| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2000;69:1222-1228
© 2000 The Society of Thoracic Surgeons
a Division of Cardiothoracic Surgery, University of California, San Francisco, California, USA
b Division of Pediatric Cardiology, University of California, San Francisco, California, USA
Address reprint requests to Dr Hanley, University of California, San Francisco, 505 Parnassus Ave, M593, San Francisco, CA 94143–0118
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsCommentReferences |
|---|
Methods. To evaluate risk factors for, and significance of, such collaterals, we reviewed angiographic and clinical data for all 76 patients who underwent BCPA between January 1990 and June 1996 and had follow-up catheterization during or before 1997.
Results. The median age at BCPA was 10 months, and the median duration from BCPA to follow-up catheterization was 18 months. Arterial collaterals were detected on follow-up catheterization in 45 patients (59%). Factors associated with collateral development included a prior right-sided systemic-to-pulmonary arterial shunt, lower pre-BCPA end-diastolic ventricular pressure and pulmonary vascular resistance, and use and duration of cardiopulmonary bypass during the BCPA operation. Fourteen of the 45 patients (30%) underwent coil embolization of the collaterals. Forty-three patients have undergone extracardiac conduit Fontan, with 1 early and 1 late death. Collaterals were present in 22 of these patients, 7 of whom underwent pre-Fontan embolization. The duration from BCPA to Fontan was longer in patients with collaterals, but these patients were not more likely to have prolonged effusions than those without, and the duration of tube thoracostomy was significantly shorter in patients with collaterals. Embolization of collaterals did not affect the duration of effusions.
Conclusions. Systemic-to-pulmonary arterial collaterals are common after BCPA. In contrast to prior reports, collaterals were not associated with a higher incidence of prolonged effusions after the Fontan procedure in our experience, and did not correlate with poor outcome.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
|
Angiographic and hemodynamic evaluation
All patients had cardiac catheterization performed before and after BCPA (follow-up catheterization in 1997 or before), with the angiograms and catheterization reports available for review. Hemodynamic and oximetric data were measured at pre-BCPA and post-BCPA cardiac catheterization. Aortic, superior caval venous, and atrial pressures and oxygen saturations were recorded. Aortography, superior caval venography, and pulmonary arteriography were performed in all patients. From 1991 to mid 1993, selective angiography of the brachiocephalic and thyrocervical vessels was performed sporadically (10 of 20 patients). From late 1993 through 1997, subclavian arteriography was performed routinely, and angiography of the thyrocervical trunks and/or internal thoracic arteries was performed in most patients (48 of 56 patients).
Systemic-to-pulmonary arterial collaterals were identified on review of systemic arteriograms. They were included in the present study if they had a readily identifiable origin, supplied the lung parenchyma, and opacified the pulmonary arteries and/or veins. Collaterals in each patient were characterized as single or multiple, to the left or right lung, and by size, as small, moderate, or large. Pulmonary filling defects were identified on review of pulmonary arteriograms.
In patients who underwent coil embolization of collaterals at the time of follow-up angiography, standard techniques of embolization were employed. The collaterals were evaluated angiographically following the embolization procedure to assess the success of occlusion.
Data analysis
Preoperative and perioperative data were collected on retrospective review of patient records, and are expressed as median and range or mean ± standard deviation unless otherwise specified. Patients were analyzed as 2 separate cohorts: 1) all patients, and 2) patients who underwent completion of a modified Fontan circulation during the study period. In the first cohort, the primary dependent variable was the presence of systemic-to-pulmonary arterial collaterals on follow-up angiography. The unit of analysis was not the collateral, but the patient, so patients with multiple collaterals were given the same weight as those with a single collateral. Independent variables analyzed for the overall cohort included age at BCPA, primary diagnosis, number of previous procedures, prior systemic-to-pulmonary artery shunts and the laterality of such shunts, hemodynamic and oximetric values measured at catheterization prior to BCPA, the presence of an additional (usually antegrade) source of pulmonary blood flow with the BCPA, bilateral BCPA, use and duration of cardiopulmonary bypass at the time of BCPA, duration of effusions following BCPA, hemodynamic and oximetric variables in the early postoperative period and at follow-up catheterization, regional pulmonary perfusion deficits, and the duration between BCPA and follow-up catheterization. In the second cohort, the primary outcome variable was early survival after Fontan completion. The secondary outcome variable analyzed was duration of chest tube drainage (pleural effusions). Independent variables evaluated for this cohort included those listed above, as well as duration between BCPA and Fontan completion, laterality, size, and number of systemic-to-pulmonary arterial collaterals, the presence of systemic venous collaterals, embolization of arterial collaterals at the time of pre-Fontan catheterization, fenestration of the Fontan, pulmonary arterioplasty at the time Fontan operation (indicating subjective assessment of pulmonary arterial stenosis and hypoplasia), and use and duration of cardiopulmonary bypass during the Fontan procedure. Fisher’s exact test and independent samples Students’ t-test were used to compare frequencies of dichotomous independent variables and mean values of continuous independent variables, respectively.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
|
Systemic-to-pulmonary arterial collaterals and outcome after modified Fontan
A modified Fontan operation was completed in 43 of the 76 patients followed, a median of 22 months (1 to 64 months) after BCPA. An extracardiac conduit Fontan was placed in 42 patients, while the remaining patient had a lateral tunnel Fontan performed elsewhere. There was 1 early death from a stroke after the Fontan procedure, with concomitant repair of a stenotic pulmonary venous orifice. The only late death occurred 5 years post-Fontan following enlargement of a restrictive bulboventricular foramen. The other 41 patients are alive with an intact Fontan circulation.
