Impact of transvenous ventricular pacing leads on tricuspidregurgitation in pediatric and congenital heart diseasepatients

Gregory Webster & Renee Margossian &

Mark E. Alexander & Frank Cecchin &

John K. Triedman & Edward P. Walsh & Charles I. Berul

Received: 29 August 2007 /Accepted: 18 October 2007 /Published online: 27 November 2007# Springer Science + Business Media, LLC 2007

AbstractIntroduction Transvenous ventricular pacing leads across thetricuspid valve may cause or exacerbate tricuspid regurgita-tion (TR). The literature in adults is inconclusive and nostudies have investigated the association between pacing leadsand TR in children or congenital heart disease patients.Methods and results A retrospective chart review was con-ducted at a large children’s hospital, yielding 123 patientswith initial placement of a transvenous lead across theirtricuspid valve that had adequate echocardiographic datafor review. The median age was 16 years (range 2–52) at timeof lead placement. The pre-procedure echo was comparedboth to the first echo after lead placement and the most recentecho. Median time was 242 days from implant to first echo,and 827 days to most recent echo. There was no difference inTR between the pre-procedure echo and first follow-up echo(p=NS). However, TR was more likely to progress mildlybetween the pre-procedure echo and the most recent echo(p<0.02) with a mean increase from 1.54 to 1.69 on a 0 to 4ordinal scale. There were 76 pts (62%) with CHD. Meanpre-procedure TR was 1.82 in right-sided valvular CHD(e.g., tetralogy of Fallot, repaired AV canal) vs. 1.43 withoutright-sided CHD (p<0.01).Conclusions In patients with transvenous ventricular leadsacross the tricuspid valve, echocardiography demonstrates asmall, but statistically significant change in TR. The detectedchange is minimal, suggesting that there is little impact of

transvenous leads on TR, even in growing children orpatients with right-sided structural heart disease.

Keywords Pediatric . Transvenous . Pacemaker .

Implantable cardioverter–defibrillator (ICD) .

Tricuspid regurgitation . Cardiac growth . Echocardiogram

1 Introduction

The first description of endocardial transvenous leadplacement was written in 1959 [1]. The procedure becamewidespread shortly thereafter and since then, the prevalenceof intracardiac leads has grown enormously. Although themajority of leads are still placed in adults, children representan important subgroup of patients who receive transvenoussystems and consensus indications have been developed forlead placement in pediatrics [2]. As the annual number of pe-diatric pacemaker and implantable cardioverter–defibrillatorleads increases, the long-term effects of these leads inpediatrics merit further investigation. With the exception ofrare anatomy with ventricular transposition or leads placedin the coronary sinus, all intracardiac ventricular leads crossthe tricuspid valve (TV). However, despite the increasingnumber of patients who live with leads across their valve,the literature in adults is inconclusive on whether theseintracardiac leads cause clinically relevant tricuspid regur-gitation (TR).

Case reports in adults have noted severe TV regurgita-tion with an intracardiac lead across the valve [3, 4]. Aretrospective study of 374 adults after transvenous pace-maker lead placement across the TV showed increasedprevalence of significant TR when compared with age andsex-matched controls [5]. However, one prospective study

J Interv Card Electrophysiol (2008) 21:65–68DOI 10.1007/s10840-007-9183-0

G. Webster :R. Margossian :M. E. Alexander : F. Cecchin :J. K. Triedman : E. P. Walsh : C. I. Berul (*)Department of Cardiology, Children’s Hospital, Boston,300 Longwood Ave.,Boston, MA 02115, USAe-mail: Charles.Berul@cardio.chboston.org

suggested that transvenous pacemaker leads do not exacer-bate TR [6].

There is a risk of tricuspid regurgitation in children bymechanisms similar to those postulated in adults. Thesemechanisms include lead adherence to the valve tissue,laceration of valve tissue and changes in ventricular activationtiming causing poor closure of the tricuspid valve [4–6]. Inaddition, normal growth in children changes the heart’s size.Adhesions of the lead to the valve apparatus may cause orexacerbate tricuspid regurgitation in patients who have hadtransvenous pacemaker leads placed before their heart isfully grown. Finally, children have a much higher prevalenceof congenital heart disease than adults. The combination ofright-sided congenital heart disease and lead-induced tricus-pid regurgitation has not been published in the literature.

Transthoracic echocardiography has been the primarymodality for quantifying tricuspid regurgitation in childrenfor over 20 years [7]. It has been previously described inadults in the setting of transvenous pacemaker leads [4–6].The purpose of this study was to determine whethertransvenous pacemaker leads are associated with increasedechocardiographic findings of tricuspid regurgitation.

