Merck said on 24 September 2026 that remigromig met the primary endpoint of the pivotal Phase 2b/3 BRUNELLO study. Merck is known as MSD outside the United States and Canada. It said no new diabetic macular oedema treatment mechanism had matched anti-VEGF therapy in Phase 3 for 20 years. Remigromig was non-inferior to ranibizumab on mean change in best-corrected visual acuity at week 52. Merck disclosed no efficacy figure of any kind. It also reported higher rates of proliferative diabetic retinopathy, vitreous haemorrhage and treatment discontinuations in the remigromig arms.
What BRUNELLO Measured
BRUNELLO is a randomised, double-masked, multi-centre study of 984 participants. Merck describes it as a pivotal Phase 2b/3 trial. The registry record lists it as Phase 2/3. Participants were randomised one to one to one across three arms. Two arms received intravitreal remigromig, at either 0.5 mg or 0.8 mg. The third received ranibizumab 0.5 mg as an active control.
All three arms were dosed every four weeks through the first year. Year two moves to a personalised treatment interval algorithm. The primary endpoint was mean change in best-corrected visual acuity to week 52. Vision was measured in the study eye using standardised ETDRS testing. The registry record names EyeBiotech Ltd as sponsor. It gives a study start of 7 August 2024 and lists the United Kingdom among participating countries. The Health Research Authority also lists a UK research summary for the trial, IRAS 1010164.
Both remigromig doses met non-inferiority against ranibizumab. Merck said the findings will be discussed with regulatory authorities. It named no filing pathway and no submission date. Full results are scheduled for the American Academy of Ophthalmology annual meeting. That presentation is set for New Orleans on 10 October 2026.
The Figures Merck Did Not Publish
The announcement carries a conclusion without the arithmetic behind it. Merck published no mean letter gain for either dose. It gave no between-group difference and no confidence interval. It stated no pre-specified non-inferiority margin. It published no p value. No central subfield thickness result was released, despite that sitting in the registry record as a secondary endpoint.
Merck also published no dosing-interval data from year two. That omission matters commercially, for reasons set out below. LSDN flagged a similar gap on a recent FUS-ALS readout. Such gaps are not unusual on a topline release. This one leaves commissioners and rival sponsors unable to size the effect. It also leaves the diabetic macular oedema treatment market without a comparator benchmark.
The Safety Imbalance Merck Flagged Itself
Merck said both doses were generally well tolerated. It reported the unfavourable findings in the same release. Higher rates of proliferative diabetic retinopathy occurred in the remigromig arms. So did higher rates of vitreous haemorrhage and of discontinuation due to adverse events. The comparison in each case was against ranibizumab. Merck added that further analyses are underway to characterise these findings.
The company gave no rates, no arm-level counts and no timing. It did not state whether the events were judged treatment related. LSDN draws no inference about causation from the release. The point that stands is narrower. A sponsor flagged a retinopathy imbalance in the same release as a win.
An Earlier Dataset Told a Different Story
The first-in-human data read very differently. EyeBio reported Phase 1b/2a AMARONE results on 13 February 2024. Twenty-six patients with diabetic macular oedema gained a mean of 11.2 letters at week 12. Mean retinal thickness fell by 143 microns.
That release described the drug as “well-tolerated”. It reported “no drug-related adverse events, drug-related serious adverse events, or intraocular inflammation”. The two datasets are not comparable. AMARONE ran to 12 weeks in 26 people. BRUNELLO ran to 52 weeks in 984. The pivotal readout adds safety signals the smaller study did not report.
Why the Comparator Choice Matters
Remigromig was tested against ranibizumab 0.5 mg. Ranibizumab is a reasonable clinical comparator and remains in wide use. It is also one of the longest-established options. Several ranibizumab biosimilars are now marketed in England. NHS England places them in its lowest-cost tier, alongside aflibercept 2 mg biosimilars.
The competitive frontier has moved elsewhere. Faricimab and aflibercept 8 mg compete on how rarely a patient needs an injection. The UK summary of product characteristics for Vabysmo allows intervals of up to 16 weeks after loading. In England’s commissioning table, aflibercept 8 mg follows three monthly doses and can extend to 16, then 24 weeks. Diabetic macular oedema treatment intervals, not letters gained, are where those two products differentiate.
