Man is not made for defeat.
Ernest Hemingway, The Old Man and The Sea
THEY HAVE BEEN called many things–‘spider cells’, ‘little bags of poison’, the ‘other brain’1–but they are officially known as glia. Alois Alzheimer himself knew about glia; under the microscope they looked like scars bordering plaques and dead neurons. But like other scientists of his generation, he considered them little more than structural filler, and so for nearly a century they were overlooked. About thirty years ago, however, when scientists realised that glia constitute over half the human brain, they decided to take a closer look.
There are three types of glia.
Astrocytes: Greek for ‘star cells’ due to their shape, astrocytes are the largest, most numerous type. They control brain functions by mediating how neurons ‘talk’ to one another. In the hippocampus, for example, a single astrocyte contacts up to 140,000 neuronal synapses. In The Other Brain, astrocyte expert Douglas Fields argues that this behaviour is more complex than neurotransmission, meaning that higher mental faculties like consciousness, thoughts and feelings may actually be governed by astrocytes. They divide and die like other cells, and can grow uncontrollably in the most lethal type of brain cancer–glioblastoma.
Oligodendrocytes: Greek for ‘few branched cells’, they are the cellular factories of myelin, a fatty substance that insulates neurons by wrapping around their axons like the plastic sheath of copper wire. Myelin is white, hence the term ‘white matter’ as opposed to ‘grey matter’, referring to neurons themselves. A human’s need for oligodendrocytes is darkly illustrated by multiple sclerosis, the devastating and common neurological condition caused by the widespread destruction of myelin. Without myelin, nerve impulses are disrupted, leading to fatigue, muscle weakness, visual problems and cognitive dysfunction.
And then there’s microglia, literally ‘small glia’, the third and most important type for our story; these are the brain’s immune cells. Swarms of these comparatively tiny cells orbit neurons in a surveillance state, constantly scanning them for signs of distress using long, antennae-like projections. In this mode they’re dubbed ‘resting’ microglia. Once a threat is found, however, they transform from guard to soldier-at-arms. These ‘activated’ microglia then unload a payload of toxic chemicals to rid the brain of unwelcome guests such as meningitis and malaria.
This is important because in the late 1980s post-mortems revealed that Alzheimer plaques were often completely surrounded by microglia.2 At the time no one quite knew what to make of this. But by 2001 advances in neuroimaging had made it possible to see activated microglia in living brains. In healthy people, the images nearly always showed a dim glow of activity across the brain. In Alzheimer’s patients, though, the brain was lit up like a Christmas tree.
At face value it looked like a classic immune response–as if the microglia were attacking the plaques in a bid to eliminate them from the brain. The idea that the immune system could be involved in this way had radical and conflicting implications. On the one hand it suggested the brain was trying to remedy the problem from within; that we had, in effect, an ally on the inside. From this came the idea that perhaps the microglia just needed a helping hand, and that by artificially ramping up their healing power, the brain’s immune system could be harnessed to treat Alzheimer’s.
But this rose-tinted outlook was counterbalanced by something far less optimistic. Mounting evidence from cell culture research showed that microglia also appeared to kill neurons if their activation wasn’t controlled.3 I’ve done this experiment myself: leave microglia in a dish with neurons and a low dose of immune stimulant–fragments of bacteria or dead cells, for instance–and they will eventually turn on their neuronal neighbours.4 Through the microscope the neurons look like a satellite image of a city at night, with the lights getting dimmer as clusters of circular bodies slowly blot them out.
This fact was once the bane of my existence. For two years I tested whether an experimental drug could hold off the microglia and allow the neurons to flourish. Every morning, including weekends, I entered the university, slipped on a white lab coat and blue rubber gloves, doused myself in sterilising ethanol, and had a look at how my cells were doing. What came next was usually the sight of dead neurons, an abrupt expletive (the standard response of a young scientist) and a phone call to cancel whatever social obligation I had planned. Others continue to work on the drug, but it’s still unclear why the microglia do this to the neurons–whether the response is deliberate or just collateral damage.
