p(leaRTcaAcafpwcp

UaRr

5

0d

Usefulness of Transcatheter Patent Foramen Ovale Closure inMigraineurs With Moderate to Large Right-to-Left Shunt and

Instrumental Evidence of Cerebrovascular Damage

Marco Papa, MDa, Achille Gaspardone, MPhil, MDb,*, Gabriele Fracasso, MDa, Silvia Ajello, MDa,Gaetano Gioffrè, MDb, Maria Iamele, MDb, Cesare Iani, MDc, and Alberto Margonato, MDa

Transcatheter patent foramen ovale (PFO) closure might be effective in improving mi-graines. To assess the efficacy of PFO closure in migraineurs with a moderate to largeright-to-left shunt and instrumental evidence of embolic cerebral damage, 76 highly symp-tomatic migraineurs were prospectively investigated. The presenting clinical syndrome wasstroke in 16 patients, repeated transient ischemic attack in 32 patients, and lone migraineassociated with cerebral ischemic lesions on magnetic resonance imaging in 28 patients.Migraine severity was assessed before PFO closure and monthly for 6 months afterdiscontinuation of antiplatelet therapy. At the end of 12 months of follow-up, the averagedpostprocedural total score was compared with the baseline score. Transcatheter PFOclosure was successful in all patients, and the 12-month PFO closure rate was 97%. Thebaseline total migraine score was similar in patients with stroke, transient ischemic attack,and lone migraine (6.8 � 1.6, 6.7 � 1.4, and 6.9 � 1.7 respectively, p � NS). After a meanfollow-up of 13.7 � 2.4 months, no recurrent cerebrovascular episodes had occurred. At theend of the follow-up period, a significant reduction in the total migraine score was observedin all groups, regardless of the initial clinical presentation. Migraine was completelyabolished in 35 patients (46%), improved in 27 (36%), and unchanged in 14 (18%). Theproportion of patients with migraine suppression and improvement was similar in the 3groups. In conclusion, in highly symptomatic migraineurs with previous ischemic cerebralevents and instrumental evidence of cerebral embolism, transcatheter PFO closure canresult in improvement of migraine severity in a high percentage of patients. © 2009

Published by Elsevier Inc. (Am J Cardiol 2009;104:434–439)

sd

M

vhMRPtp(irwa1cyvtit

Cryptogenic stroke and migraine might share the sameathogenetic mechanism mediated by right-to-left shuntingRLS) across a patent foramen ovale (PFO).1–6 The preva-ence of PFO with RLS in patients with cerebrovascularvents and migraine is 2 to 3 times greater than in controls,nd the prevalence of migraine in patients with PFO andLS is 2 times greater than in the general population.7–11

he possibility exists that migraine might represent the finalommon clinical phenotype of different pathogenetic mech-nisms, of which only some might be related to RLS.12

lthough migraine associated with RLS seems to be clini-ally similar to migraine without RLS, the attempt to selectsubpopulation of migraineurs with common instrumental

eatures might help in identifying specific subgroups ofatients that could benefit from PFO closure. To this aim,e prospectively assessed the effects of percutaneous PFO

losure on migraine in a very selected group of patients withrevious cryptogenic cerebral ischemic events and in

aCardio-Thoracic Department, Congenital and Structural Heart Diseasenit, Istituto Scientifico, San Raffaele, Milano; bDivisione di Cardiologia

nd cDivisione di Neurologia, Ospedale S. Eugenio, ASL RMC Roma,ome, Italy. Manuscript received January 18, 2009; revised manuscript

eceived and accepted March 21, 2009.*Corresponding author: Tel: (�39) 06-5100-2320; fax: (�39) 06-

100-2836.

cE-mail address: a_gaspardone@yahoo.com (A. Gaspardone).

