Alzheimer's therapeutics: Continued clinical failures question the validity of the amyloid hypothesis—but what lies beyond?

Abstract

The worldwide incidence of Alzheimer's disease (AD) is increasing with estimates that 115 million individuals will have AD by 2050, creating an unsustainable healthcare challenge due to a lack of effective treatment options highlighted by multiple clinical failures of agents designed to reduce the brain amyloid burden considered synonymous with the disease.

The amyloid hypothesis that has been the overarching focus of AD research efforts for more than two decades has been questioned in terms of its causality but has not been unequivocally disproven despite multiple clinical failures, This is due to issues related to the quality of compounds advanced to late stage clinical trials and the lack of validated biomarkers that allow the recruitment of AD patients into trials before they are at a sufficiently advanced stage in the disease where therapeutic intervention is deemed futile.

Pursuit of a linear, reductionistic amyloidocentric approach to AD research, which some have compared to a religious faith, has resulted in other, equally plausible but as yet unvalidated AD hypotheses being underfunded leading to a disastrous roadblock in the search for urgently needed AD therapeutics. Genetic evidence supporting amyloid causality in AD is reviewed in the context of the clinical failures, and progress in tau-based and alternative approaches to AD, where an evolving modus operandi in biomedical research fosters excessive optimism and a preoccupation with unproven, and often ephemeral, biomarker/genome-based approaches that override transparency, objectivity and data-driven decision making, resulting in low probability environments where data are subordinate to self propagating hypotheses.

Introduction

Alzheimer's disease (AD), the most common form of dementia in the elderly, affects 5.4 million patients in the US alone, with 1 in 8 individuals 65 or older suffering from the disorder. It is the fifth leading cause of death in this group [1]. Particularly sobering is that the number of patients with AD is increasing [2], [3], with predictions of 115 million being affected by 2050 (http://www.alz.co.uk/research/statistics, http://www.alz.org/downloads/facts_figures_2012.pdf) due to: the aging of the population; the continuing lack of progress in identifying effective treatment modalities; and the lack of predictive diagnostic techniques. In 2010, the worldwide prevalence of AD was estimated at approximately 36 million individuals with an annual societal cost of US$604 billion, numbers that probably underestimate AD incidence as approximately 50% of cases are undiagnosed, especially in the early stages of the disease.

At any given time, it has been estimated that 25% of all hospital patients, 65 years or older, have AD. In 2004, these individuals had 828 hospital stays per 1000 Medicare beneficiaries, compared to 266/1000 for patients without AD. Additionally, annualized Medicaid costs for patients over 65 years of age are nine times higher if the patient has AD ($8419 vs. $915). Indeed, provision of care for AD patients has the potential to bankrupt many western healthcare systems if effective treatments are not found [2], [3]. Since drugs to slow or reverse AD progression have been estimated to have a potential market in excess of $20 billion per year, they are a high priority for both the pharmaceutical and biotechnology industries.

The actual pathological process(es) underlying AD causality are thought to begin 20–25 years before overt clinical symptoms become apparent [4], [5], [6], [7] making accurate diagnosis of the disease at as early a stage as possible a critical priority. As current diagnostic tools based on behavioral outcomes [8] and biomarkers (brain volume and metabolism, neuronal loss, brain/CSF amyloid load, CSF tau) have yet to be convincingly validated [9] with AD diagnosis arguably only possible when the disease is well advanced, new diagnostic guidelines proposed in 2011 [10], [11], [12] are currently under validation.

While both familial and sporadic forms of AD have been identified [13], [14], the ability of the familial form, which represents less than 1% of total AD incidence [15], [16] to usefully inform the causality of the sporadic disease state, is debatable. Thus, despite large investments by both academia and the pharmaceutical industry over the past 25 years, the cause(s) of AD remains largely unknown.