The brain

THROUGH THE FLOOR-TO- CEILING windows of the Parker Institute for Cancer Immunotherapy in San Francisco are the windswept headlands of the Golden Gate Bridge, the Pacific Ocean, and a frothy coral rotunda called the Palace of Fine Arts.

“Would you like a water?” asks the centre’s publicist when I visit. “Still or sparkling?”

Of all the cancer centres I visited, the Parker Institute seemed the most like the future of medicine. The office, a few doors from Lucasfilm, has one of those pristine, snack-filled tech startup kitchens with glass jars and a microwave that pulls out like an oven. On a table in the reception area sits a set of glittery silver pamphlets the size of small yearbooks explaining the mission.

The man behind the Parker Institute is serial entrepreneur Sean Parker, the cofounder of Napster and intermittent recipient of richly deserved tabloid jabs. Parker doesn’t have the most sterling humanitarian reputation: In the movie The Social Network, Justin Timberlake portrayed him as a narcissistic party boy who screws over one of Facebook’s cofounders and is arrested for cocaine possession. Parker was fined R33 million by the California Coastal Commission for building the set of his R130 million Lord of the Rings themed wedding (complete with fake ruins, waterfalls and a cottage) in an ecologically sensitive area. Yet, a little over a year ago, the same man donated more than R3 billion to fund the study of immunotherapy at a lavish backyard gala featuring performances by John Legend, Lady Gaga, and the Red Hot Chili Peppers.

The public story about Parker’s philanthropic effort is that it stemmed from the death of his close friend, film producer Laura Ziskin, to recurrent breast cancer. According to Jeff Bluestone, the Parker Institute’s president and CEO (and, incidentally, the researcher who characterised CTLA-4 around the same time as Jim Allison), Parker was interested in cancer long before he met Ziskin. “Sean’s been interested in the immune system for much of his life, because he’s got asthma and he’s had a serious immunological imbalance,” he says, sitting at a polished rawwood conference table half again as long as a normal conference table. (Parker is extremely allergic to peanuts.) “As long ago as 2004, before Laura got sick again, he thought the immune system was going to be the answer. He deeply understands a lot of the science. We joke, is he a second-year graduate student? A third-year postdoc? Should he just go get a PHD?”

Parker is not the first very wealthy per- son who has used his money to combat disease. Several people at the Institute took care to explain how they were different from the Howard Hughes Medical Institute, a science-funding organisation founded by the reclusive airman in 1953. A more influential predecessor might be Michael Milken, the Wall Street financier who founded a charity dedicated to family medicine with his brother Lowell in 1982 that supported, among other things, the research that led to Gleevec, the precisionmedicine drug. Milken’s funds also supported Jim Allison during an important time in his pre-checkpoint-inhibitor-therapy research when his National Institutes of Health grant had briefly lapsed. In 2003, Milken cofounded Fastercures with Greg Simon with the goal of increasing the pace of cures to “all serious diseases”.

Some would argue that technology entrepreneurs are exactly the people who should be constructing the immaculate future of cancer research conceived by people like Joe Biden and Greg Simon. For one thing, tech entrepreneurs have already disrupted everything else. They understand the fast-moving, coin-chasing world of biotech development. Parker himself has already succeeded at convincing hardheaded institutions to work together. While he was an early investor and board member in the music streaming service Spotify, he negotiated with Universal and Warner to convince them to participate.

The Parker Institute’s fundamental accomplishment thus far has been to do exactly this in cancer research. From the beginning, six academic research institutions signed on to work together under the Parker Institute’s umbrella: Memorial Sloan Kettering; MD Anderson; Penn Medicine; Stanford Medicine; University of California, Los Angeles; and University

of California, San Francisco. The six, along with independent investigators at a few other research institutions, agree to share research data and work together on goals and projects without getting hung up on institutional constraints, such as intellectual property. In return, they get two things: money, which every cancer researcher needs; and guidance, which is equally pressing but not as obvious.

