|
|
||||||||
Ann Thorac Surg 2006;82:1286-1291
© 2006 The Society of Thoracic Surgeons
a Division of Cardiothoracic Surgery, Medical University of South Carolina, Charleston, South Carolina
b Division of Pediatric Cardiology, Medical University of South Carolina, Charleston, South Carolina
Accepted for publication April 19, 2006.
* Address correspondence to Dr Bradley, Division of Cardiothoracic Surgery, Medical University of South Carolina, 96 Jonathan Lucas St, Charleston, SC 29425 (Email: bradlesm{at}musc.edu).
Presented at the Fifty-first Annual Meeting of the Southern Thoracic Surgical Association, Cancun, Mexico, Nov 24, 2004.
| Abstract |
|---|
|
|
|---|
METHODS: From January 1996 to December 2005, 112 patients younger than 6 months of age underwent complete repair of tetralogy of Fallot or truncus arteriosus. An interatrial communication was used in 80 of 112 patients (71%). Hemodynamic data were determined during the first 48 hours after surgery.
RESULTS: In patients with an atrial communication, mean oxygen saturation reached a nadir of 94% ± 6%, and mean arterial PO 2 a nadir of 73 ± 25 mm Hg at 16 to 24 hours after surgery; both increased during the second 24 hours. At hospital discharge, median oxygen saturation was 98% (range, 86% to 100%). During the first 48 hours, mean oxygen saturation was less than 90% in 13 patients; the only multivariate risk factor was younger patient age. Mean right atrial pressure was greater than 10 mm Hg in 30 patients; multivariate risk factors were older patient age and repair with a transannular patch.
CONCLUSIONS: After complete repair of conotruncal defects using an interatrial communication, systemic oxygenation reaches a nadir at 24 hours after surgery, and improves by the time of hospital discharge. Clinically significant desaturation occurs in a small minority of patients. Infants undergoing repair before 2 months of age are at higher risk for systemic desaturation. The effects of an interatrial communication on systemic oxygenation should not be considered a contraindication to complete repair in early infancy.
| Introduction |
|---|
|
|
|---|
The primary aim of this study is to describe postoperative systemic oxygenation and hemodynamic status in infants undergoing repair of tetralogy of Fallot or truncus arteriosus before 6 months of age, with or without an interatrial communication. In particular, we sought to define the incidence, extent, and time course of systemic desaturation in the presence of an atrial communication. We also sought to identify risk factors for clinical indicators of right ventricular dysfunction in this group of patients.
| Patients and Methods |
|---|
|
|
|---|
|
Operations were carried out using cardiopulmonary bypass (mean, 153 ± 38 minutes) and aortic cross-clamping (mean, 72 ± 20 minutes). Periods of circulatory arrest were used in 34 patients (mean, 27 ± 14 minutes).
Analysis
Data were acquired by a retrospective review of hospital charts, intensive care unit records, operative notes, and echocardiogram reports. Hemodynamic data were determined during the first 48 hours after surgery. Right and left atrial pressures (RA, LA pressure) were measured by means of transthoracic catheters placed in the operating room. Right atrial pressure data were available in 103 of 112 patients (91%), and LA pressure data in 83 of 112 (74%). Systemic oxygen saturation was measured by extremity pulse oximetry. Inotropic score was calculated as dose of (dopamine) + (milrinone x 10) + (epinephrine x 100) [13].
Patients were analyzed for two indicators of postoperative right ventricular dysfunction: mean systemic oxygen saturation during the first 48 hours less than 90%, and mean right atrial pressure during the first 48 hours greater than 10 mm Hg. Only 2 patients had both a mean oxygen saturation less than 90% and a mean right atrial pressure greater than 10 mm Hg. Therefore, variables associated with each of these two indicators were analyzed independently.
Data are shown as mean ± standard deviation, or median (range). Data analysis was by Student's t test, Mann-Whitney U test,
2 test, linear regression, or repeated measures analysis of variance, as appropriate. Multivariable analysis was by stepwise logistic regression, retaining variables with a probability value less than 0.1 by univariate analysis (SPSS version 12.0; SPSS Inc, Chicago, IL). The variables analyzed are listed in Tables 2 and 3.
|
|
| Results |
|---|
|
|
|---|
An interatrial communication was deliberately left open at the end of surgery in 80 of the 112 patients. The atrial communication was preexisting in 78 and surgically created in 2. An atrial communication was left open in 64 of 68 (94%) of patients younger than 3 months of age, and in 71 of 77 (92%) of patients younger than 4 months of age at the time of surgery. As a consequence, the patients with an atrial communication were significantly younger than those with no atrial communication (median age, 1.1 versus 5.6 months; p < 0.001). An atrial communication was left open in 61% of the TOF/PS patients, 90% of the TOF/PA patients, and all patients with truncus arteriosus (Table 1).
