| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2005;79:596-606
© 2005 The Society of Thoracic Surgeons
a Department of Cardiac Surgery, Children's Hospital Boston, Harvard Medical School, Boston, Massachusetts
b Department of Cardiovascular Surgery, Children's Hospital Boston, Harvard Medical School, Boston, Massachusetts
c Department of Cardiology, Children's Hospital Boston, Harvard Medical School, Boston, Massachusetts
d Department of Biostatistics, Children's Hospital Boston, Harvard Medical School, Boston, Massachusetts
Accepted for publication July 6, 2004.
* Address reprint requests to Dr Hancock Friesen, #2269 New Halifax Infirmary, 1796 Summer St, Halifax, NS B3H 3A7, Canada (E-mail: camillehf{at}hotmail.com).
Presented at the Fortieth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 26–28, 2004.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDISCUSSIONAcknowledgmentsReferences |
|---|
METHODS: All patients with surgically corrected TAPVC from November 1989 to December 2000 were included. Charts were reviewed for patient demographics, operation variables, and postoperative course.
RESULTS: There were 123 patients in the cohort, of which 72 (59%) were male. The median age and weight at operation were 10 days and 3.6 kg, respectively. Sixty-eight (55%) patients presented with pulmonary venous obstruction, and 65 (53%) underwent emergent TAPVC repair. Thirty-nine (32%) had single-ventricle anatomy, and 84 (68%) had two-ventricle anatomy. Thirty patients (24%) died. Kaplan-Meier survival at 1 month was 65% (95% confidence interval [CI], 55% to 75%) for single-ventricle patients versus 90% (95% CI, 90% to 100%) for two-ventricle patients; at 36 months it was 47% (95% CI, 35% to 59%) versus 87% (95% CI, 81% to 93%), respectively. By Cox multivariable regression analysis, a single ventricle (p < 0.001, hazard ratio, 4.8; 95% CI, 2.5 to 9.2) was an independent mortality risk factor. Prerepair pulmonary venous obstruction was a multivariate risk factor for death among single-ventricle patients. Postrepair pulmonary venous obstruction occurred in 11%. If year of operation is used as a predictor, two-ventricle patient survival has significantly improved (p < 0.05).
CONCLUSIONS: Despite current interventions, single-ventricle patients continue to have a worse prognosis than two-ventricle patients.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDISCUSSIONAcknowledgmentsReferences |
|---|
Surgical repair of TAPVC has historically been associated with a significant risk of early mortality, with reports in the literature ranging from 10% to 50% [1–4]. Worse outcomes have been noted in patients with a single ventricle and, specifically, heterotaxy anatomy [1, 5, 6]. The most serious complication after TAPVC repair is pulmonary vein stenosis. This can be intrinsic to the pulmonary veins or at the anastomosis and is reported to occur in 8% to 54% of cases [4, 5, 7–9]. Earlier studies indicated that two-ventricle patients with preoperative pulmonary venous obstruction were at a higher risk of postrepair pulmonary venous obstruction, but that has been neutralized as a risk factor in some more recent studies [4].
Nitric oxide therapy has been found to be efficacious in treating the postoperative reactive pulmonary arterial hypertension after the repair of an obstructed TAPVC [10, 11]. There is some evidence that nitric oxide is useful in treating pulmonary arterial hypertension that is secondary to obstructed TAPVC [12]. However, whether nitric oxide use perioperatively will have any impact on long-term outcomes such as postrepair pulmonary vein stenosis is not clear at present.
This review was performed to assess the impact of current management strategies on the outcomes of pulmonary vein stenosis (early and late mortality, and postrepair pulmonary venous obstruction).
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDISCUSSIONAcknowledgmentsReferences |
|---|
Definitions
Echocardiography is our preferred diagnostic test, and 94 of the 123 patients (76%) proceeded to operation with only echocardiographic data. The indications for catheterization preoperatively were anatomy that was unresolved by echocardiography, characterization of pulmonary venous obstruction, or other associated anatomy that required delineation or intervention.
Preoperatively, pulmonary veins were deemed obstructed if there was echocardiographic (57/120 patients) or angiographic data (11/29 patients) that indicated a significant gradient between the pulmonary veins and their point of drainage (flow acceleration > 2 m/s by echocardiography; or a pressure gradient > 4 mm Hg, or a reduction of > 50% of the diameter of a single pulmonary vein by catheter and angiography). An operation was classified as emergent if the patient was taken to the operating room within the first 24 hours after presentation to Children's Hospital for hemodynamic or ventilatory compromise.
