| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Pediatric Cardiac Surgery, British Columbia Children's Hospital, Vancouver, BC, Canada
Received 28 December 2007; received in revised form 27 May 2008; accepted 2 June 2008.
* Corresponding author. Address: McGill University Health Center, Cardiac Surgery Department, 687 Pine Avenue West, Montreal, QC, Canada H3A 1A1. Tel.: +1 514 843 1463; fax: +1 514 843 1602. (Email: siamakmohammadi{at}yahoo.com).
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
Key Words: Systemic-pulmonary shunt • Pulmonary atresia • Pulmonary blood flow
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
|
2. Patients and methods |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
The surgical approach was by thoracotomy (group T) in 180 patients (79.6%) and sternotomy (group S, performed since 1997) in 46 patients (20.4%). The systemic side of the shunt insertion was either the innominate artery (n = 38) or the ascending aorta (n = 8) in group S, and the right (n = 151) and left (n = 29) subclavian artery in group T. The site of the distal anastomosis was: (a) the right pulmonary artery (n = 189) in patients with proximal anastomosis from the innominate and or right subclavian artery, (b) the left pulmonary artery (n = 29) in patients with proximal anastomosis from the left subclavian artery, and finally (c) the main pulmonary artery (n = 8) in all patients with proximal anastomosis from the ascending aorta. The median shunt size was 4.0 mm (range 3.0–6.0 mm). The mean indexed shunt size (mm/kg) was 1.16 ± 0.37. Type of surgical approach and shunt diameters were the surgeon's preference (Four surgeons in total during the study). Examined variables included cardiac diagnosis, age at operation, weight at operation, the year of operation (before or after 1996), the presence of antegrade PBF, preoperative urgent intubation for any reason, type of surgical approach, indexed shunt size, surgeon (A, B, C, D), duration of intubation, intensive care unit (ICU) stay, peripheral oxygen saturation (SaO2) at discharge, hospital stay, need for re-operation for shunt-related complications (thrombosis or narrowing), ASA use after discharge, in-hospital and interim mortality (death after hospital discharge and before complete repair operation). Shunt thrombosis or narrowing was defined by: (a) need for an urgent repeated SPS as a result of prohibitive cyanosis; (b) absence of murmur in conjunction with increasing cyanosis; (c) absence and or decrease of flow on echo Doppler, angiography, computerized tomography, or magnetic resonance imaging in conjunction with increasing cyanosis; (d) surgical or autopsy finding. All inpatient and outpatient medical records were reviewed. In cases in which an autopsy was performed, the report was reviewed, and correlated with clinical information.
2.1 Statistical analysis
Data were collected retrospectively. Summary statistics were expressed using mean ± 1 standard deviation for continuous variable or as percentage for categorical data and median. According to the distribution of continuous data, Student's t-test or Wilcoxon's rank-sum test was used to compare groups (T and S). The analysis of categorical variables was performed using Fisher's exact test. Multivariate analysis was performed using stepwise logistic regression including only factors identified to be significant during univariate analysis (p
< 0.15) to assess the independent risk factors for in-hospital and interim mortality. The data were analyzed using the statistical package program SAS 9.1 (SAS Institute Inc., Cary, NC). A p-value less than 0.05 was considered significant.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
|
60% of cases) to the choice of a sternotomy approach were functional single ventricle anatomy-PS/PA, and surgeon ‘C’. The main factors (60% of cases) leading to choice of a thoracotomy approach were PS/PA with VSD/tetralogy of Fallot, and surgeon ‘A, B’. Retrospective analysis also showed that factors leading mainly (60% of cases) to the choice of small shunt size (
3.5 mm) were neonates, low body weight (3.5 kg), PA-IS pathology, functional single ventricle anatomy-PS/PA, and operations undertaken more recently (between 1998 and 2005). The factors leading to choice of large shunt size (>3.5 mm) levels were higher body weight (>3.5 kg), PS/PA with VSD/tetralogy of Fallot, surgeon ‘B’, and operations undertaken during earlier years (between 1988 and 1997).
