Analgesic triptan action in an animal model of intracranial pain A race against the development of central sensitization.код для вставкиСкачать
Analgesic Triptan Action in an Animal Model of Intracranial Pain: A Race against the Development of Central Sensitization Rami Burstein, PhD,1–3 and Moshe Jakubowski, PhD1 We have shown that the development of cutaneous allodynia (exaggerated skin sensitivity) during migraine is detrimental to the anti-migraine action of the 5HTIB/ID receptor agonists known is triptans. Because cutaneous allodynia is a manifestation of sensitization of central trigeminovascular neurons, we examined whether triptan treatment can intercept such sensitization before its initiation or after its establishment in our rat model for cutaneous allodynia induced by intracranial pain. Single-unit recordings were obtained from spinal trigeminal neurons that proved to received convergent inputs from the dura and facial skin. The effects of sumatriptan (300 g/kg i.v.) on central sensitization induced by topical application of inflammatory soup (IS) on the dura were determined when the drug was administered either 2 h after IS (late intervention) or at the same time as IS (early intervention). Late sumatriptan intervention counteracted two aspects of central sensitization: dural receptive fields, which initially expanded by IS, shrunk back after treatment; neuronal response threshold to dural indentation, which initially decreased after IS, increased after sumatriptan. On the other hand, late sumatriptan intervention did not reverse other aspects of central sensitization: spontaneous firing rate and neuronal response magnitude to skin brushing which initially increased after IS, remained elevated after sumatriptan; response threshold to heating of the skin, which initially dropped after IS, remained low after sumatriptan. Early sumatriptan intervention effectively blocked the development of all aspects of central sensitization expected to be induced 2 h after IS application: dural receptive fields did not expand; neuronal response threshold to dural indentation and skin stimulation did not decrease; spontaneous firing rate did not increase. The early treatment results suggest that triptan action provides a powerful means of preventing the initiation of central sensitization triggered by chemical stimulation of meningeal nociceptors. The late treatment results suggest that triptan action is insufficient to counteract an already established central sensitization. Thus, triptan action appears to be exerted directly on peripheral rather than central trigeminovascular neurons. Ann Neurol 2004;55:27–36 On the basis of our findings that a brief application of inflammatory mediators to the rat dura sensitizes both peripheral and central trigeminovascular neurons, we have proposed that during migraine, peripheral and central sensitization are manifested, respectively, as throbbing headache and cutaneous allodynia (ie, pain resulting from a nonnoxious stimulus to normal skin).1 Our preclinical studies demonstrated that, after sensitization, perivascular meningeal nociceptors, as well as central trigeminovascular neurons that receive convergent input from the dura and skin, became hyperresponsive to low-force dural indentation to which they were essentially mute before sensitization.2– 4 These preclinical studies further demonstrated that sensitized central trigeminovascular neu- rons became extremely sensitive to otherwise mild tactile and thermal stimulation of the periorbital skin. Indeed, subsequent clinical studies showed that as many as 75% of our patients developed cutaneous allodynia during migraine attacks.5,6 There are remarkable similarities between the temporal changes in the development of peripheral and central sensitization in the rat, and their corresponding clinical manifestations during the course of migraine. The induction of peripheral sensitization in the rat occurs rapidly within 5 to 20 minutes after applying inflammatory soup (IS) onto the dura,2,7 whereas central sensitization develops between 20 to 120 minutes before it becomes firmly established after IS.3 Similarly in patients, throbbing transpires some 5 to 20 minutes From the 1Departments of Anesthesia and Critical Care, Beth Israel Deaconess Medical Center; and 2Department of Neurobiology and the 3Program in Neuroscience, Harvard Medical School, Boston, MA. Address correspondence to Dr Burstein, Department of Anesthesia and Critical Care, Harvard Institutes of Medicine, Room 830, 77 Avenue Louis Pasteur, Boston, MA 02115. E-mail: firstname.lastname@example.org Received Apr 11, 2003, and