Kruit, Mark Christian
Citation
Kruit, M. C. (2010, January 20). Migraine and brain lesions. Data from the population-based CAMERA Study. Department of Radiology, Faculty of Medicine, Leiden University Medical Center (LUMC), Leiden University.
Retrieved from https://hdl.handle.net/1887/14585
Version: Not Applicable (or Unknown) License:
Downloaded from: https://hdl.handle.net/1887/14585
Note: To cite this publication please use the final published version (if applicable).
C H A P T E R 1
GENERALINTRODUCTIONANDOUTLINE
MIGRAINESYMPTOMATOLOGY
Migraine is a common, multifactorial
neurovascular brain disorder, typically
characterized by recurrent attacks of
throbbing or pulsating unilateral head
ache. Migraine headaches are accompa
nied bysymptoms of autonomicnervous
system dysfunction, including nausea,
vomiting, and/or sensitivity to light,
sound, and/or movement (migraine
without aura; MO). Untreated attacks
typically last 472 hours; the median
durationofanattackis24hours.Acom
bination of features is required for the
diagnosis,butnotallfeaturesarepresent
ineveryattackorineverypatient.24,* Uptoonethirdofmigrainepatients
also have transient focal neurological
aura symptoms, that precede or accom
pany some or all of their headache at
tacks (migraine with aura; MA).3;5 Aura
* For diagnostic criteria for the common forms of
migraine from The International Classification of
HeadacheDisorders,seeAppendixA.
symptoms nearly always include visual
disturbances (99% of patients; blurred
vision, black dots, scintillating scotoma),
togetherwithsensory(31%;paresthesia)
or speechrelated (18%; dysarthria,
aphasia) symptoms; rarely (6%), motor
symptomsarepartoftheaura(onesided
weakness). The transient symptoms
occur usually at alternating body sides,
and typically progress over minutes and
last up to 60 minutes. Different symp
toms mostly succeed one another, and
together may last longer than 60 min
utes.6
AccordingtothecriteriafromtheIn
ternational Headache Society (IHS),
migraine patients are defined as indi
viduals who in their lifetime have had at
leasttwoattackswithauraoratleastfive
attacks without aura.2;3 However, over
50% of patients have never consulted a
physician for their migraine headaches,
and–depending on headache severity
and frequency– many ‘patients’ may be
unaware of their status as ‘migraine
patient’.
EPIDEMIOLOGY
Migraine is one of the most prevalent
neurological diseases in adults. In the
Netherlands,thelifetimeprevalencewas
estimated at 33% in females and 13% in
males;the1yearprevalencewas25%in
females, and 7% in males.7 In Denmark,
lifetime prevalences were 8% for males
and 25% for females; 1year prevalences
were 6% and 16% respectively.8 In the
USA, these figures are similar: 6% and
17%.9 The prevalence of migraine varies
considerablybyageandishighestinboth
menandwomenbetweentheagesof35
to 45 years. These high prevalence fig
ures represent the size of the public
healthproblemduetomigraine.
Headache attack frequency varies
widely; median attack frequencies have
beenestimatedat11.5permonth,with
25% of sufferers having 2 attacks per
month, and 10% having 1attack every
week.7;10 The otherside ofthisspectrum
consistsofpatientsexperiencingfewerto
only incidental attacks. Migraine is for
many patients highly disabling, and
almost all people with migraine experi
ence reductions in social activities and
work capacity.11 The World Health Or
ganization has reported that migraine is
inthetop20ofcausesofhealthylifeloss
todisabilityworldwide.12Apatientwitha
severemigraineattackwasconsideredto
beasdisabledasonewithactivepsycho
sis,dementiaortetraplegia.
Besides the burden to the individu
alsandtheirfamilies,migrainealsoleads
to high socioeconomic costs. The direct
costs of migraine include costs of health
care utilization and prescribed and over
thecounter medication, which in total
has been estimated in the United States
at about 2 billion USD in 1999; indirect
costs resulting from loss of productivity
were estimated at 13 billion USD.13 In
Europe, annual direct and indirect costs
togetherwereestimatedin2005at€590
per migraine patient;14 based on this, in
the Netherlands costestimates range
from0.6to1billionEurosperyear.
MIGRAINEPATHOPHYSIOLOGY
Althoughinthelast2decadesbothbasic
science, human physiological investi
gations and (functional) neuroimaging
havediscoveredimportantaspectsofthe
migraine pathophysiology, the complete
picture is still incompletely understood.
