An image of a piece of DNA.
There’s more to history than an interesting story. Studying the past allows us to learn from it, and understand what the future may hold.
The same is true of cancer.
By studying the disease’s past, scientists hope to predict the path that cancers may take in patients in the future. And this could help edge us towards more personalised treatment.
For lung cancer, that tantalising possibility was highlighted last year by a duo of landmark studies from the Cancer Research UK-funded TRACERx project, which is tracking the disease’s genetics over time. Now, a trio of scientific papers published in the journal Cell has unravelled the same details in kidney cancer, plotting the differences between patients’ cancers as they develop.
This could help us to predict how an individual’s cancer may progress.
– Dr Samra Turajlic
“What we’ve done is work out the exact order in which the critical genetic events occur, as well as their timing – whether they occur early on in the life of the tumour or very late after the bulk of the tumour has grown,” says Dr Samra Turajlic, from the Francis Crick Institute and the Royal Marsden Hospital, one of the lead researchers behind the studies.
“In doing so we’ve shown how a tumour’s evolutionary path influences its behaviour. In the future this could help us to predict how an individual’s cancer may progress, distinguishing those that are likely to spread rapidly from those that grow slowly and may never spread.”
Reading cancer’s diary
As a cancer grows and develops its genetic make-up changes, becoming more damaged and distinct from the healthy tissue it arose from. Over time, the cancer cells even become different to one another, giving rise to an evolving ‘patchwork’ rather than a uniform bundle of cancer cells. This presents both a challenge and an opportunity to researchers.
To produce a genetic profile for a cancer scientists need a sample, or biopsy, of the tumour. But they usually only have access to one sample, which could give a misleading snapshot of these ‘patchwork’ tumours.
So, if scientists have more samples of a tumour, the genetic diversity can be fully mapped, acting as a hidden record of the tumour’s evolutionary history. If a genetic fault is found in every sample, then it must have appeared early in the tumour’s history, when a small number of cells were multiplying into a larger pool of genetically identical copies. Whereas those gene faults that are only found in one or two tumour samples must have popped up later in the cancer’s evolutionary journey. It’s these genetic rules that the latest research is documenting.
Secrets to sequences
Led by Professor Charles Swanton, Cancer Research UK’s chief clinician and group leader at the Francis Crick Institute, the scientists began by looking at the DNA from over 1,000 samples taken from the primary tumours of 100 patients with the most common type of kidney cancer, clear cell renal cell carcinoma. This allowed them to turn back the clock and work out the different evolutionary journeys these tumours took as they grew.
“We found that we could classify kidney cancer into seven evolutionary subtypes, according to a set of rules based on the sequence and timing of genetic changes,” says Turajlic.
“But we could also more broadly group them into three categories, based on two characteristics: the extent of genetic diversity seen across the tumour, and how much large-scale damage – where large chunks of DNA are lost or gained – the tumour harbours.” And by marrying this information with patient data, the researchers made connections between the complex genetics and how each patient’s disease behaved.
We found that we could classify kidney cancer into seven evolutionary subtypes.
– Dr Samra Turajlic
At one end of the spectrum were tumours that look very similar with little genetic damage. These tumours were slow-growing and unlikely to spread, meaning patients tended to have a very good outlook.
“This suggests that these tumours could be monitored over time, rather than treated upfront,” says Turajlic, meaning that some people could potentially be spared surgery.
At the other end of the spectrum, the scientists found tumours that aren’t genetically diverse, but which show extensive damage to packages of DNA – called chromosomes – found inside the cancer cells. These tumours tended to grow and spread – or metastasise – very quickly, and patients had a poor outlook.
“These tumours aren’t diverse, but they acquire everything they need to become aggressive early on in their development, through large-scale genetic changes like whole chromosomes being lost or gained,” explains Turajlic.
“These cells grow rapidly and don’t evolve further. They are ‘born to be bad’ and are probably metastatic before they are diagnosed.”
Once again, this finding could have implications for treatment in the future. Because the disease is likely to spread widely and rapidly, removing the primary tumour even when it seems to have only spread locally is unlikely to help the patient survive. On top of that, surgery in these circumstances could delay potentially more effective drug treatments, which may help slow down disease progression. These tumours also highlight the need for ways to detect the disease early.
So what about the tumours that fall into neither of these categories? Sitting in the middle of this evolutionary spectrum were genetically diverse tumours that vary in how much damage their DNA carries. And it’s this diversity that may direct the disease down several different courses.
During their evolution, some cells within these tumours may develop genetic changes that allow them to spread, while others never accrued this damage. That means the patterns of how patients with these tumours fared varies, since only certain populations of cells within the tumours can spread.
The team found that people with these tumours benefited from surgery to remove the original tumour, even after it had spread. This included removing those tumours that had spread, because in these patients the disease often made a new home in just one other part of the body.
But is it possible to spot cells likely to spread before they do? By comparing patients’ primary tumours with samples taken from where they’d spread to, the researchers were also able to identify the cells, and patterns of genetic changes, that were common to both.
“We identified that what distinguished cells that were capable of metastasis was the presence of large-scale genome damage, and in particular two chromosomal regions that when damaged gave rise to cells that are highly capable of spreading,” says Turajlic.
“If we could look for these changes in patients’ original tumours, then in the future we would be able to predict which patients have a high risk of metastasis and offer them additional treatment and close surveillance,” she adds. In doing so, the team hopes this may stop a patient’s disease from worsening.
Marking the beginnings
To paint a complete picture of kidney cancer evolution, the team also wanted to unravel the earliest genetic events in a tumour’s history. To do this, a team led by the Wellcome Sanger Institute rewound the evolutionary clock on key genetic events, revealing a surprisingly drawn-out journey.
The majority of these kidney cancers carry a distinctive genetic fault where part of chromosome 3 (we have 23 pairs of chromosomes inside our cells) gets chopped off and lost. It turns out that this event often happens at the same time as another chromosomal mix-up: part of chromosome 5 gets accidentally copied and duplicated. And the research showed these events are the earliest triggers for a tumour to grow, happening up to 50 years before a patient is diagnosed.
It’s possible that within 10 years we’ll be using this evolutionary information to help guide treatment decisions. That’s what we want to achieve.
– Dr Samra Turajlic
“It’s actually likely that most people have a small number of cells in their kidney that have lost this part of chromosome 3, which happens during their teenage years,” says researcher Dr Tom Mitchell from the Wellcome Sanger Institute.
“But these people will only go on to develop the disease many years later if their DNA gets a second hit, a genetic fault in a gene called VHL.”
Such a long gap between these genetic events and cancer forming could again suggest an opportunity to intervene. But with this research still very much in its infancy, there is a long way to go before that could become a reality.
For now, the researchers have plenty more digging to keep them occupied.
“We are continuing to analyse the tumours from the patients recruited into the TRACERx Renal study – 300 in total – but we are also looking at additional groups, including patients that have been in active surveillance for very early stage tumours,” says Turajlic.
“We’re also beginning to look at how the tumours evolve in response to treatment, and whether we might be able to predict which tumours will be sensitive to drug therapy. It’s possible that within 10 years we’ll be using this evolutionary information to help guide treatment decisions, both medical and surgical, in the clinic.
“That’s what we want to achieve.”
Mitchell. T. et al. Timing the landmark events in the evolution of clear cell renal cell cancer: TRACERx Renal. Cell (2018)
Turajlic, S. et al. Deterministic evolutionary trajectories influence primary tumour growth: the TRACERx Renal study. Cell (2018)
Turajlic, S. et al. Tracking renal cancer evolution reveals constrained routes to metastases, results from the TRACERx Renal study. Cell (2018)