- Email: [email protected]
Single Ventricle Palliation in Low Weight Patients Is Associated With Worse Early And Midterm Outcomes
Single Ventricle Palliation in Low Weight Patients Is Associated With Worse Early And Midterm Outcomes
CONGENITAL HEART
Bahaaldin Alsoufi, MD, Courtney McCracken, PhD, Alexandra Ehrlich, MPH, William T. Mahle, MD, Brian Kogon, MD, William Border, MBChB, MPH, Christopher Petit, MD, and Kirk Kanter, MD Division of Cardiothoracic Surgery, and Sibley Heart Center, Children’s Healthcare of Atlanta, Emory University School of Medicine, Atlanta, Georgia
Background. While low weight is an established risk factor for operative mortality after single ventricle (SV) palliation, its influence on late outcomes is not well understood. We examined current-era effects of low weight at time of surgery on hospital mortality, progression through palliative stages, and survival. Methods. Five hundred and thirty infants with SV underwent first-stage palliation (2002 to 2012). Competing risk analysis modeled events after initial surgery and after Glenn. Regression models examined the effect of low weight 2.5 kg or less (n [ 77 of 530, 14.5%) on early and late outcomes. Results. Initial palliation was Norwood (n [ 284, 54%), modified Blalock-Taussig shunt (n [ 173, 33%), and pulmonary artery band (n [ 73, 14%). Competing risk analysis showed that at 6 months after initial palliation the proportion of patients who had died or received transplantation was 40% in patients 2.5 kg or less and 20% in patients greater than 2.5 kg (p < 0.001). Consequently, the proportion of patients who had progressed to Glenn was
33% in patients 2.5 kg or less and 59% in patients greater than 2.5 kg (p < 0.001). Subsequent to Glenn, progression toward Fontan was unaffected by initial weight. In addition to increased hospital mortality (odds ratio 1.86, 95% confidence interval [CI] 0.93% to 3.70%, p [ 0.08); adjusted hazard analysis showed that weight 2.5 kg or less was associated with diminished late survival (hazard ratio 1.65, 95% CI 1.085% to 2.53%, p [ 0.02) and that was evident for all palliation types and most SV morphologies. Conclusions. Low weight at time of first-stage SV palliation is associated with an increase in both hospital mortality and interstage attrition, with subsequently fewer patients progressing toward the Glenn operation. The increased death hazard in low weight SV patients persists for almost 1 year after initial palliation, suggesting the need for more vigilant monitoring and outpatient care in those high-risk patients.
M
Nonetheless, the majority of SV patients require neonatal palliation. The principles of first-stage palliation are to achieve unobstructed systemic cardiac output, a controlled and reliable source of pulmonary blood flow, and unobstructed pulmonary venous return to the heart. The type of initial palliation is dictated by the SV anatomy and involves placement of a modified Blalock-Taussig shunt (BTS), placement of a pulmonary artery band (PAB) with or without concomitant arch repair, or a Norwood-type operation [1–3]. The incidence of low weight is high in children born with congenital heart disease due to numerous factors such as prematurity, poor intrauterine growth, and associated genetic and extracardiac malformations [4–10]. First-stage palliation in low weight neonates is challenging due to numerous technical and physiologic concerns. Despite several recent improvements in the management of neonates with complex congenital heart disease, low weight (2.5 kg or less) at time of surgery continues to be a significant risk factor for operative mortality after different SV palliative surgeries [6, 7, 10–12]. While few reports have
ultistage palliation is the current basis of the surgical management of neonates and infants born with various single ventricle (SV) cardiac malformations. The anatomy of SV anomalies varies and affects the mode of presentation and subsequently the timing and type of initial palliation [1–3]. Anatomic factors that influence presentation and subsequent intervention are largely related to the degree of obstruction of the pulmonary venous return to the heart, pulmonary outflow tract, systemic outflow tract, or aortic arch [1–3]. Occasionally, SV patients present with a balanced physiology allowing deferment of first palliation surgery for a few months with the Glenn operation, traditionally the second stage surgery, becoming the first surgery performed [3].
Accepted for publication Sept 12, 2014. Address correspondence to Dr Alsoufi, Division of Cardiothoracic Surgery, Emory University School of Medicine, Children’s Healthcare of Atlanta, 1405 Clifton Rd NE, Atlanta, GA 30322; e-mail: balsoufi@ hotmail.com.
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2015;99:668–76) Ó 2015 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2014.09.036
Ann Thorac Surg 2015;99:668–76
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
669
Table 1. Patient Characteristics and Operative Details Stratified by Weight Characteristic
Weight 2.5 kg n ¼ 77 (14.5%)
Weight > 2.5 kg n ¼ 453 (85.5%)
3.1 6 212 76 58
(2.8–3.5) (4–10) (40.0%) (14.4%) (10.9%)
2.3 5 37 37 15
(2.1–2.4) (4–10) (48.1%) (49.3%) (19.5%)
3.2 6 175 39 43
(2.9–3.6) (4–12) (38.6%) (8.6%) (9.5%)
219 79 64 53 34 29 21 31
(41.3%) (14.9%) (12.1%) (10.0%) (6.4%) (5.5%) (4.0%) (5.8%)
27 12 16 7 4 2 4 5
(35.1%) (15.6%) (20.8%) (9.1%) (5.2%) (2.6%) (5.2%) (6.5%)
192 67 48 46 30 27 17 26
(42.4%) (14.8%) (10.6%) (10.2%) (6.6%) (6.0%) (3.7%) (5.7%)
p Value <0.0001 0.53 0.12 <0.0001 0.009 0.28
0.57 284 (53.6%) 173 (32.6%) 73 (13.8%) 333 (62.8%) 152 49 62 21
39 (50.7%) 29 (37.7%) 9 (11.7%) 49 (63.6%) 149 46 63 22
245 (54.1%) 144 (31.8%) 64 (14.1%) 284 (62.7%) 153 50 62 21
0.87 0.90 0.94
* Data is presented as median with interquartile range.
suggested that low weight at time of first-stage surgery might have an additional impact on late outcomes, little information is available about the effect of low weight on later progression toward subsequent palliative surgeries and overall survival [6, 11, 12]. Therefore, we aim in the current study to report our current outcomes and examine risk factors for multistage palliation of various SV anomalies with a focus on the effect of low weight 2.5 kg or less on hospital survival, progression toward the Glenn and later the Fontan operation, and mid-term survival.
Patients and Methods Inclusion Criteria From 2002 to 2012, 530 infants born with various SV anomalies underwent first-stage palliation at Children’s Healthcare of Atlanta, Emory University. Our patient cohort included infants who had a PAB, BTS, or Norwood as their first-stage surgery. Neonates who underwent bilateral branch pulmonary artery banding for hypoplastic left heart syndrome (HLHS) were excluded. The SV infants in whom the first procedure was a Glenn bidirectional cavopulmonary connection were also excluded. Patients were identified using our institutional surgical database. Demographic, anatomic, clinical, operative, and hospital details were abstracted from medical records for analysis. Approval of this study was obtained from our hospital’s Institutional Review Board and requirement for individual consent was waived for this observational study.
Follow-Up Follow-up was obtained from the recent electronic charts at Children’s Healthcare of Atlanta for children followed in our system or from direct correspondence with other pediatric cardiologists outside the system. Mean followup duration after first-stage surgery was 4.9 3.9 years and was 91% complete.
Statistical Analysis Data are presented as means with standard deviations, medians with interquartile ranges (IQR), or frequencies as appropriate. Comparisons between patients 2.5 kg or less and those greater than 2.5 kg were performed using the Fisher exact c2 and Student t tests assuming unequal variance between groups (Satterthwaite method). Comparisons for postoperative ventilation time, postoperative intensive care unit (ICU) stay, and postoperative hospital stay were performed after natural log transformation. Linear regression models were used to determine the association between weight and duration of postoperative ventilation, postoperative ICU stay, and postoperative hospital stay. Models were adjusted for age at procedure, sex, premature birth, chromosomal and extracardiac anomalies, STAT surgical risk category, and use of cardiopulmonary bypass. These analyses were performed twice; once dichotomizing weight as 2.5 kg or less versus greater than 2.5 kg and a second time using weight at surgery as a continuous variable. Associations between weight and unplanned cardiac reoperation, extracorporeal membrane oxygenation (ECMO) support, and
CONGENITAL HEART
Weight (kg) Age (days)* Female sex Prematurity 36 weeks Chromosomal/extracardiac anomalies Single ventricle morphology Hypoplastic left heart syndrome Tricuspid atresia Atrial isomerism Pulmonary atresia/intact ventricular septum Double inlet left ventricle Double outlet right ventricle Unbalanced atrioventricular septal defect Other First palliation surgery Norwood Modified Blalock-Taussig shunt Pulmonary artery band Cardiopulmonary bypass use Cardiopulmonary bypass duration (minutes) Aortic cross-clamp duration (minutes)
All Patients n ¼ 530
CONGENITAL HEART
670
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
in-hospital death were modeled in univariate logistic regression models adjusted for the factors previously mentioned. Time-dependent outcomes (death or transplantation and survival to Glenn) after first-stage palliative procedure and time-dependent outcomes (death or transplantation and survival to Fontan) after Glenn, in addition to late survival were parametrically modeled. Parametric probability estimates for time-dependent outcomes uses models based on multiple, overlapping phases of risk (available for use with the SAS system at http://www.clevelandclinic.org/heartcenter/hazard). The HAZARD procedure uses maximum likelihood estimates to resolve risk distribution of time to event in up to 3 phases of risk (early, constant, and late). Models were created using both the dichotomization mentioned above and with weight as a continuous variable, and were adjusted for the factors mentioned above. Competing risk analysis was performed to model the probability over time of each of the following 2 mutually exclusive endpoints after first palliative surgery: death or transplantation and survival to Glenn; the remaining patients being alive without Glenn. After the Glenn, mutually
Ann Thorac Surg 2015;99:668–76
exclusive endpoints were death or transplantation and survival to Fontan; the remaining patients being alive awaiting Fontan. Finally, the effect of weight 2.5 kg or less on mortality in different subgroups of patients was assessed using the parametric hazard regression modeling described above (without adjustment) and hazard ratio with 95% confidence interval for weight 2.5 kg or less in each subgroup is presented in a forest plot. Mean imputation was used instead of missing values for multivariable regression models. All statistical analyses were performed using SAS v9.3 (SAS Institute, Cary, NC).