Systemic-to-pulmonary arterial collaterals were documented prior to Fontan completion in 22 of the 43 patients (51%), 7 of whom underwent coil embolization with complete occlusion of the collaterals. The duration from BCPA to Fontan was significantly longer in patients with than without collaterals (33 ± 15 vs 20 ± 12 months; p = 0.003). Due to the small number of deaths, the relationship between collaterals and post-Fontan mortality could not be adequately assessed. The patient who died early had no evidence of collaterals, while small collaterals were present in the patient who died after reoperation. The duration of chest tube drainage after Fontan was significantly shorter in patients with than without collaterals (8 ± 6 vs 19 ± 15 days; p = 0.007), and the likelihood of chest tube drainage for 14 days or longer was lower in patients with than without collaterals (9% vs 44%; p = 0.02). When patients who underwent catheterization before 1994 (prior to routine use of selective subclavian branch angiography) were excluded, there was no difference in the frequency of prolonged effusions or the average duration of effusions (p > 0.55 for both). Patients with collaterals were less likely to have the Fontan fenestrated than those without (32% vs 59%; p = 0.07), and were more likely to have pulmonary arterioplasty at the time of Fontan (67% vs 24%; p = 0.04). Embolization of collaterals at pre-Fontan catheterization had no significant effect on these outcomes. Among patients undergoing Fontan completion, none of the other factors analyzed were found to correlate with the presence of collaterals.
|
Comment |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
Factors related to the development of systemic-to-pulmonary arterial collaterals after BCPA
It is not understood why systemic-to-pulmonary arterial collaterals develop after BCPA. Abnormal physiologic characteristics of blood flow in a pulmonary circulation that depends entirely or primarily on a BCPA include subnormal oxygen content, decreased absolute volume and velocity of flow, decreased or nearly absent pulsatility, and lack of hepatic venous effluent. Some of these may be factors in the development of systemic-to-pulmonary arterial collaterals. Arterial collaterals form in circumstances other than a simple BCPA, including univentricular physiology palliated with a Fontan connection and no prior BCPA [8, 11, 12], various types of acquired lung pathology [14], and experimental banding and ligation of the pulmonary artery [15].
In previous studies, investigators have found a number of factors that correlate with the presence of collaterals. Unfortunately, the lack of serial studies before and after the development of collaterals in these and the present study obscure the causal axis. Triedman and associates, in the most comprehensive investigation of this issue, showed that patients with collaterals after BCPA had significantly higher oxygen saturations in the superior caval vein and in the aorta, a higher maximum step-up in oxygen saturation from the superior caval vein to the pulmonary arterial system, and a higher frequency of upper lobe filling defects than patients without collaterals [8]. Salim and colleagues, in a study that included only 17 patients, found the same association between collaterals and saturations in the superior caval vein and aorta [13]. Increased systemic venous and arterial oxygen saturations in this setting are most likely the result of, rather than a cause of, collateral flow (that is, increased pulmonary blood flow). Indeed, Salim and associates found that the pulmonary-to-systemic blood flow ratio was significantly higher in patients with than without collaterals [13]. They also found that pressure in the superior caval vein was lower in patients with than without collaterals.
