THE JOURNAL OF PEDIATRICS

Children’s Hospital Medical Center, MLC-3021 3333 Burnet Ave.

Cincinnati, OH 45229-3039
513-636-7140 FAX: 513-636-7141
Embargoed until 12:00 AM, ET, March 31, 2011

For more information or to obtain a copy of the article, contact: Brigid Huey, 513-636-7140,
journal.pediatrics@cchmc.org

Study Suggests a Relationship between Migraine Headaches in Children

and a Common Heart Defect

Penelitian sebuah hubungan antara sakit kepala migrain pada anak dan
kerusakan jantung

Cincinati,OH,31 maret 2011,sekitar 15%dari anak2 ygg menderita migrain,dan

tepatnya 1/3 dari anak2 ini memiliki migrain dengan aura,didapati gejala2 spt
lemah,bintik buta,dan bahkan (even) halusinasi.Meskipun penyebab migrain
tidak jelas,sebuah penelitian terbaru muncul jurnal anak ttg hubungan antara
sakit kepala migrain pada anak dan kerusakan jantung yg dikenal dengan patent
foramen ovale,yg mana berdampak pada 25% org di US.
Dr. Rachel McCandless and colleagues from the Primary Children’s Medical Center and the University of Utah

Meneliti anak usia 6-18 tahun yang didiagnosis dengan migrain antara thn 2008
dan 2009.Sebanyak 109 anak2 dilibatkan dlm penelitian yg diterapi pada RS
anak yg melayani anak2 dari utah,idaho,monta,colorado dan sebagian wyoming.

Peneliti2 mengambil echocardiogram 2 dimensi dari masing2 jantung

anak,mencari a patent foramen ovale ( PFO),kerusakan umum pada dinding
antara 2 katup atas pada jantung.Meskipun sebuah PFO tidak membahayakan
,ini dapat meloloskan darah yg tidak terfilter(unfiltered blood) melewati paru2
dan sirkulasi keseluruh tubuh.Seperti yang diterangkan oleh dr.McCandless ,
beberapa peneliti2 menyarankan /menduga sebuah hubungan antara pemilik
PFO dengan migran.

Didalam penelitian anak2 yg mendapat migrain dengan aura,50% juga

mendapat PFO,ini merupakan nilai PFO hampir 2xnya dari populasi
umum.Bagaimanapun juga hanya 25%dari anak2 yg mendapat migrain tanpa
aura mendapat PFO. Hipotesa Dr. McCandless and colleagues menyatakan jika
hubungan penyebab bisa ditegakan , penutupan dari PFO dengan
cateter mungkin dapat menolong sebagai terapi yg penting apa jenis2
migrain terutama migrain dengan aura.ini menjadi harapannya bahwa
penelitian kami dapat membantu penelitian mendatang ttg kesulitan
masalah ini .

The study, reported in "Patent Foramen Ovale in Children with Migraine

Headaches" by Rachel T. McCandless, MD, Cammon B. Arrington, MD,
Douglas C. Nielsen, James F. Bale, Jr., MD, and L. LuAnn Minich, MD, appears
in The Journal of Pediatrics, DOI 10.1016/j.jpeds.2011.01.062, published by
Elsevier.

Patent Foramen Ovale in Children with Migraine

Headaches

Objective:Untuk determine? Prevalensi dari PFO pada anak dengan migrain

Studi design: Anak2 berusia 6 hingga 18 tahun dengan sakit kepala migrain yg
dievaluasi untuk PFO dan shunting right-left dengan scanning doppler
berwarna,echocardiography transthoracic dgn contras saline solution dan
scanning doppler contras transcranial.

Hasil:Dalam populasi yg terdiri dari 109 anak dengan migrain , 38 org ( 35%)
dengan aura dan 71 org ( 65%) tanpa aura .Semua prevalensi PFO adalah
35%,hampir sama dengan populasi umum ( 35% vs
25%,p=13).Bgmpun,perbandingan dengan populasi umum ( 25%),prevalensi
PFO lebih byk pada subyek dgn aura ( 50%,p=0,004) tapi mirip dengan yg
tanpa aura (27%,p=73).ukuran Atrial shunt ( sambungan atrium?)tidak
berhubungan dengan hasil ada atau tanpa aura.