Results Patient Characteristics During the study period, 530 infants born with various SV anomalies underwent first-stage palliation at our institution. Of those, 453 (85.5%) weighed greater than 2.5 kg while 77 (14.5%) weighed 2.5 kg or less at time of surgery. Median age at time of surgery was 6 days (IQR 4 to 10). Overall there were 318 males (60%), 76 children (14.4%)
Fig 1. (A) Association between unplanned cardiac reoperation and weight at surgery. (B) Association between extracorporeal membrane oxygenation (ECMO) requirement and weight at surgery. (C) Association between hospital mortality and weight at surgery.
Ann Thorac Surg 2015;99:668–76
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
671
were born prematurely 36 weeks or less gestation, and 58 (10.9%) had associated chromosomal or major extracardiac anomalies. Specific underlying SV anomalies included HLHS (n ¼ 219, 41%), tricuspid atresia (n ¼ 79, 15%), atrial isomerism associated with heterotaxy syndrome (n ¼ 64, 12%), pulmonary atresia with intact ventricular septum (pulmonary atresia with intact ventricular septum [PA/IVS], n¼53, 10%), double inlet left ventricle (n ¼ 34, 6%), double outlet right ventricle not amenable to biventricular repair (n ¼ 29, 5%), unbalanced atrioventricular septal defect (n ¼ 21, 4%), and other (n ¼ 31, 6%). Overall, 357 (67%) had unrestricted pulmonary blood flow while 173 (33%) had pulmonary stenosis or atresia. Arch obstruction was evident in 248 patients (47%). Table 1 depicts the differences in patients’ characteristics between those less than and greater than 2.5 kg at time of cardiac surgery. Of note, the incidence of prematurity and chromosomal or extracardiac malformations was significantly higher in infants 2.5 kg or less.
Operative Details Overall, cardiopulmonary bypass was used in 333 cases (63%). The type of first-stage palliation was Norwood (n ¼ 284, 54%), BTS (n ¼ 173, 33%), and PAB (n ¼ 73, 14%), including 29 (5%) who required simultaneous arch repair at time of PAB. Distribution of STAT categories were category 4 (n ¼ 231, 44%) and category 5 (n ¼ 299, 56%). Concomitant cardiac surgery was needed in 106 patients (20%), including repair of total anomalous pulmonary venous connection in 21 patients (4%). Table 1 depicts the differences in operative details between those 2.5 kg or less and those greater than 2.5 kg at time of cardiac surgery.
The distribution of the types of first-stage palliative surgeries was not different between the 2 groups of patients.
Hospital Outcomes We examined the effect of weight on the following 3 early clinical outcomes: unplanned cardiac reoperation; requirement for ECMO support; and hospital mortality. Overall, unplanned cardiac reoperations were performed in 65 infants (12%); 13 infants 2.5 kg or less (17%) and 52 infants greater than 2.5 kg (12%), p ¼ 0.18. In logistic regression models there was no significant association between weight group and unplanned reoperation (odds ratio [OR] 0.55, 95% confidence interval [CI] 0.26% to 1.17%) for 2.5 kg or less, p ¼ 0.12; OR 1.7 per 1 kg increase in weight, 95% CI 1.01% to 2.88%, p ¼ 0.05) (Fig 1A). Postoperative ECMO support was needed in 62 infants (12%); 10 infants 2.5 kg or less (13%) and 52 infants greater than 2.5 kg (12%), p ¼ 0.70. In logistic regression models, there was no significant association between weight group and ECMO requirement (OR 1.11, 95% CI 0.49% to 2.50%) for 2.5 kg or less, p ¼ 0.80; OR 0.75 per 1 kg increase in weight (95% CI 0.43% to 1.30%), p ¼ 0.31 (Fig 1B). Overall hospital mortality was 77 infants (14.5%); 21 infants 2.5 kg or less (27%) and 56 infants greater than 2.5 kg (12%), p ¼ 0.0006. On logistic regression models, there was a trend for diminished hospital survival in the low weight group (OR 1.86, 95% CI 0.93% to 3.70%) for 2.5 kg or less, p ¼ 0.08; OR 0.70 per 1 kg increase in weight (95% CI 0.42% to 1.18%, p ¼ 0.18) (Fig 1C). Additionally, we examined the effect of weight on resource utilization; postoperative ventilation requirement, ICU stay, and hospital stay. Overall postoperative
CONGENITAL HEART
Fig 2. (A) Difference in postoperative ventilation, intensive care unit (ICU), and hospital stay between patients 2.5 kg or less (in dark gray) and greater than 2.5 kg (in light gray). (B) Association between postoperative ventilation duration and weight as a continuous variable. (C) Association between postoperative intensive care unit (ICU) stay and weight as a continuous variable. (D) Association between postoperative hospital stay (Post-OR) and weight as a continuous variable.
CONGENITAL HEART
672
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
mechanical ventilation duration was 124 hours (IQR 70 to 276); 166 hours (IQR 94 to 430) in infants 2.5 kg or less and 123 hours (IQR 66 to 260) in infants greater than 2.5 kg, p ¼ 0.14. Postoperative ICU stay was 193 hours (IQR 117 to 368); 216 hours (IQR 114 to 365) in infants 2.5 kg or less and 192 hours (IQR 109 to 373) in infants greater than 2.5 kg, p ¼ 0.11. Postoperative hospital stay was 14 days (IQR 8 to 26); 14 days (IQR 9 to 27) in infants 2.5 kg or less and 14 days (IQR 8 to 26) in infants greater than 2.5 kg, p ¼ 0.53 (Fig 2A). While the difference in resource utilization was not significant between the 2 weight groups, there was a significant effect of weight as a continuous variable on postoperative ventilation requirement (OR -0.18 per 1 kg increase in weight, 95% CI -0.33 - (-0.03), p ¼ 0.02), ICU stay requirement (OR -0.15 per 1 kg increase in weight (95% CI -0.28 - (-0.02), p ¼ 0.03), and trend for increased hospital stay requirement (OR -0.10 per 1 kg increase in weight (95% CI -0.22 - (-0.02), p ¼ 0.10) in adjusted logistic regression models; suggesting that the very small patients were associated with the longest ventilation, ICU, and hospital stay durations (Figs 2B, 2C, 2D).
Competing Risk Analysis for Events After First-stage Palliation Competing risk models showed that the proportion of patients who underwent Glenn started to rise around 3 months and peaked around 5 months after first palliation surgery. The hazard function for death or transplantation prior to Glenn was characterized by the presence of an early hazard phase during the initial 3 months after first palliation surgery that significantly decreased after that period. Competing risk analysis showed that at 6 months after first palliation surgery 22% of patients had died or received transplantation, 55% had undergone Glenn, and 23% were alive without Glenn. At 2 years after first palliation surgery 25% of patients had died or received transplantation, 75% had undergone Glenn, and less than 1% were alive without Glenn. Competing risk models showed that the proportion of patients who underwent Fontan started to rise around 1 year and peaked around 2 years after Glenn. The hazard function for death or transplantation prior to Fontan was characterized by the presence of an early hazard phase during the initial 4 months after surgery and a persistent
Fig 3. (A) Competing risk analysis of events after first palliation surgery (death or transplantation before Glenn, Glenn) in infants 2.5 kg or less at time of first surgery. (B) Competing risk analysis of events after first palliation surgery (death or transplantation before Glenn, Glenn) in infants greater than 2.5 kg at time of first surgery.
Ann Thorac Surg 2015;99:668–76
low hazard that continued with time. Competing risk analysis showed that at 2 years after Glenn 7% of patients had died or received transplantation, 34% had undergone Fontan, and 59% were alive awaiting Fontan. At 5 years after Glenn 11% of patients had died or received transplantation, 71% had undergone Fontan, and 18% were alive awaiting Fontan. Figure 3 depicts the difference in the competing risk models after first palliation surgery between patients 2.5 kg or less (Fig 3A) and greater than 2.5 kg (Fig 3B). At 6 months after first palliation surgery, the proportion of patients who had died or received transplantation was 40% in patients 2.5 kg or less and 20% in patients greater than 2.5 kg at time of first surgery (p < 0.001). Consequently, the proportion of patients who had progressed to Glenn was 33% in patients 2.5 kg or less and 59% in patients greater than 2.5 kg at time of first surgery (p < 0.001), while the proportion of patients who were alive without Glenn was not significantly different; 27% in patients 2.5 kg or less and 21% in patients greater than 2.5 kg at time of first surgery (p ¼ 0.21). Subsequent to Glenn, the progression toward Fontan was comparable between the 2 weight groups (Figs 4A, 4B).
