For many people living with dementia and their caregivers, the worst symptom of the disease is not memory loss, but agitation. Whether it comes out as verbal insults or physical violence, agitation in people with Alzheimer's disease is often what drives distraught carers to institutionalize their loved one. No safe, effective treatment exists. Against this bleak backdrop, positive findings from Phase 3 studies of brexpiprazole, a drug approved for schizophrenia, received a warm welcome at the Clinical Trials in Alzheimer’s Disease meeting, held November 29 to December 2 in San Francisco. According to Nanco Hefting of Lundbeck in Copenhagen, the drug curbed agitation in AD, meeting the primary endpoint in a third Phase 3 trial. Researchers welcomed the positive findings, but also called for continued monitoring to ensure the drug is safe in this vulnerable population.

  • Brexpiprazole lessened agitation in people with Alzheimer's.
  • Agitation scores dropped steeply in the placebo group, too.
  • In SYMBAD trial, mirtazapine did not reduce agitation over placebo, but both groups benefited.

Curiously, the benefit of brexpiprazole over placebo paled in comparison to the effect of merely participating in the trial, as agitation eased substantially in both the treatment and placebo groups. This finding jibes with results of another agitation trial presented at the conference, which tested the anti-depressant mirtazapine in people with AD. Though that drug worked no better than placebo, agitation dropped substantially in both groups throughout that trial, as well. Together, the findings point to the power of quality medical care and social support. These factors constitute a level of attention rarely received outside of a clinical trial.

The lack of proven treatments to calm agitation in AD has led to much off-label use of sedatives. These come with a bevy of side effects in people with dementia, including falls and fractures, worsening cognitive decline, and cerebrovascular problems, not to mention low effectiveness. “When considering pharmacotherapy for these patients, the goal is to treat the agitation, and not to sedate the patient,” Hefting told the audience.

To that end, Lundbeck teamed up with Otsuka Pharmaceuticals back in 2013 to evaluate brexpiprazole for the treatment of agitation in people with AD. Brexpiprazole is a successor to Otsuka’s aripiprazole, marketed as Abilify or Aripiprex, a dopamine D2 receptor agonist that is widely used to treat schizophrenia and related affective disorders. Compared to its predecessor, brexpiprazole interacts with a broader range of neurotransmitter receptors. In people with schizophrenia, the drug is considered neither activating nor sedating.

Over the past decade, the companies have completed three Phase 3 trials for AD-associated agitation. Results from the first two trials were reported in 2018 and subsequently published together (Grossberg et al., 2020). The first trial tested safety, efficacy, and tolerability of 1 or 2 mg brexpiprazole, or placebo, daily for three months in 420 participants in the United States, Europe, and Russia. The second enrolled 270 participants and used a flexible dose design, with doses ranging from 0.5 mg to 2 mg per day. Participants with AD dementia were included if they scored at least four points on the aggression/agitation domain of the NIH neuropsychological inventory (NPI-NIH-A/A). Change on the Cohen-Mansfield Agitation Inventory (CMAI) at 12 weeks served as the primary endpoint for both trials.

Trial 1 met its primary endpoint, curbing CMAI scores at 12 weeks. The second trial missed the endpoint, although a post hoc analysis suggested an easing of agitation among those who had received the 2 mg dose. Brexpiprazole also appeared to be effective among participants who had displayed frequent aggressive behaviors—either verbal or physical—at baseline. In both trials, participants taking brexpiprazole fared better than those receiving placebo on the Clinical Global Impression-Severity of Illness (CGI-S), a secondary measure of agitation completed by the clinician.

Based on their learnings from the first two trials, the researchers initiated a third Phase 3 trial that compared 2 mg and 3 mg daily doses of brexpiprazole to placebo. In addition to the inclusion criteria for the first two trials, the third trial required participants to meet criteria for CMAI Factor 1, meaning they frequently displayed aggressive behaviors. The trial enrolled 345 participants, and randomized 117 to placebo and 228 to brexpiprazole, of whom 75 received 2 mg and 153 received the 3 mg daily dose. CMAI scores dropped substantially in the placebo and brexpiprazole groups throughout the trial. By eight weeks, the brexpiprazole group had edged out the placebo group. By the end of the trial, CMAI scores had dropped by 22 points in those taking either dose of brexpiprazole, and by 17 points in the placebo group. The same pattern was observed for scores on the CGI-S, a key secondary endpoint.

