- Email: [email protected]
Current Results of Single Ventricle Palliation of Patients With Double Inlet Left Ventricle
Current Results of Single Ventricle Palliation of Patients With Double Inlet Left Ventricle Bahaaldin Alsoufi, MD, Courtney McCracken, PhD, Kirk Kanter, MD, Subhadra Shashidharan, MD, and Brian Kogon, MD CONGENITAL HEART
Division of Cardiothoracic Surgery, Children’s Healthcare of Atlanta, Emory University School of Medicine, Atlanta, Georgia
Background. Double inlet left ventricle (DILV) is a heterogeneous single ventricle anomaly in which initial presentation, and consequently, timing and palliation mode vary based on morphology and degree of pulmonary or systemic outflow obstruction. Very few reports, mostly old, focused on palliation outcomes of DILV. We report current-era results and examine whether morphologic and subsequently surgical factors influence survival. Methods. Fifty-eight infants with DILV underwent single ventricle palliation. Echocardiographic examination showed pulmonary (n [ 29, 50%), systemic outflow tract (n [ 11, 19%), and arch (n [ 17, 29%) obstruction. Factors associated with death or transplantation were examined. Results. Forty-four patients (76%) required neonatal first-stage palliation: modified Blalock-Taussig shunt (n [ 15, 26%), Norwood (n [ 15, 26%), or pulmonary artery band (n [ 14, 24%), whereas 14 (24%) received primary Glenn. There was 1 hospital death (2%) and 2 interstage deaths before Glenn, in addition to 1 late death
that was noncardiac. Overall 10-year survival was 94% and was comparable for different palliative surgeries (p [ 0.49). Three patients (6%) underwent heart transplantation after first-stage palliation (n [ 1) or after Glenn (n [ 2) for ventricular noncompaction (n [ 1), ventricular and atrioventricular valve dysfunction (n [ 1), and pacemaker-induced cardiomyopathy (n [ 1). Overall 10-year freedom from death or transplantation was 87% and was comparable for different palliative surgeries (p [ 0.58). On regression risk analysis, none of the tested morphologic or surgical variables was associated with the risk of death or transplantation. Conclusions. Current outcomes of multistage palliation of DILV are relatively good compared with published reports of other single ventricle anomalies. Survival is not greatly affected by cardiac morphology or initial palliative surgery type.
D
pulmonary blood flow require a modified Blalock-Taussig shunt; patients with unrestrictive pulmonary blood flow require a pulmonary artery band; and patients with systemic outflow tract obstruction or arch obstruction require either a modified Norwood operation, palliative arterial switch, or arch repair plus concomitant pulmonary artery band [4–8]. Finally, a number of patients have a balanced circulation that allows deferment of surgery for several months until a primary Glenn shunt can be performed [9, 10]. Double inlet left ventricle is one of the less common single ventricle anomalies and accounts for nearly 5% of those anomalies in published reports in the literature [11, 12]. Although the recent advances in all management aspects of single ventricle patients have led to remarkable improvements in early and late palliation results, most available reports have focused on the more common single ventricle anomalies and very few reports, mostly old, have emphasized palliation results in patients with DILV [12, 13]. We aim in the current series to describe the morphologic and clinical spectrum of patients with DILV and examine the effects of anatomic and subsequently surgical factors on palliation outcomes.
ouble inlet left ventricle (DILV) is a single ventricle anomaly that is characterized by the connection of both atria to the left ventricle through two separate atrioventricular valves. The dominant left ventricle connects to the rudimentary right ventricle through a bulboventricular foramen [1–3]. Double inlet left ventricle is a heterogeneous anomaly based on ventricular looping (D loop or L loop), ventriculoarterial discordance, size of bulboventricular foramen, and semilunar valve hypoplasia or stenosis [1–3]. Consequently, the clinical picture of DILV might be one of single ventricle with restrictive pulmonary blood flow, unrestrictive pulmonary blood flow with or without systemic outflow tract obstruction, or balanced circulation. As a result, timing and mode of initial palliative surgery is related to morphology and associated physiology. Patients with restrictive
Accepted for publication April 12, 2017. Presented at the Sixty-third Annual Meeting of the Southern Thoracic Surgical Association, Naples, FL, Nov 9–12, 2016. 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; email: balsoufi@hotmail. com.
Ó 2017 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2017;104:2064–71) Ó 2017 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2017.04.031
Ann Thorac Surg 2017;104:2064–71
Patients and Methods Inclusion Criteria
Follow-Up Time-related outcomes were determined from recent office visits documented in the electronic chart of the Children’s Healthcare of Atlanta system or from direct correspondence with pediatric cardiologists outside the system. Mean follow-up duration was 7.9 4.0 years and was 96% complete.
Statistical Analysis Data are presented as mean with standard deviation, median with interquartile range (IQR), or frequencies and percentages, as appropriate. Time-dependent outcomes after first-stage palliation surgery and after Glenn were summarized using Kaplan-Meier survival curves and included overall survival and freedom from death or heart transplantation. For the Kaplan-Meier survival curves, equal-band 95% confidence intervals are provided. Risk factors associated with time-dependent outcomes were determined using Cox proportional hazards regression, and hazard ratio with associated 95% confidence interval are provided. Given the small number of events, multivariable analyses were not considered. Statistical analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC), and p less than 0.05 was considered statistically significant.
