Lilly's hotly anticipated Alzheimer's data set a new high score in slowing cognitive decline

Eli Lilly’s Alzheimer’s disease candidate has achieved a 35% slowing of cognitive decline and a 39% lower risk of advancing to the next stage of the disease when compared to placebo.

The data represent a mic drop moment for the Alzheimer’s community, which has long been desperate for options, and a drug research industry that has dealt with decades of high-profile, devastating failures.

“Just being able to reproduce the data would have been a thrill itself. Indeed it has slightly overperformed our expectations, I would say,” John Sims, M.D., head of medical, Alzheimer’s disease development at Lilly, told Fierce Biotech with a slight chuckle.

The phase 3 TRAILBLAZER-ALZ 2 data come seven and a half months after Eisai and Biogen showed that lecanemab, now approved under the accelerated pathway as Leqembi, slowed cognitive decline by 27% when compared to placebo. There are key differences between the studies, and Sims carefully warned against making direct comparisons—but Lilly says donanemab has set a new standard.

“We’ve had a very strong showing with donanemab here,” said Dawn Brooks, head of global brand development, neuroscience, at Lilly.

But there is one big caveat: No matter how you slice the data, donanemab had a higher rate of a known side effect of the amyloid-lowering monoclonal antibody class called ARIA than was reported with Leqembi. This is a key safety flag that has had Eisai on its back foot defending the efficacy results since releasing the groundbreaking data in September 2022.

Lilly's shares rose 4.5% to $422.80 in premarket trading Wednesday morning, compared to $404.20 at close the prior day. 

The data

Donanemab was compared to placebo in the late-stage study that featured 1,182 patients in the primary analysis population. Participants in this group had intermediate levels of tau and clinical symptoms of Alzheimer’s, meaning they were at a more advanced disease state. For comparison, Eisai’s study involved patients with mild cognitive impairment due to Alzheimer's and mild Alzheimer’s, also known as early Alzheimer’s.

The primary endpoint was based on the integrated Alzheimer's Disease Rating Scale, or iADRS, which combines two widely used measures of Alzheimer’s disease to detect disease progression and treatment effect. The scale considers activities of daily living such as driving, hobbies and other things key to quality of life.

The main population had a 35% slowing of cognitive decline. Donanemab also led to a 36% slowing of decline over 18 months on a key secondary endpoint measured by the Clinical Dementia Rating-Sum of Boxes, or CDR-SB, which measures disease severity. That is the endpoint that Eisai used for its phase 3 confirmatory study, which showed the 27% rate.

Other secondary analysis of the donanemab study showed that 47% of patients had no decline on CDR-SB at one year compared to 29% on placebo. More than half of patients, 52%, completed treatment by one year and 72% by 18 months after achieving plaque clearance. The participants also had a 40% slowing of decline in ability to perform activities of daily living at 18 months, and, finally, they had a 39% lower risk of progressing to the next stage of disease when compared to placebo.

All of these secondary endpoints were an effort to establish donanemab’s clinical benefit. Biogen and Eisai previously received accelerated approval from the FDA for Leqembi’s ill-fated predecessor Aduhelm based on data showing that it reduced amyloid plaques, not that it actually improved cognition or halted decline. The decision was highly controversial. Lilly, on the other hand, wanted to be able to underscore the clinical meaningfulness of the efficacy data and throw a few exclamation points in for good measure.

“I think the important thing that we've tried to do also is capture the clinical meaningfulness because I know there's been a big debate in the field about what is clinically meaningful,” Sims said. What Lilly ultimately did was new to the field. He specifically points to the nearly half of patients that showed no decline on the CDR-SB. “We thought that was an important way to characterize clinical meaningfulness because no one understands the scale point changes.”

Another is that donanemab lowered the risk of progression to the next disease stage, which can mean a huge drop in quality of life for patients.

