Multiple sclerosis: two decades of progress

The past 20 years have seen remarkable advances in multiple sclerosis, including a sea change in understanding the fundamental immune drivers mediating CNS demyelination and neurodegeneration; identification of risk genes; a more precise account of epidemiology and incidence; and development of highly effective therapeutics. Today, multiple sclerosis relapses can be completely and safely eliminated in most patients, and the more treatment-resistant progressive symptoms partially slowed. Together, these breakthroughs represent a profound achievement of modern molecular medicine.

Considered for many years an exclusively T-cell-mediated autoimmune disorder, the new millennium coincided with the beginning of a more sophisticated understanding of multiple sclerosis that builds on complex interactions between infiltrating immune system cells, including B cells and CNS-resident cells, especially microglia and astrocytes.¹ Importantly, B cells were recognized as key drivers of tissue damage. B cells located within aggregates of inflammatory cells (ectopic lymphoid follicles) in the meninges,^2,3 and near lesions,⁴ are pathogenic probably through antigen presentation and secretion of pro-inflammatory cytokines and soluble neurotoxic factors, and possibly through autoantibody production. Elevated levels of immunoglobulins (IgG) and oligoclonal bands in the CSF were the earliest indications of B cell involvement in multiple sclerosis, and multi-omic studies identified the source of these IgGs as antigen-educated, highly selected clonal B cell populations.⁵ Single-cell sequencing confirmed that CSF B cells are driven to an NF-kB mediated proinflammatory memory and plasmablast/plasma cell phenotype, further supporting their pathogenic role.⁶ Antibodies produced by CSF B cells appear to recognise diverse intracellular targets, without any unique antigen or autoantigen profile identified thus far.⁷ Scrutiny of peripheral B cells revealed roles in anti-CNS antibody reactivity,⁸ antigen presentation, and regulation of brain homing autoreactive CD4+ T cells.⁹ With respect to antigen presentation, a model for molecular mimicry was advanced whereby peptides derived from HLA molecules associated with multiple sclerosis risk were self-presented, stimulating T cells cross-reactive with peptides from Epstein-Barr virus (EBV) and the gut bacterium Akkermansia, both associated with multiple sclerosis risk, and also myelin components.⁸

The pathology of multiple sclerosis also was recast, with a new understanding that its’ two faces–inflammation leading to relapses, and neurodegeneration mediating progression–represent a continuum across the course of the disease. From a clinical perspective, this promoted a unitary view, in which relapses and progression are considered overlapping features, with relapses mediated by B and T cells that move into the CNS from the bloodstream, and progression driven by diverse immune cells “trapped” in the CNS and relatively resistant to peripherally administered therapies. Pathologies that contribute to progressive symptoms include chronic active plaques consisting of inflammation, progressive axonal loss, and large numbers of microglia at the leading edge of enlarging lesions in the absence of blood-brain barrier disruption. Primary injury to the cerebral cortex also has an important role¹⁰, associated with adjacent meningeal lymphoid follicles, and damage resulting from glutamate toxicity, oxidative injury, iron accumulation, and mitochondrial failure.

In the early 2000s, there was only rudimentary knowledge of the genes that contribute to multiple sclerosis. The HLA region on chromosome 6 had been identified as a risk factor in the 1970s, but early genetic maps, comprised of only a few hundred size-based markers, were inadequate to identify other disease genes. This changed with development of a new class of dense genome maps based on single nucleotide polymorphisms (SNPs), and it now became possible to interrogate the genome from large numbers of sporadic affected individuals. However, the low power of single SNP variants required huge data sets, accelerating collaboration. The International Multiple Sclerosis Genetics Consortium collected more than 72,000 DNA samples from people with multiple sclerosis, and many thousands more from controls. The first three non-HLA multiple sclerosis susceptibility genes were identified in 2007, the map was expanded to 95 loci in 2011, and today 233 genomic regions are firmly associated with risk.¹¹ These associations explain approximately half of disease heritability. Importantly, the results confirmed key roles for adaptive and innate immunity,¹² and are also consistent with participation of brain resident cells such as microglia and astrocytes, as well as vitamin D metabolism, in pathogenesis.

