Brain (1998), 121, 1231-1237
`
`Retinal plasma extravasation in animals but not in
`humans: implications for the pathophysiology of
`mtgratne
`
`A. May, 1•3 S. L. Shepheard,2 M. Knorr, 3 R. Effert,4 A. Wessing,4 R. J. Hargreaves,2 P. J. Goadsbyl and
`H. C. Diener3
`
`1/nstitute of Neurology, London, 2Department of
`Pharmacology, Merck, Sharp and Dohme Research
`Laboratories, Neuroscience Research Centre, Harlow, UK
`and Departments of 3 Neurology and 40phthalmology,
`University of Essen, Germany
`
`Correspondence to: Arne May, Institute of Neurology,
`The National Hospital for Neurology and Neurosurgery,
`Queen Square, London WCJN 3BG, UK.
`E-mail: amay@ion.ucl.ac.uk
`
`Summary
`High-intensity electrical stimulation of the trigeminal
`ganglion is accompanied by mast cell degranulation,
`vasodilatation, increased endothelial permeability and
`leakage of albumin from postcapillary venules within the
`dura mater. Overall, the histological appearance suggests
`an
`evolving
`sterile
`inflammatory
`response. This
`neurogenic inflammation within the meninges has been
`suggested as a model to explain the pain in migraine and
`cluster headache, and has been used to characterize the
`pharmacology of anti-migraine compounds. Using the rat
`model of neurogenic
`inflammation,
`the
`albumin
`extravasation ratio (stimulated : unstimulated side) in
`vehicle-treated animals in the dura and retina was 1.60 ±
`0.11 and 1.76 ± 0.18, respectively (n = 10; values are
`mean ± SEM). Pretreatment with sumatriptan (n = 9)
`produced a highly significant reduction in the ratio of
`
`extravasation within the dura to 1.10 ± 0.06 (P = 0.002)
`and in the retina to 0.96 ± 0.06 (P ± 0.001), as did the
`neurokinin-! receptor antagonist RP 67580 (n = 12) in
`the dura (1.04 ± 0.11, P = 0.002) and retina (1.08
`± 0.06, P = 0.001). These data demonstrate increased
`endothelial permeability and leakage of albumin not only
`in the dura but also in the retina. In a second stage we
`investigated possible extravasation in the human retina
`in acute migraine (n = 8) and cluster headache (n = 5)
`using fluorescein or indocyanine angiography. No in(cid:173)
`creased endothelial permeability or leakage of dye could
`be found in the human retinal or choroidal vessels during
`headache attacks or in the headache-free interval in
`persons suffering from both migraine and cluster head(cid:173)
`ache. These data raise the possibility that neurogenic
`inflammation is not a major factor in headache attacks
`in migraine or cluster headache.
`
`Keywords: migraine attack; neurogenic inflammation; trigeminovascular system; plasma extravasation; fluorescein
`angiography
`
`Introduction
`The underlying pathophysiology of migraine remains unclear.
`Scientific investigation is difficult because migraine can only
`be studied directly in humans. Traditionally, migraine has
`been linked to pathological changes in the diameter of cranial
`blood vessels. As early as 1931, Rodella reported widening
`of the retinal vessels, which he observed by examining the
`fundi of his patients during the headache. Three years later
`Critchley (1934) also noted engorgement and tenderness of
`the temporal vessels during headache, with congestion of the
`conjunctival and retinal vessels. He concluded that these
`findings supported a vascular origin of migraine, although
`later in life Critchley no longer held that view (personal
`communication to P.J.G.). Wolff (1963) suggested that
`
`© Oxford University Press 1998
`
`diameter changes in extracranial and most likely intracranial
`arteries were the cause of headache. This hypothesis seemed
`to explain the therapeutic benefit of substances with vaso(cid:173)
`constrictor effects, such as ergotamine, dihydroergotamine
`and sumatriptan, although other substances with even
`stronger vasoconstrictor activity, such as angiotensin and
`noradrenaline, had no effect. The suggestion that the dura
`mater and its small vessels is an important source of headache
`pain,
`indeed
`resulting
`in pain when electrically or
`mechanically stimulated (Cushing, 1908; Ray and Wolff,
`1940), led to the hypothesis that additional mechanisms may
`be involved (Strassmann eta!., 1986; Moskowitz, 1993).