002-9149/09/$ – see front matter © 2009 Published by Elsevier Inc.oi:10.1016/j.amjcard.2009.03.061

trongly symptomatic migraineurs with instrumental evi-ence of asymptomatic cerebral damage.

ethods

Of a cohort of 182 patients with documented cerebro-ascular events or lone migraine referred to 2 differentospitals (Cardiothoracic Department, S. Raffaele Hospital,ilan; and Division of Cardiology, S. Eugenio Hospital,ome) from March 2006 to June 2007, for transcatheterFO closure evaluation, 76 were migraineurs according to

he criteria of the International Headache Society.13 Theresenting clinical syndrome was stroke in 16 patientsmean age 47 � 10 years, range 25 to 66; 9 women), transientschemic attack (TIA) in 32 patients (mean age 44 � 11 years,ange 24 to 70; 22 women), and lone migraine associatedith cerebral ischemic lesions on magnetic resonance im-

ging (MRI) in 28 patients (mean age 39 � 12 years, range9 to 65; 21 women; Table 1). The prespecified study entryriteria were a �1 year history of migraine and age �50ears at onset; �3 migraine attacks per month in the pre-ious 6 months; and the lack of a, or a partial reduction inhe, response to �3 types of preventive medications, includ-ng topiramate, with no contraindications to antiplateletreatment.

On admission, all patients underwent full neurologic and

ardiologic examinations, including brain MRI, routine co-

www.AJConline.org

ase((cuastittTrpmjtdttDtceosdsapdsaloroaptcfi

T(bmwshaapmas

rieerte

TS

V

MMMAPBAB

TA

S

I

D

F

A

435Miscellaneous/Transcatheter Patent Foramen Ovale Closure

gulation screening, Holter electrocardiographic monitoring,upra-aortic vessel Doppler ultrasonography, transthoracicchocardiography, and transesophageal echocardiographyTEE). Brain MRI scans were executed on a 1.5-T scannerGiroscan NT, Philips, Eindhoven, Holland). The MRI proto-ol included baseline fast spin echo T1-weighted, fluid-atten-ated inversion recovery T2-weighted, T1-weighted sagittalnd T2-weighted coronal sequences, and T1-weighted axial,agittal, and coronal sequences after gadolinium diethylene-riamine penta-acetic acid intravenous injection. For eachmage series, 20 slices covering the entire brain (matrix 240o 256 � 256 to 320; field of view 22 to 24 cm; slicehickness 5 mm; interslice thickness 1.5 mm) were obtained.he MRI scans were evaluated on digital images by 2 expe-

ienced neuroradiologists unaware of the clinical data and theresence of RLS. Only patients with brain imaging evidence ofultiple bilateral ischemic lesions in cortical, subcortical, and

unctional gray-white matter territories were selected, becausehese lesions are highly suggestive of an embolic origin.14–17 Aetailed recording of the risk factors for atherosclerosis andhromboembolism was performed. The exclusion criteria werehe presence of even minimal carotid artery disease on carotidoppler scanning, transthoracic echocardiographic and/or

ransesophageal echocardiographic evidence of a left-sidedardiac or aortic potential source of peripheral embolism, andvidence of supraventricular/ventricular rhythm disturbancesn Holter electrocardiographic monitoring. As a part of thetandard protocol, all patients underwent a microbubbles testuring the preclosure TEE by injecting agitated saline-bloodolution (1 ml air, 1 ml blood, and 8 ml saline) into a rightntecubital vein in the patient who was in the left lateral supineosition. Of the 76 patients, 37 required light sedation. Aetailed atrial septum anatomy description, including atrialeptal aneurysm (defined as abnormally redundant inter-trial septum with an excursion of �10 mm into the right oreft atrium and a base span of �15 mm), and an evaluationf microbubbles appearing in the left atrium during normalespiration and after provocative maneuvers (Valsalva with-ut previous inspiration and/or in noncooperative patientsfter intense coughing 2 to 3 times) were assessed in allatients. RLS was semiquantitatively graded according tohe amount of microbubbles detected in the left atrium afterrossing the interatrial septum on a still frame during the

able 1tudy population basal characteristics

ariable

Stroke(n � 16)

ean age (yrs) 47 � 10en 4 (25%)igraine with aura 10 (62%)verage migraine duration (yrs) 14 � 4revalence of �1 cardiovascular risk factor 9 (56%)aseline maximal RLS 2.5 � 0.5trial septal aneurysm 2 (12%)aseline migraine score 6.8 � 1.6

* Lone migraine versus stroke.† Lone migraine versus stroke and TIA.