2065
remaining 7 (12%) had ventriculoarterial concordance. In patients with ventriculoarterial discordance, the ventricles were D-looped in 23 (40%) and L-looped in 28 (48%; Table 1). Forty-four patients (76%) required neonatal first-stage palliation including modified Blalock-Taussig shunt (n ¼ 15, 26%), Norwood (n ¼ 15, 26%; the source of pulmonary blood flow was aortopulmonary shunt in 12, left ventricle to pulmonary artery conduit in 3), and pulmonary artery band (n ¼ 14, 24%, with 4 of 14 requiring concomitant coarctation repair). In the remaining 14 patients (24%) who had a balanced circulation, primary Glenn bidirectional cavopulmonary shunt was the initial palliative surgery. Concomitant surgery at time of firststage palliation was performed in 11 of 44 patients (25%) and included pulmonary artery augmentation (n ¼ 6), aortic coarctation repair (n ¼ 4), and enlargement of the bulboventricular foramen (n ¼ 1).
Outcomes After First-Stage Palliation After first-stage palliation in 44 patients, there was 1 hospital death (2%). That patient underwent the Norwood operation and had postoperative cardiac arrest requiring extracorporeal membrane oxygenation support owing to a clotted shunt, underwent several shunt revisions, and died of multiorgan failure. Subsequent to hospital discharge, there were two interstage deaths (5%): the one was a patient who underwent the Norwood operation and had late heart block requiring permanent pacemaker implantation and associated ventricular dysfunction; and the other was a patient who had a modified BlalockTaussig shunt and had sudden death at home, presumably shunt related. In addition, 1 patient (2%) who underwent initial pulmonary artery band had progressive ventricular dysfunction due to noncompaction and required subsequent shunt and eventually underwent heart transplantation. The remaining patients (n ¼ 40, 91%) progressed to receive Glenn shunt (Fig 1).
Outcomes After Glenn Shunt Results Patient Characteristics, Morphologic, and Operative Details Fifty-eight infants with DILV underwent their initial palliation surgery. There were 30 males (52%). Median age at surgery was 12 days (IQR: 4 to 90) and median weight was 3.4 kg (IQR: 3.0 to 5.0), with 7 patients (12%) weighing 2.5 kg or less. There were 12 patients (21%) who were born prematurely at 36 weeks or less gestation. Only 3 patients (5%) had associated extracardiac anomalies (sickle cell, XYY chromosomal anomaly, and jejunal atresia; n ¼ 1 each). Echocardiographic examination of those patients showed the following: antegrade pulmonary blood flow was absent in 6 (10%), restricted in 23 (40%), and unrestricted in 29 (50%). Aortic arch obstruction was present in 17 patients (29%), and systemic outflow tract obstruction was present in 11 patients (19%). Ventriculoarterial discordance was present in 51 patients (88%), and the
The Glenn shunt was performed in 54 patients, either subsequent to first-stage palliation (n ¼ 40) or as a primary procedure in patients who had a balanced circulation with mild restriction to their pulmonary blood flow (n ¼ 14). Concomitant surgery at time of Glenn shunt (n ¼ 19) was performed in 14 of 54 patients (26%) and included pulmonary artery augmentation (n ¼ 6), Damus-KayeStansel connection (n ¼ 6), atrioventricular valve repair (n ¼ 2), permanent pacemaker implantation (n ¼ 2), residual arch obstruction repair (n ¼ 1), partial anomalous pulmonary venous connection repair (n ¼ 1), and enlargement of the bulboventricular foramen (n ¼ 1). After the Glenn shunt, there was no operative mortality. One patient (2%) was not considered to be a candidate for subsequent Fontan due to poor ventricular function and several cerebrovascular accidents and died at 11 years of age; 2 patients (4%) received heart transplants, 1 for progressive ventricular dysfunction and atrioventricular valve regurgitation requiring repair at time of Glenn, and 1 for ventricular dysfunction due to pacemaker-
CONGENITAL HEART
Between 2002 and 2015, 58 consecutive infants with DILV underwent their first palliative surgery at Children’s Healthcare of Atlanta, Emory University. Patients were identified using our institutional surgical database. Demographic, morphologic, clinical, operative, and hospital details were abstracted from the 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.
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
2066
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
Ann Thorac Surg 2017;104:2064–71
Table 1. Characteristics and Anatomic Details of Patient Cohort
CONGENITAL HEART
Characteristics
Overall
Weight, kg Weight 2.5 kg Prematurity Female Genetic syndrome or extracardiac anomalies Aortic arch obstruction Systemic outflow obstruction Pulmonary stenosis VA relationship Concordance Discordance, D loop Discordance, L loop
3.4 7 11 28 3 17 11 25
(3.0–4.6) (12) (19) (48) (5) (29) (19) (43)
7 (12) 23 (40) 28 (48)
Glenn (n ¼ 14) 6.5 0 2 7 0 0 0 10
(5.5–7.0) (0) (15) (50) (0) (0) (0) (71)
3 (21) 5 (36) 6 (43)
BT Shunt (n ¼ 15) 2.7 4 5 8 2 0 0 15
(2.5–3.5) (27) (33) (53) (13) (0) (0) (100)
2 (13) 6 (40) 7 (47)
PA Band (n ¼ 14) 3.4 0 2 6 1 2 0 0
(3.3–3.7) (0) (14) (43) (7) (14) (0) (0)
2 (14) 5 (36) 7 (50)
Norwood (n ¼ 15) 3.3 3 2 7 0 15 11 0
(2.8–3.9) (20) (13) (47) (0) (100) (73) (0)
0 (0) 7 (47) 8 (53)
Values are median (interquartile range) or n (%). BT ¼ Blalock-Taussig;
PA ¼ pulmonary artery;
VA ¼ ventriculoarterial.
