“It’s difficult to make predictions – especially about the future.” – Yogi Berra

A decade ago, Dr. Aaron Miller, Mount Sinai, New York, took a game swing at the crystal ball and wrote the article, “Multiple sclerosis: where will we be in 2020?” (Miller A. Mt Sinai J Med 2011;78:268-279).

How well do his predictions hold up?

1. “I do not believe we will see further significant revisions in the diagnostic process over the next decade.” At the time of writing, the McDonald 2010 criteria had just been published (Polman et al. Ann Neurol 2011;69:292-302). Still to come, however, was the 2017 iteration (Thompson et al. Lancet Neurol 2018;17:162-173). McDonald 2017 introduced several improvements: it was no longer necessary to differentiate cortical and juxtacortical lesions or between symptomatic and asymptomatic lesions to meet dissemination-in-space criteria; cortical lesions and symptomatic brainstem or spinal cord lesions could be used to show DIS; and oligoclonal bands in CSF made a come-back as a substitute for dissemination-in-time.

2. “I do not think we will be substantially further along in a decade in our understanding [of] the root cause of MS, because I do not believe there is a single “cause” of MS.” As predicted, the view in 2020 is largely unchanged. Miller’s wording of “a complex interplay of genetic and environmental factors” is still routinely heard.

3. “[T]he clinician in 2020 will still find little, if any, meaningful guidance from the genetic information about individual patients.” At the time of writing, the International Multiple Sclerosis Genetics Consortium had just published the first genome-wide association study (GWAS) in MS (IMSGC. N Engl J Med 2007;357:851-862) and GWAS have since provided a wealth of information. But uncovering the genetic bases of MS has been like looking for hay in a haystack: over 200 risk loci have now been identified. Most risk loci encode for immune-related genes, providing support for the idea of MS as an autoimmune disorder. However, practical questions – Why is MS more common in females? What determines the phenotype? – have remained unanswered. Genetics has provided little guidance for prognosis or treatment selection.

4. “The quest for biomarkers as surrogates that would contribute to accurate forecasting of individual prognosis is a vigorous area of investigation… I suspect that this holy grail will remain elusive over the next decade with regard to application to a specific MS patient.” Numerous potential biomarkers have been investigated over the past decade. The current front-runner is neurofilament-light chain (NfL). NfL as a marker of neuroaxonal damage is not specific to MS and has been investigated in a broad range of neuropathies and neurodegenerative diseases. It will likely have limited utility as a diagnostic tool in MS since it cannot differentiate MS mimics, but it may have a role as a secondary endpoint in progressive MS trials. It is unclear at this point if NfL will be the holy grail or the Shroud of Turin.

5. “[O]ral treatment of MS will be the mainstay of therapy… Fingolimod may be the most widely used therapy… I also expect the monoclonal antibodies to have a smaller role in the overall management of MS… I anticipate that by 2020 the use of self-administered injectable medications – the IFNβs or GA – will be minimal…” Fingolimod became available in the same year of Miller’s prediction but never became the most widely used DMT. Other oral therapies (teriflunomide, dimethyl fumarate) have proved more popular than fingolimod; one monoclonal (ocrelizumab) will likely surpass fingolimod as well. Injectable have declined in popularity but one-third of MS patients (in Canada) are still taking an IFN or GA.