Personalizing health care in Multiple Sclerosis using systems medicine tools
Onset project: May 2016
Project ends: June 2019
Development of personalized health care for complex diseases like Multiple Sclerosis (MS) is hindered by a poor understanding of the biological processes underlying the disease and their interactions, as well as by the heterogeneity between patients. These shortcomings also represent a significant limitation in terms of monitoring or predicting the disease course, as well as in the prescription of the most efficacious or safer therapies. By integrating clinical information with omics data and mathematical models of MS, we aim to develop algorithms that can be used in clinical practice to define the prognosis of the disease and that will help in selecting the best therapeutic approach based on the patient’s phenotype. We shall focus on 5 different levels of biological complexity to capture and integrate the most relevant information:
1) genomics to evaluate the individual’s genetic predisposition
2) phosphoproteomics to capture the activity of signalling pathways involved in the immune response
3) cytomics to capture the dynamics of the autoimmune response
4) imaging to quantify the damage of the central nervous system
5) clinical phenotype (clinical scales, comorbidities, drug usage, quality of life, health economics) to define the clinical outcomes to be reached.
Such an approach will benefit from previous work carried out by members of the consortium in the application of systems medicine to MS, including the development of network models of signalling pathways, mathematical models of the dynamics of immune cells, and the harmonization of multilevel and complex databases to develop clinical decision support systems. By testing such tools in small clinical studies, we shall improve the usefulness of systems medicine tools in clinical practice.
The Sys4MS consortium aims to develop new tools based on systems medicine to improve and personalize the management of patients with a complex disease, such as Multiple Sclerosis (MS). This requires developing mathematical models into which clinical information and omics data from diverse sources can be integrated and which can be used to generate algorithms that can predict the disease course and future disability in specific subgroups of MS patients, as well as aid the selection of the best therapy for each individual.
The validity, utility and cost-effectiveness of ‘omics’ based health promotion and disease prevention programmes, will be addressed by Sys4MS, allowing informed decisions on the organisation of health and care systems.
Our hypothesis is that by integrating selected molecular and cellular information into mathematical models of MS pathogenesis, we should be able to simulate the dynamics of disease pathogenesis at the individual level, taking into consideration genetic susceptibility and environmental exposition. By using quantitative endophenotypes, such as CNS atrophy (provided by MRI and OCT), we will be able to associate the dynamics of biological damage to the evolution of clinical activity and disability. Our hypothesis is based on previous work in which a close relationship between mathematical models of T cell dynamics and the clinical course of the animal model of MS and the clinical course of the disease in patients was established (Fig. 1). Thus, we postulate that the models we will derive should provide predictive information that can be applied to specific subgroups of patients at the clinical level.
Figure 1. Signaling pathways associated with MS
Figure 2. Pipeline for the identification of new therapies based on the modeling of signaling pathways associated with MS and MS drugs
Thus, the specific goals of this project are:
1. to integrate clinical, omics and imaging information into computational models of MS and to develop algorithms that predict disease activity, future disability and response to therapy.
This objective will be achieved by bringing together four prospective cohorts of MS patients (n=400) that are being followed at each of the participating clinical centres (Barcelona, Berlin, Genova and Oslo). We shall obtain specific omics data from this cohort by performing genomics (SNPs, HLA typing, RNA seq), phosphoproteomics and cytomics studies on blood samples from the patients (PBMCs – peripheral blood mononuclear cells). Genomics would provide the genetic susceptibility, phosphoproteomics the signalling information of immune cells responding to the autoimmune process and cytomics the dynamics of inflammatory cells. This information will be used to stratifying MS patients with greater precision than can be currently achieved (WP 1). By performing simulations and statistical analysis, we can define the configurations that have a good predictive capacity for specific clinical outcomes (e.g., disease activity, disability progression, response to therapy - month 18) (WP 2).
2. to validate the clinical algorithms and evaluate their benefits in short clinical studies, as well as in prospective database studies.
Based on the selection of the most accurate predictors and their relevance for supporting the decision making process in patients with MS, we shall validate the algorithms through short clinical studies in specific patient subgroups (WP2). We will define the diagnostic accuracy of the algorithms in predicting clinical outcomes. In addition, we will assess the ethical implications of using the information obtained in a format of CDSS, as well as assessing the clinical usefulness of such a CDSS through the use of questionnaires administered to patients and their physicians. In addition, we will perform a Health Economic Analysis of the CDSS developed to evaluate its potential impact on National Health Systems.
A second set of algorithms will be defined that are dedicated to the search for new therapies and combination therapies for MS, based on the effects of current drugs in our simulations and taking advantage of the models of combination therapies we generated previously based on phosphoproteomic networks (WP3). Validation of these proposed therapies will be performed in ex vivo assays using PBMCs from MS patients treated with a given immunomodulatory drug and exposed to the predicted combination therapy in in vitro assays. The effects of the reprofiling drugs will be tested in the phosphoprotein network (we use the phosphoproteomic network profile as the signature of efficacy compared with disease networks). Successful candidates will be proposed and the aim will be to test these in academic sponsored trials, which would take place after the completion of this project.
In summary, in this project, we plan to translate the concepts of systems medicine into real applications to personalize the clinical care of patients with MS. Our approach will identify the tools and data that are truly useful to answer specific clinical questions that are of value to patients and physicians (e.g., prognosis of disease severity, response to therapy, etc.). Therefore, we aim to provide a clear application of systems biology concepts to a specific medical field.
Image: Pert chart
Dr. Pablo Villoslada
|P2||Prof. Julio Saez-Rodriguez
Joint Research Center for Computational Biomedicine (RWTH Aachen University)
Faculty of Medicine
MTI2 Wendlingweg 2 D-52074
Tel.: +49 241 80 89 347
|P3||Dr. Maria Cristina Mingari
IRCCS Azienda Ospedaliera Universitaria San Martino/ IST
|Largo R. Benzi,10
Tel.: +39 0105737210
Fax: +39 0103538639
|P4||Prof. Hanne Flinstad Harbo
University of Oslo
Tel.: +47 99546680
Fax: +47 23027455
|P5||Prof. Paul Friedemann
Charity University Medicine Berlin
Tel.: +49 30 450 539775
Mobile: +49 151 1065 3018
Fax: +49 30 450 539921
Image: Task distribution