Midterm Survival The effect of weight group on mid-term survival was examined. Unadjusted survival analysis showed that weight 2.5 kg or less was a significant risk factor for mortality (hazard ratio [HR] 2.09; 95% CI 1.44% to 3.03%), p < 0.001). Adjusted regression models continued to show that weight 2.5 kg or less was a significant risk factor for mortality (HR 1.65; 95% CI 1.08% to 2.53%, p ¼ 0.02). Similarly, the effect of weight as a continuous variable on survival was examined. Unadjusted hazard analysis showed that lower weight was a significant risk factor for mortality (HR 0.48 per kg increase in weight; 95% CI 0.36% to 0.63%, p < 0.001). Adjusted regression models continued to show that lower weight was a significant risk factor for mortality (HR 0.55; 95% CI 0.40% to 0.76%, p < 0.001). Figure 5A depicts the parametric model for survival after surgery stratified by weight groups 2.5 kg or less and greater than 2.5 kg. The hazard function for mortality for both groups over time is depicted in (B).
Ann Thorac Surg 2015;99:668–76
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
673
Fig 4. (A) A parametric model depicting the progression to Fontan after Glenn stratified by weight at time of first palliation surgery. (B) Hazard function for progression to Fontan after Glenn stratified by weight at time of first palliation surgery.
Comment The reported incidence of low weight at time of congenital cardiac surgery in the literature is 8% to 18% and it seems that this rate is comparable between patients with SV and biventricular cardiac pathologies [4–10]. At our institution, the incidence of weight 2.5 kg or less in all infants who underwent cardiac surgery during the study period was 13.4% and was not different between patients who had SV (14.5%) and those who had biventricular (12.8%) cardiac anomalies [6]. The higher frequency of low weight in children with congenital heart disease is due to the higher incidence of poor intrauterine growth rate, prematurity, and presence of associated extracardiac congenital and genetic abnormalities [4–10]. Given the worse early and late outcomes in low weight children undergoing cardiac surgery, a great deal of controversy exists about the optimal management of low weight infants with biventricular cardiac malformations
and whether complete repair versus initial palliation should be offered [4–6, 10]. This issue is obviously irrelevant to our patient population as intracardiac repair is not an option and multistage palliation is the current mainstay of surgical management of SV patients. The important question remains whether low weight babies should have their first-stage palliative surgery (BTS, PAB, or Norwood) early on or if they should have this firststage palliative surgery deferred until they gain weight, with the hope of decreasing their postoperative complications and improving their survival. Nonetheless, the reported experience with this delayed intervention approach has not been favorable and emergence of problems such as ventilator dependency, necrotizing enterocolitis, renal dysfunction, sepsis, and neurologic complications has been of concern [4, 5, 10, 13]. Moreover, desired adequate weight gain is infrequently achieved leading to unnecessary prolonged delays in surgical treatment and development of those complications [4, 5, 13]. The SV physiology further complicates the issues, especially in patients with unrestrictive pulmonary blood flow who require an early intervention (PAB, Norwood, or Damus-Kay-Stansel þ shunt) to prevent pulmonary over-circulation and subsequent development of pulmonary vascular disease that might prevent the children from progressing toward final palliative stage, and have a negative effect on early and late outcomes [1, 2]. Advanced age at time of Norwood operation has been repeatedly demonstrated to be a risk factor complicating Fig 5. (A) Parametric model for late survival stratified by weight group 2.5 kg or less and greater than 2.5 kg. (B) Hazard function for mortality stratified by weight group 2.5 kg or less and greater than 2.5 kg.
CONGENITAL HEART
The effect of weight 2.5 kg or less (vs > 2.5 kg) on postoperative survival among different selected patient subgroups is depicted in Figure 6. Weight 2.5 kg or less was found to be associated with increased hazard of mortality in most subgroups of patients except for premature babies and those with SV types associated with the worst prognosis at our institution (ie, atrial isomerism and PA/IVS). In those high-risk patients, it seems that the effect of low weight on survival is neutralized.
CONGENITAL HEART
674
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
Ann Thorac Surg 2015;99:668–76
Fig 6. Association between weight 2.5 kg or less (vs > 2.5 Kg) and postoperative survival among selected patient subgroups. Reported is the hazard ratio (HR) for mortality (with 95% confidence interval) for weight 2.5 kg or less in each patient subgroup. Weight 2.5 kg or less was found to be associated with increased hazard of mortality in most subgroups of patients. (Atrial isomerism in heterotaxy [HTX]; BT ¼ modified Blalock-Taussig; CPB ¼ cardiopulmonary bypass; DILV ¼ double inlet left ventricle; HLHS ¼ hypoplastic left heart syndrome; PAB ¼ pulmonary artery band; PA/IVS ¼ pulmonary atresia with intact ventricular septum; TA ¼ tricuspid atresia.)
postoperative recovery and affecting early and late results due to elevated pulmonary vascular resistance and prolonged volume loading effects on the systemic ventricle prior to palliation [14, 15]. Therefore, we believe that firststage palliation should not be delayed in low weight neonates with SV abnormalities except in those with major comorbidities such as intracranial hemorrhage, severe pulmonary infection, necrotizing enterocolitis, sepsis, significant extracardiac malformation requiring immediate intervention, or multiorgan failure related to shock at the time of initial presentation. Nonetheless, the completion of those palliative procedures in low weight infants is challenging due to technical difficulties related to small-sized cardiac structures and tissue friability, and postoperative management issues related to prematurity, pulmonary dysfunction, and pulmonary hypertension. Low weight has persistently been demonstrated as a risk factor for mortality after the Norwood operation, including in recent series from experienced institutions [11, 12, 14]. Similarly, low weight has been associated with increased mortality risk after BTS due to shunt occlusion related to the small sized shunts or systemic steal due to excessive pulmonary blood flow that is likely more common in smaller babies [16, 17]. In a recent review of the Society of Thoracic Surgeons database, weight less than 3 kg at surgery was
an independent risk factor for hospital mortality after BTS [16]. Finally, PAB in low weight and premature patients is challenging due to the presence of lung disease, elevated pulmonary vascular resistance, and future growth of the baby, which may affect the need for reintervention after PAB [18]. Alternative treatment strategies such as ductal stenting for SV patients with restricted pulmonary blood flow or hybrid palliation (ductal stenting, bilateral branch pulmonary artery banding, and atrial septectomy) in patients with excessive pulmonary blood flow and systemic outflow tract obstruction, have been proposed by some groups with promising results and might prove to be advantageous in high-risk patients including those with low weight [19–22]. In addition to early death, low weight at time of firststage surgery was suggested to have an adverse effect on late survival [3, 11, 12]. Gelehrter and colleagues [11] reviewed outcomes of 47 neonates with HLHS patients who were 2.5 kg or less at time of Norwood palliation. Fifty-one percent of patients survived to initial hospital discharge and overall survival from birth through Fontan was 36%. In their series, they had low interstage mortality and their operative survival after Glenn and Fontan was 91% and 94%, respectively. Pizarro and colleagues [12] studied outcomes of 20 neonates with HLHS and other SV
Ann Thorac Surg 2015;99:668–76
References 1. Tweddell JS, Litwin SB, Thomas JP Jr, Mussatto K. Recent advances in the surgical management of the single ventricle pediatric patient. Pediatr Clin North Am 1999;46:465–80.
