CcT

ELO

FIS

BoFs

OEfc

MpPt

Rhpa(01gB

IFubarHasi

AAMc

1

linical, electrocardiographic, and electrophysiologicharacteristics of patients with a fasciculoventricular pathway:he role of PRKAG2 mutation

duardo Back Sternick, MD, PhD, FHRS,* Antonio Oliva, MD,† Luiz Márcio Gerken, MD,*uiz Magalhães, MD,‡ Ricardo Scarpelli, MD,* Frederico Soares Correia, MD,* Silvia Rego, MD,*to Santana, MD, PhD,‡ Ramon Brugada, MD, PhD,¶ Hein J.J. Wellens, MD, PhD, FACC§

rom the *Arrhythmia Unit, Biocor Instituto, Nova Lima, Brazil, †Institute of Legal Medicine, Catholic University, Rome,taly, ‡University Hospital, Federal University of Bahia, Salvador, Brazil, ¶Genetics Center, University Hospital, Girona,

§

pain, and Cardiovascular Research Institute, Maastricht, The Netherlands.

afNe

Ccecmm

KAGk

ADpb

(

ACKGROUND The ECG, clinical, and electrophysiologic profilesf patients with a fasciculoventricular pathway are well described.asciculoventricular pathways occurring in the setting of glycogentorage cardiomyopathy possess unique features.

BJECTIVE The purpose of this study was to compare the clinical,CG, and electrophysiologic characteristics of patients with aasciculoventricular pathway, with or without glycogen storageardiomyopathy.

ETHODS Two groups of patients with a fasciculoventricularathway were compared: group A consisted of 10 patients with theRKAG2 mutation (Arg302gln), and group B consisted of 9 pa-ients without the mutation.

ESULTS Thirty percent of group A patients had left ventricularypertrophy, and none had an additional accessory pathway. Group Batients had no structural heart disease, and 33% had an additionalccessory pathway. Group A patients had a slower resting heart rate56 � 7 vs 75 � 10 bpm, P �0.0001), a wider QRS complex (0.15 �.01 vs 0.11 � 0.02 ms, P � .0004), and a longer HV interval (34 �vs 25 � 3 ms, P � .0003). During long-term follow-up, 50% of

roup A patients developed complete AV block versus none in group

iA

adpPtnivttapeived July 24, 2010; accepted September 15, 2010.)

547-5271/$ -see front matter © 2011 Heart Rhythm Society. All rights reserved

trial fibrillation. No Group B patient had any arrhythmia duringollow-up after successful ablation of additional arrhythmia circuits.o sustained ventricular arrhythmia was induced in any patient fromither group.

ONCLUSION Patients with a fasciculoventricular pathway asso-iated with the PRKAG2 mutation have distinct clinical, ECG, andlectrophysiologic profiles and should be correctly identified be-ause of their ominous long-term prognosis. Patients without theutation have an excellent arrhythmia-free prognosis after treat-ent of additional circuits.

EYWORDS Accessory atrioventricular pathway; Atrial fibrillation;trial flutter; Atrioventricular block; Fasciculoventricular pathway;lycogen storage cardiomyopathy; PRKAG2 mutation; Wolff-Par-inson-White syndrome

BBREVIATIONS AF � atrial fibrillation; AH � atrio-His interval;NA � deoxyribonucleic acid; ECG � electrocardiogram; EPS � electro-hysiologic study; LV � left ventricle; RBBB � right bundle branchlock; WPW � Wolff-Parkinson-White

Heart Rhythm 2011;8:58–64) © 2011 Heart Rhythm Society. All rights

. Eighty percent of group A patients developed atrial flutter and/or reserved.

