Powering Alzheimer Drug Development With Biomarkers and Composite Instruments

Jeffrey Cummings, MD, ScD

This is a 2-part interview. For part 1, click here.

With just 5 drugs—all for symptomatic treatment only—achieving FDA approval for Alzheimer disease (AD) since 1995, it begs the question how effectively private funding is being used in the pipeline. To quantify the financial aspects of AD drug development in the private sector and propose ways to reduce research and development costs, Jeffrey Cummings, MD, ScD, conducted a large review of 1099 AD clinical trials conducted in the past quarter century.

The major findings showed that an estimated $42.5 billion in private expenditures had been used during this period, with little regulatory success to show for it and several failures coming in later phases of development. Cummings et al’s analysis included the recent June 2021 approval of aducanumab (Aduhelm; Biogen), which broke finally broke the dry spell of approvals for disease-modifying agents since 2003. While drug development has hit numerous walls, Cummings believes there is a great deal of momentum within the space, mainly fueled by aducanumab and its ability to show change in disease progression using clinical biomarkers and composite instruments.

Cummings, professor of brain science, and director, Chambers-Grundy Center for Transformative Neuroscience, University of Nevada–Las Vegas, sat down to provide insight on how the clinician community can learn from successes such as aducanumab and whether a greater consensus on efficacy measures would increase the chances of clinical success. He also gave his “call to action” for clinicians within the community on what they should do to eliminate time and waste from AD clinical trials.

NeurologyLive®: Would a greater standardization of efficacy measures improve the likelihood of clinical success?

Jeffrey Cummings, MD, ScD: To a certain degree that’s a reasonable position to take. I believe that the biomarkers are going to give us far more information and clinical data. This is such a slow disease that even the most sensitive instrument is going to have a challenge in terms of establishing a drug–placebo difference. It’s not like epilepsy, where you have 10 seizures a month and then you go to 5 seizures a month and things look pretty good at the end.

Here, we’re looking at a disease which changes by 3 Mini-Mental State [Examination] points per year. That’s important cognition for the individual, but it’s difficult to show a drug–placebo difference. I think we’re moving in the right direction of better instruments, such as these composite instruments, which appear to perform better than the individual instruments. ADCOMs (Alzheimer Disease Composite Score) is a combination of the ADAS-Cog (Alzheimer’s Disease Assessment Scale cognitive subscale), the CDR (Clinical Dementia Rating), and the Mini-Mental, as well as the ADRS (Alzheimer Disease Rating Scale), which is a combination of the ADL (Activities of Daily Living) scale and the ADAS-Cog. Both of those seem to be performing better than the individual instruments do that make up the composite. These composites represent one way forward. Of course, we’re not sure how the FDA is going to view the composites, and that’s a discussion that has to go on. But I think we’re making progress in measurement, and its mainly around novel composites.

What are ways the clinician community can learn from drug successes, including aducanumab?

The FDA opened a new pathway­—accelerated approval—based on a biomarker that was thought to be reasonably likely to predict clinical benefit. That was a great lesson. It showed that there is some competence in amyloid lowering in having predictive value, and we’re seeing that now. In donanemab study, we saw that lowering aducanumab was associated with slowing of integrated ADRS. With lecanemab, we saw that amyloid lowering was associated with a slowing on the ADAS-Cog and the ADCOMS, one of the novel composites.

I think the FDA correctly identified a pathway that was appropriate for the approval of disease-modifying agents. The FDA had an enormous history of doing that with cancer drugs. They were familiar with it, but the neurology community was unfamiliar with it and somewhat skeptical about it. But as more data comes in, the more convincing it becomes.

Biomarkers can be helpful in informing us in predicting clinical benefit and will be used by the FDA then for the approval pathway. I would love to see that happen with a small molecule. Having an infusion every month and having to monitor for ARIA create great demands on the healthcare system. We want to be able to have fewer demands on the healthcare system and therefore also to distribute the drugs more broadly and globally. We’re just not going to be able to reach a global population with an infusion drug that requires frequent MRI monitoring. It’s just not going to happen in much of the underdeveloped world. But we don’t want our breakthroughs to be limited only to Western populations. We still need to invest in small molecules, and I think the monoclonals may well be informative in terms of what the targets are and how to monitor the success of the drug.

Do you have an overlying call to action to improve drug development and eliminate waste?

I have 2 responses to that. First, we need to be more aggressive in stopping drugs during phase 2. That would, in short, save an enormous number of resources that could then be reinvested in other promising drugs. That’s an important lesson.

The other is that biomarkers are being very informative, and the progress in the plasma biomarkers is going to be even more helpful. If we don’t have to do imaging, don’t have to do lumbar punctures, and can measure the progression of the disease and the effect of the drug through plasma biomarkers, we will be much better positioned in terms of being able to accelerate drug development. I see a big emphasis on biomarkers as being one of the lessons that comes out of this study, and particularly promising are the plasma biomarkers.

Transcript edited for clarity.