Author + information
- Received September 4, 2012
- Revision received December 12, 2012
- Accepted January 3, 2013
- Published online March 26, 2013.
- Joy T. Johnson, MD,
- Ian Lindsay, MD,
- Ronald W. Day, MD,
- Charlotte S. Van Dorn, MD,
- James Hoffman, MD,
- Melanie D. Everitt, MD and
- Anji T. Yetman, MD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Anji T. Yetman, Adult Congenital Cardiology Program, Primary Children's Medical Center, 100 North Mario Capecchi Drive, Salt Lake City, Utah 84113
Objectives This study sought to determine whether survival in this cohort of patients was adversely affected by increased residential altitude.
Background The success of the Fontan procedure depends in large part on low pulmonary vascular resistance (PVR). Factors that increase PVR, including an increase in residential altitude, may adversely affect long-term outcome. Higher altitude has been shown to affect functional well-being in patients with a Fontan circulation.
Methods Databases from a tertiary cardiac care center in the Intermountain West (elevation 5,000 feet) were analyzed for patients born with single-ventricle anatomy who would now be of adult age. Complete data were then collected on all identified patients who subsequently underwent the Fontan operation. Correlates of, and time to, adverse outcome, defined as death, cardiac transplantation, or clinical decompensation requiring a move to sea level, were determined.
Results Of 149 patients with single-ventricle anatomy, 103 underwent the Fontan procedure, with 70 surviving to adulthood at moderate altitude. Adverse outcome occurred in 55, with death in 24 (23%), cardiac transplantation in 18 (17%), and clinical decompensation requiring move to sea level in 13 (13%). There was no relationship between type, age at, or era of Fontan procedure and long-term outcome. Correlates of long-term, transplant-free survival at moderate altitude included lower residential altitude (4,296 vs. 4,637 feet, p < 0.001), and lower pulmonary artery pressures before the Fontan procedure (13 vs. 15 mm Hg, p = 0.01), and after (14 vs. 18 mm Hg, p = 0.01).
Conclusions Long-term outcome after the Fontan procedure is adversely impacted by higher residential altitude.
The Fontan procedure is the anticipated final surgical intervention in children with single-ventricle anatomy. The procedure, first described by Dr. Fontan in 1971, revolutionized the outlook of patients with single-ventricle anatomy as it allowed them to survive into adulthood. The surgical procedure requires rerouting of superior and inferior caval flow to the pulmonary arteries without an intervening pump, and thus, low pulmonary vascular resistance is critical to the success of the procedure (1).
With increased elevation, there is a reduction in the partial pressure of oxygen (pO2), which serves as the stimulus for a number of physiological changes including an increase in plasma catecholamines and renin activity, resulting in an increase in systemic and pulmonary vascular resistance, as well as heart rate (2). These physiological adaptations may prove particularly detrimental to the patient who has undergone Fontan palliation. The negative impact of increased altitude on early Fontan hemodynamics has been described (3), as has acute clinical deterioration of sea level residents with a Fontan circulation who have been exposed to moderate altitude (4,5). The long-term outcome of patients residing at moderate altitude after the Fontan procedure, however, is unknown. We sought to assess morbidity and mortality in a group of patients with a Fontan circulation followed up at a tertiary center located at moderate altitude and to determine what, if any, effect residential elevation has had on long-term survival.
An institutional review board–approved retrospective review of surgical, cardiac catheterization, and clinical databases from a tertiary cardiac care center located at an elevation of 5,000 feet and serving a 3-state region in the Intermountain West was performed. Patients with single-ventricle anatomy who would now be of adult age (>18 years) were identified. Patients from this initial review were then included in the study for further analysis if they had undergone a Fontan operation.
Records were reviewed, and the following demographic and surgical variables were recorded: date of birth, sex, date of Fontan procedure, type of initial Fontan procedure (atriopulmonary, lateral tunnel, or extracardiac conduit), presence or absence of surgical fenestration, presence or absence of prior cavopulmonary anastomosis, interval between cavopulmonary anastomosis and subsequent Fontan procedure, presence and date of surgical conversion to an alternate type of Fontan, and need for surgical revision of the primary Fontan connection. Presence and absence of a pacemaker was noted. Cardiology variables including systemic ventricular morphology (right, left, or indeterminate), pulmonary artery (PA) pressures on pre-Fontan cardiac catheterization, subsequent percutaneous or spontaneous fenestration closure, and PA pressures at cardiac catheterization at last follow-up were noted.
The following outcomes were noted to be present or absent: protein-losing enteropathy (PLE) defined by an elevated stool alpha-1 antitrypsin and associated hypoalbuminemia, plastic bronchitis, thrombosis defined as a clinically symptomatic thromboembolic event, death, listing for cardiac transplantation, and clinical decompensation prompting relocation to lower altitude. New York Heart Association (NYHA) classification and oxygen saturations at rest and with exercise were recorded at last follow-up before death, transplantation, or move from moderate altitude to sea level. For the purposes of this study, moderate altitude was defined as >3,000 feet elevation (4).
Subjects were categorized as: 1) those with composite adverse outcome; and 2) those without composite adverse outcome, defined as death, cardiac transplantation or listing for cardiac transplantation, or move to sea level for cardiac decompensation. Adverse outcomes of death and cardiac transplantation were also analyzed individually. Date of composite adverse outcome and cause of composite adverse outcome were recorded. Cause of adverse outcome was identified as heart failure, Fontan failure, thrombosis, PLE, hepatic dysfunction, sudden death, perioperative death, or other. Heart failure was defined as NYHA class III or greater symptoms in the face of qualitatively impaired systemic ventricular function on echocardiogram. Fontan failure was defined as NYHA class III or greater symptoms in the face of qualitatively normal systemic ventricular function on echocardiogram and elevated PA pressures on cardiac catheterization. Hepatic dysfunction was defined as functional abnormalities noted on serology in association with structural alterations present on liver biopsy. Sudden death was defined as death within 3 h of the onset of symptoms and perioperative death was defined as death within 28 days of cardiac surgery. The “other” category included patients who died of noncardiac reasons or patients for whom the cause of death could not be elucidated.
Residential elevation was determined by using an adjusted average of the elevation of the zip codes at which the patient resided between the Fontan operation and the clinical endpoint (death, cardiac transplantation, move to sea level, last visit).
Patients were classified by initial Fontan type defined as atriopulmonary connection, lateral tunnel type, or extracardiac conduit, depending on the surgical procedure performed. Repeat analyses using final Fontan type was also performed to assess for clinical effect in patients undergoing conversion from one type of Fontan to another.
All patients were followed up in either a pediatric cardiology or an adult congenital cardiology academic practice. Patients were seen with the frequency that their clinical condition demanded, with recommended follow-up at least annually. Echocardiograms were obtained at the discretion of the patients' primary cardiologist but were performed at least annually. Patients with symptoms of heart failure or functional deterioration underwent cardiac catheterization. If the source of heart failure was myocardial dysfunction, heart transplant work-up was considered and discussed with the patient. In light of prior publications noting an almost universal decline in functional status of sea level Fontan patients traveling to moderate altitude (4), patients with preserved myocardial function but elevated PA pressures were given a trial of oxygen and medical therapies for pulmonary hypertension. If they continued to have incapacitating symptoms of heart failure, or persistent symptomatic desaturation, and if they had noted improvement in these symptoms at lower elevations, a move to sea level was discussed. For patients unwilling or unable to relocate, cardiac transplantation was discussed. Patients were listed for cardiac transplantation at the discretion of the heart failure/transplant team at our center. Medications, including anticoagulants, were prescribed at the discretion of the patient's primary cardiologist and were not standardized across providers. All patients were treated with either aspirin or warfarin for anticoagulation therapy. All patients with prior thromboembolic events or sustained atrial arrhythmias were managed on warfarin.
Statistical analysis was conducted using SAS version 9.2 (SAS Institute, Cary, North Carolina). Continuous data were expressed as means with standard deviations or medians with ranges as appropriate, and categorical data were tabulated. A 2-sided p value of < 0.05 was considered statistically significant. Demographic and clinical data for adult survivors with adverse outcome, and survivors without an adverse outcome were compared using a chi-square test, for dichotomous or categorical variables, and a Student t test or Wilcoxon rank-sum test for continuous variables depending on their normality. Patient data were censored on the basis of their last known visit if they were lost to follow-up. Kaplan-Meier survival curves were generated to assess time to composite adverse outcome, as well as time to death or transplantation alone. Correlates of the presence of, and time to, composite adverse outcome were sought initially on univariate analyses. Multivariate analyses with Cox proportional hazards regression were performed to estimate hazard of the outcome. Factors that had a p value < 0.2 were included in the model to determine the adjusted odds ratio. Variables were retained if they changed the point estimate by >10%.
Of 149 patients born between 1965 and 1991 with single-ventricle physiology, 103 underwent a Fontan procedure during the period 1978 to 2005; the remainder either died or were not considered to have hemodynamics suitable for a Fontan procedure and underwent palliation with a bidirectional cavopulmonary anastomosis alone or cardiac transplantation. Of those undergoing the Fontan procedure, 70 patients (68%) survived into adulthood at moderate altitude without heart transplantation, with a median age of these survivors of 29.2 (18.1 to 40.4) years.
Outcomes of the Fontan cohort are depicted in Figure 1. Of the initial 103 patients, 24 (23%) died at a median age of 13.5 (0.5 to 36.9) years. An additional 18 patients were listed for cardiac transplantation at a median age of 18.2 (3.5 to 40.8) years. Thirteen (13%) additional patients moved to sea level due to cardiac decompensation at a median age of 16.5 (8.2 to 28.8) years, and 3 moved out of state, to lower altitude, for noncardiac reasons, leaving 45 patients (43%) from the initial Fontan cohort currently alive as adults at moderate altitude without listing for cardiac transplantation. Within this group of survivors, oxygen saturations were 89.0 ± 3.5% at rest and fell to 84.6 ± 5.1% with exercise.
Cause of composite adverse outcome
Cause of adverse outcome is listed in Figure 1. There was significant overlap of comorbidities, with many patients having >1 (Table 1). Of 85 patients surviving into adolescence, 41 (48%) had a clinically significant thrombotic event (stroke, pulmonary embolus, renal artery thrombosis, or coronary artery thromboembolus with myocardial infarction). Twenty-six (44%) of the 59 patients who underwent hepatic evaluation were found to have physiologic or anatomic hepatic abnormalities. Protein-losing enteropathy was diagnosed in 19 (27%) of the 71 patients who underwent evaluation for such. Transplant-free long-term survival was quite low among patients with PLE, with only 3 of 19 (15%) still alive. Of the 3 survivors, 1 is being evaluated for heart and liver transplantation, 1 is in hospice care, and the other remains stable with a recurrent relapsing course on medical therapy. All nonsurviving patients with PLE died of fatal thrombotic events.
Correlates of adverse outcome
Initial Fontan procedure was an atriopulmonary connection in 54 patients (52%), lateral tunnel in 36 (35%), and extracardiac conduit in 13 (13%). Hemodynamic and anatomic variables categorized by type of initial Fontan surgery are displayed in Table 2. As in many centers in North America, our center underwent a surgical revolution throughout the years, transitioning from the initial atriopulmonary connection, to the lateral tunnel, and subsequently to the extracardiac conduit. Concomitant with this transition was a trend toward performing a cavopulmonary anastomosis before Fontan completion. This is reflected in Table 2 as a significant difference in the presence of a cavopulmonary anastomosis between the different types of Fontan procedure. Another historical trend that occurred during the study period was the practice of palliating infants with hypoplastic left heart variants. This, too, is reflected in Table 2 as patients undergoing a more modern version of the Fontan procedure were more apt to have a systemic right, rather than left, ventricle.
Demographic and clinical variables for patients with and without composite adverse outcome are depicted in Table 3. Underlying ventricular morphology, age, and type of surgery did not correlate with long-term outcome. The initial presence and persistence of a patent fenestration was not correlated with adverse outcome. Fourteen patients underwent Fontan conversion; 12 patients had conversion to an extracardiac Fontan, 9 from an atriopulmonary connection and 3 from a lateral tunnel. Two patients were converted from an atriopulmonary connection to a lateral tunnel Fontan. Four additional patients underwent 6 revisions without subsequent conversion, 3 had revision of an atriopulmonary connection and 1 a lateral tunnel revision. One of these 4 patients underwent 3 revisions of an atriopulmonary Fontan, with intraoperative death during the third revision. Patients having undergone Fontan conversion were equally likely to have a long-term adverse outcome (Table 3). Age at last follow-up did not differ between those with versus without adverse outcomes (28.9 [0.5 to 40.8] vs. 25.7 [16.9 to 40.3]; p = 0.52).