Postrepair pulmonary venous obstruction was diagnosed by clinical presentation or during routine radiographic surveillance. If the clinical presentation suggested postrepair recurrent, persistent, or new onset pulmonary venous obstruction, the diagnosis was confirmed radiographically by using echocardiographic or angiographic criteria as previously defined for the prerepair diagnosis of pulmonary venous obstruction. Postrepair pulmonary venous obstruction was defined as early if it occurred 30 days or less after the initial TAPVC repair operation and late if it occurred more than 30 days after. The descriptions of angiographic, echocardiographic, and surgical findings were used to classify the postrepair pulmonary venous obstruction as anastomotic stenosis or intrinsic pulmonary venous stenosis (or a combination of both).
Isolated or simple TAPVC was diagnosed if the patient had TAPVC in association with a secundum atrial septal defect or a patent ductus arteriosus, or both.
Patients were classified as having single-ventricle or two-ventricle physiology by whether their anatomy was deemed suitable for a two-ventricle or single-ventricle repair. Single-ventricle patients were further subdivided into groups with heterotaxy syndrome (defined as patients with complex anomalies of visceral–cardiac relations in association with intracardiac pathology) [13] or other forms of single-ventricle anatomy. Because of the small number of single-ventricle patients that were nonheterotaxy (n = 2), all patients with a single ventricle were grouped for analysis.
A simple operation was defined as any operation that entailed rerouting the pulmonary venous return to the left atrium and included repair of an atrial septal defect (primary suture closure or patch repair) or ligation of a patent ductus arteriosus, or both. An operation was defined as complex if any other additional concomitant procedures were performed, such as repair of a ventricular septal defect, repair of a coarctation, or augmentation of a hypoplastic aortic arch. Likewise, the operation was classified as complex if the TAPVC repair was combined with the palliative procedures for a single ventricle, such as Blalock-Taussig shunt, Stage 1 Norwood operation, Damus-Kaye-Stansel with Blalock-Taussig shunt, pulmonary artery banding, bidirectional cavopulmonary connection, or Fontan.
Surgical variables that were identified and analyzed were length of cardiopulmonary bypass time, use of profound hypothermic circulatory arrest, and use of absorbable versus nonabsorbable suture for the common pulmonary vein-to-left atrial anastomosis.
We also noted two distinct operative techniques that are used to perform the anastomosis for supracardiac, infradiaphragmatic, or mixed TAPVC. The type 1 repair (external) is performed in a normally sited heart by lifting the apex of the left ventricle toward the right shoulder, marking the proposed site of anastomosis on both the front wall of the pulmonary venous confluence and the back wall of the left atrium, and then lowering the apex to confirm the point of closest apposition is marked appropriately. The left atrium and confluence are incised, and the anastomosis performed from an extracardiac position [9].
The alternate type 2 (internal) technique is performed transatrially–transeptally and is analogous to the type II repair described by Wilson and colleagues [9]. With the heart in an anatomic position, a right atriotomy is performed, and the incision is extended across the interatrial septum into the back wall of the left atrium. The front wall of the pulmonary vein confluence is then incised and anastomosed to the left atrium, visualized by the surgeon through the back wall incision in the left atrium.
Early mortality was defined as death that occurred in the first 30 days after the operation. Late mortality was defined as death that occurred more than 30 days after surgery or after discharge from hospital (in the case of extended hospitalization).
Statistical Analysis
Data are presented as mean ± the standard deviation or median and interquartile range (IQR) as appropriate. Estimated survival and freedom from postrepair pulmonary venous obstruction were determined with the Kaplan-Meier method by using the product-limit estimator, and 95% confidence intervals (CI) were constructed around the curves according to Greenwood's formula [14].
The following variables were evaluated in the univariable and multivariable analysis to determine significant predictors of early death, overall mortality, and reoperation because of pulmonary venous obstruction: age at surgery, weight at surgery, gender, total pump time, cross-clamp time, circulatory arrest time, preoperative pulmonary venous obstruction, single-ventricle (vs two-ventricle) anatomy, emergent (vs elective) operation, complex (vs simple) operation, type 1 (vs type 2) repair, and nonabsorbable (vs absorbable) suture. Reoperation for pulmonary venous obstruction was also tested as a possible predictor of death.
The primary outcomes of the study were mortality (early or late) and postrepair pulmonary venous obstruction. Variables were evaluated using the log-rank test in univariable analysis and by the likelihood ratio test in the Cox proportional-hazards regression model for multivariable analysis [15]. Hazard ratios with 95% CIs were constructed for the significant multivariable predictors. The final selection of models was determined by the backward stepwise procedure, with variables having p of less than 0.1 in the univariable analysis entered as candidates into the Cox regression model.
Simple binomial proportions were compared by Fisher's exact test. All reported p values are two-tailed. A power analysis indicated that sample sizes of 39 single-ventricle and 84 two-ventricle patients would provide 90% power (
= 0.05, ß = 0.1) to detect a significant difference in survival between the cohorts by using the log-rank test for comparing the area under the survival curves, assuming a hazard ratio of 3.0 (version 5.0, nQuery Advisor, Statistical Solutions, Boston, MA). Statistical analysis was performed with the SPSS software package (Version 12.0, SPSS Inc, Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDISCUSSIONAcknowledgmentsReferences |
|---|
|
The TAPVC anatomy was supracardiac in 70 (57%) cases, cardiac in 8 (7%), infradiaphragmatic in 35 (28%), and mixed in 10 (8%) (Tables 1 and 2). Of the 84 two-ventricle patients, 46 (55%) were obstructed at the time of operation compared with 22 (56%) of the 39 the single-ventricle patients. When proportions of total mortality, early death, and postrepair pulmonary venous obstruction among the four anatomic subgroups were tested, there were no significant differences by the Pearson 
2 test. In subgroup analysis, a comparison of mortality in single-ventricle patients with supracardiac TAPVC anatomy (n = 24) to infracardiac (n = 9) by Fisher's exact test showed the mortality rate was statistically higher in the infracardiac subgroup (8/9, 89%) than in the supracardiac group (9/24, 38%), p = 0.02. The events in the other subgroups were too infrequent to sustain statistical analysis.
|
|
Most of the two-ventricle patients had minor associated defects, including patent ductus arteriosus in 32 patients and secundum atrial septal defect in 81. One patient had an associated coarctation, and 1 patient had complex anatomy identified segmentally as {ADD} transposition of the great arteries, complete atrioventricular canal, pulmonary stenosis, and hypoplastic pulmonary arteries. Of the 46 patients with prerepair pulmonary venous obstruction, 19 had evidence of obstruction of the ascending vertical vein. Twenty-two had evidence of intrahepatic obstruction to pulmonary venous flow. Four patients had intrinsically small pulmonary veins with no other discrete stenotic sites, and 1 patient had unilateral or single pulmonary venous obstruction (intrahepatic obstruction of the left pulmonary vein).
Preoperative Physiology
Fifty-three patients required preoperative inotropic support, and 17 cyanotic infants were treated with intravenous prostaglandin E1 preoperatively in an effort to stabilize them before diagnosis of TAPVC. The peak preoperative serum creatinine level was 0.84 ± 0.58 (range 0.2 to 4.5). The minimum recorded pH averaged 7.34 ± 0.12 (range 6.84 to 7.58) and the average pH at induction was 7.41 ± 0.11 (range 7.10 to 7.70). In difference in pH at induction between single-ventricle (7.39 ± 0.11) and two-ventricle patients (7.42 ± 0.11, p = 0.20) was not significant. Likewise, there was no significant difference in minimum preoperative pH (7.35 ± 0.10 vs 7.32 ±.15, p = 0.28) or preinduction pH (7.40 ± 0.11 vs 7.41 ± 0.12, p = 0.92) between survivors and nonsurvivors.
The median (interquartile range) total cardiopulmonary bypass time for all patients was 94 minutes (80 to 118 minutes), and median cross clamp duration was 41 minutes (36 to 51 minutes). Profound hypothermic circulatory arrest was used in 102 (83%) patients with a median duration of 37 minutes (28 to 45 minutes).
Operative Technique: Simple Versus Complex Repair
The TAPVC repairs were simple (isolated) in 74 (60%) patients and complex in 49 (40%). In 29 patients complex repairs included the TAPVC repair plus a single additional procedure: pulmonary arterioplasty in 1; Blalock-Taussig Shunt/Stage I Norwood procedure in 16; bidirectional cavopulmonary connection in 2; aortic arch augmentation in 2; Warden procedure in 3 [16]; central shunt in 1; Fontan in 2; ventricular septal defect repair in 1; and atrial septectomy in 1. In the other 12 complex repairs, TAPVC was performed with two or more additional procedures: pulmonary artery band and atrioventricular valvuloplasty in 1 patient, Fontan with takedown Blalock-Taussig shunt in 4 (3 with a simultaneous bidirectional cavopulmonary connection, and 1 with a pacemaker), bidirectional cavopulmonary connection with takedown Blalock-Taussig shunt in 5 (two performed concomitantly with an atrioventricular valve replacement or repair), central shunt with pulmonary arterioplasty in 2, and ventricular septal defect repair with pulmonary artery band takedown and pulmonary arterioplasty in 1 patient.
