Injected drug beat laser in diabetic retinopathy, with an eye-inflammation signal

An injectable eye drug, brolucizumab, was more effective than the decades-old standard of laser treatment for proliferative diabetic retinopathy in a large phase 3 trial — but it came with a clear excess of intraocular inflammation, the very complication that has dogged this drug since its launch [s1]. The result, from the company-sponsored CONDOR trial, sharpens a choice rather than settling it: better disease control on one side, a known and potentially sight-threatening safety signal on the other [s1].

Proliferative diabetic retinopathy (PDR) is the advanced stage of diabetic eye disease, in which chronically high blood sugar damages retinal vessels and the retina grows fragile new ones that can bleed or detach. For decades the treatment has been panretinal photocoagulation (PRP) — laser burns across the peripheral retina that halt the abnormal growth, at the cost of some peripheral and night vision. Injections of drugs that block vascular endothelial growth factor (VEGF), the signal driving the new vessels, are the newer approach, and CONDOR tested one of them head to head against laser.

What the trial did

CONDOR was a 96-week, two-arm, single-masked, multicentre phase 3 trial run across 152 sites in 16 countries and sponsored by Novartis, which markets brolucizumab [s1][s2]. It randomised 689 patients with PDR and no previous laser treatment 1:1 to brolucizumab 6 mg (347 patients) or PRP (342), and 572 completed the pre-specified week-54 assessment [s1]. The brolucizumab group received three loading doses six weeks apart, then injections every 12 weeks, with the option from week 48 to stretch the interval up to 24 weeks if the disease was quiet; the laser group had one to four sessions and further laser as needed [s1]. The primary outcome was the change in best-corrected visual acuity — the standard eye-chart letter score — at week 54 [s1].

One design feature is worth flagging: the trial was single-masked, meaning patients and treating staff could tell laser from injections. That is hard to avoid when comparing a procedure with a drug, but it can nudge subjective assessments and follow-up.

What it found

On the primary measure, brolucizumab was both non-inferior and superior to laser. Vision was essentially unchanged in the drug group (a mean gain of 0.2 letters) while it fell in the laser group (−4.2 letters), a difference of 4.4 letters (95% confidence interval 2.4 to 6.4; P<.001) [s1]. The gap in disease control was larger: 63.6% of brolucizumab-treated eyes (187 patients) had no proliferative retinopathy at week 54, versus 22.4% (65 patients) after laser — a 39.4-percentage- point difference [s1].

The safety picture is where caution enters. Overall ocular adverse events were actually less common with the drug (34.3%, 119 patients) than with laser (49.1%, 168 patients) [s1]. But intraocular inflammation, including retinal vasculitis — an inflammatory reaction that can damage retinal vessels and threaten vision — occurred in 5.2% of the brolucizumab group (18 of 347) against 0.6% of the laser group (2 of 342) [s1]. That roughly ninefold difference is consistent with the inflammation risk that has defined brolucizumab's safety profile in other retinal diseases.

How to read it

CONDOR makes a genuine efficacy case: over the first year, regular brolucizumab injections preserved vision and cleared abnormal new vessels more reliably than laser, in patients who had not been treated before [s1]. For a disease where laser trades long-term stability for a permanent dent in peripheral vision, a drug that spares the retina that damage is attractive — provided the eye tolerates it.

The inflammation signal is the counterweight, and it is not a footnote. A 5.2% rate of intraocular inflammation is the kind of number that shapes real-world prescribing, because the worst cases can cause irreversible vision loss and because injections must continue indefinitely, each carrying that risk [s1]. The trial also reflects only 54 weeks of a 96-week study, was industry-funded, and could not be fully masked [s1]. A reader should treat this as evidence that brolucizumab works well in PDR and that its use is a benefit-versus-risk judgement made eye by eye — not a blanket verdict that injections should replace laser.

What to watch

The full 96-week results will show whether the vision advantage and the inflammation risk both hold with longer treatment, and how many patients managed to extend to the longest dosing intervals. Comparisons with other anti-VEGF drugs that carry less inflammation risk will matter for where brolucizumab lands in practice.

This article describes research and is not medical advice. Treatment of diabetic retinopathy is a decision for treating clinicians.

Sources

• Brolucizumab in the Treatment of Proliferative Diabetic Retinopathy: The CONDOR Randomized Clinical Trial — JAMA Ophthalmology, 23 April 2026
• CONDOR trial record (NCT04278417) — ClinicalTrials.gov