Insights from Eli Lilly’s TRAILBLAZER-ALZ Program

John Sims

Dawn Brooks

In this interview, John R. Sims, senior vice president and head of global development at Eli Lilly and Company, speaks about the design of the TRAILBLAZER-ALZ program, which is aimed at treating Alzheimer's disease, and the challenges faced during the clinical trials. Sims discusses the two key design elements that contributed to the success of the program, including patient population enrichment and finding a better measurement of disease progression. He also talks about the iADRS scale that was used to assess patient progression and how it was selected. Later, Dawn A. Brooks, director of clinical development at Eli Lilly and Company, talks about the TRAILBLAZER-4 study, which was an active comparator trial of donanemab versus aducanumab, and the top-line results.

Moe Alsumidaie: I’d like to understand the challenges that you went through with your clinical trial. Before we speak about the TRAILBLAZER-ALZ 4 head-to-head study, perhaps you can first start by telling us about the overall design and the goal of the TRAILBLAZER-ALZ program. We all know that Alzheimer’s has been a tough indication for anyone to get approval and you guys have made a breakthrough there. What was so special about the trial’s design that got this to work?

John R. Sims: All the failures that we learned from in our previous trials really informed the final design of TRAILBLAZER-ALZ. One of the first things we realized was that the clinical diagnosis of Alzheimer’s disease was likely not sufficient alone in terms of trial inclusion criteria. Those early trials showed us that 20% - 30% of patients didn’t have the pathology that our drug intended to treat—that was a key learning from an amyloid perspective.
During the amyloid-focused trials, we started studying tau as well. We learned in those studies that not all amyloid patients were “equal”—some people who had a lot of tau were very fast progressing, and those who had very little to low tau were very slow progressing. Since our goal was to design drugs that slow the progression of the disease, the placebo arm of the trial had to include some kind of decline, or you can't separate treatment with placebo versus treatment with the drug. So that was the key design aspect—to enrich TRAILBLAZER-ALZ with a population whose disease progression was not too slow nor too fast, and then we could run a smaller trial to get a signal (which was almost impossible in the past).

So those are probably the two key design elements, and then of course dose was a key issue, so we had to run long Phase I trials to understand the dose and how to optimize that aspect. The other key learning was to find a better measurement of disease progression, so that's where our primary outcome came from. The learnings from the semagacestat trials, the solanezumab trials, and even some of our BACE (an enzyme that aids in forming amyloid beta) trials—all of those helped inform the use of the iADRS as our primary endpoint and increased our confidence that we thought it was the best measure to follow patients consistently over time and understand how their disease is progressing.

MA: Can you tell me about the scale, iADRS, that you used to assess patient progression? How did you figure out that was the best scale to use?

JS: It was derived from two well-validated scales that had been used in Alzheimer’s disease (AD) trials, the ADAS-COG13 and ADCS-iADL, with the goal of adequately and meaningfully assessing the impact of cognitive loss on the ability to perform daily activities in early AD. The iADRS addresses concepts identified as most important by individuals with or at risk for AD and their care partners, directly mapping to more concepts than any other measure assessed within the What Matters Most (WMM) research initiative. To date, the iADRS has demonstrated reliable, stable and consistent detection of treatment effects in an early symptomatic AD population.

MA: Turning science into reality here, let’s turn now to the TRAILBLAZER-4 study, which was an active comparator trial of donanemab versus aducanumab. Could you tell me a bit about the top line results, which compared each therapy’s ability to lower amyloid in early symptomatic Alzheimer’s disease. Was there a statistical association between the lowering of amyloid and an improvement in patient cognition based on the tests that you used? Can you tell us a bit more about those results?

Dawn A. Brooks: At the Clinical Trials on Alzheimer's Disease Congress (CTAD), we reported the six-month primary endpoint results from the first active comparator trial with amyloid targeting agents—donanemab compared to aducanumab. It measured amyloid removal, or amyloid reduction, but did not use a cognitive endpoint scale as we had in the Phase II TRAILBLAZER-ALZ study, and as you'll see in our Phase III study with donanemab, TRAILBLAZER-ALZ 2.

This trial was open label—so, people knew what they were assigned to and aducanumab patients were dosed according to the aducanumab label while donanemab patients were dosed until evidence of plaque clearance (defined as <24.1 CL) via positron emission tomography (PET) at 6, 12, or 18 months. The point of the study was to continue learning more on the biomarker front and to offer a chance to show donanemab’s unique speed and depth of amyloid plaque removal, even at the six-month point. We were able to show that 38% of donanemab-treated participants achieved brain amyloid clearance, even at that six-month point,which we define in the field as 24 centiloids. Relatively few have achieved that. It was less than 2% for those taking aducanumab.

