After this week’s release of phase 3 trial results for donanemab, a monoclonal antibody that treats Alzheimer’s disease by reducing amyloid beta plaques, I wanted to see how it compared to the last drug with positive results, lecanemab. I made a table comparing the trials side-by-side.
The trials matched identically in a few ways. They lasted 18 months. They enrolled patients with early symptomatic Alzheimer’s disease including mild cognitive impairment and mild dementia with confirmed presence of amyloid pathology. And they measured decline in Clinical Dementia Rating – Sum of Boxes (CDR-SB).
| Lecanemab | Donanemab | ||||
| Subjects | 1,795 | 1,734 | |||
| Treatment duration | Full trial | Until “they reached a prespecified level of amyloid plaque clearance” | |||
| Outcome | 27% reduction | 29% reduction | |||
| Side Effects (treatment vs placebo) | |||||
| Brain swelling (ARIA-E) | 12.5% vs 1.7% | 24.0% | |||
| Brain swelling with symptoms | 2.8% vs 0.0% | 6.1% | |||
| Brain bleeding (ARIA-H) | 17.0% vs 8.7% | 31.4% vs 13.6% | |||
Glossary:
CDR-SB – Clinical Dementia Rating-Sum of Boxes – “CDR-SB is a numeric scale used to quantify the various severity of symptoms of dementia. Based on interviews of people living with AD and family/caregivers, qualified healthcare professionals assess cognitive and functional performance in six areas: memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care.”
ARIA-E – amyloid-related imaging abnormalities-edema/effusion
ARIA-H – ARIA with cerebral microhemorrhages or macrohemorrhages
So the drugs had nearly identical efficacy. Donanemab, the newer drug, had a higher absolute rate of the most serious side effects, but the patients in that control group did too, at least for ARIA-H, so the relative risk of the brain swelling is close to 2x the control group in both trials. The donanemab press release doesn’t include the ARIA-E rates in the placebo group, so I’ll be keen to get more details from the full results.
The donanemab trialists sliced their data in an additional way that I haven’t seen from the lecanemab trial. They split the patients into 552 with high levels of tau and 1,182 with intermediate levels. The data I showed above are for the combined groups, which I believe is most comparable to the lecanemab trial. The headline result Lilly is touting – 36% slowing of CRD-SB decline – is for the intermediate tau group. They never state the results for the high-tau group, but with some basic algebra we can guess they saw about a 14% slowing. That supports the theme I’ve often heard over the last year or two that amyloid beta drugs help more if they’re given earlier.
I’m excited to see the progress in disease-modifying Alzheimer’s treatments. I look forward to seeing results after patients have been treated longer, which I expect to show larger benefits on both an absolute and relative basis. I’m also optimistic about newer versions of amyloid drugs that are more effective with lower side effects, as well as drugs that treat tau.
How to 115 