Introduction
Univentricular hearts with total cavopulmonary circulation (TCPC) may present challenges
during transvenous pacing owing to limited venous access to the right side of the
heart. Most patients undergo surgical palliation during early years of life, which
enables shunting of blood to the pulmonary circulation. Several different iterations
of this operation include Glenn, bidirectional Glenn, hemi-Fontan, classic and modified
atriopulmonary Fontan, and extracardiac and lateral tunnel (TCPC). The overall incidence
of sinus node dysfunction (SND) in the immediate postoperative period is higher in
lateral tunnel total cavopulmonary connection (TCPC) compared with extracardiac tunnel
TCPC (11% vs 28%); however, the longer-term clinical outcomes showed similar prevalence
with both techniques after 1 year.1, 2, 3
AAI pacing is usually the preferred transvenous pacing option for patients with SND;
however, the varied cardiac anatomy and surgical shunts may add a layer of complexity
when considering pacing in this population.
Case report
A 36-year-old male patient with congenital heart disease presented with exertional
breathlessness and peripheral edema over the course of the preceding 6 weeks. His
cardiac anatomy consisted of situs solitus, absent left atrioventricular (AV) valve
connection, hypoplastic left ventricle, double-outlet right ventricle (RV) with an
anterior aorta, and a systemic RV. He underwent pulmonary artery (PA) banding at 6
months, a classic Glenn shunt at 7 years of age, and an inferior vena cava (IVC) to
left PA procedure (lateral tunnel TCPC) at 12 years old (Figure 1).
Figure 1
A: Cardiac magnetic resonance imaging showing Fontan connections: dilated inferior
vena cava (IVC) connected to an intra-atrial tunnel (white arrow) to the left pulmonary
artery (LPA) only. The superior vena cava (SVC) connects directly to the right pulmonary
artery (RPA; yellow arrow). B: Diagram of patient anatomy: Fontan operation (lateral
tunnel). The total cavopulmonary connection circulation is created by connecting the
IVC to the pulmonary artery (PA) with an intra–right atrial conduit, to separate it
from the common atrium. A Glenn shunt (black arrow) is created by connecting the SVC
to the RPA. AO = aorta; LV = left ventricle; PV = pulmonary vein; RV = right ventricle.
C: Electrocardiogram (ECG) post ablation 1 year earlier. D: ECG on admission showing
junctional rhythm.
His comorbidities included type 1 diabetes, chronic kidney disease, and diabetic retinopathy.
He had a history of paroxysmal atrial tachycardia resulting in decompensated heart
failure and previously underwent catheter ablation at the age of 29, followed by a
redo procedure 6 years later for recurrence. During the first procedure, diagnostic
catheters were placed within the tunnel (IVC–PA) and burst stimulation performed to
induce atrial tachycardia. This was hemodynamically poorly tolerated and therefore
overdrive pacing was performed to convert the patient back to sinus rhythm. Remote
magnetic navigation was used (Niobe, Stereotaxis Inc, St. Louis, MO) alongside the
CARTO 3 system for 3-dimensional visualization (Biosense Webster, Irvine, CA). As
femoral venous access would lead through the TCPC tunnel to the PA, access to the
common atrium was only possible via a retrograde aortic approach using magnetic navigation.
Atrial tachycardia of the same cycle length was induced again and was mapped rapidly.
The local activation mapping was performed in both the tunnel of the TCPC and the
common atrium. This demonstrated a focal area of firing from the mid portion of the
crista terminalis (located within the tunnel of the TCPC) and radiofrequency ablation
from this point terminated the tachycardia with the first energy application. No further
tachycardias were inducible afterwards.
Redo ablation 6 years later demonstrated low-voltage tissue in the lateral aspect
of the residual right atrium (RA) / tunnel from the TCPC. Atrial tachycardia was readily
inducible from burst pacing within the TCPC and activation mapping demonstrated a
macro-reentry mechanism with a circuit rotating around the right AV valve. Therefore,
a linear inferior isthmus line was performed leading to tachycardia termination during
ablation. The patient was discharged home in sinus rhythm with normal AV conduction
(Figure 1C) and had no arrhythmia recurrence afterwards.
