Reassessing Asthma: Biomarkers, Business and Big Data

22 March 2017

Affecting millions and costing billions worldwide every year, asthma needs to be better understood to enable better treatment and management of this frequently distressing, sometimes fatal family of respiratory diseases.

With SES members Southampton University and Imperial College London at the heart of its conception and execution, the multi-million-Euro U-BIOPRED project – the world’s biggest collaboration in severe asthma – is driving discoveries that will improve targeting of drugs and equip pharma businesses to develop potentially game-changing new ones.

Above all, in pinpointing key ‘biomarkers’ – molecules revealing the precise type of asthma present in patients – U-BIOPRED has produced huge volumes of data (e.g. generated by cutting-edge proteomic and lipidomic technology at Southampton) posing a ‘big data’ challenge successfully addressed by Imperial’s computing experts.

The Anatomy of Asthma

Asthma isn’t what it seems. Far from being a single disease with a single cause, it’s an umbrella term for a range of chronic airway conditions resulting in symptoms that we recognise as asthma: attacks of wheezing and breathlessness. Collectively, these lead to around 500,000 hospital admissions and 15,000 deaths each year in Europe alone. Severe cases account for around 5-10% of the population of people with asthma, and for over 50% of asthma-related costs which, just in Europe, exceed €19bn annually.“The key to more cost-effective treatment of severe asthma is the ability to define asthma better and, specifically, to identify its ‘phenotypes’ or sub-types”, says Ratko Djukanovic, Professor of Medicine at Southampton University and Director of the National Institute of Health Research (NIHR) Southampton Respiratory Biomedical Research Unit. “One approach is to use ‘omics technologies, such as genomics and proteomics, to pinpoint – and explore the roles of and relationships between – biomarkers associated with different phenotypes.”

This approach has provided the basis for U-BIOPRED (Unbiased Biomarkers in Prediction of Respiratory Disease Outcomes). Launched in 2009, the project is built on:

A consortium of 20 academic institutions, 11 pharma companies (including some of the sector’s biggest names), three small to medium-sized enterprises (SMEs), one industry partner and six patient organisations (see full list).
Clinical and biomarker data collected from around 1000 adults and children with asthma across Europe, with over 400 adults and 100 children suffering from severe forms of the disease.

“By applying a specific analytical technique to a specific type of sample, we can produce a specific biomarker ‘fingerprint’”
Professor Ratko Djukanovic, University of Southampton

Funded by the Innovative Medicines Initiative (an EU joint undertaking with the European Federation of Pharmaceutical Industries and Associations), with in-kind support from the pharma sector, U-BIOPRED was conceived by Professor Peter Sterk (University of Amsterdam), Professor Djukanovic, Professor Ian Adcock (Imperial) and Professor Kian Fan Chung (Imperial), with excellent support from the NIHR-funded Biomedical Research Units.

“We recognised that studying biomarkers offered a route to a clearer understanding of different asthma phenotypes and how the mechanisms driving asthma can differ markedly from person to person,” Professor Djukanovic comments. “It was evident that extensive identification of relevant biomarkers would represent a huge breakthrough in terms of

Pharma for the Frontline

The close involvement of pharma businesses in U-BIOPRED ensures a clear pathway between front-end research and ultimate real-world impact in years to come, with better use of better medicines – based on improved knowledge of which sub-types/phenotypes of asthma best respond to which drugs – offering improved patient outcomes as well as big cost savings for the healthcare sector.

At the core of U-BIOPRED is painstaking research using an array of cutting-edge analytical techniques – in some cases designed as part of U-BIOPRED – to identify asthma biomarkers. For example, Southampton already had in place excellent technology platforms that could be used to analyse lipids (fatty substances found extensively in the human body), and protein biomarkers in sputum samples required use of a novel technique based on mass spectroscopy. Southampton teamed up with experts in exhaled breath (Amsterdam) and urine metabolomics (Stockholm) to establish a formidable ‘omics laboratory.

“We’ve used a wide variety of methods to analyse samples such as blood, breath, urine, sputum and tissue obtained via bronchial biopsies,” Professor Djukanovic explains. “By applying a specific analytical technique to a specific type of sample, we can produce a specific ‘fingerprint’ highlighting the biomarkers detected there. We can then combine fingerprints based on individual ‘omics methods to generate ‘handprints’ – comprehensive sets of biomarkers relating to each distinct phenotype. These handprints provide the basis for exploring the phenotypes, their specific features and the specific mechanisms that drive them, which can then be targeted by new therapies whose outcomes can be predicted.”

Inside Information

Underpinning this step change in tackling asthma is the mountain of U-BIOPRED data generated by testing samples collected from hundreds of patients and collating clinical questionnaires providing insights into their health issues and lifestyles. The sheer volume of this data resource – enough to cover 60 football pitches – is quite staggering.

Critical to U-BIOPRED’s success, then, has been the ability not simply to store the data securely but also to integrate it efficiently and interrogate it quickly to find tell-tale data clusters, identify specific phenotypes and establish their characteristics. On behalf of the consortium, computing specialists at Imperial have taken the lead on this challenge. This has involved fine-tuning the industry-standard tranSMART knowledge management platform to U-BIOPRED’s specific needs, creating a unique data resource that pharma and academic partners can mine readily and easily in order to better understand asthma.

“U-BIOPRED partners can undertake detailed data analysis and explore all sorts of correlations across data types”
Professor Yi-Ke Guo, Imperial College London

“This is a very big, very complex resource that can only be fully exploited through a highly sophisticated but highly responsive knowledge management platform,” says Yi-Ke Guo, Professor of Computing Science at Imperial’s Department of Computing. “Basing our solution on tranSMART means U-BIOPRED partners can undertake detailed data analysis, explore all sorts of correlations across data types and easily map their own findings against those of others in the consortium and beyond.”“Every aspect of U-BIOPRED has been characterised by a team ethic and a fantastic level of collegiality,” Professor Djukanovic concludes. “It’s been an object lesson in how a big consortium can work as if it were one institution, with academics, company scientists, patients and patient organisations sharing expertise and supporting each other in pursuit of a very clear objective: to improve quality of life for people with asthma and equip the pharma industry with the insights they need to make it happen.”