“To become a leader in this field, to be a Carl June or a Jim Allison, you usually have to be a bit, not myopic, but a little blind,” says Fred Ramsdell, the Parker Institute’s vice president of research. This is common in science. To understand and work on a complicated concept, a researcher has to shut out the noise of everything except his exact area of expertise. Someone who works on checkpoint inhibitor therapies in melanoma, for example, might not see much use in reading about ovarian cancer detectors made out of nanocarbon until suddenly it’s the exact bridge to his own next level of progress.

“If a person knows nothing about nanoparticles, I can step in and say, ‘Hey, this nanoparticle thing might be exactly what you need’,” says Ramsdell. “I spend a lot of time trying to develop relationships between people who might not always do so on their own.” Some of those relationships are between researchers themselves. Others are between MDS and Phds or between researchers and drug companies, or engineering companies, or the US Patent Office. It doesn’t really matter, so long as the arrangement furthers knowledge.

Up the coast in Seattle, another tech company is attempting to help cancer researchers cross entrenched divides. Microsoft’s Project Hanover has already made considerable progress on creating a combined, searchable repository of the scientific news released every month by cancer researchers all over the world. The idea is that artificial intelligence will do a better job of parsing the vast landscape of scientific papers (those paper aeroplanes flying between ships) for insights. Rather than fallible humans trying to catch every valuable new detail as papers fly out of scientific clearinghouses, the system will do it for them, considering every possible combination of targeted drugs and genetic interactions in less time and more detail than it would take a team of educated humans.

Microsoft calls this the reasoning bottleneck. In a way, they’re tackling it the same way the Parker Institute is. The same way the human body does: they’re adding a brain.

INTERLUDE

San Francisco. It’s late. At the restaurant, there is a man seated at the chef ’s table when I arrive, drinking a balloon glass of red wine.

“How’s the food?” the man asks after a good half hour. It is delicious: a buttery bucatini with lamb ragu and bread crumbs. The man has lived down the street from this restaurant for years. He’s a former tech entrepreneur who is now a project manager for a retail company. I tell him what I am writing.

“That’s a hell of a coincidence,” the man says. “I just flew home from watching my father die of cancer.” “Jesus, I’m sorry.” “He’s still there. With my sister. He told me he was tired of feeling like he was on death watch. He told me I should just go. So I went.”

He sips his wine. WHAT YOU SEE after a person has been debilitated by cancer and lived are the scars. The missing jaw or breast. The colostomy bag. Hair that has grown back curly or coarse or grey in patches. Tattoos that mark the paths of radiation beams. The disease that contains all of human biology leaves no one unchanged. There is before cancer, and then there is after.

Above Patrick Garvey’s desk, on the top shelf of a bookcase, sits a stack of brown resin jawbones; dozens of them, mostly the mandible, or bottom jaw, which is commonly replaced with a bit of lower leg bone when it has to be removed because it is shot through with cancer. Every jawbone above Garvey’s desk is a relic from a surgery he has performed at MD Anderson over the course of three years: more than 200 patients whose faces are forever altered by their interaction with the disease.

Later today, Garvey will operate on a man with a more difficult case – a large

tumour in the maxilla, or top jaw – as part of two surgical teams. The first team will remove the tumour and most of the bone, including the man’s eye, and then Garvey’s team will remove a piece of the man’s fibula along with its blood supply and use it to reconstruct the man’s face. “We’ll be here into the night,” Garvey says.

This type of surgery is called microvascular reconstruction surgery. It drastically improves life for patients who would otherwise, like late film critic Roger Ebert, no longer be able to eat or talk without support. When it fails, however, it fails impressively: the transferred bone must have the correct blood supply or the body will simply reabsorb it, leaving only the bare metal scaffold the doctor implanted. Human bone is far better suited to the long-term mechanics of chewing and talk- ing than metal is, and a plate without bone to protect it will eventually snap, like a paper clip bent back and forth over and over. Garvey has had to reconstruct jaws that have failed before, leaving patients disfigured and unable to chew properly. For a patient who has already undergone treatment for cancer, the impact of having to have multiple reconstructive face surgeries is harrowing.

