Introduction

In the late 1960s Tom Sears and I showed that either complete or partial conduction block resulted from demyelination of central nerve fibres induced by diphtheria toxin. When conduction survived it was slow, and insecure; the refractory period of transmission (a term coined by Tom in the course of these experiments) was prolonged and the damaged fibres were unable to conduct long trains of impulses at high frequencies (McDonald & Sears, 1970).

The introduction of evoked potential methods for assessing transmission in afferent pathways in man in the 1970s then made it possible to interpret some of the clinical phenomena of demyelinating disease (and in particular of multiple sclerosis) on the basis of the earlier experimental work. Multiple sclerosis (MS) is characterized by four main pathological changes: demyelination with preservation of axons, Wallerian degeneration (scanty in the early stages, more marked later), astrocytic proliferation and varying amounts of inflammation. A variable amount of remyelination also occurs. In considering the mechanism of the conduction changes, it seemed likely that demyelination per se made an important contribution as it does in experimental demyelination in the peripheral nervous system (McDonald, 1963; Rasminsky & Sears, 1972). Whether this was the whole explanation remained an unexplored issue until the technical advances of the 1980s led to the application of high-resolution magnetic resonance imaging (MRI) to the study of MS. These advances allowed us to tackle the questions ‘What factors contribute to relapse?’ and ‘What factors contribute to remission?’