rst 5 cardiac cycles of contrast entering the right atrium. l

he grading system was as follows: grade 0, none; grade 1minimal), 1 to 10 bubbles; grade 2 (moderate), 10 to 20ubbles; and grade 3 (severe), �20 bubbles.18 The maxi-um RLS severity was used for the analysis. Only patientsith RLS of grade 2 or greater were included in the present

tudy. The admission protocol also included standardizedistory taking and administration of a detailed questionnaireddressing headache severity in the previous 2 monthsccording to Anzola et al19 (Table 2). On discharge, theatients received a diary in which they were to report theonthly intensity, duration, frequency, and occurrence of

ura of their migraine attacks, if any, using the migraineeverity scoring system.

All PFO closure procedures were performed with fluo-oscopic monitoring, and intraoperative ultrasound monitor-ng was performed with TEE in 60 patients and intracardiacchocardiography in the remaining 16 patients. When trans-sophageal echocardiographic monitoring was used, patientsequired light sedation. Intracardiac echocardiographic moni-oring was performed using a 9F, 9-MHz ultrasound cath-ter (Ultra ICE, EP Technologies, Ohringen, Germany) with

Group

TIA(n � 32)

Migraine(n � 28)

p Value

45 � 11 39 � 12 0.035*9 (29%) 7 (25%) NS

20 (64%) 18 (64%) NS13 � 5 15 � 3 NS16 (50%) 8 (29%) 0.038†

2.4 � 0.5 2.3 � 0.4 NS3 (9%) 3 (11%) NS

6.7 � 1.4 6.9 � 1.7 NS

able 2nzola migraine severity score

core Description

ntensity0 No pain1 Mild (not interfering with daytime activity)2 Severe (interfering with daytime activity)3 Unbearable (confined to bed)uration (h)0 No pain1 �62 6–123 �12requency (migraines/mo)0 No pain1 1–42 5–103 �10ura0 No aura1 Aura in �1 migraine attack

ocal anesthesia provided by way of the left femoral vein.

BAttTpGMSpdhc

dadduw1Hrpaw6tum1spmfaam

daiatrueKnu

pctpac

R

phrsw7smbdwmvppgri

s

FCrsmb

TB

G

STM

436 The American Journal of Cardiology (www.AJConline.org)

alloon sizing of the PFO was performed in 4 patients only.ll patients received heparin 70 IU/kg at the beginning of

he procedure, followed, if necessary, by additional boluseso maintain an activated clotting time of �200 seconds.ranscatheter closure of the interatrial communication waserformed using the Amplatzer device (AGA Medicalolden Valley, Minnesota) in 45 patients, STARFlex (NMTedical, Boston, Massachusetts) in 22 patients, and PFO-

TAR (Applied Biometrics, Burnsville, Minnesota) in 9atients. Procedural success was defined as successfulevice implantation without major periprocedural or in-ospital major complications. Immediate PFO successfullosure was defined as a RLS of grade 1 or less on TEE.

On discharge, aspirin was given at the dose of 5 mg/kg/ay for 6 months until October 2006. Since October 2006,spirin at a dose of 100 mg/day for 6 months and clopi-ogrel at a dose of 75 mg/day for 3 months was the standardischarge treatment. All patients were allowed to take thesual antimigraine preparation as continuous therapy orhenever necessary. A clinical evaluation was performed at, 3, 6, and 12 months after PFO closure. Ambulatoryolter electrocardiography and transthoracic echocardiog-

aphy were done at 1 month to evaluate for correct deviceositioning. TEE with the microbubbles test was performedt 12 months in all patients. The presence of residual shuntas recorded and classified as previously reported.18 At the-month follow-up visit, the patients were asked to stopheir antiplatelet therapy and to record the migraine severitysing the migraine score system previously described everyonth (from the seventh to the twelfth month). At the

