Using Data to individualise patient treatments

The Multiple Sclerosis Brain Health Initiative has highlighted the importance of effective early intervention in preventing long-term disabilities in MS. However, as everybody’s MS is different and each person can respond to medications in a vastly different manner, it is difficult to match the right medication to the right person early in the disease. This often means patients go through a period of trial and error in an attempt to find the right treatment, potentially resulting in neurological damage.

International researchers have been using a large worldwide clinical database of MS patients called MSBase to predict an individual’s response to treatments. MSBase houses clinical information from more than 52,000 MS patients worldwide. This study was led by Dr Tomas Kalincik from the University of Melbourne, Australia, and used the data from almost 9000 patients in MSBase to develop an algorithm that predicts an individual’s response to treatment.

Using data from 117 MS hospitals and clinics from 34 countries, the scientists looked at age, sex, disease course, disease duration, previous therapies, changes in disabilities, types of MS symptoms, and the number and location of lesions. Using all this information, they tried to predict the outcome of various MS treatment options.

Once they built an algorithm, they applied this algorithm to an independent patient data set from a Swedish Multiple Sclerosis Registry to see whether it worked. They discovered they could reasonably predict an individual’s response to their treatment in over 80% of cases. Specifically, they could predict disability and relapse outcomes. The researchers had some success predicting the conversion of relapsing remitting to secondary progressive MS.

For example, this model showed that, for injectable therapies, there was a higher probability/chance of disability progression during treatment if there was greater disability when the treatment started, and that the previous therapies used also influenced the outcome. For fingolimod, natalizumab or mitoxantrone, disease progression was mainly associated with lower pre-treatment disease activity. They also showed that the combination of the factors they looked at had a higher predictive value than that of each individual variable.

The scientists are making their model available to treating neurologists through a web-based tool, allowing rapid translation of this research into clinical benefit. Whilst this algorithm does not give doctors a definitive answer on the best treatment, it does help them to make more informed decisions that may reduce the impact of the disease on patients.

Dr Matthew Miles, CEO of MS Research Australia, said MS was a very complex and varied disease, but treating it early leads to much better long-term outcomes. “This latest study has resulted in an innovative tool that will allow neurologists the ability to predict the most suitable medication in consultation with their patients.”

Whilst the predictions may not be 100% accurate, it will drastically improve the odds of matching the right patient with the right treatment at the right time, which will have a positive effect for patients.

With thanks to MS Research Australia – the lead provider of research summaries on our website.