ntroductionasciculoventricular pathways are a rare variant of ventric-lar preexcitation. They take off from the His bundle orundle branch and insert into the ventricular septum. Theyre not capable of sustaining reentry, do not participate ineentrant circuits, and are considered an ECG curiosity.1

owever, it has been shown that up to 40% of patients withfasciculoventricular pathway may have additional acces-

ory atrioventricular pathways.2 Therefore, it is important todentify fasciculoventricular pathways and to avoid target-

ddress reprint requests and correspondence: Dr. Eduardo Sternick,lameda do Morro 85, Torre IV, 1900, CEP 34000-000, Nova Lima,inas Gerais, Brazil. E-mail address: eduardosternick@terra.com.br. (Re-

ng them with catheter ablation, which could damage theV conduction system.3

Familial Wolff-Parkinson-White (WPW) syndrome is oftenssociated with glycogen storage cardiomyopathy, a geneticisorder caused by PRKAG2 gene mutation. A substantialercentage of familial accessory pathways4 associated withRKAG2 mutation have decrementally conducting proper-

ies.5 Among them are fasciculoventricular pathways6 andodoventricular fibers7 but not atriofascicular fibers. The clin-cal picture of glycogen storage cardiomyopathy includes leftentricular (LV) hypertrophy, sinus bradycardia, AV conduc-ion disturbances, and atrial tachyarrhythmias.8 It is possiblehat the true incidence of PRKAG2 mutation is underestimatednd that many cases are wrongly diagnosed as idiopathic hy-

ertrophic cardiomyopathy; therefore, it is important to cor-

. doi:10.1016/j.hrthm.2010.09.081

rtd

alamr

MTctdawt

EP1mMibwa

ttun

FT�

igtbnawdp

SIsopwGpi

T

G

G

Aas

59Sternick et al Fasciculoventricular Pathways and PRKAG2 Mutation

ectly identify these patients because of their ominous long-erm prognosis due to the high incidences of sudden cardiaceath and heart failure.

The purpose of this study was to compare the clinical, ECG,nd electrophysiologic characteristics of patients with a fascicu-oventricular pathway with or without the PRKAG2 mutationnd to determine predictors of the presence of a PRKAG2utation that could help the clinician correctly identify patients

equiring further investigation and genetic counseling.

ethodshis was a retrospective cohort study of patients with fas-iculoventricular pathways. Inclusion criteria were (1) elec-rophysiologic study (EPS) showing standard criteria for theiagnosis of a fasciculoventricular pathway; and (2) geneticssessment for the PRKAG2 mutation. All patients under-ent 12-lead surface ECG, 24-hour Holter monitoring, and

ransthoracic echocardiography.

lectrophysiologic studyrogrammed electrical stimulation and recordings of the2-lead surface ECG and intracardiac electrograms wereade using the EP Tracer or MS System (CardioTek BV,aastricht, The Netherlands). Patients challenged with

ntravenous adenosine received a 6-�g bolus, followedy 12 �g if no significant effect on AV nodal conductionas seen. Programmed electrical stimulation included

trial and ventricular pacing at increasing rates and ex-

able 1 Clinical characteristics of patients with fasciculoventri

Family Age (yr) GenderPresentingsymptoms Arrhythmia

roup A1 I 32 M Palpitations A. Flutter2 I 28 M tachycardia A. Flutter3 I 23 M tachycardia AF4 I 26 M No SB5 I 29 F Tachycardia A. Flutter6 I 19 M No7 I 60 M Syncope AF � AVB8 II 28 M Tachycardia A. Flutter9 II 26 F Tachycardia A. Flutter10 III 20 M Syncope AF

29 � 11roup B1 51 M No2 19 M Palpitations PAC3 19 M Cardiac arrest AF � VF4 21 M No5 54 M Tachycardia AVNRT6 13 M Tachycardia AVRT7 27 F Tachycardia AVNRT � A8 32 F Tachycardia AVRT9 51 M No

31 � 16

A. Flutter � atrial flutter; AF � atrial fibrillation; AVB � intermittenVRT � atrioventricular reentrant tachycardia; HCM � hypertrophic cardiotrioventricular block; LL � left free wall; LPS � left posteroseptal; PAC �

inus bradycardia; VF � ventricular fibrillation; WPW � Wolff-Parkinson-White sy

rastimuli during sinus rhythm and during atrial and ven-ricular pacing. The ventricular stimulation protocol withp to three extrastimuli was repeated during isoprotere-ol infusion.

