CaftfittgB(sQtmmapp2

Tm(tnmiocpcstvwi

M

t

FU22

8

8

Early Electrical and Geometric ChangesAfter Percutaneous Closure of Large

Atrial Septal DefectGiuseppe Santoro, MD, Marco Pascotto, MD, Berardo Sarubbi, MD, PhD,

Maurizio Cappelli Bigazzi, MD, Raimondo Calvanese, MD, Carola Iacono, MD,Carlo Pisacane, MD, Maria Teresa Palladino, MD, Giuseppe Pacileo, MD,

Maria Giovanna Russo, MD, and Raffaele Calabro, MD

pc1�fvtt(arycogiI

ardiac arrhythmias and right chamber enlargementre well known long-term sequelae of atrial septal de-ect (ASD). Surgical ASD closure relieves patient symp-oms but often fails to revert cardiac volume overloadndings. Transcatheter ASD closure might be an attrac-ive alternative to surgery, also because of the possibilityo study the amount and time-course of the electro-eometric modifications following shunt disappearance.etween March 2000 and December 2002, 24 patientsage 22.7 �16.8 years) underwent percutaneous clo-ure of large ASD (stretched diameter >20 mm and/orP/QS ratio >1.5:1). ASD closure was performed with

he Amplatzer Septal Occluder device (mean 25 � 7m), achieving a complete occlusion in all patients at 1onth. In 6 patients, right ventricular (RV) monophasicction potential was recorded during the procedure. Allatients underwent standard 12-lead electrocardiogra-hy and transthoracic echocardiography before and at

4 hours and 1 month after ASD closure. After the

th

fecec

tieeo

itioumthrea

aatey aancoe.

paal

uTtaismdeatoo(ddtfihPati

pmv4131-Salerno, Italy. E-mail: santoropino@tin.it.

76 ©2004 by Excerpta Medica, Inc. All rights reserved.The American Journal of Cardiology Vol. 93 April 1, 2004

rocedure, monophasic action potential length in-reased from 359 � 27 to 372 � 27 ms (p <0.0001). Atmonth, QT dispersion decreased from 54 � 25 to 4117 ms (p <0.05), RV diastolic diameter decreased

rom 42 � 6 to 34 � 5 mm (p <0.00001), and leftentricular (LV) diastolic diameter increased from 39 � 5o 44 � 5 mm (p <0.0001), resulting in a decrease inhe RV/LV ratio from 1.11 � 0.22 to 0.79 � 0.11�28.8%, p <0.00001). Electrocardiographic changes,s well as the amount and time-course of RV overloadelief, did not significantly differ between pediatric (<16ears of age; n � 11) and adult patients (n � 13). Inonclusion, regardless of age at procedure, percutane-us ASD closure results in early striking electrical andeometric cardiac changes that may be beneficial dur-

ng long-term follow-up. �2004 by Excerpta Medica,nc.

(Am J Cardiol 2004;93:876–880)

tion.5

ealsoV)an-racicourscri-nceiated;ar-aseringphic

emhy.

ec-ord-

saphicde-

o date, only a few studies have evaluatedelectrocardiographic1,2 and echocardiographic3–6

odifications after nonsurgical atrial septal deASD) closure. However, none of them have prospively addressed these changes in the single paor have they evaluated the ventricular electro-getric changes after shunt disappearance. In add

t has been suggested that the longer the volverload due to the atrial shunt, the less likelyardiac size normalization.5,7–11However, in youngeatients, postponing the procedure for a few yould make the difference between transcatheterurgical closure. This study prospectively investighe cardiac electro-geometric modifications, mainlentricular level, after percutaneous ASD closure,e evaluate the time-course of these changes ac

ng to patient age at the moment of the procedur

ETHODSBetween March 2000 and December 2002, 85

ients with isolated ASD and/or patent foramen ov

rom the Division of Pediatric Cardiology, A.O. “Monaldi,” 2ndniversity of Naples, Naples, Italy. Manuscript received July 22,003; revised manuscript received and accepted December 4,003.