induced cardiomyopathy (required pacemaker implantation for heart block complicating enlargement of the bulboventricular foramen at time of primary Glenn). The remaining patients either progressed to receive the Fontan operation (n ¼ 35, 65%) or are alive and considered proper candidates to receive future Fontan operation (n ¼ 16, 30%). Of note, concomitant surgery at time of Fontan operation (n ¼ 13) was performed in 11 of 35 patients (31%) and included pulmonary artery augmentation (n ¼ 7), permanent pacemaker implantation (n ¼ 3), Fig 1. Progression of events after initial palliative surgery in patients with double inlet left ventricle (DILV).
atrioventricular valve repair (n ¼ 1), Maze operation (n ¼ 1), and enlargement of the bulboventricular foramen (n ¼ 1; Fig 1). Subsequent to the Fontan operation, none of the patients in our series died or received heart transplants.
Overall Survival and Risk Factors Overall survival after the initial palliative surgery for the entire cohort was 96% and 94% at 1 and 10 years, respectively (Fig 2). The initial palliative surgery type did not significantly affect 10-year survival: 100% after band,
Ann Thorac Surg 2017;104:2064–71
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
2067
Fig 3. Overall freedom from death or heart transplantation after initial palliative surgery in patients with double inlet left ventricle. Blue-shaded area represents 95% equal precision band.
100% after primary Glenn, 88% after shunt, and 86% after Norwood (p ¼ 0.4939). Univariable analysis that examined multiple patient characteristics, anatomic, and surgical details was performed and is shown in Table 2. The only factor associated with mortality on univariable analysis was weight 2.5 kg or less at time of surgery (hazard ratio 20.1, 95% confidence interval: 1.8 to 227.9, p ¼ 0.016). Overall freedom from death or heart transplantation after the initial palliative surgery for the entire cohort was 94% and 87% at 1 and 10 years, respectively (Fig 3). The initial palliative surgery type did not significantly affect 10-year freedom from death or heart transplantation: 94% after band, 94% after primary Glenn, 88% after shunt, and 79% after Norwood (p ¼ 0.5770; Fig 4). Univariable
analysis that examined multiple patient characteristics, anatomic, and surgical details was performed and is shown in Table 3. The only factor associated with mortality or heart transplantation on univariable analysis was weight 2.5 kg or less at time of surgery (hazard ratio 10.3, 95% confidence interval: 2.1 to 51.7, p ¼ 0.005).
Table 2. Univariable Analysis of Overall Survival After Surgery of Infants With Double Inlet Left Ventricle Variables Weight 2.5 kg Initial palliative surgery Glenn BT shunt PA band Norwood (reference) Prematurity Female Extracardiac anomalies Aortic arch obstruction Systemic outflow obstruction Pulmonary obstruction VA concordance versus discordance D loop versus L loop ventricles a
HR
95% CI
p Value
20.05
1.77–227.85
0.016
0 0.42 0.44 . 1.73 3.24 4.94 8.05 6.70 0.42 0
Not estimateda 0.04–4.63 0.04–4.93 . 0.18–17.02 0.34–31.17 0.51–48.17 0.84–77.56 0.91–49.46 0.04–4.00 Not estimateda
0.996 0.477 0.509 . 0.637 0.309 0.169 0.071 0.062 0.448 1.00
0.74
0.10–5.31
Comment Despite the diversity in cardiac morphology, clinical presentation, and subsequent initial surgery type, palliation outcomes for DILV were generally good and were comparable for the various anatomic subtypes or initial palliative procedures. This finding is likely restricted by the small cohort size and low incidence of late events (death or heart transplantation) that limited the power of our statistical analysis. Those events might have been a
0.768
No estimate was provided, given the lack of deaths in that group.
BT ¼ Blalock-Taussig; CI ¼ confidence interval; HR ¼ hazard ratio; PA ¼ pulmonary artery; VA ¼ ventriculoarterial.
Fig 4. Freedom from death or heart transplantation after initial palliative surgery in patients with double inlet left ventricle stratified by initial palliative surgery type: modified Blalock-Taussig (BT) shunt (blue line); pulmonary artery (PA) band (green line); Norwood procedure (red line); or primary Glenn procedure (brown line).
CONGENITAL HEART
Fig 2. Overall survival after initial palliative surgery in patients with double inlet left ventricle. Blue-shaded area represents 95% equal precision band.