asciculoventricular pathway diagnostic criteriahe baseline HV (H-delta) interval during sinus rhythm is35 m. During atrial pacing at increasing rates, the HV

nterval does not change. Atrial premature beats cause pro-ressive prolongation of the AH interval but no change inhe HV interval and QRS configuration, unless they causedlock in the fasciculoventricular pathway resulting in aormal HV interval and narrow QRS complex. Response todenosine triphosphate suggests a fasciculoventricular path-ay when prolongation of the PR interval (AH interval)oes not change the degree of preexcitation or when com-lete AV block occurs after the P wave.2

tudy populationnformed and written consent was obtained from studyubjects. Group A consisted of 10 patients with a mutationf the �2-subunit of the adenosine monophosphate-activatedrotein kinase (PRKAG2 gene mutation). Nine patientsere from two families (I and II) previously reported.9

roup B consisted of 9 patients with a fasciculoventricularathway from among a cohort of 16 patients from ournstitution who agreed to undergo genetic assessment.

athways

ow-up

Device WPW HCMPRKAG2mutationptoms Arrhythmia

cope AVB � AF DDD PM Yes Arg302gincope AVB DDD PM No Arg302giniac arrest AF?/VF No Arg302ginr-syncope Int AVB DDD PM No Arg302gincope AVB DDD PM No Arg302gin

No Arg302ginVVI PM No Arg302gin

iac arrest AF DDD ICD Yes Arg302ginitations AF No Arg302gin

Yes Arg302gin50% MP 30%

NoNo

LL � RMS NoNoNo

LL NoNo

LPS NoNo

27% 0%

ventricular block; AVNRT � atrioventricular nodal reentrant tachycardia;thy; ICD � implantable cardioverter-defibrillator; int AVB � intermittentture atrial contractions; PM � pacemaker; RMS � right midseptal; SB �

cular p

Foll

Sym

SynSynCardNeaSyn

CardPalp

T

t atriomyopaprema

ndrome.

GTvtfdbSbqwCdS

FAf(1f1Gd

SVfet1

RCTo

GMarbdfi

ao1msat

npataPtAsd

pcytclAl

60 Heart Rhythm, Vol 8, No 1, January 2011

enetic analysishe study was approved by the Regional Institutional Re-iew Board, and the patients gave written consent to par-icipate in the study. Blood samples (10 ml) were obtainedrom participating family members and spouses. Genomiceoxyribonucleic acid (DNA) was isolated from peripherallood leukocytes using a commercial kit (Puregene, Gentraystems, Minneapolis, MN, USA). Exons and intron–exonoundaries of PRKAG2 were analyzed using direct se-uencing. Polymerase chain reaction products were purifiedith a commercial reagent (ExoSAP-IT, USB Corporation,leveland, Ohio) and were directly sequenced from bothirections using the ABI PRISM 3100 Automatic DNAequencer (AB Applied Biosystems, Foster City, CA).9

ollow-upll patients with the PRKAG2 mutation underwent regular

ollow-up every 6 months. In family I of group A, the probandpatient 1) and relatives have been undergoing follow-up since996, except for patient 3, who died suddenly in 1998. Inamily II of group A, patients have been seen regularly since999. Patient 10 of group A was diagnosed in January 2010.roup B patients were followed-up for 70 � 21 months. Theyid not have any recurrent arrhythmia.

tatistical analysisalues are given as mean � SD. The significance of dif-

erences (P �.05) between groups of clinical, ECG, orlectrophysiologic parameters was assessed by Student’s-test (paired and unpaired) or Fisher exact test using Stata1 software (STATA Corp LP, TX).

esultslinical findingsable 1 lists data on age, gender, presenting symptoms, typef arrhythmia, and subsequent course.

roup Aost of the group A patients were symptomatic. Only patients 4

nd 6 were initially asymptomatic. They were referred based uponecognition of a familial condition in most of their family mem-ers. Eight (80%) of the 10 patients presented with palpitationsue to atrial flutter, with 2:1 AV block in 5 patients and atrialbrillation (AF) in 3 patients (Table 1).