Address for reprints: Giuseppe Santoro, MD, Via Vito Lembo, 14,

e

t-

nt,-n,e

e

rsndstdrd-

-e

nderwent transcatheter closure at our instituwenty-four patients (age 22.7� 16.8 years, range

o 66) had large ASD (“stretched” diameter�20 mmnd/or invasive QP/QS�1.5:1). All patients gave th

nformed consent for ASD closure and 6 patientsigned a different consent for right ventricular (Ronophasic action potential (MAP) recording. Stard 12-lead electrocardiograms and transthochocardiograms were obtained before and 24 hnd 1 month after device implantation. Exclusion

eria for the study analysis were: (1) clinical evidef heart failure; (2) history of symptomatic arrhythmr syncope; (3) arrhythmia at rest or exercise-rela4) recent or current treatment with diuretics or cioactive drugs; and (5) electrolytic and acid-bisturbances at baseline laboratory work-up. Du

he procedure, all patients showed echocardiograndings of RV volume overload and 80.9% of thad RV conduction delay during electrocardiograpatients were divided in pediatric (�16 years, n� 11)nd adult (n� 13) groups, to also analyze the el

rocardiographic and echocardiographic data accng to age at the moment of the procedure.

Device implantation procedure: ASD closure waerformed under transesophageal echocardiogronitoring using the Amplatzer Septal Occluder

ice (AGA Medical Corp., Golden Valley, Minne-

0002-9149/04/$–see front matterdoi:10.1016/j.amjcard.2003.12.027

sbwdspac

cCJrtctpceMm5ttoos

gtDevsa

aQ

wcVmdcwm4swttcsbsp

Sa

FMs

ota). The transcatheter ASD closure technique haseen previously described.1–6 The occluding deviceas chosen within 2 mm of the balloon-stretchediameter and delivered through the appropriate-sizedheath, previously positioned inside the left upperulmonary vein. Aspirin treatment (3 to 5 mg/kg) wasdministered 24 hours before the ASD closure andontinued for 6 months.

MAP recording: MAP recording was obtained by aontact electrode catheter (Steerable MAP-Pacingatheter, Boston Scientific EP Technologies, San

ose, California) introduced through a separate femo-al vein access and positioned into the RV apex duringhe interventional procedure. Catheter position wasonsidered satisfactory once a stable contact was es-ablished, with the electrode tip approximately per-endicular to the endocardium and the distal elasticatheter portion allowed to flex back and forth withach cardiac contraction–relaxation cycle. OptimalAP signals showed stable amplitudes of not �10V, recorded with the bandwidth setting of 0.05 to

00 Hz, along with the intracardiac and surface elec-rocardiogram, throughout the procedure. MAP dura-ion was usually determined at the repolarization levelf 90% of the MAP amplitude. Measurements werebtained by averaging 5 consecutive beats with atable baseline.

Electrocardiographic measurements: Electrocardio-raphic analysis was blindly performed with a digi-izer system (Calcomp Drawing Board III, Calcompigitizer Product Division, Anaheim, California), av-

raging each parameter (PR, QRS, QT and JT inter-als) from all the available leads. Ventricular disper-ion of refractoriness was evaluated through the

TABLE 1 Hemodynamic and Echocardiographic DataAccording to the Patient’s Age at Procedure

Variable

Age Subgroup (yrs)

�16 (n � 11) �16 (n � 13) p Value

Age (yrs) 9.0 � 3.8 35.3 � 13.9 –Body surface area (m2) 1.18 � 0.36 1.69 � 0.08 �0.001Nonstretched ASD

diameter (mm)17 � 4 19 � 4 NS

Mean PA pressure(mm Hg)

24 � 6 21 � 6 NS

QP/QS ratio 2.1 � 0.7 2.1 � 1.0 NSRV diastolic diameter

(bidimensional)(mm)Before 40 � 4 44 � 6 NS1 mo after 33 � 5 36 � 4 NS

Indexed RV diastolicdiameter (bidimensional)(mm/m2)

Before 35 � 10 26 � 4 �0.011 mo after 28 � 6 21 � 3 �0.01

RV size decrease (%) 19 � 10 16 � 12 NSRV/LV diameter ratio

(bidimensional)Before 1.12 � 0.2 1.12 � 0.2 NS1 mo after 0.81 � 0.1 0.80 � 0.1 NSRV/LV ratio decrease (%) 26 � 14 26 � 15 NS

PA � pulmonary artery.

nalysis of QT dispersion and JT dispersion, defined v

CONGENITAL HEART DISEASE/

s the difference between the maximum and minimumT and JT values, respectively.