In any case, the evidence suggested that the internal chaos of plaques and tangles might cause microglia to become dangerously overactive. These sabotaging cells could then start a deadly, self-perpetuating cycle of toxic inflammation, besieging the brain and driving the disease into a downward spiral. If that was the case, scientists had to wonder: would powering down the immune system help Alzheimer’s patients?
Both scenarios were theoretical, of course, and arguably too simplistic. Microglia might be both good and bad for the brain. The answer depended on several unknowns, such as what state they were in when the disease started, how long they had been activated, and the role of genetic and environmental influences. But there was, unfortunately, precedent for the darker alternative. And two other types of dementia exemplify that point.
The first is neuroAIDS. In 1983, one year after the US Center for Disease Control and Prevention established the term AIDS to describe the opportunistic infection affecting young homosexual males, it was noticed that some AIDS patients also developed nervous system abnormalities similar to those seen in Alzheimer’s. After experiencing a severe decline in memory, concentration, attention and language, they would end up bedbound and incontinent, usually dying three to six months later. It was often the earliest and sometimes only indication of a patient suffering from HIV. By 1987 researchers introduced the term ‘AIDS dementia complex’ to highlight the virus’s impact on cognition, but it is now most commonly referred to simply as neuroAIDS. Although the anti-retroviral therapies released in the 1990s did appear to ameliorate neuroAIDS symptoms for some patients, exactly what combination of drugs was best suited for the purpose remained unclear. Today it’s estimated that 10–25 per cent of HIV-infected patients develop this kind of dementia.5
But from a purely scientific standpoint neuroAIDS provides a powerful lens for pinpointing what brain changes are perpetuating the symptoms of Alzheimer’s. NeuroAIDS is a disease with a known primary cause–a virus–and this gives scientists an opportune place to anchor their thoughts, because viruses are fixed entities in space and time and abide within well-defined parameters. Exploring this, scientists soon learned that after entering the brain the first cell type the HIV virus infects is microglia; neurons are spared until much later in the disease. That suggests that microglia, not neurons, are the core perpetrators of the symptoms of dementia.
Another form of dementia that gives a clue is called Nasu-Hakola, named after the Japanese pathologist and Finnish physician who first reported it in the early 1970s. This condition remains puzzling to this day. A patient with Nasu-Hakola first experiences severe bone fractures, usually during adolescence, in their hands, feet and knees. This can persist despite bone transplantation. Bizarrely, the young patient then develops a slowly progressing dementia, involving memory loss, personality changes, indifference and apathy towards those around them, problems with speech, and disorientation. Again, strikingly similar to Alzheimer’s. By the mid-1980s researchers spotted that most cases were either of Japanese or Finnish descent, suggesting a genetic cause. And by 2000 geneticists had zeroed in on two culprit DNA mutations in the genes DAP12 and TREM2, both of which code for receptors on–you guessed it–microglia.
Suddenly microglia were in the spotlight. Experiments focusing on what their normal functions were and how mutations affected them were devised. Comparisons between Nasu-Hakola microglia and microglia in Alzheimer’s were drawn. Therapeutic strategies to rein in their suspected overzealous behaviour were discussed. Scientists still weren’t sure if they were allies or saboteurs; many ominously called them the brain’s Jekyll and Hyde.
So when Dale Schenk, at a new company in San Francisco called Athena Neurosciences, wanted to empower microglia for an immune-based therapy on Alzheimer’s patients, it was, to say the least, a bold move.
‘I just had a simple idea,’ Schenk told me over the sound of stirring teaspoons and soft chatter. Scientists don’t usually meet under golden chandeliers and marble columns, but I’d managed to catch fifty-eight-year-old Schenk at a biotechnology conference being held at New York’s illustrious Waldorf Astoria Hotel. His ‘simple idea’ was one so innovative that many scientists admitted they would never have thought of it. A vaccine for Alzheimer’s.
We normally associate the term ‘vaccine’ with viruses and bacteria. Outbreaks of bird flu, Ebola and now Zika saturate the press with stories about the race to invent vaccines. But in pure biological terms, vaccines are any kind of agent that stimulates an organism to develop immunity. They can be dead or weakened forms of the threat itself–like Jonas Salk’s polio vaccine and GlaxoSmithKline’s chickenpox vaccine–or antibodies: blood cell proteins that label pathogens for destruction.