675
2. Alsoufi B. Management of the single ventricle and potentially obstructive systemic ventricular outflow tract. J Saudi Heart Assoc 2013;25:191–202. 3. Alsoufi B, Manlhiot C, Awan A, et al. Current outcomes of the Glenn bidirectional cavopulmonary connection for single ventricle palliation. Eur J Cardiothorac Surg 2012;42:42–8. 4. Reddy VM. Low birth weight and very low birth weight neonates with congenital heart disease: timing of surgery, reasons for delaying or not delaying surgery. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu 2013;16:13–20. 5. Chang AC, Hanley FH, Lock JE, Castaneda AR, Wessel DL. Management and outcome of low birth weight neonates with congenital cardiac disease. J Pediatr 1994;124:461–6. 6. Alsoufi B, Manlhiot C, Mahle WT, et al. Low weight infants are at increased mortality risk following palliative or corrective cardiac surgery. J Thoracic Cardiovasc Surg 2014; pii: S0022-5223(14)01019-8. http://dx.doi.org/10.1016/j.jtcvs. 2014.07.047. [Epub ahead of print]. 7. Curzon CL, Milford-Beland S, Li JS, et al. Cardiac surgery in infants with low birth weight is associated with increased mortality: analysis of the Society of Thoracic Surgeons Congenital Heart database. J Thoracic Cardiovasc Surg 2008;135:546–51. 8. Reddy VM, McElhinney DB, Sagrado T, Parry AJ, Teitel DF, Hanley FL. Results of 102 cases of complete repair of congenital heart defects in patients weighing 700 to 2500 grams. J Thorac Cardiovasc Surg 1999;117:324–31. 9. Oppido G, Napoleone CP, Formigari R, et al. Outcome of cardiac surgery in low birth weight and premature infants. Eur J Cardiothorac Surg 2004;26:44–53. 10. Hickey EJ, Nosikova Y, Zhang H, et al. Very low-birth-weight infants with congenital cardiac lesions: is there merit in delaying intervention to permit growth and maturation? J Thorac Cardiovasc Surg 2012;143:126–36. 11. Gelehrter S, Fifer CG, Armstrong A, Hirsch J, Gajarski R. Outcomes of hypoplastic left heart syndrome in low-birthweight patients. Pediatr Cardiol 2011;32:1175–81. 12. Pizarro C, Davis DA, Galantowicz ME, Munro H, Gidding SS, Norwood WI. Stage I palliation for hypoplastic left heart syndrome in low birth weight neonates: can we justify it? Eur J Cardiothorac Surg 2002;21:716–20. 13. Wernovsky G, Rubenstein SD, Spray TL. Cardiac surgery in the low birth weight neonate: new approaches. Clin Perinatol 2001;28:249–64. 14. Mahle WT, Spray TL, Wernovsky G, Gaynor JW, Clark BJ. Survival after reconstructive surgery for hypoplastic left heart syndrome. Circulation 2000;102(19 Suppl 3):III136–41. 15. Alsoufi B, Manlhiot C, Al-Ahmadi M, et al. Older children at the time of the Norwood operation have ongoing mortality vulnerability that continues after cavopulmonary connection. J Thorac Cardiovasc Surg 2011;142:142-7. 16. Petrucci O, O’Brien SM, Jacobs ML, Jacobs JP, Manning PB, Eghtesady P. Risk factors for mortality and morbidity after the neonatal Blalock-Taussig shunt procedure. Ann Thorac Surg 2011;92:642–51. 17. Myers JW, Ghanayem NS, Cao Y, et al. Outcomes of systemic to pulmonary artery shunts in patients weighing less than 3 kg: analysis of shunt type, size, and surgical approach. J Thorac Cardiovasc Surg 2014;147:672–7. 18. Alsoufi B, Manlhiot C, Ehrlich A, et al. Results of palliation with an initial pulmonary artery band in patients with single ventricle associated with unrestricted pulmonary blood flow. J Thorac Cardiovasc Surg 2014; pii: S0022-5223(14)01077-0. http://dx.doi.org/10.1016/j.jtcvs.2014.08.007. [Epub ahead of print]. 19. McMullan DM, Permut LC, Jones TK, Johnston TA, Rubio AE. Modified Blalock-Taussig shunt versus ductal stenting for palliation of cardiac lesions with inadequate pulmonary blood flow. J Thorac Cardiovasc Surg 2014;147:397–401. 20. Galantowicz M, Cheatham JP, Phillips A, et al. Hybrid approach for hypoplastic left heart syndrome: intermediate results after the learning curve. Ann Thorac Surg 2008;85: 2063–70.
CONGENITAL HEART
variants who were 2.5 kg or less at time of Norwood palliation. In their series, 55% survived to initial hospital discharge and on follow-up they had 1 interstage mortality, while the remaining patients progressed to Glenn operation. In our current study, we have demonstrated that low weight was not only associated with higher hospital mortality but also with a higher interstage mortality and subsequently lower percentage of patients reaching the Glenn operation. After Glenn, the progression toward Fontan operation was similar between the 2 weight groups suggesting that the effect of low weight at time of initial palliation was neutralized once the balanced circulation was achieved with the Glenn operation. Similarly, the death hazard in low weight patients was more pronounced and more prolonged for almost a year after first-stage palliation before normalizing, suggesting that those low weight babies might require a more vigilant monitoring and care after hospital survival. Similar to other reports in the literature, the SV morphologies that were associated with the worst prognosis in our current series were atrial isomerism in heterotaxy patients and PA/IVS. The higher risk of early and late mortality in heterotaxy patients is due to the high incidence of associated intracardiac defects, particularly total anomalous pulmonary venous connection, atrioventricular valve regurgitation, and conduction abnormalities in addition to extracardiac anomalies affecting outcomes beyond hospital discharge [23–25]. In patients with PA/ IVS, the increased mortality is attributed to the higher prevalence of ventricular to coronary arteries sinusoids and right ventricle dependant coronary circulation with subsequent tenuous physiologic state and higher risk of ischemia and cardiac arrest prior to or after palliation [16, 26, 27]. While the effect of low weight on mortality was evident in different patient categories in our series, that effect was less definite than in other SV anomalies associated with better prognosis, including HLHS. It is likely that low weight complicates surgery and recovery after palliative surgery in those patients, but the effect of low weight is blunted in the presence of other physiologic and anatomic risk factors affecting mortality in those complex patients. Despite recent advances in the perioperative care of SV patients undergoing multistage palliation, low weight at time of first-stage surgery continues to adversely affect hospital survival after all types of SV palliative surgeries. Additionally, the effect of low weight on outcome extends beyond hospital discharge with significant increase in interstage mortality, decrease proportion of patients progressing toward the Glenn operation, and persistently elevated death hazard up to a year after initial palliation. Consequently, methods to improve outcomes in this high-risk SV population should address perioperative management, outpatient surveillance, and follow-up care.
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
CONGENITAL HEART
676
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
21. Guleserian KJ, Barker GM, Sharma MS, et al. Bilateral pulmonary artery banding for resuscitation in high-risk, singleventricle neonates and infants: a single-center experience. J Thorac Cardiovasc Surg 2013;145:206–13. 22. Venugopal PS, Luna KP, Anderson DR, et al. Hybrid procedure as an alternative to surgical palliation of high-risk infants with hypoplastic left heart syndrome and its variants. J Thorac Cardiovasc Surg 2010;139:1211–5. 23. Jacobs JP, Pasquali SK, Morales DL, et al. Heterotaxy: lessons learned about patterns of practice and outcomes from the congenital heart surgery database of the society of thoracic surgeons. World J Pediatr Congenit Heart Surg 2011;2: 278–86. 24. Hancock Friesen CL, Zurakowski D, Thiagarajan RR, et al. Total anomalous pulmonary venous connection: an analysis
Ann Thorac Surg 2015;99:668–76
of current management strategies in a single institution. Ann Thorac Surg 2005;79:596–606. 25. Lodge AJ, Rychik J, Nicolson SC, Ittenbach RF, Spray TL, Gaynor JW. Improving outcomes in functional single ventricle and total anomalous pulmonary venous connection. Ann Thorac Surg 2004;78:1688–95. 26. Ashburn DA, Blackstone EH, Wells WJ, et al. Determinants of mortality and type of repair in neonates with pulmonary atresia and intact ventricular septum. J Thorac Cardiovasc Surg 2004;127:1000–7. 27. Guleserian KJ, Armsby LB, Thiagarajan RR, del Nido PJ, Mayer JE Jr. Natural history of pulmonary atresia with intact ventricular septum and right-ventricle-dependent coronary circulation managed by the single-ventricle approach. Ann Thorac Surg 2006;81:2250–7.
Thoracic Surgery Foundation for Research and Education: For 25 Years, Turning Today’s Research Into Tomorrow’s Patient Care Our patients don’t follow the details of our research. They don’t discuss unexpected breakthroughs or technical setbacks. They are not always aware of how changes in health care policies impact research funding and lab time. Nonetheless, the advances we make in cardiothoracic surgery touch each and every one of them. New surgical techniques and potent new drugs improve patient health and extend patient lives. TSFRE is a pivotal force for growth and vitality in cardiothoracic surgery, especially for research and academic career development. TSFRE was first established in 1988 as a 501(c)(3) not-for-profit charitable foundation with support from the four major thoracic surgery organizations—the American Association for Thoracic Surgery, The Society of Thoracic Surgeons, the Southern Thoracic Surgical Association, and the Western Thoracic Surgical Association. On October 1, 2014, TSFRE became an official charitable arm of The Society of Thoracic Surgeons. The foundation represents thoracic surgery in the United States and its research and educational initiatives support the broad spectrum of thoracic surgery. The mission of TSFRE is to foster the development of surgeon scientists in cardiothoracic surgery; increasing knowledge and innovation to benefit patient care.
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
Funding basic biomedical research to expand our understanding of thoracic surgery is vitally important. Still, it’s only half the battle. The results of such research must be successfully transferred to patient care. There is a compelling need to continue to develop the skills of cardiothoracic surgeons as scientists and as health policy leaders. TSFRE is an organization that is having a steering effect on cardiothoracic surgery through the sponsorship of research projects and the education of surgeons in health care policy. In its 25-year history, TSFRE support has enabled many young cardiothoracic surgeons to develop successful, independent research careers and improve the care of our patients. To date, TSFRE has funded 146 cardiothoracic surgery research grants, fellowships, and simulation awards, and funded more than 260 Alley-Sheridan Scholarships. Because of your support, many of these individuals are now leaders in the cardiothoracic surgery community. This is a TSFRE legacy in which we can all take pride. Please help TSFRE continue its legacy by making a donation today. Visit www.tsfre.org or contact Priscilla Kennedy, TSFRE Executive Director, at 312-2025868, or by e-mail at [email protected].