Systemic Oxygenation
In the study group as a whole, systemic oxygen saturation was significantly lower in patients with an atrial communication, than in those with no atrial communication (Fig 1). Similarly, arterial PO
2 was lower in patients with an atrial communication (Fig 2). Over this time period, mean fraction of inspired oxygen decreased from 93% ± 16% at 0 hours, to 50% ± 16% at 24 hours, and 45% ± 11% at 48 hours.
|
|
|
The mean heart rate decreased during the first 48 hours, and was higher in patients with an atrial communication than in those without (mean during the first 48 hours, 160 ± 18 versus 151 ± 15 beats per minute; p < 0.001). Mean blood pressure was lower in those patients with an atrial communication left open (mean during the first 48 hours, 59 ± 10 versus 67 ± 10 mm Hg; p < 0.001). The mean inotropic score increased to a peak at 8 to 12 hours after surgery, and declined thereafter (p < 0.001; Fig 4). Throughout the first 48 hours, inotropic score was higher in the patients with an atrial communication than in those without (p < 0.001; Fig 4).
|
Among the 13 patients with mean oxygen saturation less than 90%, all 13 had an atrial communication left open, and 12 of 13 were younger than 2 months of age. By univariate analysis, variables significantly associated with a mean oxygen saturation less than 90% were younger age (analyzed as either a continuous or a categorical variable), the presence of an atrial communication, and shorter cross-clamp time (Table 2). By multivariate analysis, the only independent predictor was younger age (p = 0.01; odds ratio, 0.5; 95% confidence limits, 0.3 to 0.8).
Among the 30 patients with a mean RA pressure greater than 10 mm Hg, 26 of 30 were older than 2 months of age, 29 of 30 had TOF/PS, and 22 of 30 were repaired with a transannular patch. By univariate analysis, variables significantly associated with a mean RA pressure greater than 10 mm Hg were older age (analyzed as either a continuous or a categorical variable), repair with a transannular patch, a diagnosis of TOF/PS, repair without a homograft, and longer cross-clamp time (Table 3). By multivariate analysis, independent predictors were older age (p < 0.001; odds ratio, 1.7; 95% confidence limits, 1.3 to 2.2), and repair with a transannular patch (p = 0.004; odds ratio, 5.0; 95% confidence limits, 1.7 to 14.6).
Outcome
Overall operative mortality was 5 of 112 patients (4%; Table 1). Two patients required extracorporeal membrane oxygenation, both initiated in the operating room, with one survivor. Four of the 5 operative deaths were in patients with an atrial communication. None appeared to be caused by low systemic oxygen saturation. One (with TOF/PS) died of low cardiac output on day 4; during the first 48 hours her mean systemic oxygen saturation was 97%, and lowest PO
2 was 52 mm Hg. The second (with TOF/PA) died of an arrhythmia on the first postoperative night, with an oxygen saturation of 95%. The third (with truncus arteriosus) failed to wean from cardiopulmonary bypass, was brought out of the operating room on extracorporeal membrane oxygenation, and did not recover myocardial function. The fourth (with truncus arteriosus) died 2 months after surgery of chylous effusions and truncal valve insufficiency. The fifth (with TOF/PS) died 1 month after surgery of aspiration while feeding.
Resource utilization was greater in the patients with an atrial communication than in those without. The median duration of mechanical ventilation was 3 days (range, 1 to 53 days) versus 1 day (range, 0 to 3 days; p < 0.001). Median intensive care unit stay was 5 days (range, 2 to 71 days) versus 3 days (range, 1 to 48 days; p < 0.001). Median hospital stay was 11 days (range, 2 to 83 days) versus 6 days (range, 3 to 27 days; p < 0.001).
| Comment |
|---|
|
|
|---|
This study indicates that systemic desaturation resulting from an atrial communication is most pronounced approximately 24 hours after surgery. In the study group as a whole, both mean systemic oxygen saturation and arterial PO 2 were significantly lower in patients with an atrial communication than in those with no communication. The nadirs of both oxygen saturation and arterial PO 2 also occurred at a time at which the mean fraction of inspired oxygen was 50%. These findings suggest that desaturation is caused by right-to-left shunting at the atrial level rather than pulmonary venous desaturation. After the first 24 hours after surgery, systemic oxygenation improves, through the time of hospital discharge. This study also shows that clinically significant desaturation occurs in a minority of patients. The mean oxygen saturation during the first 48 hours was less than 90% in 13 of 80 patients (16%) with an atrial communication. The lowest mean oxygen saturation during the first 48 hours was 77%, in a patient with TOF/PS operated on at 9 days of age and discharged on postoperative day 13 with a saturation of 91%. This oxygen saturation is in a range that is routinely well tolerated in infants with cyanotic heart defects.