2068
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
Table 3. Univariable Analysis of Death or Heart Transplantation After Surgery in Infants With Double Inlet Left Ventricle
CONGENITAL HEART
Variables Weight 2.5 kg Initial palliative surgery Glenn BT shunt PA band Norwood (reference) Neonatal palliation Prematurity Female Extracardiac anomalies Aortic arch obstruction Systemic outflow obstruction Pulmonary obstruction VA concordance versus discordance D loop versus L loop ventricles
HR
95% CI
p Value
10.32
2.06–51.71
0.005
0.30 0.29 0.63 . 2.07 1.93 0.82 4.94 3.50 4.43 0.50 0
0.03–2.89 0.03–2.78 0.11–3.80 . 0.25–17.24 0.37–10.03 0.18–3.58 0.51–48.17 0.78–15.65 0.97–20.17 0.10–2.58 Not estimateda
0.298 0.282 0.617 . 0.437 0.798 0.169 0.103 0.054 0.407 0.995
0.54
0.12–2.42
0.420
a
No estimate was provided given the lack of deaths or transplantations in that group.
BT ¼ Blalock-Taussig; CI ¼ confidence interval; HR ¼ hazard ratio; PA ¼ pulmonary artery; VA ¼ ventriculoarterial.
little more common in patients who have undergone the Norwood operation, which is naturally the most complicated procedure. The patients who needed the Norwood operation were those with ventriculoarterial discordance who had anatomic substrates for systemic outflow tract obstruction due to restricted bulboventricular foramen supplying the blood to the rudimentary right ventricle, occasionally associated with aortic valve hypoplasia and commonly associated with aortic arch obstruction. The procedure choice for patients with DILV, ventriculoarterial discordance, and aortic arch obstruction is either a Norwood operation or coarctation repair plus pulmonary artery band [10, 11, 14]. At our institution, patients with restrictive bulboventricular foramen, small aortic annulus, ascending aorta, or proximal arch diameters were more likely to undergo a Norwood operation whereas patients with aortic arch obstruction limited to coarctation and distal arch hypoplasia without significant restriction at the level of the bulboventricular foramen or significant hypoplasia of the aortic annulus, ascending aorta, or proximal aortic arch were considered for coarctation repair plus pulmonary artery band. In our current series, 15 patients with DILV underwent the Norwood operation, and only 4 received coarctation repair plus pulmonary artery band. That might have been related to the anatomic concerns about systemic outflow tract obstruction but it might also have been related to our institutional bias. Review of our institutional experience indicates, however, that coarctation repair plus pulmonary artery band strategy is a valid and potentially less complicated option for patients with DILV, unrestricted bulboventricular foramen, and arch obstruction
Ann Thorac Surg 2017;104:2064–71
that is limited to coarctation and distal arch hypoplasia [4, 14]. All those patients in our series underwent a Damus-Kaye-Stansel connection at time of Glenn shunt. Tricuspid atresia is a more common single ventricle anomaly that has been better studied in the literature. There are some similarities between DILV and tricuspid atresia that include the following: (1) left dominant ventricle morphology; (2) frequent ventriculoarterial discordance with the aorta originating from the rudimentary right ventricle; and (3) various clinical presentations that are related to the degree of associated pulmonary, systemic outflow, and arch obstruction [15]. The findings from the current series are in line with other studies of tricuspid atresia that have similarly shown that outcomes were comparable for various anatomic subtypes and palliation procedures, suggesting that the current advances in single ventricle management have mitigated the effects of surgical factors on palliation outcomes [15]. The good palliation outcomes of DILV in the current series are parallel to those in published reports from our center and other institutions. When compared with other single ventricle anomalies, DILV palliation outcomes seem to be superior after all palliative surgery types (shunt, band, Norwood, or Glenn) [4, 6, 8, 9, 11]. The superior outcomes in DILV patients might be related to the low incidence of genetic syndromes and extracardiac malformations, seen in only 5% of patients in our series. Genetic syndromes have been clearly linked to more complicated hospital course and higher operative mortality after neonatal cardiac surgery [16, 17]. In addition, the effect of genetic syndromes on survival after congenital cardiac surgery has been noted to extend beyond the operative period, with continuous attrition risk that persisted for at least 1 year after initial cardiac surgery [17]. We are currently studying the effects of genetic syndromes on palliation outcomes in neonates born with single ventricle anomalies. Our analysis indicates that genetic syndromes was the lowest among patients with DILV (compared with as high as 100% among patients with atrial isomerism and heterotaxy syndrome). We found that genetic syndromes and extracardiac malformations were associated with longer durations of mechanical ventilation, intensive care unit stay, and hospital stay after first-stage palliation, in addition to higher hospital mortality, lower progression to subsequent Glenn, and lower overall survival. Those inferior results are likely related to the morbidity associated with the extracardiac anomalies and the higher prevalence of additional risk factors such as low weight and prematurity [17]. The other potential explanation for the superior palliation outcomes among DILV patients is that those patients have left dominant ventricle morphology. There are several differences between the left and right heart structures involving the ventricles, atrioventricular valves, and the coronary supply that make the right ventricle and tricuspid valve less equipped to support the systemic circulation long term. The available reports from the literature vary, and although some have demonstrated
Ann Thorac Surg 2017;104:2064–71
Author Interview: The Author Interview can be viewed in the online version of this article [http://dx. doi.org/10.1016/j.athoracsur.2017.04.031] on http:// www.annalsthoracicsurgery.org.