Kesimpulan

Conclusion
Children with migraine with aura have a significantly higher prevalence of PFO compared with those without aura
or the general population. These data suggest that PFO may contribute to the pathogenesis of migraine with aura
in children and have implications for clinical decision making.

Migraine, a recurrent headache disorder of children and adults, significantly affects quality of life and results in a
substantial financial and social burden.1, 2 Migraine occurs in about 15% of the pediatric population, with
approximately one-third of cases associated with an aura.3 Patency of the foramen ovale (PFO), a normal fetal
connection between the atria allowing blood from the placenta to bypass the lungs, has been implicated in the
pathogenesis of migraine. Agitated saline solution contrast echocardiographic studies and autopsy reports have
found that the foramen ovale remains patent in 10% to 25% of the general population, and most studies of adults
with migraine with aura find a significantly higher prevalence of PFO ranging from 41% to 62% (composite of
56% by meta-analysis).4, 5, 6, 7, 8, 9, 10, 11, 12

The high prevalence of PFO in adults with migraine with aura has led to the hypothesis that a right-to-left shunt
across the atrial septum allows metabolic or microembolic triggers that would normally be cleared by the lungs to
pass unfiltered into the cerebral circulation, leading to the migraine. However, studies exploring the effectiveness
of PFO closure on the frequency and severity of migraine headaches in adults have had conflicting results.13, 14,
15 16 17 18 19
, , , ,

No comparable studies have been published in children. We hypothesized that children with migraine with aura
also have a higher prevalence of PFO compared with the general population and children with migraine without
aura. The purposes of this study were to determine the prevalence of PFO in pediatric patients with migraine and
to investigate the relationships between the amount of right-to-left atrial shunting and migraine subtype.

Back to Article Outline

Methods 
From February 2008 to September 2009, children 6 to 18 years of age diagnosed with migraine by the pediatric
neurologists at Primary Children’s Medical Center were invited to participate in this study. Primary Children’s
Medical Center is the major referral center for children living in the Intermountain West, a large geographic region
encompassing Utah and parts of Wyoming, Idaho, Montana, Nevada, and Colorado. The diagnosis of migraine
with or without aura was made according to the revised International Headache Society criteria.20 Children with
known congenital heart disease and patients unable to cooperate with imaging modalities were excluded.
Parental consent and participant assent were obtained as appropriate. The study was approved by the
Institutional Review Boards of the University of Utah and Intermountain Healthcare.

Complete 2-dimensional and Doppler transthoracic echocardiography (TTE) and transcranial Doppler (TCD)
scanning were performed on all participants with either an Acuson Sequoia C512 (Siemans, Mountain View,
California) or a Philips iE33 (Philips, Bothell, Washington). The protocol (Table I; available at www.jpeds.com)
was developed a priori, and all sonographers were trained and certified in its performance.21 Shunting across the
atrial septum was evaluated with two established methods: 2-dimensional TTE by use of agitated saline solution
contrast and agitated saline solution contrast TCD. Recordings were obtained with both normal breathing and
with a Valsalva maneuver to increase right atrial pressure and maximize detection of right-to-left shunting through
a PFO. For contrast TTE and TCD, sterile normal saline solution 5 mL was agitated between two syringes
connected by a three-way stopcock and injected by rapid bolus into a catheter inserted into the forearm. The
foramen ovale was considered patent by TTE if contrast was seen crossing the atrial septum within four cardiac
cycles after full opacification of the right atrium (Figure 1).22, 23 Contrast TCD was performed by placing a 4- or 5-
MHz probe over the temporal bone and recording flow velocity in the middle cerebral artery during injection of
agitated saline solution. Contrast TCD was considered positive for right-to-left atrial shunting if microcavitations
(transient spikes superimposed on the velocity curve) were detected within 10 seconds of agitated saline solution
injection. Shunt size was quantified according to previous standards with TCD.21 Briefly, this includes four
possible results (Figure 2): (1) no spikes (no right-to-left shunt); (2) 1 to 10 spikes (small shunt); (3) >10 individual
spikes but no shower of spikes (medium shunt); and (4) shower of spikes where individual spikes cannot be
identified (large shunt). Although a PFO is the most-frequent cause of a right-to-left shunt, other sources
(intrapulmonary, for example) may also result in a positive contrast TCD result. Therefore, contrast TTE with
direct visualization of microcavitations crossing the atrial septum was chosen a priori to be the best method to
detect PFO if TTE and TCD results were discrepant (ie, contrast TTE–negative but contrast TCD–positive). For
each study participant, the contrast TTE and contrast TCD results were interpreted by the same
echocardiographer. All echocardiographers were blinded to migraine subtype.


 View Large Image
  Download to PowerPoint
 Figure 1 

A, Contrast TTE (apical four-chamber view) showing contrast (white arrows) filling the right atrium (RA)
and right ventricle (RV). No contrast is seen in the left atrium (LA) or left ventricle (LV). B, Apical four-
chamber images from a study participant showing contrast crossing the atrial septum through a PFO
into the LA.


 View Large Image
 Download to PowerPoint
 Figure 2 

TCD recordings from study participants showing each of the four standardized results. A, No
microcavitations (no right to left shunt). B, One to 10 microcavitations (small shunt). C, >10
microcavitations but no “shower” (medium shunt). D, Shower of microcavitations (large shunt).
Microcavitations appear as bright spikes (white arrow in B) superimposed on the velocity curves.

Study subjects completed the Pediatric Migraine Disability Assessment (PedMIDAS) to determine migraine
frequency and measure migraine disability. The PedMIDAS scoring system is a validated tool to quantify the
impact of migraine on the individual’s quality of life.24 A score >30 indicates at least moderately disabling
headaches.25

Using previously published data, we estimated that the PFO prevalence in older children and adults is between
10% and 25%.4, 5, 6 For this study, we assumed the PFO prevalence in otherwise healthy children to be at the
upper end of this range. The PFO prevalence was compared between several groups: (1) all pediatric migraine
participants versus the general population; (2) pediatric migraine participants with aura versus the general
population; (3) pediatric migraine participants without aura versus the general population; and (4) pediatric
migraine participant subtypes—aura versus no aura.

As an additional contemporary control, we reviewed TTE results of consecutive age-matched children referred
during the study period for evaluation of a heart murmur and found to have no other cardiac abnormalities.
Because children with murmur do not routinely undergo intravenous catheter placement for contrast
echocardiography, only color-flow Doppler results were available to compare PFO prevalence between subjects
with migraine and these control subjects with murmur.

Statistics 
Continuous data between groups were compared by use of the Student t test or Mann Whitney rank sum test.
We used a binomial test of proportion to compare the literature control group with the PFO prevalence
demonstrated by contrast Doppler TTE in pediatric participants with migraine with and without aura. The Fisher
exact test was used to compare the PFO prevalence between groups and to compare differences in PFO
detection among the various imaging modalities. Parametric linear regression was used to compare the log-
transformed PedMIDAS scores between pediatric participants with migraine with and without aura, as well as
comparing the scores for those with and without a PFO.

Back to Article Outline

Results 
The study group (Table II) consisted of 109 children with migraine; 38 (35%) with aura and 71 (65%) without
aura. Two-thirds of pediatric migraine participants were girls. Participants with and without aura were similar in
age and sex.

Table II. Demographics of the study population

Migraine (all) (n = Aura (n = No aura (n = Control subjects with murmur (n =

109) 38) 71) 120)

Female 73 (67%) 29 (76%) 44 (62%) 46 (38%)†

Age 12.2 ± 2.9 13.0 ± 2.8 11.9 ± 3.0‡ 10.8 ± 3.1†

(years)

Weight 47.8 ± 18.1 52.3 ± 19.6 45.4 ± 17.0 42.4 ± 18.1

(kg)

P = .07, females in aura vs no aura group.