To make the surgery simpler, Garvey’s team uses 3D-printed cutting guides and robotically milled metal plates to create the most precise reconstruction possible. This is how the brown resin jawbone graveyard above his desk got started. After a patient has a CT scan, a company called Materialise in Plymouth, Michigan, prints the jaw models as well as bolt-on cutting guides that show the surgeons exactly where to saw and reconnect fibula bones to match the person’s original bone structure. Another company, in New York, creates a metal scaffold that is meticulously bent so as to recreate the original face angles, so MD Anderson’s surgeons don’t have to bend an off-theshelf part into position during the reconstruction.

By all accounts, using 3D-printed guides to reconstruct a human face is an advance at the very edge of cancer medicine, and yet it is still disheartening to look at the statistics. Last year, another 1,7 million Americans were diagnosed with cancer, and almost 600 000 died. Since 2004, according to the latest data available, the overall decline in death rates has been just 1,8 percent in men and 1,4 percent in women year over year. The five-year survival rate for pancreatic cancer, which most doctors consider the worst of the worst, sits stubbornly at just 8,2 percent.

Perhaps the cure for cancer seems so elusive because it’s a failure of semantics. “Curing cancer” is impossible, and the statistics reflect that; cancer kills more Americans every two years than those who died in every war the country has fought. However, helping some cancer patients, the luckiest of the unlucky, live in relative normality for years is not just possible. It is happening. The five-year overall cancer survival rate is up from 50 per cent in 1975 to 67 per cent today. For melanoma, it’s 91,7 percent. For prostate cancer, it’s 98,6. It will take time for the most promising treatments to trickle down to everyone they might be able to help, but in the meantime, the march continues.

What this has to do with Patrick Garvey is that, even subtly, using 3D- cutting guides to improve plastic surgery shifts the focus of cancer treatment from emergency battlefield triage to matters of aesthetics and psychology that matter months and years down the line. Without saying it, exactly, the field of cancer treatment is acknowledging that cancer could one day become a survivable disease – that even a stage four metastatic cancer patient could survive long enough for normalcy to matter.

There are others on the front lines: at hospitals, women with breast cancer can wear a scalp-cooling system called Digni- Cap during chemotherapy treatments to reduce the likelihood of hair loss. At MD Anderson, a neuroscientist retrains patients’ brains to improve altered nerve sensation caused by chemotherapy. St Jude hired a psychologist to help teen cancer patients plan to save their eggs or sperm, in case their treatments render them infertile and they want to have a family in the future.

Future. A tricky word for a cancer patient. Who gets to have one is still a function of blind fortune. But all these ideas are starting to come together and progress is suddenly accelerating. We are at what Crystal Mackall calls “the end of the beginning” and the hope is that one day soon, the miracles will no longer be miracles. They will just be what happens. Until then, we pin our hopes on the incremental or unpredictable improvements, the half measures, the better outcomes. It will always be true that once a person has had that most frightening of conversations with chance, life will be split into two parts: the time before cancer, and the time after it. But for a fortunate few, perhaps the second part can be as good, and even as rich and wonderful and as great as the first.

PEOPLE AT THE University of Cape Town faculty of health sciences don’t know Mhlangalab by name. This is a problem when you’re trying to find your way around, because all the corridors and building entrances look the same in a way that makes you wonder whether it’s purposely designed to confuse students. People do recognise the room number, though. This recognition is followed by the discovery that they actually do know about the lab, just they refer to it as “the new lab”.

Dr Musa Mhlanga heads up the CSIR’S synthetic biology programme, but wants Mhlangalab’s work to stand up to scrutiny from the global scientific community. His PHD in cell biology from New York University and contributions to gene expression research have pioneered now widely adopted methods and technology. Right now, though, he has a problem.

“If you came tomorrow, this would all be fixed. The problem is that when they moved this microscope they took something apart, but now they can’t put it back. Maybe you can find a solution.” The problem in question: trying to fit a circular plate on top of an expensivelooking piece of equipment. The screw holes don’t line up with the bracket.