2 Methods

2.1 Study design

The study was a retrospective chart review at a tertiarychildren’s hospital. We reviewed all patients at our hospitalthat had a transvenous lead placed across their tricuspidvalve from January 1, 2000 until December 31, 2006. Fromthis group, patients qualified for inclusion if they had at leastone transthoracic echocardiographic study (“echo”) beforethe placement of their transvenous lead and at least one echoafterwards. Although evaluating TR is a standard componentof our echo protocol, patients that did not have tricuspidregurgitation evaluated by echo on at least one pre-procedureecho and at least one post-procedure echo were excluded.

2.2 Data collection

Permission for data review was obtained from our Institu-tional Review Board. The paper and electronic medicalrecords, internal departmental charts, and departmentaldatabases were cross-referenced for each patient. Informa-tion on clinical details, lead placement, echocardiograms,and electrocardiograms were tabulated for review.

All echocardiograms at our institution have a final reporton file that grades the severity of tricuspid regurgitation asan ordinal variable: 0—Absent, 1—Trivial, 2—Mild, 3—Moderate, 4—Severe. The date of lead placement wasidentified for each patient and the pre-procedure echo that

was closest in time to that echo was identified. The firstpost-procedure echo was also identified. Finally, the mostrecent echo in our database was identified. A TR value wasobtained from each of these echos based on the scale notedabove. A subset of 10% of all echos evaluated in our studywas independently reviewed by a blinded pediatric echo-cardiographer (RM) to ensure internal consistency.

All electrocardiograms (ECGs) at our institution have afinal report on file. From those final reports, the ECGs can becategorized as demonstrating native AV conduction, dual-chamber pacing, ventricular-pacing, or an intrinsic ventricularrhythm. For each echocardiogram included in our study, weevaluated the ECG obtained at or near the time of echo. EachECG has been read by a pediatric cardiologist and a 10%subset was independently reviewed by a blinded pediatricelectrophysiologist (CIB) to ensure internal consistency.

Height and weight were gathered from the clinical dataon the echocardiogram report. If height and weight werenot available, the clinical chart was examined for a recordedheight and weight near that time. Rarely, a height was notrecorded at a clinic or echo visit. In these cases, heightswere tabulated from other visits and data was extrapolatedusing standard growth charts available from the Centers forDisease Control and Prevention [8]. Body surface area(BSA) was calculated using the Mosteller formula [9].

2.3 Statistical analysis

Because regurgitation is an ordinal variable, changes wereevaluated over time using the nonparametric Wilcoxonsigned-rank test. A paired t-test was used to compareheight, weight and BSA. For comparisons of TR betweenmultiple groups, the Kruskal-Wallis test was applied.

3 Results

Patient demographics At our institution, 1106 patients hadpacemakers or ICDs placed between 2000 and 2006. Ofthose, 123 patients underwent an initial transvenousplacement of a pacemaker or ICD lead across their tricuspidvalve without previous leads in place across the TVand metthe predefined echocardiography inclusion criteria. Themedian age was 16 years at time of lead placement with arange from 2 years to 52-years-old. ICD leads were placedin 68 of these patients (55%) and pacemaker leads wereplaced in 55 patients. Only 2 of these pacemaker leads werepart of a biventricular lead system (1.6% of all procedures).The median time of follow-up was 242 days from theimplantation procedure to the first echo and 827 days fromthe implantation procedure to the most recent echo. Table 1shows median values for height, weight and BSA at thetime of lead implantation and at the time of the last echo.

66 J Interv Card Electrophysiol (2008) 21:65–68

Echocardiography results There was no difference in TRbetween the pre-procedure echo and first follow-up echo(p=NS). However, TR was more likely to progress betweenthe pre-procedure echo and the most recent echo, with amean increase from 1.54 to 1.69 on the 0 to 4 ordinal scale(p<0.02). There is no difference in mean duration of follow-up for those whose TR gets worse versus those whose TRstays the same or gets better (p=NS). There is no differencein TR when ICD or pacemaker leads are consideredindependently. No patients with moderate or severe TR attheir final echo had significant right ventricular hypertension.

The TR increased by two ordinal levels in four patients(3%). In all four patients, it increased from absent TR tomild TR. The TR increased by one ordinal level in 27patients (22%). Of these 27, the TR changed from absent totrivial in 1 patient, from trivial to mild in 18 patients andfrom mild to moderate in eight patients. There were nopatients whose TR increased to severe.

The TR was unchanged in 77 patients (63%). Of thisgroup, 1 patient had no TR, 29 patients had trivial TR and47 patients had mild TR.

In 15 patients (12%), the TR improved. Three patients(2.4% of all patients) improved by two ordinal levels, allfrom mild TR to absent TR. Twelve patients improved byone ordinal level. Of these, 10 improved from mild totrivial TR, 1 improved from moderate to mild TR, and 1improved from severe to moderate TR.