References 1. Becker AE, Anderson RH. Double inlet ventricles. In: Pathology of congenital heart disease. London: Butterworths; 1981. 2. Jacobs ML, Anderson RH. Nomenclature of the functionally univentricular heart. Cardiol Young 2006;16(Suppl 1):3–8.
2069
3. Jacobs ML, Mayer JE. Congenital heart surgery nomenclature and database project: single ventricle. Ann Thorac Surg 2000;69(Suppl):197–204. 4. 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 2015;149:213–20. 5. Mosca RS, Hennein HA, Kulik TJ, et al. Modified Norwood operation for single left ventricle and ventriculoarterial discordance: an improved surgical technique. Ann Thorac Surg 1997;64:1126–32. 6. Alsoufi B, Slesnick T, McCracken C, et al. Current outcomes of the Norwood operation in patients with single-ventricle malformations other than hypoplastic left heart syndrome. World J Pediatr Congenit Heart Surg 2015;6:46–52. 7. Jacobs ML, Rychik J, Murphy JD, Nicolson SC, Steven JM, Norwood WI. Results of Norwood’s operation for lesions other than hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 1995;110:1555–61. 8. Alsoufi B, Gillespie S, Kogon B, et al. Results of palliation with an initial modified Blalock-Taussig shunt in neonates with single ventricle anomalies associated with restrictive pulmonary blood flow. Ann Thorac Surg 2015;99:1639–46. 9. 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. 10. Alsoufi B, Al-Wadai A, Khan M, et al. Outcomes of DamusKaye-Stansel anastomosis at time of cavopulmonary connection in single ventricle patients at risk of developing systemic ventricular outflow tract obstruction. Eur J Cardiothorac Surg 2014;45:77–82. 11. d’Udekem Y, Xu MY, Galati JC, et al. Predictors of survival after single-ventricle palliation: the impact of right ventricular dominance. J Am Coll Cardiol 2012;59:1178–85. 12. Alsoufi B, McCracken C, Ehrlich A, et al. Single ventricle palliation in low weight patients is associated with worse early and midterm outcomes. Ann Thorac Surg 2015;99: 668–76. 13. Tabbutt S, Ghanayem N, Ravishankar C, et al. Risk factors for hospital morbidity and mortality after the Norwood procedure: a report from the Pediatric Heart Network Single Ventricle Reconstruction trial. J Thorac Cardiovasc Surg 2012;144:882–95. 14. Alsoufi B, McCracken C, Shashidharan S, Kogon B, Border W, Kanter K. Palliation outcomes of neonates born with single-ventricle anomalies associated with aortic arch obstruction. Ann Thorac Surg 2017;103:637–44. 15. Alsoufi B, Schlosser B, Mori M, et al. Influence of morphology and initial surgical strategy on survival of infants with tricuspid atresia. Ann Thorac Surg 2015;99: 2117–22. 16. Patel A, Hickey E, Mavroudis C, et al. Impact of noncardiac congenital and genetic abnormalities on outcomes in hypoplastic left heart syndrome. Ann Thorac Surg 2010;89: 1805–13. 17. Alsoufi B, Gillespie S, Mahle W, et al. The impact of noncardiac and genetic abnormalities on outcomes following neonatal congenital heart surgery. Semin Thorac Cardiovasc Surg 2016;28:105–14. 18. Backer CL. The functionally univentricular heart: which is better—right or left ventricle? J Am Coll Cardiol 2012;59: 1186–7. 19. Julsrud PR, Weigel TJ, Van Son JA, et al. Influence of ventricular morphology on outcome after the Fontan procedure. Am J Cardiol 2000;86:319–23. 20. McGuirk SP, Winlaw DS, Langley SM, et al. The impact of ventricular morphology on midterm outcome following completion total cavopulmonary connection. Eur J Cardiothorac Surg 2003;24:37–46. 21. Alsoufi B, Gillespie S, Kim D, et al. The impact of dominant ventricle morphology on palliation outcomes of single ventricle anomalies. Ann Thorac Surg 2016;102:593–601.
CONGENITAL HEART
superior single ventricle palliation outcomes in patients with left dominant ventricle morphology, others have not shown significant difference at midterm follow-up [11, 18–23]. At our institution, we have demonstrated that patients’ characteristics and underlying single ventricle anomaly were more associated with outcomes than dominant ventricle morphology. For example, pulmonary atresia and intact ventricular septum was associated with the lowest survival among single ventricle anomalies despite those patients having left dominant ventricle [21]. Moreover, palliation outcomes of patients with atrial isomerism and heterotaxy syndrome were related more to associated cardiac findings than dominant ventricle morphology [24]. Therefore, we believe that, in the current era, the advances in single ventricle management strategies might have neutralized the effects of dominant ventricle morphology on early and midterm outcomes, but those effects might become evident with longer follow-up, similar to patients with other cardiac anomalies such as corrected transposition of the great arteries. Despite being one of the largest series examining palliation results in DILV patients, our analysis was limited by the series size and the low number of events. The only risk factor that was associated with death or heart transplantation on univariable analysis was weight 2.5 kg or less at time of initial surgery. Low weight is an established risk factor for single ventricle palliation and has been found to be associated not only with higher hospital death, but also with higher interstage mortality, decreased proportion of patients progressing toward the Glenn shunt, and continuously higher death hazard for as long as a year after initial palliation [25]. The persistent increased risk in low weight patients despite the management advances indicates that a more vigilant followup might be necessary to improve outcomes in this high-risk group of patients. In summary, current outcomes of multistage palliation of DILV are relatively good compared with those reported for other single ventricle anomalies. Possible reasons include the low association with extracardiac anomalies or genetic syndromes and left dominant ventricle morphology. Survival is not greatly affected by cardiac morphology, clinical presentation, or type of initial palliative surgery. This information is helpful for decision making and family discussions.
