Patient data use case studies

professor_crispin_miller.jpg

At the heart of the CRUK Scotland Institute is a team of brilliant scientists, led by Professor Crispin Miller, who specialise in using super computers to understand how cancer behaves. 

This diverse and talented team - which includes mathematicians, physicists, biologists and software engineers - are using a range of techniques to dig deep into the huge amounts of data we have on the genetics and anatomy of tumours.   

“Biology is exciting” says Crispin. “The questions of modern cancer research, like how the DNA in one cell can define an entire living person – to me that’s as exciting as understanding the first few microseconds of the Big Bang.”  

Crispin’s goal is to identify patterns in a tumour’s DNA or its structure that might help us better understand why and how cancers develop. And ultimately this could inform how best to treat them. 

But spotting those crucial patterns is difficult because tumours are so diverse. And so Crispin’s team turn to supercomputers. These powerful tools can sift through vast amounts of information to find patterns that would be impossible for a human to spot. Then the scientists can do what they do best – work out how the patterns are driving cancer progression or making a tumour more or less sensitive to an anticancer drug.   

Professor Miller explains: “Data gives us insight into things you can’t see down a microscope.  It has the potential to start developing truly targeted therapies. To find the right therapy for the right person at the right time  that is the goal.”  

His team are using a data approach on many different cancer-related questions.  In one project, patients have kindly allowed their doctors to share anonymised scan images of tumours and the tumour genetic data. Crispin’s team are seeing whether they can build a sophisticated AI approach that has the power to use this data to predict how tumours are going to respond to treatment. This could help doctors pick the best treatment for each person.   

What nearly all Professor Miller’s work has in common is that it is only possible due to the data that people with cancer have shared.  

“Data is central to the research we are doing.  The more tumours we study, the more we understand how they differ from each other and how we can use this information to improve outcomes for people with cancer.  I am incredibly grateful to the patients who donate their data and tissue, I couldn’t do my work without it.” 

“Data is everything” – how patient data is helping us understand how cancer spreads and resists treatment 

For Dr Irene Lobon, everything is data and data is everything. From calculating nutrients in new recipes to studying tumour samples, analysing data is central to her life and work. While in her spare time Dr Lobon’s analytical skills help her to optimise her favourite recipes, in the laboratory, these skills could lead to lifesaving discoveries. 

Dr Lobon is a biomedical researcher in the Cancer Dynamics Laboratory, led by Professor Samra Turajlic at The Francis Crick Institute. She works on a multidisciplinary team featuring scientists and clinicians, all working together to better understand cancer.  

The team are studying advanced, spreading melanoma, called metastatic melanoma. This is a type of skin cancer that is particularly hard to treat and often becomes resistant to drugs. Professor Samra’s group want to know how melanoma changes over time, what happens inside the tumour as it spreads and what features lead to drug resistance. To answer these questions, the team relies on patient data. Specifically, samples from patients who have passed away from cancer. 

To access these samples, the team makes use of the PEACE (Posthumous Evaluation of Advanced Cancer Environment) study, a huge collaborative project, set up to help researchers collect samples from patients who have died from cancer. In this study, researchers take samples from participants during their treatment journey and after they have died. This huge assembly of data has allowed clinicians and scientists to study cancer in a way that was not possible before. 

Professor Turajlic’s team employed sophisticated genetic sequencing techniques to study nearly 600 separate samples from 14 different patients with metastatic melanoma from the PEACE study. “What we’ve observed is that there are many variables that contribute tiny bits to the cancer progression” said Dr Lobon.  

Uncovering these small contributions allowed the team to piece together a complex timeline of how the cancer develops.  “It’s like having a fossil. We can study cancers at different points in time.”  

The team shared the results of this extensive study with the scientific community earlier this year. The incredible information uncovered has helped scientists to understand the complex ways that melanomas use to evade drugs. The researchers hope that it could also lead to the development of completely new drugs that could provide hope for those diagnosed with metastatic melanoma.  

While Professor Samra’s team are using the PEACE study to study metastatic melanoma, there are many other scientists using the valuable data collected to investigate other cancer types. There are researchers investigating lung cancer, renal cancer and many more. What they have in common, is their reliance on patient data. 

Dr Lobon emphasised that her work would be impossible without the generous people who agree to donate their samples after they pass. “It’s extremely important. There is no other way we could get such specific information without these samples.”  

“Data is everything.”  

professor_rebecca_fitzgerald.jpg

Oesophageal cancer is one of the biggest challenges in cancer research today. The underlying biology of the disease is poorly understood and survival remains very low, with only 1 in 10 people surviving their disease for 10 years or more. It’s a big challenge – and one that researchers and clinicians can’t tackle alone. 