Ann Thorac Surg 2015;99:676
0003-4975/$36.00
CONGENITAL HEART
Bahaaldin Alsoufi, MD, Courtney McCracken, PhD, Alexandra Ehrlich, MPH, William T. Mahle, MD, Brian Kogon, MD, William Border, MBChB, MPH, Christopher Petit, MD, and Kirk Kanter, MD Division of Cardiothoracic Surgery, and Sibley Heart Center, Children’s Healthcare of Atlanta, Emory University School of Medicine, Atlanta, Georgia
Background. While low weight is an established risk factor for operative mortality after single ventricle (SV) palliation, its influence on late outcomes is not well understood. We examined current-era effects of low weight at time of surgery on hospital mortality, progression through palliative stages, and survival. Methods. Five hundred and thirty infants with SV underwent first-stage palliation (2002 to 2012). Competing risk analysis modeled events after initial surgery and after Glenn. Regression models examined the effect of low weight 2.5 kg or less (n [ 77 of 530, 14.5%) on early and late outcomes. Results. Initial palliation was Norwood (n [ 284, 54%), modified Blalock-Taussig shunt (n [ 173, 33%), and pulmonary artery band (n [ 73, 14%). Competing risk analysis showed that at 6 months after initial palliation the proportion of patients who had died or received transplantation was 40% in patients 2.5 kg or less and 20% in patients greater than 2.5 kg (p < 0.001). Consequently, the proportion of patients who had progressed to Glenn was
33% in patients 2.5 kg or less and 59% in patients greater than 2.5 kg (p < 0.001). Subsequent to Glenn, progression toward Fontan was unaffected by initial weight. In addition to increased hospital mortality (odds ratio 1.86, 95% confidence interval [CI] 0.93% to 3.70%, p [ 0.08); adjusted hazard analysis showed that weight 2.5 kg or less was associated with diminished late survival (hazard ratio 1.65, 95% CI 1.085% to 2.53%, p [ 0.02) and that was evident for all palliation types and most SV morphologies. Conclusions. Low weight at time of first-stage SV palliation is associated with an increase in both hospital mortality and interstage attrition, with subsequently fewer patients progressing toward the Glenn operation. The increased death hazard in low weight SV patients persists for almost 1 year after initial palliation, suggesting the need for more vigilant monitoring and outpatient care in those high-risk patients.
M
Nonetheless, the majority of SV patients require neonatal palliation. The principles of first-stage palliation are to achieve unobstructed systemic cardiac output, a controlled and reliable source of pulmonary blood flow, and unobstructed pulmonary venous return to the heart. The type of initial palliation is dictated by the SV anatomy and involves placement of a modified Blalock-Taussig shunt (BTS), placement of a pulmonary artery band (PAB) with or without concomitant arch repair, or a Norwood-type operation [1–3]. The incidence of low weight is high in children born with congenital heart disease due to numerous factors such as prematurity, poor intrauterine growth, and associated genetic and extracardiac malformations [4–10]. First-stage palliation in low weight neonates is challenging due to numerous technical and physiologic concerns. Despite several recent improvements in the management of neonates with complex congenital heart disease, low weight (2.5 kg or less) at time of surgery continues to be a significant risk factor for operative mortality after different SV palliative surgeries [6, 7, 10–12]. While few reports have
ultistage palliation is the current basis of the surgical management of neonates and infants born with various single ventricle (SV) cardiac malformations. The anatomy of SV anomalies varies and affects the mode of presentation and subsequently the timing and type of initial palliation [1–3]. Anatomic factors that influence presentation and subsequent intervention are largely related to the degree of obstruction of the pulmonary venous return to the heart, pulmonary outflow tract, systemic outflow tract, or aortic arch [1–3]. Occasionally, SV patients present with a balanced physiology allowing deferment of first palliation surgery for a few months with the Glenn operation, traditionally the second stage surgery, becoming the first surgery performed [3].
Accepted for publication Sept 12, 2014. Address correspondence to Dr Alsoufi, Division of Cardiothoracic Surgery, Emory University School of Medicine, Children’s Healthcare of Atlanta, 1405 Clifton Rd NE, Atlanta, GA 30322; e-mail: balsoufi@ hotmail.com.
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2015;99:668–76) Ó 2015 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2014.09.036
Ann Thorac Surg 2015;99:668–76
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
669
Table 1. Patient Characteristics and Operative Details Stratified by Weight Characteristic
Weight 2.5 kg n ¼ 77 (14.5%)
Weight > 2.5 kg n ¼ 453 (85.5%)
3.1 6 212 76 58
(2.8–3.5) (4–10) (40.0%) (14.4%) (10.9%)
2.3 5 37 37 15
(2.1–2.4) (4–10) (48.1%) (49.3%) (19.5%)
3.2 6 175 39 43
(2.9–3.6) (4–12) (38.6%) (8.6%) (9.5%)
219 79 64 53 34 29 21 31
(41.3%) (14.9%) (12.1%) (10.0%) (6.4%) (5.5%) (4.0%) (5.8%)
27 12 16 7 4 2 4 5
(35.1%) (15.6%) (20.8%) (9.1%) (5.2%) (2.6%) (5.2%) (6.5%)
192 67 48 46 30 27 17 26
(42.4%) (14.8%) (10.6%) (10.2%) (6.6%) (6.0%) (3.7%) (5.7%)
p Value <0.0001 0.53 0.12 <0.0001 0.009 0.28
0.57 284 (53.6%) 173 (32.6%) 73 (13.8%) 333 (62.8%) 152 49 62 21
39 (50.7%) 29 (37.7%) 9 (11.7%) 49 (63.6%) 149 46 63 22
245 (54.1%) 144 (31.8%) 64 (14.1%) 284 (62.7%) 153 50 62 21
0.87 0.90 0.94
* Data is presented as median with interquartile range.
suggested that low weight at time of first-stage surgery might have an additional impact on late outcomes, little information is available about the effect of low weight on later progression toward subsequent palliative surgeries and overall survival [6, 11, 12]. Therefore, we aim in the current study to report our current outcomes and examine risk factors for multistage palliation of various SV anomalies with a focus on the effect of low weight 2.5 kg or less on hospital survival, progression toward the Glenn and later the Fontan operation, and mid-term survival.
Patients and Methods Inclusion Criteria From 2002 to 2012, 530 infants born with various SV anomalies underwent first-stage palliation at Children’s Healthcare of Atlanta, Emory University. Our patient cohort included infants who had a PAB, BTS, or Norwood as their first-stage surgery. Neonates who underwent bilateral branch pulmonary artery banding for hypoplastic left heart syndrome (HLHS) were excluded. The SV infants in whom the first procedure was a Glenn bidirectional cavopulmonary connection were also excluded. Patients were identified using our institutional surgical database. Demographic, anatomic, clinical, operative, and hospital details were abstracted from medical records for analysis. Approval of this study was obtained from our hospital’s Institutional Review Board and requirement for individual consent was waived for this observational study.
Follow-Up Follow-up was obtained from the recent electronic charts at Children’s Healthcare of Atlanta for children followed in our system or from direct correspondence with other pediatric cardiologists outside the system. Mean followup duration after first-stage surgery was 4.9 3.9 years and was 91% complete.
Statistical Analysis Data are presented as means with standard deviations, medians with interquartile ranges (IQR), or frequencies as appropriate. Comparisons between patients 2.5 kg or less and those greater than 2.5 kg were performed using the Fisher exact c2 and Student t tests assuming unequal variance between groups (Satterthwaite method). Comparisons for postoperative ventilation time, postoperative intensive care unit (ICU) stay, and postoperative hospital stay were performed after natural log transformation. Linear regression models were used to determine the association between weight and duration of postoperative ventilation, postoperative ICU stay, and postoperative hospital stay. Models were adjusted for age at procedure, sex, premature birth, chromosomal and extracardiac anomalies, STAT surgical risk category, and use of cardiopulmonary bypass. These analyses were performed twice; once dichotomizing weight as 2.5 kg or less versus greater than 2.5 kg and a second time using weight at surgery as a continuous variable. Associations between weight and unplanned cardiac reoperation, extracorporeal membrane oxygenation (ECMO) support, and
CONGENITAL HEART
Weight (kg) Age (days)* Female sex Prematurity 36 weeks Chromosomal/extracardiac anomalies Single ventricle morphology Hypoplastic left heart syndrome Tricuspid atresia Atrial isomerism Pulmonary atresia/intact ventricular septum Double inlet left ventricle Double outlet right ventricle Unbalanced atrioventricular septal defect Other First palliation surgery Norwood Modified Blalock-Taussig shunt Pulmonary artery band Cardiopulmonary bypass use Cardiopulmonary bypass duration (minutes) Aortic cross-clamp duration (minutes)
All Patients n ¼ 530
CONGENITAL HEART
670
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
in-hospital death were modeled in univariate logistic regression models adjusted for the factors previously mentioned. Time-dependent outcomes (death or transplantation and survival to Glenn) after first-stage palliative procedure and time-dependent outcomes (death or transplantation and survival to Fontan) after Glenn, in addition to late survival were parametrically modeled. Parametric probability estimates for time-dependent outcomes uses models based on multiple, overlapping phases of risk (available for use with the SAS system at http://www.clevelandclinic.org/heartcenter/hazard). The HAZARD procedure uses maximum likelihood estimates to resolve risk distribution of time to event in up to 3 phases of risk (early, constant, and late). Models were created using both the dichotomization mentioned above and with weight as a continuous variable, and were adjusted for the factors mentioned above. Competing risk analysis was performed to model the probability over time of each of the following 2 mutually exclusive endpoints after first palliative surgery: death or transplantation and survival to Glenn; the remaining patients being alive without Glenn. After the Glenn, mutually
Ann Thorac Surg 2015;99:668–76
exclusive endpoints were death or transplantation and survival to Fontan; the remaining patients being alive awaiting Fontan. Finally, the effect of weight 2.5 kg or less on mortality in different subgroups of patients was assessed using the parametric hazard regression modeling described above (without adjustment) and hazard ratio with 95% confidence interval for weight 2.5 kg or less in each subgroup is presented in a forest plot. Mean imputation was used instead of missing values for multivariable regression models. All statistical analyses were performed using SAS v9.3 (SAS Institute, Cary, NC).