in revised form Jul 16. Accepted for publication Jul 16, 2003. © 2003 American Neurological Association Published by Wiley-Liss, Inc., through Wiley Subscription Services 27 after the onset of headache, whereas cutaneous allodynia starts between 20 and 120 minutes and becomes firmly established only 120 to 240 minutes after the onset of headache.6 According to our scenario, meningeal nociceptors become sensitized during migraine a few minutes after their initial activation, resulting in throbbing headache and its exacerbation by bending over. The continued barrage of impulses arriving from sensitized meningeal nociceptors gradually stimulates the development of central sensitization in spinal trigeminovascular neurons, resulting in cutaneous allodynia in the same area as the referred head pain. Eventually, as we observed in the rat,3 central trigeminovascular neurons change their physiological properties and remain inveterately sensitized, independent of incoming impulses from meningeal nociceptors. We recently have conducted a clinical study showing that triptans (5HT1B/1D receptor agonists), the best studied and most commonly used family of antimigraine drugs, can render migraine patients pain-free depending on whether or not they exhibited cutaneous allodynia at the time of treatment.8 Patients who had never developed allodynia were rendered pain-free by triptan therapy at any time after the onset of migraine pain. However, patients bound to develop cutaneous allodynia during migraine were rendered pain-free and allodynia-free if triptan therapy was given before, but not after, the establishment of allodynia (ie, within 1 hour of the onset of pain). In this study, we used our animal model of intracranial pain and cutaneous allodynia to determine (1) whether early sumatriptan administration can prevent the development of central sensitization and cutaneous allodynia, and (2) whether late sumatriptan administration can reverse inveterate central sensitization and cutaneous allodynia. Materials and Methods Surgical Preparation Male Sprague-Dawley rats (350 –500gm) were anesthetized with urethane (1.2gm/kg, IP) and treated with atropine (0.04mg, IP). The femoral vein was fitted with indwelling cannula for later administration of sumatriptan. The trachea was fitted with an indwelling metal tube for artificial ventilation, and the rat head was affixed in a stereotaxic apparatus. End-tidal CO2 and core temperature were continuously monitored and kept within physiological range. The ventral half of the occipital bone and the laminar process of C1 were removed to lower a recording electrode into the trigeminal nucleus caudalis. Dura mater overlying the dorsal surface of the left hemisphere and transverse sinus was exposed by carefully removing the frontal and parietal bones. To keep the moisture, we covered the exposed brainstem with mineral oil, and the exposed dura was flooded with synthetic interstitial fluid, consisting of 10mM Hepes, 5mM KCl, 1mM MgCl2, 5mM CaCl2, and 135mM NaCl, pH 7.3. 28 Annals of Neurology Vol 55 No 1 January 2004 Neuronal Identification and Selection Central trigeminovascular neurons receiving converging input from the dura and facial skin were first identified by mapping their intracranial and extracranial receptive fields, using electrical stimulation of the dura and mechanical stimulation of the skin as we described elsewhere.3 Next, each neuron was characterized for its baseline responses to a series of stimuli. Dural stimulation was performed by indentation with a series of calibrated von Frey hairs. Stimulation of the facial skin included the following: brushing with paintbrush; pressing with loose arterial clip; incremental heating using a contact thermal stimulator. Each series of stimuli was repeated at least three times at 5 to 10-minute intervals, and only neurons showing less than 10% variability in responses were selected to be further studied. Selected neurons then were sampled for their baseline rate of spontaneous firing over a period of 10 minutes. To induce sensitization in central trigeminovascular neurons, we bathed the hotspot of the dural receptive field for 5 minutes in IS (1mM histamine, serotonin, bradykinin, 0.1mM prostaglandin E2 in 10mM Hepes buffer, pH 5.5; adapted from Steen and colleagues9,10). In the initial part of the study, we found that application of IS onto the dura produced activation in all neurons studied but resulted in sensitization in 14 of 15 neurons that received both A␦- and C-fiber input and in none of the six neurons that received only A␦-fiber input. Therefore, we selected only neurons that got activated upon IS application onto the dura and proved to receive both A␦- and C-fiber input. Sumatriptan Treatment Sumatriptan solution (300g/ml) was infused intravenously at a rate of 15g/minute over 7 to 10 minutes, to a final dose of 300g/kg body weight. The choice of injectable sumatriptan treatment in this study was based on the evidence that, of all the triptans, this formulation provides the highest therapeutic gain in migraine patients.11 Animal studies have shown that intravenous administration of sumatriptan at doses ranging from 100 to 300g/kg produces constriction of dilated blood vessels,12,13 inhibition of dural plasma protein extravasation,14 neurogenic inflammation,15 neurogenic vasodilatation,16 arterial blood flow,17 and release of calcitonin gene–related protein18,19 and substance P.20 The higher sumatriptan dose of 300g/kg was chosen to maximize potential inhibitory action of the drug on central sensitization and to minimize the probability of negative results due to lower ineffective dose. Experimental Protocol Two groups of neurons were used to examine whether intravenous infusion of the 5HT1B/1D receptor agonist sumatriptan (GlaxoSmithKline, Research Triangle Park, NC) can prevent the initiation of IS-induced central sensitization (early treatment) or reverse ongoing central sensitization (late treatment). Group 1 (n ⫽ 9) was designated for late sumatriptan treatment. At 1 and 2 hours after the application of IS onto the dura, we remapped dural receptive field size, reexamined neuronal responses to stimulation of dura and skin, and resampled spontaneous firing rate. Resulting values for each measure were then compared with the respec- tive baseline values obtained before IS application to ascertain that central sensitization took place. As described in our earlier work,3 mandatory criteria for central sensitization included expansion of receptive fields, enhanced sensitivity to dural indentation, decreased thresholds and increased response magnitude to mechanical and thermal stimulation of the skin, and increased spontaneous activity. Once central sensitization was ascertained, sumatriptan was infused intravenously, and all measurements described above were repeated again at 3 and 4 hours after the application of IS (⫽ 1 and 2 hours after administration of sumatriptan). Group 2 (n ⫽ 9) was designated for early sumatriptan treatment, such that sumatriptan was infused intravenously at the same time as IS was applied onto the dura. Resulting receptive field size, neuronal response threshold or magnitude to each stimulus, and spontaneous firing rate were remeasured 1 and 2 hours later and compared with the respective baseline values obtained before any treatment to determine whether or not sumatriptan prevented the induction of central sensitization by IS. Statistical Analysis Data are presented as mean ⫾ standard error and analyzed using nonparametric statistics.21 Spontaneous activity and neuronal responses to dural or skin stimulation were analyzed over three time points (0, 1, 2 hours), using Friedman two-way analysis of variance. Post hoc paired comparisons were performed using one-tailed Wilcoxon matched-pairs signed ranks test. Level of significance was set at 0.05. Results Recording Sites and Cutaneous Receptive Fields of Central Trigeminovascular Neurons Recording sites and maps of cutaneous receptive fields did not differ between neurons tested for early sumatriptan intervention and those tested for late intervention (Fig 1). All identified recording sites were localized to laminae IV and V of the first cervical segment. In most cases, cutaneous receptive fields were restricted to the periorbital/ophthalmic area. This area remained the most sensitive spot even in cases in which the receptive field extended over the entire face. Spontaneous and Inflammatory Soup–Induced Activity of Central Trigeminovascular Neurons Late sumatriptan intervention did not reverse the longterm increase in neuronal spontaneous firing induced by IS (Fig 2A). Neuronal firing rate, which increased sixfold ( p ⫽ 0.002) within 2 hours of IS application, remained elevated fivefold after sumatriptan infusion compared with the initial rate of spontaneous activity ( p ⫽ 0.008; see Fig 2C). Early sumatriptan intervention effectively prevented the long-term increase in spontaneous activity seen in the late intervention group, although it did not block neuronal activation (the initial burst in neuronal firing) that typically takes place upon IS application (see Fig Fig 1. Recording sites and cutaneous receptive fields of trigeminovascular neurons in the late (A) and early (B) intervention groups. The photomicrographs of electrolytic lesions show the recording locations of neurons in the deep laminae (IV, V) of the dorsal horn at first cervical segment. The schematic mapping of facial receptive fields (light gray areas) show that the periorbital skin contains the most sensitive region of the receptive field (dark gray area). Numbers identify individual cases. 