Abstract

Patient Stratification in Neurodegenerative Diseases

Ahmed Enayetallah, Associate Medical Director for Alzheimer’s Disease, Biogen Idec

Neurodegenerative diseases such as Multiple Sclerosis (MS) are multifactorial and heterogeneous, in regard to clinical presentation and disease course. Post-hoc analysis of clinical trial data may afford insights into the molecular, clinical and radiological variables associated with this heterogeneity, facilitating individual prognosis formulation and treatment decision-making. Here we examined MS patient molecular profiles to determine their association with clinical and radiological measures of disease activity and severity. Whole blood genome-wide expression data from 570 MS patients from the placebo arms of two large phase III studies (DEFINE and CONFIRM) were analysed. Integrative and analytical methodologies were used to contextualize the gene expression data to generate individual-specific putatively-mechanistic disease profiles. Unsupervised clustering of patients by their mechanistic profiles was performed. The molecularly defined patient clusters were tested for associations with baseline clinical or MRI disease characteristics and on-study relapses or disability progression. Finally, the analysis was replicated in a treated cohort at baseline and the impact of treatment was compared in the molecularly-defined patient groups. Baseline mechanistic profiles from MS patients provided distinct molecularly defined patient groups. The molecularly-defined patient groups showed distinct patterns of immune signals, oxidative stress, and differential B-cell and T-cell biology. These prospectively defined patient groups exhibited a distinct range of clinical and radiological variables with up to 20-fold differences between the extremes. In the treated cohort certain molecularly-defined groups at baseline were enriched for patients who were more likely to benefit from treatment.