- Email: [email protected]
Congenital heart defect with associated malformations in children
Journal of Pediatric Surgery (2005) 40, 1675 – 1680
www.elsevier.com/locate/jpedsurg
Congenital heart defect with associated malformations in children Michall Wojtalika, Wojciech Mro´wczyn´skia,*, Jacek Henschkea, Krzysztof Wroneckie, Aldona Siwin´skac, Maciej Piaszczyn´skia, Malgorzata Pawelec-Wojtalikd, Bartllomiej Mrozin´skic, Mallgorzata Bruskab, Michall Bllaszczyn´skib, Rafall Surmaczd a
Department of Paediatric Cardiac Surgery, Karol Marcinkowski University of Medical Sciences, Poznan 60-572, Poland Department of Paediatric Surgery, Karol Marcinkowski University of Medical Sciences, Poznan 60-572, Poland c Department of Paediatric Cardiology, Karol Marcinkowski University of Medical Sciences, Poznan 60-572, Poland d Department of Paediatric Radiology, Karol Marcinkowski University of Medical Sciences, Poznan 60-572, Poland e Department of Paediatric Cardiac Surgery, bSilesian Cardio-Surgical Centre,Q Wrocl aw 51-124, Poland b
Index words: Children; Complex congenital defects; Congenital heart defects; Multiorgan malformations
Abstract Background: Children with multisystem involvement including congenital heart defect (CHD) are a very salient problem. The purpose of this study was to evaluate the incidence of CHD associated with malformations of other systems and to assess the modalities of treatment and perioperative mortality among patients referred to the department of pediatric cardiac surgery. Methods: The medical records of 1856 children were reviewed retrospectively from 1997 to 2002 to establish CHD and types of associated malformations. The connections between CHD and other lesions were investigated. Furthermore, the influence of patient and perioperative variables on mortality risk was scrutinized. Univariate and multivariate analyses were used. Results: Eighty-four children (4.53%) had CHD and associated malformations. The malformations of digestive (35.7%), urinary (22.4%), and nervous (14.3%) systems were the most frequently observed associated defects. No relation was found between CHD and concomitant lesions. The results of multivariate logistic regression showed significant influence of patient age, primary cardiac procedure, and CHD type on mortality (ca 19%) in children with multiorgan lesions. Conclusions: The treatment of children with CHD and associated multiple lesions is connected with higher mortality risk. The following factors: younger age, urgency of surgical procedure, and primary surgical procedure had negative impact on patient’s outcome. However, these risks in certain cases are inevitable. The cardiac procedure preceding the surgical operation may improve the overall effect of treatment because of circulatory stabilization, provided that the condition of the patient does not preclude any intervention at all. D 2005 Published by Elsevier Inc.
This paper was presented at the 11th Congress of the Polish Association of Paediatric Surgeons, September 17 to 20, 2003, Wroclaw, Poland. T Corresponding author. Tel.: +48 61 8491 277, +48 61 8491 479; fax: +48 61 8669 130. E-mail address: [email protected] (W. Mro´wczyn´ski). 0022-3468/$ – see front matter D 2005 Published by Elsevier Inc. doi:10.1016/j.jpedsurg.2005.06.004
The incidence of congenital heart defects (CHDs) varies from 6 to 8 per 1000 liveborn neonates [1]. The frequency of CHD associated with congenital lesions affecting other systems is much lower, and yet, the incidence of certain
1676
M. Wojtalik et al.
associations is increasing [2]. Because of unsatisfactory outcomes, patients with congenital defects compromising several systems still remain an issue that is worth being considered. The etiology of these complex lesions is not clearly elucidated. Some of them are described as various associations such as CHARGE, VACTERL, and Cantrell pentalogy [3]. Others are combined with chromosomal aberrations, for example, Down’s and Edwards syndromes. Continuous progress in neonatology and perinatology, and appropriate prenatal and postnatal diagnosis can change significantly the natural history of CHD associated with other malformations. A multiteam and multidisciplinary approach may enable an early effective surgical palliation or correction in most patients [3].
1. Aim of the study The purpose of this study was to evaluate the incidence of CHD associated with other malformations, as well as to assess the modalities of treatment and perioperative mortality in patients referred to the department of pediatric cardiac surgery.
2. Patients and methods This retrospective observational study was carried out in the department of cardiac surgery of the tertiary pediatric center. From October 1997 till December 31, 2002, 1856 consecutive patients were admitted to our department for surgical treatment of CHD. Their medical records were reviewed to establish CHD, associated malformation type, the number of associated malformations, the presence of chromosomal aberrations (trisomy 21, Turner syndrome), and patient’s age (patient variables). Perioperative variables were the following: whether patient underwent surgical procedure, cardiac procedure, or both; if the surgical procedure was urgent; and whether surgical procedure was performed as a primary operation. A 30-day-perioperative mortality was the main outcome variable. Patients were divided according to their age (neonates, infants, children N1 year old) and according to the CHD type (Table 1). Table 1 VSD ASD PDA ToF TGA AS/AI SV CoA Other
Congenital heart defects Ventricular septal defect, including atrioventricular septal defect Atrial septal defect Patent ductus arteriosus Tetralogy of Fallot Transposition of great arteries Aortic valve stenosis and/or insufficiency Various forms of single ventricle including hypoplastic left heart syndrome Coarctation of aorta Other less frequent defects
Fig. 1 Incidence of associated defect in specific CHD groups. ASD indicates atrial septal defect; SV, single ventricle; AS/AI, aortic valve stenosis and/or insufficiency; TGA, transposition of great arteries.