Easing the Mind. Over a 12-week trial, agitation scores dropped in participants in both the placebo and brexpiprazole groups. For the final timepoints, brexpiprazole was significantly more effective than placebo. [Courtesy of Nanco Hefting, Lundbeck.]

While brexpiprazole eased all manner of agitated behaviors, it was most effective at calming the most severe ones: the drug reduced the frequency of overtly aggressive acts more than of verbal agitation or physically nonaggressive behaviors.

Were these benefits clinically meaningful? To address this, the researchers compared the number of participants in the placebo versus the combined treatment group whose CMAI scores dropped by at least 20, 30, or 40 percent. They found that a significantly greater proportion of participants in the treatment groups reached each of these response thresholds. For example, CMAI scores dropped by at least 20 percent in 68 percent of participants receiving brexpiprazole, compared with 47 percent of participants in the placebo group.

Brexpiprazole was generally well-tolerated and safe. This was true whether analyzing the third trial alone, or pooled safety data from all three Phase 3 trials. No adverse event had an incidence of more than 5 percent, and none occurred significantly more frequently in the brexpiprazole group. One notable exception was death—across all three trials, six people (0.9 percent) on brexpiprazole died, compared to one person, or 0.3 percent, on placebo. Hefting said that the deaths were deemed unrelated to the drug. Importantly, brexpiprazole did not sedate participants.

In a panel discussion following Hefting’s presentation, panelists largely commended the findings. Noting the extensive off-label use of drugs in people with AD and agitation, Alireza Atri of the Banner Sun Health Research Institute in Arizona, said “Anything that is safe and effective is welcome.” “Generally, patients and families are on knife’s edge when this happens,” Atri said, adding that agitation is often what breaks a family's resolve. Atri was encouraged by the safety data, but on efficacy would have liked to see data on the drug’s effect size, and on how much it improved the quality of life of individual patients and their caregivers.

Clive Ballard of the University of Exeter, U.K., was pleased to see a trial in a population that is notoriously difficult to recruit, and that included institutionalized patients. Ballard, too, wanted more effect size data, noting that a Cohen’s d value would have been helpful. Ballard added that while the overall safety profile of brexpiprazole appeared superior to other atypical antipsychotic drugs, he remained concerned about the elevated mortality reported among those taking brexpiprazole. He suggested monitoring mortality and safety among those prescribed the drug in the future. He also noted that at an average of 74, the trial participants were a decade younger than the average nursing home resident with dementia. “I would like to see safety data in an older, frailer group,” he said.

A more positive view came from Pierre Tariot, also at Banner, who said he thought the jury had been a bit sober so far. “We had efficacy without all the toxicity that has plagued the other drugs we’ve been studying for agitation,” he said. “This sounds like a major advance.” Tariot asked what distinguished brexpiprazole from its predecessor, aripiprazole, which has an unclear track record in easing agitation in people with AD and causes more side effects. Hefting said brexpiprazole’s edge may come from its broader, more balanced relationship with neurotransmitter receptors. It partially agonizes serotonin 5-HT1A receptors and dopamine D2 receptors, while quelling signaling from serotonin 5-HT2A and noradrenaline receptors.

At the end of the discussion, Anton Porsteinsson, University of Rochester, New York, pointed out the elephant in the room: While much focus was given to the five-point difference between the placebo and drug groups on the CMAI, a much larger effect came from simply participating in the trial, as CMAI scores dropped by 17 points in the placebo group. “We have a very substantive placebo response,” Porsteinsson said, “and we won’t be able to replicate it in clinical practice because we can’t give the degree of attention that we do in clinical trials.”

