Donanemab slows cognitive impairment in mild Alzheimer’s disease – an important step forward

The results of the phase 3 clinical trial of the anti-amyloid monoclonal antibody donanemab were released simultaneously today at the Alzheimer’s Association Clinical Conference (AAIC) in Amsterdam and in a paper published in JAMA. I have been looking forward to these results because donanemab is the most effective monoclonal antibody yet for removing amyloid from the brain. Also, this study is the first trial of an anti-amyloid monoclonal antibody to study effects on brain tau, both by tau-PET scans and measurement of plasma P-tau217, a sensitive and specific blood biomarker for amyloid-associated tau phosphorylation and secretion.

As shown in the figure above, beta-amyloid seen in amyloid PET scans dropped dramatically in subjects receiving donanemab compared to placebo.  About three quarters of subjects receiving donanemab totally cleared plaques on PET scans at 76 weeks, and the medication could be stopped. This is far and away the most effective clearance of amyloid in the brain yet seen with anti-amyloid  medications.

The effects on slowing of cognitive impairment were assessed using the integrated Alzheimer’s Disease Rating Scale (iADRS). Among subjects with milder Alzheimer’s who had a low or medium PET tau burden, donanemab produced a 35.1% slowing of cognitive decline, a benefit that seemed to persist even after treatment with donanemab was stopped.  However, in subjects who started out with high tau burden on PET, those with more severe disease, there was much less if any benefit.

What about the effect of donanemab on tau? The results were mixed.  While there was a significant drop in plasma tau217 in subjects receiving donanemab, there was little if any reduction in abnormal tau seen in PET scans. Perhaps plasma tau217 is just a more sensitive biomarker than tau PET scans.

What about side effects? Swelling and/or bleeding in the brain, ARIA-E and ARIA-H occurred in 24.0% and 31.4%.  Most of the time these ARIA were mild and resolved as in other trials. However, clinically serious ARIA-E or ARIA-H occurred respectively in 2.9% and 0.8%, including 3 deaths. As in other trials, carriers of the APOE-4 allele were more likely to have ARIA. Among carriers of one copy of the APOE-4 allele, 22.8% experienced ARIA, whereas 40.6% of homozygotes had ARIA. Interestingly, APOE-4 carriers appeared to have less cognitive protection than non-carriers in response to donanemab, but the numbers were too low to assess any statistical significance.

Would I be willing to take donanemab? No, for the same reasons I mentioned for lecanemab.  I have two copies of the APOE-4 allele. That alone puts me at increased risk of ARIA as well a smaller chance to benefit from anti-amyloid medication. More importantly, I have already had life-threatening ARIA after only four doses of aducanumab. This suggests that I am likely to have cerebral amyloid angiopathy. However, for carefully screened patients with early Alzheimer’s disease who do not carry the APOE-4 allele, donanemab appears to be reasonable option now. It’s not perfect, but it is an important step forward. In an editorial accompanying the paper in JAMA, Gil Rabinovici and Renaud La Joie at UCSF caution: “while the slowing of clinical decline seen in this trial represents an important start, and may be deemed clinically meaningful for some patients, development of more impactful and safer treatments is still needed.”