Author + information
- Received April 2, 2016
- Revision received May 25, 2016
- Accepted June 21, 2016
- Published online September 20, 2016.
- Alexander C. Egbe, MD, MPHa,∗ (, )
- Heidi M. Connolly, MDa,
- Christopher J. McLeod, MB Bch, PhDa,
- Naser M. Ammash, MDa,
- Talha Niaz, MBBSb,
- Vidhushei Yogeswaran, BScc,
- Joseph T. Poterucha, DOd,
- Muhammad Y. Qureshi, MBBSd and
- David J. Driscoll, MDd
- aDivision of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota
- bDepartment of Pediatrics, Mayo Clinic, Rochester, Minnesota
- cMayo Medical School, Mayo Clinic, Rochester, Minnesota
- dDivision of Pediatric Cardiology, Mayo Clinic, Rochester, Minnesota
- ↵∗Reprint requests and correspondence:
Dr. Alexander C. Egbe, Division of Cardiovascular Diseases, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905.
Background There are limited data about the risk of thrombotic and embolic complication (TEC) in adults with atrial arrhythmia after Fontan operation.
Objectives This study sought to determine the risk of TEC in this population and the role of anticoagulation therapy in TEC prevention.
Methods This was a retrospective review of adults with atrial arrhythmia after Fontan operation who were evaluated at the Mayo Clinic between 1994 to 2014. TEC was classified into 2 groups: systemic TEC, defined as intracardiac thrombus, ischemic stroke, or systemic arterial embolus; and nonsystemic TEC, defined as Fontan conduit/right atrial thrombus or pulmonary embolus. Patients were divided into 3 groups: anticoagulation, antiplatelet, and no therapy cohorts.
Results We followed 278 patients, mean age 31 ± 9 years, for 88 ± 14 months (1,464 patient-years). Patient groups included antiplatelet (n = 181), anticoagulation (n = 91), and no therapy (n = 6). There were 97 TEC in 81 patients (29%); 32 were systemic, yielding an event rate of 2.1 systemic TEC per 100 patient-years, and 65 were nonsystemic TEC, yielding an event rate of 4.4 nonsystemic TEC per 100 patient-years. Prevalence of TEC was 18% and 55% at 5 and 10 years, respectively. Atriopulmonary connection was a risk factor for TEC (hazard ratio: 2.31; 95% confidence interval: 1.61 to 4.64), and TEC were associated with higher risk of death and hospitalization (p < 0.0001). Anticoagulation was protective against TEC and resulted in a reduction of TEC risk by 2.5 TEC per 100 patient-years. Anticoagulation was also associated with lower risk of death and hospitalization (p = 0.02). Bleeding complications occurred in 21 (7%) patients and were similar in all groups.
Conclusions Anticoagulation was associated with lower TEC rate and lower risk of death and hospitalization, without a significant increase in bleeding risk. Perhaps anticoagulation should be the preferred preventive strategy.
Atrial arrhythmias occur with a prevalence of more than 50% at 20 years after a Fontan procedure, with the highest risk occurring in the patients with atriopulmonary connection (1,2). In general, the presence of atrial arrhythmia is associated with an increased risk of thrombotic and embolic complications (TEC); this risk is heightened in the patient with history of a Fontan operation (3,4). This is related to the Fontan physiology and its associated low-flow state in the systemic and pulmonary circulation, atrial stasis, elevated central venous pressure resulting in hepatic dysfunction, coagulation abnormalities, ventricular dysfunction, and the common presence of right-to-left shunt (5–7).
The prevalence of TEC after a Fontan operation has ranged from 5% to 33% based on data derived from mixed cohorts of pediatric and adult patients with different rhythm status (8–14). The risk of TEC in adults with atrial arrhythmia after Fontan surgery is unknown, and there is no consensus on the most appropriate preventive strategy. This study sought to determine the risk of TEC in this population and the role of anticoagulation therapy in the prevention of TEC.
We identified adult patients (>18 years) with a history of a Fontan operation and atrial arrhythmia (paroxysmal or persistent) followed at the Mayo Clinic from January 1, 1994 to June 30, 2014. Patients were excluded if they had a history of previous TEC or <12 months of follow-up from the time of atrial arrhythmia diagnosis.
Atrial arrhythmia was defined as atrial flutter/intra-atrial re-entry tachycardia, atrial fibrillation, or ectopic atrial tachycardia documented on electrocardiogram, Holter monitor, event monitor, rhythm strip, and pacemaker/defibrillator intracardiac electrograms. Given the slow atrial re-entry cycle length that is commonly present in this group of patients, it was not possible to differentiate between a focal atrial tachycardia and re-entrant atrial tachycardia/atrial flutter; these diagnoses were pooled. Paroxysmal atrial arrhythmia was defined as atrial arrhythmia <7 days in duration, whereas persistent atrial arrhythmia was defined as occurring for >7 days.
