Skip to main content

CC-BY: Ben Gilbert/Wellcome Images


CC-BY: Ben Gilbert/Wellcome Images

Alok Jha talks to Harold Varmus, Nobel Prize-winning cancer researcher and current Director of the US National Cancer Institute.

The night before Harold Varmus received a call from Sweden telling him he’d won the Nobel Prize, he had been out drinking with friends. He was a little hungover when the phone rang in the dead of night.

“I’m sure I was smiling,” he says, thinking back to that call. In only a few minutes, the news had spread and phone calls began pouring in. “I have an image of myself, because there was a picture taken of us, you know, just standing in the hallway talking on my phone for about three hours; endless calls kept coming in from various people. Then we trotted off to a press conference…it’s an exciting day.”

Harold Varmus with a bike wheel

Varmus on... cycling

Each time we fix up a Mosaic conversation, we ask our subject to bring with them an item.

The Nobel committee had already tried to get hold of Mike Bishop, Varmus’s long-time research partner, and left a message on what they thought was his answering machine to tell him he had also won the most prestigious prize in science. But when they asked Varmus for his colleague’s phone number, they soon realised they had called the wrong man. The Nobel committee had called another Mike Bishop, a faculty member in the psychiatry department at the University of California, San Francisco. “So they called back the first Mike Bishop they had called and they left another message, saying ‘Please ignore the first message’.” No one knows how that Mike Bishop responded to that particular succession of phone messages.

Almost two decades before the Nobel committee’s call, Varmus had joined Bishop’s lab to work on better understanding how cancers develop and the part viruses might play in causing them. Their work had its ups and downs, but within a decade they had laid the foundations for our modern view of cancer as a disease of the DNA in our cells. The most advanced and most promising strategies to understand and tackle this set of diseases today – from sequencing tumours to finding targeted therapies for individuals – come from the early work of a man who was a hair’s breadth away from never studying science professionally.

Unlike the majority of his elite peers in the scientific world, Varmus had no interest in science at school. He won no high school science fairs, and he found nothing notable or inspirational about his lessons in biology and chemistry. His father was a doctor and his mother a psychiatric social worker so, at Amherst college, Varmus defaulted into a premedical curriculum with a vague idea that he would end up working in the health sciences.

The core curriculum at Amherst gave him a lot of exposure to literature, history and philosophy, and the young Varmus found himself drawn to the humanities. He immersed himself in Chaucer, Shakespeare and Dickens and got wrapped up in poetry. In his spare time, he ran the student newspaper. He ended up majoring in English literature while taking the minimum number of courses he needed to leave open the option of medical school. Such was his preference for the humanities that he almost failed his organic chemistry course. “I wasn’t getting the same level of inspiration from science,” he says. “My impression of science was that it was something that was more predictable: I didn’t see the adventure in it, I didn’t see the imagination in it.”

After Amherst, he won a fellowship to study Anglo-Saxon and metaphysical poetry at Harvard University. He found the enjoyment of his subject overshadowed, though, by friends at the medical school, who seemed to have a much greater sense of purpose (and also seemed to be having more fun than him). A year into his graduate studies in English, he gave his career a rethink: if he continued on his current path, he might become an English teacher and a critic; study medicine, on the other hand, and his career options would remain open. “If I went to medical school I could end up being a novelist or an artist or something else – a psychiatrist, a scientist, it just seemed to open all doors. And so I made the switch and went to medical school.”

Rejected by Harvard Medical School, Varmus ended up at Columbia University in New York in 1962. In 1966, he elected to work as a clinical associate at the National Institutes of Health (NIH) for two years rather than go to serve in the Vietnam War, which he “fervently opposed”.

That slight career deflection was crucial because, at the age of 28, he got his first serious exposure to laboratory science. He was working on research looking at how genes were regulated inside the bacterium Escherichia coli, and he learned very quickly that “doing science was very different than learning science and doing rote experiments in college classrooms. So the glamour, the adventurousness of it became quite apparent, and I made a decision – even though I’d already had four years of medical school, two years of hospital training – to continue.”

