Brain dynamics of scalp evoked potentials and current source densities to repetitive (5-pulse train) painful stimulation of skin and muscle: central correlate of temporal summation.

Temporal summation is a potent central somatosensory mechanism and may be a major mechanism involved in e.g. neuropathic pain. This study assessed the long-latency somatosensory evoked potentials (SEPs) in response to trains of repeated painful electrical stimulation of human skin and muscle in order to investigate the cerebral representation of temporal summation. Forty series of stimuli were delivered at stimulus intensities corresponding to moderate pain levels in 20 young men. Each series consisted of a five-burst-pulses (1 ms) train delivered at 2 Hz, known to activate temporal summation, i.e. increased pain intensity during the series of stimuli. Grand mean averaged waveforms (31 ch. EEG) were obtained in response to the skin and muscle stimulation. In the "train" SEPs, the wave morphology was characterized by four peak components after the first stimulus (100 to 450 ms) and by three components after the fifth stimulus (2100-2145 ms). The latency was significantly prolonged for muscle stimulation only. The 3D topographic maps at the peak activation time (100, 140, 250, and 450 ms) showed clear reduction in the amplitudes and their spatial extent (P4/P100-Fc2/N100, POz/P140-Fc2/N140, Cz/P250, Cz/N460) betweenthe first and the fifth stimulus. The current source density (CSD) topology exhibited markedly differential patterns changing from the first to the fifth stimulus. For the skin stimulation, the fifth stimulus was associated with a distinct emergence of the frontal negativity source at Fc2 right frontal cortex. This was consistent across the 100,140, 250, and 450 peak components but was not even visible in the first stimulus. In the muscle, the fifth stimulus was associated with a marked reduction of the frontal positivity at contralateral F4 site in the early stages at 100 and 140 ms, and with a total disappearance of positive source at Cz. In summary, this study demonstrated a clear temporal summation of psychophysical ratings, reduction of the peak amplitudes in the last of the first stimuli, dissociation from simple amplitude increase of the cerebral responses to pain, and a concurrent transformation of the CSD patterns. This change in "rapid cortical dynamics" of short-term plasticity could be an important mechanism for wind-up and pain processing in the brain.