SHE doesn’t dwell on it, but 82-year-old Lydia Knott knows what will happen to her after death. Her body will be taken to a laboratory for an unusual post-mortem.
It won’t be to find the cause of her demise. Knott was diagnosed with lung cancer five years ago. After surgery to remove part of her lung, she is now doing well: “Fine for an 82-year-old, I can’t complain.” But if the disease returns and kills her, Knott wants doctors to learn more about her cancer through a “warm autopsy”, so-called because it happens soon after someone dies.
Advertisement
Within 24 hours, a team would remove up to 80 tissue samples and preserve them using liquid nitrogen. One of the aims is to fathom cancer’s surprising ability to evolve. The same forces that shape the tree of life also drive tumours to spawn and spread, generating a vast genetic diversity of cancer cells within a single person.
Now, thanks to recent leaps in genetic sequencing, the hope is that we can trace a cancer’s evolutionary journey and create powerful treatments using this information. “We may actually have the technology to cure many cancers – we just haven’t been using the right strategy,” says Robert Gatenby at the Moffitt Cancer Center in Tampa, Florida.
We might even be able to stop tumours developing in the first place, but it won’t be easy. “We are battling natural selection, one of the fundamental laws of the universe,” says Charles Swanton at the Francis Crick Institute in London, co-leader of the study that Knott has signed up to. “But I think this is our best shot.”
A step ahead
We have known since the 1970s that tumours arise when a mutation occurs in one of the genes within a cell that control its reproduction. Our cells are continually replicating themselves – even in healthy tissues – by growing and splitting in two, to replenish those that have worn out.
This process is controlled by hundreds of genes so that cells divide at the right place and time. If a cell has a mutation in one of these genes – perhaps through exposure to cigarette smoke or ultraviolet rays or just bad luck – it may start multiplying faster, its progeny outcompeting healthy cells.
Even if the immune system attacks these cancer cells, their ability to evolve may thwart our body’s attempt to destroy them. There are parallels between the way predators shape the evolution of prey populations and how immune system cells kill the most vulnerable tumour cells, says Mel Greaves at the Institute of Cancer Research in London. Cancer cells that evolve immune system resistance can thrive. “It’s survival of the fittest. It’s not happening in a jungle or a pond – it’s a prostate or a breast. But it’s a Darwinian process.”
This knowledge hasn’t yet influenced the mostly crude and brutal way we treat cancer, which is sometimes described as slash, burn and poison, referring to surgery, radiotherapy and chemotherapy. These approaches can work if the cancer is caught early enough, which generally means before it has spread. But chemotherapy and radiotherapy work by killing all quickly dividing cells, which means they also damage the skin, gut and immune system, causing side effects such as hair loss, sickness and vulnerability to infections.
More recently, targeted therapies have been touted as the next big thing. These work by blocking molecules on cancer cells that are specific to them, so tend to have fewer side effects on healthy tissue. This requires testing someone’s cancer to identify the mutations involved, so is painted as the ultimate in personalised medicine.
We often see media coverage when one of these new treatments reaches the clinic, but the reality is they usually extend people’s lives by only a few months. That is because their developers failed to consider cancer evolution, says Swanton. A targeted treatment kills all cancer cells that carry a certain molecule – but any that don’t have that molecule survive. It therefore “selects for” growth of cells that are resistant to it, so within a few months, tumour cells without that molecule are more numerous, and the treatment no longer works. The cancer has developed drug resistance.
Doctors have long been aware that targeted therapies don’t usually extend people’s lives for long, but it is only recently that we have been able to genetically chart how resistance evolves. In a 2012 study, Swanton’s team sequenced multiple samples from four people’s kidney cancers, and found the cells diverged over time – the way different animal species branched off from each other over millions of years. “It’s not linear evolution, it’s branching evolution,” says Swanton.
Within each person, two-thirds of the mutations weren’t shared across all their tumour cells. This showed that it could be misleading to take one small sample from a cancer to predict which targeted therapy to use. “Depending on where you put your biopsy needle, you’re going to get different results,” says Swanton.
Since then, several other studies have launched with the goal of drawing up detailed evolutionary trees for individual cancers, exploiting our new DNA sequencing capacity. The largest is the one that has recruited Knott. However, the only way to get enough tumour samples from each person is to collect them after death, otherwise it would be too destructive, says Mariam Jamal-Hanjani at the Francis Crick Institute, who co-leads the study along with Swanton.
The team is working with hospitals around the UK to find 500 people with many different types of cancer. While some doctors are hesitant, says Jamal-Hanjani, “almost always, these patients are incredibly willing to donate their bodies knowing it will benefit others”. As Knott says: “I want to help other people and it seemed to be a logical next step. And when you’re gone, you’re gone.”
The 2012 kidney cancer paper was seen as , but it also showed that within each person, about a third of the mutations were present in all the tumour samples they took. These mutations must have arisen when the tumour was small, before its cells had diverged much, and are sometimes called “trunk mutations”, meaning they are in the stem of the evolutionary tree, not its branches. Any therapies targeting trunk mutations should in theory kill all cancer cells and so be less likely to trigger resistance.
Swanton thinks that the best way of doing this is by weaponising the patient’s immune system. People with cancer usually have some immune cells active against their tumour – but most target branch mutations. If someone’s immune cells could be directed at their trunk mutations, this could be enough to get rid of all the malignant cells from their body.
