The most recent iterations of the McDonald criteria have enabled an earlier diagnosis of multiple sclerosis (Lee et al. Eur J Neurol 2018; epublished October 26, 2018). However, it is unclear if MS is being diagnosed earlier in real-world practice, or if patients are then receiving earlier treatment with a disease-modifying therapy.
Certainly newly-diagnosed patients enrolled in contemporary MS trials are no younger than those in historical studies of disease-modifying therapies (DMTs). In the first trial of glatiramer acetate published 30 years ago, the average age of the study cohort was about 31 years, with a mean duration of disease of 5-6 years (Bornstein et al. N Engl J Med 1987;317:408-414). In the pivotal trial of beta-interferon-1b, the mean age at entry was about 35 years with a mean disease duration of about 8 years (IFNB MS Study Group. Neurology 1993;43:655-661). In the pivotal trial of beta-interferon-1a IM, the mean age was 36.8 years with a mean disease duration of 6.5 years (Jacobs et al. Ann Neurol 1996;39:285-294). In the PRISMS trial of beta-IFN-1a, the mean age was 34.9 years with a mean disease duration of 5.3 years (PRISMS Study Group. Lancet 1998;352:1498-1504).
In more contemporary studies, the average age at entry was generally older, with an older age at diagnosis. In the OPERA studies, the average age was 37.2 years with a mean time from diagnosis of 3.9 years (Hauser et al. N Engl J Med 2017;376:221-234). In the DECIDE trial of daclizumab, age at entry and time since diagnosis were 36.3 years and 4.1 years, respectively (Kappos et al. N Engl J Med 2015;373:1418-1428). In the pivotal trials of dimethyl fumarate, the average age at entry and time since diagnosis were 38.5 and 5.5 years for DEFINE, and 37.3 and 4.6 years for CONFIRM (Gold et al. N Engl J Med 2012;367:1098-1107; Fox et al. N Engl J Med 2012;367:1087-1097). (Time from MS diagnosis was not reported in the CARE-MS trials of alemtuzumab, the TEMSO study of teriflunomide, the two FREEDOMS trials of fingolimod or the CLARITY trial of cladribine.)
What differentiates older studies from contemporary cohorts is not the age of subjects, but the duration of the MS at entry (8 years in the case of the beta-IFN-1b trial) and the age at the time of MS diagnosis – generally < age 30 for older studies and >age 30 in contemporary studies. If these study populations reflect the current pool of new MS patients, then the average age at diagnosis is about 33 years (OPERA, DEFINE).
This is about 4 years older than the average age at diagnosis reported in natural history studies by the Lyons group for the period 1976-1997 (mean 29.4 years), and the London, Ontario group for the period 1972-2000 (mean 28.5 years) (Confavreux & Vukusic. Brain 2006;129:606–616, Scalfari et al. Brain 2010;133(Pt 7):1914-1929), both of which used the older Poser criteria.
It may be that study cohorts do not reflect real-world populations due to inclusion/exclusion criteria or other factors. However, recent epidemiologic studies suggest that current RRMS patients are indeed older at diagnosis. A study on the prevalence of MS in Iceland found that the average age at diagnosis was 35 years for women and 36 years for men (Eliasdottir et al. Neuroepidemiology 2018;51:50-56). In Ireland, for new RRMS cases identified in 2014-2015, the average age at diagnosis was 37 years (O’Connell et al. Mult Scler Relat Disord 2017;13:75-80). A Norwegian epidemiologic study also reported an average age at diagnosis of 37 years (Grytten et al. Acta Neurol Scand Suppl 2012;195:51-57). A study in Genoa, Italy, reported a mean age at diagnosis of 39 years (Solaro et al. Mult Scler 2015;21:1244-1250). Mean age at diagnosis was even older (41.6 years) in an analysis of a multiethnic cohort in the Kaiser Permanente health claims database (Langer-Gould et al. Neurology 2013;80:1734-1739).
Similar results have been reported in Canada. The British Columbia MS database for the period 1975-2009 – straddling the pre-DMT and DMT eras – reported a mean age at onset (not age at diagnosis) of 32.3 years (Shirani et al. Mult Scler 2012;18:442-450).
The distinction between age at symptom onset and age at diagnosis is an important one. A BC database study reported that the mean delay from onset to first MS clinic visit was 12.1 years; only 25.3% of patients were seen at an MS clinic within 5 years of onset (Tremlett et al. J Neurol Neurosurg Psychiatry 2008;79:1368-1375). This is important because an MS clinic visit is required in BC (and other provinces) before a DMT can be initiated. Accordingly, a study using BC Health Claims data for the period 1996-2008 found that the average age at DMT initiation was 41.5 years after a mean disease duration of 9.9 years (Zhang et al. BMJ Open 2017;7:e018612). A subsequent health database analysis for the provinces of BC, Saskatchewan and Manitoba for the period 1996-2014 reported that the average age at initiation of a front-line injectable DMT was 40.2 years (Evans et al. Mult Scler Relat Disord 2016;8:78-85).
A more contemporary epidemiologic study in Saskatchewan, which defined the index date as either the first medical contact for MS or a diagnosis of an MS-like demyelinating condition, reported that the average age at incidence was 43 years (Al-Sakran et al. Can J Neurol Sci 2018;45:295-303).
Thus, it does not appear that the introduction of the McDonald criteria in 2001 led to earlier diagnosis and treatment (McDonald et al. Ann Neurol 2001;50:121-127). There is a paucity of data for the period 2010-2018, so it is unclear whether subsequent revisions to the criteria have effected a change in practice. In a recent German phase IV study, the average age of DMT initiation was 38 years (Gold et al. Ther Adv Neurol Disord 2018;11:1756286418768775). A Canadian study reported that the mean age when starting second-line fingolimod was 41 years (Lapierre et al. Can J Neurol Sci 2016;43:278-283). A US study reported that the average age when switching to an oral therapy (fingolimod or dimethyl fumarate) was 44 years (Ontaneda et al. Mult Scler Relat Disord 2018;27:101-111).
The observation that relapse reduction with a DMT is most effective in the first five years after diagnosis popularized the notion of a 5-year window of opportunity (Tremlett et al. Neurology 2009;73:1616-1623). The above data suggest that although MS can be identified earlier, many patients are not diagnosed and treated until their late thirties, when the therapeutic window has already closed.