Diagnosis of CHD was established with the use of echocardiography. In uncertain cases, angiography was applied. Presence of accompanying malformations was detected by physical examination, cranial/abdominal ultrasound, x-rays, or bronchoscopy. Selected patients required additional x-ray contrast studies. Sometimes the diagnosis of associated defects was made during the hospitalization prior or after the cardiac surgery. Admission to the hospital was elective or influenced by compromised hemodynamic condition, which was not Table 2
Associated defects (n = 84)
System
Defect
Quantity
Alimentary (n = 35, 35.7%)
Palatoschisis Other Anal atresia Esophageal atresia Duodenal atresia Diaphragmatic hernia Kidney lesions Other Ureteral lesions Urethral lesions Myelomeningocele Other
13 6 5 4 4 3 8 6 5 3 10 4
37.1 17.1 14.3 11.4 11.4 8.6 36.4 27.3 22.7 13.6 71.4 28.6
Scoliosis Other
7 4
63.6 36.4
Hernias
9
100
Other Lung hypoplasia Choanal atresia Presence of lesion
2 1 1 3
50 25 25 100
Urinary (n = 22, 22.4%) Nervous (n = 14, 14.3%) Skeletal (n = 11, 11.2%) Surface (n = 9, 9.2%) Respiratory (n = 4, 4.1) Celosomies (n=3, 3.1%)
% (system)
Congenital heart defects with malformations
Fig. 2 Relation between CHD and associated lesions of different systems.
rare in the case of patients with associated malformations, particularly in the neonatal or infantile period. These children were referred to the intensive care unit in the department of cardiac surgery. When patient’s condition became stable, the decisions concerning the surgery were made. Patients who underwent only a cardiac procedure were scheduled for further surgical treatment in other departments. The incidence of multiorgan malformations was calculated as a percentage of all children, in each specific CHD group, and in respect to all patients with associated lesions. The incidence of CHD, the presence of additional defect, and its relation to other variables were depicted in categorized histograms. The statistical significance in the case of 2-group comparisons was calculated with the use of v 2 and v 2 with Yates correction tests. Fisher-FreemanHalton test was applied for multigroup comparisons. The association of CHD with the lesion of other systems was investigated with the use of univariate analyses and multivariate log-linear analysis. To investigate the influence of the patient and perioperative variables on mortality, both the univariate analyses and
Fig. 3 Modality of treatment–percentages of patients. no op indicates patients not operated at all; cardio op, patients after cardiac procedure; chir op, patient after surgical procedure; both op, patients after staged surgical–cardio-surgical treatment.
1677
Fig. 4 Urgent surgical intervention in associated lesion groups. N indicates central nervous system; A, alimentary system; C, celosomies; R, respiratory system; Sk, skeletal system; U, urinary system; S, surface.
the logistic regression analysis were performed. Variables that significantly influenced mortality entered the model of logistic regression and were selected again by backward stepwise elimination to obtain the best model. Similarly, the impact of CHD type and presence of associated lesion on the mortality in the whole cohort of patients were examined. Univariate analyses and a second model of logistic regression with 2 independent variables were applied. The influence of predictive variables on death risk in these 2 models was expressed in odds ratio (OR) with 95% confidence interval (CI). A 2-tailed P value less than .05 was considered significant in all tests.
3. Results The most frequent CHDs in the investigated population were ventricular septal defect (VSD, 20.5%), atrial septal defect (16.7%), and patent ductus arteriosus (PDA, 13.5%). From the entire cohort of patients, 84 of them (4.53%) had associated malformations. Additional defects were
Fig. 5
Mortality according to treatment modality.
1678
M. Wojtalik et al. Table 4 Multivariate logistic regression: mortality in patients with multiorgan malformations (3-variable model, P b .0001).
Fig. 6
Mortality according to CHD type.
not distributed uniformly (Fisher-Freeman-Halton test, P = .0069) throughout all CHD groups and occurred most frequently in VSD group (v 2 test, P = .0012) (Fig. 1). Among the patients with multiple system involvement, there were 20 (23.8%) neonates, 35 (41.7%) infants, and 29 (34.5%) children who were more than 1 year old. Type of associated lesion and its incidence are presented in Table 2. The malformations of digestive (35.7%), urinary (22.4%), and nervous (14.3%) systems were the most frequently present associated defects. Among 71 patients (84.5%), only 1 additional defect was detected; among 12 (14.3%), 2 systems were compromised, and in only 1 patient with PDA (1.2%) lesions of 3 systems were observed. Clusters of various system lesions were Table 3 Influence of patient and perioperative variables on mortality Variable Patient Sex Age group Down’s syndrome Turner syndrome Performed number of associated lesions Specific system involvement Nervous Urinary Respiratory Alimentary Surface Celosomies Skeletal Perioperative Primary surgical procedure Primary cardiac procedure Performed cardiac procedure Performed surgical procedure Urgent surgical procedure NS indicates nonsignificant.
P (v 2) NS b.0001 NS NS NS NS NS NS NS NS NS NS
.032 .00012 b.0001 NS .011
Predictive variables
Age group
Presence of primary cardiac procedure
CHD type
P OR 95% CI +95% CI
.0052 0.049 0.0061 0.4
.0026 0.025 0.0023 0.26
.02 2.13 1.11 4.09
observed in 7 different combinations with the most frequent alimentary-urinary association (Fig. 4). The relation between specific CHD and the lesion of other system is shown in Fig. 2. No significant correlation was found (univariate analyses and log-linear analysis). The modality of treatment is shown in the Fig. 3. The necessity of urgent surgical intervention previously or at the moment of referral for CHD correction was noted among 30 patients (35.7%). They required urgent surgical treatment because of myelomeningocele, alimentary tract atresia, lobar emphysema, diaphragmatic hernia, various forms of celosomies, and choanal atresia. The urgency of the surgical procedure was noted mainly in the neonatal group (11 patients, 65% of the group) and with frequencies of 31% and 20% in infants and older children, respectively. Groups of patients with associated single lesions or with clusters of lesions differed significantly in the number of children requiring an urgent surgical treatment (FisherFreeman-Halton test, P = .0059). Patients with alimentary tract malformations had significantly higher number of urgent surgical interventions (v 2, P = .01) (Fig. 4). The total mortality in the cohort was 8.9% (166 patients). The logistic regression analysis did not show any influence of CHD type on mortality. The presence of associated lesions (OR, 2.57; 95% CI, 1.45-4.55; P = .0012) had a negative impact on death risk. Univariate comparisons between children with and without additional defect in specific CHD groups revealed higher mortality rates in patients with coarctation of aorta (CoA; v 2, P = .0029) and bother Q cardiac malformations (v 2, P = .0054). Significantly lower mortality was observed in the PDA group (v 2 with Yates correction, P = .0256). Patients with multiorgan malformations had 19.01% (16 patients) mortality; in the group of neonates, the mortality rate was 50%. None of the patients older than 1 year died. The mortality, according to the way of treatment and associated lesion type, is presented in the Figs. 5 and 6. There was no uniform spectrum of mortality rates in CHD groups (Fisher-Freeman-Halton test, P = .0353). Patients with CoA (v 2, P b .04) and bother CHDQ group (v 2, P b .05) revealed significantly increased mortality. The influence of patient and perioperative variables on mortality (univariate analyses) is presented in Table 3. There was no impact of associated defect type on mortality. The results of logistic regression analysis are shown in Table 4.
Congenital heart defects with malformations
4. Discussion The aim of our study was to investigate the incidence of lesions associated with CHDs requiring cardiac surgery. We have examined only patients with associated congenital malformations that could be qualified to urgent or elective correction. Children with both acquired surgical problems, for example, necrotizing enterocolitis and chromosomal aberrations without multisystem involvement, were excluded from the study. The frequency of different CHD groups in the investigated cohort is similar to other Polish observations [4] with respect to the more common CHD types. The spectrum of CHD in the group of 84 children with associated lesions was comparable to the cohort with exception to VSD. This defect was more frequent among patients with multiple system involvement. Incidence of patients with CHD with associated malformations reaches 4.53% in the present study. Data gathered from the other articles show frequencies from 7.7% [5] to 25% [6]. Incidence of multiple system involvement detected in prenatal screening varies from 6% [7] to 48.6% [8] depending on world region. Frequency of 7.2 % of CHD accompanied by additional lesions was revealed during the autopsy studies [9]. The most frequent noncardiac malformations in the investigated cohort of patients were lesions of digestive, urinary, and nervous systems, which correspond with other studies [6]. This finding is also reflected in autopsy material [9]. However, there is a difference in the frequency of urinary tract anomalies that represent 22.4% of extracardiac defects in the current study, whereas a 12% incidence was observed by other groups [10,11]. Associations such as CHARGE, VATER, or VACTERL were not identified in this 5 year study; there was only 1 patient with PDA accompanied by alimentary, respiratory, and nervous system lesions. Univariate and multivariate analyses (log-linear analysis) failed to detect any statistically significant relation between CHD type and associated malformations. This is probably because of a variety of multiple lesions in a small number of patients. Neither do other researchers find any relation [11,12]. Some statistically important correlations between cardiac and extracardiac malformations in the cohort of patients undergoing cardiac catheterization were revealed [13]. In the current study, patients with VSD constituted the only group with significantly increased number of different concomitant lesions. However, neither no statistically significant prevalence of single associated lesion nor cluster of malformations was found in VSD group (univariate analyses and log-linear analysis). The data in literature point to various relations of CHD with other defects, for example, preponderance of CHARGE to right-sided heart lesions and conotruncal defects as well as to atrioventricular septal defect connected with tetralogy of Fallot (ToF) [3]. The latter one is likely to be associated with esophageal atresia
1679 [14], tracheoesophageal fistula [3,14], frontonasal dysplasia [15], and choanal atresia [16]. There were only 2 children with cleft palate associated with ToF in the investigated cohort. Furthermore, 7 patients manifested urinary system abnormalities in the presence of VSD [3]. It was observed that children with associated extracardiac anomalies tend to have a significantly higher incidence of renal tract anomalies [10]. The treatment of patient with multiple system involvement is in most cases a complex procedure and depends on patient’s current condition and type of accompanying lesions. To avoid additional anesthesia, the preoperative planning of staged surgical interventions [17] is proposed as well as multiteam coordination of procedures [3]. A 2-staged management was proposed for Cantrell malformation associated with ToF. Congenital heart defect is operated in second turn [18]. The 2-stage approach was applied to our patients as well. However, only ca 30% of children received full treatment in the period preceding death or the discharge from the department of cardiac surgery. In most cases, a surgical procedure was performed primarily. The deferral of cardiac procedure was possible owing to prostaglandin administration (in the case of ductal dependency) or