Roche eyes diabetes success where Regeneron failed

At Roche’s fourth-quarter financial conference the Swiss company highlighted RG7716 as one of two mid-stage pipeline projects to look out for. Yesterday investors found out why, when in the Boulevard trial in diabetic macular oedema the anti-VEGF/Ang2 bispecific became the first project to beat VEGF inhibition alone.

This achievement appears to be one in the eye for Regeneron, which had failed with the same mechanistic approach last November. But this is largely a battle to extend Regeneron and Roche’s respective Eylea and Lucentis franchises, and in Roche’s Boulevard study RG7716 has not beaten Eylea.

It is also unclear why two near-identical mechanistic approaches should have had opposite results. Regeneron’s failure involved REGN910-3, a combination of two antibodies, one targeting VEGF and the other Ang2; RG7716 is what Roche calls the first bispecific MAb to hit both of these targets in the ophthalmology setting.

There were also design differences: Regeneron’s Ruby trial had sought to beat Eylea in terms of best corrected visual acuity (BCVA) between weeks 12 and 36, while Roche’s Boulevard study measured BCVA from week zero to 24, versus Lucentis.

Success?

Be that as it may, Roche yesterday celebrated RG7716’s success in increasing 24-week BCVA by 13.9 chart letters, versus 10.3 letters for Lucentis (p=0.03). A lower RG7716 dose gave an 11.7-letter improvement, allowing Roche to claim a “clear, dose-dependent” response.

Sascha Fauser, Roche’s head of ophthalmology, told an investor call today that a bispecific approach could yield a more predictable response than two antibodies, which might accumulate and be cleared at different rates. He said to improve outcomes additional pathways had to be found beyond VEGF, which had “hit the ceiling”.

Jason Ehrlich, Roche’s head of ophthalmology product development, called the Boulevard data “clinically meaningful”, saying a 35% improvement over Lucentis was something many patients would be happy with. But Leerink analysts, who cover Regeneron, were sceptical. “At best we believe the signal for RG7716 is modest,” they wrote, adding that the small, 3.6-letter difference versus Lucentis “could easily disappear in phase III”.

Moreover, superiority to Lucentis does not necessarily imply that RG7716 is better than Eylea, which currently generates almost twice as much revenue as the Roche blockbuster. For the killer blow, Leerink said RG7716 was administered via intraocular injection every four weeks; Regeneron has just filed a 12-week version of Eylea for US approval.

Asked about comparing RG7716 directly versus Eylea in phase III Mr Ehrlich told the investor call that all options were on the table. Regeneron’s own franchise-extension strategy now hinges on the Panorama study of Eylea monotherapy in diabetic retinopathy, whose readout is due by the mid-year.

A threat to Eylea?
Global sales ($m)
Product Company Mechanism 2016 2022e
Eylea Regeneron/Bayer VEGFr kinase inhibitor 5,539 6,967
Lucentis Roche/Novartis Anti-VEGF MAb 3,262 2,175
RG7716 Roche Anti-VEGF/Ang2 bispecific 54
REGN910-3 Regeneron Eylea + anti-Ang2 MAb (no consensus)
Source: EvaluatePharma.

The Roche and Regeneron follow-on projects have had an intriguing parallel existence as franchise-extending strategies.

Both were studied in wet age-related macular degeneration (AMD) as well as in diabetic macular oedema, a condition in which fluids build up on the macula, and which can lead to diabetic maculopathy and retinopathy. REGN910-3’s Onyx study in wet AMD also failed, while RG7716’s Avenue trial in this condition has yet to read out, though it should have been completed last September.

The mechanistic rationale builds on the involvement in eye disease of angiogenesis – blood vessel growth – which brought about drugs that hit VEGF, like Eylea and Lucentis. The second piece of the puzzle is Ang2, a ligand that blocks the Tie2 receptor on endothelial cells.

Tie2 activity is thought to be essential for maintaining vascular stability, a function assisted by blocking Ang2. Aerpio Pharmaceuticals is separately developing AKB-9778, a small-molecule that activates the Tie2 pathway, and has shown some early promise.

While Roche states that RG7716 is the first anti-VEGF/Ang2 bispecific specifically designed for the eye, it is not alone in pursuing these targets: Boehringer Ingelheim and Ablynx are developing the anti-VEGF/Ang2 bispecific BI 836880 for oncology use. By analogy, the active ingredient in Lucentis is based on the cancer drug Avastin.

Off-label use of Avastin has already cut into Lucentis, making Roche’s need for a fresh start in ophthalmology particularly pressing. Results of Avenue, and RG7716’s move into phase III, just became even more important.

Study Trial ID Indication Company Project Result
Boulevard NCT02699450 Diabetic macular oedema Roche RG7716 Met primary endpoint vs Lucentis
Avenue NCT02484690 Wet AMD Roche RG7716 Patients in follow-up
Ruby NCT02712008 Diabetic macular oedema Regeneron REGN910-3 Failed vs Eylea
Onyx NCT02713204 Wet AMD Regeneron REGN910-3 Failed vs Eylea

To contact the writer of this story email Jacob Plieth in London at jacobp@epvantage.com or follow @JacobPlieth on Twitter

Share This Article