How Sophia Genetics uses genomic healthcare to diagnose NHS patients

© SOPHiA GENETICS
© SOPHiA GENETICS

The healthtech company has merged genomics and radiomics for the first time and is opening up access to personalised medicine

Share

Healthtech company Sophia Genetics is bringing genomic data-driven medicine to the NHS through a proprietary AI platform that can analyse the genetic data in a single drop of blood to diagnose a patient.

The Swiss company uses next-generation sequencing (NGS) to study patients' genomic profiles and unearth actionable insights that help clinicians treat the causes of diseases rather than the symptoms.

The company's eponymous Sophia technology conducts the genomic analysis that is the foundation of its products. The flagship of these is the software-as-a-service Sophia DDM platform - or Data-Driven Medicine - which interprets and visualises the genetic variations.

"When we launched Sophia Genetics, we had a simple but big ambition, which was to build something that would basically help better diagnose and treat the maximum number of patients worldwide," Dr Jurgi Camblong, the CEO and cofounder at Sophia Genetics, tells Computerworld UK.

"Given our expertise, and what was happening in the field at that time, we took advantage of leveraging genomics, which was really the first digital technology that would be deployed from a research setting into a diagnostic setting in hospitals, due to the fact that sequencing the genome had decreased dramatically in terms of price."

Read next: How the NHS is transitioning to the new Health and Social Care Network

In January 2018, the company announced that radiomics capabilities had been added to Sophia, the first time that genomics and radiomics had been merged together.

Radiomics allows the software to predict the evolution, the size, volume and location of a tumour by analysing medical images and correlating the features with genomic patterns.

"Combining the two, we will be able to better predict how a cancer will evolve, and eventually we will be able as well to see some things in the images that are not seen by the human eye, but that we could teach algorithms to detect," says Dr Camblong.

NGS in the NHS

Sophia Genetics was founded in 2011 in Switzerland and launched the Sophia DDS platform in 2014. That year it also signed its first deal with the NHS.

It has since worked with a range of partners in the UK including St George's Hospital in London, one of the largest teaching hospitals in the country and the home of the South West Thames Regional Genetics Service.

The service was using DNA sequencing to study the genomic profiles of patients but faced a problem that had become widespread: advanced genetic testing produces an enormous quantity of complex data.

Sophia Genetics offered a solution through AI. The company helped them develop algorithms to analyse this vast amount of information and convert it into actionable insights for clinicians.

The unique genomic variants that Sophia has been exposed to allows it to understand complex DNA patterns.

Within two hours of uploading a file of the patient's raw genomic data to the platform, it can detect, annotate and classify all the variants that are unique to the patient.

This can help the hospital quickly discover disease-causing genomic details, such as a mutation in the BRCA1 gene associated with a high risk of breast cancer. Clinicians can then arrange personalised treatment.

Cost and use cases of Sophia

Genomics and AI are already both big business and investors are betting that the combination pays off. They've already ploughed a total of almost $60 million (£43 million) into the company, including $30 million (£22 million) from a 2017 Series D funding round led by Balderton Capital.

"I think it's pretty clear that for rare disorders and in oncology, genetic testing will become a de facto standard, and I think the two most important questions here are how can we make it efficient and how can we make it cost-effective?" says Camblong.

The company believes its business model can achieve this. It charges a hospital every time it uses Sophia, which typically costs between $50 (£36) and $100 (£72per patient.

Sophia works in five disorder areas: oncology, hereditary cancer, cardiology, metabolism and paediatric conditions.

It can analyse a wide range of genetic data on these diseases regardless of the sequencing, enrichment kits, and capture-based technologies that are used to analyse it.

This allows Sophia to provide versatile and affordable data-driven genomic healthcare to a range of companies such as Queen Anne Street Medical Centre, a private clinic near Harley Street in London.

The centre offers several NGS tests in collaboration with Sophia, which gives it access to its genomic analysis without the need to invest in extra laboratory equipment and personnel.

Data protection in Sophia

The use of patient data by healthtech companies is a controversial issue, particularly since news emerged that the NHS had shared 1.6 million patient records with Google's DeepMind.

Camblong says Sophia has embedded data protection in the Sophia DDM platform and the company ethos. It encrypts all that sensitive data that is uploaded to the platform and stores it redundantly in secured and private data centres.

He says the company is the first to achieve compliance with both the ISO 13485 (Medical Devices Quality Management) and ISO 27001 (Information Security Management) certifications.

The Sophia team anticipated the requirements of GDPR, which sets strict rules on data processors and controllers. Sophia acts only as the latter, while the hospital acts as the former.

Read next: GDPR explained: How to prepare for GDPR

"The only thing we can do is compute the data to help the hospitals to better diagnose and treat patients," says Camblong. "That's it. We cannot sell the data. We cannot give access to the data to anyone else than the controller."

Growing a genetic knowledge database

Sophia currently helps diagnose a patient every five minutes across more than 400 hospitals in 55 countries that use the platform.

Each new partner brings extra genetic information that can be fed into the genetic knowledge base that the community can use to develop better diagnosis treatment. 

"By adding more and more data and solving more and more complex cases, we're going to be able to create a collective intelligence that will be useful to the current users, but as well to any other hospital that may want to join our network through the use of our platform," says Camblong. 

Read next: The barriers to using IoT in healthcare: What's stopping the Internet of Things from transforming the industry?

The molecular biologist believes that we have entered an "era of real-time epidemiology" that will create a new model for healthcare.

"We can leverage this collective knowledge to continuously further improve the diagnosis and the treatment options that are being given to the patients on the basis of real-world evidence data."

"Recommended For You"

European Bioinformatics Institute to cut storage footprint by 70 percent IBM aims for personalised medicine with DNA chip