`In 1987 Moskowitz and her co-authors (Markowitz eta!.,
`
`EX2269
`Eli Lilly & Co. v. Teva Pharms. lnt'l GMBH
`IPR2018-01422, -01423, 01424, -01425,
`-01426, -01427
`
`

`

`in four cluster patients was triggered using inhalation of
`nitroglycerine during their bout; one patient suffered from a
`spontaneous attack. The headache attack in the eight migraine
`patients was untreated and spontaneous. Excluded were
`patients who received prophylactic
`treatment or other
`vasoactive drugs, patients with drug or analgesic abuse and
`patients who suffered from chronic tension-type headache or
`any other known illness other than migraine. All patients
`were studied within 6 h of the onset of untreated headache
`symptoms.
`Each patient had two examinations: (i) during the acute
`attack and (ii) during the headache-free interval 3 days to 4
`months later.
`Each patient's visual acuity was tested, then examination
`by ophthalmoscopy and slit lamp was performed; the patient
`was then subjected to fluorescein or indocyanine angiography.
`In all patients the eye on the side ipsilateral to the headache
`side was examined. A 5 ml intravenous injection (right
`brachial vein) of 10% sodium fluorescein (four patients) or
`25 mg indocyanine green dye, dissolved in 5 ml aqueous
`solvent (10 patients) (Table 1) was immediately followed by
`angiography (Flower, 1973; Wessing, 1975; Bischoff and
`Flower, 1985; Nielsen, 1985; Panzardi et al., 1992; Lim and
`Flower, 1995), using a modified fundus camera and a TV
`system for image reception. Angiographies were recorded on
`videotape and images were simultaneously digitized and
`saved on disc. Two migraine patients received both treatments.
`The late pictures were taken I 0-12 min after this procedure.
`The complete procedure was repeated in the headache-free
`interval. The immediate angiogram of the headache attack
`was then compared with the late pictures to detect vascular
`leakage, as well as with the early and late angiograms of the
`headache-free interval. The person reporting the angiograms
`was different from the person involved in the actual study
`and was blinded to the clinical data.
`
`Pathophysiology ofmigraine
`
`1233
`
`0 Control (n = 1 0)
`• Sumatriptan (1 mg kg-1, n = 9)
`~ RP67580 (1 mg kg-1, n = 12)
`
`2.5
`
`2,0
`
`0
`-~
`<::
`0
`'a 1.5
`
`~ x
`
`UJ
`
`Dura
`
`Retina
`
`0.6
`
`Fig. 1 Extravasation ratio in the retina and dura following
`unilateral electric stimulation of the trigeminal ganglion (white
`bars) and after pretreatment with sumatriptan (black bars) or
`RP67580 (hatched bars).
`
`Table 2 Percentage change of vital parameters from
`baseline (before unilateral ganglion stimulation) in
`experimental rats (n = 3)
`
`During
`stimulation
`
`I min after
`stimulation
`
`5 min after
`stimulation
`
`pC02
`p02
`pH
`Blood pressure
`Heart rate
`
`-4.92%
`31.93%
`0.38%
`44.49%
`18.29%
`
`-1.37%
`15.12%
`--0.22%
`16.80%
`10.75%
`
`0.13%
`18.70%
`--0.28%
`11.92%
`7.38%
`
`In three additional animals without pretreatment, the
`arterial blood pressure and heart rate showed a moderate
`increase during ganglion stimulation. The arterial p02 also
`increased whereas pC02 and pH moderately decreased,
`indicating mild hyperventilation (Table 2).