He subsequently presented after almost 1 year since the ablation with clinical features
of congestive cardiac failure. A 12-lead electrocardiogram demonstrated a junctional
bradycardia with heart rate of 45 beats per minute (bpm) (Figure 1D). Ward telemetry
showed his heart rate consistently at 40–45 bpm; this was a junctional rhythm with
minimal heart rate variation. He was highly symptomatic from simply mobilizing on
the ward and therefore was unable to undergo cardiopulmonary exercise testing. It
was felt that his junctional bradycardia had contributed to his clinical decompensation
and following diuresis, he was consented for a single-chamber pacemaker implantation.
Owing to the lack of superior central venous access, a femoral approach was chosen.
The right femoral vein was cannulated and a wire inserted up to the tunnel via the
IVC. A venogram demonstrated a lateral tunnel communicating with the left PA (Figure 2A).
During the pacemaker implantation, his heart rate was 40–45 bpm in junctional rhythm.
The atrial lead was manipulated and placed into the native lateral wall of the RA.
Interestingly, normal sinus node activity was sensed at 70 bpm (Figure 2, yellow and
orange arrows). However, owing to prior congenital heart surgery and the prior ablation
lesions, the native RA appeared electrically bisected. Consequently, impulses from
the sinus node were prevented from conducting through to the septal low RA (Figure 2D).
When pacing inferoseptal to this area of block, local nodal capture was demonstrated
with an intact AV conduction (with a delay of 120 ms, Figure 3B).
Figure 2
Procedure summary. A: Venogram of right atrial (RA) tunnel to pulmonary artery (PA).
B: Electrograms during procedure: yellow arrow shows the high RA signals (atrial sense
“AS,” sinus rhythm), green arrow shows trace coming from the lower RA lead showing
only right ventricle far field, orange arrows show surface electrocardiogram. C,D:
Final position in left (C) and right (D) anterior oblique projection. E: Diagram showing
the position of both leads within the total cavopulmonary connection tunnel (black
arrow) connected to the left PA (LPA) and Glenn shunt connecting the superior vena
cava (SVC) and the right PA (RPA) branch. Red arrow shows line of previous atriotomy
and blue arrow shows position of the inferior ablation line. IVC = inferior vena cava.
Figure 3
Postoperative pacing check: A: During upper “right atrial (RA) pacing” (upper RA lead,
EGM1) threshold test, there is 1:1 atrioventricular (AV) conduction but with long
PR intervals. Of note, there is an absence of signals on the lower RA lead (EGM2).
B: During lower “RA pacing” (EGM2) threshold test, there is 1:1 AV conduction with
comparatively shorter and more physiological PR intervals. Interestingly, there is
evidence of retrograde conduction to the upper RA lead (“AS” signals on lead III and
corresponding electrogram on EGM1). C: Electrocardiogram before discharge.
During the case when positioning the first atrial lead in the superoposterior TCPC
tunnel, there was predominantly no evidence of “AV” conduction. There was, however,
occasional “AV” conduction with a very long AV delay. This was the reason for placing
a second lead, which when positioned inferiorly in the TCPC tunnel showed 1:1 AV conduction
when tested. Postoperatively on a pacemaker check, when the upper atrial lead was
retested, there was more abundant evidence of AV conduction, albeit with a very long
AV delay (Figure 3).
Atrial electrograms from the lower right atrial lead at the beginning of the procedure
showed far-field junctional or ventricular signal matching the patient’s junctional
rhythm (Figure 2, green arrow). We therefore had 1 upper RA/TCPC tunnel lead (85 cm
active fixation CapSureFix Novus MRI SureScan 5076; Medtronic, Minneapolis, MN) for
sensing of the sinus node activity and a lower RA/TCPC tunnel lead (65 cm active fixation
CapSureFix Novus MRI SureScan 5076; Medtronic) to bypass the intra-atrial block and
facilitate consistent 1:1 AV conduction. The upper lead was attached to the RA port
and the lower lead to the RV port (Figure 2C and D). Lead parameters were acceptable:
upper lead threshold 0.5 V at 0.4 ms, amplitude 0.8 mV, and impedance 437 ohm; and
lower lead threshold 1.5 V at 1 ms, impedance 304 ohm. A subcutaneous pocket was fashioned
in the right groin and the device was programmed DDDR 70 bpm. Of note, the AV delay
was intentionally set to a very short 30 ms, in order to facilitate intrinsic AV conduction.