“This is called a baseport camera and it is crazy. It’s a single photon counting camera, so the chip is made in such a

“I HAVEN’ T ALWAYS WORKED IN THIS AREA OF CANCER RESEARCH. PRECISION MEDICINE IS DEFINITELY A NEW TERM THAT’S VERY MU CHINA HYPE PHASE AN DISUSED TO DESCRIBE P HARM A CO GENOMIC S .”

way that the camera cools the sensor to -80˚C so that there’s no electronic noise. It’s not supercooled, but it’s cooled to a point where molecules don’t move that much so it allows us single photon counting sensitivity. These guys took it apart and now we can’t get the plate back on. It doesn’t fit. We’ve tried all kinds of things.”

Mhlanga demonstrates engineer-like levels of intimate mechanical knowledge when explaining how the microscope works: “There are four different wavelength lasers going in here. They become one beam and hit a spinning disc that has thousands of holes and a micro lens on each hole so that the laser light can homogeneously illuminate the whole area. This allows you to look at the entire area under laser illumination. Which means you can see everything really quickly. The point scanning confocal microscope usually uses something called raster scanning, which sees only a point of the light. The only way to see with raster scanning is to move the light point, but that takes a lot of time. If an object is in motion, you won’t catch that motion with raster scanning.”

He also explains how the Nipkow disc inside this microscope is responsible for the creation of early mechanical televi- sions. In this application, the disc has around 10 000 holes, allowing the researchers to gaze inside many cells to see DNA, RNA and proteins.

“I haven’t always worked in this area of cancer research. Precision medicine is definitely a new term that’s very much in a hype phase and is used to describe pharmacogenomics. Cancer is a disease of the immune system, but it also has a genetic entomology,” explains Mhlanga. “Essentially, in cancer we have disregulation of when genes come on and go off. Our lab works on how genes are regulated: how do genes go from the on state to being in the off state and vice versa? That’s important, because understanding that process is really the first step in understanding the disregulation of gene expression that brings on cancer. For non-experts in the field, gene expression can be very intimidating and highly technical.”

Mhlangalab uses cluster computing to manage the data derived from its research. That data is the platform on which scientists can build a body of evidence to efficiently serve oncologists in their treatment needs. Precision medicine is basically matching cancer cells to the specific drugs that can eradicate it.

“Our work in this field really started with our interest in how genes are regulated. It’s called 3D chomatin architecture and how that contributes to how genes turn on and off. The way to understand the gene regulation problem intrinsically is that we have trillions of cells and in each of these cells we have exactly the same DNA, more or less. What happens is that in your skin cells, for instance, the portion of the DNA that’s needed to make a skin cell a skin cell – the elements of the genetic code – are on in the skin cell, but they’re not the same ones that are on in a neuron,” says Mhlanga. “This difference is because, even though we have the same genetic code, or recipe book, the recipe book needed to make a neuron has different ingredients and requires different specialised genes to be on. When we have a cancer cell, what happens is that we have a gene expression program in a particular cell type that is different from the normal program of that specific cell type, even though it has the same DNA. And that could be because some alteration has occurred to the DNA. Let’s say I stay in the Sun a lot and that one skin cell gets a mutation that doesn’t get fixed and it’s in a key region or in a hotspot where mutations can happen easily, that then leads to bad programming that leads to us getting the cell dividing uncontrollably. And then that may attract the attention of the immune system. If the imune system doesn’t recognise it, the immune system almost protects that cell or that group of cells and eventually they can’t contain themselves in that space and become metastatic and spread to other body parts. It’s a big advance in our understanding of cancer to link genes and abhorrent gene programs to drugs that can reverse these activities.”

In the future, all the work from the likes of Mhlangalab will be fed into computer neural networks alongside population genome sequencing data and we will, through deep machine learning, have a detailed picture of what cancers affect which people. This could theoretically put doctors and scientists in a position where they can be more proactive. The machines can then tackle the data science behind what causes cancer and work on a possible vaccine that will send the cancer the way of smallpox. We’ll stop cancer by grinding it into food for our AI machines. This is how we win. PM

PRECISION MEDICINE

Hope