News and Analysis | Published:

Pharma pumps up anti-tau Alzheimer pipeline despite first Phase III failure

Nature Reviews Drug Discovery volume 15, pages 591592 (2016) | Download Citation

AbbVie, Bristol-Myers Squibb, Genentech, Janssen and Merck & Co. have all advanced anti-tau candidates into Phase I trials in the hunt for drugs that can modify the course of Alzheimer disease.

Image: Ikon Images/Alamy Stock Photo

Because disease-modifying candidates for Alzheimer disease (AD) have had a 100% failure rate in Phase III development to date, it came as little surprise that TauRx's candidate LMTM recently flopped in the first-ever pivotal trial of a tau-targeting therapy. The drug, a derivative of methylene blue developed to reduce the formation of the tau tangles that are one hallmark of AD, showed only a hint of possible efficacy in a small and controversial subgroup analysis presented at the Alzheimer's Association International Conference in July. But many researchers are ready to dismiss the drug's failure as a minor speed bump, unconvinced that the trial properly explored the role of tau in neurodegeneration.

“I need more evidence to throw out this mechanism”

“There are so many different reasons a drug could fail to hit its end points. It could be that the drug candidate wasn't good enough, it could be that the trial design was incorrect, or it could be that the mechanism is wrong,” says James Hendrix, Director of Global Science Initiatives at the Alzheimer's Association. “But I need more evidence to throw out this mechanism.”

Industry researchers agree. “The results of this trial don't change our enthusiasm for going after tau,” says Sean Smith, Director of Neuroscience at Merck & Co.

Fuelled by the lack of success with candidates that target the formation of the amyloid-β plaques, as well as the emergence of tau-imaging tools that can track the accumulation of tau in living brains, the anti-tau pipeline is growing. Big pharmaceutical companies have advanced at least five anti-tau therapies into Phase I trials in recent years, experimenting with an assortment of vaccines, antibodies and small molecules.

Anti-tau alternatives

Tau is a microtubule-associated protein that forms tangles inside neurons in several neurodegenerative diseases, including AD. Whereas the field was once divided over whether tau or amyloid-β species were the more important driver of neuronal death in AD, it is increasingly embracing the idea that the two might interact in a toxic pas de deux (Nat. Rev. Neurosci. 12, 67–72; 2011). Because tau exists in multiple intracellular and extracellular forms, traffics through the brain in a predictable manner and needs to be hyperphosphorylated before it can form tangles, drug developers are investigating a range of different tau-targeting strategies.

One of the most advanced of these approaches are vaccines that can induce immune responses to help patients clear their own tau. Janssen and AC Immune's ACI-35 raises an antibody response to a phosphorylated form of tau, and first entered into Phase I in 2013. Results are expected next June. Axon Neuroscience recently advanced its AADvac1 into a Phase II trial, having reported that their candidate, designed to elicit a response to pathologically modified forms of tau, was safe in a Phase I trial.

A potential strong point of vaccines is that they could provide long-term protection. And given that tau tangles also start to accumulate years before the diagnosis of disease, cost-effective preventive approaches offer a clear appeal.

Another approach is to focus on therapeutic antibodies, which can be more precisely fine-tuned to target different forms of tau and to control the subsequent immune response.

In 2014, Bristol-Myers Squibb advanced its anti-tau antibody BMS-986168 into a first Phase I trial in healthy volunteers. The antibody, acquired from the biotech iPierian, was discovered by using induced pluripotent stem cells derived from AD patients (Nat. Rev. Drug Discov. 14, 589–591; 2015). BMS-986168 targets extracellular N-terminally fragmented forms of tau.

The company has since launched another Phase I trial testing the antibody in patients with progressive supranuclear palsy (PSP). PSP is a 'pure' neurodegenerative tauopathy that is not accompanied by amyloid aggregation, and is considered to be a clean way to test the efficacy of anti-tau antibodies in a less complex disease setting. The trial is due to complete in the middle of next year.

AbbVie, similarly, started testing its anti-tau antibody C2N-8E12 in the summer of 2015 in a Phase I trial in PSP. Its antibody, licensed from C2N Diagnostics, targets aggregated, extracellular tau.

And this July, Genentech announced that it had advanced an anti-tau antibody into a first Phase I trial. The trial will test the antibody RO7105705, developed with AC Immune, in patients with mild to moderate AD and in healthy volunteers.

The use of antibodies against central nervous system (CNS) targets, as always, raises a specific challenge: getting enough antibody to cross the blood–brain barrier to exert a therapeutic effect. In the case of tau, this is further complicated by the fact that toxic tangles are intracellular, typically beyond the reach of antibody therapies.