Nonsurgical variables associated with long-term composite adverse outcome were initially explored with univariate analyses and included residential elevation, pre-Fontan mean PA pressures at cardiac catheterization, post-Fontan mean PA pressures at most recent cardiac catheterization, resting oxygen saturation, presence of PLE, presence of hepatic dysfunction, age at Fontan, type of initial Fontan, or age at last follow-up (Table 4). Factors associated with adverse outcome on univariate analysis also included the presence of a pacemaker, and lack of a bidirectional cavopulmonary anastomosis before Fontan surgery. However, when adjusted for era of surgery as defined by 1970 to the 1980s versus 1990 to the 2000s in multivariate analysis, the presence of a bidirectional cavopulmonary anastomosis no longer correlated with outcome (Table 5).
Cox proportional hazard regression modeling was used to estimate the hazard ratio of a composite adverse outcome. Thrombosis, PLE, residential elevation, pre-Fontan PA pressure, recent post-Fontan PA pressure, and resting O2 saturation were adjusted for in multivariate analysis. In Table 5, the top part shows the results of the multivariate analysis and hazard ratio determination; and the bottom part shows the Cox proportional hazard regression modeling estimating the hazard ratio of presence of a bidirectional cavopulmonary anastomosis before Fontan and is adjusted for residential elevation and era of surgery, with only residential elevation showing a statistically significant association.
Figure 2 demonstrates the cumulative survival, and freedom from death, transplant, or move to sea level. Freedom from death, transplant, or move to sea level at 5, 10, 15, and 20 years was 84.7%, 67.3%, 55.1%, and 26.5%, respectively.
In analysis of outcome as death or transplant, 93 patients had data for analysis. Forty-six (49.5%) had an outcome of death or transplant, which was not significantly associated with sex, presence of hepatic dysfunction, age at last follow-up, age at Fontan, type of Fontan, or presence of surgical fenestration. A summary of significant associations can be seen in Table 6. Era of surgery was also analyzed by a Cox proportional regression model. There was a hazard ratio of 0.998 with a confidence interval of 0.552 to 1.802, and a p value of 0.99 when composite adverse outcome was analyzed by surgical era divided into 1970 to 1980 and 1990 to the 2000s.
When analyzed for freedom from transplant or death, freedom from death or transplant at 5, 10, 15, and 20 years was 84.9%, 69.9%, 57%, and 25.8%, respectively. Figure 3 shows the Kaplan-Meier survival curve for freedom from death or transplant.
As patients with the Fontan procedure have entered into adulthood, the long-term complications of this passive circulation have become evident. Despite this, the Fontan procedure remains the definitive surgical intervention of choice for patients with single-ventricle anatomy at most centers throughout the country, including centers at moderate altitude. To date, there are minimal data on actuarial survival and clinical complications of the Fontan procedure in patients residing at moderate altitude. This study provides an analysis of long-term follow-up and presence of comorbidities for a cohort of patients who are, or would have been, of adult age with single-ventricle physiology palliated with a Fontan procedure at moderate altitude.
When compared to previously published studies on sea level Fontan patient cohorts, our study demonstrates markedly reduced transplant-free survival. Khairy et al. (6) documented 15-year and 20-year survival of 72% and 65%, respectively, in a sea level cohort, in contrast to our survival rates of 57% and 25%. The reduction in survival is not particularly surprising if one considers the physiological adaptations to increased elevation present at moderate altitude. The rise in pulmonary vascular resistance will lead to impedance to forward flow through the Fontan circuit, predisposing to Fontan failure, hepatic congestion, and PLE. It is, therefore, not surprising that the incidence of PLE in our patient cohort was more than double that previously reported among patients living at low elevation (7–9). Our finding of impaired survival in patients with PLE is consistent with previously published reports (6,8). Patients with a persistent fenestration may be spared from such gastrointestinal and hepatic complications but will have erythrocytosis in association with their increased cyanosis, thus potentially predisposing them to thromboembolic complications. Increasing elevation further increases hematocrit, which may be particularly disadvantageous to these patients. Even in the absence of a patent fenestration, desaturation was almost universal in our patient cohort, and the risk of thrombosis was significantly higher than previously reported estimates of 21% to 25% in low-altitude patients (10–12). Although all patients with a documented thromboembolic event were placed on a regimen of warfarin, the clinical benefit of warfarin versus aspirin therapy for primary prophylaxis of thrombus formation in this patient population remains unknown. In a review of the available literature in 2002, it was considered that there was insufficient evidence to make recommendations regarding thromboembolic prophylaxis for this population (13). Given the very high rate of thromboembolism in our patient cohort, however, consideration may need to be given to a multicenter study of primary thromboprophylaxis of Fontan patients living at moderate or greater altitude. The higher baseline incidence may allow for better ascertainment of clinical effect.