`
`Results
`Animal study
`Following electrical stimulation of the trigeminal ganglion
`in vehicle-treated animals, the extravasation ratio in dura and
`retina was 1.60 ::!:: 0.11 and 1.76::!:: 0.18 respectively (n =
`10; values are mean::!:: SEM). Pretreatment with sumatriptan
`(n = 9) produced a highly significant reduction in the ratio
`of extravasation within the dura to 1.10 ::!:: 0.06 (P = 0.002)
`and in the retina to 0.96 ::!:: 0.06 (P = 0.001), as did RP
`67580 (n = 12) in the dura (1.04::!:: 0.11, P = 0.002) and
`retina (1.08 ::!:: 0.06, P = 0.001) (Fig. 1). The extravasation
`ratio in the facial tissues in the control group was 4.00 ::!::
`0.51 (eyelid), 3.57 ::!:: 0.79 (lip) and 1.34::!:: 0.26 (conjunctiva).
`Sumatriptan had no significant effects on the extravasation
`ratio in the facial tissues but pretreatment with RP 67580
`diminished the ratio in the eyelid to 1.71 ::!:: 0.21 (P = 0.001),
`in the lip to 1.34 ::!:: 0.15 (P = 0.006) and in the conjunctiva
`to 1.24 ::!:: 0.08 (not significant).
`
`Human study
`No significant changes in retina background and especially
`no leakage of plasma extravasation markers (i.e. fluorescein
`or indocyanine green) from the retinal or choroidal vessels
`were seen in either the acute migraine attack or the headache(cid:173)
`free interval. This held true immediately after injection of
`the dye as well as in the late pictures taken I 0-12 min after
`injection of the bolus (Fig. 2A-D] All patients had a normal
`retinal angiogram. In summary, no increased endothelial
`permeability or leakage of dye could be detected in any of
`the 13 patients.
`
`Discussion
`In 1987 Moskowitz and co-authors introduced an animal
`model of neurogenic inflammation in the dura mater where
`plasma protein extravasation, vasodilatation,
`increased
`endothelial permeability and mast cell degranulation were
`
`

`

`effective in blocking dural plasma extravasation in this
`animal model. In humans, however, we failed to show any
`extravasation in the acute headache attacks of migraine and
`cluster headache patients.
`The methods used in the animal and human studies are
`compatible in that indocyanine green dye binds to albumin
`(Panzardi eta/., 1992; Lim and Flower, 1995) and fluorescein
`binds incompletely to albumin and to a considerable extent
`to even smaller proteins (Hodge and Dollery, 1964; Wessing,
`1968), and should be extravasated
`into
`the
`tissue if
`extravasation takes place in the human retina during the
`acute migraine or cluster attack (Lange and Boyd, 1944;
`Panzardi et a/., 1992). If the model of neurogenic
`inflammation is transferable to humans, proteins at least as
`large as albumin (which is the marker for extravasation in
`this model) should be extravasated. We cannot exclude the
`possibility of a different physiology or trigeminal innervation
`and sensitivity of the dura and retina in humans compared
`with animals. Several aspects of the model of trigeminal(cid:173)
`induced plasma extravasation clarity its limitations with
`regard
`to migraine. Serotonin does not block plasma
`extravasation following
`trigeminal ganglion stimulation,
`although it is effective in humans (Kimball et a/., 1960;
`Lance et al., 1967). This issue is confused by the potentially
`proinflammatory effects of serotonin at the 5-HT2A receptor
`(Verheyen et a/., 1987; Bryant et a/., 1996). Neurologists
`would not generally consider indomethacin or valproate as
`first-line abortive agents in migraine yet they are highly
`effective in neurogenic inflammation. Furthermore, avitriptan,
`another 5-HT18110 agonist that is clearly an effective anti(cid:173)
`migraine agent (Couch et al., 1996) of similar efficacy to
`sumatriptan, is much less potent than sumatriptan in the
`neurogenic plasma extravasation model (Yocca eta/., 1997).
`Moreover, plasma extravasation
`following
`trigeminal
`ganglion stimulation can be blocked by pretreatment with
`sumatriptan (Buzzi et a/., 1991 b), while pretreatment of
`headache in patients using sumatriptan during the aura does
`not block headache (Bates, 1993). Consistent with these
`observations, and with our new data, plasma extravasation
`has not been observed in humans during migraine (Nissila
`et a/., 1996).