Sube Banerjee of the University of Plymouth, U.K., calls this apparent “placebo effect” a “treatment-as-usual effect,” referring to the benefits that come with the care given to all participants in a clinical trial. Few people have access to the level of care deemed “usual” in the context of a clinical trial, and when they do, it can have a stronger effect on agitation than any drug, he suggested (see full comment below).

Making Treatment-as-Usual More Usual
At CTAD, Banerjee presented published findings from the SYMBAD trial, which evaluated mirtazapine in people with AD and agitation (Banerjee et al., 2021). This antidepressant is widely prescribed for off-label use in this population. The trial enrolled, from 26 U.K. centers, 204 people with AD whose agitation was unresponsive to nondrug treatment. Nearly half lived in a care home; their average age was 82. Over 12 weeks, CMAI scores dropped by 12 points in both the placebo and mirtazapine groups. “There was considerable improvement, but it had nothing to do with being given mirtazapine,” Banerjee told the audience.

Why did being in the trial work so well? For one, Banerjee noted that agitation in AD can arise from many sources. They can be unmet physical needs, such as hunger, thirst, or pain. They can be an overcrowded, noisy, too stimulating, or too boring environment. They can be psychosocial issues such as stress, loneliness, depression, or lack of purpose. They can be relational issues with the family, caregivers, and other residents in a care home. Dementia itself was last on his list of sources of agitation, and Banerjee said this was by design. “The assumption that the agitation is caused by the dementia itself, rather than societal, family, or environmental factors to do with dementia, must be tested,” he said. “When you start looking at all of these causes of agitation, what is the likelihood that a drug of any kind would help with these unmet needs?”

Banerjee believes agitation in dementia requires a multipronged approach, noting that small studies have demonstrated that nondrug interventions, which take time to engage with the person with dementia, help reduce agitation (Livingston et al., 2014). The trouble is, such mental and social support services are underfunded and unavailable to most people with dementia who live at home or in facilities.

“Health systems would be well advised to focus on providing good-quality, nondrug psychosocial care and support for those with agitation in dementia rather than seeking to use medication to deal with these complex states,” Banerjee wrote to Alzforum. “Drugs should be a last line treatment in all but the most severe cases.”—Jessica Shugart


  1. It was good to see the set of well-conducted and careful trials. The observed difference from the placebo group looked in line with the observed effects of other antipsychotics. The trials were of short duration and longer-term safety follow-up data were not presented. I was a little concerned that what looked like a three-fold increase in deaths in the intervention groups was presented as unconnected, when the deaths associated with antipsychotics in dementia that are of such concern are not of a specific type. We need more data on safety.

    In terms of the "placebo effect," I would suggest that is likely to be a "treatment as usual effect" of all the other things that happen when an individual is identified as having agitation in dementia by the sort of services that would have referred into the trial. The effect size of this treatment as usual effect is very much larger than that of the drug and suggests, as we concluded in the Lancet paper describing our SYMBAD trial, that watchful waiting while investigating and addressing the causes of agitation is the best first step. Health systems would be well advised to focus on providing good quality nondrug psychosocial care and support for those with agitation in dementia rather than seeking to use medication to deal with these complex states which have multiple aetiologies. Drugs should be a last line treatment not a first line one in all but the most severe cases.

Make a Comment

To make a comment you must login or register.


Therapeutics Citations

  1. Brexpiprazole
  2. Aripiprazole

Paper Citations

  1. . Efficacy and Safety of Brexpiprazole for the Treatment of Agitation in Alzheimer's Dementia: Two 12-Week, Randomized, Double-Blind, Placebo-Controlled Trials. Am J Geriatr Psychiatry. 2020 Apr;28(4):383-400. Epub 2019 Oct 1 PubMed.
  2. . Study of mirtazapine for agitated behaviours in dementia (SYMBAD): a randomised, double-blind, placebo-controlled trial. Lancet. 2021 Oct 23;398(10310):1487-1497. PubMed.
  3. . A systematic review of the clinical effectiveness and cost-effectiveness of sensory, psychological and behavioural interventions for managing agitation in older adults with dementia. Health Technol Assess. 2014 Jun;18(39):1-226, v-vi. PubMed.

Further Reading

No Available Further Reading