Advances in Alzheimer’s disease diagnosis have led to the characterization of a preclinical stage, opening the door for therapeutic interventions designed to delay or even prevent the onset of symptomatic cognitive decline. Advocates for such a "pre-emptive strike" cite the ineffectiveness in symptomatic patients of seemingly well-conceived therapeutic trials using advanced biologic agents that are able to clear amyloid from the brain and yet fail to halt or even significantly slow clinical progression. A few prevention trials to date using agents such as Ginkgo biloba have unfortunately been unsuccessful, but the new trials, planned to begin enrollment in 2014, will be using monoclonal antibody–based agents with a demonstrated ability to clear amyloid, and will be targeted at genetically high-risk, presymptomatic individuals, particularly those with pathogenic autosomal dominant mutations (Biomark. Med. 2010;4:3-14) .

Although it seems clear why a presenilin 1 carrier might be willing to undergo such therapy, there are others who advocate widespread dementia screening programs so as to improve patient outcomes using existing therapies with the logic that earlier intervention must be best. A study presented by Dr. Carol Brayne at this year’s Alzheimer’s Association International Conference noted that the evidence showing beneficial clinical or psychosocial outcomes for population-based dementia screening is lacking. Screening programs have traditionally taken the form of simple mental status–styled tests, but a more recent development that has already become a thriving business involves genetic screening. The Risk Evaluation and Education for Alzheimer’s Disease (REVEAL) study has shown that under carefully controlled conditions – including psychological screening, patient education, and follow-up – disclosure of apolipoprotein E allele status results can be done safely (N. Engl. J. Med. 2009;361:245-54).

However, the paradigm followed by most who wish to pursue genetic testing is quite different. Interested individuals can simply go onto the websites of companies that offer genetic testing, purchase a DNA test kit, and receive their genetic results in the mail. This is not at all the model used by REVEAL, yet it is a rapidly growing practice.

People want predictive tests whether there is an immediate tangible impact on their health or not, and this seems true for a variety of conditions including, but not limited to, Alzheimer’s disease (Health Econ. 2012;21:238-51). There are a variety of reasons why such testing may be deemed valuable by participants, including a general dislike for uncertainty, greater insight into one’s health risks, nonmedical decision making, future planning, and, of course, actual disease prevention. To what degree people would use this information for making healthy lifestyle choices is at best unclear, and at worst discouraging.

Those who propose parallels between cardiovascular disease and Alzheimer’s development reason that control of risk factors for Alzheimer’s may prevent or at least delay the onset of symptoms, much like the control of CV risk factors may prevent or delay atherosclerosis-related complications, such as heart attack, stroke, and peripheral vascular disease. The analogy is highly imperfect given the far greater correlation between CV risk factors with CV outcomes. Yet, in a cohort of 11,993 followed for more than 11 years, only 0.2% had all seven ideal metrics for CV health (smoking, body mass index, physical activity, healthy diet, total cholesterol, blood pressure, and fasting plasma glucose). Ideal diet, for example, was met in only 4.2% (Mayo Clin. Proc. 2012;87:944-52). The level of noncompliance with proven, indisputable risk factors for the leading cause of death should at least raise concern that the public’s desire to know correlates very imperfectly with optimizing healthy behaviors, the primary point of predictive testing.

Another possible outcome of screening programs would be the identification of individuals eligible for research programs. Preclinical Alzheimer’s is a research construct, not a clinical diagnosis, and is primarily based upon biomarker evidence of Alzheimer’s disease pathology in asymptomatic individuals. Group analyses leave little doubt that biomarker evidence of pathology correlates with future clinical outcomes (Ann. Neurol. 2012;71:765-75; Lancet Neurol. 2013 Sept. 4 [doi:10.1016/S1474-4422(13)70194-7]), but at the level of the individual (the clinical level), outcomes vary enough to make prognostication for any given patient unreliable. However, this should not invalidate preclinical Alzheimer’s as the research tool for which it was intended. Rather than declare a stalemate between those who advocate for and against preclinical testing, both sides undoubtedly agree that better therapies are needed, and preclinical Alzheimer’s, as a research construct, offers the opportunity to harness the current "free-for-all" state of genetic screening into a more thoughtfully designed and implemented program with the goal of pairing results with an action plan that, for some, could include participation in a therapeutic trial. What form such screening programs should take, how widespread they should be, and what rules should govern disclosure of results are all questions still in need of answers, but new research opportunities provide a compelling rationale to redesign rather than abandon screening programs.

Dr. Caselli is a professor of neurology and associate director and clinical core director of the Alzheimer’s Disease Center at the Mayo Clinic in Scottsdale, Ariz. He has no relevant financial disclosures.

Redesign needed for Alzheimer’s screening programs