Pain in multiple sclerosis

Abstract

Background: Pain is frequently reported by people with multiple sclerosis (MS). It has been associated with decreased quality of life, psychiatric morbidity, interference with day to day activities, and frequent healthcare attendance. It has been reported by people with multiple sclerosis to be one of their most important symptoms, and available treatments are limited in their effectiveness. Despite this, our understanding of the epidemiology and mechanisms of pain in people with MS are limited. Our understanding of the interactions of central nervous system mechanisms and pain states overall is growing. However, the application of this knowledge to MS is incomplete. Previous studies have shown that the descending pain modulatory system (DPMS) is an endogenous network of cortical and subcortical brain structures which act to limit, or accentuate, an individual’s perception of pain, via descending brainstem pathways. Associated clinical measures include depression, anxiety, and cognitive flexibility. Our understanding of the function or dysfunction of this system in MS is limited. We do not know if the MS disease process may adversely affect the structure or function of the DPMS. Hypothesis: In people with neuropathic limb pain in relapsing remitting MS (RRMS), compared to people with RRMS who do not have pain, there will be disruption of the endogenous descending pain modulatory system. This will manifest as impaired descending inhibition of pain. Aims and Methods Establishing the background using systematic reviews: The first aim of this thesis was to establish the prevalence, natural history and associations of pain (and pain syndromes) occurring in people with MS. The second aim was to explore existing knowledge of how the MS disease process may contribute to pain states, using a systematic review of neuroimaging studies. Prospective clinical study: A case-control study of 47 people with RRMS was then carried out. 31 of these had neuropathic pain in the limbs, and 16 did not have pain. Using targeted assessments, function of the descending pain modulatory system was assessed in the following ways: First: Detailed clinical, behavioural and neuropsychological assessment, focussing on cognitive, behavioural and affective features known to be closely related to the DPMS. Second: MRI imaging of brain structure, focussing on the volume and location of MS lesions, as well as the volume of key grey-matter structures involved in the DPMS. Third: Resting state functional MRI imaging of the brain, focussing on functional connectivity between the rostral anterior cingulate cortex and two other key DPMS structures (dorsolateral prefrontal cortex, and periaqueductal gray). Results: Systematic reviews: Meta-analysis of existing prospective studies confirmed that pain is very common in MS, affecting about 63% of people with MS on average (95%CI between 55 and 70%). Many different types of pain contribute to this overall estimate. No significant associations with disease course or stage emerged. Several neuroimaging studies have assessed people with MS-associated pain using MRI. These studies were often small, and with associated methodological issues. It is likely that location of MS lesions is implicated in aetiology of pain syndromes in some cases, though our overall knowledge is limited. Prospective study: In a prospective study, people with and without pain were matched for age and gender. Furthermore, groups were balanced for a range of other variables. The pain group more frequently received gabapentinoid medications. The presence of pain was significantly associated with increased scores for depression, fatigue and catastrophising, as well as with specific impairments at neuropsychological assessment, including cognitive flexibility. Many of these impairments are directly relevant to existing models of the DPMS. Overall volume of MS lesions was not different in people with pain, though lesions were more likely to occur in the brainstem. Some alterations of grey-matter volumes in people with pain which mirrored studies of pain disorders outside MS were found, but these did not involve structures key to the DPMS. Affected structures included trigeminothalamic nucleus (relative volume increase in pain group), posterior cingulate cortex and parahippocampal gyrus (volume decrease in pain group). Functional connectivity of the rostral anterior cingulate cortex to the periaqueductal grey matter, a key structure in the descending modulation of pain, was stronger in the group without pain. Conversely, functional connectivity to the dorsolateral prefrontal cortex, repeatedly implicated in the DPMS and thought to be involved in cognitive evaluation and flexibility, was stronger in the pain group. MS lesion volume appeared to account for some of this difference in a multivariate analysis. Limitations: Key limitations of this work include cross-sectional design, small sample size, and number of statistical comparisons carried out. Conclusions: Systematic reviews examined the prevalence, natural history and associations of pain in MS, as well as examining existing neuroimaging studies which investigated how the MS disease process could contribute to pain states. A prospective study found evidence of both emotional/affective and cognitive dysfunctions relevant to the hypothesis of dysfunction in the DPMS. Higher likelihood of MS lesions in the brainstem could be relevant to DPMS function. Separately, there were structural grey-matter volume alterations reflecting those found in many pain studies outside MS. Importantly, however, these did not affect key DPMS structures. Resting state functional MRI however demonstrated altered connectivity of core DPMS structures, which may be partly mediated by MS lesion volume. Functional connectivity findings could be consistent with the hypothesis of impaired descending pain inhibition, in people with relapsing remitting MS affected by neuropathic limb pain.