Asked when he realised that this kind of laboratory research was going to be his calling, and what it was that showed him the adventure in science that he had never appreciated at school or college, Varmus answers immediately. It came, he says, in the late 1960s when he developed a test that used radioactive markers to measure the amount of messenger RNA (molecules that genes use to communicate information to cellular machinery on how to build proteins) inside cells. “Things like measuring amounts of RNA are so trivial now that students would laugh at my finding this an emotional experience. But to be able to say ‘the gene is off in this cell, and the gene is on in that cell,’ that’s pretty thrilling. And I knew the questions that I wanted to answer could be answered because I had the bull by the horns; I had an assay that was really clean.”

He was soon to start working in Mike Bishop’s lab at the University of California, San Francisco, on the roles of viruses in cancer. “You infect a cell, and the cell changes its behaviour: nothing can be more dramatic than that,” he says about why he became interested in the topic. “And that was a very attractive notion because all of us who had watched cancer develop in our own friends or family members, or had taken care of cancer patients on a hospital ward, saw this as one of the great mysteries – why should a normal cell suddenly go crazy and outgrow the surrounding cells and ultimately kill the patient?”

Each of us came from a single cell created at our conception when sperm and egg fused and began to divide. That fertilised egg divided countless thousands of times over the following months, in multiple ways, to produce the organs, limbs, brain and hundreds of cell types that make up a baby. Those divisions continue, every day, as our bodies grow or need to repair damage or just replace old and tired cells that are malfunctioning.

The creation and functions of all of these cells are controlled by chemical signals sent around the body. These signals determine when a cell divides and when it should die, when it should grow and when it simply needs to stay put. The signals also determine how big a kidney or liver grows and when it is more or less the right shape. Our cells respond to the chemical signals and do all the growing and dying that is expected of them because they’re following the rules built into their genomes.

If those rules aren’t followed properly, cells malfunction. They might have divided too many times or collected too many errors in their DNA with age, or their genes might have been damaged by radiation or environmental pollutants. These errors could mean that the cells don’t stop growing when they are meant to, for example, or outstay their usefulness in some other way. They hang around – or worse, grow numerous and needlessly take up nutrients and space. When the rules for cell division fail, the malfunctioning cell can grow inappropriately and take over an organ. That ball of overgrown cells, a tumour, can prevent the organ from doing its job properly and, in the worst case, kill a person.

When Varmus began his research career around 45 years ago, scientists knew that cancers were growths that arose in certain organs and that those growths, if not controlled or cut out or irradiated, would spread to other parts of the body and could kill people. His mother died of breast cancer in the early 1970s and, at the time, doctors could only get a basic understanding of what was wrong with her. “She was well for a year or two after breast surgery, and then it turned up in her bones and other places and there was nothing to do to control it,” he said. “We had no idea why she had breast cancer, whether there was a genetic component, whether she’d been exposed to something environmental; it was disturbing [not knowing] whether this was a spontaneous thing that couldn’t be predicted.” 

On the heels of the success of the Apollo moon landings, the US political elite felt that scientific progress could tackle almost anything. In 1971, President Richard Nixon signed the National Cancer Act (often referred to as his ‘war on cancer’) – with the assumption that if US scientists could get to the Moon and back, they could deal with cancer too. The National Cancer Institute was given more money and power, and Varmus and Bishop became beneficiaries of a special programme to look for viruses that could cause cancer in people.

One of their early areas of study was the Rous sarcoma virus (RSV), identified in 1910 by Peyton Rous and known to trigger cancer in chickens. It was also the organism in which scientists first detected cancer-causing genes, known as ‘oncogenes’, which could transform a healthy cell into a cancerous one that grew without stopping.

“If you put cells from a chicken embryo into a normal petri dish, the cells will grow – they’ll form a single cell layer of very nicely behaved cells, all lined up, and fill the dish and stop growing,” says Varmus. “And if you introduce Rous sarcoma virus into one of those cells, it’ll make a big pile of cells that’ll spill out, they won’t stick to the plastic surface, and they will change the way the tissue looks. Now, it’s got little mountains of cells that look like little tumours.”