2B). Neuronal firing rate did not increase; in fact, it decreased slightly ( p ⫽ 0.048; see Fig 2D). Dural Receptive Fields of Central Trigeminovascular Neurons Late sumatriptan intervention effectively reversed ISinduced expansion of dural receptive field. In 89% of the cases, dural receptive field expanded within 2 hours of IS application and contracted back 1 to 2 hours after sumatriptan infusion; with the exception of two cases (Cases 2 and 34), dural receptive field returned to its original, presensitization size (Fig 3A). Burstein and Jakubowski: Sumatriptan and Sensitization 29 Fig 2. Spontaneous and ISinduced firing of central trigeminovascular neurons. (A) Mean firing rate of a neuron sampled at 1-hour intervals before (baseline/green) and after (red) application of IS, and again after a delayed administration of sumatriptan (blue). Notice that spontaneous activity increased from a baseline of 2 spikes/sec to 28 spikes/sec 4 hours after IS application and remained elevated at 18 spikes/sec after sumatriptan infusion. (B) Mean firing rate of a neuron sampled before (baseline/green) and after (purple) simultaneous application of IS and sumatriptan. Notice that spontaneous activity increased transiently (from 0.1 to 10 spikes/sec) for only 20 minutes. (C, D) Summary of spontaneous firing rate (mean ⫾ SE) in late (n ⫽ 9) and early (n ⫽ 9) sumatriptan intervention groups, respectively. Early sumatriptan intervention prevented the expansion of dural receptive fields seen in the late intervention group. In 78% of the cases, dural receptive field size remained unchanged; the remaining 22% (Cases 12 and 30) expanded transiently (see Fig 3B). Sensitivity of Central Trigeminovascular Neurons to Dural Indentation To one degree or another, late sumatriptan intervention reversed IS-induced increase in neuronal sensitiv- 30 Annals of Neurology Vol 55 No 1 January 2004 ity to dural indentation. In 56% of the cases, neuronal threshold increased back to its presensitization value, in 33% it increased only partially, and in the remaining 11% the threshold did not increase at all (Fig 4A). Overall, neuronal threshold for dural indentation plummeted significantly ( p ⬍ 0.001) within 2 hours of IS application and then increased significantly ( p ⫽ 0.005) within 2 hours of sumatriptan infusion (see Fig 4C). Early sumatriptan intervention effectively prevented IS from inducing increased neuronal sensitivity in all Fig 3. Dural receptive fields of individual central trigeminovascular neurons. (A) Receptive field sizes mapped before (baseline/green) and after (red) the application of IS, and again after a delayed administration of sumatriptan (blue). Notice that dural receptive fields that expanded by IS, contracted back to their original size by sumatriptan. (B) Receptive field sizes mapped before (baseline/ green) and after (purple) simultaneous application of IS and sumatriptan. Notice that dural receptive fields did not expand. Numbers on top identify individual cases. Inset depicts meningeal sinuses. cases. In fact, neuronal threshold for dural indentation slightly increased (rather than decreased) in 57% of the cases (see Fig 4B), and the overall increase in threshold approached the level of significance ( p ⫽ 0.056; see Fig 4D). Sensitivity of Central Trigeminovascular Neurons to Mechanical Stimuli at the Periorbital Skin Late sumatriptan intervention did not reverse ISinduced increase in neuronal sensitivity to mechanical stimulation at the periorbital skin (Fig 5A). There was a twofold increase in neuronal response magnitude to brush ( p ⫽ 0.004) and pressure ( p ⫽ 0.005) within 2 hours of IS application, and sumatriptan infusion produced no change in these elevated responses (see Fig 5C). Early sumatriptan intervention effectively prevented IS from inducing such increased neuronal sensitivity to mechanical skin stimuli (see Fig 5B). Neuronal response magnitude to brush and pressure did not show any significant increase from baseline values (see Fig 5D). Sensitivity of Central Trigeminovascular Neurons to Thermal Stimulation of the Periorbital Skin Of the 18 neurons studied, 2 (1 of each group) were heat insensitive and therefore were excluded from the analysis. Three additional