2-month follow-up visit, the monthly subscores wereummed and averaged for the previous 6 months for eachatient and compared with the preprocedural mean baselineigraine severity score that had been individually assessed

or 2 months before PFO closure. Cessation of migraine wasrbitrarily defined as the complete abolition of migrainettacks and migraine improvement as a reduction in theigraine score of �2 points.Continuous data are expressed as the mean and standard

eviation; discrete variables are given as absolute valuesnd percentages. Continuous variables were compared us-ng the 2-sided unpaired t test or analysis of variance, asppropriate. For multiple comparisons between groups, thetest with Bonferroni’s correction was used to avoid spu-

ious significances. Categorical variables were comparedsing the chi-square test, with Yates’ correction, or Fisher’sxact test, as appropriate. The Mann-Whitney U test orruskal-Wallis nonparametric test were used to compareon-normally distributed data. Correlations were evaluated

able 3aseline and 12-month migraine severity scores

roup Baseline

Intensity Duration Frequency Aura

troke 2.3 2.2 1.6 0.7IA 2.3 2.1 1.8 0.5igraine 2.4 2.1 ...1.8 0.6

* p �0.001 compared with baseline score.

sing linear regression analysis. All tests were 2-sided. A t

robability value of �5% was considered significant. Cal-ulations were performed using Statistical Analysis Sys-ems, version 8e (SAS Institute, Cary, North Carolina). Theresent study was performed without any financial supportnd was approved by the institutional independent ethicommittees. All patients provided written informed consent.

esults

The closure devices were successfully deployed in all 76atients. Periprocedural complications were limited to groinematoma in 5 patients that in all cases spontaneouslyesolved within 1 month after discharge. The mean fluoro-copic time was 10 � 6 minutes. The average hospital stayas 2 � 1 nights. No patient was lost to follow-up. Of the6 patients, 39 (51%) had basal RLS, and in all patients, ahunt could be induced during provocative maneuvers. Theean severity of RLS was 2.4 � 0.5. No differences could

e detected in the severity of the shunt among patients withifferent clinical presentations nor within groups in patientsith or without aura. Immediate residual RLS using theicrobubble test, evaluated immediately after closure de-

ice deployment and before echo-monitoring removal, wasresent in 6 patients (8%). In 2 patients, the severity of therovoked shunt was grade 2 and in the remaining 4 patients,rade 1. At 12 months, TEE demonstrated the presence ofesidual shunt greater than grade 1 in 2 of the 6 patients withmmediate postprocedural shunting (2.6%).

The prevalence of risk factors for cerebral embolism wasignificantly less in the patients with migraine alone than in

igure 1. Plot showing outcome at end of 12-month follow-up period.ured indicates migraine cessation; improved indicates migraine score

eduction of �2 points (baseline score minus 12-month follow-up score);teady indicates migraine score reduction of �2 points (baseline scoreinus 12-month follow-up score). Data within bars indicate absolute num-

er of patients.

12 mo

l Intensity Duration Frequency Aura Total

1.6 0.9 0.9 0 3.4*1.4 0.9 1.0 0.1 3.4*1.5 0.9 1.0 0.2 3.6*

Tota

6.86.76.9

he patients with stroke or TIA (Table 1). In 5 patients (6%),

ar(shme�ebs1pTppmbmtawpwpaumgtmmir1ep(wt

D

hhtmsthpvgrPsgunw

tuetrntatIp(dcbeimimdmltndswetimslohtarqgmpe

AstqasPomtamesl

437Miscellaneous/Transcatheter Patent Foramen Ovale Closure

decrease of �1 g/dl in the baseline hemoglobin level wasegistered because of procedural blood loss. Eight patients10%) reported palpitations in the 4 weeks after PFO clo-ure, and 5 patients (6%) reported a sensation of continuouseaviness in the chest. Follow-up electrocardiographiconitoring at 1 month revealed sporadic supraventricular

ctopic beats and a brief episode of atrial fibrillation lasting4 minutes in 1 patient. At the 3-, 6-, and 12-month clinical

valuation, no patient reported cardiologic symptoms. Theaseline total migraine score was similar in patients withtroke, TIA, and migraine (6.8 � 1.6, 6.7 � 1.4, and 6.9 �.7, respectively; p � NS; Table 1). Migraine with aura wasresent in 10 patients in the stroke (60%), 20 patients in theIA group (64%), and 18 in the migraine alone group (64%;� NS). All patients were regularly examined in the out-

atient clinic at the scheduled follow-up points. After aean follow-up of 13.7 � 2.4 months, no recurrent cere-

rovascular events had occurred. Overall, the severity ofigraine and the presence of aura were not associated with