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
2070
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
22. Hirsch JC, Goldberg C, Bove EL, et al. Fontan operation in the current era. A 15-year single institution experience. Ann Surg 2008;248:402–10. 23. Tweddell JS, Nersesian M, Mussatto KA, et al. Fontan palliation in the modern era: factors impacting mortality and morbidity. Ann Thorac Surg 2009;88: 1291–9.
Ann Thorac Surg 2017;104:2064–71
24. Alsoufi B, Schlosser B, McCracken C, et al. Outcomes of multistage palliation of infants with functional single ventricle and heterotaxy syndrome. J Thorac Cardiovasc Surg 2016;151:1369–77. 25. Alsoufi B, Manlhiot C, Mahle WT, et al. Low-weight infants are at increased mortality risk after palliative or corrective cardiac surgery. J Thorac Cardiovasc Surg 2014;148:2508–14.
CONGENITAL HEART
DISCUSSION DR ROSS M. UNGERLEIDER (Corpus Christi, TX): Double inlet left ventricle (DILV) is, as you say, a form of single ventricle. So the first question I would have is, what was it that made you think to look at DILV as a standalone group? What was it about that group of patients that captured enough of your intrigue that you thought, well, let us just study them?
lowest association with genetic syndromes, about 5%, as compared with other single ventricle anomalies, for example, heterotaxy at 100%. The other possibility is the fact that the dominant ventricle in DILV is of left morphology; however, whether that affects outcomes in the short to mid term is not really proven.
DR ALSOUFI: Since I joined Emory a few years ago, I have performed few clinical studies that examined results of single ventricle palliation and the effect of some risk factors such as genetic syndromes on outcomes. That is when I noticed the superior survival in patients with DILV compared with those with other single ventricle anomalies. Given that my review of the literature did not identify recent studies focusing on DILV, I decided to collect more anatomic and surgical details and examine outcomes and associated factors specific to DILV patients.
DR UNGERLEIDER: Do you think that double inlet deserves its own chapter in textbooks?
DR UNGERLEIDER: Is The Society of Thoracic Surgeons (STS) Congenital Heart Surgery Database granular enough for you to compare your outcomes with DILV in the STS cohort? Is it possible for you to gather that kind of comparison? DR ALSOUFI: Naturally, the data are as good as the person entering them and there are always inherent limitations associated with the registry databases. However, the STS database is granular enough when it comes to the underlying single ventricle anomaly, so identification of patients with DILV is possible. DR UNGERLEIDER: So I guess what I am leading to as a question is, as an outcome of this, Bahaaldin, would you suggest to Dr Jacobs and his companions who work on the STS Congenital Heart Surgery Database how to add some granularity to the fields to track these entities going forward?
DR ALSOUFI: It depends how big your textbook is. DR UNGERLEIDER: I guess what I am saying is as we start getting granular, we start taking single ventricles and now we get into all these different types, at what point does this become an entity of enough distinction because of the outcomes or because of the approaches or because of something that deserves a standalone description? DR ALSOUFI: I believe that multiinstitutional studies can allow more sophisticated analysis to improve our understanding and knowledge of rare cardiac defects. There are always limitations related to data granularity in the registry databases and to followup, for example, in the STS database. I have previously discussed a project with Drs Jeff and Marshall Jacobs hoping to study Norwood outcomes for various single ventricle anomalies, and I hope that this project might see the light if we can figure out how to link STS data with other databases to allow follow-up beyond hospital discharge. DR UNGERLEIDER: One final question. The patients with the balanced circulation, I presume those are the ones with a moderate amount of pulmonary stenosis? DR ALSOUFI: Correct.
