With the demographic shift toward an increasingly elderly population, it has been predicted that the number of people afflicted with dementia will triple by 2050, with the cost to the healthcare system estimated to be $1.1 trillion.

The complex neuropathology underlying Alzheimer’s disease (AD) is only beginning to be understood, and we have only scratched the surface in investigating the role that each underlying biology plays in propagating disease.

This complexity can be used as a weapon against disease by combining drugs of different biologies with the hope that a multipronged therapeutic attack will add up to a cure.

To consider the players in such combinations, let us review what we know today.

When Alois Alzheimer first described amyloid plaques and neurofibrillary tangles as cardinal features of the disease, it brought about a revolution in perception; dementia went from being a social stigma or sign of weak character to a physical disease of the brain that might someday be cured through development of new medicines.

It took nearly 80 years for scientists to “take apart” amyloid plaques and neurofibrillary tangles and discover compositions principally of beta-amyloid peptide (βAP or Aβ) and the microtubule-associated protein tau, respectively.

Coincident with these key discoveries, a curious schism emerged: The research community began dividing into camps that favored βAP or tau as principal causal agents in AD.

This schism became so pronounced in the 1990s that it was whimsically dubbed the “βAPtist/Tauist war.”

While science thrives on the pursuit of competing hypotheses, the βAPtist/Tauist war had an unfortunate consequence: Research seeking a more integrated understanding of AD pathophysiology was discouraged, and clinical approaches to treatment became narrowly focused.

Large-effect genes focused attention on βAP

The proponents of βAP hypothesized that a generation of βAP from a longer protein called the amyloid precursor protein (APP) is a seminal event in the development of AD.

This view was strongly boosted by the discovery that a sizable number of genetic mutations in APP and in γ-secretase – one of the enzymes that liberates βAP from APP – invariably led to an early-onset form of AD.

Enthusiasm built for the idea that slowing the rate at which βAP was liberated from APP or preventing βAP from forming aggregates believed to be toxic in the brain would prevent or delay AD from developing, and likely mitigate symptoms in AD patients.

The “amyloid hypothesis” has now been pursued in the clinic over the past 2 decades through the design of inhibitors for the APP cleaving enzymes β-secretase and γ-secretase, which are responsible for βAP generation.

Another therapeutic approach has been to design antibodies that bind to βAP or aggregates of βAP to accelerate their removal from the brain or decrease their toxicity.

The clinical trial results to date for treatments aimed at the amyloid hypothesis are disappointing at best.

First generation γ-secretase inhibitors were amenable to design, but they were plagued by safety issues. Centrally penetrant β-secretase inhibitors have been much more difficult to design and are only now entering advanced clinical trials.

Multiple antibodies against βAP have failed to meet their primary clinical endpoints. Whether they were tested too late in the disease or they just didn’t work is not clear.

So while it is premature to write an epitaph for the “amyloid hypothesis,” the clinical record to date suggests we need to be casting a wider net in the battle against AD.

Smaller-effect genes reveal the complexity of Alzheimer’s

Genes affecting the generation of βAP from APP are rare, but they increase one’s chances of getting early-onset AD to near certainty.

Over the years, other genes and gene families impacting the risk of AD have emerged. These gene variants tend to have a much broader distribution in the general population, but their impact on increasing one’s lifetime risk of developing AD is lower than for the so-called familial genes.

Consequently, it has often been necessary to study large populations – as opposed to individual families – to identify these genetic risk factors.

The oldest and best known of these “smaller-effect” genes is a specific variant of apolipoprotein E, known as ApoE4. Possessing a single ApoE4 gene increases one’s lifetime risk of developing AD by three-fold in Caucasians.

While apoE studies by AD researchers have focused on its potential role as a βAP transport protein, this emphasis likely reflects the “βAPtist bias.”

