Success Story – EuResist
EuResist – Enhancing HIV treatment with more precise patient modeling
HIV is not curable but it is treatable, states Dr Francesca Incardona, CEO of EuResist Network GEIE and research area manager at Informa Srl, the SME responsible for coordinating the EuResist project. The thirty-month initiative was devoted to providing better treatment by implementing an intelligent system that uses patients’ clinical information together with viral genetic data, thus personalising treatment, delivering better results and coping with the complexity of the virus.
‘Choosing the best treatment for a specific person can be very challenging. The antiviral drugs designed to help deal with HIV are numerous, and furthermore the HIV virus evolves very rapidly so that each person hosts slightly different mutants of the wild virus, with a different response to the various drugs. Some viruses become resistant to several drugs and doctors may encounter many difficult situations,’ says Dr Incardona.
Drug resistance is the main source of treatment failure today in western countries. There are numerous antiviral drugs that can be used, and in most cases, it is relatively simple to treat a patient at the onset by prescribing a cocktail of antiviral drugs. Such treatment success is usually short-lived. The virus evolves and becomes resistant to the medicine. When doctors choose new medicines to combat the disease, they may find that the viral population (the mutated viruses hosted by the specific patient), is already resistant to the treatment. EuResist was launched with the objective to collect as much data as possible in order to understand how patients fare under this particular virus.
Dr Incardona explains that the vision behind the EuResist project was to create a system with the ability to learn from real data through smart modelling techniques. Sufficient amounts of data and a modelling algorithm were first becoming available at the time. The system was intended to help doctors choose the best treatment for a particular patient with a given viral population and resistance profile, in a more specific and accurate way than before. The achievement of the project was based on two key outcomes.
State-of-the-art tools for effective management of HIV patient data
First, an integrated database was created, which is among the largest HIV resistance databases available today with over 62,000 patients. It is used by hundreds of countries and contains such data as patient information, drug therapies and AIDS defining events. It proved trouble-free for the project initiators to gather data from different sources and encourage people to share such details, which is in stark contrast to common practice. This surprising willingness was key to project success, she says, and it was accomplished thanks to ‘a pure collaboration of spirits.’ The database is freely available for studies, unless these studies are driven by pharmaceutical companies, which are required to provide a financial contribution.
Second, a prediction engine was developed based on the database, which examines data that is given online and foresees how treatment will work best. The engine boasts a high accuracy percentage (77-78%) in predicting the right treatment. Increasing this percentage remains challenging, as it cannot be done simply by populating the database. ‘We have to take into account other factors, especially the human factor,’ underlines Dr Incardona. ‘What we believe will drastically change the accuracy is to consider the profile of the human immune system.’ The engine is free online.
This system enables more effective patient care and considerably reduces the prohibitive international therapy management costs. Designed to offer more treatment options to patients, the EuResist system can contribute towards saving lives. By reducing the possibility of choosing wrong treatment, it contributes to better quality of life and to the decrease of expenses. The benefit derived from adoption of a treatment optimisation tool is a function of the increase in effectiveness with respect to the standard of care, the treatment failure rate and the availability of treatment options in a specific geographic area.
Increasing effective treatment
Dr Incardona puts the system’s added value into perspective. A 10% yearly increase in the number of effective treatments in an area where 10,000 patients are being treated with a success rate of 80% would result in the reduction of the final treatment failure rate from 20 to 12%. Based on this, an additional 800 patients would benefit from effective therapy every year. ‘Using a system with this modest rate of increase in successful treatment on a large scale can make a considerable difference when we consider that around eight million patients are under treatment in low- and middle-income countries and more will be put on treatment every year,’ she adds. By suggesting a number of possible treatment options, the EuResist engine also contributes to a rational, case-based cost-efficacy paradigm.
Dr Incardona emphasises that there is significant merit in EuResist. Its prediction models - the ideas behind the prediction engine - can be applied to other diseases as well, although not directly. The first area of application with similar issues is hepatitis, where antiviral drugs are now coming out on the market. The idea to use large data sets and these kinds of models can also be applied to cancer.
The EU's funding instrument was vital in getting the project off the ground. When it ended, the EuResist Network was set up, an entity which involves most of the initial project partners. The Network participates in EU co-funded research initiatives, but it is economically independent and self-sustained thanks to a successful business model. The large amount of data collected is highly valued by pharmaceutical companies. Studies on data requested by pharma companies are first evaluated by a scientific committee that consists of all the Network partners. If approved, the paid studies are carried out. This enables the EuResist Network to continue its basic activities which require the efforts of seven professionals, in addition to the involved scientists.
Dr Incardona stresses that additional public financing at this stage is necessary to continue to build on achievements. The broader research aims of the project, such as further efforts to make it applicable to third world countries, require funding which private companies and organisations cannot allocate.
‘The project was a success beyond expectations,’ says Dr Incardona. Not one to rest on her laurels, she aspires ‘to increase accuracy by considering the patient’s genetics, particularly that of the immune system.’ Furthermore, the EuResist Network plans to use similar models in other domains like hepatitis, and to adapt the system for users in third world countries who are most in need.
Participants: Italy (coordinator), Germany, Sweden, UK, Israel