A year after Roche's Alzheimer's failure, scientists reflect on what could have been

Almost a year to the day after Genentech revealed that its Alzheimer’s disease drug gantenerumab failed a phase 3 trial, the data are appearing in The New England Journal of Medicine, giving scientists a chance to reflect on what was learned.

Since the gantenerumab news, two other Alzheimer’s treatments have advanced: Eisai and Biogen’s lecanemab, now known as Leqembi, was fully approved, and Eli Lilly’s donanemab is soon to have a decision from the FDA.

Lon Schneider, M.D., of the Keck School of Medicine at the University of Southern California, expressed surprise that the Roche unit’s offering isn’t among the class of successful therapies today in a paper published along with the gantenerumab results Wednesday. 

“It is surprising that two phase 3 trials of gantenerumab, the results of which are reported in this issue of the Journal, did not show significant benefits, given that gantenerumab is similar to aducanumab and lecanemab,” Schneider wrote.

All three therapies bind to the same region of amyloid beta that is believed to trigger a neurodegenerative cascade in the brain that causes Alzheimer’s. All three “markedly reduce” the amyloid plaques that are a hallmark of the disease.

Schneider noted that nearly 1,000 people received an infusion of gantenerumab between the two trials called GRADUATE I and II every two weeks over 27 months of study. There was a difference between the treatment and placebo groups on the primary endpoint, a measure of clinical decline called the Clinical Dementia Rating-Sum of Boxes (CDR-SB). The gantenerumab group saw a favorable -0.31 difference in GRADUATE I and a -0.19 change in GRADUATE II, but these findings were not significant. Lower scores indicate smaller cognitive impairment.

“Nevertheless, the CDR-SB results observed in the gantenerumab trials are similar in magnitude to those seen in the aducanumab and lecanemab trials, which showed between-group differences of −0.39, +0.03, and −0.45 with 95% confidence intervals that overlapped with those of the gantenerumab trials,” Schneider wrote.

On secondary outcomes, gantenerumab had a small effect similar to those seen with Lilly and Biogen’s offerings.

“For half these outcomes, the between-group differences would have been significant if the primary analysis had not failed and had not been part of a hierarchical statistical analysis plan,” Schneider said.

When Roche combined the two trials for a secondary analysis, the results favored gantenerumab, which Schneider said “might have been significant.”

“If these identical trials had been designed as one trial, then gantenerumab might have met the same FDA criteria for approval that aducanumab and lecanemab had fulfilled,” Schneider said. “However, caution should be observed in drawing definitive conclusions from the results for the secondary outcomes and from the pooled analysis.”

In other words, had Genentech conducted one big trial instead of two smaller ones, perhaps the results would have landed in gantenerumab’s favor.

Gantenerumab led to a “substantially lowered amyloid load” as detected on PET scans; lowered accumulation of tau, a secondary biomarker in Alzheimer’s; boosted cerebrospinal fluid neurofilament light chain, which is associated with better cognition; and had lower rates of brain swelling than Leqembi and donanemab.

And yet, Biogen and Eisai have an approved therapy, and Lilly is at the FDA’s gates. Genentech has moved on from gantenerumab, cutting a clutch of studies in December. 

Schneider wondered whether there were population differences that may have impacted the results. Both GRADUATE studies featured patients with early Alzheimer’s but had more patients with higher CDR-SB scores and greater amyloid loads at the start of the trials than in the tests for Leqembi and donanemab.

“Although the three antibodies were each associated with approximately the same degree of amyloid plaque removal, the amount of residual plaque was higher in the gantenerumab trials,” Schneider wrote.

In a post hoc analysis, Roche looked at participants who had achieved amyloid-negative status and found they had better clinical scores at the end of the trials than those who did not achieve lowering of that degree. “Those who became amyloid negative had less impairment and lower amyloid loads at baseline than the other participants,” Schneider said.

The researcher also wondered about bias in the GRADUATE trials. Questions have been raised about whether there was adequate masking of the medications being administered. There were also high rates of injection-site reactions, swelling in the brain and bleeding in the brain, which revealed that gantenerumab had been administered and led to suspension of treatment.

Schneider, as well as Alzheimer's Drug Discovery Foundation co-founder and Chief Science Officer Howard Fillit, Ph.D., said the results show early treatment is paramount. Fillit also suggested that combination medicines will be key to future treatment regimens. 

“Scientific research is an iterative process, and it is clear from this trial and others that years of relentless work by Alzheimer’s researchers has brought us into the modern era of research,” Fillit said in a statement. 

So what do Genentech's results mean in the grand scheme of things? Schneider said the gantenerumab studies can confirm whatever side of the amyloid argument you fall on: “Depending on one’s perspective, the results of the antibody trials to date either reinforce confidence in this therapeutic approach and its clinical meaningfulness or support a view that the effects are small, unreliable, and barely distinguishable from no effect. Or, if a meaningful effect is not apparent after 1.5 to 2 years of treatment, there may be hope that it manifests in the future.”