Two of our 39 single-ventricle patients were candidates for a simple repair (TAPVC alone) at first operation because of well-balanced pulmonary and systemic blood flows. One of these patients had obstructed anomalous pulmonary venous return with an emergency operation, and 1 had elective repair for nonobstructed veins. These patients later had second or third operations as necessary to address their single-ventricle physiology. A complex repair was undertaken at the time of TAPVC repair in the remaining 37 (95%) single-ventricle patients.
As is intuitive, the frequency of complex repair is much higher in single-ventricle (37/39) than in two-ventricle patients (12/84), with virtually all single-ventricle patients requiring initial surgical palliation for single-ventricle physiology at the time of the TAPVC repair. Reflecting the outcomes in the entire single-ventricle cohort, the subgroup of single-ventricle patients that required a complex repair suffered a significantly worse overall mortality rate (67%) than those treated with a simple repair (31%, p < 0.01).
Operative Technique: Type 1 Versus Type 2 Repair
A type 1 repair was used in 72 patients, whereas the remainder (50, as one patient did not have data on operative positioning of the heart) had a type 2 repair (Fig 1). In 79 patients the anastomosis was performed with absorbable suture, in 34 the anastomosis was performed with nonabsorbable suture, and in 10 either no suture was necessary because the procedure was unroofing of the coronary sinus (n = 4) or the suture type had not been recorded (n = 6). The vertical vein was ligated in 87 patients.
The operative technique was found to have a significant impact on total mortality. The 2 analysis indicated a significant difference in the total mortality rate in single-ventricle patients with a type 1 repair (5/16, 31%) compared with type 2 repair (14/22, 64%) (p = 0.05). One patient with a single ventricle and unobstructed pulmonary veins was not captured in this analysis because the operative technique was not specified. The patient had a complex, elective repair and died early (Fig 1).
The modes of death in the type 1 repairs were sudden postoperative death, with no evidence of postrepair pulmonary venous obstruction in 3 patients, and late unknown cause of death in 2 patents. In the type 2 repaired cohort, profound hypoxia or acidosis (or both), with no evidence of postrepair pulmonary venous obstruction, accounted for four deaths. Recurrent pulmonary venous obstruction was the cause of death in 6 patients, and in 4 patients the cause of death was unknown.
In the two-ventricle patients, a trend to higher mortality in the type 2 repaired group was evident only in the supracardiac TAPVC anatomic subgroup (type 1 mortality, 0/19; type 2 mortality, 5/27; p = 0.07).
Early Mortality
The Kaplan-Meier survival curve for all patients was 85% at 1 month (95% CI, 80% to 90%) and 73% at 24 months (95% CI, 66% to 80%) (Fig 2).
|
|
|
Of the 16 patients with a concomitant Blalock-Taussig shunt and TAPVC repair, the early mortality rate was 56% (9/16). Evidence suggests that current management has significantly improved if year of operation is used as a predictor of survival among patients with two ventricles (p < 0.05) but not for those with a single-ventricle anatomy (p = 0.78).
Prerepair Pulmonary Venous Obstruction
A highly significant interaction in the multivariable Cox model indicated that prerepair pulmonary venous obstruction was not a predictor of mortality in the entire cohort but rather only among single-ventricle patients. The estimated Kaplan-Meier 1-year survival rate for single-ventricle patients with prerepair pulmonary venous obstruction was 31% (95% CI, 16% to 46%) compared with 68% (95% CI, 54% to 82%) for patients without obstruction (log-rank test, 5.69; p < 0.01) (Fig 4).
|
30 days after the initial TAPVC repair) and 10 were performed late. Only one reoperation occurred more than 1 year after the index TAPVC operation, and that was at 91 months.
|
A range of surgical and catheter-based techniques were used to treat postrepair pulmonary venous obstruction. Most recurrences were intrinsic pulmonary vein stenosis (7) or mixed intrinsic/anastomotic stenosis (4) and were addressed with unroofing the involved pulmonary veins and augmenting the pulmonary vein–left atrial confluence with pericardium (Table 4). Three early and two late deaths occurred in the patients that required reoperation for postrepair pulmonary venous obstruction.