Patient 10 was admitted due to syncope in the setting offast preexcited atrial fibrillation. Transthoracic echocardi-graphy showed LV hypertrophy (interventricular septum9 mm, LV posterior wall 15 mm). The patient underwentyocardial septal biopsy to confirm a presumptive diagno-

is of glycogen storage cardiomyopathy (Figure 1). Geneticssessment later revealed the Arg302gln PRKAG2 muta-ion, the same pattern found in all group A patients.

During follow-up, another six patients had episodes ofear-syncope, syncope, or cardiac arrest. All five group Aatients suffering from type I atrial flutter underwent righttrial cavo-tricuspid isthmus radiofrequency catheter abla-ion, with resumption of sinus rhythm. Three developedtrial fibrillation during follow-up (patients 1, 8, and 9).atient 3 had many episodes of atrial fibrillation with ven-

ricular rates up to 280 bpm during 24-month follow-up.fter 2 hours of fast palpitations with recurrent episodes of

yncope, the patient eventually lost consciousness and dieduring transportation to the hospital.

Patient 8 received an implantable cardioverter-defibrillator asrimary prevention due to a presumptive diagnose of hypertrophicardiomyopathy (LV hypertrophy and recurrent syncope). Twoears later, genetic assessment was positive for PRKAG2 muta-ion. Of interest, the patient received an inappropriate implantableardioverter-defibrillator shock due to an episode of atrial fibril-ation with a ventricular rate of 230 bpm. In total, 5 of the 10 group

patients developed complete AV block during long-term fol-ow-up (Table 1).

Figure 1 Group A, patient 10. Top:Rhythm strip (leads I, II, and III) showingvery fast atrial fibrillation (280–300 bpm).Bottom: Endomyocardial right ventricularseptal biopsy specimens (hematoxylin andeosin E staining). Bottom left: Normalmyocardial architecture with normal myo-fibrils, without myofiber disarray. Bottomright: Presence of profound vacuolization(arrows) in most of the myofibrils causedby glycogen storage (PRKAG2 Arg302glnmutation) (high-power magnification).

GG(gPtmcwadattpr

E

●

●

●

E

●

●

●

●

T

G

G

S

61Sternick et al Fasciculoventricular Pathways and PRKAG2 Mutation

roup Broup B consisted of nine unrelated individuals. Three

23%) patients had an additional accessory pathway. Noroup B patient had structural heart disease or carried theRKAG2 mutation. Three patients were asymptomatic. Pa-

ients 1 and 9 were referred for electrophysiologic assess-ent due to ventricular preexcitation found during routine

ardiologic evaluation for a noncardiac surgery. Patient 4as assessed in a cardiologic precompetitive sports evalu-

tion. Five patients presented with palpitations, which wereue to regular tachycardia in 4 patients with additionalrrhythmia circuits and premature atrial contractions in pa-ient 2. Patient 3 was resuscitated from ventricular fibrilla-ion without sequelae. He had two additional accessoryathways, one (left free wall) with a very short effectiveefractory period (�200 ms).

CG findings (Table 2)

Heart rhythm: Group A patients had on average a signifi-cantly slower mean heart rate (56 � 7 bpm vs 75 � 10 bpm,P �.0001) as assessed by 24-hour Holter monitoring.PR interval: PR interval did not differ among group A andB patients (0.11 � 0.01 ms vs 0.10 � 0.01 ms, P � NS).QRS complex: Group A patients had a wider QRS complex(0.15 � 0.01 ms vs 0.11 � 0.02 ms, P � .0004), with a lowerfrontal plane QRS axis (21° � 18° vs 48° � 24°, P � .03;Figure 2). Nine of 10 group A patients had a right bundlebranch block (RBBB)–like morphology, whereas only patient10 had a left bundle branch block–like morphology. No group