I think it was our first chance in the field to see some of that biomarker movement, and we augmented that with other biomarkers to support this cascade—like P-tau, for example. We'll continue to present additional data on biomarkers at future congresses. It also allowed us to explore the safety of the drugs, and the data showed that donanemab maintained the same safety profile as what we saw in Phase II.

So, ultimately, this trial has allowed us to advance our understanding of biomarkers and confirm the amyloid reduction that donanemab showed in Phase II. While this was the six-month primary endpoint report, we'll have 12- and 18-month data collections as well.

MA: Do you anticipate that there will be some resistance in adopting donanemab with these results, and, if so, how are you prepared to address these concerns that will come up from the public?

DB: The TRAILBLAZER-ALZ 4 data, for example, is part of our safety database. The number of people who have taken donanemab as part of our clinical trials continues to increase, and I think that's important for public and prescriber confidence—this database of individuals with experience on the drug. We’re also finishing our Phase III trial. If that data is reported positive in the second quarter of 2023, that will be the basis for our traditional FDA approval. I think that will be the point, perhaps with that additional confirmatory data, that we would expect the Centers for Medicare & Medicaid Services (CMS) to reconsider its current policy and, hopefully, provide access. And then together with all that data, I think that provides the confidence that the public and prescribers need.

MA: A lot of repair needs to be done from previous events, unfortunately.

DB: I think FDA is to be commended—they're trying to help drive innovation. We demonstrated amyloid plaque reduction in our Phase II trial, and the lecanemab data that was reported for their Phase III clinical trial is good news for the field because it reinforces the association between amyloid plaque reduction and slowing of clinical decline.

Even gantenerumab, with the negative Phase III results, also reinforces that association because they were able to learn that those with plaque clearance did see better clinical benefits. So, the pathway that FDA used for aducanumab was potentially the right thing to do. But that's where we are today, and we’ve all learned so much since then.

MA: Going back to John, why do you feel optimistic about the Phase III data for donanemab?

JS: Certainly, science is about replication, and confidence increases with replication. I think we've seen a couple of really encouraging results. Lecanemab’s data, of course, is very encouraging for the entire field. As Eisai did, we’ve also designed our Phase III trial to confirm our Phase II results. We also planned the TRAILBLAZER-ALZ 4 study comparing aducanumab and donanemab, anticipating the biomarker results and plaque removal we would see with donanemab, and that met our expectations.

I've been studying Alzheimer’s disease for a long time and can say that rarely anything has turned out as expected. We are now really starting to see things turn out like we planned.

We've removed a lot of the design issues that we were facing in the past because we were testing both whether a molecule worked and whether our trial design worked. Sometimes our trial design was the problem, and a lot of times the molecules were the problem because we weren't dosing them high enough.

Now, we better understand both. Once you get a trial design that works, that gives you the crack in the door to open it wider, and you're less worried about trial design and more focused on how to better optimize the molecules. Even though the gantenerumab data was negative, I believe it still reinforced that you ultimately must remove enough amyloid to see a clinical benefit. So, I think we saw some really encouraging results even in the negative data there.

MA: That's where the precision of science comes to the imprecision of human behavior, and trial design is kind of human behavior when it comes to how you do it.

JS: It is very challenging, and people leave that out of this discussion about whether amyloid is the right target, but there are multiple moving gears to solve for that, and I think we're getting a much better idea of how Alzheimer’s disease works.

MA: Last question, what's coming in the pipeline that Lilly’s looking forward to now that your trial designs are starting to become more streamlined and are generating the results that you're looking for?

JS: We're certainly excited about our investigational asset remternetug. It's a potent molecule as well, so we'll continue to hopefully advance that development program.

We have a Phase II trial that's ongoing with therapies that target tau. We certainly look forward to bringing other agents toward the other pathologies that we know exist in AD, with tau, of course, being a very large one there. As an example, we look forward to having our OGA inhibitor results coming out toward the end of next year. We'll continue to study some of the other aspects that we know are playing a role in inflammation, another key biology that needs more work, and continue to throw all our resources at all the pathologies until we really slow this disease and maybe one day, hopefully, reverse it.

MA: Is there anything else either of you would like to say?

DB: I’d like to end by saying that we are excited about our AD portfolio. Lilly has a long-standing commitment to fighting this disease, and we are thrilled to be able to share these scientific advancements, to have the support of the company and to have the great researchers and medical experts, like John, working hard every day to fight this disease and ultimately bring solutions to people and their loved ones who are devastated by this disease.

Moe Alsumidaie, MBA, MSF, is a thought leader and expert in the application of business analytics toward clinical trials, and regular contributor to Applied Clinical Trials.