Nature Biotechnology, August 2012.

 The first clinical trial of a drug’s ability to forestall Alzheimer’s disease in cognitively healthy individuals is preparing to launch.  Genenetch of S.San Francisco, the Banner Alzheimer’s Institute in Phoenix and the National Institutes of Health in Bethesda, Maryland will together spend $96 million in  an international effort testing Genentech’s experimental drug crenezumab, a humanized monoclonal antibody that binds the amyloid beta peptide (Aβ), in people certain to develop early Alzheimer’s disease due to autosomal dominant gene mutations. Two other drug clinical trials, one also targeting Aβ in presymptomatic familial Alzheimer’s disease and another in the general population, also aim to launch in 2013.  These pioneering studies will also test the amyloid hypothesis and multiple Alzheimer’s biomarkers and thus shape future drug development for a field desperate for a success.

  That’s because clinical trials of agents targeting Aβ have so far failed almost completely.  (Phase 3 results for Eli Lilly’s solanezumab and Pfizer and Elan’s bapinezumab were imminent as of late July.) The new approach will test drugs before or just after the onset of Alzheimer’s symptoms, thus theoretically improving the odds of success.  Fraught with their own risks, these trials may offer the best chance of making meaningful progress anytime soon against a disease that is projected to overwhelm the health systems of the developed world within a generation.

  One reason for the failure of Aβ-targeted drugs to date is that they may come too late in the disease course to make a difference.  Neuropathological studies reveal high brain amyloid burden even before Alzheimer’s symptoms develop.  “By that time there’s been considerable synapse loss, destruction, degeneration of neurons, [so] that it’s unlikely that the current disease-modifying therapies, those that go after amyloid or go after tau, are going to be hugely effective,” says Steven Paul, a former president of Lilly Research Laboratories in Indianapolis, Indiana who now directs an Alzheimer’s institute at Weill Cornell Medical College in New York, New York.  But a conventional prevention trial in Alzheimer’s disease would need to follow many thousands of people for a long period to show a drug effect.

  So academic researchers began planning smaller, shorter trials in populations certain (or very likely) to suffer Alzheimer’s disease within a relatively short time frame.  (Table 1)  Such drug trials have never before been attempted in any neurological disease, says Magali Haas, incoming chief science and technology officer for One Mind for Research, a nonprofit in Rutherford California focused on multidisciplinary approaches to brain diseases.  Researchers are counting on government funding and philanthropic support, and have recruited industry partners to provide the drugs.  Despite myriad risks, including the danger of serious side effects in cognitively normal individuals using investigational drugs, these trials have the support of regulators in the US and Europe, which in turn has convinced companies to participate.  “It’s pretty bold,” says Paul.  “But I think this is the way that one has to approach the disease.”

  The only trial of the three to secure complete funding by early July is an international trial sponsored by the Alzheimer’s Prevention Initiative (API), a collaboration organized by Alzheimer’s researchers Eric Reiman and Pierre Tariot of the Banner Alzheimer’s Institute. This randomized trial will enroll 300 members of an extended family in and around Medellin, Colombia, many of whom have a fully penetrant autosomal dominant mutation in the presenilin-1 (PS1) gene that typically triggers Alzheimer’s symptoms around age 50.(Some noncarrier controls and a small number of individuals in the US with varied mutations will also be included.) Neurologist Francisco Lopera of the Universidad de Antioquia in Medellín, who identified the family and has been following it for three decades, will head the Colombian team. The trial, using crenezumab, will last five years, with an interim analysis at 2 years to gauge the drug’s effect on a variety of Alzheimer’s biomarkers (Table 1). If one or more biomarkers move in response to treatment at two years, the trial will proceed.  “The biomarkers may be more sensitive than the cognitive measures,” explains Reiman.  The trial’s primary endpoint, though, is a positive difference in cognition.

    The second trial is sponsored by the Dominantly Inherited Alzheimer’s Network (DIAN), a research partnership funded by the National Institute on Aging (NIA) in Bethesda, Maryland.  DIAN proposes a five year, two-stage randomized trial in carriers of autosomal dominant mutations in PS1, PS2 and amyloid precursor protein (APP).  Three drugs, each employing a different anti-Aβ mechanism, will be tried in the first phase.  “This could be a window of opportunity to intervene therapeutically before there’s substantial brain damage,” says John Morris, a neurologist at Washington University in St. Louis and DIAN’s program director.