The increase in systemic vascular resistance that occurs at increasing elevations may adversely affect systemic ventricular function, atrioventricular valve regurgitation, and systemic ventricular filling, thereby predisposing patients to both myocardial dysfunction and Fontan failure, which were also common among our adult patients. The altitude-related increase in catecholamines may further predispose to arrhythmias, which may in turn be associated with thrombosis, and myocardial dysfunction. Given these changes in physiology, the population of Fontan survivors at moderate altitude may respond to medical therapies differently than their counterparts at sea level. Higher renin levels (2) and reduced inhaled nitric oxide responsiveness (14) may affect the impact of therapies such as angiotensin-converting enzyme inhibitors and phosphodiesterase-5 inhibitors in patients living at moderate altitude.
Our findings are consistent with those of other investigators assessing early post-operative outcome and functional status. In their review, Hosseinpour et al. (3) described increased early Fontan failure, defined as death, Fontan takedown, or cardiac transplantation, at higher elevations ranging up to 520 m (approximately 1,700 feet). Impaired growth (15) and reduced exercise tolerance (16) have also been noted in this patient cohort when compared to sea level controls. The Pediatric Heart Network investigators recently reported on the functional status of patients with a Fontan circulation (17), and noted greater impairment in functional status and decreased ventricular systolic function in patients with a pacemaker. Similarly, patients with a pacemaker were more likely to experience an adverse clinical outcome in our series.
It is not known, what, if any, absolute residential elevation is not detrimental to the Fontan patient. We have demonstrated on multivariate analysis that, with an altitude median of 4,500 feet (interquartile range: 4,260 to 4,699 feet), the risk of an adverse event appears to linearly increase with increasing residential altitude. Similarly, Darst et al. (16) demonstrated a continuous effect of increasing elevation on exercise performance with a linear decline in exercise performance over a range of elevations from sea level to those in excess of 6,000 feet in a large cohort of Fontan patients.
This study was retrospective in nature and thus limited to the data collected as part of routine clinical care. As such, the date of onset of each adverse event cannot be defined, but rather, only the date of diagnosis of the adverse event. However, the timing of the composite adverse outcomes was precisely known. Although in any retrospective study the potential for uncontrolled confounding exists, the primary clinical confounders such as PA pressures, other adverse events such as thrombosis, PLE or hepatic complications, as well as era of surgery, were controlled for. However, there remains the possibility of uncontrolled confounding as retrospectively immeasurable factors such as income, socioeconomic status, and access to healthcare could be associated with both residential elevation and outcome. As there was a significant span of years over which the Fontan procedure was performed, there is potentially an era effect, with patients in the earlier era having worse outcomes. However, there was no statistical difference in the number of adverse outcomes when adjusted by surgical era. Our composite adverse outcome included patients moving to lower elevation for relief of symptoms. The effect of relocation in these subjects is unknown, however. Therefore, the data were analyzed with 2 different definitions of composite adverse outcome, the one including these patients and the other limited to death or cardiac transplantation.
Patients with a Fontan circulation residing at moderate altitude have decreased rates of transplant-free survival. We have identified risk factors for adverse outcome in this group of patients, including higher PA pressures, lower oxygen saturations, presence of thrombosis or PLE, and increased residential elevation. The effect of relocation to lower elevation remains unknown but is deserving of future study. Presence of these other risk factors should prompt consideration of transplantation or a trial of living at sea level.
The authors have reported they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- bidirectional cavopulmonary shunt
- New York Heart Association
- pulmonary artery
- protein-losing enteropathy
- Received September 4, 2012.
- Revision received December 12, 2012.
- Accepted January 3, 2013.
- American College of Cardiology Foundation
- Fontan F.,
- Baudet E.
- Kollias J.,
- Buskirk E.
- Khairy P.,
- Fernandes S.M.,
- Mayer J.E.,
- et al.
- Monagle P.,
- Cocharne A.,
- Roberts R.,
- et al.