DR ALSOUFI: Again, the underlying single ventricle anomaly is already there. Adding more anatomic details to certain cardiac defects might be possible and might prove to be beneficial for future studies. I believe that the STS database is beginning to expand toward that direction and some additional anatomic data have been recently added. This is similar to the increased details that are already in place, for example, for genetic syndromes and extracardiac malformations. DR UNGERLEIDER: You mentioned in a study that you presented earlier today where non–DiGeorge genetic syndromes were a marker for poor outcome. And so it obviously begs another question, which is, do you think the good outcomes that you are having are related to that or is it related to double inlet on its own or a combination? DR ALSOUFI: It is hard to know. One possibility is the fact that there is little association between DILV and genetic syndromes. I have recently studied the effect of genetic syndromes on single ventricle palliation outcomes. What I found is that DILV has the
DR UNGERLEIDER: So when we see that, those are the ones we can begin to think could survive to the next stage of palliation without an intervention? DR ALSOUFI: Right. DR UNGERLEIDER: Thank you. DR JEFFREY JACOBS (St. Petersburg, FL): First of all, a great presentation. As usual, it is up to your typical standard. I just wanted to respond to the dialogue that Ross had with you about whether it is possible to replicate a study like this in the STS Congenital Heart Surgery Database. I think the majority of the fields and data elements are in the database, so from that point of view it is possible. The trick is the longitudinal follow-up, and follow-up in the STS database stops at hospital discharge in 30 days. It is not always easy to identify that a given patient who had stage I is the same patient who had stage II who is the same patient who had stage III. So the trick is, how do we transform
Ann Thorac Surg 2017;104:2064–71
our STS Congenital Heart Surgery Database into a platform for longitudinal follow-up. A lot of people are working on it but nobody has figured out how to do it.
2071
In your data, it looks to me like those with a Glenn or BlalockTaussig shunt or both had pulmonary outflow obstruction, presumably, and those with a pulmonary artery band or those with a pulmonary artery band and arch obstruction must have had systemic outflow obstruction. So if you made your categories larger in the ways I have suggested, then I wonder if some of these things that have marginal p values might in fact turn out to be significant. DR ALSOUFI: Thank you for your comments. In fact, that was exactly the purpose of our study and we wanted to examine the effect of anatomic and subsequently surgical variables on palliation outcomes. In my presentation, there were no representative figures for those anatomic variables as none of them was a significant factor on our analysis. However, we have specifically examined the effect of ventriculoarterial discordance, along with the presence of pulmonary, systemic, and arch obstruction on transplant-free survival. Those are demonstrated on the tables here, and as you can see, there might be a trend on univariate analysis with decreased transplant-free survival in patients with systemic outflow obstruction that did not reach statistical significance, likely owing to limitations in our study because of sample size and the small number of events.
CONGENITAL HEART
DR JOHN E. MAYER (Boston, MA): The way I thought about single ventricle patients like this is heavily influenced by where the aorta arises, which type of chamber. If it is arising from a morphologic right ventricle and the morphologic left ventricle is the dominant ventricle, then you have got single ventricle with transposition, which is a different group of patients than if you have patients with normally related great vessels. The second way that I thought about single ventricle patients is that most single ventricles have obstruction to either one circulation or the other: they either have obstruction to the pulmonary circulation or they have obstruction to the systemic circulation. Therefore, I wonder if in your analysis of your data you have looked at categories based on these two anatomic considerations? One example is single ventricle double inlet with normally related great vessels, single ventricle with transposition, so the aorta is arising from an infundibular chamber, or those with systemic outflow obstruction as opposed to pulmonary outflow obstruction. I think that this sort of categorization scheme would give us something that would be a little bit easier to understand and maybe would yield some more statistical significance.
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
Division of Cardiothoracic Surgery, Children’s Healthcare of Atlanta, Emory University School of Medicine, Atlanta, Georgia
Background. Double inlet left ventricle (DILV) is a heterogeneous single ventricle anomaly in which initial presentation, and consequently, timing and palliation mode vary based on morphology and degree of pulmonary or systemic outflow obstruction. Very few reports, mostly old, focused on palliation outcomes of DILV. We report current-era results and examine whether morphologic and subsequently surgical factors influence survival. Methods. Fifty-eight infants with DILV underwent single ventricle palliation. Echocardiographic examination showed pulmonary (n [ 29, 50%), systemic outflow tract (n [ 11, 19%), and arch (n [ 17, 29%) obstruction. Factors associated with death or transplantation were examined. Results. Forty-four patients (76%) required neonatal first-stage palliation: modified Blalock-Taussig shunt (n [ 15, 26%), Norwood (n [ 15, 26%), or pulmonary artery band (n [ 14, 24%), whereas 14 (24%) received primary Glenn. There was 1 hospital death (2%) and 2 interstage deaths before Glenn, in addition to 1 late death
that was noncardiac. Overall 10-year survival was 94% and was comparable for different palliative surgeries (p [ 0.49). Three patients (6%) underwent heart transplantation after first-stage palliation (n [ 1) or after Glenn (n [ 2) for ventricular noncompaction (n [ 1), ventricular and atrioventricular valve dysfunction (n [ 1), and pacemaker-induced cardiomyopathy (n [ 1). Overall 10-year freedom from death or transplantation was 87% and was comparable for different palliative surgeries (p [ 0.58). On regression risk analysis, none of the tested morphologic or surgical variables was associated with the risk of death or transplantation. Conclusions. Current outcomes of multistage palliation of DILV are relatively good compared with published reports of other single ventricle anomalies. Survival is not greatly affected by cardiac morphology or initial palliative surgery type.