RSV is a retrovirus – its genome is in the form of RNA. This is copied into DNA, which is integrated into the infected cells’ DNA. RSV’s genome is mostly composed of genes for the virus’s multiplication, but it also carries a gene, now known as a viral oncogene, that transforms cells. This oncogene is called SRC (pronounced 'sark’). 

The wider programme to look for cancer-causing viruses in humans, according to Varmus, unfortunately yielded very little, and some of what it did find turned out to be wrong. By the mid-1970s, many scientists were convinced that most human cancers were not caused by viruses and certainly not by retroviruses.

But in carrying out that search, Varmus and Bishop found that normal, uninfected cells from a wide range of species – from yeast to birds to humans – contained very similar genes to the oncogenes that had been observed in retroviruses. In normal cells, these ‘proto-oncogenes’ (as the pair had named them) were important for cell growth and specialisation. But they could go wrong because of mutations caused by random errors during normal cell division or when they were infected by a virus or exposed to some environmental carcinogen. When they did go wrong, these genes triggered uncontrolled cell growth, leading to tumours.

Once Varmus and Bishop had proposed the proto-oncogene idea, several teams of scientists went looking for mutant versions that might be the precursors to the oncogenes carried by retroviruses. Some of the most persuasive early evidence came in the early 1980s from Robert Weinberg at the Massachusetts Institute of Technology. “[He] showed that there were mutant forms of another gene called Ras, a gene that is found in retroviruses that come from mice and rats, not from chickens. And that gene, the Ras gene, was sometimes mutated in human cancers. Now we know that gene is actually mutated in about a quarter to a third of all human cancers; [it’s] a really big deal, whereas the SRC gene that we studied is really very, very rarely mutated in human cancers.”

In the 1976 paper for which they received the 1989 Nobel Prize in Physiology or Medicine – ‘for their discovery of the cellular origin of retroviral oncogenes’ – Varmus and Bishop had already proposed that damage to the proto-oncogenes within healthy cells played a key part in the origins of cancers, although work by many cancer researchers in the intervening years was required to establish the validity of their proposal.

Varmus and Bishop’s work gradually turned cancer from a bunch of mysterious growths in the body to a disease that had its roots (and, therefore, solutions) in DNA. This genetic basis for cancer shed light on several puzzles: why there are so many types of the disease found in the body; why the incidence of cancers rises as people get older (there is more time for genetic mutations to accrue); and why some people exposed to environmental carcinogens develop the disease while others do not (variations in individual genomes).

With the Nobel Prize, Varmus became more prominent in the leadership of his profession. Until then, he had tended to avoid administrative roles and anything that took him away from frontline research. “I’d never been a department chair. I didn’t want to be. Some people would have said as a citizen of my academic institution, I was, sort of, a shirker. I did some things, but I wouldn’t serve on all committees. I really cared about my own laboratory research.”

The Prize, and the attention that came with it, pushed Varmus into getting more involved in the politics and policy of science. At the time, federal budget deficits were threatening basic research funding, and Varmus dipped his toes in the water by becoming involved in policy discussions and committees for the National Academy of Sciences. He joined a group of scientists who supported Bill Clinton’s successful bid for the presidency in 1992, and the following year Clinton nominated him to be director of the NIH, the organisation that had kick-started his scientific career almost 25 years earlier.

Given his relative lack of administrative experience, Varmus says this offer would have been unthinkable without the public attention bestowed by the Nobel Prize. “Was I ready? No. I mean, I had never run anything anywhere near the size and complexity of something like the NIH. And I was only able to do so because I had a lot of help when I got there and I got into it.”

Despite his previous reluctance for administration, Varmus says his time at NIH was a good one. “I actually enjoyed the change of pace involved in political work, and going down to Capitol Hill and fighting for our money. Those were days when you could be successful, and I was successful. I liked taking on certain kinds of problems, like getting NIH to focus more on global health and trying to change attitudes towards publication practices. And I was there at a time when there were some big political, or ethical, issues that only the NIH director could actually take on, [such as] cloning and stem cell research…things that required that the public and Congress talked to me. And that was exhilarating.”

He brought in a raft of new talent to lead new NIH research institutes, oversaw a large part of the NIH’s work for the Human Genome Project, and tried his best to make the organisation more open to more innovative, riskier research. By the time he left its leadership, the NIH budget was set on a course to double in less than a decade.