In the presence of an atrial communication, systemic desaturation is one indicator of right ventricular dysfunction. Among our study patients, multivariate analysis found that the single independent risk factor for a mean systemic saturation less than 90% during the first 48 hours was younger age, with 12 of 13 patients being younger than 2 months. This finding suggests that an age younger than 2 months at the time of surgery may be the most useful indication for maintenance of an atrial communication during repair.
An elevated RA pressure may also be an indicator of right ventricular dysfunction. However, in this study, selective elevation of RA pressure over LA pressure was uncommon. Only 1 patient had a mean RA to LA pressure difference during the first 48 hours greater than 3 mm Hg. It is possible that patients with postoperative right ventricular dysfunction had been successfully selected for maintenance of an atrial communication, which prevented selective elevation of RA pressure over LA pressure by allowing right-to-left shunting at the atrial level. Alternatively, diastolic ventricular function may be primarily affected by factors that affect both the right and left ventricle similarly, such as intraoperative myocardial ischemia and reperfusion. Among our study patients, multivariate analysis found that independent risk factors for a mean RA pressure greater than 10 mm Hg during the first 48 hours were repair with a transannular patch and older age, with 26 of 30 (87%) patients being older than 2 months. This finding may be a result of the effects of pulmonary insufficiency in the setting of the more extensive right ventricular hypertrophy seen in patients older than 2 months of age. It is in agreement with other studies that have shown that placement of a competent pulmonary monocusp valve during right ventricular outflow tract operations may be associated with improved outcome [14].
The time courses of systemic oxygenation, atrial pressures, and inotropic scores indicate that the first 48 hours after complete repair in early infancy are characterized by physiologic compromise followed by improvement. Among our patients, RA and LA pressure as well as inotropic score reached peaks, and systemic oxygenation a nadir, at around 24 hours. Similar declines have been observed in mixed venous oxygen saturation early after a Norwood procedure [1518] and in cardiac output early after an arterial switch operation [19]. These declines may reflect a period of depressed cardiac function related to the effects of intraoperative myocardial ischemia and reperfusion.
This study has several limitations. Maintenance of an atrial communication during repair was not randomized, and significantly covaried with patient age and anatomic diagnosis. The data were collected retrospectively; complete data were not available in all patients. Right atrial pressure and systemic desaturation were used as indicators of right ventricular diastolic dysfunction. These indicators were chosen because they are clinically relevant. However, they may not reflect more direct measurements of right ventricular function. They may also be affected by other factors, such as tricuspid regurgitation, pulmonary hypertension, intrapulmonary shunting, and branch pulmonary artery stenosis.
In conclusion, complete repair of tetralogy of Fallot or truncus arteriosus using an atrial communication results in decreased systemic oxygenation during the first 48 hours. Systemic oxygenation reaches a nadir at 24 hours after surgery and improves by the time of hospital discharge. Significant desaturation occurs in a minority of patients and is of a degree that is clinically well tolerated. Infants undergoing repair before 2 months of age are at highest risk for systemic desaturation. We suggest that the effects of an interatrial communication on systemic oxygenation should not be considered a contraindication to complete repair in early infancy. Patient age younger than 2 months at the time of surgery may be the most useful indication for maintenance of an atrial communication during repair.
| Discussion |
|---|
|
|
|---|
My other question or comment is that sometimes you can do this whole operation through a ventriculotomy and never look at the PFO. Did you always look at the interatrial septum? And were there any cases when you actually thought there was a large PFO or small ASD (atrial septal defect), where you actually made it smaller?
I have had the experience where I have left a large PFO open only to have the patient desaturate significantly in the first 24 hours. After that experience I have always made it a routine to look at the atrial communication, and, if it is large, to close it down to about 4 mm.
DR LAUDITO: Thank you, Dr Austin. I appreciated your comment because it gives me the opportunity to explain that the PFO was assessed preoperatively with an echocardiogram and most of the time during surgery. In those situations when the atrial septal defect was assessed to be larger then the requested size, it was partially closed and essentially left with a dimension of 4 or 5 mm.
DR JOHN H. CALHOON (San Antonio, TX): Antonio, that was a very nicely presented talk. Just to follow up on what Dr Austin said, so the PFO or atrial communication you try to leave is what size exactly? Is it 4 to 5 mm, is that what you just said?
DR LAUDITO: That is correct.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. P. Graham Jr The Year in Congenital Heart Disease J. Am. Coll. Cardiol., July 24, 2007; 50(4): 368 - 377. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |