other-month dosing. A 2024 presentation reported the same hazard ratios with p=0.0098 and p=0.0788, and the 2023 topline release reported risk reductions of 72% (p=0.006) and 48% (p=0.064).
  • Mean change in BCVA at month 12 (prespecified visual function measure): -1.7 letters with monthly dosing, -2.6 letters with every-other-month dosing, and -4.1 letters with sham, with nominal p=0.1904 and p=0.3977 from a mixed model for repeated measures.
  • Low-luminance visual acuity (change from baseline was a prespecified visual function measure): the registry reports LLVA ≥15-letter loss events in 6 of 79, 8 of 85, and 16 of 79 participants (monthly, every other month, sham). The 2023 topline release reported 8 of 79 (10.1%), 10 of 85 (11.8%), and 16 of 79 (20.3%), with p=0.076 and p=0.14, described as a “non-significant trend”. The 2026 presentation reports 6 of 79 (7.6%), 8 of 85 (9.4%), and 16 of 79 (20.3%), with nominal p=0.0216 and p=0.0497. Mean change in LLVA at month 12 was -1.4, -2.8, and -4.8 letters, with nominal p=0.0592 and p=0.2664.
  • Ellipsoid zone loss (prespecified structural analysis, nominal p-values, Spectralis OCT subpopulation): compared with sham, pooled vonaprument was associated with a 27% decrease in total ellipsoid zone loss across the macula (107 vonaprument and 67 sham participants; nominal p=0.0457), a 48% decrease within the central 2.0 mm (nominal p=0.0218), and a 59% decrease within the central 1.5 mm (nominal p=0.0319).
  • Other loss thresholds (nominal p-values, 2026 presentation): loss of 10 or more letters occurred in 15.7% with monthly dosing, 20.7% with every-other-month dosing, and 25.8% with sham (nominal p=0.096 and p=0.408); loss of 20 or more letters occurred in 4.5%, 6.5%, and 16.9% (nominal p=0.008 and p=0.030).
  • Subgroups by lesion location: the topline combined analysis stratified by foveal and non-foveal lesions reported p=0.0019 for monthly and p=0.0439 for every-other-month dosing. Within subgroups (final data, nominal chi-square p-values), BCVA ≥15-letter loss among participants with foveal lesions occurred in 3 of 51 (5.9%) with monthly dosing (nominal p=.0088) and 8 of 49 (16.3%) with every-other-month dosing (nominal p=.3003), against 11 of 44 (25.0%) with sham; among participants with non-foveal lesions it occurred in 2 of 38 (5.3%; nominal p=.0810) and 1 of 43 (2.3%; nominal p=.0168), against 8 of 45 (17.8%).
  • Subfoveal lesions (post hoc analysis): BCVA ≥15-letter loss occurred in 3 of 51 (5.9%) with monthly dosing (nominal p=.0172) and 7 of 49 (14.3%) with every-other-month dosing (nominal p=.2930), against 10 of 44 (22.7%) with sham; the hazard ratio with monthly dosing was 0.242 (CI 0.067 to 0.881; nominal p=.0204).
  • Baseline low-luminance visual deficit and less advanced disease: the topline combined analysis stratified by baseline LLVD below 30 or 30 and above reported p=0.0016 for monthly and p=0.029 for every-other-month dosing. In additional analyses, the sponsor reports a larger response in eyes with lower LLVD, smaller lesions, and more remaining ellipsoid zone.
  • Off-treatment period: in a preliminary 6-month off-treatment analysis, the rate of BCVA ≥15-letter loss in the treated groups increased to parallel that of sham (“>1.6% per month”).
  • Number needed to treat (sponsor calculation from the prespecified secondary analysis): “1 / (21.3 – 5.6) * 100= 6.4”, reported as 7 for monthly dosing to avoid BCVA ≥15-letter loss, with 10 for loss of 10 or more letters and 8 for loss of 20 or more letters.
Summary: safety
  • ARCHER overall (12-month treatment period, registry): treatment-emergent adverse events occurred in 65 of 89 participants with monthly dosing, 68 of 92 with every-other-month dosing, and 68 of 89 with sham. Serious adverse events occurred in 21 of 89 (23.60%), 16 of 92 (17.39%), and 16 of 89 (17.98%), and all-cause mortality was 2 of 89 (2.25%), 3 of 92 (3.26%), and 2 of 89 (2.25%). Treatment was discontinued by 13 (14.6%), 11 (12.0%), and 10 (11.2%) participants.
