Reliable EEG responses to the selective activation of C-fibre afferents using a temperature-controlled infrared laser stimulator in conjunction with an adaptive staircase algorithm

Brain responses to the activation of C-fibres are obtained only if the co-activation of Aδ-fibres is avoided. Methods to activate C-fibres selectively have been proposed, but are unreliable or difficult to implement. Here, we propose an approach combining a new laser stimulator to generate constant-temperature heat pulses with an adaptive paradigm to maintain stimulus temperature above the threshold of C-fibres but below that of Aδ-fibres, and examine whether this approach can be used to record reliable C-fibre laser-evoked brain potentials. Brief CO2 laser stimuli were delivered to the hand and foot dorsum of 10 healthy subjects. The stimuli were generated using a closed-loop control of laser power by an online monitoring of target skin temperature. The adaptive algorithm, using reaction times to distinguish between late detections indicating selective activation of unmyelinated C-fibres and early detections indicating co-activation of myelinated Aδ-fibres, allowed increasing the likelihood of selectively activating C-fibres. Reliable individual-level electroencephalogram (EEG) responses were identified, both in the time domain (hand: N2: 704 ± 179 ms, P2: 984 ± 149 ms; foot: N2: 1314 ± 171 ms, P2: 1716 ± 171 ms) and the time-frequency (TF) domain. Using a control dataset in which no stimuli were delivered, a Receiver Operating Characteristics analysis showed that the magnitude of the phase-locked EEG response corresponding to the N2-P2, objectively quantified in the TF domain, discriminated between absence vs presence of C-fibre responses with a high sensitivity (hand: 85%, foot: 80%) and specificity (hand: 90%, foot: 75%). This approach could thus be particularly useful for the diagnostic workup of small-fibre neuropathies and neuropathic pain.