Mechanisms of central neuropathic pain: a combined psychophysical and fMRI study in syringomyelia

Abstract

The pathophysiology of central pain syndromes is still poorly understood and their treatment remains a major challenge. It has long been suggested that lesions of the spinothalamic pathways are necessary for developing these pain syndromes. The recently proposed thermosensory disinhibition theory suggests that reduction of the inhibition of thermal sensory afferents that affect nociceptive systems may play a major pathophysiological role. Syringomyelia, which is frequently associated with central neuropathic pain, is characterized by a selective or preferential lesion of the spinothalamic tract resulting in thermosensory deficits of various extents and magnitudes. Thus, syringomyelia represents a unique 'pathological model' particularly suited to investigating the relationship between spinothalamic tract dysfunction, thermosensory deficits and pain. Here, we systematically compared the sensory loss (thermal and mechanical), using quantitative sensory testing, between 46 consecutive syringomyelia patients with or without neuropathic pain. We then further investigated the mechanisms of evoked pains in these patients, using functional MRI (fMRI) in a subgroup of patients with cold or brush-evoked allodynia, compared with patients without pain and healthy volunteers. We found no significant difference in the magnitude or extent of sensory deficits between patients with or without neuropathic pain, suggesting that lesions of the spinothalamic pathways are not sufficient for developing central pain. However, a different pattern of sensory deficits was observed between patients with spontaneous pain only (n = 11) and patients with both spontaneous pain and allodynia (n = 20), suggesting that the mechanisms of central pain are not univocal. In patients with spontaneous pain only, the thermal sensory loss was significantly more asymmetrical and there was a direct relationship between the extent of thermosensory deficits (i.e. deafferentation) and the intensity of burning pain. In contrast, patients with allodynia had reduced thermal deficits, in terms of both magnitude and extent. In addition, the sensory deficits were different between patients with cold or tactile allodynia, suggesting distinct pathophysiological mechanisms related to the sub-modalities of allodynia. Our fMRI study further confirmed this, showing that different sub-types of allodynia were associated with distinct patterns of brain activity, which do not necessarily correspond to the 'pain matrix' involved in acute physiological pain. The prefrontal cortex was the only area consistently activated by pathological evoked pains, suggesting that alteration of high-level pain modulatory mechanisms might play a major role in allodynia due to central lesion.