Skip to main content

Together we are beating cancer

Donate now
  • Science & Technology
  • Health & Medicine

Improving brain tumour diagnosis to make treatment personal

by Justine Alford | Analysis

24 August 2018

0 comments 0 comments

An illustration of phone software to match brain tumours to treatments

In a 6-part series, we’re exploring the major challenges that are holding back progress in the field of brain tumour research. This fifth instalment focuses on how better characterising brain tumours can improve diagnosis and ultimately make treatment more personal.

Imagine a pocket watch. The face tells you the time, but only by peering inside and seeing the cogs, springs and screws can you find out how it ticks. Armed with an understanding of these inner workings, experts can diagnose what has gone wrong when they fail.

The same is true for brain tumours.

Examining what a brain tumour looks like offers a timestamp, pointing to how long it may have been growing and the type of tumour it might be. But taking these diseases at face value risks missing important information. It’s the hidden detail that researchers are now turning to, examining faults in the molecular cogs, springs and screws that make these tumours tick. Not only will this help refine brain tumour diagnosis, but experts hope it will also point to potential new fixes.

That’s why, as part of our inaugural Brain Tumour Awards, we’re calling on researchers to develop ways to do precisely this, so that improved diagnostics go hand in hand with better treatment.

The power of combination

Brain tumour diagnosis starts with a scan. Imaging techniques, such as MRI and CT scans, have become incredibly sophisticated over the years and can precisely pinpoint a mass of cells in the brain that shouldn’t be there. But these scans can’t say exactly what type of cells the mass is made of, particularly if they’re cancerous or benign. So, more detailed analysis is necessary.

Doctors or scientists often look at a sample of the tumour under the microscope and study the cells’ appearance. This is called ‘histopathology’, and it’s a mainstay of cancer diagnosis. But there’s human error at play here, because appearances can be deceiving.

“Histopathology is an important part of diagnosis; it often gives us the right broad diagnosis on a particular cancer,” says Professor Steve Clifford, a Cancer Research UK-funded brain tumour expert at the University of Newcastle.

But simply looking at the shape of cells can only tell scientists so much. The genes and faulty molecules inside those cells have a big part to play too.

“We’ve come to understand that the molecular biology of a cancer is critically important in determining the nature and clinical behaviour of the tumour,” adds Clifford.

Professor Louis Chesler, a Cancer Research UK-funded expert on children’s cancers at the Institute of Cancer Research, London, agrees: “When you combine molecular diagnostics with histopathology, it’s very powerful. I think that’s a way forward to better understanding.”

This idea is backed by a recent change in how the World Health Organisation (WHO) views brain tumour diagnosis. For the first time, molecular features of brain tumours have been included alongside histopathology in the WHO Classification of Tumours, which Clifford refers to as the pathologist’s “bible” for diagnosing cancer.

Collaboration is key

Cells may be tiny, but they’re packed with various molecular clues waiting to be discovered by scientists, including:

  • Billions of ‘letters’ that make up a cell’s DNA
  • Patterns of tags on this DNA that switch genes off or on
  • Protein molecules made from the DNA
  • Biochemical reactions taking place inside the cell

Collectively these features are called ‘biomarkers’, and they may appear different inside cancer cells than inside healthy cells. Understanding these differences, and finding ways to spot them, could help better classify brain tumours. But combing through a long list of features in every patient would be an arduous and expensive process, which may not bring any practical benefit in the clinic.

Coordination and collaboration is really important to make progress.

– Prof Chesler

Clifford says that while all these different biomarkers are potentially important, the key is to discover how to get the best answers in the most efficient way possible.

“We need a concerted approach to find out which method will give us answers,” he adds. “We need to be as open and comprehensive as possible.”

Being open and sharing resources will also help scientists learn faster, says Chesler. “We need to make sure that everyone is following the same methods as much as possible, because otherwise it becomes difficult to interpret the results.

“With any cancer where you have a very small number of patients, coordination and collaboration is really important to make progress.”

Blood, sweat, but no tears

The fact brain tumours are relatively rare makes it challenging to gather enough samples to draw firm conclusions. But this challenge is multiplied further by where these tumours grow.

The brain is a vital yet highly fragile organ, and sometimes taking biopsy samples or performing surgery is too risky for the patient. Just like how a watchmaker would struggle to find the root of a fault using just a photograph of a watch, without tissue to analyse, making an accurate brain tumour diagnosis is difficult. But both Clifford and Chesler highlight the potential of less invasive tests known as ‘liquid biopsies’.

“Brain biopsies are expensive and dangerous, particularly for children,” Chesler says. “So increasingly we’re looking at blood, spinal fluid, saliva, sweat – anything we can.

“A major question is whether you can find brain tumour DNA in these samples. If it’s present but in tiny amounts, can we make a test that’s good enough to be informative? That’s why engineers and physicists are now working together to help develop technologies to solve this problem.”

Time to get personal

By gathering these pieces of information, the aim is to make brain tumour diagnosis more accurate. This should then help guide treatment decisions in the clinic.

Medulloblastoma is a great example of this. Work by our scientists and others discovered that this childhood brain tumour can be grouped into 7 distinct types, and that these differ not only in their biology but also how well patients fare. One type has a very good outlook and most children survive, whereas the others are trickier to treat.

Figuring out which type children have is now standard practice for this disease. Clifford says that clinical trials are also underway to see if children with the better outlook can be treated less intensively. The goal would be to spare them from side effects without sacrificing the likelihood of being cured.

The bottom line is that we’re getting better and better at diagnosing brain tumours.

– Prof Clifford

Better diagnosis could also help treatment in other ways. If a tumour is found to have a biomarker that can be targeted by an existing drug, patients could get treatment tailored to their disease. Or if a potential drug is still being tested out in trials, this information will help guide patients into clinical studies that are most likely to benefit them. And, importantly, patients with similar brain tumours in the future.

In a nutshell, that’s precisely what one of Chesler’s studies is trying to achieve for children with cancer, including brain tumours.

With around 130 different types of brain tumour to study, scientists like Chesler and Clifford have their work cut out. But as time goes by, incremental progress is made, and hopefully funding boosts like the Cancer Research UK Brain Tumour Awards can accelerate this.

“The field has progressed beyond infancy, but perhaps it hasn’t yet reached full adulthood,” says Clifford.

“The bottom line is that we’re getting better and better at diagnosing brain tumours. But by no means are we entirely there yet.”