Around this time, Varmus became interested in the opportunities that the internet provided in making scientific literature freely available to researchers around the world. Around the turn of the century, he helped to launch a service called ‘E-biomed’ (later PubMed Central), which would be an archive of scientific literature. The following year, he became cofounder of the Public Library of Science (PLoS), a non-profit publisher of open access journals. PLoS now has a range of high-end specialist journals in the biomedical sciences, including PLoS Biology, PLoS Medicine and the more general PLoS ONE, which publishes on any topic that could be classed as scientific and accepts all submissions where the conclusions are supported by the presented data. All papers from these journals are available free on the web, and the costs of publication are paid by researchers.

The arguments for open access are well aired today, but in the late 1990s Varmus’s ideas were met with a lack of cooperation, if not outright hostility, from publishers and scientists who liked the publishing system (and the scientific structures it enabled) just the way it was. More than 15 years later, his support for open sharing of information is undimmed. “It seems to me [that] funding agencies have a responsibility to get the information back out there,” he says. “It’s simply wrong to say that because a lot of the work is technical, only people who work at prestigious academic institutions should have easy access to it.”

The profit margins of some of the major academic publishing houses – which have been reported at more than 35 per cent in some cases – are, he says, “on the verge of being egregious,” but he is less concerned than he used to be about the future of this particular campaign. Even the most elite of scientific journals, Nature and Science, have recently launched open-access titles, and Varmus says the large-scale move towards this method of publishing and accessing science is inevitable. “Well, there’s still journals that don’t want to do open access,” he says. “They may want to do a hybrid form; they will come around eventually. Eventually, the world will be all open access.”

In his four decades in the field, the scientific understanding – and potential treatment – of cancers has changed radically. “Probably the first thing to say is that we know that every cancer is different from every other cancer. If a cancer arises in an organ like the lung, for example, to call it ‘lung cancer’ is not fair. There are many types of cells in the lung, and at least three or four are represented as different types of cancer when cells fail to obey the usual constraints.”

Study 500 tumours that all involve the same type of cell, and each tumour will be unique; each one will have a different constellation of genetic mutations. There may be some shared mutations, but the pattern will be different in every case. On top of that, each cancer will arise in the context of a unique genome. “And all of these things mean we have a tremendous job to do to sort out what’s truly significant. If you’re going to plan either to prevent or to treat the cancer, you’ve got to think about all these variables.”

Cancer scientists around the world have produced a phenomenal arsenal of data and strategies to begin dealing with this complexity. In part, this has been enabled by the decreasing cost of genetic sequencing and other precision technologies available for molecular biology. “One of the extraordinary things about what we’re doing today is that we’re able to look at a cancer that arises in human beings and look at the entire genome, look at all the mutations that have occurred, all of the variations that were inherited, all of the RNAs that are made from that genome in the cancer cell, and even look at modifications of the DNA and proteins made from the DNA – and that can be done in a few weeks.”

All of that allows a more fine-grained ability to classify cancers and look for weak points in their armour. Varmus takes a type of cancer that his lab works on, adenocarcinoma of the lung, as an example. It is the most common form of lung cancer arising from cells that are in the periphery of the organ. “Perhaps 25 per cent [of cases] will have mutations in the Ras gene; about 10 per cent will have mutations in a gene called the epidermal growth factor receptor gene.” Not every gene that is mutated within the cancers will be unique, but there might be recognisable patterns. Those commonalities can lead to simplifying strategies in dealing with the tumours.

Suppose you had a way of treating cancers that were driven by a Ras gene, for example. Even though the pattern of mutations in a tumour may be different in each case, everyone that has the Ras mutation in their tumour might be able to benefit from your treatment. “By treating Ras alone you wouldn’t solve all the problems, but many of them. And then when you look at the vast panoply of cancers, you find that 95 per cent of adenocarcinomas of the pancreas, a big killer everywhere, 50 per cent of colon cancers, 25 per cent of adenocarcinomas of the lung, and so forth, have these strong mutations. So if you could do any single thing in cancer therapeutics, it would be a way to kill cancer cells that are driven by Ras.” Further ahead, he says, why not look at all the proteins that Ras interacts with and find drugs to disrupt those interactions and prevent the tumours from progressing?