Aflibercept 8 mg is instructive on that point. Its pivotal PHOTON study was non-inferior, but not superior, to aflibercept 2 mg on vision. It won its place on injection count rather than letters. Over 60 weeks it needed roughly seven or six injections. The older 2 mg dose needed 9.8. Any new diabetic macular oedema treatment that only matches on vision must therefore say something about burden.
The England Diabetic Macular Oedema Treatment Pathway Runs on Price
NHS England published version 3 of its commissioning recommendations for medical retinal vascular medicines on 11 March 2026. It backs treat-and-extend as standard, using aflibercept 2 mg biosimilar. That approach, NHS England says, “achieves the same patient outcomes at a lower cost”. Ranibizumab biosimilar sits alongside it in that pathway. The document describes the resulting savings as freeing resources for reinvestment. Its modelling found appointments over three years broadly similar across options. Faricimab and aflibercept 8 mg were associated with slightly fewer.
NICE arrives at the same place from a different direction. Its appraisal of faricimab recommends the drug for diabetic macular oedema on one condition about thickness. Central retinal thickness must be 400 micrometres or more at the start of treatment. The recommendation also depends on the company supplying the drug under the agreed commercial arrangement. Where several treatments are suitable, NICE tells clinicians to choose the least expensive. NICE’s 2024 diabetic retinopathy guideline, NG242, separately says to consider anti-VEGF or macular laser below 400 micrometres.
England therefore runs a diabetic macular oedema treatment pathway built on cost minimisation. LSDN reads that as a hard commercial test for any branded entrant. It is not a NICE or NHS England position on remigromig. No UK regulatory application for the molecule has been announced. LSDN found no MHRA authorisation and no NICE appraisal for remigromig as at 24 September 2026.
Capacity Is the Other Constraint
Volume is the pressure English eye services actually feel. The Royal College of Ophthalmologists set out the scale in written evidence to Parliament. It reported more than 7.5 million ophthalmology outpatient appointments in England in 2021/22. Follow-up appointments made up 69 per cent of those. The college also said permanent harm from delay is nine times more likely in follow-up patients.
Intravitreal injection clinics sit inside that follow-up load. A regimen dosed every four weeks through year one adds visits rather than removing them. Diabetic macular oedema treatment capacity is the constraint a service planner has to model. The same monitoring pressure shows up across other diabetes complications. Remigromig’s year two algorithm may change that arithmetic. Merck has published nothing yet that would allow the calculation.
What Merck Has and What It Still Needs
The mechanism claim is the substantive news here. Remigromig is a tetravalent, tri-specific antibody that activates the Wnt pathway. Merck says that pathway is involved in repair and maintenance of the blood-retinal barrier. That is a different target from every approved anti-VEGF agent.
David Guyer, Founder, Chief Executive and President of EyeBio, a wholly owned Merck subsidiary, put the unmet need at scale. He said that “up to 40% of patients with diabetic macular edema do not fully respond”. Those patients, he said, “remain at risk of continued vision loss”. He also called it the first new mechanism in 20 years to reach Phase 3 non-inferiority. Both statements are sponsor claims and LSDN has not independently verified the 40 per cent figure.
Dean Y. Li, President of Merck Research Laboratories, said the company looked forward to advancing remigromig in diabetic macular oedema. Merck acquired the asset with EyeBio, in a deal announced on 29 May 2024. It agreed to pay 1.3 billion US dollars upfront. Up to 1.7 billion more is payable in milestones.
A second pivotal study, BAROLO, is still running in diabetic macular oedema. Merck has not said whether BRUNELLO alone would support a filing. A Phase 2 proof-of-concept study, SUPER TUSCAN, covers neovascular age-related macular degeneration and retinal vein occlusion.
For a trade readership the package is incomplete in a specific way. The primary endpoint was met. The effect size, any interval benefit and the retinopathy findings all remain unquantified. BMO Capital Markets wrote that the update was “not enough to establish a differentiated profile”, BioSpace reported. Until those land, a new diabetic macular oedema treatment on this profile has a clinical story. It does not yet have an access story.