Schenk’s vaccine for Alzheimer’s was called AN-1792, and it consisted of synthetic beta-amyloid. His aim was to trick the brain into thinking that the plaques themselves were the foreign invaders, and thus stimulate a potent immune response. ‘I thought, if we vaccinate mice with beta-amyloid,’ he said, adjusting his glasses, ‘they’re going to develop antibodies to beta-amyloid and have them circulating in their blood, right? And a small fraction of those antibodies are going to get into the brain. Over time, that should disrupt beta-amyloid and dissolve the plaques.’ Schenk grew up in Pasadena, California. The son of a fire chief and newspaper columnist, he went into science because ‘it just seemed like a good idea’. After his PhD he worked on heart disease for a company called California Biotech. Then, one day in the mid-1990s, tired and frankly bored of the heart, he got talking to a colleague working on John Hardy’s amyloid hypothesis.
‘So what does beta-amyloid actually do?’ Schenk remembered asking him.
‘Oh, I don’t know,’ said his colleague.
‘What do you mean you don’t know?’ urged Schenk.
‘No one knows what it does, they just think it might cause Alzheimer’s.’
‘Well that’s stupid. All it does is stick together. How could it cause Alzheimer’s?’
The conversation continued in this vein until Schenk, finally convinced by the hypothesis, decided to join his colleague at Athena Neurosciences, where Schenk has now worked for the past twenty-eight years.
Perched on the edge of the San Francisco Bay, flanked by the brown sun-scorched slopes of San Bruno Park and Sweeney Ridge, Athena was the slick new contender that everyone thought would outsmart Alzheimer’s. It had some of the brightest minds in the field–including Dora Games, inventor of the Alzheimer’s mouse that provided support for John Hardy’s amyloid hypothesis. With the amyloid hypothesis as its foundation, the company made a list of all the therapies it wanted to test in the mice.
Schenk’s idea was low on that list. ‘Nobody wanted to do it,’ he said, laughing. ‘We had a list of thirty-three therapies we wanted to try on the mice and it was listed as number thirty-two. I couldn’t even get any animals to do the experiment.’ One of his colleagues thought the idea was so absurd that he put it on a list of bad ideas, which he stuck in the lab for all to see. As further ridicule, he gave Schenk the ‘broken clock’ award–even a broken clock is right twice a day–to which Schenk would retort, ‘Well, at least it’s exactly right twice a day.’
But with patience and dogged persistence, Schenk’s time finally came. Managing to obtain a group of mice left over from somebody else’s work, he performed the immunisation, sacrificed the mice, and sent their brains off for analysis. Then he waited. And waited–his gamble was hardly considered urgent–until, months later, he got a phone call from Dora Games herself. ‘You’re not going to believe this,’ she said. ‘But we’re—’
‘Not seeing any plaques in the mice that are vaccinated, right?’
So how did it work? I’ve spent years culturing microglia and can honestly tell you that we really don’t know. The more we learn about microglia, the more complex their character turns out to be. Nonetheless, microglia tactics are basically twofold. They can secrete chemicals that kill parasites by degrading the parasite’s DNA, or physically engulf the intruder in a process called phagocytosis (Greek, ‘to devour’). Of course, Athena’s neuroscientists didn’t care how the patient’s microglia chose to eviscerate the plaques, so long as they did it safely and stopped Alzheimer’s as a result.
This now started to feel like an achievable feat. Between 1997 and 2000 vaccinations in rabbits, guinea pigs and monkeys all supported Schenk’s discovery.6 The vaccination even appeared to improve animal cognition. With that, the pharmaceutical companies Élan and Wyeth began human trials.
Human trials, also known as clinical trials, are typically split into four phases. Phases one and two are all about safety. In phase one the drug is given to a small group of twenty to eighty people to determine a safe range of doses. If no adverse effects are spotted the drug advances to phase two, where hundreds of people are treated and tests are performed to see if it has any biological effect. Most drugs, unfortunately, do not make it past phase two.7 For those that do, phases three and four represent the peak of the mountain. In phase three thousands of people are tested over the course of several years; side effects are monitored; and if the drug appears to have worked it’s then marketed and approved by agencies such as the FDA or EMA (European Medicines Agency). Phase four is a kind of follow-up to see how the drug is faring in the general population, and if there are any long-term effects that weren’t detected earlier.