neurons in the late intervention group failed to exhibit a decrease in heat threshold after IS application and therefore could not be tested for potential inhibitory effects of late sumatriptan infusion. Late sumatriptan intervention did not reverse ISinduced increase in neuronal sensitivity to heat stimuli at the periorbital skin (Fig 6A). Neuronal response threshold for heat decreased by 4 to 6°C ( p ⫽ 0.008) within 2 hours of IS application and remained low after sumatriptan infusion compared with the initial heat threshold value ( p ⬍ 0.05; see Fig 6C). Early sumatriptan intervention prevented IS from in- Burstein and Jakubowski: Sumatriptan and Sensitization 31 Fig 4. Response threshold of central trigeminovascular neurons to dural indentation. (A) Response threshold to dural indentation of 2 neurons sampled at 1-hour intervals before (baseline/ green) and after (red) application of IS, and again after a delayed administration of sumatriptan (blue). Notice that response thresholds plummeted after IS application and recovered completely in one case and only partially in the other after sumatriptan administration. (B) Response threshold to dural indentation of two neurons sampled at 1-hour intervals before (baseline/green) and after (purple) simultaneous application of IS and sumatriptan. Notice that the response threshold remained unchanged in one case and increased in the other. (C, D) Summary of response threshold to dural indentation (mean ⫾ SEM) in late (n ⫽ 9) and early (n ⫽ 9) sumatriptan intervention groups, respectively. ducing such increased neuronal sensitivity to heat (see Fig 6B), and neuronal response threshold remained unchanged ( p ⬎ 0.07; see Fig 6D). Discussion This study demonstrates that the effects of sumatriptan on spinal trigeminovascular neurons depends on whether the drug is given before or after the establishment of central sensitization and cutaneous allodynia. Late sumatriptan intervention effectively reversed measures of central sensitization that depend on peripheral input from the meninges (ie, enlarged dural receptive field and increased sensitivity to dural indentation), but not those that reflect intrinsic activity (ie, increased spontaneous firing and increased sensitivity to periorbital skin stimulation). These results suggest that sumatriptan acts peripherally to block transmission of pain signals from the dura but does not act directly on the central neurons in spinal trigeminal nucleus. Such peripheral blockade of signal transmission from the dura to the central trigeminovascular neurons is also compatible with our finding that early sumatriptan intervention effectively prevented the induction of all of the above measures of central sensitization. 32 Annals of Neurology Vol 55 No 1 January 2004 Effects of Sumatriptan on Sensitization in Central Trigeminovascular Neurons This is the first study to our knowledge to test the effects of any triptan on long-lasting sensitization of central trigeminovascular neurons. Critical to meaningful interpretation of sumatriptan action was the integration of intracranial and extracranial sensitivities in the same central trigeminovascular neuron. When given after the establishment of central sensitization (late intervention), sumatriptan effectively reversed increased intracranial sensitivity of central trigeminovascular neurons (ie, enlarged dural receptive field; increased sensitivity to dural indentation). On face value, this could have been mistaken as evidence that sumatriptan directly counteracts ongoing sensitization of central trigeminovascular neurons, perhaps through postsynaptic inhibition. Conversely, late treatment did not reverse the increased spontaneous firing of the same sensitized central trigeminovascular neurons, nor did it reverse their increased extracranial sensitivity (ie, increased sensitivity to mechanical and thermal stimulation of the periorbital skin), suggesting that sumatriptan cannot counteract ongoing central sensitization. Considering that central trigeminovascular neurons process afferent signals from intracranial and Fig 5. Response magnitude of central trigeminovascular neurons to mechanical stimulation of the periorbital skin. (A) Responses induced by brush and pressure in a neuron studied at 1-hour intervals before (baseline/green) and after (red) application of IS, and again after a delayed administration of sumatriptan (blue). Notice that response magnitudes increased after IS application and remained elevated after delayed sumatriptan administration. (B) Responses induced by brush and pressure in a neuron studied at 1-hour intervals before (baseline/green) and after (purple) simultaneous application of IS and sumatriptan. Notice that response magnitudes did not increase from baseline. (C, D) Summary of response magnitudes (mean ⫾ SEM) to brush (top) and pressure (bottom) in late (n ⫽ 9) and early (n ⫽ 9) sumatriptan intervention groups, respectively. Gray bars in C and D indicate spontaneous activity in the 10-second interval that preceded the stimulus. extracranial organs simultaneously, it is far more reasonable to conclude that sumatriptan acts peripherally to block information flow from the dura to central trigeminovascular neurons, rather than inhibiting the latter directly. When sumatriptan was administered at the same time as IS was applied onto the dura (early interven- tion), it invariably prevented the induction of every aspect of central sensitization that we measured. Thus, central trigeminovascular neurons showed no expansion of dural receptive field, no increase in sensitivity to dural indentation, no long-term increase in spontaneous firing, and no increase in response magnitude to mechanical and thermal skin stimulation. On face Burstein and Jakubowski: Sumatriptan and Sensitization 33 Fig 6. Response thresholds and magnitudes of central trigeminovascular neurons to thermal stimulation of the periorbital skin. (A) Responses induced by raising skin temperature from 35 to 50°C in a neuron studied at 1-hour intervals before (baseline/green) and after (red) application of IS, and again after a delayed administration of sumatriptan (blue). Gray area (also in B) marks the interval between the onset of heating and the response threshold at baseline (green panel). Notice that response thresholds decreased from 46 to 42°C after IS application and remained at 42°C after delayed sumatriptan administration. Notice also that response magnitudes increased after IS application and remained elevated after delayed sumatriptan administration. (B) Responses induced by increasing skin temperature from 35 to 50°C in a neuron studied at 1-hour intervals before (baseline/green) and after (purple) simultaneous application of IS and sumatriptan. Notice that response thresholds and magnitudes did not change from baseline. (C, D) Summary of response magnitudes (mean ⫾ SEM) to heat in late (n ⫽ 5) and early (n ⫽ 8) sumatriptan intervention groups, respectively. value, these findings could be mistaken as evidence that sumatriptan acts directly on trigeminovascular neurons in the dorsal horn. Viewed together with the late intervention results, however, the more likely interpretation is that early sumatriptan intervention averted the initiation of central sensitization by blocking noxious signals generated in sensitized meningeal nociceptors from reaching central trigeminovascular neurons in the dorsal horn. Based on the integration of the findings above, we propose that the sumatriptan reduces the transmission of sensory signals from the dura to the dorsal horn by acting directly on meningeal nociceptors. Such action may be exerted on the nociceptor’s peripheral branch (in the dura), or its cell soma (in the trigeminal ganglion), or its 34 Annals of Neurology Vol 55 No 1 January 2004 central branch (in the spinal cord). We also propose that sumatriptan cannot generate enough inhibition to significantly counteract inveterate sensitization in central trigeminovascular neurons, presumably because these neurons do not express the 5HT1B/1D receptor or, alternatively, the receptor may be present but its activation falls short of generating functional inhibition when the neurons are sensitized. Finally, we rule out the possibility that the failure of sumatriptan to counteract inveterate central sensitization was caused by poor penetrability through the blood–brain barrier, because triptans with much higher lipophylicity than sumatriptan are as equally as ineffective in terminating migraine pain in the presence of cutaneous allodynia as we report in the accompanying clinical study.8 Effects of Triptans on Nonsensitized Central Trigeminovascular Neurons Unlike this study, which tested the effects sumatriptan on neurons undergoing a long-lasting state of sensitization that simulates migraine-like pathophysiology, other studies have examined the effects of triptans on naive, nonsensitized central trigeminovascular neurons. In those studies, intravenous administration of triptans such as sumatriptan,22 zolmitriptan,23 naratriptan,24 or rizatriptan25,26 reduced evoked potential magnitude and firing probability of central trigeminovascular neurons in response to electrical stimulation of meningeal blood vessels. Collectively, those