`Although our data indicate that the vital parameters, such
`as blood pressure, heart rate and blood gases, show only a
`moderate change during stimulation, it is important to
`consider that lowering a bipolar electrode directly through
`the brain and then stimulating the trigeminal ganglion with
`high current is non-physiological and may well produce a
`whole range of changes within the trigeminovascular system.
`Possible factors other than neurogenic inflammation involved
`in human migraine attacks are hyperexcitability of the cortex
`(Welch et a/., 1993), disturbances of the blood-brain barrier
`(Kaube eta/., 1993), abnormal vascular reactivity and release
`of neurotransmitters, such as calcitonin gene-related peptide
`and substance P (Goadsby and Edvinsson, 1993). The source
`of the migraine headache is still not clear. Vascular dilatation
`alone is unlikely to explain it. Activation oftrigeminovascular
`
`Pathophysiology of migraine
`
`I 235
`
`neurons causes the release of neuropeptides. The 5-HTI
`receptors on these trigeminovascular nerves may modulate
`the central transmission of nociceptive inputs (Feniuk et al. ,
`1979; Connor and O'Shaughenessy, 1993; Kaube et al., 1993;
`Hoskin et a/., 1996). The trigeminovascular system appears
`to play a pivotal role as the substrate for craniovascular
`nociception (Goadsby and Edvinsson, 1993, 1994), although
`what excites the perivascular nociceptors in the first place
`remains to be established.
`In conclusion, these new data, taken together with the facts
`that certain substances are potent inhibitors of neurogenic
`extravasation but ineffective in human migraine and that
`substances effective in migraine are relatively ineffective in
`the model of neurogenic extravasation, indicate that the
`model of inflammatory neurogenic activation of C-fibres
`might not be an appropriate model to elucidate the clinical
`efficacy of the entire range of novel substances in migraine.
`The model as it is currently employed seems to select
`substances that are active against extravasation but may not
`necessarily prove to have anti-migraine activity. Our data
`raise the possibility that neurogenic inflammation is not
`present or, if present, forms a role that is not sufficient by
`itself to produce pain in migraine or cluster headache since
`its blockade does not necessarily relieve headache. The data
`support the view that other fundamental processes, most
`likely in the central nervous system, are necessary and key
`to the pathophysiology of an acute migraine attack.
`
`Acknowledgements
`The authors with to thank Ms Debora Cook and Mr David
`Williamson for
`technical assistance, and Dr Albrecht
`Lommatzch. This paper was presented in preliminary form
`at the 8th Congress of the International Headache Society
`in Amsterdam, June 1997 (May et a!. Cephalalgia; 1997;
`17: 240).
`
`References
`Bates D. Treatment with subcutaneous sumatriptan during the
`migraine aura [abstract]. Cephalalgia 1993; 13 Suppl 13: 188.
`
`Bischoff PM, Flower RW. Ten years experience with choroidal
`angiography using indocyanine green dye: a new routine examination
`or an epilogue? Doc Ophthalmol 1985; 60: 235-91.
`
`Briindli P, Loffier BM, Breum V, Osterwalder R, Maire JP, Clozel
`M. Role of endothelin in mediating neurogenic plasma extravasation
`in rat dura mater. Pain 1996; 64: 315-22.
`
`Bryant HU, Nelson DL, Button D, Cole HW, Baez MB, Lucaites
`VL, et al. A novel class of 5-HT2A receptor antagonists: aryl
`aminoguanidines. Life Sci 1996; 59: 1259-68.
`
`Buzzi MG, Moskowitz MA. The antimigraine drug, sumatriptan
`(GR 43175), selectively blocks neurogenic plasma extravasation
`from blood vessels in dura mater. Br J Pharmacoll990; 99: 202-6.
`
`

`

`Moskowitz MA. Neurogenic versus vascular mechanisms of
`sumatriptan and ergot alkaloids in migraine. [Review]. Trends
`Pharmacol Sci 1992; 13: 307-11.
`
`Moskowitz MA. Neurogenic inflammation in the pathophysiology
`and treatment of migraine. [Review]. Neurology 1993; 43 (6 Suppl
`3): Sl6-20.
`
`Nielsen NV. The normal fundus fluorescein angiogram. V.