In our experience, pre-BCPA factors that were associated with the development of collaterals included lower pulmonary vascular resistance, lower end-diastolic ventricular pressure, and a history of a prior right-sided shunt. Operative factors at the time of BCPA that were associated with the development of collaterals included use of and longer duration of cardiopulmonary bypass, and the absence of an additional source of pulmonary blood flow. It is not clear how most of these factors may have predisposed to the formation of collaterals, or whether they were simply markers for factors of greater significance. Pulmonary blood flow in addition to the BCPA may protect against the development of collaterals through either hemodynamic or oximetric factors. We did not find any difference in arterial or systemic venous oxygen saturations between patients with and without collaterals, as did Triedman and associates [8] and Salim and colleagues [13], which may have been due in part to the fact that approximately half of our patients had an additional source of pulmonary blood flow. Similarly, the association that we observed between collaterals and pre-BCPA pulmonary vascular resistance, use of and longer duration of bypass, and prior history of right-sided shunts may implicate early postoperative pulmonary vascular function. Given the correlation between collaterals and a previous right-sided shunt, it is notable that collaterals to the right lung were significantly more common than to the left.
Correlations between the amount of collateral flow (as opposed to simply the presence of collaterals) are more difficult to discern. Calculating the total cross-sectional area of angiographically identifiable and measurable collaterals is one method of estimating collateral flow, but this is difficult to achieve with accuracy in many collaterals and does not account for very small collaterals, which may carry a significant amount of flow. Nevertheless, Triedman and colleagues found a correlation between the estimated total cross-sectional area of collaterals and both oxygen saturation in the superior caval vein and the maximum step-up in saturation from the superior caval vein to the pulmonary artery [8]. Another method of estimating the volume of flow through collaterals is to collect pulmonary venous effluent carefully during cardiopulmonary bypass (with the pulmonary arteries snared) at the time of Fontan operation and calculate the ratio of this flow to that delivered through the aortic cannula. This method may be confounded by the nonphysiologic flow conditions (nonpulsatile, relatively low pressure flow) during cardiopulmonary bypass with hypothermia, by flow through the bronchial arteries, and by inaccurate measurement. In addition, it requires intracardiac access, which is not routinely obtained in patients undergoing extracardiac modifications of the Fontan operation, as performed in our center. Using this method in 33 patients undergoing the Fontan operation (most of whom did not undergo prior BCPA), Ichikawa and coworkers found several factors to correlate with an increased percentage of pump flow returning through the pulmonary veins, including older age at Fontan and lower preoperative pulmonary artery area index. In addition, all of their patients with greater than 30% return through the pulmonary veins had a pre-Fontan systemic arterial oxygen saturation greater than 75% and a pulmonary artery area index less than 55% of normal, suggesting that collaterals tend to develop in patients with small pulmonary arteries (and presumably lower pulmonary blood flow through the pulmonary arteries) and serve to increase systemic oxygen delivery.
Clinical implications of systemic-to-pulmonary arterial collaterals after BCPA
The potential for adverse affects of systemic-to-pulmonary arterial collaterals after cavopulmonary anastomosis has been acknowledged since early in the clinical experience of Glenn and colleagues [16]. However, the significance of such collaterals has yet to be well defined. The most serious potential complications of collateral flow are an increased volume load on the functionally single ventricle heart and the effects of flow at systemic pressures in the pulmonary vascular system, which may have implications both in the critical early post-Fontan period and chronically. The volume load imposed by collaterals may be significant, with up to 55% of pump flow returning to the heart through the pulmonary veins in the study by Ichikawa and associates [11], and an estimated 8% of cardiac output in the report of Salim and colleagues [13]. Pre-Fontan end-systolic left ventricular volume was not found to correlate with collateral flow in the report of Ichikawa and coworkers [11], so the presumed cardiac volume load was not substantiated on this basis. Triedman and colleagues [8] found a higher cardiac index in Fontan patients with collaterals than those without (not statistically significant) and a significant correlation between cardiac index and the estimated total cross-sectional area of collaterals. Also in Fontan patients, they demonstrated a higher maximum step-up in oxygen saturation from the superior caval vein to the pulmonary artery and a higher superior caval venous oxygen saturation, but a lower aortic oxygen saturation, which is consistent with higher pulmonary blood flow and more extensive mismatch between ventilation and perfusion. It is not clear, however, to what extent the presence of a Fontan fenestration may have affected these results.
Evidence of pulmonary vascular complications resulting from collateral flow is lacking. In Ichikawa’s series, all 4 patients with greater than 33% of pump-flow returning through the pulmonary veins and post-Fontan systemic venous pressure greater than 17 mm Hg died [11]. They hypothesized that increased collateral flow during cardiopulmonary bypass (increased pulmonary blood flow) may lead to elevated pulmonary vascular resistance and more pronounced systemic venous hypertension in the early postoperative period, which is a well-established risk factor for early mortality after the Fontan operation [17, 18]. It is debatable whether this mechanism was truly responsible for the poor outcome in these patients, given the inaccuracy of the method and their high rates of overall early mortality (24%) and failure (27%).