he severity of the baseline maximal RLS (r � 0.36, p � NSnd r � 0.32, p � NS, respectively). Also, no correlationas found within each group. At the end of the follow-uperiod, a significant reduction in the total migraine scoreas observed in all groups, regardless of the initial clinicalresentation (Table 3). Overall, migraine was completelybolished in 35 patients (46%), improved in 27 (36%), andnchanged in 14 (18%). The proportion of patients withigraine suppression and improvement was similar in the 3

roups (Figure 1). In the 2 patients with residual shunt onhe 12-month transesophageal echocardiographic scan, theigraine was completely abolished. Although an increase inigraine intensity was observed in 3 patients, 2 without an

mmediate residual shunt and 1 with grade 1 immediateesidual shunting, in the first 2 months after PFO closure, at2 months of follow-up, no patient presented with a wors-ning of migraine status. At 12 months of follow-up, 11atients (14%) were still taking acutely (7 patients) or stably4 patients) antimigraine therapy. Finally, no correlationas found between the migraine status at follow-up and the

ype of device used.

iscussion

A number of epidemiologic and clinical observationsave supported an association between PFO and migraineeadache, although the mechanism of the migraine relationo PFO remains largely speculative.4,20 Shunting of the

igraine-triggering mediators such as activated platelet,erotonin, and/or other chemicals to the systemic circula-ion, bypassing the lung filter and catabolization, has beenypothesized, but this mechanism remains largely un-roved. Although recent systematic overviews have pro-ided no compelling evidence of a relation between mi-raine relief and PFO closure, several investigators haveeported migraine abolition or partial improvement afterFO closure.11 Since the original observation by Wilm-hurst et al,3 a variety of single-center reports have sug-ested an improvement in migraine severity in patientsndergoing percutaneous PFO closure for migraine andonmigraine indications.19,21–25 However, these studies

ere hampered by important methodologic flaws in that v

hey were mostly retrospective, included nonselected pop-lations, usually had a short follow-up period, and did notxclude potential interference from medical antiplateletreatment on migraine severity.26 The recently publishedesults of the Migraine intervention with STARFlex tech-ology (MIST) trial, the only randomized study performedo date, although confirming an association between RLSnd migraine with aura, nevertheless have clearly refutedhe encouraging results of previous observational reports.27

n that trial, no significant beneficial effect was found forrimary (migraine cessation within 180 days) and secondaryreduction of migraine severity) end points in patients un-ergoing percutaneous PFO closure compared with the pla-ebo-sham arm. Although that important trial was hamperedy methodologic flaws that could have weakened its clinicalffect, the discrepant results obtained in the MIST trial andn previous studies remain difficult to explain.28 Because

igraine associated with RLS appears to be clinically sim-lar to migraine without RLS, the possibility exists thatigraine might represent a final common phenotype of

ifferent pathogenetic mechanisms, of which only someight be related to RLS.12 In an attempt to overcome the

imitations of the previous investigations and to search fur-her specific migraine populations with common pathoge-etic features of cerebral embolism, the present study wasesigned to evaluate the efficacy of percutaneous PFO clo-ure in patients in which significant RLS (grade 2 or greater)as associated with brain imaging evidence of ischemic

vents likely due to cerebral embolism. In our investigation,o reduce the risk of including patients with a similar clin-cal phenotype but potentially different pathophysiology forigraine, we enrolled only patients who, in addition to

imilar subjective migraine symptoms, also had neurologicesions on brain imaging, suggesting a similar pathophysi-logic embolic mechanism. To this aim, we selected onlyighly symptomatic migraineurs with PFO and moderate-o-large RLS in whom the clinical symptoms were associ-ted with objective evidence of potentially postembolic ce-ebral damage. Furthermore, because recent reports haveuestioned the specificity of white matter lesions in mi-raineurs with RLS, we also enrolled patients with loneigraine and MRI cortical, subcortical, or junctional lesion

attern, which are more likely to be caused by systemicmbolism.13–17,29,30

Migraine was evaluated using the score proposed bynzola et al.19 This is a simple self-reporting evaluation

