The question to be answered for personalised treatment discovery was formulated as follows
Is it possible to analyze clinical trials conducted for multiple sclerosis (MS) and find potential signals which could help to identify candidate personalised therapies?
Identify relevant clinical trials in publicly-available databases and analyze data from these trials to discover both candidate VAMs and biomarkers which predict responders/non-responders for each candidate.
On analysis of the clinical research studies available, we identified a subset of studies that could prove useful in our computational analysis using the inVerse methodology:
The number of unique MS trials assessed in the datasets is shown on the right. Learned properties of poorly-designed clinical trials:
Volv performed analysis across the studies highlighted above and it produced 22 ‘core’ models with which to predict patient response to 15 important MS therapies and facilitate personalized treatment optimization. In the example below we focus on one of those core models.
We identified ten ‘core’ sets of models and their main purposes (e.g. optimal therapy selection, assessment of unapproved treatments), and for each core model set, we then specified type/output/predictive features for all component models and, where feasible, derive simplified versions of the models using our clinically interpretable model method from inTrigue.
Contact us for more details.
Leveraging MSBase research
An international online registry for neurologists studying multiple sclerosis and other neuro-immunological diseases. https://www.msbase.org/
here we examine whether different patients do indeed have different optimal therapies and, if so, which biomarkers would facilitate personalized MS treatment selection.
Observational/experimental data:
Predictive analysis consists of three main steps:
1 reduce MS disability progression
Therapies predicted to optimally reduce MS disability progression for each of 8513 patients (2D feature-space projection):
2 reduce conversion to progressive MS
Therapies predicted to optimally reduce conversion to progressive MS for each of 8513 patients (2D feature-space projection):
3 Predicting disability progression events
Predicting disability progression events
Risk prediction error for baseline model (red) and models obtained using our learning-from-aggregates algorithm for each of the five therapies investigated in the study.
Features predictive of response were identified using feature importance and forward/ backward feature selection [HTF 2009] and, interestingly, vary according to therapy type. Predictive features include
Analysis is informed by seven clinical trials together with data from two associated papers.
It is possible to find distinct phenotypic differences between reposnder and non reponder populations for both therapies that will
This is good news for patients but also to those responsible for the management of disease in countries around the world.
We have applied computational analysis to identify clinical trials expected to be of value for personalisation of treatment. From an initial set of ~2000 candidate trials, this procedure resulted in the collection/preprocessing of 45 project-relevant datasets involving 18 unique therapies.
Analysis of these datasets shows:
Predictive modeling yielded a number of interesting findings, including:
Volv performed analysis across the studies highlighted above and it produced 22 ‘core’ models with which to predict patient response to 15 important MS therapies and facilitate personalized treatment optimization. In the example below we focus on one of those core models.
We identified ten ‘core’ sets of models and their main purposes (e.g. optimal therapy selection, assessment of unapproved treatments), and for each core model set, we then specified type/output/predictive features for all component models and, where feasible, derive simplified versions of the models using our clinically interpretable model building from inTrigue.
inVerse - MS Case Study - the case for change
Using inVerse learning from aggregates capabilities we show here that by enabling granular insights and learnings from aggregate data such as successful and failed clinical trials, we can derive massive value for Payers and Healthcare systems.