Both the recurrent episodic character of
the disorder, as well as the associated
symptoms (sensitivity to light/sound/
movement) and aura symptoms give
clues tothe pathophysiology of migraine
attacks,andthereforeperformingofictal
imaging(e.g.withMRI)isofgreatpoten
tial value. The sudden and unannounced
natureofattackscomplicatesplanningof
preictal and ictal examinations, and a
noisy MRI examination is in many mi
graine patients with nausea and/or
photo and phonofobia not well toler
ated.
Therecurrentandvariablenatureof
migraine suggests that only when a
specific threshold is reduced, ‘normal’
triggerscaninitiate(acascadeof)physio
logical reactions leading to a migraine
attack. Genetic factors appear to be a
major factor in setting this individual
threshold. This is based on the inherited
nature of migraine, on findings in twin
studies, and on the discovery of specific
genechanges in familial hemiplegic
migraine, suggesting that migraine, or at
leasttheaurasymptoms,arecausedbya
channelopathy.15 Which genes are in
volved in ‘common migraine’ has yet to
be elucidated. Other factors, including
internal variables like hormonal fluctua
tions, substance misuse, fatigue, and
externalenvironmentalormeteorological
variables, appear to modulate the indi
vidual threshold further.5;16 This concept
seems to apply both for the ‘pain’ and
also for the ‘aura’ phenomena in mi
graine.
PAINPROCESSING
The net effect of modulation of the set
point probably results in episodic dys
function of brainstem or diencephalic
nucleithatareinvolvedintheprocessing
of craniovascular afferents from the
trigeminovascular system.4 Nociceptive
information received from a neural
plexus surrounding the large cerebral
vessels,thepialvessels,thelargevenous
sinuses and the dura mater converges
through the trigeminal ganglia and the
dorsal cervical roots of C1 and C2, to
wards the ‘trigeminocervical complex’,
which is a functional group of second
order neurons from the trigeminal nu
cleus caudalis and C1/C2 dorsal horns.
From this trigeminovascular complex,
nociceptive input is modulated through
brainstem nuclei and after decussation
projected to third order neurons in con
tralateral thalamic and cortical pain
areas.17 The modulation is probably
mostlythroughthedorsalraphenucleus,
locus coeruleus, and nucleus raphe
magnus.15;18 On brainstem and/or tha
lamic level, a defective pain modulation
seems to be in part responsible for the
increasedpainsensationinmigraineurs.4
Simultaneously, a ‘trigeminalauto
nomicreflex’maybepresentinmigraine.
After activation of the trigeminocervical
complex, this outflowing cranial para
sympathetic reflex is mediated through
the pterygopalatine, otic, and carotid
ganglia, and is suggested to result in
perivascular release of vasoactive and
painproducingneuropeptidesinthedura
mater.Thisprocessisdescribedassterile
neurogenic inflammation, and includes
local release of calcitoningenerelated
peptideandsubstanceP,leadingtomast
celldegranulation,plateletaggregationin
postcapillary venules, and vasodilata
tion. There remain many questions re
garding the mechanism of this ‘reflex’
pathway in migraine, and its relevance
andexistenceisstillunderdebate.15 Painduringamigraineattack,isthus
likely to result from a combination of (a)
direct activation of intracranial pain
receptors, (b) central facilitation or de
fective (descending) modula
tion/inhibition of these signals, and
possibly (c) effects from retrograde
perivascularneuropeptiderelease.
MIGRAINEAURA
Basedongrowingclinicalandexperimen
tal data, it is now accepted that cortical
spreadingdepression(CSD)isthephysio
logicalcorrelateofthemigraineaura.1922 CSD is a slowly propagating wave (2–3
mm/min) consisting of an initial brief
excitation followed by longerlasting
depressed neuronal bioelectrical activity
(depolarisation)ofneuralandglialmem
branes and cells. The depolarisation is
cause of a disruption of membrane ionic
gradients, leading to efflux of cytosolic
potassium and influx of calcium, which
results in release of excitatory amino
acids (neurotransmitters, glutamate)
fromnervecells.Thelocalfailureofbrain
ion homeostasis can result in depolarisa
tion of adjacent cells, and contribute to
furtherspreadofdepolarisation.23;24
These changes are accompanied by
profoundfluctuationsinregionalcerebral
blood flow (rCBF), and to some extend
loss of bloodbrainbarrier (BBB) integ
rity.25 The changes in rCBF include typi
cally an initial hyperemia lasting 3 to 4.5
minutes,followedbymildhypoperfusion
lasting60to120minutes.26Similartothe
CSD, the hyperemia/hypoperfusion
spreads across the cortex, and CSD and
flow changes seem to halt typically at
majorsulci.
