Jump to comment:

• Treatment of neuropsychiatric syndromes in Alzheimer's disease (AD): In search of evidence?
  Amir A. Sepehry
  Published on: 9 September 2015

• Published on: 9 September 2015

  Treatment of neuropsychiatric syndromes in Alzheimer's disease (AD): In search of evidence?

  • Amir A. Sepehry, Postdoctoral fellow
  • Other Contributors:
    • Philip E. Lee, and Ging-Yuek Robin Hsiung

  We read with interest the recent systematic review and meta-analysis, "Pharmacological treat-ment of neuropsychiatric symptoms in Alzheimer's disease"(1). By examining the evidence for treating heterogeneous range of neuropsychiatric syndromes in persons with Alzheimer's disease (AD), this study found favorable results for the use of cholinesterase inhibitors (ChEIs) and atyp-ical antipsychotics albeit with risk of side ef...

  Show More

  We read with interest the recent systematic review and meta-analysis, "Pharmacological treat-ment of neuropsychiatric symptoms in Alzheimer's disease"(1). By examining the evidence for treating heterogeneous range of neuropsychiatric syndromes in persons with Alzheimer's disease (AD), this study found favorable results for the use of cholinesterase inhibitors (ChEIs) and atyp-ical antipsychotics albeit with risk of side effects, but not for selective serotonin uptake inhibitors (SSRIs), mood stabilizers, or an N-Methyl-D-Aspartate receptor antagonist. The authors sug-gested that their result is generalizable across the spectrum of AD severity. While these findings are important, the results could be expanded. For example, important points including investiga-tion of the heterogeneity (in assessment scales, results, and severity of AD), and methodological limitations need addressing for better understanding of the relative significance of the findings. The Neuropsychiatric Inventory (NPI) used in these trials varies and captures a range of 10 to 12 syndromes. These variations increase the heterogeneity in the findings. Furthermore, this hetero-geneity is enhanced when different compounds with differing mechanism of action or affinity that are designed primarily for individual syndrome use (e.g., antipsychotics for psychosis or de-lusion, or antidepressant for depression, or apathy) are combined. In other words, total NPI score is not necessarily the best reflection of improvement, rather, looking at subtest of items is per-haps more relevant not only in light of studying mechanism of action for each compound, but also for better treatment planning. Same can be said in light of the NPI that aggregates many dif- ferent symptoms, although all related to Behavioral and psychological symptoms of dementia (BPSD), one cannot reasonably expect each medication to be beneficial to all BPSD - i.e. depres-sion and apathy may require something slightly stimulating, while anxiety and agitation may re-quire something slightly sedating. Combining them all under total NPI score and expecting a medication that can cure all is unreasonable. It would have been more informative if these stud-ies reported the effect of each drug on the specific BPSD syndromes and have proper scales to quantify them. Although well formulated to quantify the benefits and safety profiles of various compounds on overall neuropsychiatric syndromes, this meta- analysis could have benefited from taking into account several methodological points, including 1) whether the goal of the included studies was primary or secondary to investigate neuropsychiatric syndromes; and 2) investigated moderating clinical factors.

  The first concern of this study is the lack of explicit statement about the goal of the included studies, which could have shown experimenter bias. This is important for two reasons, A) if the primary goal of the studies was other than neuropsychiatric syndrome, the weight given to indi-vidual compounds would be different. For example, one would anticipate that the cholinesterase inhibitor trials would not be primarily set to find differences in behavioral issues versus the atyp-ical antipsychotic trials, which would be particularly designed to identify such an effect. B) This could be a power issue in that, if the secondary data come from post-hoc analysis, the result could have been underpowered. The second concern is the absence of moderating factors such as inpatient/outpatient, treatment duration, dosage, or severity of AD investigation. Teasing-apart treatment effect for mild, moderate, or severe AD would have added more strength to their find-ing. It is possible that the lack of effect observed in some of the studies is masked because of se-verity of AD, or that optimal dosage was not applied for the preferred duration of treatment.

  In conclusion, with both favorable and dissenting results, although the meta-analysis is well de-signed and carried out, it could have been more informative about the use of psychotropic medi-cation for treating neuropsychiatric syndromes in AD. It could have informed us about the use of each compound for specific AD severity, and duration and dosage of treatment. Also, it could have added notes as to whether these treatment approaches are optimal when used in combina-tion or alone. Given that this study compares itself to earlier meta-analysis and reviews, thus fu-ture studies further benefit by taking into consideration of the above- mentioned points to opti-mize findings.

  References:

  1. Wang J, Yu JT, Wang HF, et al. Pharmacological treatment of neuropsychiatric symptoms in Alzheimer's disease: a systematic review and meta-analysis. J Neurol Neurosurg Psychiatry. 2015;86(1):101-109.

  Conflict of Interest:

  None declared

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top