Please login to the form below

Not currently logged in
Email:
Password:

The virtual patient

Healthcare needs an IT revolution to support integration of molecular, physiological and anatomical models that can be tailored to the individual
Virtual patient

In the future, every one of us could have his or her own 'virtual patient' model, continuously evolving as we age and drawing on a wealth of data from our genetic make-up, physiology, anatomy and environmental variables.

The model would be used to predict lurking healthcare problems and, if treatment were required, help choose the most effective and safest option for that person's unique profile.

At the moment that sounds like science fiction and, almost certainly, will only happen many years from now, but researchers around the world are already starting to draw together and integrate the components of such a system.

One of these groups is led by Hans Lehrach, director of the Max Planck Institute for Molecular Genetics and Dahlem Centre for Genome Research and Medical Systems Biology in Berlin, Germany.

Lehrach and colleagues from 15 countries have set up a project entitled IT Future of Medicine (ITFoM), which is one of six pilot projects vying for two awards of €1bn each in funding under the auspices of the European Future Technologies (FET) programme.

Started in May 2011, the six pilots have been given 12 months and €1.5m to develop written proposals. Two winners will be selected in the second half of 2012 as full FET Flagship Initiatives, with funding of €100m a year over 10 years.

ITFoM has the ambitious goal of bringing together medicine and state-of-the-art diagnostic techniques such as medical imaging and functional genomics, virtual patient models and new data handling tools within a closely integrated IT architecture.

"The project is modelled in a similar way to the human genome project, which also cost €1-2bn and took around 10 years, and could have a similarly dramatic impact on healthcare," said Lehrach.

Like all ambitious schemes in research, the ITFoM project is starting from more humble origins.

Lehrach and his associates are currently working on a project to sequence genomes of individual cancer patients and run them through a computer model that can predict behaviour in very large networks. The model analyses more than 2,000 data points, with an emphasis on intracellular signalling pathways, in order to establish the optimum treatment for an individual patient.

"ITFoM's aim, in a sense, is to generalise this basic idea," said Lehrach. "We want to move forward from our current mostly sequence-based efforts in oncology and help to develop an integrated system using molecular, physiological and anatomical models for each individual in the healthcare system, starting from birth to old age."

ITFoM has the ambitious goal of bringing together medicine and state-of-the-art diagnostic techniques such as medical imaging and functional genomics, virtual patient models and new… tools

Beyond personalised medicine
That involves moving beyond what many describe as personalised medicine, which at present means making a treatment decision based on a single genetic marker. This is, in fact, stratified medicine, separating patients into two groups, but still falls some way below a truly individualised approach.

The foundations of the project will be the various 'omics' technologies: sequencing of genomes, proteomes, transcriptomes, metabolomes and so on, which will be combined with modern imaging techniques and sensors that will communicate with, and enliven, the virtual patient model.

Omics technologies are getting more powerful and the costs keep dropping, according to Lehrach. While a genome still costs around $10,000 to sequence at the moment, extrapolating the current downward costs would suggest a major reduction in the next 10 years.

The ITFoM team claims this is the first time that the information and communication technology (ICT) implications of worldwide individualised patient care will be addressed in combination with genomics and medical requirements.

Initial tasks will be to explore novel computing architectures, probabilistic programming languages, information processing paradigms, new mathematical models for complex biological systems and statistical machine learning methods to predict drug responses, as well as algorithms to search for optimal combinations of complex, interacting therapies.

The project launches at a time of almost unprecedented challenges for healthcare systems around the world. Governments are struggling to meet rising healthcare costs, and this is a legacy of pharmaceutical researchers trying for years to develop drugs for the average patient among a population, not the individual.

"It's a bit like a car mechanic who, using an evidence-based approach, concludes that when red cars are faulty the spark plug is usually to blame, and for green cars it is usually the battery because evidence says this occurs 5 per cent more than random," he noted.

"It's better than random, but not much better, and that is basically what we are doing at the moment in medicine."

In most cases, treatment completely disregards the molecular basis of disease, which is why cancer drugs, for example, have traditionally been approved based on the location of the tumour, not the underlying molecular profile.

"We continue to prescribe drugs which cost €50,000 or more a shot, and make 70-80 per cent of patients sicker than before because the tumour is the only tissue guaranteed not to be damaged by the drug."