†P = .0001, females and age in control subjects with murmur vs migraine group.

‡P = .13, age in aura vs no aura group.

With contrast TTE, the prevalence of PFO in all study participants (38/109) was similar to the general population
(35% vs 25%; P = .13). In contrast (Figure 3), the subtype analysis showed that children with aura (19/38) were
significantly more likely to have a PFO than the general population (50% vs 25%; P = .0004), and the PFO
prevalence in children without aura (19/71) was the same as the estimated PFO prevalence in the general
population (27% vs 25%; P = .73). Compared with 27% of those without aura, 50% of pediatric subjects with
migraine with aura had a PFO (P = .02).


 View Large Image
 Download to PowerPoint
 Figure 3 

Prevalence of PFO in children with migraine with and without aura compared to the general population.

By color-flow Doppler interrogation of the atrial septum, 10/120 (8%) control subjects with murmur and 12/109
(11%) pediatric subjects with migraine had evidence of a PFO (P = .51). When migraine subtypes were
compared with the control subjects with murmur, a PFO was seen by color-flow Doppler scanning in 7/38
participants with aura (18% vs. 8%, P = .12) and in 5/71 without aura (7% vs 8%, P = 1.0).

With contrast TCD, a PFO was detected in 37 participants, and findings correlated well with contrast TTE (P = .
99). Only one subject with a positive contrast TTE had a negative contrast TCD result, and all subjects with a
positive contrast TCD result had a positive contrast TTE result. Compared with contrast TTE, contrast TCD had a
sensitivity of 95% (95% CI: 0.81-0.99) and a specificity of 100% (95% CI: 0.94-1.0).

Of the 37 patients with contrast TCD evidence of a PFO (Figure 4; available at www.jpeds.com), 9 had small
shunts (5 with aura, 4 without aura), 17 had medium shunts (6 with aura, 11 without aura), and 11 had large
shunts (7 with aura, 4 without aura). Shunt size did not correlate with migraine subtype.

Many subjects (37/109, 34%), regardless of migraine subtype, had moderately disabling migraines.25 The mean
PedMIDAS score was similar between participants with and without aura (30.1 ± 32.1 vs 40.6 ± 43.0,
respectively; P = .18). There was also no significant difference in the mean PedMIDAS score between patients
with a PFO versus those without a PFO (mean 32.2 vs. 36.7, respectively, P = .72).

Back to Article Outline

Discussion 
The increasing number of studies in adults describing an association between PFO and migraine with aura has
led to growing public and physician interest in the use of PFO closure as a therapy for migraines, particularly for
those individuals for whom medical management has failed.13, 14, 15, 16, 17, 18, 19 Despite the absence of rigorous
evidence supporting PFO closure as a safe and effective treatment for migraines, increasing numbers of children
are being referred to our institution for PFO evaluation and closure. This study demonstrates a significantly
higher PFO prevalence in children with migraine with aura compared with the general population and thus
provides a basis for further research on the role of PFO in pediatric migraines.

Independent reports of migraine cessation or improvement in adults after PFO closure for non-migraine
indications, such as cryptogenic stroke or decompression illness, led to the hypothesis that a PFO may permit
vasoactive metabolic or microembolic triggers to enter the cerebral circulation directly, bypassing the pulmonary
circulation where they are normally filtered or metabolized into an inactive form.26, 27 This hypothesis is supported
by reports of a higher prevalence of PFO in adults with migraine with aura. In addition, several single-center PFO
closure studies have demonstrated treatment success rates, defined as migraine cessation or reduction in
migraine frequency and severity after PFO device closure, as high as 70% to 100%.15, 16, 17, 18, 19, 28 However,
the Migraine Intervention with STARFlex Technology study, a prospective, blinded (sham device control arm),
randomized trial that evaluated the effectiveness of PFO closure on headache relief in adults with migraine with
aura failed to show that PFO closure was an effective treatment option.13 The Migraine Intervention with
STARFlex Technology study cannot be considered conclusive, however, because of technical issues and
potential design flaws involving patient selection and study endpoints.13, 29, 30 Results from our study of children
ages 6 to 18 years are similar to those reported in adults, showing that the prevalence of PFO in migraine with
aura is roughly double that found in migraine without aura and suggest that spontaneous closure of a PFO is
unlikely after 6 years of age.

Some adult studies have shown that those with migraine with aura have larger right-to-left atrial shunts than
those without aura.10 Although contrast TCD data from our study showed a trend toward larger right-to-left atrial
shunting in migraine with aura, it was underpowered to identify a significant difference. We estimate that a
sample size of nearly 500 is needed to detect a significant difference in the degree of atrial shunting between
pediatric patients with migraine with and without aura.

The role of TCD in children with migraine headache is not completely clear. Contrast TCD findings correlated well
with contrast TTE results (sensitivity and specificity ≥95%) and TCD has an advantage in allowing quantification
of the amount of right-to-left atrial shunting; however, the importance of this information in the pediatric
population remains unknown. Contrast TCD imaging may have a primary role in PFO detection in children with
suboptimal TTE windows obviating the need for transesophageal imaging and the risks associated with this
technique in children. TCD could be incorporated in future trials that investigate the impact of the right-to-left
shunt size on pediatric migraine headache.

This study had several limitations. Because of ethical issues and concerns regarding recruitment, we were not
able to place intravenous lines in a healthy control group to perform contrast TTE and TCD. Consequently, we
compared our data to the upper bound estimate of PFO prevalence in the general population obtained from
previously published data.31, 32, 33, 34, 35 We also attempted to address this issue of a contemporary healthy
control group by use of color-flow Doppler scanning results in control subjects with murmur. This comparison
revealed a trend toward a higher PFO prevalence in patients with migraine with aura, but because color-flow
Doppler scanning is less sensitive than contrast TTE, it did not reach statistical significance.

Because our results are similar to those reported in most adult studies, spontaneous PFO closure must be rare
after 6 years of age. Furthermore, if a causal link is established between PFO and migraine, routine investigation
for PFO with contrast TTE may be indicated for all children with migraine headaches.

Back to Article Outline

Appendix 
 
 View Large Image
 Download to PowerPoint
 Figure 4 

Percentage of subjects with varying sizes of right-to-left shunt as determined by contrast TCD in
migraine with and without aura.

Table I. Protocol for agitated saline TTE/TCD imaging

• Placement of an intravenous catheter into a forearm vein.

• All participants are coached on how to perform a Valsalva maneuver by exhaling air against a closed glottis to
increase right atrial pressure.

• The standard 2-dimensional and Doppler echocardiography is performed by experienced pediatric sonographers
trained in this protocol.

• While obtaining an image of the heart in the apical four-chamber view, a cardiologist experienced in this protocol
infuses contrast using sterile normal saline solution 5 mL that has been agitated between two syringes connected by
a three-way stopcock. One injection is performed with the subject resting and breathing normally. A second injection
is performed while the patient is performing a Valsalva maneuver. The Valsalva maneuver is terminated suddenly
when contrast medium arrives in the right atrium. Residual atrial shunting is present if microcavitations cross the
septum within four cardiac cycles after full opacification of the right atrium.

• A TCD probe is placed over the middle cerebral artery. Agitated saline solution is injected during Doppler recording
in this vessel. Microcavitations detected within 10 seconds of the injection appear as transient spikes in the velocity
curve. Injections are recorded at rest and with Valsalva maneuver.

• Four possible results have been standardized for the TCD recordings:

No microcavitations (no right to left shunt)

1-10 microcavitations (small shunt)

>10 microcavitations but no shower of microcavitations where a single microcavitation cannot be identified
(medium shunt)

• Shower of microcavitations where individual microcavitations cannot be counted (large shunt)