A new era of genomic investigations, based on understanding the human microbiome containing a millionfold more genes than the human genome, as well as epigenetic changes associated with specific cell types such as B cells¹³ or environmental exposures, is likely to power the next phase of discovery, and hopefully will dissect the interplay between host genetics and environment. There is robust evidence for a positive association of multiple sclerosis with EBV infection,¹⁴ Akkermansia, as well as smoking, alcohol consumption, serum vitamin D, and childhood obesity. The EBV association is particularly compelling, because acute infection results in a lifelong carrier state of EBV in B cells. Combining genomic studies with established environmental effectors has the potential to transform prospects for a true understanding of multiple sclerosis biology, for precision therapy and possibly prevention.

In 2002, the diagnosis of multiple sclerosis relied on the McDonald Criteria. These were revised in 2005 to include spinal cord lesions along with juxtacortical, periventricular, and infratentorial locations that were relatively specific for multiple sclerosis. In 2010, further refinement allowed for diagnosis of MS after a single clinical attack if radiographic evidence of dissemination in time was present. The 2010 criteria also integrated pediatric-onset MS and considered diagnosis in ethnically diverse populations. The 2017 Criteria¹⁵ introduced the concept of dissemination in time based on elevated intrathecal gamma globulin synthesis. These criteria also provided specific considerations to avoid misdiagnosis. Multiple sclerosis subtypes also underwent revision in 2013 and now multiple sclerosis is classified as either starting with relapses (relapsing) or progressive disability without initial relapses (primary progressive). Multiple sclerosis disease activity can be either clinical (relapses) or radiographic (new lesions), whereas the term progression strictly refers to worsening of disability. Relapsing patients who experienced worsening disability independent of relapses are considered to have secondarily progressed, and primary progressive patients who experienced a relapse are classified as having superimposed active MS. The new scheme supports the notion that multiple sclerosis is a single disease with a phenotypic continuum.

Advances in MRI continue to profoundly shape knowledge of multiple sclerosis biology, diagnosis, disease course, and treatment response. Examples of progress include imaging of central veins within multiple sclerosis lesions; meningeal B cell rich aggregates; chronic active plaques; as well as quantitation of grey matter and spinal cord pathology.¹⁶ Blood-based protein biomarkers are also beginning to affect the field. Measurement of neurofilament light chains has become increasingly useful as a marker of neuronal loss, and it is likely that additional biomarkers sensitive to changes in myelin and other CNS cells will soon be available for clinical use.¹⁷

Most important, recent years have seen remarkable progress in multiple sclerosis therapeutics.¹⁸ In 2002, therapeutic discussions focused on the relative merits for modestly effective self-injected medications, beta interferons and glatiramer acetate, and on optimal dosing strategies. Off-label use of broad-spectrum immune suppressants was also common. In 2004, natalizumab, a VLA-4 receptor binding monoclonal antibody, came to market offering an apparent doubling of efficacy with respect to relapse rate reduction. The initial euphoria regarding this breakthrough turned to despair when the first cases of progressive multifocal leukoencephalopathy were identified. Natalizumab was withdrawn from the market only 3 months after its introduction and would not return until mid-2006 with a risk evaluation and mitigation strategy.