studies have led the authors to propose that the therapeutic actions of triptans in alleviating migraine headache are caused by direct inhibition of central23,27 and/or peripheral25,28 trigeminovascular neurons. Note that these studies reflect the physiological properties of trigeminovascular neurons only under normal conditions, rather than the pathophysiological changes they undergo during migraine attack. Cumberbatch and colleagues28 showed that the triptan L-741-604 interfered with calcitonin gene–related peptide–induced amplification of responses of central trigeminovascular neurons to vibrissal air puffs and concluded that this triptan can inhibit central sensitization. However, the increased neuronal responses to vibrissal stimulation lasted a mere 6 minutes, a period of time too short for the development of central sensitization, which should run for hours after termination of the triggering stimulus, independent of incoming impulses from peripheral trigeminovascular nociceptors.3 Short of fulfilling additional criteria for central sensitization (ie, expansion of dural and facial receptive fields, decreased threshold of mechanical and thermal skin stimulation, long-lasting increase in spontaneous activity), Cumberbatch and colleagues appear to have studied nonsensitized trigeminovascular neurons, and their results have no bearing to triptan action on central sensitization. Physiological and Clinical Correlates and Conclusions We propose that the responses of individual trigeminovascular neurons in our animal model of intracranial pain and cutaneous allodynia correspond to the neurological symptoms of migraine attacks in patients in three ways. Correlate 1: expansion of dural receptive fields and increased neuronal sensitivity to dural indentation correspond to the throbbing and exacerbation of pain by bending over. Correlate 2: increased spontaneous activity corresponds to the pain intensity. Correlate 3: increased neuronal sensitivity to brushing and heating of the rat periorbital skin corresponds to mechanical and thermal allodynia in the patient periorbital skin of the referred pain area. Based on these correlates, we propose the following interpretations for the neuronal action of sumatriptan. The peripheral action of sumatriptan that blocks transmission of pain signals from the dura to the central trigeminovascular neurons can explain our observations in migraine patients8 that the drug is highly effective, whether given early or late into the attack, in terminating throbbing and eliminating the exacerbation of pain upon bending over (correlate 1). This peripheral action also explains the results of our clinical study8 that early sumatriptan intervention is effective in terminating migraine pain (correlate 2) and blocking the development of cutaneous allodynia (correlate 3), because it disrupts the input necessary for the development of central sensitization. The evidence that the peripheral action of sumatriptan cannot counteract the inveterate, ongoing sensitization in the central trigeminovascular neurons can explain the failure of late sumatriptan intervention to terminate migraine pain (correlate 2) or reverse the exaggerated skin sensitivity (correlate 3) in attacks already associated with allodynia.8 That the benefits of triptan therapy is limited to the early stages of a migraine attack among allodynic patients8 is consistent with the evidence that triptans can prevent the initiation of central sensitization which depends on incoming impulses from meningeal nociceptors but cannot abort ongoing central sensitization once it became self-sustained. This limitation of triptans calls for new migraine drugs designed to reverse the hyperexcitable state of central trigeminovascular neurons in the later stages of migraine by bringing their membrane potential back to its normal resting value. This work was supported by a grant from GlaxoSmithKline (48573, R.B.) and by the National Institutes of Health (NIDCR, DE13347, R.B., and NINDS, NS35611, R.B.). References 1. Burstein R. Deconstructing migraine headache into peripheral and central sensitization. Pain 2001;89:107–110. 2. Strassman AM, Raymond SA, Burstein R. Sensitization of meningeal sensory neurons and the origin of headaches. Nature 1996;384:560 –564. 3. Burstein R, Yamamura H, Malick A, Strassman AM. Chemical stimulation of the intracranial dura induces enhanced responses to facial stimulation in brain stem trigeminal neurons. J Neurophysiol 1998;79:964 –982. 4. Yamamura H, Malick A, Chamberlin NL, Burstein R. Cardiovascular and neuronal responses to head stimulation reflect central sensitization and cutaneous allodynia in a rat model of migraine. J Neurophysiol 1999;81:479 – 493. 