J Riley McCarten

Pharmacologic Interventions to Prevent Cognitive Decline, Mild Cognitive Impairment, and Clinical Alzheimer-Type Dementia: A Systematic Review

Dementia is a clinical syndrome in which an acquired cognitive deficit interferes with a persons independence in daily activities (1). It adversely affects patient quality of life, burdens caregivers, increases institutionalization, and is costly to families and society (2). Alzheimer disease is the most common cause of dementia, although differentiation between when Alzheimer disease is present in isolation and when it is combined with another cause of dementia may not be possible in clinical settings. Mild cognitive impairment (MCI) is an acquired cognitive deficit in the absence of functional impairment (3, 4) and often leads to dementia (5). Analyses have suggested that behavioral and pharmacologic interventions targeting potentially modifiable risk factors could substantially reduce Alzheimer disease prevalence (6). We did a systematic review of clinical trials to assess the most current evidence about the efficacy and safety of pharmacologic interventions for preventing or delaying cognitive decline, MCI, and dementia and whether effects differ by patient characteristics. Methods We developed and followed a standard protocol (7). The full technical report (8) contains detailed findings. Data Sources We searched MEDLINE, PsycINFO, Embase, and the Cochrane Library for studies published between January 2009 and July 2017 (Part A of the Supplement). We identified earlier studies from a 2010 Agency for Healthcare Research and Quality (AHRQ) report on interventions for preventing Alzheimer disease and cognitive decline (9). We also searched reference lists of eligible studies and articles suggested by experts. Supplement. Supplementary Material Study Selection We included English-language randomized and nonrandomized controlled trials of adults without dementia (that is, with normal cognition or MCI) that compared U.S. Food and Drug Administrationapproved prescription medications versus placebo or any control. Minimum follow-up was 6 months. Primary outcomes were cognitive diagnoses of MCI or dementia (Alzheimer disease or unspecified type, but excluding other specific causes, such as dementia attributable solely to a clinically recognized stroke). Secondary outcomes were measures of cognitive performance assessed by validated instruments. Two reviewers independently examined titles, abstracts, and full articles for eligibility and resolved discrepancies by consensus. Data Extraction and Quality Assessment For each study, 2 reviewers directly and independently rated risk of bias for outcomes of interest and the study overall as low, medium, or high on the basis of AHRQ criteria (10, 11). One reviewer extracted the study design, participant characteristics, pharmacologic interventions, and funding source from all eligible studies but extracted outcomes and adverse events only from studies with low or moderate risk of bias. A second reviewer checked the accuracy of the extracted data. Data Synthesis and Analysis We organized results by intervention type and categorized study participants by baseline cognitive status (normal or MCI). We grouped cognitive tests into memory tests; global cognitive screening tests (such as the Mini-Mental State Examination [MMSE]) or multidomain neuropsychological tests (such as the Alzheimers Disease Assessment ScaleCognitive Subscale); and tests of executive function, attention, and processing speed (Part B of the Supplement). We analyzed and reported cognitive test results by direction of effect and statistical significance because different cognitive tests were used, tests were analyzed and reported in many ways, and we were frequently unable to determine effect size or assess whether between-group differences in scores were clinically meaningful. Because of heterogeneity in study designs and outcomes, results were not pooled in meta-analyses. When a treatment comparison had at least 2 studies or 1 large study (500 participants), 2 reviewers graded strength of evidence for each outcome on the basis of study limitations, directness, consistency, and precision (12); strength of evidence for single smaller trials was graded as insufficient. Assessments were confirmed by consensus. Role of the Funding Source This review was funded by the National Institute on Aging and AHRQ. These agencies and members of the National Academies Committee on Preventing Dementia and Cognitive Impairment helped refine the scope and reviewed the draft AHRQ report. The authors are solely responsible for the content of the manuscript and decision to submit it for publication. Results We identified 102 eligible studies of pharmacologic treatments; 51 unique trials had low or medium risk of bias and were included in analyses (Appendix Figure). Of these, 32 received at least some industry funding. Trials in which baseline cognitive test results were essentially normal and those providing no baseline cognitive data were categorized as enrolling participants with normal baseline cognition. Most studies that enrolled participants with cognitive impairment (1320) specified that participants had MCI defined by the Petersen criteria (13, 1720). The Table shows main efficacy results for all interventions. Details on study and participant characteristics, including cognitive test characteristics, age, baseline cognition, risk-of-bias ratings, funding source, and strength-of-evidence ratings, are presented by pharmacologic intervention in Parts C to H of the Supplement. Appendix Figure. Evidence search and selection. NSAID = nonsteroidal anti-inflammatory drug; ROB = risk of bias. * Some studies were excluded for >1 reason. 1 eligible trial separately compared antihypertensive treatment and lipid-lowering medication versus placebo. Table. Effect of Pharmacologic Interventions Versus Control on Cognitive Outcomes in Adults With Normal Cognition and MCI Table. Continued. Dementia Medications Among 15 eligible references (13, 14, 19, 2132), 3 unique trials had low to moderate risk of bias (13, 14, 19, 32) (Part C of the Supplement). These trials provided mostly low-strength evidence that acetylcholinesterase inhibitors do not reduce risk for dementia or improve cognitive test performance versus placebo in older adults with MCI; they gave insufficient evidence about persons with normal cognition. Most evidence came from 1 trial of 512 adults (mean age, 73 years) with MCI who were randomly assigned to receive donepezil or placebo (13, 19). After 3 years, the treatment groups did not differ in progression to Alzheimer disease (HR, 0.80 [95% CI, 0.57 to 1.13]) (low strength of evidence) or in any cognitive test (insufficient to low strength of evidence). In participants with at least 1 apolipoprotein E 4 allele, those assigned to donepezil progressed to Alzheimer disease less often than those assigned to placebo, although results were not statistically significant after correction for multiple comparisons. Participants who received donepezil were more likely to have adverse gastrointestinal symptoms, sleep disturbances, and arthritis. Antihypertensive Medications Antihypertensive Medication Versus Placebo Among 14 eligible references (3346), 8 unique trials had low to moderate risk of bias (3344) (Part D of the Supplement). These trials enrolled 31287 participants (mean age among 7 trials reporting, 70.2 years). Low-strength evidence in adults with normal cognition showed that antihypertensive treatment versus placebo does not reduce risk for dementia, and moderate-strength evidence showed no difference between these treatments on global cognitive screening tests. No trials reported data about persons with MCI. Evidence on risk for dementia was based on 4 trials. With antihypertensive treatment, the Syst-Eur (Systolic Hypertension in Europe) trial reported a halving of incident dementia of marginal statistical significance (35, 36), although 3 subsequent and larger trials each reported no reduced risk for dementia (38, 42, 43). TRANSCEND (Telmisartan Randomised Assessment Study in Angiotensin Converting Enzyme Inhibitor Intolerant Subjects with Cardiovascular Disease) reported no difference between antihypertensive treatment and placebo in incident cognitive impairment, defined as a composite of incident dementia, incident cognitive impairment, or MMSE score less than 24 (low strength of evidence) (33). Four trials reported no difference between antihypertensive treatment and placebo in change on global cognitive screening tests (moderate strength of evidence) (33, 35, 36, 38, 43), whereas weaker evidence suggested no difference between treatment groups for other cognitive tests (34, 3740). In the only trial that reported subgroup analyses, treatment effects on risk for incident cognitive impairment and cognitive test performance did not differ as a function of age or history of hypertension, stroke, or transient ischemic attack (33). In the 3 studies that reported adverse effects (38, 41, 43), other than for methyldopa (41) no consistent differences existed between active treatment and placebo. Intensive Versus Standard Antihypertensive Treatment One unique trial with moderate risk of bias randomly assigned 1439 adults (mean age, 62.4 years) with normal cognition, diabetes, and heightened cardiovascular risk to a systolic blood pressure target of less than 120 mm Hg or less than 140 mm Hg for 40 months. It reported no data on MCI or dementia outcomes and no between-group difference in any measured cognitive test (low strength of evidence) (47). Participants aged 70 years or older did better with intensive than standard treatment on a single evaluated memory test and 1 of 2 tests of executive function, attention, and processing speed; however, performance did not differ by sex, baseline cognition, cardiovascular disease history, or diabetes duration. Adverse events data were not reported. Antihypertensive Medication Comparisons Among 11 eligible references (15, 16, 33, 34, 41, 4853), 9 unique trials (15, 16, 33, 34, 41, 4