Over the course of a few days, the patient lost over 10 kg with diuretics, renal function
stabilized, and he was discharged home a few days later. Device check in the pacing
clinic showed upper atrial pacing 44% and lower atrial pacing 100%. This pacing set-up
facilitated sensing of intrinsic sinus node activity and preserving of AV conduction,
and we felt this was preferable and more physiological than rate response single-chamber
mode pacing in a complex congenital patient with a univentricular heart.
Discussion
This case is an important description of dual-site atrial pacing to overcome anatomical
and electropathophysiological block in a surgically corrected univentricular heart.
Anatomical depiction of the conduction system in the normal heart have been well described
by James.
4
These sino-nodal connections course in the interatrial septum and are organized as
3 bundles: anterior, median, and posterior. This last one, corresponding to the crista
terminalis, is detached from the posterior end of the sinus node, drops vertically
by forming the posterior edge of the interatrial septum, and ends in the posterior
section of the AV node. The anterior and median bundles are situated anterior to the
fossa ovalis. Analysis of atrial tissue collected during cardiac surgery has shown
the development of collagen fibrosis in the extracellular compartment alongside intracellular
changes (glycogen deposits, myofibril destruction, sarcoplasmatic reticulum abnormalities)
among patients with evidence of intra-atrial conduction block.
5
Complex atrial scars and abnormal atrial wall stress after classic Fontan atriopulmonary
connection led to a high prevalence of atrial arrhythmias. The intra-atrial baffling
procedure, such as the lateral tunnel, was consequently developed to try and obviate
this problem.
6
However, this technique also involves extensive suture lines in the atrium and has
the potential to cause atrial distension, which can still lead to SND and atrial tachyarrhythmia.
7
The extracardiac conduit procedure, the most recent modification of the technique,
leaves the entire atrium at low pressure and avoids significant suture lines.
After intra-atrial baffling, patients often show prolonged P-wave duration
8
with larger dispersion associated with SND, suggesting a propensity to arrhythmia,
although less progressive than that seen in those undergoing classic atriopulmonary
connection. Although patients undergoing extracardiac conduit have a similar prevalence
of SND, prolonged P-wave duration has not been demonstrated in this cohort, suggesting
this surgical approach may be preferable to avoid intra-atrial conduction delay.
9
The hemodynamic sequelae of surgically corrected univentricular hearts are challenging.
Atrial pacing is often required to treat SND and, where possible, efforts should be
made to avoid the deleterious effects associated with long-term right (subpulmonic)
ventricular pacing. Long AV delays can be programmed to facilitate intrinsic conduction;
however, this increase in the total atrial refractory period will naturally limit
the maximum tracking rate of the pacemaker. This limitation is important among young
patients, who may be very active otherwise. Additionally, given the absence of a subpulmonic
pump in Fontan patients, cardiac output is constrained by preload and atrial pacing
at higher rates may not improve cardiac output owing to a concomitant fall in stroke
volume.
Dual-site right atrial pacing has been described in several studies aiming to achieve
better control of the arrhythmia burden in patients with atrial fibrillation, but
it has not been reported in this context. This case illustrates both the underlying
mechanism causing bradycardia and heart failure in surgically corrected univentricular
hearts but also the feasibility and effectiveness of dual-site transvenous pacing.
10
A history of radiofrequency ablations
11
,
12
forming scars inside the atria makes the mechanism described even more likely.
Conclusion
Functional intra-atrial block can be an important electropathophysiological consequence
of surgical correction for patients with a univentricular circulation. This may be
overcome by dual-site right atrial pacing.
Key Teaching Points
•
Patients with surgically corrected univentricular hearts have important hemodynamic
and electropathophysiological sequelae including sinus node disease requiring atrial
pacing.
•
Postsurgical intra-atrial conduction delay or block may be pro-arrhythmic and some
patients may require catheter ablation for tachyarrhythmias.
•
Post–total cavopulmonary connection intra-atrial (tunnel) block may prove challenging
to overcome with single-site (atrial) pacing and markedly prolonged atrioventricular
(AV) delays do not provide favorable hemodynamic effects in patients with univentricular
hearts.
•
Dual-site atrial pacing can circumvent intra-atrial conduction block and facilitate
physiological AV conduction in this patient group.
•
Femorally sited pacemakers are an alternative approach when superior venous access
to the heart is limited.