Echocardiographic recordings: Echocardiographyas performed with an Acuson Sequoia C256 echo-

ardiographic machine (Acuson Corp., Mountainiew, California) using 3V2c or 7V3c probes. M-ode and bidimensional measures were obtained in 3

ifferent frames recorded at the same phase of theardiac cycle. M-mode ventricular measurementsere obtained in the parasternal long-axis view. Bidi-ensional cardiac dimensions were calculated in the

-chamber view. Atrial size was calculated as end-ystolic medio-lateral and supero-inferior diameters,hereas ventricular inlet diameters were calculated at

he tip of the atrioventricular valves at the end-dias-ole. RV/left ventricular (LV) diameter ratio was cal-ulated either from M-mode or bidimensional mea-ures. RV diastolic diameter was also indexed forody surface area, and the time-course of the postclo-ure diameter changes was evaluated according to theatient age at the moment of the procedure.

Statistical analysis: Results are expressed as mean �D. Statistical analysis was performed by 2-tailed pairednd unpaired Student t tests and chi-square tests. A p

IGURE 1. Electrophysiologic modifications after ASD closure. (A)AP length before and shortly after ASD closure; (B) QT disper-

ion before and 1 month after ASD closure.

alue �0.05 was considered statistically significant. Sta-

EARLY CHANGES AFTER TRANSCATHETER ASD CLOSURE 877

1gs(s2

atb4dmdtdve2i(lcAsntmaRsarer

Fhchange in right chamber size early after the device implantation.

878 THE AMERICAN JOURNAL OF CARDIOLOGY� VOL. 93

tistical correlation between dependentvariables was assessed with Pearson’sregression analysis.

RESULTSHemodynamic data: Nonstretched

ASD size was 18 � 6 mm (range 8 to29), resulting in a QP/QS ratio of 2.1� 0.9 (range 1.2 to 5). Mean pulmo-nary artery pressure was 22 � 6 mmHg (range 10 to 35). ASD diameterand hemodynamic data did not differbetween pediatric and adult patients(Table 1). ASD closure was success-fully performed in all patients usingthe Amplatzer Septal Occluder de-vice (mean size 25 � 7 mm, range 11to 38, median 24). Immediate ASDcomplete occlusion was achieved in83.3% of patients (20 of 24 patients),increasing to 100% at 1 month afterthe procedure.

Electrophysiologic changes: ASDclosure resulted in a sudden increaseof the MAP duration, from 359 � 27to 372 � 27 ms (p �0.0001; Figure

), without any significant change in electrocardio-raphic intervals (Table 2). At 1 month, QT dispersionignificantly decreased from 54 � 25 to 41 � 17 msp �0.05; Figure 1), with a similar trend, although nottatistically significant, for JT dispersion (from 55 �9 to 44 � 18 ms, p � 0.066).

Echocardiographic changes: Atrial shunt disappear-nce caused early striking changes in cardiac geome-ry (Figure 2). Twenty-four hours after the procedure,idimensional RV diastolic diameter decreased from2 � 6 to 36 � 6 mm (�14.3%, p �0.00001) and LViastolic diameter increased from 39 � 5 to 41 � 4m (�5.1%, p �0.001). This resulted in a significant

ecrease in RV/LV diameter ratio, from 1.11 � 0.22o 0.91 � 0.22 (�18.0%, p �0.001). At 1 month, RViameter decreased to 34 � 5 mm (�19.0%, p � NSs 24 hours, p �0.000001 vs baseline) and LV diam-ter increased to 44 � 5 mm (�12.8%, p �0.001 vs4 hours, p �0.0001 vs baseline; Figure 3), resultingn a final RV/LV diameter ratio of 0.79 � 0.11�28.8%, p �0.05 vs 24 hours, p �0.00001 vs base-ine; Figure 3). Similar findings in cardiac geometrichanges were found using M-mode analysis (Table 2).t 1 month, the decrease in right atrial size became

tatistically significant, although the left atrial size didot significantly change. A significant correlation be-ween the preprocedural RV volume overload and theagnitude of RV size reduction was found, either as

bsolute RV diameter (r � 0.51, p �0.02) or asV/LV diameter ratio (r � 0.74, p �0.0002). No

ignificant relation between preprocedural RV sizend QT or JT dispersion was found. Similarly, noelation between postprocedural RV volume changes,ither as absolute RV diameter or RV/LV diameter

and After

p Value*

NSNSNSNS

�0.050.066

�0.0001�0.01

�0.00001�0.00001

NSNS

�0.00001�0.0001

3 �0.000011 �0.00001

TABLE 2 Electrocardiographic and Echocardiographic Data BeforeAtrial Septal Defect (ASD) Closure

Before 24 h 1 mo

PR (ms) 137 � 25 134 � 19QRS (ms) 102 � 9 102 � 9QTc (ms) 433 � 18 431 � 18JTc (ms) 329 � 26 329 � 18QTD (ms) 54 � 25 41 � 17JTD (ms) 55 � 29 44 � 18RA diameter (mm)

supero-inferior 52 � 9 47 � 8 45 � 6lateral 43 � 8 36 � 8 35 � 8

RV diastolic diameter (mm)M-mode 34 � 8 31 � 7 26 � 6Bidimensional 42 � 6 36 � 6 34 � 5

LA diameter (mm)supero-inferior 48 � 10 47 � 8 46 � 10lateral 34 � 6 33 � 5 33 � 5

LV diastolic diameter (mm)M-mode 41 � 6 43 � 7 46 � 6Bidimensional 39 � 5 41 � 4 44 � 5

RV/LV diameter ratioM-mode 0.85 � 0.17 0.72 � 0.12 0.56 � 0.1Bidimensional 1.11 � 0.22 0.91 � 0.22 0.79 � 0.1

*Versus preclosure.

IGURE 2. Bidimensional echocardiography before (A) and 24ours after (B) percutaneous ASD closure. Note the remarkable

atio, and QT or JT dispersion was recorded.

APRIL 1, 2004

pvydgoodsi

D

etc

reuosmsIRopbdu

c

FRV/LV diameter ratio (B) after transcatheter ASD closure.

Electrogeometric changes according to age atrocedure: Before ASD closure, the absolute rightentricle size was slightly larger in older than inounger patients. However, at 1 month from proce-ure, no significant difference between the 2 ageroups was found in terms of extent and time-coursef the RV size and RV/LV ratio decrease. Regardlessf patient age at the moment of the procedure, the RViameter reversed to normal range12 shortly afterhunt disappearance, either in terms of absolute orndexed values (Table 1).

ISCUSSIONThe “ ideal” ASD treatment should aim at both

liminating the intracardiac shunt and at revertinghe electrical and geometric changes caused by the

ardiac overload. ASD surgery results in a significant d

CONGENITAL HEART DISEASE/

improvement of symptoms,11,13 butoften fails to completely revert theelectrical and geometric cardiac ab-normalities due to the long-standingvolume overload. Despite significantchanges in electrocardiographic signsof volume overload,14,15 most post-surgical patients continue to experi-ence arrhythmias.16,17 In addition,persistence of RV enlargement hasbeen reported in �80% of pediatricand up to 77% of adult patients overa midterm follow-up.9,18,19 Each time,these findings have been ascribed tochronic myopathic changes fromlong-lasting volume overload, func-tional postsurgical anomalies, or geo-metric modifications due to pericar-dial opening.6,9,22,23 Although arecent comparative study found com-parable effects on cardiac geometryof percutaneous versus surgical ASDclosure over a midterm follow-up,17

the surgical option seems to affectRV function more extensively,6 mak-ing the percutaneous approach defin-itively preferable over surgical ther-apy.