Genentech's Gai Ayalon, however, thinks that access problems are manageable. His recent work suggests that, contrary to previous reports, anti-tau antibodies bind to extracellular tau, as it is trafficked between neurons, rather than to intracellular tangled tau (Cell Rep. 16, 1–11; 2016). “Anti-tau antibodies don't enter neurons and they don't bind intracellular tau. We've invested a lot of careful rigorous work to try and understand this and I hope that the field will agree that we can put to rest that question,” says Ayalon. Because extracellular tau levels in the CNS are thought to be relatively low, he adds, even poorly penetrant anti-tau antibodies may get past the blood–brain barrier in high enough quantities to mop up significant amounts of the trafficking tau that might otherwise seed the accumulation of further tau tangles. This is different from the situation with amyloid-β, he adds, in which antibodies act on brains that are riddled with extracellular amyloid-β.

This same paper also found that an antibody can deplete tau even when its effector function — which activates subsequent immune cell responses in the brain — is turned off. (Effector function was turned off by mutating the antibody's Fc region.) “This was an exciting and surprising finding,” says Ayalon, because it suggests that anti-tau antibodies need not induce chronic inflammation in already inflamed AD brains.

Merck & Co. is taking a different approach, working with partners Alectos to develop a small-molecule-based indirect attack on tau. Because tangle formation depends on the hyperphosphorylation of tau, their strategy is to control the post-translational processing of tau with the sugar tag O-linked N-acetylglucosamine (O-GlcNAc). O-GlcNAc is physiologically added to tau at the same residues at which hyperphosphorylation occurs. By inhibiting protein O-GlcNAcase (OGA), the enzyme that cleaves O-GlcNAc from post-translationally modified proteins, their plan is to leave the tag in place, prevent the hyperphosphorylation of tau and thereby reduce the development of tangles.

The partners have completed a Phase I trial of their OGA inhibitor, MK-8719, in healthy subjects. Joe Herring, Executive Director of Clinical Neuroscience at Merck, says the next steps will include Phase II trials in PSP. Success in that indication could pave the way for further research in AD and other tauopathies.

Several companies have worked to develop small-molecule drugs against glycogen synthase kinase 3 (GSK3), which phosphorylates a number of the hyperphosphorylated tau sites, as an alternative means of modulating tau's post-translational modification. Noscira made it as far as a failed Phase II trial in 2012 with its GSK3 inhibitor, and subsequently went into liquidation. Other companies, including Sanofi and AstraZeneca, have discontinued projects against this target.

Picking patients

As with anti-amyloid treatments, researchers still need to figure out which AD patients might benefit from an anti-tau treatment. Repeated failures in Phase III trials of anti-amyloids have pushed trials into earlier and earlier stages of disease (Nat. Rev. Drug Discov. 11, 657–660; 2012). Amyloid starts building up around 20 years before AD is diagnosed, and brain levels of amyloid are close to being maxed out by the time cognitive decline begins. Anti-amyloid treatments, the theory goes, therefore need to be started years if not decades before the onset of cognitive decline.

Tau might offer a different treatment window because its accumulation tracks more closely with cognitive and functional decline. Tau levels are still on the rise at the time of disease diagnosis. “That makes tau a more attractive target,” says Keith Fargo, Director of Scientific Programs at the Alzheimer's Association. “By the time you see somebody with a headful of amyloid it may be too late to slow down the amyloid accumulation, but by going after tau in these patients we may still be able to stave off dementia symptoms.”

At the same time, cautions Ayalon, anti-tau drugs (and anti-amyloids, for that matter) aren't designed to bring dead neurons back to life; at best, they might slow the decline of ailing axons. “We have to be realistic. As patients progress through the disease, the therapeutic effect of drugs is likely to decline,” says Ayalon.

With most anti-tau candidates still only in Phase I trials, companies have not yet disclosed where they think the sweet spot is for mid- and late-stage trials of these drugs. But all options are open. Randall Bateman, of Washington University in St Louis, USA, says that he is considering adding an anti-tau therapeutic arm to the DIAN-TU trial, which is testing drug candidates as preventive therapies in a population of largely asymptomatic people at high risk of early-onset disease.

“I'm very much looking forward to seeing what the future holds with all these approaches,” says Ayalon. “I hope our hypotheses for tau pan out.”

About this article

Publication history




  1. Search for Asher Mullard in:

Further reading

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing