pacemaker in complicated and refractory breath-holding spells: when to think about it?

$: Pacemaker in complicated and refractory breath-holding spells: When to think about it?$
www.elsevier.com/locate/braindev

Brain & Development xxx (2014) xxx–xxx

Review article

Pacemaker in complicated and refractory breath-holding spells:When to think about it?

Stefano Sartori a,⇑,1, Margherita Nosadini a,1, Loira Leoni b, Luca de Palma a,Irene Toldo a, Ornella Milanesi c, Alessia Cerutti c, Agnese Suppiej a

a Pediatric Neurology Unit, Division of Pediatrics, University of Padua, Padua, Italyb Cardiology Division, University of Padua, Padua, Italy

c Pediatric Cardiology Unit, Division of Pediatrics, University of Padua, Padua, Italy

Received 15 July 2013; received in revised form 2 February 2014; accepted 5 February 2014

Abstract

Background: Breath-holding spells (BHS) are benign non-epileptic paroxysmal events of infancy, rarely occurring with high fre-quency and complicated by prolonged syncope, convulsions and even status epilepticus. In these cases response to medical treatmentis often unsatisfactory. Pacemaker implantation is a possible therapeutic option, but its indications, efficacy and complications havenot been clarified yet. Objective: To report a new case of BHS treated with pacemaker and to review its indications and efficacy inpatients with severe BHS. Methods: We extensively searched the literature in PubMed on cardiac pacing in patients with BHS andwe described a new case. Results: A previously healthy boy presented at the age of 4 months with frequent BHS inconstantly asso-ciated to prolonged syncope and post-anoxic non-epileptic and epileptic seizures. Parental reassurance, iron supplementation andpiracetam were ineffective. After cardiac pacing at the age of 16 months, BHS and their complications disappeared. We identified47 patients with BHS treated with pacemaker in the literature. Based on the available data, in all patients asystole or marked bra-dycardia were documented during BHS or stimulating maneuvers; syncope complicated BHS in 100% of cases and post-anoxic con-vulsions in 78.3%. Medical treatment before pacing, when administered, was ineffective or poorly tolerated. After pacing, BHScomplications disappeared in 86.4% of cases, and decreased in 13.6%. Technical problems with the device were reported in25.7% of patients and mild medical complications in 11.4%. Conclusions: Pacemaker could be reasonably considered in subjects withfrequent and severe BHS, poor response to medications, and demonstration of cardioinhibition during spells.� 2014 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.

Keywords: Breath-holding spells; Pacemaker; Cardiac pacing; Asystole; Syncope; Reflex anoxic seizures; Post-anoxic convulsions; Post-anoxicconvulsive epileptic seizures

1. Introduction

Breath-holding spells (BHS) are well-known benignnon-epileptic paroxysmal events occurring in about

http://dx.doi.org/10.1016/j.braindev.2014.02.004

0387-7604/� 2014 The Japanese Society of Child Neurology. Published by E

⇑ Corresponding author. Address: Pediatric Neurology Unit,Department of Pediatrics, University of Padua, Via Giustiniani 3,35128 Padova, Italy. Tel.: +39 049 821 8094; fax: +39 049 8215430.

E-mail address: [email protected] (S. Sartori).1 Stefano Sartori and Margherita Nosadini are to be considered first

authors; they contributed equally to this work.

Please cite this article in press as: Sartori S et al. Pacemaker in complicatedDev (2014), http://dx.doi.org/10.1016/j.braindev.2014.02.004

0.1–4.6% of healthy children with onset generallybetween 6 and 18 months of age [1]. BHS are usuallyelicited by provocation, pain or frustration and are clin-ically characterized by inconstant cry, apnea, change inskin color (cyanosis or pallor), brief loss of conscious-ness and of postural tone, with spontaneous resolution[1,2]. Iron-deficiency anemia [3–6] and a possibleunderlying genetically determined autonomic dysfunc-tion [7–12] have been hypothesized in their pathogenesis.

lsevier B.V. All rights reserved.

and refractory breath-holding spells: When to think about it?. Brain


mailto:[email protected]



2 S. Sartori et al. / Brain & Development xxx (2014) xxx–xxx

The prognosis is excellent, with disappearance aroundschool age and normal neurologic development [7].

Despite their benign nature, BHS may be compli-cated by prolonged loss of consciousness and in 15%of cases by the occurrence of non-epileptic post-anoxicconvulsions as a consequence of relative cerebral ische-mia or hypoxia due to marked bradycardia or asystole[7,13]. In rare cases, a prolonged and severe cerebralhypoxia may also precipitate true epileptic post-anoxicconvulsive events and even status epilepticus [14,15].

Given their benign nature and favorable outcome,parental reassurance and counseling are the mainstayof therapy in non-complicated forms [7]. By contrast,in frequent or severe BHS several medications such asiron, piracetam, atropine, antiepileptic drugs and fluox-etine, have been used with variable results [3–6,16–20].In rare complicated cases, refractory to parental coun-seling and medical treatment, pacemaker (PM) implan-tation has been considered a therapeutic option in thelast two decades. Unfortunately, clinical indications,efficacy and complications of cardiac pacing have notbeen systematically studied in the population of childrenaffected by complicated BHS yet.

The aim of the present study was to evaluate, throughliterature review and analysis of a personal case, theclinical indications to PM implantation in BHS, its effi-cacy and complications. The reported case is docu-mented with video, EEG and ECG recordings.

2. Methods

The literature search was conducted through MED-LINE, using the following search term combinationswithout additional filters: “pacemaker” [All Fields]AND “breath-holding spells” [All Fields]; “cardiac pac-ing” [All Fields] AND “breath-holding spells” [AllFields]; “pacemaker” [All Fields] AND “reflex anoxicseizures” [All Fields]; “cardiac pacing” [All Fields]AND “reflex anoxic seizures” [All Fields].

We searched in all the identified cases a comprehen-sive set of clinical and instrumental data including gen-der, type of BHS (cyanotic, pallid or mixed), age atonset, frequency and severity of the episodes (occurrenceof bradycardia, asystole, loss of consciousness, non-epi-leptic and epileptic convulsions), medical treatment, ageat PM implantation, type of PM, follow-up, efficacy ofthe pacemaker on BHS and on their complications,and PM-related complications. When not available inthe articles, we marked as “NR” the data not reportedby the authors. In some reports information was avail-able for a group of cases with no specification for eachpatient.

The classification of BHS type was based on the inter-pretation of the authors or on the description of the epi-sodes in the original articles. The age at BHS onset andat PM implantation was reported as specified in the ori-


ginal papers, but in cases where the age could only beinferred by the text, it was reported at the best of ourinterpretation and marked with the symbol “�“. Weregarded as clinical BHS complications episodes of lossof consciousness (LOC) and post-BHS convulsions; thelatter were subspecified as anoxic non-epileptic seizures(ANES), anoxic epileptic seizures (AES) or status epi-lepticus (SE).

In order to classify the epileptic or non-epileptic nat-ure of post-BHS convulsive phenomena, we consideredas epileptic only the convulsive phenomena interpretedas such by the authors in the original papers, or the casesconfirmed by ictal EEG. Criteria for classifying post-BHS motor phenomena as epileptic seizures were notspecified in the original papers except for Horrocksand coworkers [15]. According to the currently acceptedoperative definition of status epilepticus [21], weregarded and listed as status epilepticus the epileptic sei-zures longer than 5 min.

Maximum length of asystole or bradycardia werereported by the authors in different clinical settings: dur-ing BHS, oculocardiac response, tilt testing, or cardiacmonitoring. We searched for medications used prior tocardiac pacing, their efficacy and side effects, when avail-able. The type of PM implanted (epicardial vs. endocar-dial; VVI vs. DDD) and the length of follow-up werealso reported, if available in the original papers. Basedon the efficacy of PM implantation, we identified the fol-lowing outcome groups: BHS persistence, with cessationof BHS complications (A1); BHS persistence, withdecreased BHS complications (A2); disappearance ofBHS and of their complications (B1); decrease of BHSand disappearance of their complications (B2); notreported (NR).

3. Results

3.1. Case report

A previously healthy boy with family history of BHS(mother and maternal grandfather) presented at the ageof 4 months with BHS occurring after mild provocation,such as undesired postural changes: the child cried vig-orously, with subsequent apnea, marked pallor with per-ioral cyanosis and inconstant brief loss of consciousness.The frequency at onset ranged between 2 and 8 episodesmonthly (most often 2 per month, with higher frequencyobserved during febrile infective illnesses).

Since the age of 12 months BHS frequency increasedup to 2–3 per day, with lowering triggering threshold;beside the usual spells, parents reported also some epi-sodes of BHS followed by a prolonged phase (up to15 min) described as characterized by decreased con-sciousness and sometimes purposeless limbs movementsand inconstant trismus followed by subsequent deep andprotracted sleep.



S. Sartori et al. / Brain & Development xxx (2014) xxx–xxx 3

At age of 16 months the boy underwent neurologicalexamination by a pediatric neurologist at our Depart-ment. The evaluation of a home video recordingsshowed the boy during one of the major events reportedby the parents immediately after a BHS triggered by atrivial scolding. The video documented a prolonged epi-leptic convulsive seizure: the boy appeared unconsciouswith upper limbs rhythmic clonic movements, upwardeye deviation, guttural sounds (Video 1). The episodelasted about 10 min and was followed by a protracteddeep sleep. Neurological examination was normal,except for mildly oppositional behavior. EEG, routineECG and echocardiography were normal; blood testsonly revealed mild iron-deficiency anemia. The boywas admitted to our Department and a continuousEEG videopolygraphic monitoring was set, soon allow-ing to capture a complicated BHS characterized insequence by cry, apnea, initial opisthotonus, subsequentgeneralized limpness, and abrupt loss of consciousnesslasting about 2 min, associated with an isolated myo-clonic movement involving upper limbs (Video 2). Sub-sequently, the child started to cry again, opened his eyesand finally went down into a deep sleep. The simulta-neous ECG tracing recorded sinus rhythm interruptedby a pause of 22 s (during the cry–apnea–syncope–myo-clonia event) interrupted by 2 escape junctional beats(Figs. 1 and 2). On the EEG, hypersynchronousbackground slowing of activity was followed by diffusebackground electrical decrementation for 15 s andhigh-amplitude symmetrical rhythmic discharges onthe anterior regions of the scalp for about 8 s.

Fig. 1. ECG recording during a complicated BHS characterized by cry, apneaof consciousness lasting about 2 min, associated with an isolated myoclondocuments sinus rhythm interrupted by a pause of 22 s interrupted by 2 esc


Background activity slowing ensued, followed by renor-malization (Fig. 2). Further similar episodes occurred inthe following days, occasionally complicated by pro-longed convulsive epileptic seizures that requiredsedation with benzodiazepines; they couldn’t berecorded due to the marked opposition of the boy andthe anxiety of the family, that didn’t allow to protractthe EEG–ECG monitoring.

The off-line time domain analysis of the heart ratevariability, performed over a 24-h period, demonstratedthe following results: Standard Deviation of Normal-to-Normal intervals (SDNN) = 78 ms; Standard Deviationof the Successive Differences between adjacent R-Rintervals (SDSD) = 84 ms; number of pairs of adjacentNN intervals differing by more than 50 ms divided bytotal number of NNs (pNN50) = 8%; Root MeanSquare of the Successive Difference between adjacentR-R intervals (rMSSD) = 36 ms. The off-line frequencydomain analysis of the heart rate variability, performedwith the power spectral analysis of 5-min ECG record-ings, showed a LF/HF ratio of 2,75 and 0,28, respec-tively in quiet wakefulness at rest and during sleep.

Reassurance, behavioral and educational counselingto the parents and the boy, iron therapy for 3 weeksand a trial with piracetam up to the dosage of 800 mgdaily (80 mg/kg/day) yielded no improvement.

Considering the severity and the high frequency ofthe events, the occurrence of prolonged syncope andconvulsive status epilepticus, the parental anxiety, theunlikelihood of a spontaneous and short-term resolutionof the episodes given the early age of the boy, the patient

, initial opisthotonus, subsequent generalized limpness, and abrupt lossic movement involving upper limbs. The simultaneous ECG tracingape junctional beats.



Fig. 2. ECG–EEG recording during a complicated BHS characterized by cry, apnea, initial opisthotonus, subsequent generalized limpness, andabrupt loss of consciousness lasting about 2 min, associated with an isolated myoclonic movement involving upper limbs. The simultaneous ECGdocuments sinus rhythm interrupted by a pause of 22 s interrupted by 2 escape junctional beats. On the EEG, hypersynchronous background slowingof activity is followed by diffuse background electrical decrement for 15 s. Background activity slowing ensued, followed by renormalization (EEG:10 uV/mm, 1 min/page).


was referred for cardiac pacing. An epicardial singlechamber PM (VVI) was inserted (Medtronc Sensia DRepicardio), set at a back up rate of 80 beats per minutewith hysteresis at 40 beats per minute. Since the day ofPM implantation and activation, the child has had nomore events: neither major fits complicated by syncope,post-anoxic convulsions, or epileptic seizures, nor simpleuncomplicated BHS. In the following months, PM activ-ity recording showed intervention in 2% of time; sponta-neous mean HR was 95 beats per minute. The devicewas monitored every 3 months by Home monitor Med-tronic Care Link. At a 26-month neurologic and cardio-logic follow-up, the boy shows a normal psychomotordevelopment, leads a normal life with no limitations,and the family’s quality of life is excellent.

3.2. Literature review

The review of the existing literature led to the identi-fication of 47 children with BHS treated with PMimplantation, reported in 14 distinct papers publishedbetween 1996 and 2013 [15–17,22–32]. In two of the arti-cles found in the literature [22,24], an additional previ-ous case of PM implantation is mentioned; despiteefforts, though, this paper was impossible to track inPubmed, and therefore we didn’t take it into account.The clinical and demographic features, data regardingPM implantation and effectiveness of the treatment ofthe 47 patients whose data were available for analysis,and including our personal case (total 48 cases), aredetailed in Table 1.


Female gender was prevalent (29/48 patients, 60.4%).Only in 1 case an underlying neurological disorder (men-tal retardation and epilepsy) was reported [31]. Dataregarding age at BHS onset were available for 34/48patients; this ranged between 3 days and 3 years.Reported frequency of the spells ranged between 3 permonth and 40 per day; it was not reported in 27/48.According to available data, 27/47 (57.4%) patientshad pallid BHS, 2/47 (4.3%) had cyanotic spells and18/47 (38.3%) had mixed type spells (type of spells notspecified in 1/48 patients). Syncope complicated BHSin 47/47 (100%) patients (data not reported in 1/48patients) and convulsive phenomena in 36/46 (78.3%)(data not reported in 2/48 patients). Epileptic convul-sions or status epilepticus were recognized and/orreported in 10/36 patients reported to have convulsivephenomena.

Medical treatment of BHS was tried prior to cardiacpacing in 27/30 (90%) but was mostly ineffective orpoorly tolerated (no treatment tried in 3/30, 10%; datanot reported in 18/48 patients); among most commonlyused medications were antiepileptic drugs (carbamaze-pine, valproate, phenobarbital, levetiracetam, topira-mate, phenytoin, clonazepam), atropine, theophylline,fluoxetine, primidone, piracetam, and ironsupplementation.

When done, echocardiography, interictal EEG andneuroimaging were normal. In all patients, at restECG was normal. Extreme bradychardia or cardiacpauses were recorded in 47/47 (100%) patients duringspontaneous BHS or provocation tests (data not



Table 1Clinical and instrumental data of the patients with breath-holding spells (BHS) treated with pacemaker (PM) implantation reported in the literature and present case, and summary of the data. A1:BHS persistence; cessation of BHS complications. A2: BHS persistence; decreased BHS complications. ANES: anoxic non-epileptic seizures. B1: Disappearance of BHS and of their complications.B2: Decrease of BHS and disappearance of their complications. BHS: breath-holding spells. Bradyc.: bradycardia. C: cyanotic BHS. CBZ: carbamazepine. Compl.: complications. d: day(s). DDD:dual-chamber pacing and sensing with either triggered or inhibiting pace output [25]. ECG electrocardiogram. EEG electroencephalogram. Endoc.: endocardial. Epic.: epicardial. EOL: end of life.F-U: follow-up. Freq.: frequency. HR: heart rate. Ineff.: ineffective. LOC: loss of consciousness. LVT: levetiracetam ES: epileptic seizures. m: month(s). Not avail.: not available. NR: not reported.OCR: oculocardiac reflex. P: pallid BHS. PB: Phenobarbital. PHT: phenytoin. PM: pacemaker. SE: status epilepticus (epileptic activity > 5 min). SEf: side effects. Sev.: several. TPR topiramate. Tx:treatment. VPA: valproate. VVI: ventricular pacing and sensing only, with inhibition (if endogenous ventricular activity occurs) [25]. w: week(s). y: year(s). �: about. �: three of the six patients withconvulsive phenomena received AED for the finding of bouffees of spike waves at the EEG.

Sex TypeofBHS

Age atBHSonset

MaxBHSfreq.

Complications of BHS Max recordedasystole orbradycardia

Medical tx priorto PM implant

Age at PMimplant

Type of PM Results(F-U)

PM-compl.LOC Post-BHS

convulsiveepisodes

ANESAESSE

Breningstall[22]

(1 case)

M C 3 d 20-30/d Yes NR NRNoNo

Bradicardia50 beats/min(during BHS)

CBZ (ineff.; SEf) NR (�2.5 y) Endocardial A1 (NR) NR

Sreeram andWhitehouse[23]

(1 case)

F M Infancy 10-40/d Yes Yes ANESAESNR

Asystole 24 s(during BHS)

Atropine(ineff.; SEf)

NR (�3 y) Endocardial,VVI

A1 (1 y) NR

McLeodet al. [24]

(12 cases)

8 M4 F

P NR NR Yes Yes ANESNRNR


NR 4.1 y Endocardial,DDD-VVI

Withdrawnfrom the study

Yes




A1(symptomaticresponse) (8 m)

Yes




A1 (8 m) 1 Yes8 No




A1(symptomaticresponse)(8 m)




A1(symptomaticresponse)v(8 m)




A1 (8 m)




A1 (8 m)




A2 (8 m)

P NR NR Yes Yes ANESNR


NR 2 y Endocardial,DDD-VVI

A1 (8 m)

S.

Sa

rtori

eta

l./Bra

in&

Develo

pm

ent

xx

x(

20

14

)x

xx

–x

xx

5

Please

citeth

isarticle

inp

ressas:

Sarto

riS

etal.

Pacem

aker

inco

mp

licatedan

drefracto

ryb

reath-h

old

ing

spells:

Wh

ento

thin

kab

ou

tit?.

Brain

Dev

(2014),h

ttp://d

x.do

i.org/10.1016/j.b

raind

ev.2014.02.004


NRP NR NR Yes Yes ANES

NRNR



A2 (8 m)




A1 (8 m)




A1 (8 m) Yes

Villain et al.[25]

(11cases)

4 M7 F

P 6 m Sev./d(8/11)NR (3/11)

Yes No NoNoNo

Asystole 6-25 s orseveral brief pauses(2-3 s) causingmarkedbradychardia

(during OCR in 11/11)Asystole12-25 s in 4/11

(during Holter ECG)Bradycardia<30 beats/min in 3/11

(during Holter ECG)Sinus arrhythmia in4/11

(during Holter ECG)

Diphemanil (11/11) Propranololor Pindolol (3/11)AED (3/11)Disopyramide (1/11)(all ineff.)

3.5 y Epicardial, VVI A1 (7 y) No

M <1 m Yes No NoNoNo

2.8 y Epicardial, VVI A2 (9 y) Yes

P 2 m Yes No NoNoNo

2 y Epicardial, VVI A1 (2 y) No

M 2 y Yes Yes ANESNR�NR

5.5 y Epicardial, VVI B1 (8 y) No

P 3 y Yes Yes ANESNR�NR

5.5 y Epicardial, VVI A1 (9 y) Yes

M 7 m Yes Yes ANESNR�NR

2.5 y Endocardial,VVI

B1 (0.9 y) No

M 4 m Yes No NoNoNo

2 y Endocardial,atrial pacing

B1 (10 y) Yes

P 1 m Yes No NoNoNo

3 y Epicardial, VVI A2 (14 y) Yes

M 7 d Yes Yes ANESNR�NR

5 y Epicardial, VVI A1 (10 y) No

P 2 m Yes Yes ANESNR�NR

4 y Endocardial,VVI

B1 (3 y) No

P 6 m Yes Yes ANESNR�NR

1.7 y Endocardial,VVI

A1 (8 y) No

Kelly et al.[26]

(10cases)

M M 2 w NR Yes Yes ANESNRNR

Asystole15 s(during BHS)

PBCZPPHTPrimidone

1 y 9 endocardial,VVI1 epicardial,VVI

B1 (14 y) Yes

F M 6 m NR Yes No NoNoNo

Asystole 1.7 s(during BHS)

NR 1 y B1 (11 y) No

6S

.S

arto

riet

al./B

rain

&D

evelop

men

tx

xx

(2

01

4)

xx

x–

xx

x

Please

citeth

isarticle

inp

ressas:

Sarto

riS

etal.

Pacem

aker

inco

mp

licatedan

drefracto

ryb

reath-h

old

ing

spells:

Wh

ento

thin

kab

ou

tit?.

Brain

Dev

(2014),h

ttp://d

x.do

i.org/10.1016/j.b

raind

ev.2014.02.004




PropanthelineTheophylline(minorimprovement)

10 m A1 (10 y) Yes

F P 4 m NR Yes No NoNoNo


PropanthelineBelladonnaPB

4 y A1 (8 y) Yes

F(sister ofthefollowingcase)

M 4 m NR Yes Yes ANESAESSE


PB(persistence ofBHS, stopconvulsions; SEf)CBZ(persistence ofBHS withcyanosis andLOC, stopconvulsions)DZP

5 y B1 (5.5 y) No

F(sister oftheprecedingcase)

M 12 m NR Yes Yes ANESNRNR


BelladonnaPB

16 m A1 (5 y) No

F P 12 m NR Yes Yes ANESNRNR


Theophylline 19 m B1 (5 y) Yes



None 13 m B1 (3.5 y) No

M M 7 m NR Yes Yes ANESNRNR


Atropine 1 y A2 (3.2 y) No

M M 6 d NR Yes Yes ANESNRNR

Bradycardia36 beats/min(during BHS)

CBZ 2.4 y A2 (3.2 y) No

Legge et al.[27]

(2 cases)

F M 11 m �2/w Yes Yes ANESNRNR


IronPiracetamAtropine(all ineff.)

2.4 y Epicardial, VVI A1 (1 y) NR

F M 8 m Sev./w Yes Yes ANESAESNo


None 11 m Epicardial, VVI A1 (6 m) NR

Chehabet al. [28]

(1 case)

F M 6 m 10-15/d Yes No NoNoNo


NR 18 m Epicardial, VVI A1 (NR) NR

Wilson et al.[29]

(1 case)

F P 13 m 2-3/w Yes Yes ANESNRNo


NR NR (�19 m) Epicardial,DDD

A1 (1 y) Yes

S.

Sa

rtori

eta

l./Bra

in&

Develo

pm

ent

xx

x(

20

14

)x

xx

–x

xx

7

Please

citeth

isarticle

inp

ressas:

Sarto

riS

etal.

Pacem

aker

inco

mp

licatedan

drefracto

ryb

reath-h

old

ing

spells:

Wh

ento

thin

kab

ou

tit?.

Brain

Dev

(2014),h

ttp://d

x.do

i.org/10.1016/j.b

raind

ev.2014.02.004


Horrockset al. [15]

(2 cases)

F P 42 m 3/m Yes Yes ANESAESSE

Asystole 30 s(during tilt testing)

None NR NR A1 (NR) NR

F P 6 d 4/d Yes Yes ANESAESSE

Asystole 20 s(cardiac monitoring)

DZP (ineff.)ParaldehydeVPA (effective)

NR NR A1 (NR) NR

Di Pinoet al. [30]

(1 case)

F P 4 m Almostdaily

Yes Yes ANESNRNR


NR NR �14 m Epicardial, VVI? A1 (26 m) NR

Lukkarinenet al. [16]

(1 case)

F P NR 1-3/d Yes Yes ANESNRNR


AtropinePropranolol(ineff.)LVT(effective; SEf)

NR � 12 m NR NR NR

Ergul et al.[31]

(2 cases)

M(brotherof thefollowingcase)

C 2 m 1-2/d Yes Yes ANESNRNR


NR NR � 3 m Epicardial B2 (NR) NR

M(brotherof theprecedingcase)

NR NR NR NR NR NRNRNR

NR NR NR2.5 y (epilepsy andmentalretardationpreexistent to PMimplant)

NR NR NR

Walsh et al.[17]

(1 case)

F P 4 m NR Yes Yes ANESAESNR

Asystole;bradichardia inducedtorsade de pointes(during BHS)

PBTPRFluoxetine(all ineff.)

NR(�2 y)

NR NR NR

de Almeidaet al. [32]

(1 case)

F P 8 m Sev./d Yes Yes ANESNRNR


Atropine(partial efficacy)Beta-blocker(ineff.)

NR(�12 m)

Epicardial,DDD

B1 (NR) NR

Present case M M 4 m 2–3 d Yes Yes ANESAESSE


IronPiracetam(all ineff.)

19 m Epicardial, VVI B1(1.5 y) No

Total 48cases(14articles +present case)

Sex

19 M29 F

BHStype

27 P2 C18 M1 NR

BHSonset

34Range3d–3y14 NR

BHSfreq.21Range3 m –40 d27 NR

LOC47Yes1NR

Post BHSmotorphenomena36 Yes10 No2 NR

ANES36 Yes10 No2 NRAES10 Yes11 No27 NR

SE4 Yes13 No31 NR

Recorded asystole orbradycardia duringBHS, OCR orcardiac monitoring45 Asystole (range1.7-70s)5 Bradyc.4 Sinus arrhythmia1 NR

Medical tx priorto PM implant27 Yes3 No18 NR

Age at PMimplant45 Range 3 m –14.7 y2 NR

Type of PM27 Endoc.16 Epic.5 NR26 VVI2 DDD12 VVI-DDD1 atrial7 NR

Results26 A16 A211 B11 B23 NR1 withdrawnfrom the studyF-U40 Range 6 m–14 y8 NR

PM-relatedcompl.13 Yes22 No13 NR

8S

.S

arto

riet

al./B

rain

&D

evelop

men

tx

xx

(2

01

4)

xx

x–

xx

x

Please

citeth

isarticle

inp

ressas:

Sarto

riS

etal.

Pacem

aker

inco

mp

licatedan

drefracto

ryb

reath-h

old

ing

spells:

Wh

ento

thin

kab

ou

tit?.

Brain

Dev

(2014),h

ttp://d

x.do

i.org/10.1016/j.b

raind

ev.2014.02.004



reported in 1/48 patients); according to the availableinformation, recorded cardiac pauses ranged between1.7 and 70 s.

Age at PM implantation ranged between 3 monthsand 14 years. An endocardial PM was implanted in27/43 patients (62.8%), and an epicardial PM in 16/43(37.2%) (type of PM implanted not specified in 5/48patients).

Data regarding outcome after cardiac pacing wereavailable for 44 patients. Of these, at a reported followup ranging between 6 months and 14 years, completedisappearance of BHS complications was reported in38/44 of cases (86.4%), and their decrease in the remain-ing 6/44 (13.6%). Among the 38 in whom BHS compli-cations disappeared after pacing, complete resolution ofBHS episodes was also seen in 11, and their decrease in1.

Data about complications of PM implantation wereavailable for 35 patients. Nine patients (25.7%) hadtechnical problems in the device (dysfunction of stimula-tion threshold, battery end of life, exteriorization of thedevice, failure to capture, lead retraction or displace-ment, inadequate atrial sensing). Of these, the devicewas removed in 1 and replaced in 2. Four additionalpatients (11.4%) experienced PM-related medical com-plications: 2 reported poor tolerance to the device (diz-ziness and awareness of the PM), 1 needed PMremoval due to infection at the generator site and 1due to the complaint of intestinal symptoms; interest-ingly, in this latter case the device was reimplanted4 months later for recurrence of spells [26].

4. Discussion

Breath-holding spells (BHS) are well-known benignnon-epileptic paroxysmal events occurring in childhood.According to the change in skin color during the event,BHS have been traditionally divided into cyanotic andpallid. These two clinical forms were originally thoughtto be due to different physiopathological sequences bothleading to cerebral hypoxia/ischemia and subsequentloss of consciousness and of postural tone: in cyanoticBHS, initial violent crying has been advocated to causehypocapnia and consequent forced expiratory apnealeading to hypoventilatory cerebral and systemic hypox-emia (cyanosis), worsened by the possible simultaneousoccurrence of other phenomena (decreased venousreturn and right ventricular cardiac output, adrenergic-and hypoxia-mediated pulmonary arterial vasoconstric-tion, and putative intrapulmonary right to left bloodshunting); by contrast, a marked bradycardia (withinconstant concomitant vasoparalysis) due to pain- ordistress-induced excessive vagal discharge, with subse-quent low cardiac output, has been considered the pri-mum movens in the physiopathogenesis of pallid BHS[22], that some authors also call “reflex anoxic syncope”


or “reflex anoxic seizure” [15,29,32–34]. At present it isacknowledged that this distinction should be less rigidand that the two sequences might overlap in mixed BHS.

Several authors suggested the existence of an underly-ing genetically determined autonomic dysfunction inpatients with BHS. Interestingly, though, the excessivevagal tone (parasympathetic over-activity) hypothesizedespecially in the pallid BHS [7–12] was often establishedbased on clinical and neurophysiologic tests, such asoculocardiac reflex, that actually measure instantaneousvagal over-reactivity rather than tonic vagal over-activity

[35]. Conversely, as demonstrated by our patient, a nor-mal vagal tone doesn’t rule out the possible occurrenceof abnormal vagal reactions causing symptomatic bra-dycardia or asystole, considering the absence of a directrelationship between cardiac vagal tone and phasic car-diac vagal reactivity [35].

Despite their usual benign nature, rarely BHS may becomplicated by prolonged loss of consciousness andnon-epileptic and epileptic post-anoxic convulsionssometimes leading to status epilepticus, with a potentialrisk for secondary brain injury and further life-threaten-ing cardiorespiratory compromise. In light of this, in thelast two decades PM implantation has been used as atherapeutic option in patients with complicated BHSunresponsive to medical treatment. To the best of ourknowledge, 48 patients have been reported to haveundergone PM implantation, taking into account ourpersonal case (Table 1) [15–17,22–32]. Most of thepatients paced had pallid BHS, and all had severeBHS, variably identified by the authors as spells associ-ated to at least one of the following phenomena: markedbradycardia or prolonged asystole; prolonged loss ofconsciousness; protracted generalized stiffness, opisthot-onus, clonic, myoclonic or tonic–clonic jerks; epilepticseizures or status epilepticus. In particular, BHS-induced syncope was present in all patients undergoingPM implantation (100%), and convulsive phenomenain most (78.3%). The severity and the frequency of thespells (up to 40/day in some cases) was reported as suchas to affect the patient’s and the family’s quality of life.None of the patients was reported to suffer from cardio-logic conditions. At rest ECG and echocardiogram werealways normal, but evidence of pathological cardioin-hibitory events (prolonged asystole or prolongedmarked bradycardia) clinically associated to spontane-ous BHS or provoked syncopes were documented inall patients on ECG monitoring. PM proved effectivein all the cases, with complete disappearance of BHScomplications in 86.4% of patients and their reductionin the remaining, providing further evidence to thehypothesis that reflex cardioinhibition played a pivotalrole in BHS physiopathogenesis in the reportedimplanted cases, as in our patient.

PM has been an established intervention in childrenwith cardiac rhythm disorders since the late 1960s, both




for controlling symptoms and for providing a life-sus-taining cardiac rhythm. In the light of the present liter-ature review, cardiac pacing seems effective also inpreventing syncope and post anoxic convulsions in thespecific subset of patients with complicated and drugrefractory BHS in whom cardioinhibition is demon-strated to play a major physiopathological role, regard-less of the change in skin color [8,9,12]. In these cases,PM may activate and pace at a back-up rate wheneverBHS cause a sudden drop in heart rate below an age-dependent pre-determined value.

Even though definitive criteria for PM implantationin this condition cannot be established, based on litera-ture review and physiopathological rationale, it is possi-ble to suggest that cardiac pacing could reasonably beconsidered in subjects with BHS when all of the follow-ing features are documented: (1) frequent and severe epi-sodes, that is spells complicated by prolonged syncopewith subsequent lethargy and/or post-anoxic non-epi-leptic or epileptic prolonged convulsions; (2) poorresponse to medical treatment, family counseling andpsychological support, especially in case of marked dif-ficulty coping with the spells or managing the child ineveryday life; (3) demonstration of cardioinhibition onECG monitoring during at least one spontaneous orprovoked BHS episode, in the absence of clinical datasuggesting a pivotal role of primary hypotensive phe-nomena (vasoparalysis) as main causal factor of thebrain ischemia. In fact, demonstrating a consistent andprimitive hypotension during the complicated BHSshould suggest caution in considering PM implantationin these patients, because this device acts on cardiacrhythm but not on vessel tone.

In conclusion, when BHS are complicated by pro-longed loss of consciousness, post-anoxic non-epilepticand/or epileptic convulsions as a consequence of relativebrain ischemia or hypoxia due to a severe and pro-tracted cardioinhibition, PM should be considered asan effective intervention to prevent potential life-threat-ening events and/or neurological consequences.

Appendix A. Supplementary data

Supplementary data associated with this article canbe found, in the online version, at http://dx.doi.org/10.1016/j.braindev.2014.02.004.

References

[1] Lombroso CT, Lerman P. Breathholding spells (cyanotic andpallid infantile syncope). Pediatrics 1967;39:563–81.

[2] DiMario FJ. Breath-holding spells in childhood. Am J Dis Child1992;146:125–31.

[3] Colina KF, Abelson HT. Resolution of breath-holding spells withtreatment of concomitant anemia. J Pediatr 1995;126:395–7.


[4] Daoud AS, Batieha A, Al-Sheyyab M, Abuekteish F, Hijazi S.Effectiveness of iron therapy on breath-holding spells. J Pediatr1997;130:547–50.

[5] Mocan H, Yildiran A, Orhan F, Erduran E. Breath holding spellsin 91 children and response to treatment with iron. Arch DisChild 1999;81:261–2.

[6] Boon R. Does iron have a place in the management of breathholding spells? Arch Dis Child 2002;86:77–8.

[7] DiMario FJ. Prospective study of children with cyanotic andpallid breath-holding spells. Pediatrics 2001;107:265–9.

[8] DiMario Jr FJ, Chee CM, Berman PH. Pallid breath-holdingspells. Evaluation of the autonomic nervous system. Clin Pediatr(Phila) 1990;29:17–24.

[9] DiMario Jr FJ, Burleson JA. Autonomic nervous system functionin severe breath-holding spells. Pediatr Neurol 1993;9:268–74.

[10] DiMario Jr FJ, Bauer L, Volpe J, Baxter D. Respiratory sinusarrhythmia in children with severe cyanotic breath-holding spells.J Child Neurol 1997;12:260–2.

[11] DiMario Jr FJ, Bauer L, Baxter D. Respiratory sinus arrhythmiain children with severe cyanotic and pallid breath-holding spells. JChild Neurol 1998;13:440–2.

[12] Anil BG, Nedunchezian K, Jayanthini V, Pathmanabhan M.Breath-holding spells: evaluation of autonomic nervous systemfunction. Indian Pediatr 2005;42:923–7.

[13] Stephenson JB. Anoxic seizures: self-terminating syncopes. Epi-leptic Disord 2001;3:3–6.

[14] Stephenson J, Breningstal G, Steer C, Kirkpatrick M, Horrocks I,Nechay A, et al. Anoxic-epileptic seizures: home video recordingsof epileptic seizures induced by syncopes. Epileptic Disord2004;6:15–9.

[15] Horrocks IA, Nechay A, Stephenson JP, Zuberi SM. Anoxic-epileptic seizures: observational study of epileptic seizures inducedby syncope. Arch Dis Child 2005;90:1283–7.

[16] Lukkarinen H, Virtanen I, Arikka H, Arola A, Peltola M, EkbladH. Recurrent sinus arrest and asystole due to breath-holding spellin a toddler; recovery with levetiracetam-therapy. Circulation2010;122:637.

[17] Walsh M, Knilans TK, Anderson JB, Czosek RJ. Successfultreatment of pallid breath-holding spells with fluoxetine. Pediat-rics 2012;130:e685–9.

[18] Donma MM. Clinical efficacy of piracetam in treatment ofbreath-holding spells. Pediatr Neurol 1998;18:41–5.

[19] Azam M, Bhatti N, Shahab N. Piracetam in severe breath holdingspells. Int J Psychiatry Med 2008;38:195–201.

[20] Carano N, Bo I, Zanetti E, Tchana B, Barbato G, Agnetti A.Glycopyrrolate and theophylline for the treatment of severe pallidbreath-holding spells. Pediatrics 2013;131:e1280–3.

[21] Brophy GM, Bell R, Claassen J, Alldredge B, Bleck TP, GlauserT, et al. Neurocritical Care Society Status Epilepticus GuidelineWriting Committee. Guidelines for the evaluation and manage-ment of status epilepticus. Neurocrit Care 2012;17:3–23.

[22] Breningstall GN. Breath-holding spells. Pediatr Neurol1996;14:91–7.

[23] Sreeram N, Whitehouse W. Permanent cardiac pacing for reflexanoxic seizure. Arch Dis Child 1996;75:462.

[24] McLeod KA, Wilson N, Hewitt J, Norrie J, Stephenson JB.Cardiac pacing for severe childhood neurally mediated syncopewith reflex anoxic seizures. Heart 1999;82:721–5.

[25] Villain E, Lucet V, Do Ngoc D, Bonnet D, Fraisse A, Kachaner J.Cardiac pacing in children with breath-holding spells. Arch MalCoeur Vaiss 2000;93:547–52.

[26] Kelly AM, Porter CJ, McGoon MD, Espinosa RE, Osborn MJ,Hayes DL. Breath-holding spells associated with significantbradycardia: successful treatment with permanent pacemakerimplantation. Pediatrics 2001;108:698–702.




http://refhub.elsevier.com/S0387-7604(14)00042-4/h0005









































































[27] Legge LM, Kantoch MJ, Seshia SS, Soni R. A pacemaker forasystole in breath-holding spells. Paediatr Child Health2002;7:251–4.

[28] Chehab G, Shalak W, Haddad J, Sacy R, Hatem J, Chouairi S.Cardiac pacing in children with severe form of breath-holdingspells, an exceptional treatment. A case report (in French). J MedLiban 2004;52:171–4.

[29] Wilson D, Moore P, Finucane AK, Skinner JR. Cardiac pacing inthe management of severe pallid breath-holding attacks. JPaediatr Child Health 2005;41:228–30.

[30] Di Pino A, Calabro MP, Gitto P, Bianca I, Oreto G. Permanentcardiac pacing for severe pallid breath-holding spells. Pacing ClinElectrophysiol 2007;30:280–2.


[31] Ergul Y, Otar G, Nisli K, Dindar A. Permanent cardiac pacing ina 2.5 month-old infant with severe cyanotic breath-holding spellsand prolonged asystole. Cardiol J 2011;18:704–6.

[32] de Almeida T, Pennock V, Skinner JR. Medical image. Reflexanoxic seizures in a toddler. N Z Med J 2013;126:83–5.

[33] Iyer A, Appleton R. Management of reflex anoxic seizures inchildren. Arch Dis Child 2013;98:714–7.

[34] Garg AK, Paul SP. Reflex anoxic seizures are not epileptic fits: amanagement dilemma. Indian J Pediatr 2013;80:531–3.

[35] Stephenson JB. Clinical diagnosis of syncopes (including so-calledbreath-holding spells) without electroencephalography or ocularcompression. J Child Neurol 2007;22:502–8.





























pacemaker in complicated and refractory breath-holding spells: when to think about it?

Documents