Alzheimer’s Drug Trials Should Report Effect Sizes, Researchers Say

Derick Alison
Derick Alison
6 Min Read

Standard effect sizes are more meaningful outcomes in Alzheimer’s disease drug trials than the widely used metric of percent slowing of decline, an analysis suggested.

In recent trials of Alzheimer’s anti-amyloid agents, effect sizes (Cohen’s d) for Clinical Dementia Rating-Sum of Boxes (CDR-SB) scores were below 0.24, while percent slowing of decline ranged from 22% to 29%, reported Terry Goldberg, PhD, of Columbia University Irving Medical Center in New York City, and co-authors.

Effect size is potentially independent of percent slowing of decline, Goldberg and colleagues wrote in the Journal of Neurology, Neurosurgery and Psychiatry.

“We showed that in the recent studies of anti-amyloid immunotherapies, Cohen’s d effect sizes for the CDR-SB were small, despite seemingly impressive percent slowing-of-decline metrics. These had not been previously examined in these studies,” Goldberg told MedPage Today.

“We then showed that number needed to treat [NNT], another measure of efficacy and derived from effect size, ranged from 14 to 18,” he added.

“The very recent gantenerumab phase III trials are in line with these results,” Goldberg said. “Taken together, these provide important information for clinicians as they explain the effects of these drugs to patients.”

Alzheimer’s clinical trials often report that a drug slowed decline by x% at an 18-month endpoint compared with placebo, the researchers noted. “Such a comparison may be attractive to clinicians and investors because it implies a slower disease progression rate for the active treatment compared with placebo.”

But the magnitude of a drug’s effect cannot be determined from this kind of reporting, they pointed out.

“For example, a 20% less decline at the endpoint could mean scores of 40 versus 50 for placebo with an absolute difference of 10, or mean scores of 8 versus 10, with a difference of 2,” they wrote. “Relative change cannot consider the magnitude or variance of the outcome measures as would a widely established statistic (e.g., t-test, F test, beta coefficient) or a standardized effect size.”

The researchers computed effect sizes and NNTs for three anti-amyloid drug trials: the aducanumab (Aduhelm) EMERGE (high-dose) trial, the lecanemab (Leqembi) CLARITY AD trial, and the TRAILBLAZER-ALZ 2 trial of investigational donanemab. All three phase III trials included patients with early Alzheimer’s disease and spanned 18 months.

Percent slowing on the CDR-SB was reported as 22% for aducanumab, 27% for lecanemab, and 29% for donanemab. Effect size was a critical determinant of NNT, “such that NNT was uniformly smaller when d was larger,” Goldberg and co-authors noted.

The Cohen’s d for CDR-SB was 0.16 for the aducanumab trial, 0.21 for the lecanemab trial, and 0.23 for the donanemab trial. The NNT was 18 for aducanumab, 15 for lecanemab, and 14 for donanemab.

“A d of 0.15-0.19 can only be described as small, but it is also important to understand the measure from which this was derived,” noted Tomas Kalincik, MD, PhD, and Amy Brodtmann, MBBS, PhD, both of the University of Melbourne in Australia, in an accompanying editorial.

“The Clinical Dementia Rating Sum of Boxes is a standard metric for dementia trials: a cognitive and functional impairment scale scored via interview with participant and informant,” they explained. “Many trialists acknowledge its inherent flaws — ordinal nature and susceptibility to informant and rater bias.”

Results of clinical studies rarely can be condensed into a single value, Kalincik and Brodtmann observed. “It is substantially more informative to consider differences between compared interventions from multiple angles,” they suggested. “Clinically, both absolute and relative effects help neurologists understand the importance of findings in the context of a minimum clinically meaningful difference and the natural course of Alzheimer’s disease.”

While their analysis focused only on efficacy data, Goldberg and colleagues said “the risk of adverse effects such as ARIA [amyloid-related imaging abnormalities] and cerebral volumetric reductions may further lower the benefit-to-risk ratio for new treatments.”

  • Judy George covers neurology and neuroscience news for MedPage Today, writing about brain aging, Alzheimer’s, dementia, MS, rare diseases, epilepsy, autism, headache, stroke, Parkinson’s, ALS, concussion, CTE, sleep, pain, and more. Follow


This work was funded by National Institute on Aging (NIA) grants.

Goldberg had no conflicts of interest to disclose. Co-authors reported relationships with AC Immune, Alpha-Cognition, Athira, Corium, Cortexyme, BioVie, Eli Lilly, GW Research, Lundbeck, Merck, Neurim, Novo Nordisk, Otsuka, Roche/Genentech, Cognition Therapeutics, Takeda, Biohaven, Biogen, Eisai, Novartis, the NIA, the Alzheimer’s Association, Acadia, TauRx, and BioXcel.

The editorialists reported relationships with the MS International Federation, the WHO, Bristol Myers Squibb, Roche, Janssen, Sanofi Genzyme, Novartis, Merck, Biogen, WebMD Global, Eisai, Teva, BioCSL, Celgene, and Eli Lilly.

Primary Source

Journal of Neurology, Neurosurgery and Psychiatry

Source Reference: Goldberg TE, et al “Comparison of relative change with effect size metrics in Alzheimer’s disease clinical trials” J Neurol Neurosurg Psychiatry 2023; DOI: 10.1136/jnnp-2023-331941.

Secondary Source

Journal of Neurology, Neurosurgery and Psychiatry

Source Reference: Kalincik T, Brodtmann A “How effective is effective enough?” J Neurol Neurosurg Psychiatry 2023; DOI: 10.1136/jnnp-2023-332311.

Source link

Share this Article
Leave a comment