Roche’s age-related macular degeneration (AMD) blockbuster-to-be has failed a phase 3 trial. The failure of lampalizumab to improve AMD-related geographic atrophy (GA) prompted Roche to stop dosing patients until it gets data from a second phase 3 trial.

Lampalizumab was widely tipped to rack up blockbuster sales in a hard-to-treat subset of patients with AMD. But those sales, which Jefferies tipped to top out at $2 billion a year, now look set to slip through Roche’s fingers, leaving it without a source of revenue to mitigate biosimilar attacks on some of its core franchises.

The deterioration of lampalizumab’s prospects follows the publication of top-line data from the first of two phase 3 trials, Spectri. Neither dosing regimen of complement factor D-binding antibody fragment lampalizumab reduced GA lesion area by significantly more than placebo in the one-year, 936-patient trial.

Faced with a “lack of efficacy,” Roche has stopped dosing patients until it has data from a second phase 3 trial. Roche had planned to keep the clinical trial going to generate data against multiple, two-year secondary endpoints.

“While this result is disappointing, we will continue to evaluate results from Spectri to get a clearer understanding of the data as we await the results of our second phase 3 study, Chroma, anticipated in November,” Roche CMO Sandra Horning, M.D., said in a statement.

Expectations for the phase 3 were dialed up by midstage data that suggested lampalizumab could move the needle in a hard-to-treat population. That earlier trial, Mahalo, reported a 20% drop in lesion area progression among patients who received the experimental antibody fragment. The figure jumped to 44% when Roche focused on a subset of patients with complement D variants.

Spectri enrolled a broad population of GA patients, not just those with complement D variants, but the tone of Roche’s statement suggests there is little hope of salvaging the program by focusing in on the genetic subgroup.

Analysts at Jefferies put a brave face on the likely loss of a drug that accounted for about 2% of their midterm valuation forecast for Roche.

“Whilst news of this setback is disappointing, we had considered it to be one of Roche’s higher risk programs. Meanwhile, other new product development and launches, such as Ocrevus, Tecentriq and ACE910 look set to drive continued growth for the company despite approaching biosimilar headwinds,” Jeffrey Holford, Ph.D. wrote in a note to investors.

Holford pointed to upcoming clinical trial readouts that will shape the fates of two of those drugs, Tecentriq and ACE910. PD-L1 checkpoint inhibitor Tecentriq is closing in on data in first-line non-small cell lung cancer, an indication in which competing drugs from AstraZeneca and Bristol-Myers Squibb have floundered over the past year or so.

The second Roche trial on Holford’s radar is Haven 3, which is assessing ACE910 in non-inhibitor hemophilia patients. ACE910, also known as emicizumab, has its own problems. Most notably, the bispecific antibody has been dogged by safety doubts during clinical development and is the subject of a row with Shire.

Holford is optimistic these trials can tee Roche up to cope with declining sales from old products as biosimilar competitors come online. But with the phase 3 lampalizumab trial likely to wipe out one source of mitigating revenue, Roche’s margin for error just got smaller.