`Intraobserver-, interobserver- and interocular variation of the fundus
`fluorescein angiogram. Acta Ophthalmol (Copenh) 1985; 63: 463--6.
`
`Nissila M, Parkkola R, Sonninen P, Salonen R. Intracerebral arteries
`and gadolinium enhancement in migraine without aura [abstract].
`Cephalalgia 1996; 16: 363.
`
`Panzardi G, Donati MC, Longobardi G, Poggi P. Choroidal
`angiography with
`indocyanine green dye:
`absorption and
`fluorescence techniques. Eur J Ophthalmol 1992; 2: 83-5.
`
`Ray BS, Wolff HG. Experimental studies on headache: pain(cid:173)
`sensitive structures of the head and their significance in headache.
`Archs Surg 1940; 41: 813-56.
`
`Rodel\a A. Migraine and cutaneus vessels. [German]. Schweiz Med
`Wschr 1931; 61: 1256.
`
`Roon K, Diener HC, Ellis P, Hettiarachchi J, Poole P, Christiansen
`I, et a!. CP-122,288 blocks neurogenic inflammation, but is not
`effective in aborting migraine attacks: results of two controlled
`clinical trials [abstract]. Cephalalgia 1997; 17: 245.
`
`Saito K, Markowitz S, Moskowitz MA. Ergot alkaloids block
`neurogenic extravasation in dura mater: proposed action in vascular
`headaches. Ann Neurol 1988; 24: 732-7.
`
`Shepheard SL, Williamson DJ, Hill RG, Hargreaves RJ. The
`non-peptide neurokinin 1 receptor antagonist, RP 67580, blocks
`neurogenic plasma extravasation in the dura mater of rats. Br J
`Pharmacol 1993; 108: 11-2.
`
`Pathophysiology of migraine
`
`123 7
`
`Shepheard SL, Williamson DJ, Williams J, Hill RG, Hargreaves
`RJ. Comparison of the effects ofsumatriptan and the NKl antagonist
`CP-99,994 on plasma extravasation in dura mater and c-fos mRNA
`expression
`in
`trigeminal
`nucleus
`caudalis
`of
`rats.
`Neuropharmacology 1995; 34: 255-61.
`
`Strassmann A, Mason P, Moskowitz M, Maciewicz R. Response of
`brainstem trigeminal neurons to electrical stimulation of the dura.
`Brain Res 1986; 379: 242-50.
`
`Verheyen A, Vlaminckx E, Lauwers F, Borgers M, Wouters L.
`Radioactive colloidal gold as a tool to quantify extravasation of
`macromolecules in the rat cremaster muscle. Int J Microcirc Clin
`Exp 1987; 5: 321-33.
`
`Welch KM, Barkley GL, Tepley N, Ramadan NM. Central
`neurogenic mechanisms of migraine. [Review]. Neurology 1993;
`43 (6 Suppl 3): S21-5.
`
`Wessing A. Fluoreszenzangiographie der Retina. Lehrbuch und
`Atlas. Stuttgart: Georg Thieme Verlag; 1968.
`
`Wessing A. Principles of diagnostic progress through fluorescence
`angiography. [German]. Ber Zusammenkunft Dtsch Ophthalmol Ges
`1975; 73: 566-8.
`
`Williamson DJ, Shepheard SL, Hill RG, Hargreaves RJ. The novel
`anti-migraine agent rizatriptan inhibits neurogenic dural vasodilation
`and extravasation. Eur J Pharmacol 1997; 328: 61-4.
`
`Wolff HG, editor. Headache and other head pain. 2nd ed. New
`York: Oxford University Press; 1963.
`
`Yocca FD, Gylys JA, Smith DW, Ruediger E, Yevich JP, Mahle C,
`et a!. BMS-181885: a clinically effective migraine abortive with
`peripherovascular and neuronal 5HT ID antagonist properties
`[abstract]. Cephalalgia 1997; 17: 404.
`
`Received October 31, 1997. Revised January 13, 1998.
`Second revision February 13, 1998. Accepted February 16, 1998.
`
`

`

`
`
`