In a small phase one clinical trial of just twenty-four patients, a single dose of Schenk’s vaccine appeared to be safe. Multiple doses were then given to more than seventy patients. Again, no adverse effects. Confident it was safe and worth pursuing, Élan and Wyeth moved to phase two, where 300 patients received the vaccine.
Scientists the world over were on tenterhooks. This was the first real-world test of the amyloid cascade hypothesis.
It was a disaster. Seventeen patients developed a dangerous form of brain inflammation called encephalitis, causing confusion, fever, and, if left untreated, seizures, stroke and death. In January 2002 AN-1792 testing was immediately aborted. After all the early promise it was a severe blow. And a hard lesson.
Schenk was stunned. ‘Because we didn’t see anything in the animals…’ he lamented to me in a quiet, solemn tone, ‘but it may have come and gone without us being able to detect it.’ Determined the tragedy wouldn’t be in vain, scientists around the world began an extensive follow-up of every patient from the trial. The findings were a mixture of encouraging, disappointing and just plain weird: encouraging because the first post-mortem showed a brain almost entirely clear of plaques; disappointing because Sid Gilman–Élan’s elected expert to chair the trial’s safety-monitoring committee–reported that of the 300 patients only 59 actually mounted an immune response; and plain weird because some patients did show a flicker of improved memory, even though a subsequent MRI scan showed that their brains had actually shrunk. How could a shrinking brain be found alongside increased cognition? ‘We still don’t know,’ said Schenk. ‘We may never know.’ Only one thing was unanimously agreed: to watch every vaccinated patient like a hawk.
Four years later it was revealed that 159 patients did eventually show some improvements in cognition. Schenk’s idea was alive again, and so if scientists could overcome the side effects then the makings of an effective treatment were in sight. Fortunately someone had already stumbled across a possible solution. Shortly after the trial ended, a Swiss psychiatrist named Christopher Hock found that patients who had made antibodies to beta-amyloid did better on tests of language, attention, memory and self-care than those whose immune system had not made antibodies.8 Switching tactics, Schenk devised an antibody-based vaccine: Bapineuzumab, or ‘Bapi’, was an antibody made in mice but artificially tweaked for humans. The genius of this strategy was that patients would not have to mount a full-blown immune response because the antibodies to the plaques would already be present. It was less aggressive, and thereby reduced the risk of encephalitis.
In 2006 clinical trials were under way and the only side effect was a little water in the brain–cerebral oedema, quickly remedied by lowering the dose. By December 2007 the drug made it to phase three and more than 2,000 patients spanning North America and Europe, aged fifty to eighty-five, were enrolled.
It was the largest, most audacious attempt to combat dementia. Alois Alzheimer could only describe what he saw, William Summers could only delay the symptoms of the disease, but now Dale Schenk might be able to prevent the disease from taking hold. As the trial got under way, US pharmaceutical giants Pfizer and Johnson & Johnson stepped in, investing hundreds of millions of dollars in Bapi–the silver bullet to end an epidemic.
But it too failed. In August 2012 the results were in and Bapi showed no therapeutic benefit in all phase three trials, its effect on memory no better than a placebo. With the financial cost of the failure staggeringly high, Johnson & Johnson and Pfizer swiftly halted development.
There was widespread doubt over the amyloid hypothesis. Even early, highly influential supporters like Zaven Khachaturian–director of the NIA and recruiter of amyloid mastermind George Glenner–began to express reservations. ‘The amyloid hypothesis became such a strong scientific orthodoxy that it began to be accepted on the basis of faith rather than evidence,’9 he told one reporter, adding, ‘no one has stepped back to ask whether our basic premise about the disease is the correct one.’