TEC were classified into 2 groups: 1) systemic TEC, defined as intracardiac thrombus, ischemic stroke, or systemic arterial embolus; and 2) nonsystemic TEC, defined as Fontan conduit/right atrial thrombus or pulmonary embolus.
The imaging modalities used in this study were transthoracic echocardiogram (TTE), transesophageal echocardiogram, computed tomography (CT), or magnetic resonance imaging, including cardiac magnetic resonance imaging. TTE was considered the basic cardiac imaging modality, whereas other modalities were considered supplementary. The imaging reports of all patients were reviewed and the images for all patients with TEC were retrieved and viewed to confirm the diagnosis.
Patients were categorized based on the type of preventive medical therapy received during the study period: 1) no therapy group (did not receive any anticoagulation or antiplatelet therapy at the time of TEC or throughout the study period for patients without TEC); 2) antiplatelet therapy group (received antiplatelet therapy alone at the time of TEC or throughout the study period for patients without TEC); and 3) anticoagulation therapy group (received anticoagulation therapy with or without antiplatelet therapy at the time of TEC or throughout the study period for the patients without TEC).
For patients on warfarin therapy who experienced TEC, a subtherapeutic international normalized ratio (INR) was defined as INR below designated target range based on the last INR assay performed within 48 h of the TEC.
The primary objective was to determine the incidence and predictors of TEC in adult patients with Fontan palliation. Secondary objectives were to determine the risk of bleeding and the impact of TEC on the composite adverse event endpoint of heart failure hospitalization and all-cause mortality.
Major bleeding was defined as intracranial bleeding, pericardial or pleural hematoma requiring drainage, or any bleeding requiring transfusion of blood product. Minor bleeding was defined as cutaneous bleeding, epistaxis, gastrointestinal bleeding, or any bleeding event that did not meet the criteria for major bleeding.
All statistical analyses were performed with JMP (version 10.0 software, SAS Institute Inc., Cary, North Carolina). Categorical variables were expressed as number (%) and continuous variables were expressed as mean ± SD or median (interquartile range) for skewed data. Categorical variables were compared using the chi-square test or Fisher exact test; continuous variables were compared with a 2-sided unpaired Student t test or Wilcoxon rank sum test, as appropriate. Cox proportional hazards models were used to identify predictors of TEC. Only the variables that achieved statistical significance on univariable analysis were used in the multivariable model. The risk associated with each variable was expressed as a hazard ratio (HR) and 95% confidence interval (CI). The TEC rates and composite adverse event rates were calculated with the Kaplan-Meier method and compared using the log-rank test. Any p value <0.05 was considered significant.
There were 278 patients who met the inclusion criteria; mean age was 31 ± 9 years and the mean duration of follow-up was 88 ± 14 months (1,464 patient-years) (Table 1).
The presenting arrhythmia was atrial flutter/intra-atrial re-entry tachycardia/atrial tachycardia in 200 patients (72%), atrial fibrillation in 9 (3%), and a combination of both atrial flutter/intra-atrial re-entry tachycardia and atrial fibrillation in 69 (25%). Atrial arrhythmia was paroxysmal in 246 patients (88%) (Table 1).
Anticoagulation and antiplatelet therapy
There were 181 patients (65%) in the antiplatelet therapy group; the antiplatelet agents used were aspirin (n = 164), clopidogrel (n = 16), or aspirin plus clopidogrel (n = 1). There were 91 patients (33%) in the anticoagulation therapy group, 33 of whom received warfarin only, 55 received warfarin plus antiplatelet therapy, and 3 received rivaroxaban plus antiplatelet therapy. Of the 88 patients on warfarin, 29 (33%) had subtherapeutic INR.
There were 6 patients in the no-therapy group. The reason for not receiving anticoagulation or antiplatelet therapy was due to a history of gastrointestinal bleeding in 5 patients and intracranial bleeding in 1 patient. (These events occurred prior to the beginning of the study period and, as a result, were not counted as bleeding events for the study).
Thrombotic and embolic complications
Overall, 97 TEC occurred in 81 patients during follow-up, including Fontan conduit/right atrial thrombus (n = 33), pulmonary embolus (n = 32), intracardiac thrombus (n = 14), ischemic stroke (n = 15), and systemic arterial embolus (2 renal infarctions and 1 splenic infarction). Table 2 shows the imaging modality used at the time of initial diagnosis for each TEC.
