A powerful new data platform unlocks the possibility of personalised medicine

As you might expect from the name, rare diseases are rare. But the numbers soon start to stack up: while any one rare disease affects fewer than one in two thousand people, there are thousands of different conditions. This adds up to around 7 per cent of the population or one in 13, affecting an estimated 300 million people around the world.

Adding to the problem, most patients with rare diseases take up to five years to receive a diagnosis for their illness, during which time they and their family have to live with uncertainty and delays with starting treatment.

One of the key issues lies with the fact that a doctor might only see a handful of patients with rare diseases in any year and might only encounter a particular condition once in their entire career. Although clinicians may write case reports about individual patients, much of the information about the diagnosis and treatments of rare diseases remains locked away in confidential medical records and obscure scientific papers spread all around the world rather than being organised and shared in a way that could benefit patients.

“Around five years ago, we decided that this had to stop,” says Ivo Gut, director of the CNAG-CRG. “These patients are often young children, which puts a huge burden on their parents – caring for their child and running from doctor to doctor in search of answers takes a huge amount of time, money and emotional energy.”

“We realised the need to build an online system that would enable doctors to work faster and more effectively to deliver a diagnosis or potential treatment for rare diseases, and to find other similar cases elsewhere that might provide vital information and end this diagnostic odyssey.”

Gut teamed up with Sergi Beltran, head of the CNAG-CRG Bioinformatics Unit, to develop the Genome-Phenome Analysis Platform within the RD-Connect project, an international research collaboration for rare diseases. The cutting-edge global data platform allows doctors to store, analyse, and securely and ethically share all kinds of information about their patients with rare diseases.

Genomic data plays a significant role, as most rare diseases are caused by specific inherited genetic variations, but that’s not all. Other types of information include the levels of various molecules in the blood, as well as physical symptoms and measurements. All of these datasets come together to provide a detailed picture of exactly how the changes in an individual patient’s genes affect their body and cause their disease.

RD-Connect now contains information about more than 5,000 individuals and has led to the identification of hundreds of new genes that are implicated in rare diseases, slashing the time taken to diagnose most rare diseases to around a year. Importantly, it has also brought certainty for many families that have been desperately waiting for answers.

Having a specific diagnosis for a rare disease provides a name and a cause for a condition, unlocking access to financial help and other types of support. In some cases, although not all, a confirmed diagnosis may point towards a specific treatment, such as drugs, dietary changes or high-dose vitamin supplements. And it can provide a reason for something that was previously a frustrating mystery.

“Knowing what is wrong is very important for families,” explains Gut. “They may worry that their child is sick because it’s something they’ve done – they think it’s their fault or they made a mistake. We can say, ‘no, you were just unlucky,’ which helps to calm their spirits.”

Importantly, the RD-Connect platform provides a way to safely, securely and ethically share and search data, allowing clinicians to find information that might help their own patients.

“For diseases that are very are, it’s hard to know how best to treat diseases that are extremely rare,” says Gut. “Our platform effectively provides access to all the world’s doctors. Clinicians can find information about what works – and what doesn’t work – for similar patients by sharing details about best practice and treatments.”

The platform’s technology allows users in other locations to ‘dial in’ and search for similar cases – or even help to diagnose patients in other hospitals – without having to move data between institutions or across borders. It also provides a standardised way of gathering genetic and clinical information and reporting results back to doctors and genetic counsellors, ensuring that everyone is speaking the same ‘data language’ when it comes to diagnosing and treating patients.

Five years on, there are currently more than 400 doctors using the RD-Connect platform, inputting and analysing data to help diagnose and treat their patients with rare diseases. The CRG-CNAG team is also working with collaborators in Finland and other countries to establish the system as part of their national healthcare.

It is also being used in two major European projects on rare diseases, Solve-RD and the EJP-RD, connecting doctors and researchers across Europe to analyse, share and diagnose thousands of their most challenging unsolved rare disease cases.

In Spain, the CNAG-CRG is involved in several regional projects related to Personalised Medicine. URD-Cat (Undiagnosed Rare Disease Program of Catalonia), led by  Luís Pérez-Jurado, is working to diagnose 1,000 patients with rare neurological diseases in seven Catalan hospitals, and the Navarra 1,000 Genomes Project, coordinated by Ángel Alonso, is a pilot project for diagnosing rare diseases and treating cancer within the health service, based on using genome sequencing as a front-line diagnostic test.

“The Navarra 1,000 Genomes Project is a genome-first approach, seeing how we can make this kind of genetic testing available to anyone in the healthcare system,” explains Beltran. “We have customised the platform for them, collecting phenotypic and genomic information from 1,000 people with rare diseases or cancer, processing the data in a standard way and making it available through the platform, so doctors can work out what best to do for their patients.”

However, the potential benefits of the tools that the CNAG-CRG team have built extend beyond rare diseases and into other health conditions, such as cancer.

The CNAG-CRG researchers have teamed up with Núria López-Bigas at the IRB Barcelona and David Torrens at the Barcelona Supercomputing Centre for the Bioinformatics developments in MedPerCan – a pilot project for personalised medicine in Cancer in Catalonia.

Run across three major hospitals in Barcelona, the project is using an adapted version of the underlying RD-Connect technology to set up a workflow for diagnosing and treating cancer patients in a more personalised way, cataloguing the underlying gene faults that are driving the disease and identifying potentially useful therapies.

All the DNA sequencing for URD-Cat, MedPerCan and the Navarra 1,000 Genomes project is being done at CNAG-CRG, highlighting the vital importance of the centre in supporting precision medicine within Spain.

Not only could this new platform help to diagnose rare diseases and treat cancer more effectively, it also looks at genetic variations that affect how people respond to and break down certain drugs (known as pharmacogenomics). And it can also provide information about the broader pattern of genetic changes that are present within an individual’s genome, which could be relevant when planning a family.

“What is medicine for one person is poison for another, so what makes the difference?” Gut asks. “Is it genetics? Psychology? Food and lifestyle? Exercise? Sleep? Systems that can correlate all of this information and crunch it all together, including data from patients themselves, would really move us forward. That’s the future of where this will go – collecting as much information about an individual’s life and health as possible and figuring out what is best for them and why.”

“The system is all about organising and interrogating data about genes and their effects on health, rather than focusing on any particular disease,” Beltran points out. “We are now adapting the RD-Connect platform for providing insights about the best way to treat someone based on the genetic and molecular makeup of their disease – this is personalised or precision medicine.”

For example, while there may be many patients affected by the same type of cancer, each case is an individual disease with its own biological quirks. Currently, patients are given treatments according to the physical and molecular characteristics of their tumour. But these are still relatively large groups, and some people’s cancers respond well to therapy while others do not.

Understanding the complex interplay between the underlying genetic and molecular characteristics of the disease is likely to point towards therapies that are more likely to be effective – something that the underlying technology of RD-Connect is exactly designed to do. Yet despite the success of their system today, Beltran and Gut were convinced that they were lagging behind the field when they first came up with the idea to build RD-Connect.

“We thought we would be playing catch-up with other platforms that were already being built,” Gut says. “But even now, although lots of people are talking about doing something like this, there is still nothing comparable out there.”

“I’m extremely proud of our team,” add Gut. “It’s been a huge process and a lot of work for our engineers, and I’m also very grateful to the first users who helped us to improve and debug the system. Five years on from having started this project, it has turned out to be exactly the right thing to do and I’m delighted with what we have built.”