ApoE has been studied more broadly for its causal role in vascular/metabolic disease. Viewed from this perspective, one needs to look no further than the APP gene that sparked the βAPtist movement to find AD’s vascular connections: Mutations within βAP lead to vascular diseases of the brain, and the best established function for one variant of APP is as a regulator of the blood coagulation cascade.

These and other findings in recent years have led AD to be viewed as a vascular/metabolic disease, even prompting AD to be referred to as “type 3 diabetes.”

The role of innate immune system genes in AD

Genome-wide association studies approach the genetics of AD from the opposite end of the spectrum as single family studies. These studies explore the full genome of very large numbers of individuals to look for genes contributing to risk of disease.

Such studies have identified novel genes and gene families that underlie the pathophysiology of AD. Among the more prominent of these players are genes of the innate immune system.

Studies that explore changes in transcriptional activity across the entire genome also implicate the innate immune system in AD.

One interesting example of note is TREM2, an innate immune gene that – similar to ApoE4 – increases lifetime risk of developing AD by three-fold in certain populations. TREM2 genetics extend the curious vascular connection of AD in a way similar to APP: While certain mutations increase risk of AD, other mutations lead to Nasu-Hakola disease – a vascular dementia.

Circling back to tau, there is still no identified tau genetic mutation leading to AD. While this fact has given tau a decided disadvantage in the βAPtist/Tauist war, other properties of tau pathology clearly implicate tau in dementia.

First, there are genetic mutations in tau known to lead to non-AD dementia. As we’ve seen with the TREM2 and APP examples, such mutations offer important clues about AD biology even if they do not specifically lead to AD.

Second, pathologic changes in tau are associated with a wide variety of central nervous system(CNS) disorders, collectively called tauopathies.

Third, the appearance of tau-related pathology in AD correlates much better with the onset of dementia than does the appearance of amyloid plaques (which can precede clinical AD by decades). Thus, the absence of a direct genetic link between tau and AD is a poor argument for de-emphasizing the potential role of tau in the pathogenesis of AD.

An integrated approach to AD is emerging

The recent string of clinical trial failures in AD will teach us little if they are used only to resurrect old βAPtist/Tauist rivalries. Rather, the emerging science reminds us that AD is a complicated disease with multiple stages of development.

Researchers are likely to learn much more about the biology of AD by investigating the common links among pathologies implicated in AD rather than studying those pathologies in isolation.

Four biologies ripe for investigating these common links in AD are βAP/amyloid pathology, tau/neurofibrillary tangle pathology, vascular/metabolic dysregulation, and innate immune dysregulation/neuroinflammation. Just a few of the many known intersection points for these biologies are mentioned here.

Putting all the biological puzzle pieces together will take time. Unfortunately, the AD epidemic facing the aging baby boomer generation is fast approaching with no time to spare.

In the absence of an integrated understanding of how AD develops and what treatments may work best at different disease stages, it is important, like in other complex diseases, to consider strategies such as combination therapy sooner than we might traditionally pursue these.

Apart from the many βAP-focused treatments in later stage clinical development for AD, TRx-0237 is a tau-focused compound in development.

A large number of tau-directed monoclonal antibodies are also in preclinical development. Agents having a vascular/metabolic disease focus include intranasal insulin, the PPARγ agonist pioglitazone, and the calcium channel blocker nilvadipine.

Treatments targeting the immune system include intravenous immunoglobulin and the RAGE antagonist TTP-488.

These agents and others that are more comprehensively reviewed elsewhere, together with symptomatic approaches such as the 5HT6 antagonist idalopirdine – an investigational compound designed to improve neurocognitive function – are possible contenders for combinations.

To make combination therapy in AD a reality, it will – to borrow a phrase – “take a village” to make it happen.

Leaders from industry, the Food and Drug Administration (FDA), the National Institutes of Health (NIH), academia, and patient groups will need to come together to enable funding, trial design, and solve a host of complex issues associated with delivering combination therapy.

But there is hope. Such public-private partnerships like the Global Alzheimer’s Platform and others have already begun to take up this important challenge.