  The largest of the three trials is the “Anti-Amyloid Treatment in Asymptomatic AD,” or A4 trial.  Its 1,000 enrollees will be people in their 70s and early 80s in the general population selected on brain amyloid burden as measured by amyloid PET rather than by the presence of autosomal dominant mutations. But the randomized trial won’t be large enough to detect a difference in number of dementia cases that develop, just their relative latency.  “What we’re really powered to do, with 500 per arm, is to see a difference in the slope of decline, in the rate of decline,” says Reisa Sperling, a Harvard neurologist who heads the A4 trial.

  Taken together, these aggressive trials will test the validity of the amyloid cascade hypothesis, that Aβ deposition  leads to senile plaques, neurofibrillary tangles, neuronal cell death and dementia. And tracking biomarkers and clinical outcomes in parallel could enable future trials.    “We’re including all of the best established biomarkers, so that we can find out which ones might qualify for use as a reasonably likely surrogate endpoint in the accelerated approval of… future preclinical Alzheimer’s treatments,” says Reiman.  But clinical effect is the main goal of all these trials.

  Regulators seem prepared to approve drugs if they’re successful in a single trial, at least for API and DIAN.  Russell Katz, head of the FDA’s neurology products division, said in an April meeting that a change in a single cognitive test in DIAN could be enough for registration.  “We are open to that,” Katz said.  “We have not signed off on that formally, but have asked sponsors to flesh that out for us.  We recognize that very early on, when people are asymptomatic, the usual rules are tough to apply.” (A typical dementia trial would also require showing a change in daily function.)  In the case of A4, Sperling says that the trial by itself would probably not be enough for drug approval, but that the FDA and EMA have shown  flexibility in discussions.  “They’ve come an amazingly long way in thinking about how we could do this,” she says.

   First the drugs must work, hardly a sure thing.  Crucially, the drug mechanism must be right for the disease.  API selected crenezumab in part because it targets diverse forms of Aβ: monomers, oligomers and fibrils, and thus is capable of clearing free Aβ as well as plaques.  Crenezumab also depends in part on microglial engulfment of Aβ.  This mechanism differs from other anti- Aβ antibodies.For example, Lilly’s solanezumab only binds soluble Aβ and works by altering the balance between plaque and plasma Aβ. In theory, crenezumab should work better in early disease, but no one knows for sure.

  Also, safety is a major concern, because cognitively normal people will be taking these drugs.  Some anti-Aβ mAbs have caused vasogenic edema, the leakage of proteinaceous fluid from blood vessels, and microhemorrhages.  According to Reiman, as of early July more than 250 patients had received crenezumab with no reported cases of treatment-related vasogenic edema or microhemorrhage.  (The antibody was engineered with an IgG4 backbone to limit microglial activation, a likely contributor to vasogenic edema.)

  Other risks are political.  Familial Alzheimer’s disease may be more difficult to treat than sporadic disease, notes Ronald Petersen, director of the Mayo Alzheimer’s Disease Research Center in Rochester, Minnesota. So failure in API or DIAN might prematurely doom a drug that would otherwise work in the general population.  Conversely, success in early-onset Alzheimer’s does not mean that a drug will work in the much more common late-onset variety, Petersen says, since the familial form is caused by too much Aβ deposition, and the late-onset disease is more about deficient Aβ clearance.  “There likely are similarities, but it’s not the same illness,” agrees Morris.

  Joshua Boger, founder and former CEO of Vertex Pharmaceuticals in Cambridge, Massachusetts, worries that a drug effective in API or DIAN, though approved only for rare cases of familial Alzheimer’s, would cause “a real stampede” of off-label use in the general population.  “It could create quite a clamor for what would still be a very hypothetical and experimental treatment,” Boger says. But everyone’s first concern is finding a drug that’s effective. “We sure hope that we see a positive signal from these studies, because the field otherwise I think will have a hard time recovering,” says Morris. 

  Boger feels that all these risks are necessary.  “There’s a bigger risk in not doing anything,” he says, calling Alzheimer’s “a species emergency”:  “We can do the demographics–it’s ugly.  Trillions of dollars a year and crippling the entire medical system, and crippling families.   So this is a crisis.  Risk taking is prudent in a crisis.”

Table: Presymptomatic Alzheimer’s disease clinical trials