Drugs which block the epidermal growth factor (EGF) pathway are among the most modern of cancer treatments and are used to treat non-small cell lung, colon and head and neck cancers and have become very successful products. Blocking the EGF pathway in the entire body, however, can have significant side effects, but a ras mutation in the tumour can make it completely insensitive to the drug.

"That means we spend a lot of money to mistreat the patient, and then more money treating side effects," said Lehrach.

The cost of our current medicine is unsustainable as it is predicated on throwing money and manpower at the problem, and a reason why the US has to devote around 27 per cent of its gross national product to healthcare, which, overall, is not very effective.

"Changes in the cost of analytical technologies and computing means that it is getting easier to individualise medicine, based on all the data we can collect from the patient and the information we have collated over decades of research," said Lehrach.

The technology already exists to tackle a large part of the problem, in the form of high-throughput data production, sensors, heavy computing and modelling, and the challenge now lies in putting all the pieces together.

It is just not sustainable for the industry to continue to put so much money into running unstratified trials just to keep the market as large as possible

While the hope is that this approach will dramatically improve the efficacy and safety of cancer therapy in the clinical setting, he is also intrigued by the possibilities it opens up in the development of new drug treatments, and potentially rescuing compounds which have failed in trials.

That could rescue pharma from the costly paradigm of trying to develop mass market, blockbuster drugs, that lots are trying to shake off but which is still entrenched in many companies.

Using virtual patient models may allow drug developers to run many virtual clinical trials, but only require 20 patients or so for a development programme. That would reduce the cost of development dramatically and, assuming the models are effective, should give a much lower chance of failure.

Add to that the ability to rescue the large number of failed drugs in pharma's pipeline, shorter development times which could lead to longer patent lives, plus the opportunity to charge premium pricing for drugs that may only affect a small percentage of patients with a form of cancer, for example, and the pharmacoeconomics may begin to add up.

"In the long run, this could be positive, not only for patients and taxpayers, but also for the pharma industry itself," asserted Lehrach. "It is just not sustainable for the industry to continue to put so much money into running unstratified trials just to keep the market as large as possible."

This scenario clearly depends on some seismic changes not only within pharmaceutical research itself, but also in other business functions, such as manufacturing and, of course, regulatory systems, which would have to evolve to allow a very different development process.

The wealth of information available on cancers means that oncology is a good place to start in developing a virtual patient approach, but Lehrach believes that, in time, even polygenic diseases such as hypertension and diabetes could be addressed using such a model.

The Max Planck Institute has spun out a company called Alacris Theranostics to try to commercialise its computer modelling approach, and funding has already started to flow in from other non-EU grant sources and private investors. 

One two-year project involving a clinical trial in melanoma is being funded by a medical charity, while another is using the approach to develop biomarkers for colon cancer and test therapy selection.

"We find this model works extremely well in oncology, and would expect that it would have similar promise in many other areas of healthcare," said Lehrach.

The approach could also lead to dramatic developments from an economic perspective outside medicine, and create huge new opportunities for the information and communication technology (ICT) industry.

Lehrach believes widespread adoption of integrated ICT in healthcare could almost double the size of the current ICT market, thanks to the opportunities opened up for developing and managing the equipment and software needed to individualise medicine.

"Just 20 per cent of the current global healthcare budget corresponds to the current ICT market," he pointed out.

"It would be transformative to see that amount of money directed towards data-rich, computation-intensive approaches."

 

Phil Taylor

The Author

Phil Taylor is a freelance journalist specialising in the pharmaceutical industry

To comment on this article, email pme@pmlive.com

 

 

 

25th August 2011

Share

Featured jobs

Subscribe to our email news alerts

PMHub

Add my company
Real Science Communications

Real Science Communications is a scientific centre of excellence, infusing credibility and scientific rigour into the conversations around health and...

Latest intelligence

Solaris Health looks at the gene therapy revolution
Gene therapies are bringing new hope for many people with diseases caused by genetic disorders...
PM Society Digital Awards – the power of together
Our chief executive, Emma Statham, writes about the value of awards and the power of together....
Seduction_feature_image_thumb.jpg
Seduce anyone in four simple steps
You know the health of the global economy is dependent on our ability to seduce one another – don’t you? And you know that we need to be able to...

Infographics