preoperative hemodynamic stability. There was also a group of patients operated only by surgical team. However, none of them survived until the stage of CHD palliation or correction. This is the result of severe condition of this group. During the treatment in our department, only 57% of patients underwent a cardiac procedure. In most cases, these children did not require urgent surgical operation, which could be postponed. Because of severe condition, 4 children were not qualified neither to corrective nor to palliative surgical procedure; nevertheless, the medical treatment was applied. Discontinuation of treatment of severely compromised patients is another alternative solution described in literature [3]. This point raises the following questions: In which situations should one stop the treatment? Who is to decide? Finally, what are the limits of resistance to multiple surgical trauma? The results of treatment of CHD with associated lesions are worse in comparison to patients with cardiac malformations only; it has been proven in this study. Multiple system involvement is described as a risk factor for CHD treatment [3,19]. Association of esophageal atresia and tracheoesophageal fistula with ToF, presence of VACTERL, or choanal atresia usually contributes to a poor outcome [2,3,14]. In addition, the presence of diaphragmatic hernia [20] increases perioperatvie mortality in children operated for CHD. No relation of associated lesion with increased mortality was found. Although mortality in multiorgan malformation is high, some authors find that only one fifth of the patients die because of cardiac problems, which has been proven by observations of patients with CHARGE [3]. As far as case
1680 of associated esophageal atresia is concerned, only 6% of patients die from cardiac defects [2]. This may correspond with our findings; there were patients operated in the first turn by surgical team who did not reach the second stage and died. Moreover, the mortality among patients operated only for CHD was lower in comparison with the mortality of 2-stage management. On the other hand, increased mortality among formerly mentioned groups can be the result of suboptimal condition created by CHD presence before its correction. The correction of cardiac lesion had a positive influence on mortality in our logistic regression model. Age factor had a similar impact; older patients had greater chances of survival. We should bear in mind that there are patients who cannot wait because of their severe condition. Children with CoA and other defects had increased death risk, which is converse to that of the PDA group in which no cases of death were observed. The presence of CoA could be an aggravating factor, especially among patients with late diagnosis or with deferral of cardiac correction. The group of other defects consists of rare and complex heart malformations whose correction poses apparent risk. The positive impact of PDA can be explained by the possibility of rapid stabilization of patient hemodynamics. This is accomplished by simple ligation procedure and may facilitate further management. The positive effect of patient’s age was observed; only younger patients usually required urgent surgical correction that could jeopardize the outcome. No influence of chromosomal aberrations on mortality was observed. Down’s syndrome with atrioventricular septal defect and associated lesions was diagnosed among only 5 patients. Management of children with CHD with associated multiple lesions is connected with higher mortality risk. The following factors have negative impact on treatment’s outcome: younger age, urgency of surgical procedure, and primary surgical procedure. However, these risk factors are sometimes inevitable because of the patient’s poor condition. The cardiac procedure preceding the surgical operation may improve the overall effect of treatment because of circulatory stabilization, provided that the condition of the patient does not preclude any intervention at all. Further studies are needed to assess long-term outcome in these severely compromised patients.
M. Wojtalik et al.
References [1] Friedman WF. Congenital heart disease in infancy and childhood in heart diseases. In: Braunwald E, editor. A textbook of cardiovascular medicine. Philadelphia7 WB Saunders Company; 1997. [2] Driver CP, Shankat KR, Jones MO, et al. Phenotypic presentations and outcome of esophageal atresia in the era of the Spitz classification. J Pediatr Surg 2001;36:1419 - 21. [3] Wyse RKH, Al-Mahdawi S, Burn J, et al. Congenital heart disease in CHARGE association. Pediatr Cardiol 1993;14:75 - 81. [4] Annual report of Polish cardiac surgeons’ club. Warszawa. 2002. [5] Kramer HH, Majewski F, Trampisch HJ, et al. Malformation patterns in children with congenital heart disease. Am J Dis Child 1987;141: 789 - 95. [6] Stoll C, Alembik Y, Roth MP, et al. Risk factors in congenital heart disease. Eur J Epidemiol 1989;5:382 - 91. [7] Grech V, Gatt M. Syndromes and malformations associated with congenital heart disease in a population-based study. Int J Cardiol 1999;68:151 - 6. [8] Stoll C, Garne E, Clementi M. Evaluation of prenatal diagnosis of associated congenital heart diseases by fetal ultrasonographic examination in Europe. Prenat Diagn 2001;21:243 - 52. [9] Reinhold-Richter L, Fischer A, Schneider-Obermeyer J. Congenital heart defects. Frequency at autopsy. Zentralbl Allg Pathol 1987;133: 253 - 61. [10] Murugasu B, Yip WC, Tay JS, et al. Sonographic screening for renal tract anomalies associated with congenital heart disease. J Clin Ultrasound 1990;18:79 - 83. [11] Derchi L, Giusti G, Reginato E, et al. The incidence of urinary tract malformations in congenital heart disease. Thorac Cardiovasc Surg 1984;32:78 - 80. [12] Pradat P. Noncardiac malformations at major congenital heart defects. Pediatr Cardiol 1997;18:11 - 8. [13] Pongiglione G, Marasini M, Ribaldone D, et al. Extracardiac malformations and congenital heart disease: frequency and patterns of associations. G Ital Cardiol 1987;17:419 - 25. [14] Saydam TC, Mychaliska CB, Harrison MR. Esophageal lung with multiple congenital anomalies: conundrums in diagnosis and management. J Pediatr Surg 1999;34:615 - 8. [15] De Moor MMA, Baruch R, Human DG. Frontonasal dysplasia associated with tetralogy of Fallot. J Med Genet 1987;24:107 - 9. [16] Hall BD. Choanal atresia and associated multiple anomalies. J Pediatr 1979;95:395. [17] De Ugarte DA, Boechat MI, Shaw WW, et al. Parasitic ophalopagus complicated by omphalocele and congenital heart disease. J Pediatr Surg 2002;37:1357 - 8. [18] Abdallach HI, Marks LA, Balsara RK, et al. Staged repair of pentalogy of Cantrell with tetralogy of Fallot. Ann Thorac Surg 1993;56:979 - 80. [19] Roodpeyma S, Kamali Z, Afshar F, et al. Risk factors in congenital heart disease. Clin Pediatr (Phila) 2002;41:653 - 8. [20] Cunniff C, Jones KL, Jones MC. Patterns of malformation in children with congenital diaphragmatic defects. J Pediatr 1990;116:258 - 61.