The National Cancer Institute has just set up a programme to target Ras, based at a facility in Frederick, Maryland. Varmus gets animated about his role in organising this kind of work – identifying problems, bringing leading scientists on board to run research groups, explaining to the scientific community why Ras was a tough (but potentially achievable) goal, moving money around and getting the programme funded in a climate in which tax dollars are getting more scarce.

All of that research is, of course, geared towards finding ways to tackle cancer that work better for individuals without the nasty side-effects of modern drugs. Because of a distaste for euphemisms, he doesn’t like the term ‘personalised medicine’ when it comes to thinking about the cancer treatments of the future. “Precision medicine is a term I prefer.”

The current triumvirate of cancer treatments – surgery, radiotherapy and chemotherapy – are not going anywhere any time soon, he says, but he predicts two major forms of therapy that will become more common in future. “The first is what we call targeted therapy. You’ve been hearing a lot about that, where you identify the mutations and you have specific drugs like Gleevec, which is the classic – the poster child for targeted therapies directed against a mutant Abl protein.”

There are several drugs like Gleevec, he adds, that will allow doctors to promise cancer patients a relatively normal lifespan. “They’ll be taking drugs all their life, very likely, but the drugs are relatively non-toxic, they’re taken orally, and they’re going to be quite cheap. I think most cancers will become chronic disorders that we take care of and sometimes cure.”

The other avenue for cancer therapy, he says, is to recruit the body’s immune system. This is already in some use, of course: monoclonal antibodies, such as Herceptin, targeted towards specific proteins on the surface of a cancer cell can be used to reduce the size or number of cells in a tumour.

“But the two new forms of immunotherapy are even more interesting scientifically, and possibly will be of much wider use in cancer practice in the next 10 or 20 years. One is to take advantage of the deep knowledge we now have of how the immune system works.”

The immune system doesn’t sit around waiting for a foreign body to trigger a response: there are both positive and negative control elements, says Varmus. In general, the immune system doesn’t respond too strongly to cancer because of the brakes on it that prevent it from responding to the body’s own cells. “And we now know how to interfere with those brakes on the immune system by using antibodies that interfere with those suppressor systems. Some of those have been approved by the FDA [Food and Drug Administration], some are on a fast track to approval… When we learn how to use these antibodies – that is, what the right context is for using them – and develop better ones, this will be very important. It’s already pretty important.”

Another option is to programme T cells, the immune system’s killing cells, to respond to the proteins on the surface of cancer cells. “That means putting into T cells what’s called a chimeric, a fabricated receptor that recognises something on the surface of a cancer cell and leads the T cell to attack it. That’s being used in the USA. It’s stuff that’s not FDA-approved yet, but I’m sure that there will be FDA approvals for some of these.”

The opportunities scientists and doctors have to fight cancer in the decades ahead, so many of which were sparked by the virus work of Varmus and Bishop, are vast. This frightening, mysterious, shape-shifting set of diseases is giving up its secrets, base by base. But that does not mean we should fool ourselves into thinking victory is around the corner.

Addressing the assembled dignitaries at the Blue Hall in Stockholm during the Nobel banquet in 1989, Varmus used his love of literature to indicate the ground that lay before him and his scientific colleagues.

He quoted from the epic poem Beowulf, where the hero’s victory over Grendel is celebrated in the hall of Hrothgar, and he likened that gathering to the one taking place that evening in Stockholm. Unlike Beowulf’s victory, however, he told his audience that humanity had not yet slain its enemy, the cancer cell, or even figuratively torn the limbs from his body. “In our adventures, we have only seen our monster more clearly and described his scales and fangs in new ways – ways that reveal a cancer cell to be, like Grendel, a distorted version of our normal selves. May this new vision and the spirit of tonight’s festivities inspire our band of biological warriors to inflict much greater wounds tomorrow.”

The interview and photoshoot for this article took place on Friday 21 March 2014.

Return back to top of the page