3.1 In-hospital mortality
There were 13 in-hospital deaths (5.7%), which was significantly higher (p
= 0.03) in group S patients (n
= 7, 15.2%) compared to group T patients (n
= 6, 3.3%). The reasons of death are depicted in Table 2
. Univariate analysis revealed younger age (p
= 0.01), lower body weight (p
= 0.04), a diagnosis of PA with intact septum (p
= 0.005), preoperative intubation (p
=
0.03), length of intubation (p
< 0.0001), longer ICU stay (p
< 0.0001), and group S (p
= 0.03) as risk factors for in-hospital death. In the multivariate analysis, younger age (p
=
0.04), PA with intact septum (p
= 0.01), and length of intubation (p
= 0.01) were independent risk factors of in-hospital mortality.
|
3.5 mm) shunt size (4 patients with 3 mm and 6 patients with 3.5 mm shunt size), and 5 patients had large (>3.5 mm) shunt size (2 patients with 4 mm, 2 patients 5 mm, and 1 patients with 6 mm shunt size). The nature of death was sudden in the majority of our patients (n
= 12, 80%), 9 out of which (75%) had small shunt size (3.5 mm). Two patients had admission for severe pulmonary infection; both were in hospital at the time of death. One premature neonate died from sepsis. Among them univariate analysis revealed younger age (p
=
0.03), a diagnosis of PA with intact septum (p
=
0.04), group T (p
=
0.03), and small (3.5 mm) shunt size (p
=
0.04) as risk factors for death prior to second-stage or complete repair surgery. In the multivariate analysis, younger age (p
= 0.03), and group T (p
=
0.03) were independent risk factors of interim mortality. ASA had been started in 128 survivors from hospital (60.1%), which was not significantly different in the 2 groups (p
= 0.4). Absence of antiplatelet therapy was not a risk factor for interim mortality. Autopsy has been performed in nine patients in whom the diagnosis of shunt thrombosis was confirmed.
3.3 Early and interim shunt-related complications
There were 19 re-operations in the early (n
= 12; 8 in group T, and 4 in group S) and interim (n
= 7; all in group T) period. There were 12 episodes (5.3%) of early shunt-related complications, with 8 (4.4%) episodes in group T and 4 (8.7%) episodes in group S (p
= 0.3). Four out of 12 patients (33.3%) with early shunt-related complications (thrombosis) died during hospitalization, 2 patients in each group (p
= 0.2). All four patients had undergone surgical revision of the shunt. All other early shunt-related complications except one (six episodes in group T and one episode in group S) were thrombosis or stenosis that needed shunt revision and/or replacement. There was one patient of group S who underwent reduction of the shunt diameter with the use of a metallic clip on day 3 postoperation for excessive PBF and cardiac failure. Univariate analysis revealed younger age (p
= 0.03) and lower body weight (p
= 0.04) as risk factors for early shunt-related complications. Surgical approach was not a risk factor for early shunt thrombosis or stenosis (p
=
0.3).
In this interim period among the survivors discharged from the hospital, there were seven shunt re-operations. None of them were in group S. Five of those patients had another SPS on the opposite side in order to delay the complete repair between 1988 and 1993, and two patients had SPS for early shunt stenosis.
The interval between the first SPS and complete repair were 19.9 ± 9.6 months and 22.3 ± 17.3 months for group S and T, respectively (p = 0.04). The pulmonary artery required repair at the site of the SPS shunt in 27 patients (11.9%). There was a trend towards having more PA distortion needing repair in patients with thoracotomy (n = 23, 12.7%), compared to patients with sternotomy (n = 4, 8.7%) SPS (p = 0.12) during complete repair.
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
4.1 Early outcomes
In the present study, univariate analysis showed that general patient characteristics (younger age, and lower body weight), original pathologic feature (in particular patients with a PA-intact septum), surgical approach (sternotomy with SPS originating from the innominate artery and/or ascending aorta), and finally complicated postoperative course with longer intubation periods and ICU stay have an impact on in-hospital death. However only the PA-intact septum, longer intubation period, and ICU stay were independent risk factors for in-hospital mortality.