For over 10 years, Professor Rebecca Fitzgerald at the University of Cambridge has been spearheading a huge project that brings together scientists, doctors and nurses from across the UK with the goal of better understanding oesophageal cancer.  

Known in the research community as OCCAMS (Oesophageal Cancer Clinical and Molecular Stratification), this impressive programme of work has seen researchers collecting tumour and blood samples from people with oesophageal cancer and decoding the cancer’s genetic sequence so that they have a complete map of the cancer’s DNA.  

These samples provide a treasure trove of information, allowing scientists to see how changes in DNA sequence affects progression of oesophageal cancer. And every person’s data provides a new piece of the puzzle. 

“This really is a team effort,” says Rebecca. “By sharing information, we can see the patterns and the trends and that’s what allows the breakthroughs to happen. Right now we can’t predict how well someone will do just by how they look when we first see them in the clinic. Some people do well, and sadly some people do badly, and we need to know why.” 

And there are a lot of questions that need answering. How do genetic changes in the cancer change the way it develops over time? Are there specific risk factors that could predict if oesophageal cancer is going to recur? Why do some people respond well to treatment and some people don’t? All these questions rely on matching up clinical data (data collected by doctors about how each person’s cancer progresses) with laboratory data (data that comes from samples that scientists take away to analyse in the lab).    

“And we need a lot of data,” explains Rebecca. “Oesophageal cancer is really complicated. We can’t find the needle in the haystack by just using data from a few people. The current problem my team is working on – working out whether all oesophageal cancers start from the precursor condition Barrett’s oesophagus – is using data from 4,000 patients. And that is why we are so grateful to all the patients who allow us to do this work.” 

In the end, this work is all driven by the desire to make the situation better for people with oesophageal cancer. Already the team have been able to describe for the first time some of the DNA mutations that they think cause cancer, and there are likely to be many more findings to come. 

 “We have seen some improvements over the 20 years I have been working on this,” reflects Rebecca, “but we’ve got a long way to go. Solving this problem is what gets me out of bed in the morning. And I want to get to the end of my career and see that it is really different. I can’t do what I do without patient data – and I am not done yet.”   

Personalising radiotherapy – how patient data is helping reduce the side effects of treatment 

Radiotherapy is the gold standard of treatment for many types of cancers. In fact, more than 130,000 patients benefit from radiotherapy every year in the UK. Today, most people receive image-guided radiotherapy – that is, using imaging such as x-rays and MRIs to target the beam of radiation to the tumour site, to make it as accurate as possible.  

“Many people think about imaging at the point of diagnosis to find a tumour, or to follow up on treatment to see if its working,” says Dr Rajesh Jena, a clinician scientist based in Cambridge who is finding ways to improve radiotherapy for people with tumours of the brain and spine. “But imaging can be so much more: it can be used to personalise and even predict response to treatment.”  

With radiotherapy there can be side effects because the beam of radiation can also affect healthy tissue surrounding the tumour. Combining imaging while delivering radiotherapy can help minimise these side effects, and getting better at combining these approaches is where patient data can be so important.  

By studying images taken during radiotherapy, researchers can determine which healthy tissues were affected by the radiation, and then use that information to help predict how a new patient might respond to treatment, potentially altering the treatment plan to make it as side-effect free as possible. 

“The wealth of information an image holds comes with an incredible potential,” says Dr Jena. Such potential to inform patient care led Dr Jena and his team to develop the first continuously learning artificial intelligence (AI) medical device in his hospital. Using patient data to train the AI, they were able to develop a tool that speeds up analysis of images. The technology helps doctors by cutting the amount of time spent ‘sketching’ around healthy organs, as they create radiotherapy plans, a painstaking yet vital step in radiotherapy treatment to ensure healthy tissue is protected.  

And this data offers value far beyond direct medical treatment. “It’s not only doctors that need access to imaging data like this,” Dr Jena adds. “Patients are so generous allowing us to use this data, so we try to maximise its value by working with mathematicians, physicists, biologists and many other experts.” Bringing together different specialisms like this gives the scientists the best chance of understanding a disease as complex as cancer. 

“And we also look back at historical patient images – this is important to provide context, improve our understanding and build up a valuable database of information.” 

Thanks to patient data, Dr Jena and his team have already been able to create a technology that frees up valuable time and ultimately enable patients to receive treatment sooner. Over the next year, Dr Jena and his team are working on improving this technology by providing the AI with even more examples of patient images to learn from, allowing it to become better than ever at identifying healthy tissue.  

“We’re so grateful to have access to patient imaging data because without it we wouldn’t be able to develop technologies like this, which can make a real difference to people with cancer. We’re very careful with how we use patient data, we take care to ensure that all data is anonymised and is treated with respect.”