Results Patient Characteristics During the study period, 530 infants born with various SV anomalies underwent first-stage palliation at our institution. Of those, 453 (85.5%) weighed greater than 2.5 kg while 77 (14.5%) weighed 2.5 kg or less at time of surgery. Median age at time of surgery was 6 days (IQR 4 to 10). Overall there were 318 males (60%), 76 children (14.4%)
Fig 1. (A) Association between unplanned cardiac reoperation and weight at surgery. (B) Association between extracorporeal membrane oxygenation (ECMO) requirement and weight at surgery. (C) Association between hospital mortality and weight at surgery.
Ann Thorac Surg 2015;99:668–76
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
671
were born prematurely 36 weeks or less gestation, and 58 (10.9%) had associated chromosomal or major extracardiac anomalies. Specific underlying SV anomalies included HLHS (n ¼ 219, 41%), tricuspid atresia (n ¼ 79, 15%), atrial isomerism associated with heterotaxy syndrome (n ¼ 64, 12%), pulmonary atresia with intact ventricular septum (pulmonary atresia with intact ventricular septum [PA/IVS], n¼53, 10%), double inlet left ventricle (n ¼ 34, 6%), double outlet right ventricle not amenable to biventricular repair (n ¼ 29, 5%), unbalanced atrioventricular septal defect (n ¼ 21, 4%), and other (n ¼ 31, 6%). Overall, 357 (67%) had unrestricted pulmonary blood flow while 173 (33%) had pulmonary stenosis or atresia. Arch obstruction was evident in 248 patients (47%). Table 1 depicts the differences in patients’ characteristics between those less than and greater than 2.5 kg at time of cardiac surgery. Of note, the incidence of prematurity and chromosomal or extracardiac malformations was significantly higher in infants 2.5 kg or less.
Operative Details Overall, cardiopulmonary bypass was used in 333 cases (63%). The type of first-stage palliation was Norwood (n ¼ 284, 54%), BTS (n ¼ 173, 33%), and PAB (n ¼ 73, 14%), including 29 (5%) who required simultaneous arch repair at time of PAB. Distribution of STAT categories were category 4 (n ¼ 231, 44%) and category 5 (n ¼ 299, 56%). Concomitant cardiac surgery was needed in 106 patients (20%), including repair of total anomalous pulmonary venous connection in 21 patients (4%). Table 1 depicts the differences in operative details between those 2.5 kg or less and those greater than 2.5 kg at time of cardiac surgery.
The distribution of the types of first-stage palliative surgeries was not different between the 2 groups of patients.
Hospital Outcomes We examined the effect of weight on the following 3 early clinical outcomes: unplanned cardiac reoperation; requirement for ECMO support; and hospital mortality. Overall, unplanned cardiac reoperations were performed in 65 infants (12%); 13 infants 2.5 kg or less (17%) and 52 infants greater than 2.5 kg (12%), p ¼ 0.18. In logistic regression models there was no significant association between weight group and unplanned reoperation (odds ratio [OR] 0.55, 95% confidence interval [CI] 0.26% to 1.17%) for 2.5 kg or less, p ¼ 0.12; OR 1.7 per 1 kg increase in weight, 95% CI 1.01% to 2.88%, p ¼ 0.05) (Fig 1A). Postoperative ECMO support was needed in 62 infants (12%); 10 infants 2.5 kg or less (13%) and 52 infants greater than 2.5 kg (12%), p ¼ 0.70. In logistic regression models, there was no significant association between weight group and ECMO requirement (OR 1.11, 95% CI 0.49% to 2.50%) for 2.5 kg or less, p ¼ 0.80; OR 0.75 per 1 kg increase in weight (95% CI 0.43% to 1.30%), p ¼ 0.31 (Fig 1B). Overall hospital mortality was 77 infants (14.5%); 21 infants 2.5 kg or less (27%) and 56 infants greater than 2.5 kg (12%), p ¼ 0.0006. On logistic regression models, there was a trend for diminished hospital survival in the low weight group (OR 1.86, 95% CI 0.93% to 3.70%) for 2.5 kg or less, p ¼ 0.08; OR 0.70 per 1 kg increase in weight (95% CI 0.42% to 1.18%, p ¼ 0.18) (Fig 1C). Additionally, we examined the effect of weight on resource utilization; postoperative ventilation requirement, ICU stay, and hospital stay. Overall postoperative
CONGENITAL HEART
Fig 2. (A) Difference in postoperative ventilation, intensive care unit (ICU), and hospital stay between patients 2.5 kg or less (in dark gray) and greater than 2.5 kg (in light gray). (B) Association between postoperative ventilation duration and weight as a continuous variable. (C) Association between postoperative intensive care unit (ICU) stay and weight as a continuous variable. (D) Association between postoperative hospital stay (Post-OR) and weight as a continuous variable.
CONGENITAL HEART
672
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
mechanical ventilation duration was 124 hours (IQR 70 to 276); 166 hours (IQR 94 to 430) in infants 2.5 kg or less and 123 hours (IQR 66 to 260) in infants greater than 2.5 kg, p ¼ 0.14. Postoperative ICU stay was 193 hours (IQR 117 to 368); 216 hours (IQR 114 to 365) in infants 2.5 kg or less and 192 hours (IQR 109 to 373) in infants greater than 2.5 kg, p ¼ 0.11. Postoperative hospital stay was 14 days (IQR 8 to 26); 14 days (IQR 9 to 27) in infants 2.5 kg or less and 14 days (IQR 8 to 26) in infants greater than 2.5 kg, p ¼ 0.53 (Fig 2A). While the difference in resource utilization was not significant between the 2 weight groups, there was a significant effect of weight as a continuous variable on postoperative ventilation requirement (OR -0.18 per 1 kg increase in weight, 95% CI -0.33 - (-0.03), p ¼ 0.02), ICU stay requirement (OR -0.15 per 1 kg increase in weight (95% CI -0.28 - (-0.02), p ¼ 0.03), and trend for increased hospital stay requirement (OR -0.10 per 1 kg increase in weight (95% CI -0.22 - (-0.02), p ¼ 0.10) in adjusted logistic regression models; suggesting that the very small patients were associated with the longest ventilation, ICU, and hospital stay durations (Figs 2B, 2C, 2D).
Competing Risk Analysis for Events After First-stage Palliation Competing risk models showed that the proportion of patients who underwent Glenn started to rise around 3 months and peaked around 5 months after first palliation surgery. The hazard function for death or transplantation prior to Glenn was characterized by the presence of an early hazard phase during the initial 3 months after first palliation surgery that significantly decreased after that period. Competing risk analysis showed that at 6 months after first palliation surgery 22% of patients had died or received transplantation, 55% had undergone Glenn, and 23% were alive without Glenn. At 2 years after first palliation surgery 25% of patients had died or received transplantation, 75% had undergone Glenn, and less than 1% were alive without Glenn. Competing risk models showed that the proportion of patients who underwent Fontan started to rise around 1 year and peaked around 2 years after Glenn. The hazard function for death or transplantation prior to Fontan was characterized by the presence of an early hazard phase during the initial 4 months after surgery and a persistent
Fig 3. (A) Competing risk analysis of events after first palliation surgery (death or transplantation before Glenn, Glenn) in infants 2.5 kg or less at time of first surgery. (B) Competing risk analysis of events after first palliation surgery (death or transplantation before Glenn, Glenn) in infants greater than 2.5 kg at time of first surgery.
Ann Thorac Surg 2015;99:668–76
low hazard that continued with time. Competing risk analysis showed that at 2 years after Glenn 7% of patients had died or received transplantation, 34% had undergone Fontan, and 59% were alive awaiting Fontan. At 5 years after Glenn 11% of patients had died or received transplantation, 71% had undergone Fontan, and 18% were alive awaiting Fontan. Figure 3 depicts the difference in the competing risk models after first palliation surgery between patients 2.5 kg or less (Fig 3A) and greater than 2.5 kg (Fig 3B). At 6 months after first palliation surgery, the proportion of patients who had died or received transplantation was 40% in patients 2.5 kg or less and 20% in patients greater than 2.5 kg at time of first surgery (p < 0.001). Consequently, the proportion of patients who had progressed to Glenn was 33% in patients 2.5 kg or less and 59% in patients greater than 2.5 kg at time of first surgery (p < 0.001), while the proportion of patients who were alive without Glenn was not significantly different; 27% in patients 2.5 kg or less and 21% in patients greater than 2.5 kg at time of first surgery (p ¼ 0.21). Subsequent to Glenn, the progression toward Fontan was comparable between the 2 weight groups (Figs 4A, 4B).