In a study looking for factors associated with prolonged effusions following Fontan, Spicer and associates found collaterals to be the only significant predictor [12]. Hospital stay was also longer in these patients, but no other differences in outcome were noted. It did not appear that other implications of collaterals were assessed directly, as the purpose of the study was to identify correlates of prolonged effusions. Accordingly, the hemodynamic basis underlying the difference in effusion rates between patients with and without collaterals cannot be inferred from their results. Ichikawa and coworkers found higher collateral flow to correlate with higher postoperative systemic venous pressure [11], which might contribute to increased effusions, insofar as pleural drainage is primarily through the lymphatic system, and systemic venous hypertension in a Glenn or Fontan circulation tends to produce lymphatic hypertension and consequently impaired drainage.
In our study, systemic-to-pulmonary arterial collaterals did not correlate with adverse outcome. In fact, the average duration and overall frequency of prolonged (> 14 days) chest tube drainage was less in patients with than without collaterals, and there was no difference when only patients who underwent follow-up catheterization and Fontan completion in 1994 or after were analyzed. It is important to note that detecting and making sense of differences in the severity of effusions may be confounded by the subjective nature of decisions regarding tube thoracostomy. Our strategy in Fontan patients is to leave chest tubes in place until drainage is less than 1 mL/kg/day for at least 2 days, in an effort to prevent the more debilitating problem of chronic effusions, which have been uncommon in our experience. Thus, duration of chest tube drainage in our patients may be longer than in other studies. Another potentially confounding factor is the decrease in pleural drainage over the course of the study period. In fact, when dates of Fontan or follow-up catheterization were controlled for, there was no difference in the duration of effusions.
Management of systemic-to-pulmonary arterial collaterals after BCPA
In our experience and that of others, coil embolization appears to be effective for occluding angiographically identifiable systemic-to-pulmonary arterial collaterals. However, there still may be significant collateral flow in patients who have undergone successful coil embolization (ie, complete occlusion), and there is little evidence that occlusion has a salutary effect on outcomes [11]. Spicer and associates showed that patients with collaterals who underwent embolization had a significantly shorter duration of effusions after modified Fontan operation than those who did not have their collaterals embolized [12]. Our experience was not consistent with this finding. This may be confounded by the subjective nature of decisions to embolize collaterals and to continue postoperative tube thoracostomy, as discussed above. In addition to uncertainty about the efficacy of embolization at reducing collateral flow, it is not known whether significant collateral flow will recur following embolization. If the factor(s) that led to their development in the first place are still present, it stands to reason that collaterals may reform, although any such factors may have been eliminated with the completion of the Fontan circulation. Among a small number of Triedman’s patients with serial angiograms and evidence of collaterals on the first angiogram, 70% of those who underwent embolization had collaterals on the subsequent study, though there was a trend towards a decrease in the number of vessels, while patients who did not undergo embolization tended to have an increased number of collaterals [8]. Ultimately, angiographic follow-up after embolization of post-BCPA collaterals is limited, and additional data will be required before the long-term efficacy of such interventions can be determined.
All of these caveats notwithstanding, it seems advisable to embolize significant collaterals when they are detected. The primary rationale for aggressive embolization of systemic-to-pulmonary arterial collaterals before (or after) Fontan is that they may have adverse implications for ventricular function and/or the pulmonary vasculature, which can have a critical impact on the function of a Fontan circulation, both in the perioperative period and chronically. The potential benefits of reducing the ventricular volume load and high pressure flow in the pulmonary arteries are substantial, even if such benefits have not been demonstrated definitively. Thus, we recommend embolization of all large or moderate collaterals in patients undergoing pre-Fontan evaluation, as well as discretely identifiable collaterals in patients with multiple small vessels. Interventional cardiologists are becoming more aggressive with vascular embolization, as it has become clear that the procedure is highly successful, with potential complications that are minor and uncommon for the most part [19]. Our experience has paralleled this trend, as our tendency to embolize collaterals has increased over time: 4 of 26 patients with collaterals detected prior to 1996 underwent embolization versus 10 of 19 in 1996 to 97. Ultimately, the role of embolization in the management of collaterals after BCPA will require additional experience and the test of time.
