Assessment of heat sensation following spinal cord injury - a pilot study

We recruited participants with complete established (greater than or equal to 12 months ) thoracic spinal cord injury who were medically stable. The extent of spinal cord injury (SCI) damage was recorded according to the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) and classified by the American Spinal Injury Association Impairment Scale - AIS. Each participant was examined by the same clinician (PW). Quantitative Sensory Testing (QST) Quantitative Sensory Testing (mechanical and thermal stimuli) was undertaken to establish the degree of sensory loss or gain in participants with SCI above and below their neurological level of injury (arm, medial knee and dorsum of foot) in comparison with healthy controls. The German Research Network on Neuropathic Pain standardised protocol was used to determine sensory thresholds for cold detection, cold pain, warm detection, warm pain, mechanical detection, mechanical pain, pressure pain and vibration detection. Contact Heat Evoked Potentials (CHEPs) Participants were placed in a 22°C±2°C room and asked to keep their eyes open during testing and to fixate on a stationary point to minimise eye movement artefact. Fifteen contact heat stimuli (27 mm diameter thermode, Medoc Pathway System, Ramat Yishai, Israel) using a brief heat pulse (500 ms, 51°C , 70°C/s increasing rate, 40°C/s return rate) were applied in a pseudo random fashion (8-12s intervals) across a small area of skin (7.5x7.5cm) marked at each test site. To reduce peripheral desensitisation the thermode was moved within this area to ensure no two successive stimuli were delivered to the same area of skin. A stimulus-triggered EEG was recorded (V-amp 8, Brain Products GmbH, Munich, Germany) using disposable Ag-AgCl electrodes (bandpass 0.5-35 Hz, sample rate 500 Hz) from Fz, Cz, Pz, T3 and T4 [international 10-20 system, referenced to linked earlobes]. Blink artefacts were monitored by an electrooculogram (EOG) recorded from lateral and infraorbital electrodes. Electrode impedance was maintained below 5 kO by cleaning the skin with alcohol wipes and an abrasive skin prepping gel. A familiarisation series of 15 stimuli was carried out on the contralateral (generally left) forearm and the same sites used for QST were also used for CHEPs testing. In the SCI-BLNP group, CHEP stimuli were administered within their area of pain if it was not already tested using the routine sites. Optimising contact heat evoked potential recording To optimise the acquisition of cortical potentials during CHEP stimulation we tested two baseline temperatures: 38°C and 42°C. All participants were tested with both baselines temperatures at the arm, healthy controls also received both baselines at knee and foot sites. Given their profound loss of sensation, participants with SCIs were tested using the 42°C baseline at test sites below their neurological level of injury. The effect of pre-treatment with a chemical sensitising agent was also assessed. Three millilitres of capsaicin solution (0.6% (w/v) capsaicin solution in 45% ethanol) was applied to a 7.5 x 7.5 cm gauze patch that was covered and held in place with a transparent film dressing for 15 minutes. CHEP testing series were carried out in the following order: 38°C baseline (where appropriate); 42°C; capsaicin at either 38°C or 42°C (based on the highest temperature required to elicit an evoked potential). For each test site stimulus-triggered cortical traces (2.5 s including 1s pre stimulation) were recorded from Cz and averaged to produce the final waveform for analysis. Automatic eye-blink detection and correction was used, as was a semi-automated process of trace acceptance or rejection. Each stimulus-linked trace was visually examined and rejected if significant artefact was present (muscular contraction, movement artefact, significantly unstable baseline), otherwise they were included in the analysis. The number of readable traces, out of 15, as well as the number of traces containing an evoked potential were recorded. For the final averaged waveform we recorded presence or absence of an EP, N2-P2 amplitude, and N2 and P2 latency. Perceived pain intensity of the CHEP stimulus was expressed verbally using the Numerical Rating Scale (NRS-101, where 0 = no pain and 100 = most intense pain imaginable), three seconds after each stimulus to avoid interfering with the EEG recording. If the CHEP stimulus was not felt (NF) this was recorded. All of the above tests were completed in a single visit of approximately 3 hours.