Some illnesses follow a set course, making them easy to treat. With measles, your doctor will likely tell you to: “stay home, keep away from people you might infect, and drink lots of fluids. Take paracetamol or ibuprofen if you have a temperature or any pain and try a steam bath if you get a stuffy nose. You should feel better in 7 to 10 days.”
If you are unlucky enough to be diagnosed with prostate cancer, however, the treatment plan won’t be nearly so clear-cut. As Dr Justine Alford of Cancer Research UK explains: “A major hurdle to making further progress against prostate cancer is the lack of ways to accurately predict how a person’s disease will progress, making it challenging to know which treatment is best for each patient.”
But what if you could personalise medical treatments? Even now, clinicians are working with data scientists to analyse massive health data sets, hoping to identify genetic, environmental or lifestyle indicators that can be used to customise future treatments. It’s a practice called precision medicine.
In the case of cancer, big data is being used to analyse the genetic makeup of patients, with the aim of identifying molecular drivers of disease. One large study combined the genetic information from 112 men with data from other studies, covering 930 subjects. Big data was used to map out the genes of the patients, then that information was fed into the world's largest database for cancer drug discovery (canSAR) to try and identify pathological mutations and pair them with suitable drugs.
Paul Workman, a co-author of the study run by the London Institute of Cancer Research, explains its significance: “This study has uncovered a remarkably large number of new genes [80 in all] that drive the development of prostate cancer and given us vital information about how to exploit the biology of the disease to find potential new treatments.” He hopes that further studies will enable the development of bespoke drugs to fight specific disease mutations and save lives.
One person who is sure to be eagerly awaiting such advances is Intel employee Bryce Olson, who was diagnosed with Stage 4 prostate cancer in 2014. Soon he had exhausted all possible treatments. “Facing my own mortality, I decided to go back to work at Intel, transferring into the health and life sciences division in the hopes that I could make my final days count. It was a move that literally saved my life,” he says .
On learning that his division was working with the OHSU Knight Cancer Institute, using computers to find ways to sequence genomes quickly, cheaply and securely, Bryce offered himself up as a test subject. The sequencing revealed the genes responsible for driving Bryce’s particular form of cancer, allowing his doctors to get him into suitable drug trials.
Big data could help refine targeted treatments and potentially head off disease before it starts.
Although Bryce’s cancer is now back, as his body has developed resistance to the drugs being used, he considers himself lucky to have bought a few extra years. He is also excited about the potential of the research – conducted through a mixture of data analytics, artificial intelligence, and clinical advances – to not only refine targeted treatments but potentially head off disease before it starts.
“In my case,” says Bryce, “what would have been nice is if they would have been able to say, ‘You don't have prostate cancer yet, but we can see a molecular driver or a mutation that is associated with a high probability that you're going to get this, so let's take action early.’”
Big data may not be able to prevent cancer from occurring yet, but it is already being used to help people to manage other diseases and chronic conditions. Diabetes, for example, can be significantly impacted by diet and exercise, but sufferers often lack the motivation and know-how to make the necessary lifestyle changes.
Dr Maxim Osipov works for Sentimoto Ltd., the developers of My Kin, an app that delivers precision healthcare as opposed to precision medicine. As Osipov explains, My Kin aims to prevent people from getting sick by analysing unique factors of their body or life, rather than tailoring treatments to match their pathology.
It’s an uphill battle. We are complex creatures. “Everything – from the stress experienced by [your] mother during pregnancy, to exposure to natural environments in early years, education, and the entire history of [your] health… (can) contribute to risks of developing complex health conditions later in life,” Osipov says.
That said, “these factors are modifiable – that means behaviour change may lead to reduction of health risks. As much as 90% for the risks of type 2 diabetes, more than 50% for coronary heart disease and even 30% for dementia.”
Sentimento’s app provides advice based on a continuous flow of information collected from smartphones and wearable technology. As well as physical activity, the company monitors social activity, sleep, and environmental factors.
“Our vision [at Sentimoto] is that very soon people will understand how each of their actions and behaviours affect their health and wellbeing. And this will mean longer, healthier and more meaningful lives for billions.”
Speaking of saving billions, one data-driven project by scientists in the US, hopes to do exactly that. The aim is to review mathematical modelling research to estimate the global impact of antibiotic resistance, a problem that the World Health Organization has declared “one of the biggest threats to global health, food security, and development today.”
At a recent health symposium (“Data-Powered Strategies to Counteract Antibiotic Resistance”), Harvard assistant professor of biomedical informatics Maha Farhat said that drug-resistant microbes threaten “one of the most core therapies that define modern medicine.” But she feels the use of big data could be the key to fighting antibiotic resistant superbugs.
Clinicians hope to get the issue referenced in the Global Burden of Disease Study, a report of global patterns of mortality that aims to identify the biggest threats to worldwide health. If accepted, mass funding will likely be provided by policymakers to tackle the problem.
It looks certain that the future of healthcare will be shaped by clinicians and data scientists working together to develop new and impactful treatments. As Intel’s Bryce Olsen explains: “Precision medicine is all about big data… Analytics help decode what's fuelling that disease… Then you move to doing the clinical interpretation, to identify the right treatment path for an individual. It takes massive amounts of data, and going forward, a lot of artificial intelligence, to make this possible.”
And as the tools used to analyse the massive pools of medical data improve, such studies will become ever faster, cheaper, and easier to run.