able 2 ECG and electrophysiologic characteristics of patients

HR PR QRSBBB-like(QRS �0.12 sec.) QRS axis AH

roup A1 58 0.12 0.16 RBB 15 452 60 0.11 0.12 RBB 30 503 54 0.1 0.16 RBB 15 354 52 0.12 0.16 RBB 30 505 72 0.1 0.15 RBB 60 546 50 0.13 0.16 RBB 0 457 �30 0.15 RBB8 62 0.1 0.16 RBB 10 359 54 0.1 0.16 RBB 30 3810 48 0.12 0.18 LBB 0 35

56 � 7 0.11 � 0.01 0.15 � 0.01 21 � 18 43 �roup B1 72 0.1 0.1 80 602 62 0.09 0.12 70 503 86 0.12 0.12 50 754 58 0.11 0.12 60 805 72 0.09 0.09 15 556 86 0.11 0.15 LBB 30 607 84 0.11 0.15 LBB 45 608 76 0.11 0.08 15 659 80 0.12 0.08 70 60

75 � 10 0.10 � 0.01 0.11 � 0.02 48 � 24 61 �P �.0001 P � NS P � .0004 P � .03 P �

2:1 � 2:1 AV block; BBB � bundle branch block; HR � heart rate; IHB � infrahiB � sinus bradycardia; VT � ventricular tachycardia.

B patient had an RBBB-like morphology.

lectrophysiologic parameters (Table 2)

AV nodal conduction: Group A patients had a shorter AVnodal conduction time (AH: 43 � 7 ms vs 61 � 9 ms, P �.002). Besides having an average shorter AH interval, group Apatients had a much shorter maximal AH interval (maximal 1:1AH interval during rapid atrial stimulation minus baseline AHinterval; 63 � 23 ms vs 148 � 38 ms, P � .0008). The AVnodal Wenckebach point during incremental atrial pacing didnot differ between the two groups of patients (332 � 102 ms vs325 � 48 ms, P � NS). However, the AH increment during theWenckebach sequence differed significantly (105 � 23 ms vs210 � 44 ms, P � .0002).Infranodal conduction: A split His was recorded duringEPS in two group A patients (no. 1 and 4). The HVinterval was significantly longer in group A patients(34 � 1 ms vs 25 � 3 ms, P � .0003). During atrialpacing at increasing rates, all patients showed a fixed HVinterval, except for patient 10 of Group A, who had acycle length–dependent increase in the HV interval,probably caused by decremental conduction in the fas-ciculoventricular pathway (Figure 3).Programmed ventricular stimulation: All patients under-went programmed ventricular stimulation. No sustainedventricular tachyarrhythmia was inducible in any patient,including group A patients 1, 8, and 10, who had LVhypertrophy (all of these patients probably had a glyco-gen storage cardiomyopathy without myocardial architec-ture disarray or true LV hypertrophy, as found on myo-cardial biopsy in patient 10).Adenosine: Six group A patients were challenged with

sciculoventricular pathways

ax. AHHH(split His) HV PW Max. (�AH)

AHAdenosine

VTInduction

85 Yes 35 360 40 no00 35 320 50 SB no65 34 300 30 SB no95 Yes 35 280 45 2:1 no44 34 270 90 SB no05 35 600 60 SB no

IHB15 32 300 85 no28 32 320 90 no15 34 280 80 SB no05 � 23 34 � 1 332 � 102 63 � 23

80 25 320 120 no10 25 380 160 Long PR no15 25 290 140 Long PR no20 19 350 240 2:1 no15 29 360 160 no80 20 320 120 no70 28 250 110 Long PR no05 31 270 150 no00 25 390 140 Long PR no10 � 44 25 � 3 325 � 48 148 � 38� .0002 P � .0003 P � NS P � .0008

k; LBB � left bundle branch; PW � Wenckebach point; RBB � right bundle branch;

with fa

M

1

11

111

7 1

122321122

9 2.002 P

sian bloc

adenosine. Five of six patients responded with further

GDmfi(fmnvp

DAwP

TTh

pgn

AGAirwlbAtavaetact

FiG

62 Heart Rhythm, Vol 8, No 1, January 2011

slowing of the sinus rate. Only patient 4 developed atransient 2:1 AV nodal block (Figure 4A). Five group Bpatients received an intravenous bolus of adenosine. Fourshowed an increase in PR interval without change in QRSmorphology and width (Figure 4B). Patient 4 respondedwith 2:1 AV nodal block.