core system that has been previously validated and permitshe separate recording of the intensity, duration, and fre-uency of the migraine attacks, as well as the occurrence ofura. In our study, the baseline self-evaluation of migrainetatus was performed prospectively for 2 months beforeFO closure, thus ruling out the possibility of bias from theverestimation of preprocedural migraine severity. Further-ore, to exclude the potential interference of antiplatelet

herapy with migraine status, migraine status was evaluatedfter the interruption of antiplatelet therapy at the sixthonth for an additional period of 6 months.19,26 The trans-

sophageal echocardiographic evaluation of the residualhunt was also performed at the twelfth month, a periodong enough to permit complete endothelization of the de-

ice in most patients.28

sdhsttdsstbcr

aobmftsbotbtatAbmbd

aMismaMoRsirewao

tpwtbocvw

cdeniglatcttcmtrpdhups

ALtp

1

1

1

438 The American Journal of Cardiology (www.AJConline.org)

Our results have indicated that in a selected populationharing similar clinical and brain imaging features likelyue to systemic embolism, percutaneous PFO closure isighly effective in reducing migraine severity. Migraineuppression or improvement was reported by about 80% ofhe patients at 12 months and this percentage was consis-ently observed, regardless of the presenting clinical syn-rome. These findings, which have confirmed previous ob-ervational studies, reopen the possibility that in highlyelected patients, PFO closure might represent an effectiveherapeutic strategy. The 12-month closure rate as evaluatedy TEE was 97%, and the periprocedural and 12-monthomplication rate was virtually absent, showing that in expe-ienced centers, this procedure is both safe and effective.

No correlation was found in our study between themount of RLS and the severity of migraine in all patientsr in the specific groups. This finding could be partiallyiased by having selected only homogenous patients withoderate to large shunts, using the semiquantitative method

or migraine grading and by selecting only highly symp-omatic patients, thus greatly reducing the variability of ourtudy population. Nevertheless, a clear temporal associationetween PFO closure and migraine improvement could bebserved, thus confirming a causal link between this ana-omic-functional feature and migraine associated with cere-rovascular damage. In this context, in at least some pa-ients with migraines, shunting to the systemic circulation of

substantial amount of vasoactive mediators might facili-ate headache attacks by lowering the migraine threshold.fter a mean follow-up of 13.7 months, no recurrent cere-rovascular events occurred, indicating that PFO closureight represent an effective tool to prevent ischemic em-

olism in selected patients even after antiplatelet therapyiscontinuation.

One relevant issue in the evaluation of migraine head-che therapy is the potential presence of a placebo effect.igraine is known to have one of the largest placebo effects

n medicine, and previous investigations have reported aignificant incidence of placebo effect in migraine improve-ent with different therapeutic strategies.31 The presence ofsubstantial placebo effect was further confirmed by theIST trial results in sham-operated patients and by the

bservation that migraine relief occurred despite residualLS after transcatheter PFO closure.27,32 However, in our

tudy, about 80% of the patients showed objective migrainemprovement, a result that is substantially greater than theeported placebo threshold of 30% to 40% for a therapeuticffect recommended for migraine studies.33 Furthermore,e selected only patients with evidence of definite or prob-

ble embolic cerebral damage, thus reducing the possibilityf functional somatization of migraine symptoms.

Some limitations of our study should be considered. Inhe present study, we selected only highly symptomaticatients with instrumental evidence of cerebral damage,hich likely represented only a portion of the total popula-

ion of patients with migraine. Thus, our findings should note directly extrapolated to all patients with migraine. Sec-nd, the low rate of complications and the high rate of PFOlosure obtained were the consequence of the previous high-olume case load performed at the 2 centers with experience

ith �600 patients. The lack of an established independent

ore laboratory for echocardiographic and brain imagingata analysis was definitely a limitation of our study. How-ver, the brain images were evaluated by 2 experiencedeuroradiologists who were unaware of the presenting clin-cal syndrome, and the potential bias in the echocardio-raphic demonstration of RLS was reduced by having se-ected only patients with moderate to large shunts. RLS wasssessed by TEE; this technique is less sensitive than con-rast-enhanced transcranial Doppler but probably more spe-ific, and, when properly done, it permits a careful evalua-ion of the septal anatomy.18 Additionally, the follow-up inhe present study was relatively short; thus, no definitiveonclusions about the long-term effect of PFO closure onigraine status could be drawn. Finally, and most impor-

antly, the present study was not randomized, and the recentesults of the MIST trial have clearly emphasized the im-ortance of performing controlled investigations beforerawing final conclusions. However, we tried to select aomogenous study population and to control the methodsed very carefully to overcome many of the flaws ofrevious studies. This approach could represent a suitabletarting point for future randomized trials.

