Abstract
Dear Sir The recent paper by Coste et al. (1) on the role of wind-up in trigeminal sensory processing was most interesting, and their findings may have great relevance for a number of neurovascular headaches. Wind-up is a frequency-dependent increase in the excitability of spinal cord neurons, evoked by electrical stimulation of afferents, and which produces an increase in the number of discharges in second-order neurons in response to arriving stimuli. The mechanism of wind-up is unclear, but it has been seen as a system for the amplification in the spinal cord of nociceptive messages (2–4).
The most prominent intracranial pain-sensitive structures are the large penetrating arteries and the small arteries of the dura mater, a discovery made many years ago (5). Naturally, these arteries pulse in time with the heartbeat, and consequently so does the pain of migraine headache. This throbbing nature of the pain has usually been attributed to the activation of stretch receptors with the systolic pressure wave in the large arteries (6). The pulsating nature of the pain can often be abolished by compression of a carotid artery (6, 7). A minority of neurons in the trigeminal nucleus of cats that receive sensory input from stretch receptors in the walls of these arteries show rhythmical activity, also synchronized with the heart beat (8, 9), and this too can be eliminated by carotid compression. In humans, the frequency of the stimulus arising from vascular pulsation is about 1.2 Hz. In cats, first-order trigeminal sensory neurons are ‘self-inhibitory’ inasmuch as neuronal discharge, whether spontaneous or initiated by a sensory stimulus, often produces an extended refractory period of 50–1000 ms, during which no further discharge can occur (10, 11). It has been hypothesized that this refractoriness is mediated by recurrent inhibition occurring in local circuitry in the dorsal horn (10). In our experiments (11), this refractoriness seems to be more frequent in neurons that receive sensory input from the dura mater and is more prominent in neurons in the cervical spinal cord than in those in the trigeminal nucleus. The reason for the difference between the trigeminal nucleus and its spinal equivalent is completely unknown, but it may be relevant to the slightly different findings made by Coste et al. (1) in the trigeminal nucleus and Bolton et al. (12) in the cervical spinal cord. This refractory period means that vascular stimuli that repeat at frequencies substantially above 1 Hz will not always produce a neuronal response, and possibly not produce wind-up. The frequency of stimulation in the experiments reported by Coste et al. (admittedly in rats) is 0.66 Hz. In another report, also in rats, Bolton et al. (12) were able to elicit wind-up of cutaneous facial afferents, but not of dural afferents; they used frequencies of 1 and 2 Hz.
These observations suggest that trigeminovascular neurons might have properties that predispose them towards wind-up of sensory signals arriving at them only at frequencies equivalent to the heart rate in humans and thus predispose them towards amplifying the sensations produced by vascular pulsations, thus accentuating the throbbing nature of vascular headaches.