“It means going from like mild cognitive impairment to dementia. Or going from mild dementia to moderate dementia. So those are really big step changes,” Sims said. “I think a powerful message that we do want to get across [about] these medicines is they have some risk but they have really powerful clinically meaningful effects for patients.”


Which brings us to the safety data. Amyloid-related imaging abnormalities, or ARIA, has been seen with the class of amyloid-clearing monoclonal antibodies like donanemab, Leqembi and Aduhelm. They are indicators of brain swelling (ARIA-E), small brain bleeds (ARIA-H) or other complications seen on diagnostic imaging. Typically, ARIA is asymptomatic, but in rare cases can cause serious or life-threatening events.

Two deaths due to ARIA were reported in the TRAILBLAZER-ALZ 2 study and a third patient died after a serious ARIA, according to Lilly’s release. ARIA-E occurred in 24% of the treated patients in the donanemab group, with 6.1% reported as symptomatic. For ARIA-H, the rate was 31.4% in the donanemab group and 13.6% in the placebo arm.

Lilly said that the majority of the ARIA cases were mild to moderate and resolved or stabilized with appropriate management. The rate of serious ARIA, including the noted deaths, was 1.6%.

“We are encouraged by the potential clinical benefits that donanemab may provide, although, like many effective treatments for debilitating and fatal diseases, there are associated risks that may be serious and life-threatening,” said Mark Mintun, M.D., group vice president of neuroscience research and development at Lilly, in a statement.

Lilly, as well as Eisai, has been underscoring that this is a risk patients will have to accept for treatment with amyloid-reducing therapies. They will have to make the decision in close consultation with their care provider and have the risks clearly explained to them.

“If a person doesn't have amyloid, they can't have ARIA,” Sims said. “The risks are there. They're common to this class of medications, and they're an area that will need education for appropriate use and to minimize that risk.”

But the risk should be balanced with efficacy—and Lilly hopes that today’s data can tip the scale toward treatment. While Sims cautioned against making direct comparisons of the ARIA rates, patients and their doctors will need to consider both options before they decide which treatment to go on.

“What's important is that this is a life-threatening disease itself,” he said. “Most people realize that they are on a one-way street, that this is an aggressively progressive disease and the sixth leading cause of death in the U.S.”

The differences

Another way that donanemab is differing from Leqembi is on dosing. Lilly dosed patients until their scans revealed that the amyloid plaques had cleared. This means that patients can stop dosing. Amyloid accumulates at about three to four centiloids a year, Brooks explained. The baseline accumulation for patients in the trial was 100 centiloids, and clearance is defined as 24 centiloids.

“If someone has been cleared of amyloid, it may be three or four years until they might be positive again,” Brooks said. Plaques typically build up for 10 to 20 years before causing cognitive issues. “Over the course of this trial, we're removing a decade or more worth of deposited plaque and slowing the progression of the disease.”

Lilly does not yet have the long-term data to definitively say when or how re-dosing might happen. The company is conducting extension trials now to find out.

With the data in hand, Lilly plans to race to regulators to secure the coveted full approval for donanemab. The company missed out on an accelerated nod earlier this year after the FDA asked for more long-term exposure data. At the time, Lilly said that long-term data weren’t available because of the short-term dosing that was a feature, not a bug, of the TRAILBLAZER-ALZ 2 trial. Brooks hopes to see the approval arrive in the first half of 2024.

“In light of confirming our phase 2 results and seeing such very strong p-values here, I think our results—we hope—will be looked at favorably by the regulators,” Brooks said.

And patients, too. Sims, reflecting on the moment this represents for the Alzheimer’s community, said it’s been a long road. “People have been desperate for some kind of therapeutic hope, and I think this is a big deal for everyone,” he said.

Not only will patients have a second option if donanemab is approved, but Sims said that the scientific community now has a beacon to point to for further research.

“It opens the door for new avenues to be studied. These therapies will let us understand the disease even more and come up with even better therapies in the future,” he said.