Limitations of this and previous studies
There are a number of limitations of this and previous studies on this topic [7, 10–13]. Nonuniform angiography practices almost certainly resulted in underdetection of collaterals, as a number of technical factors have been shown to correlate with the sensitivity of angiography for detecting second and third order branches from the aorta [8, 20]. Undersurveillance almost certainly biased the results of this and previous studies, and indeed, when we divided patients according to the date of follow-up angiography, the incidence of collaterals detected was significantly higher in patients undergoing angiography in 1994 or after. Another limitation of this and previous studies is that they do not allow for inference of the natural history of collaterals or a causal nexus between collaterals and factors found to be associated with their development. Serial hemodynamic assessment and imaging studies will be necessary to document the time frame of collateral development, to determine factors predisposing to (as opposed to resulting from) collaterals, and to characterize the hemodynamic perturbations that they may engender. Also, existing studies have not characterized with accuracy the burden of ventricular volume load and the pulmonary vascular effects of systemic-to-pulmonary collaterals in a single ventricle circulation. Ultimately, it is these factors that are of most concern as potential complications of collaterals, and the lack of information to this effect is a significant shortcoming in our understanding of this condition. Finally, the fact that limited data have been collected on the long-term efficacy of embolization, and the subsequent development of new collaterals after embolization, hinder our ability to make evidence-based decisions on who should (or should not) undergo embolization before or after Fontan completion.
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
Related Article
Ann. Thorac. Surg. 2000 69: 1228.
This article has been cited by other articles:
|
|
 |
|
  A. J. Powell Aortopulmonary Collaterals in Single-Ventricle Congenital Heart Disease: How Much Do They Count? Circ Cardiovasc Imaging, May 1, 2009; 2(3): 171 - 173. [Full Text] [PDF]
|
|
|
|
|
|
L. Grosse-Wortmann, A. Al-Otay, and S.-J. Yoo Aortopulmonary Collaterals After Bidirectional Cavopulmonary Connection or Fontan Completion: Quantification With MRI Circ Cardiovasc Imaging, May 1, 2009; 2(3): 219 - 225. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. W. Brown, K. Gauvreau, A. J. Powell, P. Lang, S. D. Colan, P. J. del Nido, K. C. Odegard, and T. Geva Cardiac Magnetic Resonance Versus Routine Cardiac Catheterization Before Bidirectional Glenn Anastomosis in Infants With Functional Single Ventricle: A Prospective Randomized Trial Circulation, December 4, 2007; 116(23): 2718 - 2725. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Januszewska, A. Stebel, and E. Malec Consequences of Right Ventricle to Pulmonary Artery Shunt at the First Stage for the Fontan Operation Ann. Thorac. Surg., November 1, 2007; 84(5): 1611 - 1617. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Sallehuddin, A. Mesned, M. Barakati, M. A. Fayyadh, F. Fadley, and Z. Al-Halees Fontan completion without surgery Eur. J. Cardiothorac. Surg., August 1, 2007; 32(2): 195 - 200. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R Kaulitz and M Hofbeck Current treatment and prognosis in children with functionally univentricular hearts Arch. Dis. Child., July 1, 2005; 90(7): 757 - 762. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. S. Ro, J. Rychik, M. S. Cohen, W. T. Mahle, and J. J. Rome Diagnostic assessment before Fontan operation in patients with bidirectional cavopulmonary anastomosis: Are noninvasive methods sufficient? J. Am. Coll. Cardiol., July 7, 2004; 44(1): 184 - 187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Gupta, C. Daggett, S. Behera, M. Ferraro, W. Wells, and V. Starnes Risk factors for persistent pleural effusions after the extracardiac Fontan procedure J. Thorac. Cardiovasc. Surg., June 1, 2004; 127(6): 1664 - 1669. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. W. Brown, K. Gauvreau, A. M. Moran, K. J. Jenkins, S. B. Perry, P. J. del Nido, and S. D. Colan Clinical outcomes and utility of cardiac catheterization prior to superior cavopulmonary anastomosis J. Thorac. Cardiovasc. Surg., July 1, 2003; 126(1): 272 - 281. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Kaulitz, G. Ziemer, T. Paul, M. Peuster, H. Bertram, and G. Hausdorf Fontan-type procedures: residual lesions and late interventions Ann. Thorac. Surg., September 1, 2002; 74(3): 778 - 785. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Bradley, M. M. McCall, J. J. Sistino, and W. A.K. Radtke Aortopulmonary collateral flow in the Fontan patient: does it matter? Ann. Thorac. Surg., August 1, 2001; 72(2): 408 - 415. [Abstract] [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 |