Among these complications, 32 were systemic (yielding an event rate of 2.1 systemic TEC per 100 patient-years) and 65 were nonsystemic (4.4 nonsystemic TEC per 100 patient-years). Overall TEC prevalence was 29%, for an event rate of 6.5 TEC per 100 patient-years. Patients with atriopulmonary connection (APC) had more TEC than did patients with other types of Fontan connections: 7.3 versus 3.9 TEC per 100 patient-years (Figure 1). Of the 97 TEC, 43 (44%) presented with symptoms and 38 (39%) required hospitalization for anticoagulation and imaging evaluation. The indications for hospitalization were heart failure symptoms (ascites, pedal edema, dyspnea, hypoxia), palpitation, and neurological symptoms.
Sixteen patients had more than 1 TEC, including the following: Fontan conduit/right atrial thrombus and pulmonary embolus (n = 8); Fontan conduit/right atrial thrombus and intracardiac thrombus (n = 3); intracardiac thrombus and ischemic stroke (n = 3); intracardiac thrombus and renal infarction (n = 2). The prevalence of TEC (systemic and nonsystemic) in this cohort was 7%, 18%, and 55% at 3, 5, and 10 years, respectively (Central Illustration).
All patients with TEC underwent TTE; some also experienced supplemental cardiac imaging with transesophageal echocardiography (n = 76), cardiac/chest CT scan (n = 68), or cardiac/chest cardiac magnetic resonance scan (n = 14). Noncardiac supplementary imaging included brain magnetic resonance imaging scan (n =22), brain CT scan (n=36), or abdominal CT scan (n = 14). Indications for supplemental imaging were symptoms (chest pain/discomfort, dyspnea, abdominal pain, hematuria, neurologic symptoms) in 37 patients, pre-procedural screening for cardioversion in 16 patients, abnormal finding noted on TTE in 24 patients, and unknown in 4 patients.
There were 92 TEC (95%) with available electrocardiograms performed at the time of diagnosis; atrial arrhythmia was present in 59 electrocardiograms (64%). There was no difference in TEC rates in the patients with history of atrial fibrillation versus the patients with other forms of atrial arrhythmia (22% vs. 15% at 60 months; p = 0.56). The anticoagulation therapy group had a lower rate of TEC than did the antiplatelet therapy group (11% vs. 22% at 60 months; p = 0.016) (Central Illustration).
The multivariable risk factor for TEC was an APC (HR: 2.31; 95% CI: 1.61 to 4.64; p = 0.037), whereas anticoagulation therapy was protective against TEC (HR: 0.63; 95% CI: 0.32 to 0.79; p = 0.003) (Table 3).
There were 21 bleeding events (7%; 6 major and 15 minor bleeding events), for a major bleeding incidence rate of 0.4 per 100 patient-years. The major bleeding events were intracranial hemorrhage (n = 2) and gastrointestinal bleeding requiring blood transfusion (n = 4). Preventive medical therapy (anticoagulation and/or antiplatelet therapy) was discontinued in all 6 patients because of the bleeding events.
Minor bleeding events were epistaxis (n = 4), gastrointestinal bleeding (n = 9), menorrhagia (n = 1), and hemoptysis (n = 1). Preventive medical therapy was discontinued in 3 patients because of their minor bleeding events. There was no significant difference in the incidence of bleeding events in the anticoagulation therapy group versus the antiplatelet therapy group (9% vs. 7%; p = 0.081).
The composite adverse event endpoint of heart failure hospitalization and all-cause mortality was reached in 161 patients (58%). The occurrence of a composite adverse event was significantly higher in the patients with TEC (70% vs. 49% at 60 months; p < 0.0001) (Central Illustration). Patients receiving anticoagulation therapy had a lower adverse event rate than did those in the antiplatelet therapy group (68% vs. 54% at 60 months; p = 0.02).
This study reviewed a large cohort of adult Fontan patients with atrial arrhythmia to determine the incidence of TEC and the efficacy of anticoagulation therapy in preventing these events. The major findings were as follows: 1) TEC occurred in 29% of Fontan patients with atrial arrhythmia and most patients were asymptomatic at the time of TEC diagnosis; 2) TEC were more common in APC than other types of connections; 3) the use of anticoagulant therapy (with warfarin or a direct oral anticoagulant) was associated with a lower risk of TEC when compared with antiplatelet therapy alone; and 4) the occurrence TEC was associated with increased risk for heart failure hospitalization and all-cause mortality.
Complications and their risk factors
Thrombolic and embolic complications are again highlighted as being especially common in this group of patients, and extrapolation from this study suggested that TEC will occur in 55% of Fontan patients with atrial arrhythmia at 10 years, with an incidence of 6.5 TEC per 100 patient-years. Most of the studies reporting TEC prevalence in the Fontan population were derived from a general Fontan population irrespective of their underlying rhythm (8–14).
A large retrospective review of 592 Fontan patients had reported a 9% risk of TEC (15); however, in contrast to this investigation, the lower TEC incidence was likely due to a younger cohort (median age: 4 years), lower proportion of patients with APC, and shorter follow-up. Plus, the previous review used a definition of TEC that was restricted to intracardiac thrombus identified on echocardiography and atrial arrhythmia was present in <15% of their cohort. Other studies have shown higher TEC rates in cohorts of older Fontan patients or in case series of Fontan patients with atrial arrhythmia (11,12,16).
The diagnosis of TEC requires a high index of suspicion because it appears that most patients are asymptomatic at the time of diagnosis or they present with nonspecific symptoms. Less than one-half of our cohort with TEC presented with symptoms and the majority of supplementary imaging studies used for TEC diagnosis were performed for pre-procedural screening or because of abnormal findings on the TTE. Consistent with our finding, 43% of patients with TEC were asymptomatic at the time TEC diagnosis in another cohort of 70 Fontan patients (11). Several other studies demonstrated a prevalence of “silent” pulmonary embolism and Fontan conduit thrombus of 17% and 13%, respectively, based on CT scans of asymptomatic Fontan patients (13,17).
The presence of APC was a risk factor for TEC. This contrasted with previous studies that reported no relationship between TEC and type of Fontan connection (10,11,15). All these studies were derived from a younger population with median age <13 years, compared with our cohort with mean age of 31 years. Although the presence of APC did not increase the risk of TEC in the early years of Fontan physiology, we speculate that right atrial dilation occurred over time, leading to abnormal flow dynamics and blood stagnation resulting in the increased TEC risk seen in our study. A recent study elucidating the underlying mechanism of thrombus formation in the Fontan pathway using 2-dimensional computer hemodynamic simulation demonstrated that the APC model had the highest incidence of thrombus formation compared with total cavopulmonary models because of slower blood flow at rest and significant blood flow stagnation in the atrium (18).
Because APC was the predominant type of Fontan connection in our cohort and in several other studies of adults with history of Fontan operation (1,2), we should consider most adults with APC and atrial arrhythmia to be at a high risk for TEC. Therefore, there should be a low threshold for initiating preventive therapy for TEC.
Interestingly, arrhythmia type was not predictive of TEC in our study. It is worthwhile to note that there was some overlap in arrhythmia type with some patients having a history of more than 1 type of arrhythmia, the most common being a history of atrial flutter and atrial fibrillation. Also, the presence of protein-losing enteropathy, fenestration, or liver disease was not predictive of TEC, which is similar to the findings of other studies (12,15).
Anticoagulant and antiplatelet therapy
Previous studies have demonstrated abnormal levels of procoagulant and anticoagulant factors in Fontan patients (5–7), creating a paradox of increased bleeding risk in the setting of increased clotting risk. As a result, there is a general reluctance to initiate anticoagulation therapy in Fontan patients even in the setting of paroxysmal atrial arrhythmia (19–21).
Our data suggested that anticoagulation therapy with or without antiplatelet therapy was associated with lower risk of TEC than antiplatelet therapy alone. There are few studies comparing the efficacy of anticoagulation therapy (warfarin therapy) to antiplatelet therapy (aspirin) in the prevention of TEC, and these studies have provided conflicting results (8–10,22,23). TEC rates in the range of 5% to 15% have been reported in small cohorts of Fontan patients (10,22), with those patients receiving warfarin sustaining less TEC without any increase in bleeding events compared to the patients on aspirin or no therapy. A different study reported that anticoagulation had no protective benefit against TEC based on the review of a cohort of Fontan patients with a TEC rate of 5%. The limitation of that study was that the entire cohort had very low TEC occurrence (4 cases); thus, it was underpowered to detect a difference in treatment outcome. On the other hand, the current study characterized a higher-risk Fontan cohort, with atrial arrhythmia and a higher TEC rate, than previous reports had; hence, our study was sufficiently powered to detect difference in treatment outcome.
A randomized controlled trial comparing aspirin to warfarin for the prevention of TEC immediately after Fontan operation reported no difference in outcome (23). Importantly, this trial was based on a pediatric cohort (mean age was 4 years), and the efficacy of aspirin and warfarin therapy were only compared for the first 2 years after Fontan operation. Moreover, the patients in the treatment arm were in therapeutic range only for a short time, which is in stark contrast to our adult group of Fontan patients (mean age 31 years) with atrial arrhythmia followed more than 7 years.
Risk of bleeding is a major concern in the decision to initiate anticoagulation therapy. These findings illustrated an important feature to aid clinical decision making in this arena, with the data revealing 21 bleeding events (7%), of which 6 (2%) were major bleeding events requiring discontinuation of therapy. The bleeding risk in this study was comparable to results of previous studies (10,11). Of note: there was no difference in bleeding risk in the patients on anticoagulation therapy compared with those on antiplatelet therapy alone.
Impact of thrombotic and embolic complications
Our data suggested that the occurrence of TEC was associated with a higher risk of heart failure hospitalization and death; yet, this provocative finding raises the question as to whether the occurrence of TEC directly increases this risk for heart failure hospitalization and death, or is the presence of TEC just a marker of a much sicker patient subset? This inference cannot be drawn from our study; a different study design is required to answer this question. Notwithstanding, other studies have suggested an increased mortality risk in Fontan patients with thromboembolism (14,24).
The pathophysiology for heart failure and death in Fontan patients with TEC might be due to the effect of microembolization into the pulmonary vasculature. The absence of nonpulsatile pulmonary blood flow in Fontan physiology is associated with impairment of the nitric oxide pathway and endothelial dysfunction (25,26). We speculated that microembolization from systemic venous thrombi superimposed on this underlying pulmonary capillary endothelial dysfunction will increase pulmonary vascular resistance at rest and also impair pulmonary capillary recruitment required to accommodate increased pulmonary blood flow during exercise. The downstream effect will be worsening systemic venous hypertension, right heart failure, and death. Additionally, the presence of fenestration and right-to-left shunt will increase the risk of embolization to the systemic circulation, resulting in morbidities such as stroke, myocardial infarction, and ultimately death.
Recommendations for Fontan patients
Based on these findings, we propose the following: 1) aggressive screening for paroxysmal atrial arrhythmia using Holter and event monitors because of the increased risk of TEC in patients with atrial arrhythmias; 2) a high index of suspicion for TEC in Fontan patients with atrial arrhythmia, and a low threshold to employ supplementary imaging modalities such as transesophageal echocardiogram, CT, and magnetic resonance imaging (cardiac or noncardiac); and 3) initiation of anticoagulation therapy with or without antiplatelet therapy as the preferred preventive strategy for TEC especially, in the subset of patients with APC.
This was a retrospective review of patients in a single tertiary center, and it is therefore subject to the bias inherent in this type of study design. There were very few patients with isolated atrial fibrillation, and a subset of patients with overlap of atrial flutter and atrial fibrillation. This might affect the generalizability of our finding that TEC risk was independent of arrhythmia type. Finally, our study was unable to completely separate the protective effect of anticoagulation therapy alone from that of antiplatelet therapy alone because most of our anticoagulation therapy patients were also on antiplatelet therapy. We were also not able to comment on the superiority of various anticoagulation strategies. However, our results suggested that the addition of anticoagulation therapy was associated with lower risk of TEC versus antiplatelet therapy alone.
Fontan patients with atrial arrhythmia are at increased risk of TEC and this risk is higher in the subset with an APC. The occurrence of TEC was associated with increased risk of heart failure hospitalization and death. Anticoagulation therapy with or without antiplatelet therapy was associated with lower risk of TEC without a significant increase in bleeding risk and might be considered the preferred preventive strategy in these patients.
COMPETENCY IN MEDICAL KNOWLEDGE: Among adults with atrial arrhythmias following Fontan palliation, the risk of clinical thromboembolism is 6.5% per 100 patient-years.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Prophylactic anticoagulation with warfarin reduces the incidence of thromboembolism, hospitalization for heart failure, and all-cause mortality.
TRANSLATIONAL OUTLOOK: Further studies are needed to identify the clinical features and imaging findings associated with higher or lower risks of thromboembolism in this patient population.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- atriopulmonary connection
- confidence interval
- computed tomography
- hazard ratio
- international normalized ratio
- thrombotic and embolic complication
- transthoracic echocardiogram
- Received April 2, 2016.
- Revision received May 25, 2016.
- Accepted June 21, 2016.
- American College of Cardiology Foundation
- Pundi K.N.,
- Johnson J.N.,
- Dearani J.A.,
- et al.
- Stoddard M.F.
- Varma C.,
- Warr M.R.,
- Hendler A.L.,
- Paul N.S.,
- Webb G.D.,
- Therrien J.
- Walker H.A.,
- Gatzoulis M.A.
- Monagle P.,
- Cochrane A.,
- Roberts R.,
- et al.,
- for the Fontan Anticoagulation Study Group