The quest for oral options led to the approval in 2010 of fingolimod, a sphingosine-1-phosphate receptor modulator that sequesters lymphocytes within lymphoid tissue; use was hampered by cumbersome pre-treatment screening. Teriflunomide, a dihydro-orotate dehydrogenase inhibitor that inhibits DNA metabolism in proliferating lymphocytes received regulatory approval in 2012; however, concern about hepatoxicity and teratogenesis limited use. By contrast, when dimethyl fumarate, an anti-psoriatic treatment that alters lymphocyte gene expression profiles, was approved in 2013, there was rapid adoption because of its favorable safety and tolerability profile and minimal monitoring requirements. Alemtuzumab, a lymphocyte depleting monoclonal antibody, was approved first in Europe and later by the US Food and Drug Administration (FDA) in 2014; alemtuzumab’s impressive efficacy was plagued by safety concerns that included de novo autoimmunity, vascular events and malignancies, such that this product is now uncommonly used in the USA. Cladribine, a lymphotoxic small molecule, was initially rejected by the FDA in 2011 but eventually found approval in 2019 as a second-line treatment. Other recently approved products that share mechanisms of action with earlier therapies include three sphingosine-1-phosphate receptor modulators (siponimod, ozanimod and ponesimod), fumarate preparations (diroximel fumarate and monomethyl fumarate), and pegylated interferon beta. In addition, biosimilars for glatiramer acetate, and generic versions of dimethyl fumarate became available.

B cell depletion therapy with anti-CD20 monoclonal antibodies proved to be game-changers. In 2017, ocrelizumab, a monoclonal administered by infusion every 6 months, was approved by the FDA for both relapsing and primary progressive multiple sclerosis. Ocrelizumab remains the only therapy approved for primary progressive multiple sclerosis. This was followed in 2020 by approval of another anti-CD20 monoclonal antibody, subcutaneous ofatumumab, for relapsing MS. The B cell depleting monoclonal antibodies offer a combination of high efficacy, excellent tolerability and highly favorable safety profiles that are leading to a paradigm shift in prescribing practice away from the fix-on-fail escalation treatment paradigm toward high-efficacy, frontline therapy.

Earlier and more accurate diagnosis leading to initiation of effective disease modifying therapies has changed the face of the illness.¹⁹ What was once a chronic neurodegenerative disease leading to profound disability, social isolation, and early death is now largely treatable. Nonetheless, unmet needs persist for patients with all forms of progressive multiple sclerosis, including those with relapsing multiple sclerosis who also experience insidious worsening (silent progression) independent of relapsing activity.²⁰

With effective control of relapsing multiple sclerosis, more effective means to treat or prevent progressive symptoms have become the holy grail of current therapeutic studies. The partial benefits of B cell depleting therapies on progression stimulated efforts to treat with higher doses, target treatment-resistant CNS B cells, or treat incident (e.g. new-onset) cases with highly effective therapy to prevent treatment-resistant CNS inflammatory niches from developing. Another promising area is to develop therapeutics against activated microglia, and in this regard the Bruton’s tyrosine kinase (BTK) inhibitors offer a distinct advantage that they target both B cells and microglia. The strong implication of EBV in multiple sclerosis pathogenesis is also leading to innovative approaches to eliminate EBV from B cells or selectively kill EBV-infected B cells. Neurorestorative therapies, that might reduce or reverse disability through repair mechanisms such as remyelination, axonal sprouting, or neurogenesis, were once a matter of fictional speculation but are beginning to come into focus, and it seems plausible that the first such treatments with broader applications in neurology will be derived from multiple sclerosis research.

Figure: Advances in Multiple Sclerosis 2002–2022.

Overview of research (blue) and therapeutic (orange) advances in multiple sclerosis over the past 20 years. Left side: year indicates publication date of scientific advance. Right side: year indicates date of first regulatory approval; when designated by an asterisk, indicates year of publication of key clinical trial of an unapproved therapy. ^a Cross-presentation of DR peptide fragments with EBV, bacterial and myelin proteins. ^b Specificities directed against a diverse array of (mostly) ubiquitous intracellular targets. *Currently off-label in the USA.

APC, antigen-presenting cell; CNS, central nervous system; CSF, cerebrospinal fluid; EBV, Epstein-Barr virus; HLA, human leukocyte antigens; MRI, magnetic resonance imaging: MS, multiple sclerosis; NfL, Neurofilament light chain protein; OCB, oligoclonal bands; SPMS, secondary-progressive multiple sclerosis