cknowledgment: We thank Lisa Befani, MD, Rosariaicitra, MD, Sara Rossi, MD, and Arianna Giardina, MD,

he catheterization laboratory staff, and all patients partici-ating in the present study.

1. Del Sette M, Angeli S, Leandri M, Ferriero G, Buzzone GL, FinocchiC, Gandolfo G. Migraine with aura and right-to-left shunt on trans-cranial Doppler: a case-control study. Cerebrovasc Dis 1998;8:327–330.

2. Anzola GP, Magoni M, Guindani M, Rozzini L, Dalla Volta G.Potential source of cerebral embolism in migraine with aura: a trans-cranial Doppler study. Neurology 1999;52:1622–1625.

3. Wilmshurst P, Nightingale S, Walsh KP, Morrison WL. Effect onmigraine of closure of cardiac right-to-left shunts to prevent recurrenceof decompression illness or stroke or for haemodynamic reasons.Lancet 2000;356:1648–1651.

4. Wilsmhurst P, Nightingale S. Relationship between migraine and cardiacand pulmonary right-to-left shunts. Clin Sci 2001;100:215–220.

5. Schwedt T, Dodick D. Patent foramen ovale and migraine: bringingclosure to the subject. Headache 2006;46:663–671.

6. Wilmshurst P, Nightingale S, Pearson M, Morrison L, Walsh K.Relation of atrial shunts to migraine in patients with ischemic strokeand peripheral emboli. Am J Cardiol 2006;98:831–833.

7. Lamy C, Giannesini C, Zuber M, Arquizan C, Meder JF, Trystram D,Coste J, Mas JL. Clinical and imaging findings in cryptogenic strokepatients with and without patent foramen ovale: the PFO-ASA studyatrial septal aneurysm. Stroke 2002;33:706–711.

8. Milhaud D, Bougousslavsky J, van Melle G, Liot P. Ischemic strokeand active migraine. Neurology 2001;57:1805–1811.

9. Stajzel R, Genoud D, Roth S, Mermillod B, Le-Floch-Rohr J. Patentforamen ovale, a possible cause of symptomatic migraine: a study of 74patients with acute ischemic stroke. Cerebrovasc Dis 2002;13:102–106.

0. Mas JL, Arquizan C, Lamy C, Zuber M, Cabanes L, Derumeaux L,Coste J. Recurrent cerebrovascular events associated with patent fo-ramen ovale, atrial septal aneurysm, or both. N Engl J Med 2001;345:1740–1746.

1. Schwedt TJ, Demaerschalk BM, Dodick DW. Patent foramen ovaleand migraine: a quantitative systematic review. Cephalalgia 2008;28:531–540.

2. Anzola GP, Meneghetti G, Zanferrari C, Adami A, Dinia L, Del SetteM; SAM Study Group. Is migraine associated with right-to-left shunt

a separate disease? Results of the SAM study. Cephalalgia 2008;28:360–366.

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

439Miscellaneous/Transcatheter Patent Foramen Ovale Closure

3. Headache Classification Committee of the International Headache So-ciety. Classification and diagnostic criteria for headache disorders,cranial neuralgias and facial pain. Cephalalgia 2004;363:281–391.

4. Romano A, Cipriani V, Bozzao A. Neuroradiology and headaches.J Headache Pain 2006;7:422–432.

5. Santamarina E, González-Alujas MT, Muñoz V, Rovira A, Rubiera M,Ribó M, Alvarez-Sabin J, Molina CA. Stroke patients with cardiacatrial septal abnormalities: differential infarct patterns on DWI. J Neu-roimaging 2006;16:334–340.

6. Kruit MC, Launer LJ, Ferrari MD, van Buchen MA. Infarcts in theposterior circulation territory in migraine: the population-based MRICAMERA study. Brain 2005;128:2068–2077.

