Building a ‘Google Street View’ for cancer

Cancer Research UK recently set out to find the current major challenges that need combatting in order to beat cancer. They presented seven challenges that would need to be overcome, and in this blog, originally posted on their website, they focus on the fifth: mapping cancer.


‘You have reached your destination’: five words that have become synonymous with how technology helps us get around.

And whether it’s via satellites orbiting earth or small cars carrying cameras, complex digital mapping has made navigating our world much simpler. With a few taps of our fingers we can zoom in on our nearest post office, or out to the borders of continents.

For years researchers have used fancy microscopes and scans to map tumors in similar ways, hoping that this could reveal new avenues to treat the disease.

Alongside this, scientists have also meticulously broken tumors down to study the masses of cells that form the rogue ‘towns’ and ‘cities’ that make up cancer.

These two different research worlds have led to big strides in the ways we both diagnose and treat cancer. With the latest imaging techniques, doctors can see the size of a tumor, helping to plan procedures like surgery or radiotherapy.

And with samples of cells (biopsies) we can begin to understand the faulty molecules that have led groups of cells down the path towards cancer, pointing to those which could also be the targets of new treatments.

But putting all this complex information together in enough detail has been tough. Our map isn’t complete. We can see the border of the city, but we have no idea how the different boroughs within it are connected. And while we know there are bad neighbourhoods, pinpointing the rogue cellular communities and understanding how they corrupt their neighbors has been just out of reach.

Our fifth Grand Challenge hopes to change that.

Through a combination of next-generation technology, and expertise ranging from physics and maths to biology and computing, we want to build the ultimate map of cancer.


A new type of organ

“We’ve recognized for a long time that a tumor isn’t just a collection of cancer cells,” says Dr Rick Klausner, former Director of the US National Cancer Institute and chair of our Grand Challenge Advisory Panel.

Instead, he sees a tumour as much more like an organ, growing from our own cells gone bad.

“For a century, all of our scientific research has aimed to describe cancer by isolating the cells,” he says.

“That makes perfect sense. And we’ve made great progress.”

But in the background, says Klausner, we’ve always known that to truly understand how the disease behaves, we need to see it as an organ.

By working with isolated samples, researchers are actually measuring an average of the genes and faulty molecules inside it, potentially missing some of the finer details about how the cells carrying these faults work. So while they have been aware of the scale of the problem, the technology has been missing to fully understand it.

“The reality is the vast majority of our technologies in research involve measuring molecules in bulk,” says Klausner.

“What we do is we measure the total amount of say, DNA, or the total changes in DNA in a tumor. The tumor, with all of its cells and all of its neighbor relationships, is lost. We do everything to lose the mapping, and we just ask what the total content is.”

Rick Klausner
Rick Klausner

He likens this to the logistics of warfare: “Imagine it like a map of a country you were trying to attack. There’s an army somewhere in the country, and you know that some people in the army are carrying guns. But, most importantly, you don’t know exactly where they are.”

In terms of cancer, you might imagine that the soldiers carrying weapons are the more aggressive tumor cells that we need to understand and target. Without knowing where they are we can’t truly understand how they work, and also collude with other cells.

So the challenge is to find a way to accurately measure the size of an army’s arsenal – the combination of faulty genes and molecules that may fuel a tumor’s growth and spread – while also taking stock of where it’s located within the tumor’s ever-changing terrain. And, crucially, Klausner is confident that the latest technology is now able to start homing in on the challenge.

“We have the beginning of technologies that allow us to look at all these aspects of the organ without destroying its structure,” he says.

And he offers a tantalizing glimpse of what this technology could uncover. “I think there’s a great chance that we’re going to discover that, in tumors, there are types of cells that today we don’t even know exist in our body.”

Klausner speculates that these could be: “normal cells that have been re-educated and changed by the tumor itself. The value of this would be extraordinary”.

But you can’t draw a map without looking at the terrain. And being able to measure the extent of an army’s arsenal only answers one part of this challenge.

It’ll take new ways of looking at tumors – through advanced microscopy and scanning techniques – to help pin down the locations.

A new revolution

“If you think about the history of medical imaging and its use in the clinic, we’ve been quite good at adopting new technologies,” says Dr Sarah Bohndiek, a physicist and expert in cancer imaging from our Cambridge Research Institute. “For example, ultrasound was developed following the use of sonar in the First World War.”

And according to Bohndiek, it’s thanks to this post-war revolution that we have the tools we need to gather images of a patient’s entire body. For cancer, this has offered a clearer picture of where a tumor is, which is vital for surgery and radiotherapy. But that picture could be telling us more. And Bohndiek predicts we’re on the cusp of a new revolution that could give us the answers.

“We’re now having a second revolution, which is imaging particular areas of tumors or cells at very high resolution,” she says.