A number of case reports described
unilateral,hemisphericMRIchangesafter
(prolonged) aura, affecting regions cor
responding to patients’ symptoms. In
these cases cortical thickening, sulcal
effacementandincreasedsignalonfluid
attenuated inversion recovery (FLAIR)
andT2imagesalongthe(mostlyparieto
occipital) cortex, and to a less extent in
the subcortical white matter, were de
scribed.27;28 In one case increased signal
intensity on diffusionweighted images
without reduction of apparent diffusion
coefficient was noted (vasogenic
edema),28 whereas others found evi
denceforpresenceofcytotoxicedema.29 Concurrentleptomeningealbutalsogyral
parenchymal enhancement has been
reported in a number of cases, including
typical migraine aura, prolonged aura,
spontaneous and familial hemiplegic
migraineattacks.27;2935Followupimaging
was in most cases unremarkable, but
some reports described laminar cortical
necrosisand/orcorticalatrophy.29
Findings indicate that occasionally
migraineauraoccurswithMRIdetectable
cortical edema (either vasogenic or
cytotoxic), and breakdown of BBB, lead
ing to vasogenic leakage and cortical
and/or meningeal enhancement. Dis
turbed ion homeostasis and metabolism
due to CSD, may thus lead to BBB
dysfunction, and can account for tempo
rary impairment of cortical function.
Vasogenicleakagecould delay spontane
ous recovery of neuronal suppression.35
Similar changes may thus also occur
during normal migraine and/or aura
attacks, but remain below the current
MRIdetectionlevel.
Why CSD develops is still unknown,
but the propensity to develop CSD is
likelytobeinfluencedbyseveralfactors.
Studies in animal models of familial
hemiplegicmigrainestronglysuggestthat
mutations in ion transporter genes lead
to increased glutamate and potassium
levels in synaptic clefts, that facilitate
excitability of involved cortical areas.36 Several other internal or external factors
may also be involved in mechanisms
leadingtoincreasedcorticalexcitability.37 BasedonthefactthatinMAaurasymp
toms almost always consist of visual
symptoms,theoccipitallobesseemtobe
mostsusceptible.However,thepresence
of CSD in other regions has been under
explored,butitcanoccurinthecerebel
lum,forinstance.38
In experimental studies in rat, but
not in humans, CSD has been shown to
beabletoactivatethetrigeminovascular
system, and to induce a longlasting
bloodflow enhancement within the
middle meningeal artery and plasma
protein leakage in the dura mater.39 In
this scenario, with CSD as the pivotal
event, dysfunction of antinociceptive
brainstem nuclei is likely permissive,
leading to ‘central trigeminal hyperexcit
ability’.17 This would provide a link be
tween the aura and the headache phe
nomena in MA3941 and may be also in
MO.Toexplainthelatter,theconceptof
‘silent CSD’ has been introduced in the
literature, as a possible pathway for the
development of headache in MO pa
tients,butevidenceforthishypothesisis
stilllacking.17;4244
MIGRAINEASARISKFACTORFOR
BRAINLESIONS
Migraine has been considered for dec
ades as an episodic disorder without
longterm consequences to the brain.
However, over the past 30 years several
studies have been carried out looking at
and delivering arguments for a possible
association between migraine and brain
changes. First, several cases of ‘migrain
ous infarction’ or ‘migraineinduced
stroke’ have been described, suggesting
thatmigrainecanactasanacuteprecipi
tantofischemicstroke.4550Insuchcases
stroke is assumed to be directly and
causally related to an acute migraine
attack. Second, data from several hospi
talbased stroke casecontrol studies
suggestthatmigraineis–atleast–arisk
factor for ischemic stroke.5164 A recent
metaanalysis summarized the evidence
for migraine as a risk factor for clinical
ischemic stroke, and calculated a pooled
relative risk of 2.2 for migraineurs.65 Third, a number of clinicbased MRI
studiesfoundanincreasedprevalenceof
cerebral white matter hyperintense
lesions (WMLs) in migraine patients.6670 Anoverviewoffindingsintheseprevious
studiesislistedinchapter2.
Although most of these studies
showedanapparentassociationbetween
migraine and ischemic brain lesions,
selection bias and various other meth
odological problems (see chapter 2,
‘Methodological issues‘) prohibited
drawing definite conclusions from those
results. It remained uncertain whether
migraineurs are at independent risk for
permanentbrainchanges,andifso,there
remained still controversies whether
specific subgroups of patients would be
most at risk, and what etiologic mecha
nisms are involved. Finding definite
answerstothesequestionsisparticularly
importbecauseofthehighprevalenceof
migraineinthegeneralpopulation.