Using Cerebrospinal Fluid Biomarker Testing to Target Treatment to Patients with Mild Cognitive Impairment: A Cost-Effectiveness Analysis

ObjectiveCerebrospinal fluid (CSF) biomarkers are shown to facilitate a risk identification of patients with mild cognitive impairment (MCI) into different risk levels of progression to Alzheimer’s disease (AD). Knowing a patient’s risk level provides an opportunity for earlier interventions, which could result in potential greater benefits. We assessed the cost effectiveness of the use of CSF biomarkers in MCI patients where the treatment decision was based on patients’ risk level.MethodsWe developed a state-transition model to project lifetime quality-adjusted life-years (QALYs) and costs for a cohort of 65-year-old MCI patients from a US societal perspective. We compared four test-and-treat strategies where the decision to treat was based on a patient’s risk level (low, intermediate, high) of progressing to AD with two strategies without testing, one where no patients were treated during the MCI phase and in the other all patients were treated. We performed deterministic and probabilistic sensitivity analyses to evaluate parameter uncertainty.ResultsTesting and treating low-risk MCI patients was the most cost-effective strategy with an incremental cost-effectiveness ratio (ICER) of US

Neuropsychological Test Descriptions

37,700 per QALY. Our results were most sensitive to the level of treatment effectiveness for patients with mild AD and for MCI patients. Moreover, the ICERs for this strategy at the 2.5th and 97.5th percentiles were US

Stroke/TIA Outcomes Tables

18,900 and US

Cognitive Outcomes Tables: Incident Cognitive Impairment Defined by Composite of Neuropsychological Tests

50,100 per QALY, respectively.ConclusionBased on the best available evidence regarding the treatment effectiveness for MCI, this study suggests the potential value of performing CSF biomarker testing for early targeted treatments among MCI patients with a narrow range for the ICER.