7. Jauss M, Wessels T, Trittmacher S, Allendorfer J, Kaps M. Emboliclesion pattern in stroke patients with patent foramen ovale comparedwith patients lacking an embolic source. Stroke 2006;37:2159–2161.

8. Seiler C. How should we assess patent foramen ovale? Heart 2004;90:1245–1247.

9. Anzola GP, Frisoni GB, Morandi E, Casilli F, Onorato E. Shunt-associated migraine responds favorably to atrial septal repair: a case-control study. Stroke 2006;37:430–434.

0. Anzola GP, Morandi E, Casilli F, Onorato E. Different degrees ofright-to-left shunting predict migraine and stroke: data from 420 pa-tients. Neurology 2006;66:765–767.

1. Morandi E, Anzola GP, Angeli S, Melzi G, Onorato E. Transcatheterclosure of patent foramen ovale: a new migraine treatment? J IntervenCardiol 2003;16:39–42.

2. Schwerzman M, Wiher S, Nedeltchev K, Mattle HP, Wahl A, Seile C,Meier B, Windecker S. Percutaneous closure of patent foramen ovalereduces the frequency of migraine attacks. Neurology 2004;62:1399–1401.

3. Post MC, Thijs V, Herroelen L, Budts WI. Closure of a patent foramenovale is associated with a decrease in prevalence of migraine. Neurol-ogy 2004;62:1439–1440.

4. Azarbal B, Tobis J, Suh W, Chan V, Dao C, Gaster R. Association ofinteratrial shunts and migraine headaches. Impact of transcatheter

closure. J Am Coll Cardiol 2005;45:489–492.

5. Reisman M, Christofferson RD, Jesurum J, Olsen JV, Spencer MP,Krabil KA, Diehl L, Aurora S, Gray WA. Migraine headache reliefafter transcatheter closure of patent foramen ovale. J Am Coll Cardiol2005;45:493–495.

6. Wilmshurst P, Nightingale S, Walsh K, Morrison W. Clopidogrelreduces migraine with aura after transcatheter closure of persistentforamen ovale and atrial septal defects. Heart 2005;91:1173–1175.

7. Dowson A, Mullen MJ, Peatfield R, Muir K, Khan AA, Wells C,Lipscombe SL, Rees T, De Giovanni JV, Morrison WL, Hildick-SmithD, Elrington G, Hillis WS, Malik IS, Rickards A. Migraine interven-tion with STARFlex technology (MIST) trial. Circulation 2008;117:1397–1404.

8. Carroll JD. Migraine intervention with STARFlex technology trial: acontroversial trial of migraine and patent foramen ovale closure. Cir-culation 2008;117:1358–1360.

9. Del Sette M, Dinia L, Bonzano L, Roccatagliata L, Finocchi C, ParodiRC, Sivori G, Gandolfo C. White matter lesions in migraine andright-to-left shunt: a conventional and diffusion MRI study. Cephalal-gia 2008;28:376–382.

0. Adami A, Rossato G, Cerini R, Thijs VN, Pozzi-Mucelli R, AnzolaGP, Del Sette M, Finocchi C, Meneghetti G, Zanferrari C; SAM StudyGroup. Right-to-left shunt does not increase white matter lesion load inmigraine with aura patients. Neurology 2008;71:101–107.

1. Diener HC, Kronfeld K, Boewing G, Lungenhausen M, Maier C,Molsberger A, Tegenthoff M, Trampisch HJ, Zenz M, Meinert R;GERAC Migraine Study Group. Efficacy of acupuncture for the pro-phylaxis of migraine: a multicentre randomised controlled clinicaltrial. Lancet Neurol 2006;5:382–383.

2. Jesurum JT, Fuller CJ, Kim CJ, Krabill KA, Spencer MP, Olsen JV,Likosky WH, Reisman M. Frequency of migraine headache relieffollowing patent foramen ovale “closure” despite residual right-to-leftshunt. Am J Cardiol 2008;102:916–920.

3. van der Kuy P-HM, Lohman JJHM. A quantification of the placebo

response in migraine prophylaxis. Cephalalgia 2002;22:265–270.