Behavioral and Psychological Symptoms—An Emerging Crisis of the Alzheimer Dementia Epidemic

Behavioral and psychological symptoms of dementia (BPSD) are implicated in a cycle of negative events including deterioration of family and professional relationships, as well as increased caregiver burden, institutionalization, and risk of death in individuals with Alzheimer dementia (AD).¹ Furthermore, BPSD, which include agitation, psychosis, depression, and apathy, affect nearly all individuals with AD over the course of the disease.² With the aging of the population leading to an epidemic of AD, effective and safe treatments for BPSD are needed. While behavioral therapies are recommended as first-line treatments for BPSD, they may be less effective with a delayed onset of action, especially for patients with the most severe agitation.

For these reasons, psychotropic medications, especially antipsychotics, are widely used to treat agitation and associated behavioral symptoms in AD. However, efficacy data for antipsychotic medications are inconsistent³ and their use is associated with increased mortality, resulting in a US Food and Drug Administration (FDA) boxed warning.^4,5 Furthermore, randomized clinical trials do not provide relative comparisons of efficacy and safety among the antipsychotic medications used to treat BPSD. Although antidepressants, particularly selective serotonin reuptake inhibitors, have fewer and less severe adverse effects than antipsychotic medications, treatment may still be complicated by cardiac conduction-delaying effects and effects on reducing agitation (40% response rate) are not evident until after 6 to 9 weeks of treatment.⁶

Yunusa and colleagues⁷ examined what were reported in randomized clinical trials to be the most effective and safe antipsychotic medications for the treatment of the behavioral and psychological symptoms of dementia. This network meta-analysis of 17 randomized clinical trials, the majority of which were conducted in nursing homes, includes 5373 patients with AD and related agitation, with a mean age of 80.8 years, nearly 70% women, and a mean duration of follow-up of 10 weeks. The medications included in this analysis (aripiprazole, olanzapine, quetiapine, and risperidone) represent the most widely used atypical antipsychotic medications (AAPs) in clinical practice for individuals with dementia and associated agitation.

The authors point out that the network meta-analysis method enables inference for all possible comparisons between interventions included in the analysis even if never studied directly in clinical trials. This approach represents an advance over existing individual trials that compare a single drug with placebo or an efficacy and safety comparison of 2 medications. Behavioral outcomes included the Neuropsychiatric Inventory, the Brief Psychiatric Rating Scale, and the Cohen-Mansfield Agitation Inventory—3 measures that span a wide variety of neuropsychiatric symptoms including agitation, aggression, psychosis, depression, and anxiety. Prior studies are also limited by use of specific outcome measures that vary across studies, limiting the comparison of findings.

Results of this network meta-analysis identified specific efficacy advantages of individual medications compared with placebo. Although no AAP was associated with superior results compared with the others across all effectiveness and safety outcomes, aripiprazole was the only 1 of the 4 AAPs studied to demonstrate efficacy advantages over placebo across all 3 behavioral measures. Quetiapine was associated with improvement on the Brief Psychiatric Rating Scale compared with placebo, and risperidone was associated with improvement on the Cohen-Mansfield Agitation Inventory compared with placebo; olanzapine did not demonstrate efficacy advantages over placebo for any of the 3 behavioral outcomes measures. The effect size of aripiprazole, risperidone, and quetiapine over placebo was small and consistent with previous findings, further highlighting the need for more-effective behavioral and pharmacotherapies for BPSD.

The network meta-analytic approach also compared important safety differences across the 4 AAPs studied. For example, none of the included AAPs were significantly different from placebo or from each other on the risk of death, although in support of findings from the FDA boxed warning, placebo had the highest probability of safety on the mortality outcome (87.3%), followed by risperidone (55.4%), aripiprazole (37.9%), quetiapine (37.1%), and olanzapine (32.4%). Cerebrovascular adverse events, noted in previous individual trials as another significant clinically adverse effect of antipsychotic medications, were more commonly identified in patients treated with olanzapine and risperidone, compared with placebo, although none of the included AAPs were significantly different from one another on cerebrovascular adverse events risk. Consistent with clinical insight and experience, quetiapine was considered the safest for extrapyramidal adverse effects, with risperidone the least safe and most likely associated with sedation and somnolence. The authors suggest that these findings support the clinical recommendation that olanzapine and risperidone should be avoided or used with caution in individuals with hypovolemia, history of cerebrovascular and cardiac diseases, and vascular dementia who are already at high risk for stroke or transient ischemic attack.

Pharma-funded clinical trials assessing safety and efficacy of AAPs essentially ended in 2005 given the concerns regarding mortality risk raised by the FDA meta-analysis and boxed warning. Subsequent studies included analyses of large national and Veterans Affairs databases,⁸ further highlighting the risks of mortality associated with AAPs and conventional antipsychotics, such as haloperidol. Unlike previous placebo-controlled studies or analyses of large clinical databases, this study’s network meta-analysis method allows for a comparison of 4 widely used AAPs providing clinical guidance on the relative risks and benefits of specific AAPs when used to treat BPSD.⁷ Without further comparison or placebo-controlled trials, clinicians are left to follow the 2016 American Psychiatric Association guidelines⁹ that highlight careful clinical assessment, behavioral interventions, and judicious use of antipsychotic medications for time-limited interventions when benefits outweigh safety concerns.

Clinical decisions are individualized and often dependent on complex subjective factors beyond the application of available clinical trial data. Research into the diverse causes of the behavioral and psychological symptoms of dementia may help inform the development of more-targeted behavioral and pharmacologic therapies. Furthermore, there is a clear need for leveraging novel technologies that can phenotype behavioral symptoms of dementia and provide a more precise clinical picture to guide decision making.¹⁰ Ultimately, our ability to reduce the burden from these symptoms will depend directly on generating both basic and clinical research and to rapidly translate this into effective interventions.

Published: March 22, 2019. doi:10.1001/jamanetworkopen.2019.0790

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2019 Forester BP et al. JAMA Network Open.

Corresponding Author: Brent P. Forester, MD, MSc, McLean Hospital, 115 Mill St, Belmont, MA 02478 (bforester@mclean.harvard.edu).

Conflict of Interest Disclosures: Dr Forester reported grants from the National Institute of Aging, grants and other funding from Biogen, grants and other funding from Eli Lilly, grants from the Rogers Family Foundation, and grants from Spier Family Foundation outside the submitted work. Dr Vahia reported receiving an editorial honorarium from the American Journal of Geriatric Psychiatry outside the submitted work.