Midterm Survival The effect of weight group on mid-term survival was examined. Unadjusted survival analysis showed that weight 2.5 kg or less was a significant risk factor for mortality (hazard ratio [HR] 2.09; 95% CI 1.44% to 3.03%), p < 0.001). Adjusted regression models continued to show that weight 2.5 kg or less was a significant risk factor for mortality (HR 1.65; 95% CI 1.08% to 2.53%, p ¼ 0.02). Similarly, the effect of weight as a continuous variable on survival was examined. Unadjusted hazard analysis showed that lower weight was a significant risk factor for mortality (HR 0.48 per kg increase in weight; 95% CI 0.36% to 0.63%, p < 0.001). Adjusted regression models continued to show that lower weight was a significant risk factor for mortality (HR 0.55; 95% CI 0.40% to 0.76%, p < 0.001). Figure 5A depicts the parametric model for survival after surgery stratified by weight groups 2.5 kg or less and greater than 2.5 kg. The hazard function for mortality for both groups over time is depicted in (B).
Ann Thorac Surg 2015;99:668–76
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
673
Fig 4. (A) A parametric model depicting the progression to Fontan after Glenn stratified by weight at time of first palliation surgery. (B) Hazard function for progression to Fontan after Glenn stratified by weight at time of first palliation surgery.
Comment The reported incidence of low weight at time of congenital cardiac surgery in the literature is 8% to 18% and it seems that this rate is comparable between patients with SV and biventricular cardiac pathologies [4–10]. At our institution, the incidence of weight 2.5 kg or less in all infants who underwent cardiac surgery during the study period was 13.4% and was not different between patients who had SV (14.5%) and those who had biventricular (12.8%) cardiac anomalies [6]. The higher frequency of low weight in children with congenital heart disease is due to the higher incidence of poor intrauterine growth rate, prematurity, and presence of associated extracardiac congenital and genetic abnormalities [4–10]. Given the worse early and late outcomes in low weight children undergoing cardiac surgery, a great deal of controversy exists about the optimal management of low weight infants with biventricular cardiac malformations
and whether complete repair versus initial palliation should be offered [4–6, 10]. This issue is obviously irrelevant to our patient population as intracardiac repair is not an option and multistage palliation is the current mainstay of surgical management of SV patients. The important question remains whether low weight babies should have their first-stage palliative surgery (BTS, PAB, or Norwood) early on or if they should have this firststage palliative surgery deferred until they gain weight, with the hope of decreasing their postoperative complications and improving their survival. Nonetheless, the reported experience with this delayed intervention approach has not been favorable and emergence of problems such as ventilator dependency, necrotizing enterocolitis, renal dysfunction, sepsis, and neurologic complications has been of concern [4, 5, 10, 13]. Moreover, desired adequate weight gain is infrequently achieved leading to unnecessary prolonged delays in surgical treatment and development of those complications [4, 5, 13]. The SV physiology further complicates the issues, especially in patients with unrestrictive pulmonary blood flow who require an early intervention (PAB, Norwood, or Damus-Kay-Stansel þ shunt) to prevent pulmonary over-circulation and subsequent development of pulmonary vascular disease that might prevent the children from progressing toward final palliative stage, and have a negative effect on early and late outcomes [1, 2]. Advanced age at time of Norwood operation has been repeatedly demonstrated to be a risk factor complicating Fig 5. (A) Parametric model for late survival stratified by weight group 2.5 kg or less and greater than 2.5 kg. (B) Hazard function for mortality stratified by weight group 2.5 kg or less and greater than 2.5 kg.
CONGENITAL HEART
The effect of weight 2.5 kg or less (vs > 2.5 kg) on postoperative survival among different selected patient subgroups is depicted in Figure 6. Weight 2.5 kg or less was found to be associated with increased hazard of mortality in most subgroups of patients except for premature babies and those with SV types associated with the worst prognosis at our institution (ie, atrial isomerism and PA/IVS). In those high-risk patients, it seems that the effect of low weight on survival is neutralized.
CONGENITAL HEART
674
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
Ann Thorac Surg 2015;99:668–76
Fig 6. Association between weight 2.5 kg or less (vs > 2.5 Kg) and postoperative survival among selected patient subgroups. Reported is the hazard ratio (HR) for mortality (with 95% confidence interval) for weight 2.5 kg or less in each patient subgroup. Weight 2.5 kg or less was found to be associated with increased hazard of mortality in most subgroups of patients. (Atrial isomerism in heterotaxy [HTX]; BT ¼ modified Blalock-Taussig; CPB ¼ cardiopulmonary bypass; DILV ¼ double inlet left ventricle; HLHS ¼ hypoplastic left heart syndrome; PAB ¼ pulmonary artery band; PA/IVS ¼ pulmonary atresia with intact ventricular septum; TA ¼ tricuspid atresia.)
postoperative recovery and affecting early and late results due to elevated pulmonary vascular resistance and prolonged volume loading effects on the systemic ventricle prior to palliation [14, 15]. Therefore, we believe that firststage palliation should not be delayed in low weight neonates with SV abnormalities except in those with major comorbidities such as intracranial hemorrhage, severe pulmonary infection, necrotizing enterocolitis, sepsis, significant extracardiac malformation requiring immediate intervention, or multiorgan failure related to shock at the time of initial presentation. Nonetheless, the completion of those palliative procedures in low weight infants is challenging due to technical difficulties related to small-sized cardiac structures and tissue friability, and postoperative management issues related to prematurity, pulmonary dysfunction, and pulmonary hypertension. Low weight has persistently been demonstrated as a risk factor for mortality after the Norwood operation, including in recent series from experienced institutions [11, 12, 14]. Similarly, low weight has been associated with increased mortality risk after BTS due to shunt occlusion related to the small sized shunts or systemic steal due to excessive pulmonary blood flow that is likely more common in smaller babies [16, 17]. In a recent review of the Society of Thoracic Surgeons database, weight less than 3 kg at surgery was
an independent risk factor for hospital mortality after BTS [16]. Finally, PAB in low weight and premature patients is challenging due to the presence of lung disease, elevated pulmonary vascular resistance, and future growth of the baby, which may affect the need for reintervention after PAB [18]. Alternative treatment strategies such as ductal stenting for SV patients with restricted pulmonary blood flow or hybrid palliation (ductal stenting, bilateral branch pulmonary artery banding, and atrial septectomy) in patients with excessive pulmonary blood flow and systemic outflow tract obstruction, have been proposed by some groups with promising results and might prove to be advantageous in high-risk patients including those with low weight [19–22]. In addition to early death, low weight at time of firststage surgery was suggested to have an adverse effect on late survival [3, 11, 12]. Gelehrter and colleagues [11] reviewed outcomes of 47 neonates with HLHS patients who were 2.5 kg or less at time of Norwood palliation. Fifty-one percent of patients survived to initial hospital discharge and overall survival from birth through Fontan was 36%. In their series, they had low interstage mortality and their operative survival after Glenn and Fontan was 91% and 94%, respectively. Pizarro and colleagues [12] studied outcomes of 20 neonates with HLHS and other SV
Ann Thorac Surg 2015;99:668–76
References 1. Tweddell JS, Litwin SB, Thomas JP Jr, Mussatto K. Recent advances in the surgical management of the single ventricle pediatric patient. Pediatr Clin North Am 1999;46:465–80.
