Next generation software solution for personalised cancer care Genetic variations between people play a significant role in determining responses to treatment for illness. SAGE-CARE combined patient data with evidence-based medical recommendations to support personalised cancer care. Digital Economy Health © Zita, Shutterstock One of the challenges now faced by doctors in the fight against cancer, is that of accessing and navigating the huge amount of medical data currently available. Along with clinical records, medical imaging and next generation sequencing technology offer data-sets which can capture the genomic profiles of patients and their tumours. While a boost for treatment, there remains the hurdle of matching this genomic data from individual patients, to the actual progression of their disease and treatment. The SAGE-CARE EU-supported project developed technology that enables researchers and clinicians to link data-sets. This allows procedures to be deployed faster, to be more effectively monitored and for lessons learned to be shared, ultimately increasing the effectiveness of treatment. This research was undertaken with the support of the Marie Curie programme. The cancer care prototype To develop a software prototype which tackled melanomas, the SAGE-CARE project brought together a mix of technical and clinical specialists including computer scientists with expertise in ontologies, machine learning and cyber security. Collectively, this multidisciplinary team worked with Hochschule Darmstadt – University of Applied Sciences (HDA) and NSilico, the project’s commercial company, and its clinical partners. After consultations with cancer specialists across hospitals and research communities, the team piloted the SAGE-CARE software with clinicians in Cork University Hospital, Ireland. Sample runs of the system were trialled in workshop settings, with feedback used to improve the prototype’s software design and build the platform. The resultant system allows clinicians to edit electronic health records to enter examination results, diagnoses and treatment, and to insert comments and treatment updates. Linking with various medical information sources – such as clinical trial registers or drug interaction databases – improves the records, keeping them up-to-date. “Giving doctors the opportunity to compare evidence-based medical recommendations allows them to collectively agree on the best treatment for the patient,” says Prof. Bernhard Humm, project coordinator. The SAGE-CARE approach wouldn’t be possible without high performance computing (HPC), due to the large quantity of data generated. Putting this in context, “The human genome contains 3 billion nucleotide characters – the famous A, T, G, C – which indicates how a person responds to disease and treatment.” says Prof. Humm. Expanding the scope SAGE-CARE generated a vast amount of highly specialised software and expertise, which can now benefit clinical care in Ireland. Indeed, over 8 000 cancer cases were managed by software enhanced by the project. Previously, clinicians had to rely on large quantities of paper records in order to discuss and plan treatment; a system prone to errors due to filing and transcription errors. Enhanced user interfaces now link medical ontologies so that clinicians can rapidly enter precise medical terms. The research and training aspect of the MSCA RISE scheme, also gave the NSilico development team skills in scalable processing and deployment. This enabled the system to be installed across many clinical settings, increasing its ability to process large numbers of patients as the user base grows. The researchers also contributed to a new standard for computing for personalised medicine, the BioCompute Object. While SAGE-CARE focussed specifically on melanoma, its findings and insights have already been applied to breast cancer management and new projects are already underway to extend the system to other clinical specialties, including lung cancer, in several EU countries. Keywords SAGE-CARE, personalised medicine, genomics, genetic variations, treatment, medical treatments, melanoma, cancer, ontologies, machine learning, data-sets, cyber security