Potential impact of age on tricuspid regurgitation Age didnot predict an increase in regurgitation. Patients who wereunder the age of 18 showed no significant differencebetween their pre-procedure echo, where the TR had amean of 1.47, versus their last echo where the mean was1.63 (p=NS). Patients age 18 or older also showed nosignificant difference with a pre-procedure mean TR of1.64 and a mean TR on the last echo of 1.76 (p=NS).

Potential impact of congenital heart disease on tricuspidregurgitation There were 76 patients (62%) with congenitalheart disease (CHD). Of those, 35 patients (46%) had right-sided disease, consisting of 27 patients with tetralogy ofFallot, 6 patients with an atrioventricular canal, and one

patient with Ebstein’s anomaly of the tricuspid valve. Thisgroup also contained one patient with L-transposition of thegreat vessels {S,L,L}, status post a Senning procedure andRV to PA conduit (i.e. “double-switch”). With this repair,the patient had a systemic mitral valve and a ventricularlead across the tricuspid valve. Before the procedure,patients with right-sided CHD had a mean pre-procedureTR of 1.82, compared to 1.43 in those patients who did nothave right-sided CHD (p<0.01). In patients with right-sidedCHD, the mean TR did not change significantly from thepre-procedure echo to the last echo (1.82 vs 1.94, p=NS);however, in patients who did not have right-sided CHD, theTR increased from 1.43 to 1.59 (p=0.04) from the pre-procedure echo to the last echo. There was no significantchange in TR from the pre-procedure echo to the firstfollow-up echo in any subgroup.

Impact of atrioventricular conduction and synchrony ontricuspid regurgitation At the time of the pre-procedureechocardiogram, 92 patients (75%) were conducting acrossthe intrinsic AV node, 21 patients (17%) were being dual-chamber paced and 10 patients (8%) were either ventricular-paced or their baseline rhythm was an intrinsic ventricularrhythm. All patients who were paced at the pre-procedureECG were paced with epicardial systems. At the time of thelast echocardiogram, 79 patients (64%) were conductingacross their own AV node, 35 patients (28%) were beingdual-chamber paced and 9 patients (7%) were eitherventricular-paced or their baseline rhythm was an intrinsicventricular rhythm. A Kruskal–Wallis test was not signif-icant for differences between patients with intrinsic con-duction, dual-chamber pacing or ventricular rhythms whencomparing the pre-procedure TR to the first post-procedureTR (p=NS) or when comparing the pre-procedure TR tothe most recent echo TR (p=NS).

4 Discussion

In patients with transvenous ventricular leads across thetricuspid valve, early echocardiography, with a medianfollow-up of less than a year, demonstrates no change in TR

Table 1 Median height, weight and body surface area (standard deviation in parenthesis)

Age (in years) Height (cm) Weight (kg) Body surface area (m2)

At time ofprocedure

At last echo p valuea At time ofprocedure

At last echo p valuea At time ofprocedure

At last echo p valuea

<18 154 (27.3) 165.5 (24.8) <0.01 48.2 (23.7) 58 (25.1) <0.01 1.48 (0.47) 1.63 (0.46) <0.01≥18 168 (11.6) 168.5 (11.6) 0.04 66 (15.9) 67.8 (17.2) <0.01 1.76 (0.22) 1.78 (0.24) 0.01All ages 164 (23.6) 167 (20.6) <0.01 59.1 (23.1) 63.5 (23.1) <0.01 1.66 (0.43) 1.72 (0.40) <0.01

a Two-tailed paired t test used for comparisons

J Interv Card Electrophysiol (2008) 21:65–68 6767

after lead placement. However, after extended follow-upwith a median of more than two years, the increase in TRis statistically significant. On the ordinal scale of 0 to 4,the pre-procedure TR had a mean of 1.54, which isbetween the trivial and mild scores. At our last follow-up,the mean score had increased to 1.69, which is still in thatsame range of trivial to mild. This is not likely to beclinically significant. There were no extreme changes fromabsent or trivial TR to moderate or severe TR in any ofour patients.

One reason we hypothesized why TR may increase inchildren is because children grow after transvenous leadplacement [10, 11]. In the present cohort, overall change inheight was 9.5 cm from lead implantation to the last echo;however, in patients less than 18 years of age at the time oflead placement, the median change in height was 10.9 cm.In patients who were 18 or older at the time of lead place-ment, the median change was only 3 cm. Not surprisingly,this suggests that most of the change in height in ourpopulation occurred in those patients who were under 18 atthe time their lead was placed. The changes in weight andBSA similarly reflect more growth before age 18. However,despite evidence for growth, the subset of patients less than18-years-old were not more likely to have an increase in theTR jet. This suggests that linear growth does not play asubstantial role in the development of TR in young patientswho have intracardiac leads across the tricuspid valve.

