Objective: The aim of this study is to investigate phantom limb pain (PLP), cortical reorganization and the modulation of PLP by visual input. Mirror therapy (viewing the intact limb in a mirror so that it appears to be the missing limb) has been shown to reduce PLP. This reduction demonstrates that vision can play a key role in modulating PLP. The study proposed here will identify brain regions critical for the success of this therapy using functional magnetic resonance imaging (fMRI).
Study population: Individuals with unilateral amputation of individual limbs and at least three episodes of PLP a week. Additionally, we will study amputees without PLP and healthy control volunteer subjects with no limb amputations.
Design: Amputee subjects with PLP will be scanned using fMRI while they undergo mirror therapy for PLP. During fMRI scans subjects will see images of hands or feet, move their limbs or experience tactile stimulation. Mirror therapy requires four weeks to be effective. Amputee subjects will be scanned prior to starting therapy, two weeks after starting mirror therapy and again after four weeks of mirror therapy. Mirror therapy sessions will be conducted daily and subjects will also complete daily questionnaires (Visual Analogue Scale, or VAS, and Short-Form McGill Pain Questionnaire or SF-MPQ) assessing their level of pain and number of PLP episodes. Amputees without PLP and one group of healthy control subjects will go through the same series of functional imaging scans over the same time period as the amputee subjects, allowing us to identify brain changes that may be associated with repeated scanning (e.g. due to practice of tasks) rather than resulting from the mirror therapy. Another group of control subjects will be scanned once to identify particular brain regions that might be involved in the integration of visual, motor and somatosensory information for use in the analysis of the amputee subject data.
Outcome measures: Changes in fMRI activation during the course of mirror therapy will be compared with data on the number of episodes of PLP, the average length of episodes, and the average intensity of pain assessed using the VAS and SF-MPQ. Analyses will focus on cortical regions known to be involved in visual, tactile and motor representations (including primary motor cortex, primary and secondary somatosensory cortex and regions of visual cortex selectively responsive during the viewing of body parts). Additional brain regions that might be involved in integrating visual, motor and somatosensory information will be identified in control subjects. Direct comparisons will be made between cortical regions (e.g. somatosensory regions) involved in the representation of the amputated limb compared with the intact limb and between amputee subject data and control subject data.