LETTER TO THE EDITOR Predicting a window of therapeutic opportunity in multiple sclerosis

Sir, The relationship between cumulative relapses and unremitting progressive disability is highly relevant to clinical practice and to randomized controlled trial design. Neuropathological studies unmentioned by Andersen (2010) have not been able to confirm the widespread belief that severe disability accumulates as a result of successive exacerbations. DeLuca et al. (2006), examining one of the largest reported pathological cohorts of multiple sclerosis cases, demonstrated lack of correlation between plaque load and axonal loss in the corticospinal tracts throughout the length of the spinal cord where the clinical phenomena characterizing the progressive phase typically localize. Disease progression and inflammatory attacks are probably driven by different mechanisms (Trapp and Nave, 2008). This became clear in the interferon and monoclonal antibody studies which showed no impact on disease evolution following relapse suppression, and MRI studies have provided but weak or non-existent correlations between lesion load and long-term disease evolution (Fisniku et al., 2008). Previous natural history studies have demonstrated predictive value of early disease features, including early relapse rate (Weinshenker et al., 1989b; Eriksson et al., 2003). Nevertheless, this was no longer reported to apply once permanent disability occurs (defined as 4 on the Disability Status Scale [DSS]; Confavreux et al., 2003). Outcome appears to be largely determined during the early stage of the disease, and once progression has begun, its rate seems independent of factors preceding it (Confavreux et al., 2003). It is unaffected by superimposed inflammatory attacks (Kremenchutzky et al., 1999) and it is homogeneous among progressive subtypes (Kremenchutzky et al., 2006). Number of relapses during the first 2 (Weinshenker et al., 1989b) and 5 years (Kantarci et al., 1998; Confavreux et al., 2003) had modest predictive value but the 5-year data did not show any additional effect beyond 2 years. Causality could not be assumed. Previous studies had examined neither the independent predictive role for long-term outcomes of relapses after the second year nor the relationship between inflammatory attacks and the onset of the progressive phase. The London Ontario database, advantaged by geographical ascertainment and 28 years follow-up, provided opportunities to address in detail these aspects of disease course. The large number of patients reaching hard outcomes (67% at DSS 6 and 48% at DSS 8) and low percent of censored information warranted high reliability to the survival estimates. In contrast to Andersen’s assertion, we have provided strong evidence that characteristics of the onset attack have little or no prognostic value. The degree of remission from first relapse had no significant impact on long-term disease evolution (Kremenchutzky et al., 2006) and the current study showed no independent effect exerted by type and number of neurological systems involved at onset. We did confirm modest predictive value of number of relapses during the first 2 years, driven by the minority (21%) having three or more attacks during this period. This appeared to operate by increasing the probability of entering the secondary progressive phase and by shortening the latency to progression, also unmentioned by Andersen (2010). Patients with more frequent early relapses seem unable to suppress mechanisms evolving into progressive disability accumulation, are more likely to convert to secondary progressive multiple sclerosis in shorter time and are therefore at higher risk of developing severe disability. The earlier onset of progression seems either to suppress or mask further relapses, explaining the inverse relationship between number of doi:10.1093/brain/awq226 Brain 2010: 133; 1–3 | e162