Systematic reviews examine Alzheimer's detection, biomarkers, pharmacological treatments

An accompanying editorial noted that while some tools are available to identify, diagnose, and treat cognitive dysfunction from Alzheimer's dementia, they are blunt and lack clinical utility.

Three systematic reviews looked at evidence for brief cognitive tests to distinguish Alzheimer's dementia from mild cognitive impairment (MCI) or normal cognition, biomarker testing for neuropathologic Alzheimer's disease, and the use of prescription drugs and supplements.

The three reviews were published April 28 by Annals of Internal Medicine.

The first review of 57 studies concluded that many brief, single cognitive tests accurately distinguish clinical Alzheimer's-type dementia from normal cognition in older adults, but are less accurate for distinguishing mild dementia from normal cognition or dementia from MCI.

Numerous standalone tests were sensitive for distinguishing Alzheimer's dementia from normal cognition. These included standalone tests such as the clock-drawing test (median sensitivity 0.79 and specificity 0.88 in eight studies), Mini-Mental State Examination (0.88 and 0.94, respectively, in seven studies), Montreal Cognitive Assessment (0.94 and 0.94 in two studies), and Brief Alzheimer Screen (0.92 and 0.97 in one study), as well as memory tests (list delayed recall, 0.89 and 0.94, in five studies), and language tests (category fluency, 0.92 and 0.89, nine studies). No studies reported on testing harms. The authors noted that the included studies were small, few test metrics were evaluated by multiple studies, and few studies directly compared different tests, scores, cut points, or test combinations.

The second review of 15 brain imaging studies and nine cerebrospinal fluid studies concluded that amyloid positron emission tomography (PET), fluorodeoxyglucose (FDG) PET, MRI, and cerebrospinal fluid tests were highly sensitive for neuropathologic Alzheimer's disease.

Median sensitivity and specificity, respectively, were 0.91 and 0.92 for amyloid PET, 0.89 and 0.74 for 18F-labeled-FDG PET, 0.64 and 0.83 for single-photon emission CT, and 0.91 and 0.89 for medial temporal lobe atrophy on MRI. Individual cerebrospinal fluid biomarkers and ratios had moderate sensitivity (range, 0.62 to 0.83) and specificity (range, 0.53 to 0.69). Single studies suggested that amyloid PET, 18F-FDG PET, and cerebrospinal fluid test combinations may add accuracy to clinical evaluation. The authors cautioned that the studies included were small, biomarker cut points and neuropathologic Alzheimer's disease were inconsistently defined, and methods with uncertain applicability to typical clinical settings were used.

The third review of prescription drugs and over-the-counter supplements included 55 studies for cognitive and functional outcomes and 12 for neuropsychiatric outcomes. Across Alzheimer's disease severity, mostly low-strength evidence suggested that, compared with placebo, cholinesterase inhibitors produced small average improvements in cognition, no difference to small improvement in function, no difference in the likelihood of at least moderate improvement in global clinical impression, and increased withdrawals due to adverse events. The differences of these drugs versus placebo were of uncertain clinical importance, the authors wrote. Also, insufficient evidence for greater benefit than harm was noted for pharmacologic treatment of neuropsychiatric outcomes or supplements on cognitive, functional, and neuropsychiatric outcomes. Risk of bias was high for the majority of trials of most drugs, especially those involving supplements, those using active drug comparisons, and those of longer duration.

An accompanying editorial stated that physicians have some tools to identify, diagnose, and treat cognitive dysfunction from Alzheimer's dementia. “However, the tools are blunt and have limited utility,” the editorialists wrote. While biomarkers in PET and cerebrospinal fluid seem up to the task, they are not actionable therapeutically, despite evidence suggesting that they lead to changes in clinical practice, the editorial continued. Available medications have modest but limited benefits for cognition and function but are ineffective for neuropsychiatric symptoms, which are also associated with caregiver burden.

“With a humble approach to lifelong learning, we hope more options for preventing, diagnosing, treating, and caring for Alzheimer's dementia will be available in primary care physicians' medicine bags a decade from now,” the editorial concluded.

For more on new technology and biomarkers in Alzheimer's, read the story from the November/December 2019 ACP Internist. For more on new technologies to scan for the biomarkers of mild cognitive impairment, read the story from the July/August 2019 ACP Internist.