Click here to download your copy of the PA 4D Translational Medicine Report.
'Translational medicine' is often referred to as a one-way translation of ideas from 'bench to bedside'. Translational medicine is far more than this. It is a two-way process of collaborative and iterative biomedical discovery and development focused on patients.
Significantly, industry, academia and clinical disciplines are now working together to share insights and strengths on the clinical utility of biomedical discoveries much earlier in the process than has previously been possible.
Historically, the benefits from translational medicine have been limited due to a number of important challenges including:
- limited access to clinical data
- the lack of appropriate informatics tools to integrate and interrogate clinical data
- the need for new business models that allow stakeholders to realise the value of translational medicine.
At PA Consulting Group's and 4D’s annual translational medicine foresighting event, we discussed how a number of government, clinical academic and private sector organisations are addressing these challenges allowing them to realise the value of adopting a translational medicine approach to the way we discover, develop and deliver healthcare.
Click here to download your copy of the PA 4D Translational Medicine Report.
Listen to the experts at the PA / 4D foresighting event:
Access to clinical data
The UK government through the UK Life Science strategy has committed to ‘making all patient data available’ for approved research, allowing patients to opt-out if they do not want their records to be used in healthcare research. This innovative and forward-facing approach to realising the value of the NHS is at the heart of the new Clinical Practice Research Datalink (CPRD). CPRD is the English NHS observational data and interventional research service, jointly funded by the NHS National Institute for Health Research (NIHR) and the Medicines and Healthcare products Regulatory Agency (MHRA) and provides researchers access to 52m medical records.
Speaking at the event, Peter Knight from the Department of Health noted that resolving issues around ethics and consent has been key. As Peter explained: “research which had been undertaken showed that less than 1% of patients would opt out of having their data used to support life sciences research. Most people are philanthropic: they know their data could help future generations and particularly their own families.”
The creation of the CPRD and the change in mind-set to an NHS focused on innovation, has cleared the path for technology companies at all levels to unlock the value of health data in the UK.
Although the UK has ‘one-NHS’, it is in fact a number of organisations working together to improve patient outcomes. The structure of the NHS and other healthcare providers (HCPs) means that it is often more difficult than one might expect to share data across multiple sites and stages of the care pathway from primary, secondary to tertiary care.
However, to improve healthcare outcomes, we need this data to be integrated to allow researchers to build holistic views of disease management.
Technological developments to enable the combination of data from multiple and often disparate sources is critical for successful translational medicine. At our foresighting event, we heard from organisations who are making real progress on local, regional and national levels across the care pathway from discovery, development to delivery.
At a local level on discovery, IDBS is working with a number of clinical academic organisations including King’s College London to integrate data from multiple hospitals. This allows clinicians to make better decisions about treatments and patient management, facilitating earlier diagnosis, prognosis and treatment.
Paul Denny-Gouldson, VP of Translational Medicine at IDBS noted that: “the fact that it’s hard is no reason not to do it. We need to get new treatments from research into the clinic much faster, and also requirements from clinic back into research much faster; but as researchers, we don’t trust data unless we have the context to interpret and analyse it ourselves. That means you need as much good quality data as possible.”
At a regional level, development initiatives such as Manchester-based North West eHealth (NWeH) are providing a secure environment for aggregating, anonymising and analysing health-related data from defined populations, to be used for developing health provider services and enabling scientific research. NWeH currently has access to over 400,000 patient records and by 2013 the figure will exceed two million. NWeH will provide the region’s population with better healthcare decision making and ultimately better healthcare outcomes.
But as Professor Iain Buchan, CSO NWeHealth, points out: ”bigger is better in databases is a myth”. In the US, there are 174 million patient years of observational medical outcomes research yet this has delivered non-reproducible results because of different data sets, different results and different modelling methodologies. Professor Buchan explains “to realise the value of this data you need systems that talk to one another and say ‘hang on, you’ll run into problems if you make that assumption’.”