5. Burstein R, Yarnitsky D, Goor-Aryeh I, et al. An association between migraine and cutaneous allodynia. Ann Neurol 2000; 47:614 – 624. 6. Burstein R, Cutrer FM, Yarnitsky D. The development of cutaneous allodynia during a migraine attack: clinical evidence for the sequential recruitment of spinal and supraspinal nociceptive neurons in migraine. Brain 2000;123:1703–1709. Burstein and Jakubowski: Sumatriptan and Sensitization 35 7. Levy D, Strassman AM. Distinct sensitizing effects of the cAMP-PKA second messenger cascade on rat dural mechanonociceptors. J Physiol 2002;538:483– 493. 8. Burstein R, Collins B, Jakubowski M. Defeating migraine pain with triptans: a race against the development of cutaneous allodynia. Ann Neurol 2003; DOI: 10.1002/ana.10785. 9. Steen KH, Reeh PW, Anton F, Handwerker HO. Protons selectively induce lasting excitation and sensitization to mechanical stimulation of nociceptors in rat skin, in vitro. J Neurosci 1992;12:86 –95. 10. Steen KH, Steen AE, Reeh PW. A dominant role of acid pH in inflammatory excitation and sensitization of nociceptors in rat skin, in vitro. J Neurosci 1995;15:3982–3989. 11. Saxena PR, Tfelt-Hansen P. Triptans, 5-HT1b/1d receptors agonists in the acute treatment of migraine. In: Olesen J, TfeltHansen P, Welch K, eds. The headaches. Vol 2. Philadelphia: Lippincott Williams & Wilkins, 2000:411– 438. 12. Friberg L, Olesen J, Iversen HK, Sperling B. Migraine pain associated with middle cerebral artery dilatation: reversal by sumatriptan. Lancet 1991;338:13–17. 13. Humphrey PP, Feniuk W. Mode of action of the anti-migraine drug sumatriptan. Trends Pharmacol Sci 1991;12:444 – 446. 14. Moskowitz MA. Neurogenic versus vascular mechanisms of sumatriptan and ergot alkaloids in migraine. Trends Pharmacol Sci 1992;13:307–311. 15. Buzzi MG, Moskowitz MA. The antimigraine drug, sumatriptan (GR43175), selectively blocks neurogenic plasma extravasation from blood vessels in dura mater. Br J Pharmacol 1990;99:202–206. 16. Shepherd SL, Williamson DJ, Beer MS, et al. Differential effects of 5-HT1B/1D receptor agonists on neurogenic dural plasma extravasation and vasodilation in anaesthetized rats. Neuropharmacology 1997;36:525–533. 17. Messlinger K, Hotta H, Pawlak M, Schmidt RF. Effects of the 5-HT1 receptor agonists, sumatriptan and CP 93,129, on dural arterial flow in the rat. Eur J Pharmacol 1997;332: 173–181. 36 Annals of Neurology Vol 55 No 1 January 2004 18. Buzzi MG, Carter WB, Shimizu T, et al. Dihydroergotamine and sumatriptan attenuate levels of CGRP in plasma in rat superior sagittal sinus during electrical stimulation of the trigeminal ganglion. Neuropharmacology 1991;30:1193–1200. 19. Goadsby PJ, Edvinsson L. The trigeminovascular system and migraine: studies characterizing cerebrovascular and neuropeptide changes seen in humans and cats. Ann Neurol 1993;33:48 –56. 20. Messlinger K, Ebersberger A, Schaible HG. Release of immunoreactive substance P in the brain stem upon stimulation of the cranial dura mater with low pH—inhibition by the serotonin (5-HT1) receptor agonist CP 93,129. Br J Pharmacol 1998;125:1726 –1732. 21. Siegel S. Nonparametric statistics for the behavioral sciences. Tokyo: McGraw-Hill Kogakusha, 1956. 22. Kaube H, Hoskin KL, Goadsby PJ. Inhibition by sumatriptan of central trigeminal neurones only after blood-brain barrier disruption. Br J Pharmacol 1993;109:788 –792. 23. Goadsby PJ, Hoskin KL. Inhibition of trigeminal neurons by intravenous administration of the serotonin (5HT)1B/D receptor agonist zolmitriptan (311C90): are brain stem sites therapeutic target in migraine? Pain 1996;67:355–359. 24. Goadsby PJ, Knight Y. Inhibition of trigeminal neurones after intravenous administration of naratriptan through an action at 5-hydroxy-tryptamine (5-HT(1B/1D)) receptors. Br J Pharmacol 1997;122:918 –922. 25. Cumberbatch MJ, Hill RG, Hargreaves RJ. Rizatriptan has central antinociceptive effects against durally evoked responses. Eur J Pharmacol 1997;328:37– 40. 26. Cumberbatch MJ, Hill RG, Hargreaves RJ. The effects of 5-HT1A, 5-HT1B and 5-HT1D receptor agonists on trigeminal nociceptive neurotransmission in anaesthetized rats. Eur J Pharmacol 1998;362:43– 46. 27. Goadsby PJ, Akerman S, Storer RJ. Evidence for postjunctional serotonin (5-HT1) receptors in the trigeminocervical complex. Ann Neurol 2001;50:804 – 807. 28. Cumberbatch MJ, Williamson DJ, Mason GS, et al. Dural vasodilation causes a sensitization of rat caudal trigeminal neurones in vivo that is blocked by a 5-HT1B/1D agonist. Br J Pharmacol 1999;126:1478 –1486.