enetic analysisirect sequencing of the PRKAG2 gene revealed the sameissense mutation in exon 7 (Arg302Gln) in six individuals

rom group A family I siblings and in their mother, in twondividuals from family II, and in patient 10 from family IIITable 1). This mutation has been described in the familialorm of the WPW syndrome, and it is the most commonutation in the WPW syndrome and LV hypertrophy phe-

otype.8 A transgenic mouse model has already been de-eloped with the characteristic phenotype.10 No group Batient had the mutation.

iscussioncomparison of patients with a fasciculoventricular path-

ay as their accessory connection with and without theRKAG2 mutation revealed several interesting findings.

welve-lead ECGhe peculiar ECG profile seen in group A patients (slower

eart rate, AV conduction disturbances, wider QRS com- g

lex [mostly RBBB]) likely were caused by the underlyinglycogen storage cardiomyopathy involving the sub–AVodal conduction system and ventricular muscle.4,5,8,9

V nodal conduction behaviorroup A patients consistently showed evidence of unusualV nodal properties as evidenced by a shorter baseline AH

nterval, limited decremental AH conduction, and lack ofesponse to intravenous adenosine. All of these parametersere significantly shorter compared to patients with fascicu-

oventricular pathways without the PRKAG2 mutation (Ta-le 2). Despite the fact that no difference in the moment ofV nodal Wenckebach block was seen during atrial pacing,

he much smaller increment in AV conduction time on ratecceleration may explain why some group A patients hadery fast ventricular rates during atrial fibrillation. In thebsence of anatomic data, we can only speculate about thexplanation for the lack of decremental AV nodal conduc-ion properties in group A patients. Our findings rule out antrio-His bypass tract and a nodoventricular fiber but areonsistent with an anatomically small AV node in additiono the fasciculoventricular pathway.

asciculoventricular pathway with near-normal HVntervalroup A patients had a longer mean HV interval than did

Figure 2 Top: Typical 12-lead ECG of a group A patientshowing short PR, wide QRS complex with right bundlebranch block–like pattern, and QRS frontal plane axis at 20°.Bottom: Twelve-lead ECG of a group B patient showingshort PR interval, minimal preexcitation, more narrow QRScomplex, and higher QRS frontal plane axis at 60°.

roup B patients. Dey et al11 reported a patient with fas-

cdpttAl

sd

AfAp

Fitimrt

FAsc

63Sternick et al Fasciculoventricular Pathways and PRKAG2 Mutation

iculoventricular pathway who showed decremental con-uction during atrial pacing at increasing rates. Group Aatient 10 had decremental conduction following a criticallyimed atrial premature beat (Figure 3). That observation andhe significantly longer HV interval and wider QRS in group

patients suggest slow conduction in the fasciculoventricu-ar pathway as well as in intraventricular conduction in the

igure 3 Group A, patient 10. Follow-ng a critically timed atrial premature beat,here was prolongation of the AH and HVntervals, without change in QRS complexorphology, which is consistent with dec-

emental conduction over a fasciculoven-ricular pathway.

igure 4 A: Group A, patient 4. The only patient in this group who respV block). PR interval of conducted P waves also enlarged, while the

uggestive of intraatrial conduction delay. B: Group B, patient 7. Followin

hange in the degree of preexcitation and QRS morphology.

etting of glycogen storage cardiomyopathy, with the mostelay or block in the right bundle branch.