This is especially important in a time when we look forward to an environment that seeks to interrogate multiple data sources. In the future making results from one group available to another will be increasingly important. If one group is investigating disease through social modelling and another is looking at patient behaviour, then we need systems that automatically flag up related findings from one group to another and make sense of the connections between the data.
Finally, at a national level of delivery, the integration of records using anonymised identifiers brings with it the possibility of creating ‘virtual’ records by matching records from multiple health and non-health sources. This large-scale data processing is being made more practicable by the use of highly scalable and relatively low-cost cloud computing platforms such as Google BigTable and BigQuerry. PA’s James Mucklow demonstrated that with cloud it is possible to analyse year-on-year A&E admissions to a large hospital in as little as twenty seconds. The ability to conduct analysis of large-scale population data in this time will change the way we deliver care.
New business models
Advances in access to clinical records, and in the IT tools to interpret them, are providing a wealth of new data that has commercial value to both the existing incumbents and to new entrants to the healthcare sector. However, in addition to providing appropriate access to the data and developing robust analytical tools, it is now necessary to develop and test new models of commercial interaction that recognise the contribution of how the data influences commercial decisions. If we are to develop outcomes-based models of healthcare then we require innovation in the way clinical data is used by businesses. Nowhere is this more evident than in the development, management and application of the data sets at the heart of translational medicine.
Life science companies have traditionally based their go-to-market propositions on a product, predicated on approved indications and defined by its efficacy benefits. Data, experience and the need to deliver targeted therapies that deliver real benefit to patients has resulted in an increased focus on the patient through translational medicine. This focus has moved the life science industry to think about integrated health value propositions that demand product, technology and service integration
The increased use of biomarkers and companion diagnostics to stratify distinct patient populations is a good illustration of this, in the future we expect many more therapies to be linked to a companion diagnostic product, gene analysis tool and potentially a monitoring device. In order to do this, pharmaceutical and biotech companies will have to work closely with device companies as well as clinical data to proactively determine patient cohorts that benefit from receiving a particular therapy.
One such example is the collaboration between Arizona State University and Pfizer who will work together over four years in an initiative sponsored by the National Institutes of Health to discover proteins, or biomarkers, to help predict cardiovascular disease and to assess potential new treatments in people with type 2 diabetes. If in the future the information that determines this cohort comes from academia or health care providers like the NHS, rather than from pharma clinical trial data, how should this contribution be valued and can traditional licensing and royalty models be used in this space?
Not only is there a need to find the mechanisms to deliver appropriate value to the contributing partners (academia, industry and healthcare provider) there is also a need to be innovative in terms of approaches the partners take to funding this research.
As increasing external investment is put into R&D, the industry will need to be more agile in the way it accesses and funds the data it requires. This will be particularly evident as we shift from a price-based reimbursement model to one of predicated outcomes and risk sharing. Here we are likely to see an upstream shift toward risk-based investment as well as in development, promoting earlier collaborations between those with the clinical data and those with the drugs.
In a world of increasing translational medicine initiatives, we need to ensure we have the systems and processes that value each of the partner’s contributions. Without these the stakeholders will find it increasingly difficult to share end-to-end data across the discovery and care pathways to the detriment of healthcare innovation.
Click here to download your copy of the PA 4D Translational Medicine Report
As organisations such as IDBS, NWeH and CPRD have shown, the real power of large data sets only emerges when we find ways to analyse it in a consistent way, allowing us to better understand disease progression, the impact of treatment interventions and ultimately the impact on healthcare outcomes.
In the future, interpretation could become more, rather than less, complex as more sources of data are combined – in particular, data from the three pipelines of R&D, clinical audit and clinical commissioning. The addition of social care data could add another important perspective but, again, adds a further layer of complexity.
While there is still work to be done, we now have working examples of translational medicine ‘in-action’ and delivering on the promise of facilitating the creation and development of effective medicines that lead to better outcomes for the right patient at the right time.