Patients with right-sided congenital heart disease havemore TR prior to receiving a transvenous lead that thosewho have no right-sided congenital heart disease, althoughthe mean for both was within the trivial-to-mild range.Despite theoretical concerns that right-sided heart diseasemight exacerbate TR progression, this did not appear to beclinically significant. Patients without right-sided diseasedid develop slightly more tricuspid regurgitation after leadplacement; however, in this group the difference was slight,changing from a mean of from 1.43 to a mean of 1.59, bothwell within the trivial-to-mild range. Even after leadplacement for a prolonged period of time, the patientswithout right-sided disease had less mean TR than thepatients with right-sided disease did at any point.

It has been hypothesized that ventricular pacing and thesubsequent changes in ventricular activation timing causepoor closure of the tricuspid valve. In our study, wecompared three groups to each other: those patients whohad intrinsic AV node conduction, those who were dual-chamber paced, and those who had ventricular pacing or anintrinsic ventricular rhythm. Among all of these groups,there were no differences in the TR when the pre-procedureechos, the first post-procedure echos and the last echoswere compared. Therefore, atrioventricular dissociation ordyssynchronous ventricular activation due to chronic

ventricular pacing is not significantly changing the degreeof TR during the follow-up interval in this group.

Overall, pediatric and CHD patients who have trans-venous pacemaker or ICD leads placed across their tri-cuspid valve have a statistically significant increase intricuspid regurgitation; however, there does not seem to bea clinically significant change in tricuspid regurgitationduring the follow-up period. These findings suggest thatthose patients with transvenous leads across the tricuspidvalve who develop a more substantial increase in tricuspidregurgitation should be evaluated carefully for otherpathologies that might exacerbate tricuspid regurgitation.

References

1. Furman, S., & Schwedel, J. (1959). An intracardiac pacemaker forStokes–Adams seizures. New England Journal of Medicine, 261,943–948.

2. Gregoratos, G., Abrams, J., Epstein, A. E., Freedman, R. A.,Hayes, D. L., Hlatky, M. A., et al. (2002). American College ofCardiology/American Heart Association Task Force on PracticeGuidelines American College of Cardiology/American HeartAssociation/North American Society for Pacing and Electrophysi-ology Committee. ACC/AHA/NASPE 2002 guideline update forimplantation of cardiac pacemakers and antiarrhythmia devices:Summary article: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines.Journal of Cardiovascular Electrophysiology, 13, 1183–1199.

3. Gibson, T. V., Davidson, R. C., & DeSilvey, D. L. (1980).Presumptive tricuspid valve malfunction induced by a pacemakerlead: a case report and review of the literature. PACE, 3, 88–93.

4. Iskandar, S. B., Jackson, S. A., Fahrig, S., Mechleb, B. K., &Garcia, I. D. (2006). Tricuspid lead malfunction and ventricularpacemaker lead: case report and review of the literature.Echocardiography, 23, 692–696.

5. Paniagua, D., Aldrich, H. R., Lieberman, E. H., Lamas, G. A., &Agatston, A. S. (1998). Increased prevalence of significanttricuspid regurgitation in patients with transvenous pacemakersleads. American Journal of Cardiology, 82, 1130–1132 (A9).

6. Leibowitz, D. W., Rosenheck, S., Pollak, A., Geist, M., & Gilon,D. (2000). Transvenous pacemaker leads do not worsen tricuspidregurgitation: a prospective echocardiographic study. Cardiology,93, 74–77.

7. Stevenson, G. J. (1984). Experience with qualitative and quanti-tative applications of Doppler echocardiography in congenitalheart disease. Ultrasound in Medicine & Biology, 10, 771–796.

8. Kuczmarski, R. J., Ogden, C. L., Grummer-Strawn, L. M., Flegal,K. M., Guo, S. S., Wei, R., et al. (2000). CDC growth charts:United States. Advance Data, 314, 1–27.

9. Mosteller, R. D. (1987). Simplified calculation of body-surfacearea. New England Journal of Medicine, 317, 1098.

10. Fortescue, E. B., Berul, C. I., Cecchin, F., Triedman, J. K., Walsh,E. P., & Alexander, M. E. (2004). Patient, procedural, and hardwarefactors associated with pacemaker lead failures in pediatrics andcongenital heart disease. Heart Rhythm, 1, 150–159.

11. Bar-Cohen, Y., Berul, C. I., Alexander, M. E., Fortescue, E. B.,Walsh, E. P., Triedman, J. K., et al. (2006). Age, size, and lead factorsalone do not predict venous obstruction in children and young adultswith transvenous lead systems. Journal of Cardiovascular Electro-physiology, 17, 754–759.

68 J Interv Card Electrophysiol (2008) 21:65–68