Donepezil and Memantine for Alzheimer’s Disease

Published - Written by Lisa Rosenbaum

All of us who have completed an internship are well-familiar with the poly-pharmacy-discharge challenge. The patient comes in on eight medications. Now it’s time to go home, and you’ve gone and added another four. Surely the patient doesn’t need all of these, but as you review her list, it’s not clear that any can be safely discontinued. But just as you are about to click, “Select All,” you start wondering whether she really needs the donepezil, a cholinesterase inhibitor often used to treat Alzheimer’s disease.  After all, she’s been taking donepezil for years, but her cognition has only gotten worse. Moreover, she’s been losing weight, which donepezil is known to cause, and now she needs marinol to stimulate her appetite. But as much as you’d like to eliminate unnecessary medications, how do you really know that the donepezil isn’t doing her some good?

You don’t. That’s why the results of the “Donepezil and Memantine in Moderate to Severe Alzheimer’s Disease trial,” (DOMINO) trial, published in this week’s NEJM, may help you, (or at least answer your question). In this placebo-controlled, two-by-two factorial trial, patients with moderate to severe Alzheimer’s disease, who had been on donepezil for an average of 2 to 3 years, were randomized to either continue donepezil, or to stop it. So far, makes sense. But then the trial got a bit more complicated.

Memantine, an NMDA receptor antagonist, is another Alzheimer’s medication that is only FDA approved for patients with moderate to severe disease. Nevertheless, as Lon Schneider points out in his accompanying editorial, many US physicians add memantine to donepezil therapy, even when patients are only mildly impaired. Thus, these investigators also wanted to investigate the combined effect of memantine and donepezil, as well as the effect of memantine alone. Hence, the 2 x 2 factorial design, where both the donepezil continuation and discontinuation groups were further randomized to the addition of either memantine or placebo.

The design may be complicated, but the results were pretty straightforward. About 300 community-living patients were followed for one year. The primary outcome included standard measures of cognitive ability, as well as ability to perform activities of daily living. The results are clear: patients who were continued on their donepezil had better cognitive function and functional status than those who weren’t. Adding memantine to donepezil did not prove synergistic; however, starting memantine among those whose donepezil had been discontinued was better than not starting memantine at all. One drug is better than none. Two drugs are not better than one. But what about no drug at all?

Unfortunately, the trial does not really answer this question. Because this is a discontinuation trial, there is always the possibility that the benefit seen with donepezil was not related to the efficacy of donepezil, but instead related to the withdrawal that may occur upon its discontinuation. Moreover, we don’t really know whether donepezil was helping any of these patients in the first place; we just know what happened when they stopped it. Whether or not to initiate donepezil among patients with moderate to severe Alzheimer’s disease is not answered by this trial. Finally, it should be noted that the benefit of donepezil continuation, overall, was quite modest in comparison to the degree of general deterioration during the follow up year. This relentless decline in cognitive and functional status suggests, more than anything, the need for better drugs to treat Alzheimer’s.

For now, however, the war against poly-pharmacy will wage on, with one more medication we should likely not discontinue. Or as a wise attending once said, “In medicine, never shoot a singing bird.”