Freedom from reoperation for pulmonary venous obstruction for the entire cohort was 93% (95% CI, 89% to 97%) at 6 months, 86% (95% CI, 80% to 92%) at 1 year, and 84% (95% CI, 78% to 90%) at 3 and 5 years after TAPVC. For two-ventricle patients, the 1-year freedom from reoperation was 92% (95% CI, 87% to 97%); in single-ventricle patients it was 68% (95% CI, 54% to 82%). The comparison between single-ventricle and two-ventricle patients indicated that reoperation for pulmonary venous obstruction occurred earlier and more frequently in patients with a single ventricle (log-rank test, 5.31; p = 0.02).
With regard to risk of postrepair pulmonary venous obstruction, both a single ventricle and use of nonabsorbable suture in the construction of the common pulmonary vein-to-left atrial anastomosis were risk factors in univariable analysis, but neither variable achieved significance as an independent risk factor (Table 3).
Late Mortality
There were 10 late deaths, 6 of which occurred in single-ventricle patients. Two of the deaths, as already described, occurred after reoperation for postrepair pulmonary venous obstruction (including 1 patient who died post heart/lung transplantation) and 8 occurred in patients who had not had a reoperation. Of this latter group, 5 patients died at home (no autopsy was performed), and 1 had an autopsy revealing multiple cerebral infarcts consistent with an arrhythmic death. Two patients had evidence of postrepair pulmonary venous obstruction and were undergoing palliative procedures to treat diffuse pulmonary stenoses at the time of death.
Nitric Oxide
After its introduction to our institution in 1990, 23 patients received nitric oxide postoperatively. Of the 23 patients treated postoperatively with nitric oxide, 20 had prerepair pulmonary venous obstruction. The mortality in this group was 15% (3/20), with 5 patients requiring reoperation for postrepair pulmonary venous obstruction. This outcome was not significantly different from the 100 patients who did not receive nitric oxide, 48 of whom had prerepair pulmonary venous obstruction, with 17 (35%) deaths and 4 patients requiring reoperation for postrepair pulmonary venous obstruction. Although the number of events is too small to afford statistical analysis, the mortality in those patients with prerepair pulmonary venous obstruction treated with nitric oxide was half that of those in the untreated group, suggesting a potential benefit.
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDISCUSSIONAcknowledgmentsReferences |
|---|
The most outstanding finding of our review is the striking and persistent difference in outcome between single-ventricle and two-ventricle patients, despite introduction of the aforementioned treatment strategies. Azakie and colleagues have reported current early mortality rates of 95% in complex single-ventricle patients with primary neonatal repair [19]. Likewise, recent reports by Gaynor and colleagues indicate that the early mortality in single-ventricle patients is 44%. Mortality is highest in those undergoing a first-stage palliation for a single ventricle with concomitant TAPVC repair [5]. It is not surprising that single-ventricle patients with obstructed pulmonary venous return fare poorly, both because of low preoperative cardiac output and elevated postoperative pulmonary vascular resistance.
In a single-ventricle patient, the effect of obstructive pulmonary veins is particularly compromising because increased pulmonary vascular resistance compromises antegrade pulmonary blood flow. This sets up a vicious circle of increased pulmonary vascular resistance, impaired pulmonary blood flow, and progressive preferential shunting to the systemic circulation with increasingly more desaturated blood flowing to the systemic vasculature (including the coronary arteries). In patients requiring concomitant neonatal TAPVC repair and single-ventricle palliation (such as a Blalock-Taussig shunt), the ratio of pulmonary blood flow to systemic blood flow is compromised because of both physiologic neonatal pulmonary arterial hypertrophy and the effect of the obstructed total anomalous pulmonary veins on pulmonary vascular resistance.
We rely principally on transthoracic echocardiography as the diagnostic test of choice in TAPVC patients (the sole preoperative anatomic assessment in 76% of our patient cohort), as it is noninvasive and highly accurate in providing the anatomic detail necessary for operative planning [20, 21]. ECMO and nitric oxide were used in a minority of patients, 2% and 19%, respectively. It is not possible to demonstrate whether ECMO or nitric oxide has had a significant impact on the long-term outcome in our patients, either in terms of postrepair pulmonary venous obstruction or late mortality, although we noted a trend to reduced early mortality in those patients with obstructed, single-ventricle TAPVC who were treated with nitric oxide.
