These diagnostic criteria were intended to accelerate research in AD throughout its stages, but there was a risk of making a clinical use of these criteria without adequate validation of their predictive value. This was particularly true with amyloid PET imaging, which became available commercially outside of academic centers where ethical supervision and the use of carefully designed disclosure protocols are available. Fortunately, good collaborations between the FDA and companies distributing the amyloid ligands have reduced the risk of catastrophic reactions from misinterpreting a “positive amyloid scan” as a definitive diagnosis of AD (Grill et al, 2017). Furthermore, the use of amyloid PET imaging (or CSF β42 measurements) is now a standard practice in the enrolment of volunteers in many AD studies, and although there is generally no disclosure of individual results, it can be done if appropriate disclosure protocols are in place.
A number of biomarkers have been established for AD, including pathophysiological markers (amyloid PET, CSF β42), neurodenerative markers (glucose PET, tau PET, MRI, CSF tau), and others that may be potential markers of disease activity (including inflammatory CSF and PET markers).
The AD field has benefited from large scale observational studies such as the Alzheimer Disease Neuroimaging Initiative (ADNI; Weiner et al, 2017) with an observational study in subjects with no cognitive decline, mild cognitive impairment or early dementia due to AD. ADNI’s data sharing policy to investigators around the world has led to important findings such as the underlying pathology of AD being present long before dementia has clinically manifested. An example of the impact of ADNI is the demonstration of a synergistic effect of β-42 and tau pathologies in asymptomatic at-risk persons as well as persons with Mild Cognitive Impairment (Pascoal et al, 2017a;b).
These biomarkers may lead to preventive treatments that are personalized, e.g. treat the pathological factors most active for that individual at the current stage of his disease. The best example would be neuro-inflammation, which is silent but very much present in the brain of individuals in the asymptomatic stage of AD, now detectable by specific PET ligands and by CSF examination, and is potentially treatable with non-steroidal, anti-inflammatory drugs.
Genetic factors in AD include the rare autosomal dominant mutations (PS1, PS2, APP) causing early onset AD, and the common (approximately 15% of population) apolipoprotein E4 genotype which increases the risk for late onset AD around age 70. Homozygous (e4/e4) persons have an even higher risk by age 65, leading to clinical trials specific to that population. Since the prevalence of ApoE-e4/e4 is relatively rare (3% of the general population; 15% of patients with AD), novel recruitment strategies are under way, such as registries where volunteers can mail a cheek-swab and agree in principle to be contacted if there is a preventive study appropriate with their risk status. Careful ethical supervision of these registries, data sharing, disclosure of genetic results is supervised by Ethical, Legal, Social Impact (ELSI) committees such as the one supported by the Canadian Institutes for Heath Research through the Canadian Consortium on Neurodegeneration in Aging.
For persons carrying PS1, PS2 or APP mutations causing early onset AD, specific research programs are underway, such as the Dominantly Inherited Alzheimer Network (DIAN) led by Ronald Bateman (Bateman et al, 2012) and the Columbia Registry (Rios-Romenets et al, 2016), both leading to randomized clinical trials with anti-amyloid drugs.
Finally, the special case of patients with Down’s syndrome (trisomy 21) should be mentioned. Since the amyloid precursor protein is located on chromosome 21 resulting in excessive amyloid production, Down’s syndrome is associated with symptomatic dementia which often occurs earlier in life (by the early to mid-50s) (Castro et al, 2017). There are studies under way with this population using anti-amyloid active immunization.
It is thus possible to imagine an individual assessment of risk for AD in later life and administer treatments proportional to their risk (Figure 5).