Some investigators [3,8] have studied the negative impact of younger age, and lower body weight in the early outcome of palliation with SPS. The higher mortality rate in these younger patients with lower body weight could be related to the small size of the pulmonary arteries. Weight at the time of the procedure has been identified by all of the larger studies as an independent risk factor [9,10]. However, a number of investigators have demonstrated early primary repair of cyanotic congenital defects with good results [11–13]. The staged approach remains a surgical option in young patients in whom definitive repair may not be possible.
It is generally agreed that in patients with a PA and intact septum, with a small right ventricle and/or ventricular-coronary artery fistulae, palliation with SPS may be the first surgical approach. In patients with this rare pathology, there are a variety of anatomic subtypes and considerable morphologic heterogeneity [14,15]. Patients with the severe form of PA and intact septum are at increased risk of death despite the type of treatment. Fenton et al. [16] showed that the interim mortality after palliation with SPS in infants with PA and intact septum and severe right ventricle hypoplasia is 24%. In the present study, PA and intact septum was an independent risk factor for in-hospital mortality in which severe right ventricle hypoplasia was present in 61.3% of patients. However, it was not a risk factor for interim mortality. The latter may be related to the small number of survivors in this subgroup of patients.
The surgical approach and technique might be one of the most important factors to modify for improving the early outcome. In our study of univariate analysis, the sternotomy approach was a risk factor for in-hospital mortality despite the similar median age, the number of neonates, body weight, the mean indexed shunt size, and the number of patients with antegrade pulmonary flow, compared with the thoracotomy group. However, this was not confirmed in the multivariate analysis. Group S patients experienced significantly longer median intubation time and ICU stay than group T patients. Group S patients had a more proximal insertion shunt site than group T patients. The characteristics of the shunt on pulmonary blood flow and the influence of shunt site on negative early outcome risk are complex. The length and diameters of the graft are also determinants of blood flow through the shunt.
The longer median intubation time and ICU stay in patients with sternotomy may be related to the pulmonary over circulation. The use of the innominate artery and/or ascending aorta as the site of shunt insertion renders the length of the graft shorter, and puts the distal anastomosis closer to the PA bifurcation. However, the latter allows even distribution of blood flow to both lungs and potentially better PA growth, but the combination of both may allow for increased pulmonary blood flow. These shunt-related physiologic factors may contribute to in-hospital mortality. The use of smaller shunt size and more distal insertion site in group T patients may result in less pulmonary blood flow. We recommended use of a 3.5 (in cases of small pulmonary arteries) or a 4.0 mm graft for full term infants (up to 4 kg) and a 3.0 mm graft for premature babies weighing less than 3.0 kg, even if the shunt comes from the innominate artery. Because most of these children go on to a complete repair at 6–12 months of age, the shunt flow is adequate and not excessive. The more favorable outcome in patients with smaller shunt size is consistent with previous studies in neonates [3], patients with tricuspid atresia [17], and canine models of univentricular hearts [18].
4.2 Interim outcomes
Interim mortality, defined as death between the first stage of palliation and complete surgical repair, has an important impact on overall outcome of patients with cyanotic congenital disease undergoing SPS. In the present study, univariate analysis showed that general characteristics of the patients (younger age, and lower body weight), smaller shunt size (3.5 mm), and surgical approach (thoracotomy approach) were risk factors for interim mortality, however only the younger age and thoracotomy approach were independent risk factors for in-hospital mortality. These data demonstrated that infants with a SPS would have an increased risk of interim mortality (7%), which is comparable with interim death in patients with hypoplastic left heart syndrome [19,20].
Thoracotomy approach was an independent risk factor for interim death. Determination of the mechanism by which thoracotomy approach portends a poor interim outcome, is beyond the scope of this observational study. However, a possible contributing factor deserves mention. The longer length of the graft and the use of a smaller insertion site of both systemic and pulmonary sites may make SPS more susceptible to thrombosis and death. The overall mortality (in-hospital plus interim mortality) are comparable in both sternotomy and thoracotomy groups. Using a sternotomy approach with a smaller shunt size to avoid pulmonary over circulation may improve the interim mortality by preventing a thoracotomy approach. More than 65% of the interim mortality happened in patients with a thoracotomy approach and a small shunt size (3.5 mm), so thus the use of a bigger shunt size in patients with the thoracotomy approach may also avoid shunt thrombosis and improve interim mortality. Pulmonary over circulation may also be avoided by a smaller diameter of the subclavian artery to some extent.