In order to find such answers, we
planned and performed a crosssectional
MRI study in an already existing popula
tionbased sample of adults with mi
graine and controls without a headache
history: the Cerebral Abnormalities in
Migraine, an Epidemiological Risk Analy
sis (CAMERA) study.7;71 In this study we
assessedthepresenceofseveraltypesof
brain changes (see below), compared
thesebetweenmigraineursandcontrols,
andcorrelatedpresenceandextensionof
these changes also to various demo
graphic, medical and specific migraine
characteristics. In addition, we collected
data from neurologic physical examina
tion and cognitive tests, to correlate
thesedatawithlesionload.Asanexten
sion to the main purpose of the study,
some other migrainerelated topics were
addressed with questions on family
history of migraine, symptoms of epilep
sia, visual sensitivity, food consumption
(like caffeine and alcohol use), fainting
history,andsymptomsofdepression.The
overall design of the study aimed as far
as possible to exclude potential sources
of bias, to minimize the risk of finding
false associations, and to remain with
unequivocalanswersaboutmigraineasa
riskfactorforbrainlesions.
AIMSANDOUTLINE
Primaryaimofthisthesisistoanswerthe
question whether migraineurs are at
independentlyincreasedriskofstructural
brain lesions. We describe the motives,
aims, methodology, and primary results
of the populationbased CAMERA study,
anddiscusstherelevanceofthefindings.
The results consist of several interictal
MRIfindingsthatarerelatedtomigraine
variables; with respect to these relation
ships and associations potential con
founding and/or etiologic factors are
evaluated.
In addition, secondary aims for this
thesisinclude:
x to review preexisting evidence of an
association between migraine, white
matter lesions and stroke, and to
summarize methodological issues af
fecting the interpretation of these
earlierstudies,
x to identify subgroups of migraineurs
thataremostatrisk,
x toevaluatecluesforetiologicmecha
nismsintherelationshipbetweenmi
graine and structural brain lesions,
and
x to examine the interrelationship(s)
between migraine, symptoms of
autonomic nervous system dysfunc
tionandbrainlesions.
Although many other types of data have
beencollected(seeabove),thesearenot
part of this thesis, and are in part still
underevaluation,orintendedasbaseline
measures for future followup research.
Below,anoutlineofthisthesisperchap
terisgiven.
InCHAPTER 2, we present an over
view of the existing literature that was
thestartingpointfortheCAMERAstudy.
In addition, this chapter describes the
complex relationship between migraine,
potential comorbid and/or confounding
factors and the occurrence of brain
lesions. It further emphasizes the rele
vance of proper methodology in assess
ing the relationship between migraine
and brain lesions, as was applied in the
CAMERAstudy.
CHAPTER 3 describes the overall
methodology and primary MRI results of
the CAMERA study. The objective is to
assess whether migraineurs from the
general population are at independently
increased risk of brain infarcts and
WMLs, and to identify migraine charac
teristics associated with these lesions.
CHAPTER 4 AND 5 focus on prevalence
and imaging features of two types of
specific brain lesions in migraine cases:
posterior circulation territory infarctlike
lesions and infratentorial hyperintense
lesions. We tried to assess whether
localization and aspect of lesions may
provide clues for pathophysiological
mechanismsbehindthem.
InCHAPTER6,wecomparedtheiron
concentrationsinsevendeepbrainnuclei
between (subgroups of) migraineurs and
controls.Thisassessmentofbrainironin
migrainewasinitiatedbecauseincreased
iron depositions in the periaqueductal
greymatterwerereportedbyothersina
preliminary study in migraineurs, sug
gesting an impaired central antinocicep
tive neuronal network.72 We add new
findingsthatfurthersupportinvolvement
of central pain processing networks in
migraine.
CHAPTER 7examinestheassociation
between migraine and autonomic ner
vous system (ANS) symptoms, including
syncope and orthostatic intoler
ance/insufficiency. CHAPTER 8 explores
the relationship between these ANS
symptomsandbrainlesions,andassesses
whether frequent syncope and or
thostaticinsufficiencyinmigraineursmay
explain a part of their increased risk of
brainlesions.
Theresultsofthisthesisaresumma
rized inCHAPTER 9, and relevance of the
findings is discussed. Finally, suggestions
for future research opportunities based
ontheresultsinthisthesisareprovided.