Little Evidence That Cholinesterase Inhibitors Prevent Progression of Mild Cognitive Impairment to Dementia, but They Are Associated With Adverse Effects
Carlos H Rojas-Fernandez
Evid Based Ment Health. 2013;16(2):39
Russ TC, Morling JR. Cholinesterase inhibitors for mild cognitive impairment. Cochrane Database Syst Rev 2012;9:CD009132.
How effective and safe are cholinesterase inhibitors for mild cognitive impairment?
Progression to dementia or Alzheimer's disease (National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's disease and Related Disorders Association (NINDS-ARDRA), International Classification of Diseases (ICD)-10 or Diagnostic and Statistical Manual (DSM-IV criteria), dementia with Lewy bodies or vascular dementia (consensus criteria) at 12, 24 and 36 months and adverse events.
Systematic review with meta-analysis.
Cochrane Dementia and Cognitive Improvement Group's Specialised Register ALOIS was searched for randomised controlled trials (RCTs). ALOIS is populated from searches in MEDLINE, EMBASE, PsycINFO, CINAHL, LILACS, Cochrane Central Register of Controlled Trials (CENTRAL), WHO trials portal (ICTRP), selected trial registries and grey literature resources. Additional searches were run in MEDLINE, EMBASE, PsycINFO, CINAHL, LILACS, CENTRAL, ICTRP and Clinicaltrials.gov to retrieve most recent papers. Reference lists of identified studies were hand searched. Search dates were not stated.
Study Selection and Analysis
Two reviewers appraised studies and selected double-blind RCTs comparing any cholinesterase inhibitor (galantamine, donepezil or rivastigmine) used for at least 1 month versus placebo in adults with mild cognitive impairment (MCI, study definitions of MCI accepted if consistent with a subjective memory complaint and essentially normal functioning). Two reviewers rated study quality and extracted data. Reporting bias was assessed with a funnel plot. Heterogeneity was explored with the χ2 test and forest plot. Mantel-Haenzel random-effects meta-analyses were conducted using RevMan software.
Nine RCTs (n=5149; donepezil: 3 RCTs; galantamine: 4 RCTs; rivastigmine: 2 RCTs) met inclusion criteria but only three RCTs could be included in the meta-analyses examining progression to dementia because of differences in the lengths of each trial. There was no significant difference in progression to dementia between cholinesterase inhibitors and placebo at 12 months (3 RCTs, n=2560; RR 0.69, 95% CI 0.47 to 1.00, p=0.05; I2=55%). At 24 months, cholinesterase inhibitors reduced conversion to dementia compared with placebo (2 RCTs, n=2048; RR 0.67, 95% CI 0.55 to 0.83, p=0.0002; I2=0%). The two RCTs that were pooled in the 24-month analysis were separate studies reported in the same article. At 36 months, there was no significant difference in progression to dementia between cholinesterase inhibitors and placebo (2 RCTs, n=1530; RR 0.84, 95% CI 0.70 to 1.02, p=0.08; I2=0%). Cholinesterase inhibitors increased overall adverse events compared with placebo (5 RCTs; RR 1.09, 95% CI 1.02 to 1.16, p=0.0007). There were no significant differences between groups for serious adverse events, cardiac problems, depression or death (see Webextra table S1). Cholinesterase inhibitors increased the risk of muscle cramps, vomiting, abnormal dreams, nausea, diarrhoea, insomnia, syncope/dizziness and headache compared with placebo (see supplementary table S1).
There is little evidence that cholinesterase inhibitors can reduce the progression of mild cognitive impairment to dementia, and they are associated with an increased risk of side effects.
MCI is a common problem in older people (10–20% of people >65 years are affected) and the incidence of dementia in those with MCI (~5–10%/year) is higher than normal controls (~1–2%/year), underscoring its importance. From the pharmacotherapeutic perspective, MCI presents a challenge, as studies of acetylcholinesterase inhibitors (ACHEIs) in MCI have been disappointing. Prior reviews yielded mixed results, leaving clinicians with no clear direction regarding ACHEI's role in MCI.
Russ and Morling conducted a review designed to overcome shortcomings of previous reviews and address the issue of whether ACHEIs might stall the progression of MCI. This is a germane clinical question, as in some countries, ACHEIs are used off-label for MCI (a practice which is ill-advised). They used stringent and appropriate methods in their Cochrane review and selected only double-blind, placebo-controlled studies. They conclude that ACHEIs should not be recommended for use in MCI as the evidence that they affect progression to dementia is weak and inconsistent, and clinically significant side effects are evident. Their conclusions are consistent with past reviews and extend findings to demonstrate that, regardless of different study time frames, the outcome is the same (ie, lack of efficacy). Appropriately, the authors note that it is not entirely possible to be certain of a lack of an effect owing to methodological considerations, a point previously noted by others. Nevertheless, when safety is considered in this context, the conclusion to not use these drugs for MCI is obvious. Although the lack of a mortality signal in this analysis was reassuring, other side effects were significantly higher: diarrhoea, nausea and vomiting were two-to-four times more common with the use of ACHEIs, whereas insomnia, syncope or dizziness, abnormal dreams and leg cramps or muscle spasms were approximately two-to-seven times higher. Such side effects (which would be expected with higher frequency among frail elderly) are not trivial, and might lead to a prescribing cascade, which would potentially lead to adverse events, affect the quality of life and increase healthcare utilisation. The findings thus provide an appropriate context for watchful waiting for those with MCI.
Sources of funding
Alzheimer Scotland, BBSRC, EPSRC, ESRC and MRC, UK
This article has been corrected since it was published Online First. The author name has been corrected from 'Carlos Rojas-Fernandez PharmD' to 'Carlos H Rojas-Fernandez'.