Prerepair pulmonary venous obstruction was not identified as a risk factor in two-ventricle patients but does, however, persist as a risk factor in single-ventricle patients. Gaynor and colleagues used pathologic studies of obstructed TAPVC to suggest that the degree of pathologic stenosis of pulmonary veins is underestimated by clinical presentation [5]. One might speculate that pulmonary venous obstruction may be uncovered by the creation of increased pulmonary blood flow, as occurs with the creation of a systemic-to-pulmonary artery shunt. Of the 22 single-ventricle patients with preoperative pulmonary venous obstruction, only 1 had intrinsic pulmonary venous obstruction, whereas 21 patients had evidence of draining vein obstruction. Of the 46 two-ventricle patients with prerepair pulmonary venous obstruction, 4 had intrinsic pulmonary venous obstruction, whereas 42 patients had evidence of obstruction of the draining veins.
We would expect that patients with pulmonary venous obstruction, on the basis of obstructed draining veins (63/68), would have an adequate pulmonary venous confluence and thus a successful TAPVC repair. It is possible that our series underestimated the incidence of intrinsic pulmonary venous obstruction because of the large proportion of patients with obstructed TAPVC. Obstruction of the drainage of pulmonary veins remote to the individual veins may cause intrinsically abnormal veins to distend and thus falsely appear normal in caliber. The reason for such a high proportion of our patients having echocardiographic or angiographic evidence of obstruction may be partly accounted for by the referral bias of more complex cases to our tertiary center.
The finding in our cohort that a transatrial repair is associated with higher early mortality directly contradicts the literature that found that a transatrial approach was in fact protective [9]. This finding may reflect the relatively small numbers of patients in the series; however, it is also possible that in type 1 repair, postrepair atrioventricular valvar regurgitation may be less because of less trauma to the atrioventricular groove.
We also hypothesize that the number of arrhythmias or localized autonomic nerve dysfunction in patients with a type 1 repair may be reduced, as there is less dissection through the posterior atrial wall fat pads bearing autonomic ganglia and less disruption of blood supply to the atrial wall, and possibly, to the sinus node. The rhythm status of patients at follow-up is poorly captured, and only 6 patients were identified at follow-up by electrocardiogram or current medications as requiring treatment for arrhythmias. Holter monitor data would be necessary to capture intermittent arrhythmias, which may materially impact on long-term ventricular function. We do not routinely obtain Holter data on our follow-up TAPVC patients, and without this data it will be impossible to draw any conclusions about the impact of arrhythmia on ventricular function.
No data in the current study support either of the two hypotheses we have discussed as to why, in our series, the type 2 operative technique appears to be associated with worse outcome than type 1. To prove these hypotheses, we will require long-term follow-up of the type 1 and type 2 repaired patients as well as a greater number of patients. We continue to use the type 2 approach selectively because of the excellent exposure that this technique affords.
For complex single-ventricle patients, which in our population were principally represented by heterotaxy but also by unbalanced atrioventricular canal and hypoplastic left heart syndrome, our current surgical strategy is dictated principally by the symptomatic lesion at presentation. If the total anomalous pulmonary veins have become symptomatic or obstructive by hemodynamic measurements, they are addressed at the time of the first operation, which if it occurs in the first 3 months of a single-ventricle infant's life with insufficient pulmonary blood flow, would be a Blalock-Taussig shunt with relief of ventricular outflow obstruction as necessary. These operations would be performed simultaneously despite the fact that we recognize the higher risk of mortality in this cohort. Likewise, if the patient has sufficient pulmonary blood flow and no prerepair intrinsic pulmonary vascular obstruction is evident, the TAPVC repair would be delayed until the child had sufficiently low pulmonary vascular resistance such that a bidirectional Glenn could be performed with simultaneous repair of the total anomalous veins.
Genetic reprogramming of pulmonary vein endothelium in the face of obstruction may lead to increased collagen production and fibrotic scar formation. This may be a stereotyped reaction in both single-ventricle and two-ventricle patients but may be manifest clinically more acutely in the volume-loaded early stages of single ventricle palliation. Or, it may be that single-ventricle patients have a different genotypic response to the stress of pulmonary venous obstruction than do two-ventricle patients. Intelligent genomic mining of pathologic tissue may provide us with a better understanding of the molecular basis for the clinical syndrome, TAPVC in the context of single-ventricle anatomy, which we have recognized as so debilitating.
Study Limitations
The median follow-up for our entire patient cohort was 22 months with a mean follow-up of 38 months. The median follow-up for the 69 patients operated on before 1995 was 56 months, and the 55 patients operated on from 1996 and after had a median follow-up of 15 months. Most of the follow-up information was obtained from a review of the medical records. One weakness of this retrospective review is that a cross-sectional method of follow-up was not performed. A more exact accounting of fatal and nonfatal time-related events would improve estimates of patient survival.
