Emerging Research & Trial Strategies for Progressive Multiple Sclerosis

Multiple sclerosis is a heterogeneous disease with an unpredictable course. We aimed to define subpopulations of patients at higher risk of disability progression and assess whether disability predictions could expedite patient recruitment in clinical trials.  We used clinical and MRI data from 17 previously published clinical trials and observational cohorts and divided them into training and external validation data sets. At study entry, 5,483 patients in the training data set and 2,668 patients in the external validation data set were classified into three groups of high-risk, medium-risk and low-risk of disability progression. In the external validation data set, patients who were predicted to be in the high-risk group had a higher chance of developing CDP compared to those in the medium-risk (HR=1.51, p<0.0001) and low-risk groups (HR=1.51, 95% CI: 1.18 to 1.92, p<0.0001). When we looked at the model-derived risk scores, for every one standard deviation increase in the risk score at the study entry, there was 0.68 (standard error=0.03) increase in EDSS at baseline (p<0.001). Including patients predicted as high risk increased the statistical power by 6% for a trial of 1,200 participants (p<0.0001).  Our model (an extension of the previously published "SuStaIn" model) can be used to expedite clinical trials by reducing sample sizes or duration and identify those who may benefit from more efficacious treatments because they have a greater risk of disability worsening.

Multiple sclerosis (MS) diagnosis depends in part on invasive methods like lumbar puncture or contrast injection. These yield a broad range of sensitivities and specificities, some reported well below 80%. Proton magnetic resonance spectroscopy (1H MRS) measures disease-relevant small molecules in vivo and may therefore offer a noninvasive route to improving MS diagnostics, particularly the identification of progressive phenotypes. To this end, we applied supervised learning to seven age-corrected metabolite concentrations, including N-acetyl aspartate, choline, myoinositol, glutamate, glutamine, GABA, and glutathione, measured via 1H MRS at 7 Tesla in normal-appearing frontal cortex voxels of individuals with progressive (N=19) relapsing-remitting (N=25), or no (N=24) MS. Resultant models distinguished progressive from relapsing-remitting MS with up to 84% sensitivity and 74% specificity, and from control with up to 79% sensitivity and 68% specificity, exceeding diagnostic accuracies of many imaging- and CSF-based methods previously reported. Feature importances indicated the disproportionate contributions of glutamate, myoinositol, and GABA to distinguishing multiple sclerosis from control, but of glutamine, total choline, and glutathione to distinguishing progressive from relapsing-remitting MS. These results demonstrate that 1H MRS data from normal-appearing cortex can inform classifiers of progressive MS exhibiting some degree of both accuracy and interpretability.

Biomarkers of MS activity have long been sought both for discovery biology to elucidate mechanisms of injury and for clinical application to predict and monitor clinical outcomes. MRI of the brain is now routinely used clinically to detect new and enlarging T2 lesions, which are associated with inflammatory disease activity but T2 lesion volume is only modestly associated with clinical measures of disability in MS and is insensitive to neuro-axonal loss. Neurofilaments are the dominant proteins of the neural cytoskeleton and are released into the extracellular space following neuro-axonal damage, and have thus been proposed as putative biomarkers of neuro-axonal injury. Neurofilament light chain (NfL) especially, has been shown to be a promising biomarker because of its high solubility, and increased NfL levels have been found in the blood and CSF in several neurological disorders with underlying neuro-axonal degeneration, including MS. NfL holds promise as a prognostic and monitoring biomarker, especially early on in the disease process when axonal transections are common in active lesions. In progressive MS the frequency of peripheral immune mediated infiltrates is reduced and secondary mechanisms of injury such as neurotoxic glial activation, synaptic pathology, metabolic dysfunction, and axon degeneration evolve with duration of disease highlighting the need for next generation biomarkers associated with slow insidious progression.

Over the past two and half decades, dozens of positive Phase III trials have been conducted in relapsing MS. Despite this success in relapsing MS, the story with progressive MS has been very different with most trials failing to meet their primary outcome. This talk will review lessons learned from those trials, including therapeutic targets, patient selection, and outcome measures.