Schenk had his own doubts, but there were three gaping holes in the trial’s design. First, how did we know these patients really had Alzheimer’s? Post-mortem was still the only way to know definitively; perhaps they had another kind of dementia. Second, there was still no way of rigorously separating early from mid-stage cases; perhaps that skewed the overall result. And third, John Hardy and many other Baptists all agreed that the dose was too low; risks aside, there could be no kid gloves in this fight.
‘We couldn’t figure out who had the disease. We couldn’t separate mild from moderate. And we were limited to low doses,’ said Schenk. Another drawback, he explained, was the trial’s benchmark for success. Because the activities of daily living vary so much between people–some people might do crosswords all the time, for instance, while others read or sew instead–the FDA demanded that trial candidates score highly on two different sets of cognitive tests, instead of the usual one. This set the bar for a positive result very high. ‘So the sad part is that this drug probably would’ve worked for some people.’
Schenk’s criticisms were soon backed up when scientists re-examined the drug using a powerful new technology. Pioneered at the University of Pittsburgh, Pennsylvania, ‘PiB’ (or Pittsburgh compound B) was a radioactive dye that bound to beta-amyloid in living subjects. Combined with conventional brain scans, PiB could actually show the spread of amyloid in the brain.10 It was also a strong diagnostic tool: APOE4 carriers were more likely to show a bright PiB signal, as were carriers of APP mutations. But to the amazement of those involved in the Bapi trials, 30 per cent of the recipients were PiB-negative. They didn’t have Alzheimer’s. They had been misdiagnosed.
And so the flaws of the trials, along with the ambiguous nature of the results, gave researchers enough reason to continue pursuing vaccine therapy. They’re still working on it today.
It was at this point that I had to ask: what did Schenk think caused Alzheimer’s? He paused and breathed a heavy sigh. ‘I don’t think there’s a single cause. I think it’s like heart disease, and having plaques and tangles is like saying you have clogged arteries. You can have clogged arteries without having heart disease. And even though everybody views me as a Baptist, it’s just that amyloid’s been the most treatable drug target. That’s why we’ve all focused on it.’
I told him about my conversations with John Hardy and Allen Roses, and how Roses had seemed disgruntled that people weren’t paying enough attention to his APOE4 discovery. Schenk gave me a knowing smile. ‘Well, John Hardy’s a geneticist. Allen Roses is a philosopher. If he’s disgruntled it’s probably because he feels it’s a target that we’ve missed, and he’s probably right. But I swear to God we all tried our hardest to figure it out. Everybody had an APOE4 programme. We had one for eight years. Nothing came of it. That doesn’t mean that APOE4 and the tangles are bad targets. It’s just been harder to make a drug for them.’
Given all that, what will the cure look like? I asked. ‘To be honest, I like to talk about conquering Alzheimer’s rather than curing it. Because it’s like asking, how do you cure heart failure? Well, you get a new heart. You can’t get a new brain. So I think there’s going to be a way to prevent it, or push it off many years. That’s how we’ll conquer it.’
I wasn’t sure how I felt about his answer. Like many families and patients, I can’t help but hope for something that reverses the symptoms altogether, rather than just holding back the disease–something that reaches beyond the gloomy sea of dead neurons and pulls a memory back into the light. This might not be a fantasy (as we shall see in chapters fourteen and fifteen), but it’s certainly further away.
One might wonder at the significance of these drug trials. With failures in the past the vaccine therapies have a lot to prove. I am yet to be convinced that a vaccine will work for everyone; Alzheimer’s seems too nuanced for that. But if there’s anything to provide real hope, it’s a scientific lesson best articulated by a woman who was neither a scientist nor a physician. Gertrude Stein said: ‘A real failure does not need an excuse. It is an end in itself.’11
We have known for a long time that the pantheon of science is decorated with failure. Failures are the moving force in science: they seal off one possibility in order to expose another; they force us to look at the problem in a new light. We owe a huge debt, therefore, to the researchers whose life’s work leads to a cul-de-sac. They are indirectly showing us the correct path.
Altogether, these so-called ‘prevention trials’ promised a sea change in the fight against Alzheimer’s. But while the concept of Alzheimer’s as a process got scientists thinking about early prevention, this, they realised, depended critically and inexorably on something else: early detection.