www.elsevier.com/locate/jpedsurg
Congenital heart defect with associated malformations in children Michall Wojtalika, Wojciech Mro´wczyn´skia,*, Jacek Henschkea, Krzysztof Wroneckie, Aldona Siwin´skac, Maciej Piaszczyn´skia, Malgorzata Pawelec-Wojtalikd, Bartllomiej Mrozin´skic, Mallgorzata Bruskab, Michall Bllaszczyn´skib, Rafall Surmaczd a
Department of Paediatric Cardiac Surgery, Karol Marcinkowski University of Medical Sciences, Poznan 60-572, Poland Department of Paediatric Surgery, Karol Marcinkowski University of Medical Sciences, Poznan 60-572, Poland c Department of Paediatric Cardiology, Karol Marcinkowski University of Medical Sciences, Poznan 60-572, Poland d Department of Paediatric Radiology, Karol Marcinkowski University of Medical Sciences, Poznan 60-572, Poland e Department of Paediatric Cardiac Surgery, bSilesian Cardio-Surgical Centre,Q Wrocl aw 51-124, Poland b
Index words: Children; Complex congenital defects; Congenital heart defects; Multiorgan malformations
Abstract Background: Children with multisystem involvement including congenital heart defect (CHD) are a very salient problem. The purpose of this study was to evaluate the incidence of CHD associated with malformations of other systems and to assess the modalities of treatment and perioperative mortality among patients referred to the department of pediatric cardiac surgery. Methods: The medical records of 1856 children were reviewed retrospectively from 1997 to 2002 to establish CHD and types of associated malformations. The connections between CHD and other lesions were investigated. Furthermore, the influence of patient and perioperative variables on mortality risk was scrutinized. Univariate and multivariate analyses were used. Results: Eighty-four children (4.53%) had CHD and associated malformations. The malformations of digestive (35.7%), urinary (22.4%), and nervous (14.3%) systems were the most frequently observed associated defects. No relation was found between CHD and concomitant lesions. The results of multivariate logistic regression showed significant influence of patient age, primary cardiac procedure, and CHD type on mortality (ca 19%) in children with multiorgan lesions. Conclusions: The treatment of children with CHD and associated multiple lesions is connected with higher mortality risk. The following factors: younger age, urgency of surgical procedure, and primary surgical procedure had negative impact on patient’s outcome. However, these risks in certain cases are inevitable. The cardiac procedure preceding the surgical operation may improve the overall effect of treatment because of circulatory stabilization, provided that the condition of the patient does not preclude any intervention at all. D 2005 Published by Elsevier Inc.
This paper was presented at the 11th Congress of the Polish Association of Paediatric Surgeons, September 17 to 20, 2003, Wroclaw, Poland. T Corresponding author. Tel.: +48 61 8491 277, +48 61 8491 479; fax: +48 61 8669 130. E-mail address: [email protected] (W. Mro´wczyn´ski). 0022-3468/$ – see front matter D 2005 Published by Elsevier Inc. doi:10.1016/j.jpedsurg.2005.06.004
The incidence of congenital heart defects (CHDs) varies from 6 to 8 per 1000 liveborn neonates [1]. The frequency of CHD associated with congenital lesions affecting other systems is much lower, and yet, the incidence of certain
1676
M. Wojtalik et al.
associations is increasing [2]. Because of unsatisfactory outcomes, patients with congenital defects compromising several systems still remain an issue that is worth being considered. The etiology of these complex lesions is not clearly elucidated. Some of them are described as various associations such as CHARGE, VACTERL, and Cantrell pentalogy [3]. Others are combined with chromosomal aberrations, for example, Down’s and Edwards syndromes. Continuous progress in neonatology and perinatology, and appropriate prenatal and postnatal diagnosis can change significantly the natural history of CHD associated with other malformations. A multiteam and multidisciplinary approach may enable an early effective surgical palliation or correction in most patients [3].
1. Aim of the study The purpose of this study was to evaluate the incidence of CHD associated with other malformations, as well as to assess the modalities of treatment and perioperative mortality in patients referred to the department of pediatric cardiac surgery.
2. Patients and methods This retrospective observational study was carried out in the department of cardiac surgery of the tertiary pediatric center. From October 1997 till December 31, 2002, 1856 consecutive patients were admitted to our department for surgical treatment of CHD. Their medical records were reviewed to establish CHD, associated malformation type, the number of associated malformations, the presence of chromosomal aberrations (trisomy 21, Turner syndrome), and patient’s age (patient variables). Perioperative variables were the following: whether patient underwent surgical procedure, cardiac procedure, or both; if the surgical procedure was urgent; and whether surgical procedure was performed as a primary operation. A 30-day-perioperative mortality was the main outcome variable. Patients were divided according to their age (neonates, infants, children N1 year old) and according to the CHD type (Table 1). Table 1 VSD ASD PDA ToF TGA AS/AI SV CoA Other
Congenital heart defects Ventricular septal defect, including atrioventricular septal defect Atrial septal defect Patent ductus arteriosus Tetralogy of Fallot Transposition of great arteries Aortic valve stenosis and/or insufficiency Various forms of single ventricle including hypoplastic left heart syndrome Coarctation of aorta Other less frequent defects
Fig. 1 Incidence of associated defect in specific CHD groups. ASD indicates atrial septal defect; SV, single ventricle; AS/AI, aortic valve stenosis and/or insufficiency; TGA, transposition of great arteries.