Separating the hemodynamic con-sequences of shunt disappearancefrom the artificial changes due to thesurgical technique, transcatheter ASDclosure can also be considered as an“unbiased” model to study the elec-trical and geometric changes follow-ing a volume overload relief. Basicresearch and human clinical modelsshowed that diastolic stretch due tovolume overload results in a signifi-cant decrease of the ventricular MAPlength, potentially triggering sponta-neous depolarization potentials.21 Inaddition, a chronic diastolic stretchhas been shown to bring about anincreased dispersion of myocardial

efractoriness and repolarization, thus resulting inlectrical instability.22,23 Therefore, a sudden RV vol-me overload relief should revert these electrophysi-logic risk factors. This is the first clinical studyhowing that a sudden relief of the volume-inducedyocardial stretch caused by ASD closure produces

ignificant changes in ventricular electrophysiology.n this model, the sudden significant increase of theV MAP duration might be considered as expressionf the “mechano-electrical feedback” reported in ex-erimental models.24,25 Furthermore, ASD closure haseen associated with a significant decrease in QTispersion, paralleling the striking changes of ventric-lar geometry over a short-term follow-up.

According to other studies,3,5 transcatheter ASDlosure also caused early and striking changes in car-

ers (A) and

IGURE 3. Time-course of changes in RV and LV inlet diastolic diamet

iac geometry, mostly within a few hours after shunt

EARLY CHANGES AFTER TRANSCATHETER ASD CLOSURE 879

dnaRuamf1RtirmtsttialTpse

1ec2Eu43td4Ms5od6s7s8

sC9Ea116C1d1m21rA1eA1vC1a81Co11as1Rd2ac2Prdo52rp2is12p2o

8

isappearance (Figure 2). RV and LV diameters sig-ificantly changed, either at bidimensional or M-modenalysis, eventually resulting in an almost completeV/LV ratio normalization over a short-term follow-p. At 1 month, RV diameter decreased by 19%,pproaching the normal range,15 regardless of theagnitude of the preclosure cardiac overload. At this

ollow-up time period, LV diameter increased by2.8%, resulting in a nearly 30% decrease in theV/LV diameter ratio. Over a short-term follow-up,

he time-course of the cardiac size normalization pos-tively correlated with the magnitude of the preclosureight chamber volume overload. In other words, theore significant the RV volume overload, the higher

he cardiac size decrease after ASD closure. Again, noignificant difference between younger and older pa-ients was found both in terms of magnitude and inime-course of the cardiac geometric changes. Finally,n our study the cardiac size reverted to normal withinfew weeks after percutaneous ASD closure, regard-

ess of the magnitude and length of volume overload.herefore, these data seem to prompt the safe post-onement of percutaneous ASD closure in potentiallyuitable, but small-sized, patients until the transcath-ter treatment may be successfully performed.

. Schenck MH, Sterba R, Foreman CK, Latson LA. Improvement in noninvasivelectrophysiologic findings in children after transcatheter atrial septal defectlosure. Am J Cardiol 1995;76:695–698.. Hill SL, Berul CI, Patel HT, Rhodes J, Supran SE, Cao QL, Hijazi ZM. EarlyCG abnormalities associated with transcatheter closure of atrial septal defectssing the Amplatzer septal occluder. J Interv Card Electrophysiol 2000;4:469–74.. Rao PS, Ende DJ, Wilson AD, Smith PA, Chopra PS. Follow-up results of

ranscatheter occlusion of atrial septal defects with buttoned device. Can J Car-iol 1995;11:695–701.. Veldtman GR, Razack V, Siu S, El-Hajj H, Walker F, Webb GD, Benson LN,cLaughlin PR. Right ventricular form and function after percutaneous atrial

eptal defect device closure. J Am Coll Cardiol 2001;37:2108–2113.. Du ZD, Cao QL, Koenig P, Heitschmidt M, Hijazi ZM. Speed of normalizationf right ventricular volume overload after transcatheter closure of atrial septalefect in children and adults. Am J Cardiol 2001;88:1450–1453.. Dhillon R, Josen M, Henein M, Redington AN. Transcatheter closure of atrialeptal defect preserves right ventricular function. Heart 2002;87:461–465.. Richmond DE, Lowe JB, Barratt-Boyes BG. Results of surgical repair of atrialeptal defects in the middle-aged and elderly. Thorax 1969;24:536–542.

. Phillips SJ, Okies JE, Henken D, Sunderland CO, Starr A. Complex of m

80 THE AMERICAN JOURNAL OF CARDIOLOGY� VOL. 93

ecundum atrial septal defect and congestive heart failure in infants. J Thoracardiovasc Surg 1975;70:696–700.. Pearlman AS, Borer JS, Clark CE, Henry WL, Redwood DR, Morrow AG,pstein SE. Abnormal right ventricular size and ventricular septal motion aftertrial septal defect closure: etiology and functional significance. Am J Cardiol978;41:295–301.0. John Sutton MG, Tajik AJ, McGoon DC. Atrial septal defect in patients aged0 years or older. Operative results and long-term postoperative follow-up.irculation 1981;64:402–409.1. Ghosh S, Chatterjee S, Black E, Firmin RK. Surgical closure of atrial septalefects in adults: effect of age at operation on outcome. Heart 2002;88:485–487.2. Schnittger I, Gordon EP, Fitzgerald PJ, Popp RL. Standardized intracardiaceasurements of two-dimensional echocardiography. J Am Coll Cardiol 1983;

:934–938.3. Doty DB, Wright CB, Hiratzaka LF, Eastham CL, Marcus ML. Coronaryeserve in volume-induced right ventricular hypertrophy from atrial septal defect.m J Cardiol 1984;54:1059–1063.4. Karpawich PP, Antillon JR, Cappola PR, Agarwal KC. Pre- and postoperativelectrophysiologic assessment of children with secundum atrial septal defect.m J Cardiol 1985;55:519–521.5. Terada Y, Mitsui T, Sato M, Horigome H, Tsukada K. Right ventricularolume unloading evaluated by tangential magnetocardiography. Jpn J Thoracardiovasc Surg 2000;48:16–23.6. Gatzoulis MA, Freeman MA, Siu SC, Webb GD, Harris L. Atrial arrhythmiafter surgical closure of atrial septal defects in adults. N Engl J Med 1999;340:39–846.7. Brandenburg RO Jr., Holmes DR Jr., Brandemburg RO, McGoon DC.linical follow-up study of paroxysmal supraventricular tachyarrhythmias afterperative repair of secundum type atrial septal defect in adults. Am J Cardiol983;51:273–276.8. Meyer RA, Korfhageh JC, Covitz W, Kaplan S. Long-term follow-up studyfter closure of secundum atrial septal defect in children: an echocardiographictudy. Am J Cardiol 1982;50:143–148.9. Ning SB, Fazal H, Cook D, Wood MM, Wood MM, Duncan WJ, Rowe RD.ight ventricular size and ventricular septal motion after repair of atrial septalefect in children. Can J Surg 1984;27:395–398.0. Thomson JD, Aburawi EH, Watterson KG, Van Doorn C, Gibbs JL. Surgicalnd transcatheter (Amplatzer) closure of atrial septal defects: a prospectiveomparison of results and costs. Heart 2002;87:466–469.1. Sarubbi B, Mercurio B, Calvanese R, Santoro G, Cappelli Bigazzi M,alladino MT, Russo MG, Carminati M, Calabro R. Monophasic action potentialecording during volume overload relief in humans: from basic to clinicalemonstration of mechano-electrical feedback. Proceedings The “New Frontiers”f Arrhythmias, XV International Congress. G Ital Aritm Cardiostimol 2002;(suppl 1):172–174.2. Satoh T, Zipes DP. Unequal atrial stretch in dogs increases dispersion ofefractoriness conducive to developing atrial fibrillation. J Cardiovasc Electro-hysiol 1996;7:833–842.3. Gatzoulis MA, Till JA, Somerville J, Redington AN. Mechano-electrical

nteraction in tetralogy of Fallot. QRS prolongation relates to right ventricularize and predicts malignant ventricular arrhythmias and sudden death. Circulation995;92:231–237.4. Dean J, Lab MJ. The effect of changes in load on the monophasic actionotential of the in-situ pig heart. Cardiovasc Res 1989;23:887–896.5. Franz MR, Cima R, Wang D, Profitt D, Kurz R. Electrophysiological effectsf myocardial stretch and mechanical determinants of stretch-activated arrhyth-

ias. Circulation 1992;86:968–978.

APRIL 1, 2004