In nine autopsies performed in patients that died during the interim period, shunt thrombosis was identified in all of them. The nature of death was sudden in the majority of our patients (80%). In another study [7], interim deaths were unexpected in 81% of patients who were doing well. Our study fails to demonstrate any beneficial effect of aspirin to prevent shunt thrombosis, probably due to the insufficient number of patients. Aspirin has been used in approximately 65% of our patients during the interim period. The use of anticoagulation in patients with SPS is a subject of debate. Li et al. [21] in a multicenter study showed infants palliated with SPS to have a poor outcome. In this study, the use of aspirin appeared to be associated with a reduction in shunt thrombosis and death, which was not confirmed in other studies. Investigations into additional mechanisms of antiplatelet and antithrombotic therapies in the interim period are needed. Considerations also must be given to planning the complete repair surgery as early as possible.
4.3 Study limitations
Limitations of our study include its retrospective nature, the wide variety of anatomic subtypes, and the absence of autopsies in some patients. The data do not allow us to extract more specific, concrete, or more beneficial measures or strategies that might favorably impact survival in this challenging group of patients.
4.4 Conclusion
Our data demonstrate that patients with SPS have a significant mortality rate before undergoing complete repair. Patients with younger age, lower body weight and pulmonary atresia and intact septum diagnosis are more susceptible to in-hospital death. Patients with a thoracotomy approach have higher interim mortality. Deploying smaller shunt and/or changing systemic to pulmonary shunt insertion origin through sternotomy may improve in-hospital and interim outcome. The latter technique may also avoid future pulmonary arteries distortion.
|
Appendix A |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
Dr M.A. Navabi (Tehran, Iran): I think in the sternotomy group it's very important to clarify the origin of the aortic site, because if you do a sternotomy and do a right Gore-Tex shunt, you may put the proximal end on the ascending aorta or you may put it on the brachiocephalic, and that determines the flow of the shunt. So in your sternotomy group you probably have to clarify the site of aortic anastomosis.
Dr Mohammadi: In the majority of patients with sternotomy, the insertion site was the innominate artery, however, in eight patients it was the ascending aorta. In our series, these patients experienced a longer postoperative intubation period and ICU stay, but whether it was really related to the pulmonary overcirculation cannot be explained by the present study. I, personally, believe systemic to pulmonary shunts with a more proximal insertion site have more pulmonary blood flow. However, the flow through the modified BT shunt is a complex issue, and for a given diameter of prosthesis, it is also dependent on the length of graft and distal bed, which is different in the different types of surgery, and pathologies.
Dr W. Daenen (Linden, Belgium): Based on your results now, what size would you recommend for a 4 kg patient treated with a systemic shunt from the brachiocephalic artery to the pulmonary artery?
Dr Mohammadi: I think the best recommendations are to perform the procedure through a median sternotomy, to use a smaller shunt size that corresponds to the weight of the patient, and to perform the graft as distal as possible in the systemic arterial tree towards the subclavian artery. The shunt size is one of the factors which plays a role in pulmonary blood flow.
Dr C. Sebening (Heidelberg, Germany): Did you heparinize or anticoagulate in the mid or interim phase? Did you anticoagulate your shunt?
Dr Mohammadi: About 60% of our patients received aspirin, which was equally distributed in both groups, and the absence of using aspirin was not a risk factor for shunt-related complications and death. However, the number of patients is not enough to make a statement in this regard.
Dr Sebening: If you say 65%, what were the criteria for the others not to receive it?
Dr Mohammadi: There were no strict criteria due to the very long duration of the study. During this period, which started in 1988 and ended in 2005, the practice evolved in terms of the use of anticoagulation and the use of aspirin in the interim phase.
|
Footnotes |
|---|
Presented at the 21st Annual Meeting of the European Association for Cardio-thoracic Surgery, Geneva, Switzerland, September 16–19, 2007. |
References |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionAppendix AReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
