The extracardiac conduit (ECC) modification of the Fontan procedure has been theorized to reduce the risk of sinus node dysfunction and atrial arrhythmia compared with the intra-atrial lateral tunnel (ILT) Fontan. This study aimed to compare the prevalence of early and late arrhythmias in patients who underwent ECC and ILT Fontan from a similar era with long-term follow-up at a single institution. A retrospective cohort study was conducted of all patients who underwent ECC or ILT Fontan from 1995 to 2005 at The Children’s Hospital of Philadelphia. Bradyarrhythmias (including sinus node dysfunction), tachyarrhythmias, and pacemaker burden prevalence was determined throughout early (<30 days) and late (>30 days) postoperative periods. Of 434 patients undergoing the Fontan procedure during the study period, a total of 87 and 106 patients who underwent ECC and ILT Fontan, respectively, met the inclusion criteria. There were no significant differences in risk of sinus node dysfunction or tachyarrhythmia in both early and late postoperative periods. Although the overall risk of late postoperative pacemaker therapy was lower for the ECC cohort (4.9% vs 15.7%, p = 0.03), when adjusting for follow-up time, no significant difference was observed (odds ratio 3.1, 95% confidence interval 0.6 to 15.2, p = 0.16). In conclusion, the overall prevalence of late postoperative arrhythmias observed after contemporary Fontan modifications is low. Intra-atrial reentrant tachycardia, a potentially fatal complication of the atriopulmonary Fontan operation was infrequently encountered in both ECC and ILT Fontan cohorts. Pacemaker use was higher in the ILT group, although this difference may be explained by differences in follow-up time. Despite the low prevalence of arrhythmias after contemporary Fontan modifications, ongoing surveillance is warranted as the onset of arrhythmias may emerge after longer follow-up time.
Postoperative arrhythmias after the Fontan operation have been described since the procedure was introduced in 1971. Currently, the 2 most commonly employed surgical forms of the total cavopulmonary connection are the intra-atrial lateral tunnel (ILT) and the extracardiac conduit (ECC). For more than a decade, the ECC connection has been the preferred Fontan modification in several institutions, yet comparative data on the incidence of arrhythmias after ILT and ECC have been limited and controversial. The ECC Fontan procedure has been theorized to reduce the risk of sinus node dysfunction (SND) and atrial arrhythmia compared with ILT Fontan because of the avoidance of extensive atrial suture lines and exclusion of the atrial chamber from elevated systemic venous pressure. Yet previous injury of the crista terminalis, large atrial wall incisions at the time of intracardiac procedures, and multiple-stage operations may predispose patients to atrial arrhythmias after the ECC Fontan procedure as well. Recent attempts to define rhythm status in this population have included patients with the traditional atriopulmonary form of the Fontan procedure. Yet results from studies such as the Pediatric Heart Network Fontan Cross-Sectional Study demonstrate a disproportionate arrhythmia burden for this older subset of patients who underwent the Fontan procedure, thereby making results difficult to interpret in the current era. Although additional investigations have excluded patients who underwent traditional atriopulmonary Fontan, these studies are also limited in their lack of sufficient follow-up time and insufficient statistical power for identification of arrhythmia predictors. From 1995 to 2005, The Children’s Hospital of Philadelphia cardiothoracic surgical experience evolved to include an overall balance of ECC and ILT modifications, performed in parallel, without using the traditional atriopulmonary Fontan procedures. We sought to better define the prevalence of arrhythmias in this population while comparing differences in arrhythmias and pacemaker burden between the 2 most commonly employed modifications of the Fontan procedure in the current era.