Diagnosis of CHD was established with the use of echocardiography. In uncertain cases, angiography was applied. Presence of accompanying malformations was detected by physical examination, cranial/abdominal ultrasound, x-rays, or bronchoscopy. Selected patients required additional x-ray contrast studies. Sometimes the diagnosis of associated defects was made during the hospitalization prior or after the cardiac surgery. Admission to the hospital was elective or influenced by compromised hemodynamic condition, which was not Table 2
Associated defects (n = 84)
System
Defect
Quantity
Alimentary (n = 35, 35.7%)
Palatoschisis Other Anal atresia Esophageal atresia Duodenal atresia Diaphragmatic hernia Kidney lesions Other Ureteral lesions Urethral lesions Myelomeningocele Other
13 6 5 4 4 3 8 6 5 3 10 4
37.1 17.1 14.3 11.4 11.4 8.6 36.4 27.3 22.7 13.6 71.4 28.6
Scoliosis Other
7 4
63.6 36.4
Hernias
9
100
Other Lung hypoplasia Choanal atresia Presence of lesion
2 1 1 3
50 25 25 100
Urinary (n = 22, 22.4%) Nervous (n = 14, 14.3%) Skeletal (n = 11, 11.2%) Surface (n = 9, 9.2%) Respiratory (n = 4, 4.1) Celosomies (n=3, 3.1%)
% (system)
Congenital heart defects with malformations
Fig. 2 Relation between CHD and associated lesions of different systems.
rare in the case of patients with associated malformations, particularly in the neonatal or infantile period. These children were referred to the intensive care unit in the department of cardiac surgery. When patient’s condition became stable, the decisions concerning the surgery were made. Patients who underwent only a cardiac procedure were scheduled for further surgical treatment in other departments. The incidence of multiorgan malformations was calculated as a percentage of all children, in each specific CHD group, and in respect to all patients with associated lesions. The incidence of CHD, the presence of additional defect, and its relation to other variables were depicted in categorized histograms. The statistical significance in the case of 2-group comparisons was calculated with the use of v 2 and v 2 with Yates correction tests. Fisher-FreemanHalton test was applied for multigroup comparisons. The association of CHD with the lesion of other systems was investigated with the use of univariate analyses and multivariate log-linear analysis. To investigate the influence of the patient and perioperative variables on mortality, both the univariate analyses and
Fig. 3 Modality of treatment–percentages of patients. no op indicates patients not operated at all; cardio op, patients after cardiac procedure; chir op, patient after surgical procedure; both op, patients after staged surgical–cardio-surgical treatment.
1677
Fig. 4 Urgent surgical intervention in associated lesion groups. N indicates central nervous system; A, alimentary system; C, celosomies; R, respiratory system; Sk, skeletal system; U, urinary system; S, surface.
the logistic regression analysis were performed. Variables that significantly influenced mortality entered the model of logistic regression and were selected again by backward stepwise elimination to obtain the best model. Similarly, the impact of CHD type and presence of associated lesion on the mortality in the whole cohort of patients were examined. Univariate analyses and a second model of logistic regression with 2 independent variables were applied. The influence of predictive variables on death risk in these 2 models was expressed in odds ratio (OR) with 95% confidence interval (CI). A 2-tailed P value less than .05 was considered significant in all tests.
3. Results The most frequent CHDs in the investigated population were ventricular septal defect (VSD, 20.5%), atrial septal defect (16.7%), and patent ductus arteriosus (PDA, 13.5%). From the entire cohort of patients, 84 of them (4.53%) had associated malformations. Additional defects were
Fig. 5
Mortality according to treatment modality.
1678
M. Wojtalik et al. Table 4 Multivariate logistic regression: mortality in patients with multiorgan malformations (3-variable model, P b .0001).
Fig. 6
Mortality according to CHD type.
not distributed uniformly (Fisher-Freeman-Halton test, P = .0069) throughout all CHD groups and occurred most frequently in VSD group (v 2 test, P = .0012) (Fig. 1). Among the patients with multiple system involvement, there were 20 (23.8%) neonates, 35 (41.7%) infants, and 29 (34.5%) children who were more than 1 year old. Type of associated lesion and its incidence are presented in Table 2. The malformations of digestive (35.7%), urinary (22.4%), and nervous (14.3%) systems were the most frequently present associated defects. Among 71 patients (84.5%), only 1 additional defect was detected; among 12 (14.3%), 2 systems were compromised, and in only 1 patient with PDA (1.2%) lesions of 3 systems were observed. Clusters of various system lesions were Table 3 Influence of patient and perioperative variables on mortality Variable Patient Sex Age group Down’s syndrome Turner syndrome Performed number of associated lesions Specific system involvement Nervous Urinary Respiratory Alimentary Surface Celosomies Skeletal Perioperative Primary surgical procedure Primary cardiac procedure Performed cardiac procedure Performed surgical procedure Urgent surgical procedure NS indicates nonsignificant.
