AVERT: Autoimmunity relapse prediction using multiple parallel data sources

In most causes of autoimmune disease, where the body's immune system attacks an individual's own body, the condition relapses and remits. This means that strong medications to suppress the immune system bring the patient into remission, but they remain at risk of suffering a flare of their disease. In the autoimmune kidney condition under study in this project, ANCA vasculitis, this risk of flare or relapse is about 50% after 5 years, and requires the use of high dose medications, which carry the threat of infection when the immune system is suppressed by them. In order to tailor these medications precisely to an individual patient we need to understand more about why they relapse. We believe that relapse occurs due to issues inside the patient (such as the kind of genes they carry) and triggers in their environment (such as an infection); when these two components meet, a relapse is triggered. The problem is that we don't know exactly what these triggers are. To address this we plan to monitor very closely a group of patients with ANCA vasculitis for up to 5 years (the first 2 years of which will be funded by this grant). This will include monitoring their level of activity, how they are feeling and where they are located, and linking these things (which of course change all the time) with detailed information about local weather and bacteria and viruses in their environment. This approach has recently been made possible by advances in "data science", namely the ability to use powerful computers and very large amounts of data to build up a detailed picture of the events surrounding a relapse.
Using this, we eventually aim to build a computer "dashboard" that will help a doctor to decide whether to increase or decrease the immune system suppressing medications.

Award Date
01 July 2016
Award Value
Principal Investigator
Professor Mark Little
Host Institution
Trinity College Dublin
MRCG-HRB Joint Funding Scheme