Methods
The medical and surgical records of all patients undergoing Fontan palliation from January 1, 1995 to December 31, 2005 at The Children’s Hospital of Philadelphia were reviewed retrospectively. The Institutional Review Board approved this study. Inclusion criteria for this study included (1) presence of the ILT or ECC Fontan procedure, (2) follow-up at The Children’s Hospital of Philadelphia, and (3) electrocardiographic testing within 2 years before the termination of data collection. Additionally, patients were required to have at least one 15-lead electrocardiogram before Fontan procedure, 1 in the early postoperative period (<30 days) and 1 in the late postoperative period (>30 days). Supporting electrocardiographic data in the form of Holter monitoring was additive; Holter monitoring was not an inclusion requirement. Exclusion criteria included (1) presence of arrhythmia before ECC or ILT Fontan, (2) previous atriopulmonary Fontan procedure, (3) concomitant arrhythmia surgery at the time of ECC or ILT Fontan procedure, and (4) any patient undergoing transplantation during the follow-up period. Age at Fontan procedure and time from Fontan to the most recent outpatient clinical appointment were evaluated. Additional clinical variables obtained from chart review included ventricular morphology (e.g., right ventricle dominant vs non–right ventricle dominant) and presence of fenestration at the time of Fontan procedure. All available electrocardiographic records and 24- or 48-hour ambulatory Holter monitor recordings were analyzed retrospectively. Any impairment of rhythm origin or conduction was assessed as an arrhythmia. The following groups of rhythm disturbances were defined as bradyarrhythmias: (1) SND, which includes sinus bradycardia, ectopic atrial rhythm or bradycardia, predominant junctional rhythm, or sinus pauses exceeding 2 seconds, and (2) complete heart block. The following groups of rhythm disturbances were defined as tachyarrhythmias: (1) supraventricular tachycardia, which included atrial fibrillation, atrial flutter, ectopic atrial tachycardia, junctional ectopic tachycardia, or atrioventricular (AV) reciprocating tachycardia, and (2) ventricular arrhythmias including ventricular tachycardia and ventricular fibrillation. Rhythm disturbances documented during the early postoperative period (<30 days) were defined as early-onset postoperative arrhythmias, and rhythm disturbances documented during the late postoperative period (>30 days) after Fontan operation were considered late arrhythmias.
Data are expressed as mean ± SD for normally distributed continuous variables, median (range) for skewed continuous variables, and count (percentage of total) for categorical variables. Testing of differences in demographic and clinical data based on Fontan type (ECC vs ILT) was accomplished with either unpaired student t test or Wilcoxon rank sum test for continuous variables and with either Pearson’s chi-square test or Fisher’s exact test for categorical variables, as appropriate. Measures of association between potential predictor variables and the primary outcome variables were determined first by univariate logistic regression. Covariates with p <0.2 in univariate testing were considered for inclusion in a multivariate model to identify factors independently associated with outcomes. Covariates were retained in the final multivariate model if the p value was <0.05 or if they showed evidence for significant confounding or effect modification. Kaplan-Meier event-free analysis was used to compare freedom from important arrhythmia outcomes between Fontan types. Statistical significance was established using a 2-tailed p value of <0.05. All statistical analyses were performed using Stata v10 (StataCorp., College Station, Texas).
Results
From January 1, 1995 to December 31, 2005, a total of 434 patients underwent Fontan palliation at our institution, with 193 patients (44%) meeting study inclusion criteria (87 ECC [45%] vs 106 ILT [55%]). Patients excluded from this study were those not monitored by our institution (n = 192, 44%), those who underwent Fontan revision (n = 28, 6.4%), those deceased at the time of data collection (n = 11, 2.5%), those with preoperative arrhythmias or pacemakers (n = 9, 2.1%), and those who underwent cardiac transplantation (n = 1). Differences in demographic and clinical variables based on Fontan type are summarized in Table 1 . There were no major differences in the lead surgeons for the 2 Fontan groups. Patients who underwent ILT and ECC Fontan underwent Fontan completion at early ages during the 1995 to 2000 era (n = 109): ILT (n = 76) median age 1.81 years (range 1.14 to 12.36) and ECC (n = 33) median age 2.06 years (1.01 to 10.95), p = 0.03. In comparison, age at Fontan procedure during the 2001 to 2005 era (n = 84) was later for both ILT (n = 30, median 2.27 years [1.5 to 3.84]) and ECC (n = 54, median 2.47 [1.48 to 4.66]), p = 0.24.
| Characteristic (n = 193) | ECC (n = 87) | ILT (n = 106) | p Value |
|---|---|---|---|
| Men | 51 (59) | 67 (63) | 0.52 |
| Age at Fontan (yrs) | 2.4 (1.0–10.95) | 1.9 (1.1–12.4) | 0.0002 |
| Follow-up after Fontan (yrs) | 7.1 (1.1–15) | 10.5 (1.7–15.8) | 0.0001 |
| Ventricular morphology | |||
| Left ventricular | 32 (37) | 23 (22) | 0.07 |
| Right ventricular | 52 (60) | 78 (74) | |
| Mixed | 3 (4) | 5 (5) |
The prevalence of early postoperative arrhythmia and differences based on Fontan type are summarized in Table 2 . The prevalence of SND in the early postoperative period was 5.2% (10 of 193) for the entire cohort with no statistically significant difference observed between ECC and ILT groups (4.6% ECC vs 5.7% ILT, p = 0.7). Early tachycardias were observed in 4.7% (9 of 193) of the overall cohort with a greater prevalence in the ECC group, although this difference did not attain statistical significance (8% ECC vs 1.9% ILT, p = 0.06). Junctional ectopic tachycardia was the most common tachycardia observed in the early postoperative period with 83% of all episodes of junctional ectopic tachycardia occurring in the ECC Fontan cohort. Only 2 patients required pacemaker implantation during the early postoperative period (1 ECC vs 1 ILT, p = 0.89).
| Variable | Overall, n = 193 (%) | ECC, n = 87 (%) | ILT, n = 106 (%) | ILT vs ECC, OR (95% CI) | p Value |
|---|---|---|---|---|---|
| Early arrhythmia or pacemaker | |||||
| All arrhythmia | 21 (11) | 11 (13) | 10 (10) | 0.72 (0.29–1.8) | 0.48 |
| Bradyarrhythmia | 12 (6) | 4 (5) | 8 (8) | 1.69 (0.49–5.8) | 0.40 |
| SND | 10 | 4 | 6 | 1.25 (0.4–4.2) | 0.7 |
| Complete heart block | 2 | — | 2 | ||
| Tachyarrhythmia | 9 (5) | 7 (8) | 2 (2) | 0.22 (0.04–1.09) | 0.06 |
| AV nodal reentrant tachycardia | 1 | — | 1 | ||
| Ectopic atrial tachycardia | 1 | 1 | — | ||
| Junctional ectopic tachycardia | 6 | 5 | 1 | ||
| Ventricular tachycardia | 1 | 1 | — | ||
| Pacemaker | 2 (1) | 1 (1) | 1 (1) | 0.82 (0.05–13.3) | 0.89 |
| Late arrhythmia or pacemaker | |||||
| Any arrhythmia | 60 (31) | 26 (30) | 34 (32) | 1.11 (0.6–2.1) | 0.74 |
| Bradyarrhythmia ∗ | 58 (30) | 25 (29) | 33 (31) | 1.12 (0.63–2.17) | 0.72 |
| SND | 58 | 25 | 32 | ||
| Complete heart block | 1 | — | 1 | ||
| Tachyarrhythmia ∗ | 7 (4) | 3 (3) | 4 (4) | 1.1 (0.24–5) | 0.9 |
| AV nodal reentrant tachycardia | 2 | 1 | 1 | ||
| Intra-atrial reentrant tachycardia | 5 | 2 | 3 | ||
| Junctional ectopic tachycardia | 1 | — | 1 | ||
| Pacemaker | 14 (7) | 2 (2) | 12 (11) | 5.4 (1.2–24.9) | 0.03 |
The prevalence of late postoperative arrhythmias and differences based on Fontan type are summarized in Table 2 . Univariate analysis revealed a higher risk of late pacemaker implantation in the ILT group compared with the ECC Fontan group (2.3% ECC vs 11.3% ILT, p = 0.03). Two patients underwent early pacemaker placement at 7 and 17 days after Fontan procedure with an additional 14 patients requiring pacemaker during the late postoperative period. Of those patients undergoing late pacemaker implantation, the median time from Fontan procedure to pacemaker placement was 2.01 years (range 0.4 to 11.52) with SND as the most common indication for pacemaker implantation.
Potential risk factors for arrhythmias and pacemaker implantation in the late postoperative period were identified by univariate logistic regression (see Table 3 ). Although duration of follow-up from Fontan operation and Fontan type were identifiable risk factors for pacemaker implantation in the late postoperative period on univariate logistic regression, only duration of follow-up (odds ratio 1.29, 95% confidence interval 1.06 to 1.57, for every year of follow-up, p = 0.01) remained an independent predictor of late pacemaker utilization on multivariate testing.