“Now we can look within single cells and see individual proteins interacting, and see the DNA being made.”

And being able to view cancer on two very different scales – where the tumor is in the body, and what’s going on inside its cells – means we have a much better chance of mapping a tumor’s inner workings.

We’re now having a second revolution, which is imaging particular areas of tumors or cells at very high resolution.

Dr Sarah Bohndiek

“We have our whole body imaging that we would classically think about in the context of cancer,” says Bohndiek, “and we also have these microscopy approaches, which can be applied to single cells and zoom right in.”

Combining these technologies could put the missing layer of the map within reach. “We’ve not been able to connect these bits of information before,” says Bohndiek. “And I think now is a great time to try.”

But there are some big technological hurdles ahead. “Imaging is always a massive trade-off,” she adds.

To understand why, imagine zooming in and out of the map on your smartphone. If you zoom in you can see each house on a street, but you can’t see where that street is located in the city. Equally, a city-wide map won’t be able to tell you the colour of the door at number 65.

“You’re either trading off your sensitivity, or you’re trading off your spatial resolution, or you’re trading-off the measurements you can make over time. You can’t really have everything, and that’s one of our biggest challenges,” says Bohndiek.

One of the key things that scientists need to overcome is settling on exactly which information is most useful. And putting that together will require some expertise in handling big data.


Big data

“The quantity of data is certainly a problem,” says Dr Andrew Steele, a computational biologist from the Francis Crick Institute in London.

Steele and his colleagues are used to handling vast amounts of genetic data, mining it for clues about what might trigger cancer and other diseases. They focus on all three billion letters of genetic code we call the human ‘genome’. So the team is all too clear on the data challenges this brings.

In the background to all this, although it’s nice to think of it as a Google Street View, it’s a Google Street View of a very rapidly changing town.

Dr Andrew Steele

“Say there’s a genome for each sample, and say that genome is approximately the same size as a normal human genome, every single one of those samples is going to be a gigabyte of data before you even start,” he says. “That’s a lot.”

According to Steele, the challenge will lie in understanding what’s changed in a tumor compared to its healthy surroundings. But he believes that connecting together data that covers the differences between cancer cells, healthy cells and even the cancer cells themselves “is as much a conceptual challenge as it is a computing one”.

“You want to create a map that somehow explains what the cancer’s doing and how the genome connects to that. That’s going be very, very challenging I think.”

“In the background to all this, although it’s nice to think of it as a Google Street View, it’s a Google Street View of a very rapidly changing town,” he adds.

But there’s no better time to try and build this map. From a technological and a financial perspective, reading all the genetic information held within a cell’s DNA has never been easier or cheaper.

“The cost of DNA sequencing has plummeted, as has the computing cost of processing that data” says Steele.

“Where previously you’d be lucky to have a DNA sample from a person’s tumor, now it’s potentially cost effective to analyze multiple samples from different parts of a patient’s tumor at different times during the cancer’s progression. And that’s just something that wasn’t financially possible before.”

The patient perspective

We won’t understand how tumors function until we understand why all the cells are there, how they got there, and what they are doing. At the moment the picture we have is fragmented. This challenge encompasses so much science, involving technology across disciplines. Everything has to work together for this challenge to be successful. Being part of the Grand Challenge is the icing on the cake for me. Not only do I have the opportunity to hear what the challenges are, but also to explore innovative solutions. From the patient and public perspective we often have insights and expertise that complement and add value to the research. We have ideas that can relate the research to problems faced by patients and the wider public; or ideas for improving how the data is collected, analyzed or reported. There is nothing as important as ensuring that the patient voice is heard.

– Helen, member of the Grand Challenge patient panel

The prize for successfully combining the technology and data is huge. “If this works, this is what all clinical pathology labs will look like. These are the tools they will be using in the future,” says Klausner.

And to do this, it will take experts from across multiple scientific disciplines.

The right people

“The research community will have to come together,” says Klausner. “We need to evolve our machinery, evolve our technology.”

“It’s about getting the people that are fantastic at imaging together with people that know how to measure molecules, and with the technologists that know how to read RNA, DNA and protein, and the specialists in cell biology, cancer biology, immunology, inflammation.

“That’s the great thing about these challenges. We will discover the true life of a tumor, which we haven’t ever done before.”

Getting this right will bring our destination into sight. And solving this challenge will help plot the routes we need to take to reach it.

View the latest posts on the On Medicine homepage

One Comment

Michael Da Pisa

As a Cancer survivor I am always interested in reading about the efforts being made by researchers and doctors to find answers in how to fight this dreadful disease. Thanks to people like Dr. Peel and Dr. Klausner we may see the day when Cancer is totally understood and defeated.

Comments are closed.