  • Endophthalmitis: serious endophthalmitis occurred in 1 of 89 (1.12%) with monthly dosing, 2 of 92 (2.17%) with every-other-month dosing, and 0 of 89 with sham. The sponsor determined the 3 events to be related to the intravitreal injection procedure and “not related to treatment”.
  • Intraocular inflammation and vascular occlusion: intraocular inflammation occurred in 2 participants (2.2%) with monthly dosing, 1 (1.1%) with every-other-month dosing, and none with sham; the topline release described 3 serious events of intraocular inflammation “suspected of being treatment-related”, not associated with retinal vasculitis. The iritis resolved with topical steroids in 2 days and the vitritis in 9 days. One retinal vascular occlusion (1.1%) occurred with every-other-month dosing, described as “Isolated cilioretinal artery occlusion; no vasculitis confirmed by DSMC and reading center”.
  • Retinal vasculitis and ischemic optic neuropathy: no events were reported in any arm.
  • Choroidal neovascularization: 3 participants (3.4%) with sham, 4 (4.5%) with monthly dosing, and 4 (4.3%) with every-other-month dosing.
  • Other events: non-serious conjunctival haemorrhage occurred in 7.87%, 6.52%, and 3.37% of participants (monthly, every other month, sham). Serious atrial fibrillation occurred in 2 of 89 (2.25%), 4 of 92 (4.35%), and 0 of 89, and serious pneumonia in 3 of 89 (3.37%), 2 of 92 (2.17%), and 0 of 89.
  • ARCHER II: no safety results have been reported.
  • Phase 1 glaucoma studies (participants with primary open-angle glaucoma; 5 mg given in 100 μL): in the open-label single-dose phase Ia study (1 mg, 2.5 mg, or 5.0 mg; 9 participants), all 9 participants (100.0%) had a treatment-emergent adverse event, most often conjunctival hyperemia (8; 88.9%); all adverse events were mild and resolved, no dose-limiting toxicities were observed, no anti-drug antibodies were detected, serum unbound drug was not detected, and serum unbound C1q was unchanged. In the double-masked, sham-controlled phase Ib study (2 injections 4 weeks apart; 17 participants treated), no serious, severe, or fatal adverse events were reported, the 6 treatment-emergent adverse events were mild, intraocular pressure returned to 21 mm Hg or lower within 45 minutes in all participants, no systemic pharmacokinetic or pharmacodynamic changes were observed, and low-level anti-drug antibody positivity was detected in 7 participants, including 1 in the sham group.
Summary: comparison with approved complement inhibitors
  • No head-to-head data: no trial has compared vonaprument with pegcetacoplan or avacincaptad pegol. In 2023 the sponsor planned ARROW, a phase 3 trial of approximately 500 participants comparing monthly vonaprument with pegcetacoplan as an injection comparator; in 2025 it stated that it “no longer plans to conduct a second injection-controlled head-to-head Phase 3 trial”. ARCHER II compares vonaprument with a sham procedure.
  • Dosing: pegcetacoplan is labeled at 15 mg (0.1 mL) by intravitreal injection once every 25 to 60 days, and avacincaptad pegol at 2 mg (0.1 mL) once monthly; ARCHER II gives vonaprument 5 mg in 25 microliters monthly.
  • Endpoints: pegcetacoplan and avacincaptad pegol were approved on a reduction in the rate of GA expansion, and a 2024 review states that functional data from two phase 3 trials for each drug “demonstrated no visual benefit to patients in the treatment groups”. ARCHER II uses BCVA ≥15-letter loss as its primary endpoint, and in phase 2 vonaprument did not significantly reduce lesion growth. The sponsor states that “Protection of RPE has not resulted in preservation of vision” with C3 and C5 inhibitors, and that the two approved products “may pose competition for vonaprument, if approved”.
  • Independent appraisal: a 2026 systematic review excluded C1q inhibition from its evidence synthesis because its results “are not yet supported by peer-reviewed publications”.