P (v 2) NS b.0001 NS NS NS NS NS NS NS NS NS NS
.032 .00012 b.0001 NS .011
Predictive variables
Age group
Presence of primary cardiac procedure
CHD type
P OR 95% CI +95% CI
.0052 0.049 0.0061 0.4
.0026 0.025 0.0023 0.26
.02 2.13 1.11 4.09
observed in 7 different combinations with the most frequent alimentary-urinary association (Fig. 4). The relation between specific CHD and the lesion of other system is shown in Fig. 2. No significant correlation was found (univariate analyses and log-linear analysis). The modality of treatment is shown in the Fig. 3. The necessity of urgent surgical intervention previously or at the moment of referral for CHD correction was noted among 30 patients (35.7%). They required urgent surgical treatment because of myelomeningocele, alimentary tract atresia, lobar emphysema, diaphragmatic hernia, various forms of celosomies, and choanal atresia. The urgency of the surgical procedure was noted mainly in the neonatal group (11 patients, 65% of the group) and with frequencies of 31% and 20% in infants and older children, respectively. Groups of patients with associated single lesions or with clusters of lesions differed significantly in the number of children requiring an urgent surgical treatment (FisherFreeman-Halton test, P = .0059). Patients with alimentary tract malformations had significantly higher number of urgent surgical interventions (v 2, P = .01) (Fig. 4). The total mortality in the cohort was 8.9% (166 patients). The logistic regression analysis did not show any influence of CHD type on mortality. The presence of associated lesions (OR, 2.57; 95% CI, 1.45-4.55; P = .0012) had a negative impact on death risk. Univariate comparisons between children with and without additional defect in specific CHD groups revealed higher mortality rates in patients with coarctation of aorta (CoA; v 2, P = .0029) and bother Q cardiac malformations (v 2, P = .0054). Significantly lower mortality was observed in the PDA group (v 2 with Yates correction, P = .0256). Patients with multiorgan malformations had 19.01% (16 patients) mortality; in the group of neonates, the mortality rate was 50%. None of the patients older than 1 year died. The mortality, according to the way of treatment and associated lesion type, is presented in the Figs. 5 and 6. There was no uniform spectrum of mortality rates in CHD groups (Fisher-Freeman-Halton test, P = .0353). Patients with CoA (v 2, P b .04) and bother CHDQ group (v 2, P b .05) revealed significantly increased mortality. The influence of patient and perioperative variables on mortality (univariate analyses) is presented in Table 3. There was no impact of associated defect type on mortality. The results of logistic regression analysis are shown in Table 4.
Congenital heart defects with malformations
4. Discussion The aim of our study was to investigate the incidence of lesions associated with CHDs requiring cardiac surgery. We have examined only patients with associated congenital malformations that could be qualified to urgent or elective correction. Children with both acquired surgical problems, for example, necrotizing enterocolitis and chromosomal aberrations without multisystem involvement, were excluded from the study. The frequency of different CHD groups in the investigated cohort is similar to other Polish observations [4] with respect to the more common CHD types. The spectrum of CHD in the group of 84 children with associated lesions was comparable to the cohort with exception to VSD. This defect was more frequent among patients with multiple system involvement. Incidence of patients with CHD with associated malformations reaches 4.53% in the present study. Data gathered from the other articles show frequencies from 7.7% [5] to 25% [6]. Incidence of multiple system involvement detected in prenatal screening varies from 6% [7] to 48.6% [8] depending on world region. Frequency of 7.2 % of CHD accompanied by additional lesions was revealed during the autopsy studies [9]. The most frequent noncardiac malformations in the investigated cohort of patients were lesions of digestive, urinary, and nervous systems, which correspond with other studies [6]. This finding is also reflected in autopsy material [9]. However, there is a difference in the frequency of urinary tract anomalies that represent 22.4% of extracardiac defects in the current study, whereas a 12% incidence was observed by other groups [10,11]. Associations such as CHARGE, VATER, or VACTERL were not identified in this 5 year study; there was only 1 patient with PDA accompanied by alimentary, respiratory, and nervous system lesions. Univariate and multivariate analyses (log-linear analysis) failed to detect any statistically significant relation between CHD type and associated malformations. This is probably because of a variety of multiple lesions in a small number of patients. Neither do other researchers find any relation [11,12]. Some statistically important correlations between cardiac and extracardiac malformations in the cohort of patients undergoing cardiac catheterization were revealed [13]. In the current study, patients with VSD constituted the only group with significantly increased number of different concomitant lesions. However, neither no statistically significant prevalence of single associated lesion nor cluster of malformations was found in VSD group (univariate analyses and log-linear analysis). The data in literature point to various relations of CHD with other defects, for example, preponderance of CHARGE to right-sided heart lesions and conotruncal defects as well as to atrioventricular septal defect connected with tetralogy of Fallot (ToF) [3]. The latter one is likely to be associated with esophageal atresia
1679 [14], tracheoesophageal fistula [3,14], frontonasal dysplasia [15], and choanal atresia [16]. There were only 2 children with cleft palate associated with ToF in the investigated cohort. Furthermore, 7 patients manifested urinary system abnormalities in the presence of VSD [3]. It was observed that children with associated extracardiac anomalies tend to have a significantly higher incidence of renal tract anomalies [10]. The treatment of patient with multiple system involvement is in most cases a complex procedure and depends on patient’s current condition and type of accompanying lesions. To avoid additional anesthesia, the preoperative planning of staged surgical interventions [17] is proposed as well as multiteam coordination of procedures [3]. A 2-staged management was proposed for Cantrell malformation associated with ToF. Congenital heart defect is operated in second turn [18]. The 2-stage approach was applied to our patients as well. However, only ca 30% of children received full treatment in the period preceding death or the discharge from the department of cardiac surgery. In most cases, a surgical procedure was performed primarily. The deferral of cardiac procedure was possible owing to prostaglandin administration (in the case of ductal dependency) or