Conclusions
Although outcomes have improved over time in two-ventricle TAPVC patients, single-ventricle patients with TAPVC continue to have worse outcomes in terms of early (36% vs 7%) and late (15% vs 5%) mortality, and reoperation for postrepair pulmonary venous obstruction (18% vs 8%). Overall mortality in this series was 24%, which is comparable to other series reporting outcomes of both single-ventricle and two-ventricle patients [1]. Mortality continues to be associated with previously identified risk factors, including prerepair pulmonary venous obstruction, longer cardiopulmonary bypass times, longer circulatory arrest duration, complex operative repairs, and single-ventricle anatomy. In multivariable analysis, however, only single-ventricle anatomy remains an independent risk factor for death. Early and late mortality continue to be significantly higher in patients with a single ventricle even though outcomes for two-ventricle patients improved during the study period.
Preoperative pulmonary venous obstruction has been neutralized as a risk factor for worse outcome in two-ventricle patients, but single-ventricle patients with obstructed TAPVC fare worse than those with unobstructed TAPVC. None of our current perioperative or operative strategies have had an impact on the incidence of postrepair pulmonary venous obstruction, implying that this process may reflect an underlying predisposittion.
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDISCUSSIONAcknowledgmentsReferences |
|---|
DR HANCOCK FRIESEN: Thank you for your questions, Dr Karl. First of all, we have obviously thought a lot about why type 2 was found to be a risk factor for mortality in this analysis, given that there is literature that actually says the opposite, which is that type 2 repair is associated with improved outcome. The first theory to explain why type 2 repair was associated with worse early outcome in our series is that we were not powered to determine the difference between the two repairs and that perhaps this is a statistical aberration. There are fat pads on the posterior aspect of the right and left atria that house autonomic ganglia and plexuses and are important regulators of cardiac autonomic efferents and afferents.
Our second theory, with no data to support it, is that by making a dissection line through the posterior right atrium and into the left atrial posterior wall, interrupting those neurons, we somehow impair ventricular function, and that is more important in single-ventricle than two-ventricle patients.
The third possibility is that by distorting the anatomy, having made the incision through the back wall of both atria, you potentially predispose those patients to increased AV valve regurgitation, which again would have a higher impact on patients with a single ventricle than two ventricles.
Although we do not have data to support these theories, by following these in larger cohorts we may be able to answer these hypotheses definitively.
To address your second question, a number of people have proposed that volume loading, such as occurs during the early stage palliation of single-ventricle patients, may uncover otherwise subclinically present pulmonary vein obstruction. So you may uncover a narrowing that would have been present if the patient had two-ventricle anatomy, but would not have been either hemodynamically or physiologically important in that patient. Whereas, in the single-ventricle patient, the pulmonary vein narrowing is exacerbated by the volume load, the elevation in pulmonary vascular resistance, secondary to the pulmonary vein obstruction, is poorly tolerated and cardiac output is compromised.
It may also be that there is a stereotyped genomic level response to the stress of pulmonary vein obstruction that is different between single- and two-ventricle patients. Molecular level analysis of this pathology, comparing gene and protein expression between patients with obstructed TAPVC and either single- or two-ventricle anatomy will be helpful in identifying if there are truly different pathologic processes initiated.
Have we abandoned type 2 repair entirely? No. Type 2 technique does provide an excellent exposure to the anterior wall of the confluence and it is still used selectively in Children's Hospital.
The superior approach you alluded to, approaching the confluence medial to the superior vena cava and posterior to the atrial wall, is used in selective cases of supracardiac TAPVC.
DR ANTONIO CORNO (Lausanne, Switzerland): Two questions. You showed a trend towards better result with nitric oxide versus without nitric oxide in patients with pulmonary venous obstruction. And you also had the survival in two children with pulmonary venous obstruction thanks to the ECMO. Is this suggesting then there is a substantial component of functional obstruction to the pulmonary venous return instead of simple anatomical obstruction?
And the second question bears on the observation of this retrospective study. What do you suggest now as a best approach for patients with single ventricle, total anomalous pulmonary venous connection and pulmonary venous obstruction?
DR HANCOCK FRIESEN: Thank you for your questions. First of all, the question of functional versus structural pulmonary vein obstruction. It is clear, from research that was done at Children's before this paper, that there was some net benefit to using nitric oxide, even in the preoperative patients who had structural pulmonary vein obstruction.
Clearly, there is a period of time after repair of obstructed TAPVC when there is a functional element to the reactivity of the pulmonary vascular bed that persists despite relief of the mechanical obstruction, as witnessed by the patients that have pulmonary arterial hypertensive crises responsive to medical interventions. So I think probably you are right, the fact that we can occasionally support a patient with ECMO (although that is a very anecdotal finding given the small proportion of its use in our cohort) and nitric oxide is an argument for the fact that there are patients who have some postoperative functional residual pulmonary vascular resistance issues that we can manage conservatively.
As for suggesting the best approach in a single-ventricle patient, I wish I could answer that question on the basis of our data. Obviously, no one has the full answer to that question. It is clear that patients who have a BT shunt at the time of TAPVC repair fare worse than other single-ventricle patients with repaired TAPVC. Thus if it is possible to stage the BT shunt and repair the TAPVC later, the overall risk of mortality will be lower. However if the patient has insufficient pulmonary blood flow coincident with obstructed TAPVC, both issues will have to be addressed simultaneously, and we have to acknowledge that these patients still have a high risk of operative mortality.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDISCUSSIONAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDISCUSSIONAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. T. Mahle, J. W. Newburger, G. P. Matherne, F. C. Smith, T. R. Hoke, R. Koppel, S. S. Gidding, R. H. Beekman III, S. D. Grosse, and on behalf of the American Heart Association Congen Role of Pulse Oximetry in Examining Newborns for Congenital Heart Disease: A Scientific Statement From the American Heart Association and American Academy of Pediatrics Circulation, August 4, 2009; 120(5): 447 - 458. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. T. Mahle, J. W. Newburger, G. P. Matherne, F. C. Smith, T. R. Hoke, R. Koppel, S. S. Gidding, R. H. Beekman III, S. D. Grosse, and on behalf of the American Heart Association Congen Role of Pulse Oximetry in Examining Newborns for Congenital Heart Disease: A Scientific Statement from the AHA and AAP Pediatrics, August 1, 2009; 124(2): 823 - 836. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Nakata, Y. Fujimoto, K. Hirose, M. Osaki, Y. Tosaka, Y. Ide, M. Tachi, and K. Sakamoto Functional single ventricle with extracardiac total anomalous pulmonary venous connection Eur. J. Cardiothorac. Surg., July 1, 2009; 36(1): 49 - 56. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. K. Chowdhury, B. Airan, A. Malhotra, A. K. Bisoi, A. Saxena, S. S. Kothari, M. Kalaivani, and P. Venugopal Mixed total anomalous pulmonary venous connection: Anatomic variations, surgical approach, techniques, and results J. Thorac. Cardiovasc. Surg., January 1, 2008; 135(1): 106 - 116. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Hirsch and E. L. Bove Total anomalous pulmonary venous connection MMCTS, May 7, 2007; 2007(0507): 2253. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. K. Chowdhury, K. G. Subramaniam, K. Joshi, S. Varshney, G. Kumar, R. Singh, and P. Venugopal Rechanneling of total anomalous pulmonary venous connection with or without vertical vein ligation: Results and guidelines for candidate selection J. Thorac. Cardiovasc. Surg., May 1, 2007; 133(5): 1286 - 1294. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Karamlou, R. Gurofsky, E. Al Sukhni, J. G. Coles, W. G. Williams, C. A. Caldarone, G. S. Van Arsdell, and B. W. McCrindle Factors Associated With Mortality and Reoperation in 377 Children With Total Anomalous Pulmonary Venous Connection Circulation, March 27, 2007; 115(12): 1591 - 1598. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. A. Latson and L. R. Prieto Congenital and Acquired Pulmonary Vein Stenosis Circulation, January 2, 2007; 115(1): 103 - 108. [Full Text] [PDF]
|
|
|
|
 |
|
 K. Takeuchi, A. Murakami, Y. Hirata, K. Kitahori, Y. Doi, and S. Takamoto Surgical Outcome of Heterotaxy Syndrome in a Single Institution Asian Cardiovasc Thorac Ann, December 1, 2006; 14(6): 489 - 494. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L.S. Morales, B. E. Braud, J. H. Booth, D. E. Graves, J. S. Heinle, E. D. McKenzie, and C. D. Fraser Jr Heterotaxy Patients With Total Anomalous Pulmonary Venous Return: Improving Surgical Results Ann. Thorac. Surg., November 1, 2006; 82(5): 1621 - 1628. [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 |