ssociation of LV hypertrophy withasciculoventricular pathway

30% incidence of LV hypertrophy was found in group Aatients. No group B patient had LV hypertrophy. We would like

to 12 �g of adenosine with development of transient AV nodal block (2:1mplexes remained unaltered. Note the enlarged P wave in this patient,�g bolus of adenosine, there was prolongation of the PR interval without

ondedQRS cog a 12-

tpphs7iiwgtn

MOml3pfit(Ouo

IptTituapai(fWpuimtcfMdrisRPr

tPfinep

IfIftsiopfiolse

R

1

1

1

1

1

1

64 Heart Rhythm, Vol 8, No 1, January 2011

o stress that the association of LV hypertrophy with ventricularreexcitation, particularly in the presence of a fasciculoventricularathway, should raise suspicion of a PRKAG2 mutation. LVypertrophy in patients with PRKAG2 is caused by glycogentorage in the myofibrils, with a reported incidence of 26% to4%.5,8 It is possible that the true incidence of PRKAG2 mutations underestimated and that many cases are wrongly diagnosed asdiopathic hypertrophic cardiomyopathy. Some cases of patientsith fasciculoventricular pathways and LV hypertrophy withoutenetic assessment have been reported.11,12 The ominous long-erm prognosis of carriers of a mutant PRKAG2 emphasizes theeed to rule out PRKAG2 mutation.

echanisms of syncope and sudden deathne group B patient had an aborted sudden death. Theechanism of his ventricular fibrillation probably was re-

ated to occurrence of fast preexcited AF. Group A patientdied suddenly at age 23 years, and the mechanism of deathrobably was AF with fast heart rate causing ventricularbrillation. Patients with the PRKAG2 mutation are known

o be at high risk for dying from fast AF at an early age20–30 years) and from complete AV block at a later age.9

f interest, no patient in both groups had inducible ventric-lar tachyarrhythmias, which suggests that the mechanismf death is not related to primary ventricular arrhythmias.

ncidence of WPW syndrome and variants ofreexcitation associated with PRKAG2 comparedo other serieshe incidence of ventricular preexcitation in PRKAG2 patients

s quite variable. Mehdirad et al13 found ventricular preexcita-ion in 100% of their 26 patients. Gollob et al5 found ventric-lar preexcitation in all 24 patients for whom an ECG wasvailable (from a cohort of 31 patients). Five (62%) of the 8atients who underwent EPS had a decrementally conductingccessory pathway. Arad et al8 found ventricular preexcitationn 65% (28/43) of patients with Arg302Gln, but in only 23%6/25) of patients with the Asn488Ile mutation. Murphy et al14

ound ventricular preexcitation in 60% (n � 27) of patients.e identified ventricular preexcitation in 10 (47%) of 21

atients from three unrelated families (group A patients). Allnderwent EPS, and a fasciculoventricular pathway was foundn 100% of the patients. Histologic studies in N488I transgenicice with ventricular preexcitation showed structural disrup-

ions of the annulus fibrosus due to glycogen-laden myo-ytes.10 Ventricular preexcitation was found in 13 subjectsrom a single Dutch family, with the R302Q mutation and a

ahaim-like fiber found in 50% of the six patients who un-erwent EPS.7 Autopsy of one patient who died suddenlyevealed three small nodoventricular tracts. In that study, nonformation was given about the sub–AV nodal conductionystem, which would be of interest given the very commonBBB-like QRS configuration in our group A patients.RKAG2 mutant transgene suppression with doxycycline can

educe glycogen accumulation and preexcitation occurrence in

ransgenic mice.15 All of these findings suggest that theRKAG2 gene may play a role in the development of annulusbrosus, and that malfunction of the gene, by causing promi-ent structural disruptions with extensive arborization, mayxplain the remarkably high incidence of fasciculoventricularathways seen in patients with the PRKAG2 mutation.

mportance of identifying patients withasciculoventricular pathway with PRKAG2n contrast to the excellent long-term prognosis of patients with aasciculoventricular pathway without structural heart disease,hose with the PRKAG2 mutation have a high incidence of sickinus syndrome and complete AV block requiring pacemakermplantation. Eighty percent developed atrial arrhythmias. Ourbservations raise the question of early AV nodal ablation andacemaker implantation because of the risk of dying suddenlyrom a fast ventricular rate during atrial flutter or fibrillation. Thendication for pacemaker implantation is also supported by the riskf dying from sudden complete AV block. We conclude thatifelong follow-up of the mutation carrier as well as genetic coun-eling of family members is mandatory because of the high pen-trance of this autosomal dominant disorder.

eferences1. Sternick EB, Wellens HJJ. Variants of Ventricular Preexcitation: Recognition

and Treatment, Malden, MA: Blackwell Futura, 2006:83–103.2. Sternick EB, Gerken LM, Vrandecic MO, Wellens HJJ. Fasciculoventricular

pathways: clinical and electrophysiologic characteristics of a variant of preex-citation. J Cardiovasc Electrophysiol 2003;14:1057–1063.

3. Sternick EB, Rodriguez LM, Gerken LM, Wellens HJJ. The electrocardiogram ofpatients with fasciculoventricular pathways. A comparative study with patients withanteroseptal and midseptal accessory pathways. Heart Rhythm 2005;2:1–7.

4. Mac Rae CA, Ghaisas N, Kass S, et al. Familial hypertrophic cardiomyopathywith Wolff-Parkinson-White syndrome maps to a locus on chromosome 7q3.J Clin Invest 1995;96:1216–1220.

5. Gollob MH, Green MS, Tang ASL, et al. Identification of a gene responsible forfamilial Wolff-Parkinson-White syndrome. N Engl J Med 2001;344:1823–1831.

6. Govindan M, Ward D, Behr E. A rare connection: fasciculoventricular pathwayin PRKAG2 disease. J Cardiovasc Electrophysiol 2010;21:329–332.

7. Tan HL, Van der Wal AC, Campian ME, et al. Nodoventricular accessory pathwaysin PRKAG2-dependent familial pre-excitation syndrome reveal a disorder in cardiacdevelopment. Circ Arrhythmia Electrophysiol 2008;1:276–281.

8. Arad M, Benson DW, Perez-Atayde AR, et al. Constitutively active AMP kinasemutations cause glycogen storage disease mimicking hypertrophic cardiomyop-athy. J Clin Invest 2002;109:357–362.

9. Sternick EB, Oliva A, Magalhâes LP, et al. Familial pseudo-Wolff-Parkinson-White syndrome. J Cardiovasc Electrophysiol 2006;17:724–732.

0. Patel VV, Arad M, Moskowitz IPG, et al. Electrophysiologic characterizationand postnatal development of ventricular pre-excitation in a mouse model ofcardiac hypertrophy and Wolff-Parkinson-White syndrome. J Am Coll Cardiol2003;42:942–951.

1. Dey S, Tschopp D, Morady F, Jongnarangsin K. Fasciculoventricular bypasstracts with decremental conduction properties. Heart Rhythm 2006;3:975–976.

2. Ghosh S, Avari JN, Rhee E, Woodard PK, Rudy Y. Hypertrophic cardiomyop-athy with preexcitation: insights from noninvasive electrocardiographic imaging(ECGI) and catheter mapping. J Cardiovasc Electrophysiol 2008;19:1215–1217.

3. Mehdirad AA, Fatkin D, DiMarco JP, et al. Electrophysiologic characteristics ofaccessory atrioventricular connections in an inherited form of Wolff-Parkinson-White syndrome. J Cardiovasc Electrophysiol 1999;10:629–635.

4. Murphy RT, Mogensen J, McGarry K, et al. Adenosine monophosphate-acti-vated protein kinase disease mimics hypertrophic cardiomyopathy and Wolff-Parkinson-White syndrome. J Am Coll Cardiol 2005;45:922–930.

5. Wolf CM, Arad M, Ahmad M, et al. Reversibility of PRKAG2 glycogen-storage

cardiomyopathy and electrophysiological manifestations. Circulation 2008;117:144–154.