preoperative hemodynamic stability. There was also a group of patients operated only by surgical team. However, none of them survived until the stage of CHD palliation or correction. This is the result of severe condition of this group. During the treatment in our department, only 57% of patients underwent a cardiac procedure. In most cases, these children did not require urgent surgical operation, which could be postponed. Because of severe condition, 4 children were not qualified neither to corrective nor to palliative surgical procedure; nevertheless, the medical treatment was applied. Discontinuation of treatment of severely compromised patients is another alternative solution described in literature [3]. This point raises the following questions: In which situations should one stop the treatment? Who is to decide? Finally, what are the limits of resistance to multiple surgical trauma? The results of treatment of CHD with associated lesions are worse in comparison to patients with cardiac malformations only; it has been proven in this study. Multiple system involvement is described as a risk factor for CHD treatment [3,19]. Association of esophageal atresia and tracheoesophageal fistula with ToF, presence of VACTERL, or choanal atresia usually contributes to a poor outcome [2,3,14]. In addition, the presence of diaphragmatic hernia [20] increases perioperatvie mortality in children operated for CHD. No relation of associated lesion with increased mortality was found. Although mortality in multiorgan malformation is high, some authors find that only one fifth of the patients die because of cardiac problems, which has been proven by observations of patients with CHARGE [3]. As far as case
1680 of associated esophageal atresia is concerned, only 6% of patients die from cardiac defects [2]. This may correspond with our findings; there were patients operated in the first turn by surgical team who did not reach the second stage and died. Moreover, the mortality among patients operated only for CHD was lower in comparison with the mortality of 2-stage management. On the other hand, increased mortality among formerly mentioned groups can be the result of suboptimal condition created by CHD presence before its correction. The correction of cardiac lesion had a positive influence on mortality in our logistic regression model. Age factor had a similar impact; older patients had greater chances of survival. We should bear in mind that there are patients who cannot wait because of their severe condition. Children with CoA and other defects had increased death risk, which is converse to that of the PDA group in which no cases of death were observed. The presence of CoA could be an aggravating factor, especially among patients with late diagnosis or with deferral of cardiac correction. The group of other defects consists of rare and complex heart malformations whose correction poses apparent risk. The positive impact of PDA can be explained by the possibility of rapid stabilization of patient hemodynamics. This is accomplished by simple ligation procedure and may facilitate further management. The positive effect of patient’s age was observed; only younger patients usually required urgent surgical correction that could jeopardize the outcome. No influence of chromosomal aberrations on mortality was observed. Down’s syndrome with atrioventricular septal defect and associated lesions was diagnosed among only 5 patients. Management of children with CHD with associated multiple lesions is connected with higher mortality risk. The following factors have negative impact on treatment’s outcome: younger age, urgency of surgical procedure, and primary surgical procedure. However, these risk factors are sometimes inevitable because of the patient’s poor condition. The cardiac procedure preceding the surgical operation may improve the overall effect of treatment because of circulatory stabilization, provided that the condition of the patient does not preclude any intervention at all. Further studies are needed to assess long-term outcome in these severely compromised patients.
M. Wojtalik et al.
References [1] Friedman WF. Congenital heart disease in infancy and childhood in heart diseases. In: Braunwald E, editor. A textbook of cardiovascular medicine. Philadelphia7 WB Saunders Company; 1997. [2] Driver CP, Shankat KR, Jones MO, et al. Phenotypic presentations and outcome of esophageal atresia in the era of the Spitz classification. J Pediatr Surg 2001;36:1419 - 21. [3] Wyse RKH, Al-Mahdawi S, Burn J, et al. Congenital heart disease in CHARGE association. Pediatr Cardiol 1993;14:75 - 81. [4] Annual report of Polish cardiac surgeons’ club. Warszawa. 2002. [5] Kramer HH, Majewski F, Trampisch HJ, et al. Malformation patterns in children with congenital heart disease. Am J Dis Child 1987;141: 789 - 95. [6] Stoll C, Alembik Y, Roth MP, et al. Risk factors in congenital heart disease. Eur J Epidemiol 1989;5:382 - 91. [7] Grech V, Gatt M. Syndromes and malformations associated with congenital heart disease in a population-based study. Int J Cardiol 1999;68:151 - 6. [8] Stoll C, Garne E, Clementi M. Evaluation of prenatal diagnosis of associated congenital heart diseases by fetal ultrasonographic examination in Europe. Prenat Diagn 2001;21:243 - 52. [9] Reinhold-Richter L, Fischer A, Schneider-Obermeyer J. Congenital heart defects. Frequency at autopsy. Zentralbl Allg Pathol 1987;133: 253 - 61. [10] Murugasu B, Yip WC, Tay JS, et al. Sonographic screening for renal tract anomalies associated with congenital heart disease. J Clin Ultrasound 1990;18:79 - 83. [11] Derchi L, Giusti G, Reginato E, et al. The incidence of urinary tract malformations in congenital heart disease. Thorac Cardiovasc Surg 1984;32:78 - 80. [12] Pradat P. Noncardiac malformations at major congenital heart defects. Pediatr Cardiol 1997;18:11 - 8. [13] Pongiglione G, Marasini M, Ribaldone D, et al. Extracardiac malformations and congenital heart disease: frequency and patterns of associations. G Ital Cardiol 1987;17:419 - 25. [14] Saydam TC, Mychaliska CB, Harrison MR. Esophageal lung with multiple congenital anomalies: conundrums in diagnosis and management. J Pediatr Surg 1999;34:615 - 8. [15] De Moor MMA, Baruch R, Human DG. Frontonasal dysplasia associated with tetralogy of Fallot. J Med Genet 1987;24:107 - 9. [16] Hall BD. Choanal atresia and associated multiple anomalies. J Pediatr 1979;95:395. [17] De Ugarte DA, Boechat MI, Shaw WW, et al. Parasitic ophalopagus complicated by omphalocele and congenital heart disease. J Pediatr Surg 2002;37:1357 - 8. [18] Abdallach HI, Marks LA, Balsara RK, et al. Staged repair of pentalogy of Cantrell with tetralogy of Fallot. Ann Thorac Surg 1993;56:979 - 80. [19] Roodpeyma S, Kamali Z, Afshar F, et al. Risk factors in congenital heart disease. Clin Pediatr (Phila) 2002;41:653 - 8. [20] Cunniff C, Jones KL, Jones MC. Patterns of malformation in children with congenital diaphragmatic defects. J Pediatr 1990;116:258 - 61.