675
2. Alsoufi B. Management of the single ventricle and potentially obstructive systemic ventricular outflow tract. J Saudi Heart Assoc 2013;25:191–202. 3. Alsoufi B, Manlhiot C, Awan A, et al. Current outcomes of the Glenn bidirectional cavopulmonary connection for single ventricle palliation. Eur J Cardiothorac Surg 2012;42:42–8. 4. Reddy VM. Low birth weight and very low birth weight neonates with congenital heart disease: timing of surgery, reasons for delaying or not delaying surgery. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu 2013;16:13–20. 5. Chang AC, Hanley FH, Lock JE, Castaneda AR, Wessel DL. Management and outcome of low birth weight neonates with congenital cardiac disease. J Pediatr 1994;124:461–6. 6. Alsoufi B, Manlhiot C, Mahle WT, et al. Low weight infants are at increased mortality risk following palliative or corrective cardiac surgery. J Thoracic Cardiovasc Surg 2014; pii: S0022-5223(14)01019-8. http://dx.doi.org/10.1016/j.jtcvs. 2014.07.047. [Epub ahead of print]. 7. Curzon CL, Milford-Beland S, Li JS, et al. Cardiac surgery in infants with low birth weight is associated with increased mortality: analysis of the Society of Thoracic Surgeons Congenital Heart database. J Thoracic Cardiovasc Surg 2008;135:546–51. 8. Reddy VM, McElhinney DB, Sagrado T, Parry AJ, Teitel DF, Hanley FL. Results of 102 cases of complete repair of congenital heart defects in patients weighing 700 to 2500 grams. J Thorac Cardiovasc Surg 1999;117:324–31. 9. Oppido G, Napoleone CP, Formigari R, et al. Outcome of cardiac surgery in low birth weight and premature infants. Eur J Cardiothorac Surg 2004;26:44–53. 10. Hickey EJ, Nosikova Y, Zhang H, et al. Very low-birth-weight infants with congenital cardiac lesions: is there merit in delaying intervention to permit growth and maturation? J Thorac Cardiovasc Surg 2012;143:126–36. 11. Gelehrter S, Fifer CG, Armstrong A, Hirsch J, Gajarski R. Outcomes of hypoplastic left heart syndrome in low-birthweight patients. Pediatr Cardiol 2011;32:1175–81. 12. Pizarro C, Davis DA, Galantowicz ME, Munro H, Gidding SS, Norwood WI. Stage I palliation for hypoplastic left heart syndrome in low birth weight neonates: can we justify it? Eur J Cardiothorac Surg 2002;21:716–20. 13. Wernovsky G, Rubenstein SD, Spray TL. Cardiac surgery in the low birth weight neonate: new approaches. Clin Perinatol 2001;28:249–64. 14. Mahle WT, Spray TL, Wernovsky G, Gaynor JW, Clark BJ. Survival after reconstructive surgery for hypoplastic left heart syndrome. Circulation 2000;102(19 Suppl 3):III136–41. 15. Alsoufi B, Manlhiot C, Al-Ahmadi M, et al. Older children at the time of the Norwood operation have ongoing mortality vulnerability that continues after cavopulmonary connection. J Thorac Cardiovasc Surg 2011;142:142-7. 16. Petrucci O, O’Brien SM, Jacobs ML, Jacobs JP, Manning PB, Eghtesady P. Risk factors for mortality and morbidity after the neonatal Blalock-Taussig shunt procedure. Ann Thorac Surg 2011;92:642–51. 17. Myers JW, Ghanayem NS, Cao Y, et al. Outcomes of systemic to pulmonary artery shunts in patients weighing less than 3 kg: analysis of shunt type, size, and surgical approach. J Thorac Cardiovasc Surg 2014;147:672–7. 18. Alsoufi B, Manlhiot C, Ehrlich A, et al. Results of palliation with an initial pulmonary artery band in patients with single ventricle associated with unrestricted pulmonary blood flow. J Thorac Cardiovasc Surg 2014; pii: S0022-5223(14)01077-0. http://dx.doi.org/10.1016/j.jtcvs.2014.08.007. [Epub ahead of print]. 19. McMullan DM, Permut LC, Jones TK, Johnston TA, Rubio AE. Modified Blalock-Taussig shunt versus ductal stenting for palliation of cardiac lesions with inadequate pulmonary blood flow. J Thorac Cardiovasc Surg 2014;147:397–401. 20. Galantowicz M, Cheatham JP, Phillips A, et al. Hybrid approach for hypoplastic left heart syndrome: intermediate results after the learning curve. Ann Thorac Surg 2008;85: 2063–70.
CONGENITAL HEART
variants who were 2.5 kg or less at time of Norwood palliation. In their series, 55% survived to initial hospital discharge and on follow-up they had 1 interstage mortality, while the remaining patients progressed to Glenn operation. In our current study, we have demonstrated that low weight was not only associated with higher hospital mortality but also with a higher interstage mortality and subsequently lower percentage of patients reaching the Glenn operation. After Glenn, the progression toward Fontan operation was similar between the 2 weight groups suggesting that the effect of low weight at time of initial palliation was neutralized once the balanced circulation was achieved with the Glenn operation. Similarly, the death hazard in low weight patients was more pronounced and more prolonged for almost a year after first-stage palliation before normalizing, suggesting that those low weight babies might require a more vigilant monitoring and care after hospital survival. Similar to other reports in the literature, the SV morphologies that were associated with the worst prognosis in our current series were atrial isomerism in heterotaxy patients and PA/IVS. The higher risk of early and late mortality in heterotaxy patients is due to the high incidence of associated intracardiac defects, particularly total anomalous pulmonary venous connection, atrioventricular valve regurgitation, and conduction abnormalities in addition to extracardiac anomalies affecting outcomes beyond hospital discharge [23–25]. In patients with PA/ IVS, the increased mortality is attributed to the higher prevalence of ventricular to coronary arteries sinusoids and right ventricle dependant coronary circulation with subsequent tenuous physiologic state and higher risk of ischemia and cardiac arrest prior to or after palliation [16, 26, 27]. While the effect of low weight on mortality was evident in different patient categories in our series, that effect was less definite than in other SV anomalies associated with better prognosis, including HLHS. It is likely that low weight complicates surgery and recovery after palliative surgery in those patients, but the effect of low weight is blunted in the presence of other physiologic and anatomic risk factors affecting mortality in those complex patients. Despite recent advances in the perioperative care of SV patients undergoing multistage palliation, low weight at time of first-stage surgery continues to adversely affect hospital survival after all types of SV palliative surgeries. Additionally, the effect of low weight on outcome extends beyond hospital discharge with significant increase in interstage mortality, decrease proportion of patients progressing toward the Glenn operation, and persistently elevated death hazard up to a year after initial palliation. Consequently, methods to improve outcomes in this high-risk SV population should address perioperative management, outpatient surveillance, and follow-up care.
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
CONGENITAL HEART
676
ALSOUFI ET AL SV PALLIATION IN LOW WEIGHT PATIENTS
21. Guleserian KJ, Barker GM, Sharma MS, et al. Bilateral pulmonary artery banding for resuscitation in high-risk, singleventricle neonates and infants: a single-center experience. J Thorac Cardiovasc Surg 2013;145:206–13. 22. Venugopal PS, Luna KP, Anderson DR, et al. Hybrid procedure as an alternative to surgical palliation of high-risk infants with hypoplastic left heart syndrome and its variants. J Thorac Cardiovasc Surg 2010;139:1211–5. 23. Jacobs JP, Pasquali SK, Morales DL, et al. Heterotaxy: lessons learned about patterns of practice and outcomes from the congenital heart surgery database of the society of thoracic surgeons. World J Pediatr Congenit Heart Surg 2011;2: 278–86. 24. Hancock Friesen CL, Zurakowski D, Thiagarajan RR, et al. Total anomalous pulmonary venous connection: an analysis
Ann Thorac Surg 2015;99:668–76
of current management strategies in a single institution. Ann Thorac Surg 2005;79:596–606. 25. Lodge AJ, Rychik J, Nicolson SC, Ittenbach RF, Spray TL, Gaynor JW. Improving outcomes in functional single ventricle and total anomalous pulmonary venous connection. Ann Thorac Surg 2004;78:1688–95. 26. Ashburn DA, Blackstone EH, Wells WJ, et al. Determinants of mortality and type of repair in neonates with pulmonary atresia and intact ventricular septum. J Thorac Cardiovasc Surg 2004;127:1000–7. 27. Guleserian KJ, Armsby LB, Thiagarajan RR, del Nido PJ, Mayer JE Jr. Natural history of pulmonary atresia with intact ventricular septum and right-ventricle-dependent coronary circulation managed by the single-ventricle approach. Ann Thorac Surg 2006;81:2250–7.
Thoracic Surgery Foundation for Research and Education: For 25 Years, Turning Today’s Research Into Tomorrow’s Patient Care Our patients don’t follow the details of our research. They don’t discuss unexpected breakthroughs or technical setbacks. They are not always aware of how changes in health care policies impact research funding and lab time. Nonetheless, the advances we make in cardiothoracic surgery touch each and every one of them. New surgical techniques and potent new drugs improve patient health and extend patient lives. TSFRE is a pivotal force for growth and vitality in cardiothoracic surgery, especially for research and academic career development. TSFRE was first established in 1988 as a 501(c)(3) not-for-profit charitable foundation with support from the four major thoracic surgery organizations—the American Association for Thoracic Surgery, The Society of Thoracic Surgeons, the Southern Thoracic Surgical Association, and the Western Thoracic Surgical Association. On October 1, 2014, TSFRE became an official charitable arm of The Society of Thoracic Surgeons. The foundation represents thoracic surgery in the United States and its research and educational initiatives support the broad spectrum of thoracic surgery. The mission of TSFRE is to foster the development of surgeon scientists in cardiothoracic surgery; increasing knowledge and innovation to benefit patient care.
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
Funding basic biomedical research to expand our understanding of thoracic surgery is vitally important. Still, it’s only half the battle. The results of such research must be successfully transferred to patient care. There is a compelling need to continue to develop the skills of cardiothoracic surgeons as scientists and as health policy leaders. TSFRE is an organization that is having a steering effect on cardiothoracic surgery through the sponsorship of research projects and the education of surgeons in health care policy. In its 25-year history, TSFRE support has enabled many young cardiothoracic surgeons to develop successful, independent research careers and improve the care of our patients. To date, TSFRE has funded 146 cardiothoracic surgery research grants, fellowships, and simulation awards, and funded more than 260 Alley-Sheridan Scholarships. Because of your support, many of these individuals are now leaders in the cardiothoracic surgery community. This is a TSFRE legacy in which we can all take pride. Please help TSFRE continue its legacy by making a donation today. Visit www.tsfre.org or contact Priscilla Kennedy, TSFRE Executive Director, at 312-2025868, or by e-mail at [email protected].
Ann Thorac Surg 2015;99:676
0003-4975/$36.00