Just as rolling yarn into a ball prevents it from becoming tangled, our cells package long strands of DNA into neat bundles called chromosomes.

This helps keep the DNA safe and organised. But in cancer cells the situation can be vastly different – particularly in lung cancer.

Order and stability are exchanged for chaos at the genetic level. Spelling mistakes litter the billions of DNA letters, and larger DNA changes tangle or stitch strands together.

It’s this chaos that makes lung cancer so hard to treat, especially as these errors develop and change over time.

But it now seems that those large-scale alterations to cancer cells’ DNA can help researchers predict the outlook for some patients with lung cancer. The greater the level of chromosomal carnage, the worse a patient is likely to do.

That’s according to some of the earliest results to come from our pioneering TRACERx study that’s meticulously documenting the evolution of lung cancer.

Headed by Professor Charles Swanton, from the Francis Crick Institute in London, this in-depth work has already given us the most detailed understanding to date of how genetic disarray can determine survival in some lung cancer patients. The results are published in the New England Journal of Medicine, and span the first 100 patients out of their 842 target.

To find out more, we caught up with the lead scientists behind this flagship study and asked the important question: what’s next?

Rearranging the genetic ‘furniture’

Cancer is caused by faults in a cell’s DNA, making it to grow out of control. Following landmark work from Swanton’s team, his lab and others have shown that over time, as a cancer progresses, more and more of these mistakes build up, appearing in different parts of the tumour. The end result is a tumour that, on the genetic level, can be vastly different from one region to the next.

These results draw striking similarities to how Charles Darwin famously explained the evolution of species through natural selection. And it’s become clear from recent work – by Swanton and others – that it’s not just small tweaks to DNA that are important in cancer evolution. Big changes affecting chunks of chromosomes have a huge part to play too, with whole lengths of DNA being chopped out, duplicated, or stuck in where they shouldn’t be.

“These complex rearrangements of chromosomes and the mechanisms generating them have largely been ignored in the field of drug development,” says Swanton. “Much research has focused on targeting single changes to the DNA code.

“But work from us and others has shown that this could be a futile approach in late stage disease, because the tumour will inevitably evolve to become resistant to the treatment.”

Studies have attempted to look into this chromosomal chaos before, but have had to do so retrospectively using only one tumour region. This means they’ve looked back at patient samples and taken a snapshot of what’s going on at a certain point in time – providing only a static measure of cancer’s chromosomal complexity.

We want to explore the relationship between a tumour’s genetic diversity and patient outcome

– Dr Mariam Jamal-Hanjani

This is where TRACERx is truly unique. Multiple samples are being collected from each lung cancer patient at diagnosis and through the disease course over time, through treatment and beyond.

This means the scientists can track the cancer’s evolution as it happens, which they hope will achieve two things.

“First, we want to explore the relationship between a tumour’s genetic diversity and patient outcome,” says lead author Dr Mariam Jamal-Hanjani from the UCL Cancer Institute.

“Second, we want to know how chemotherapy affects this diversity, in the context of disease that’s returned.”

Double the trouble

TRACERx is looking at patients with the most common type of lung cancer, non-small cell lung cancer.

For each individual, the researchers are using sophisticated lab techniques to characterise the genetic profile of their tumour. They’re looking for cancer-causing genetic changes in two different forms: changes to individual DNA letters, so called ‘point changes’, or changes to the number of large segments of chromosomes, called ‘copy number changes’.

Across each person’s tumour, they’ve so far found extensive variation in both types of genetic change. Think of this like the development of a tree. One single faulty cell – the tree’s seed – slowly grows over time, forming the tumour, or trunk. These growing cells make up the bulk of the tumour.

We’re beginning to understand the order of these genetic events, helping us create an evolutionary rulebook of cancer

– Professor Charles Swanton

As these abnormal cells continue to grow, further genetic mistakes crop up in individual cells that make them distinct from the rest of the tumour. These then seed the growth of a ‘branch’ of tumour cells with slightly different genetics from the ‘trunk’ of the tumour.

Tracing back to the roots of this diversity, the researchers found that mistakes early on in the cancer’s development were to blame. These mistakes resulted in cells having double the normal number of chromosomes.

“Part of the reason we think lung cancers undergo this process is that it leads to a greater level of chromosomal damage, speeding up the tumour’s genetic evolution,” says Swanton.

“What’s really fascinating is that we’re beginning to understand the order of these genetic events, helping us create an evolutionary rulebook of cancer.

“Genome doubling is a common early event in the tumour’s ‘trunk’, but later on we see mistakes appearing in the ‘branches’ that affect DNA repair mechanisms.”

And being unable to repair faults in DNA would further fuel genetic diversity, which Swanton thinks is “no coincidence”.

“The tumour fixes the massive strain of DNA damage that accumulates over time, for example through smoking, but later on in lung cancer evolution genes involved in the repair of DNA become faulty themselves, which may lead to further mistakes in DNA fuelling more genetic variation,” he explains.

But that’s not the only evolutionary quirk they discovered.

Drawing parallels

When the team dug deeper and looked across the individual tumours, researchers found matching genetic mistakes in different areas of the tumour that had appeared independently of one another.

“For example we could find a particular gene had been doubled on the mother’s chromosome in one particular region of the tumour, and the father’s in another,” says Swanton.

“That’s why chromosomal chaos is so problematic. It allows tumours essentially to get to the same end in numerous different ways, or ‘infinite diversity of structure for gaining the same end’ as Darwin once said, providing a fertile ground upon which evolution can act.

We’re now really getting a clearer picture of how complex these tumours are

– Dr Mariam Jamal-Hanjani

“And ultimately this results in poor outcome for these patients.”

But while it was known that genetic diversity within a tumour is linked with outlook, intriguingly the study found this only to be true for the large-scale chromosomal changes. Greater diversity at the level of small point changes in DNA wasn’t linked with worse survival, but large-scale chromosomal changes were.

“We’re now really getting a clearer picture of how complex these tumours are, and the relevance of chromosomal chaos for patient outlook,” says Jamal-Hanjani.

“And the more tumours we can analyse in TRACERx, the more likely we are to identify and create a timeline of genetic events that could have implications for treatment.

“If we can predict the next step in tumour evolution before it’s occurred, then perhaps we could treat patients before their disease progresses and spreads, or monitor them more closely.”

Keeping ahead of the game

For a cancer that’s hard to treat, the ability to stay one step ahead of its progression is an exciting idea that could dramatically change the outlook for patients. But with the study still in its infancy, at the moment that’s a faraway goal.

In the meantime though, TRACERx researchers have plenty to be getting along with.

“We’ve learnt an awful lot from the first 100 patients, but there is so much yet to come,” says Jamal-Hanjani.

“We’ll be performing imaging studies, looking for markers that could help predict outlook in blood samples collected from patients, studying the tumour environment such as the landscape of immune cells and more. Looking at these aspects over time and linking them with the genetic diversity of tumours will help us develop a global view of non-small cell lung cancer evolution.”

What’s more, the team will be sharing these data so that scientists and clinicians around the world can join in and work towards the same cause.

“This is crowd science,” says Swanton. “We’re at the tip of the iceberg in terms of the groups that will be working on these data in 5 years.

“It’s a massive gift from patients and Cancer Research UK to the clinical and scientific cancer research community.”

With millions of lives lost each year to lung cancer, it’s certainly set to be a gift worth waiting for.



Jamal-Hanjani, M. et al. (2017). TRACERx – Tracking Non-Small Cell Lung Cancer Evolution. The New England Journal of Medicine. DOI: 10.1056/NEJMoa1616288