Swanton has co-founded a firm called , which is due to start two small trials of this strategy this year in lung and skin cancer. When people have their tumour cut out, the firm will sequence many of its cells to identify the trunk mutations. Researchers will also extract immune cells from within the tumour and select those that target these mutations. The firm will then multiply the immune cells in the lab, so that many can be injected back into the patient.
“Targeting certain tumour mutations should in theory kill all cancer cells”
Great adaptations
Achilles is by no means the first to try to exploit people’s immune response against their cancer. The idea has a long history and has been recently turned into a treatment called CAR-T therapy for leukaemia and lymphoma, cancers that occur when blood cells turn malignant. Using this approach against solid tumours has proved more difficult, though. Several groups have tried and failed at this – although so far none has tried genetically identifying trunk mutations as Achilles is doing, says Swanton.
Whether or not this approach works, a focus on evolution may lead to other treatment avenues, such as tweaking the way we use existing anti-cancer drugs. Gatenby sees parallels between the evolution of drug resistance in cancer cells and that of pesticide resistance in insects. “They learned 50 years ago you can’t eradicate with huge doses of pesticide – all you do is you get resistance.” Treating intermittently is a better approach, and might work on cancers too, he says. “If you treat a little bit then take it away, the tumour will regrow over months or years. But when it comes back, there has been no selection for resistance.”
Gatenby’s institute recently found in prostate cancer. Researchers there have now started or are planning five larger trials to put this “adaptive therapy” to a more rigorous test, by comparing it with standard treatment in prostate cancer and three other types of tumours.
In a related tactic, another team at the Moffitt Cancer Center plans to tackle a rare and aggressive form of muscle cancer by switching from one drug to a second one rather than using a single drug intermittently.
The current approach for this is to give a certain drug combination for 10 months and then wait for the cancer to recur with drug-resistant cells, which it nearly always does. At that stage, people are given a second therapy. The new idea is to give the first therapy for just three months and then immediately switch to the second. The rationale is that a small number of resistant cells must have been there all along, says oncologist Damon Reed, who is planning the trial. Hitting hard and fast with the second therapy could have more chance of killing all the cells.
“The idea is there’s a moment when you can more likely induce an extinction,” says Reed. When the number of cells sensitive to the second therapy are as low as possible, “that’s your chance to do this”.
Better than improving treatments for cancer would be stopping it arising in the first place. This goal is being explored for people who are at higher risk of certain tumours, again using evolutionary principles. One such group is people with Barrett’s oesophagus, a condition in which the tube going from the mouth to the stomach becomes inflamed by leaking stomach acid, making them prone to cancer of the oesophagus. Those affected have a check-up every few years with a camera put down the throat and regular tissue samples taken to try to spot tumours early. But the best time to intervene is unclear. Doctors don’t want to cut out any of the oesophagus unnecessarily and some people end up getting cancer despite the screening. Perhaps we can be more accurate by considering evolution, says Trevor Graham at the .
Evolutionary theory says that, all else being equal, animal or plant species that are more genetically variable are more likely to branch into multiple new species under some new selective pressure, like a change in climate. In the same way, a genetically diverse group of cells in the oesophagus may be more likely to harbour one that can turn cancerous; the new selective pressure in this case could be a change in lifestyle, like starting smoking.
Graham’s team put this idea to the test by sequencing the DNA of cells taken from routine biopsies of 320 people with Barrett’s oesophagus. Twenty ended up developing cancer and those who initially had more genetically diverse cells in their oesophagus were .
That means that offering this sequencing to everyone with the condition might let us decide whether people are at high risk – and could have check-ups every few months instead of years – or low risk. “If someone’s at very high risk we can keep a close eye on them, but the real win is to send people home and tell them not to worry,” says Graham.
“There may be a better way to tell if people have aggressive or slow-growing cancers”
His team has now done the same study in people with an inflammatory bowel condition called ulcerative colitis, which puts them at higher risk of bowel cancer and is also monitored with regular biopsies. The results, Graham says, aren’t yet published but are similar. This same thing may also be the case for other types of cancer, such as breast and prostate, says Greaves. In both these tumour types, we often take biopsies from small lumps found through screening tests and don’t know if they are aggressive cancers that must be cut out straight away or slow-growers suitable for watchful waiting, sometimes characterised as “tigers” and “pussycats”.
Give us a steer
For decades, scientists have been trying to develop a test for biopsies to tell us which kind of cancer someone has. Efforts have focused on which mutations are present but that hasn’t led to a useful test so far. We could make more progress by measuring the cells’ “evolvability”, says Greaves. That could be influenced by their genetic diversity, the mutation rate or other factors yet to be discovered.
This is one of many evolution-based research programmes planned by the Institute of Cancer Research. Other avenues will include developing medicines that slow down evolution, by reducing the mutation rate, and a technique known as “evolutionary steering”. This is still a theoretical concept, one that designs drugs in such a way that cells can only develop resistance to them by mutating in ways that make them susceptible to other treatments. “We want to assess candidate drugs by assessing not their ability to kill cells in a dish, but their ability to restrain evolution,” says Greaves.
In the US, the National Cancer Institute began a major new programme focusing on evolution in 2018, setting up the . Initiatives such as these will take many years to bear fruit, but for Greaves, it is the only logical way to tackle this most fearsome of diseases.
“For decades, we’ve been getting new drugs by testing them on cells in tissue culture and it’s not good enough. Every new drug sounds great, but you still encourage resistance,” he says. “We need to change track so we see this as an evolutionary problem. Then, we can try to find an evolutionary solution.”