D
pulmonary blood flow require a modified Blalock-Taussig shunt; patients with unrestrictive pulmonary blood flow require a pulmonary artery band; and patients with systemic outflow tract obstruction or arch obstruction require either a modified Norwood operation, palliative arterial switch, or arch repair plus concomitant pulmonary artery band [4–8]. Finally, a number of patients have a balanced circulation that allows deferment of surgery for several months until a primary Glenn shunt can be performed [9, 10]. Double inlet left ventricle is one of the less common single ventricle anomalies and accounts for nearly 5% of those anomalies in published reports in the literature [11, 12]. Although the recent advances in all management aspects of single ventricle patients have led to remarkable improvements in early and late palliation results, most available reports have focused on the more common single ventricle anomalies and very few reports, mostly old, have emphasized palliation results in patients with DILV [12, 13]. We aim in the current series to describe the morphologic and clinical spectrum of patients with DILV and examine the effects of anatomic and subsequently surgical factors on palliation outcomes.
ouble inlet left ventricle (DILV) is a single ventricle anomaly that is characterized by the connection of both atria to the left ventricle through two separate atrioventricular valves. The dominant left ventricle connects to the rudimentary right ventricle through a bulboventricular foramen [1–3]. Double inlet left ventricle is a heterogeneous anomaly based on ventricular looping (D loop or L loop), ventriculoarterial discordance, size of bulboventricular foramen, and semilunar valve hypoplasia or stenosis [1–3]. Consequently, the clinical picture of DILV might be one of single ventricle with restrictive pulmonary blood flow, unrestrictive pulmonary blood flow with or without systemic outflow tract obstruction, or balanced circulation. As a result, timing and mode of initial palliative surgery is related to morphology and associated physiology. Patients with restrictive
Accepted for publication April 12, 2017. Presented at the Sixty-third Annual Meeting of the Southern Thoracic Surgical Association, Naples, FL, Nov 9–12, 2016. 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; email: balsoufi@hotmail. com.
Ó 2017 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2017;104:2064–71) Ó 2017 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2017.04.031
Ann Thorac Surg 2017;104:2064–71
Patients and Methods Inclusion Criteria
Follow-Up Time-related outcomes were determined from recent office visits documented in the electronic chart of the Children’s Healthcare of Atlanta system or from direct correspondence with pediatric cardiologists outside the system. Mean follow-up duration was 7.9 4.0 years and was 96% complete.
Statistical Analysis Data are presented as mean with standard deviation, median with interquartile range (IQR), or frequencies and percentages, as appropriate. Time-dependent outcomes after first-stage palliation surgery and after Glenn were summarized using Kaplan-Meier survival curves and included overall survival and freedom from death or heart transplantation. For the Kaplan-Meier survival curves, equal-band 95% confidence intervals are provided. Risk factors associated with time-dependent outcomes were determined using Cox proportional hazards regression, and hazard ratio with associated 95% confidence interval are provided. Given the small number of events, multivariable analyses were not considered. Statistical analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC), and p less than 0.05 was considered statistically significant.
2065
remaining 7 (12%) had ventriculoarterial concordance. In patients with ventriculoarterial discordance, the ventricles were D-looped in 23 (40%) and L-looped in 28 (48%; Table 1). Forty-four patients (76%) required neonatal first-stage palliation including modified Blalock-Taussig shunt (n ¼ 15, 26%), Norwood (n ¼ 15, 26%; the source of pulmonary blood flow was aortopulmonary shunt in 12, left ventricle to pulmonary artery conduit in 3), and pulmonary artery band (n ¼ 14, 24%, with 4 of 14 requiring concomitant coarctation repair). In the remaining 14 patients (24%) who had a balanced circulation, primary Glenn bidirectional cavopulmonary shunt was the initial palliative surgery. Concomitant surgery at time of firststage palliation was performed in 11 of 44 patients (25%) and included pulmonary artery augmentation (n ¼ 6), aortic coarctation repair (n ¼ 4), and enlargement of the bulboventricular foramen (n ¼ 1).
Outcomes After First-Stage Palliation After first-stage palliation in 44 patients, there was 1 hospital death (2%). That patient underwent the Norwood operation and had postoperative cardiac arrest requiring extracorporeal membrane oxygenation support owing to a clotted shunt, underwent several shunt revisions, and died of multiorgan failure. Subsequent to hospital discharge, there were two interstage deaths (5%): the one was a patient who underwent the Norwood operation and had late heart block requiring permanent pacemaker implantation and associated ventricular dysfunction; and the other was a patient who had a modified BlalockTaussig shunt and had sudden death at home, presumably shunt related. In addition, 1 patient (2%) who underwent initial pulmonary artery band had progressive ventricular dysfunction due to noncompaction and required subsequent shunt and eventually underwent heart transplantation. The remaining patients (n ¼ 40, 91%) progressed to receive Glenn shunt (Fig 1).
Outcomes After Glenn Shunt Results Patient Characteristics, Morphologic, and Operative Details Fifty-eight infants with DILV underwent their initial palliation surgery. There were 30 males (52%). Median age at surgery was 12 days (IQR: 4 to 90) and median weight was 3.4 kg (IQR: 3.0 to 5.0), with 7 patients (12%) weighing 2.5 kg or less. There were 12 patients (21%) who were born prematurely at 36 weeks or less gestation. Only 3 patients (5%) had associated extracardiac anomalies (sickle cell, XYY chromosomal anomaly, and jejunal atresia; n ¼ 1 each). Echocardiographic examination of those patients showed the following: antegrade pulmonary blood flow was absent in 6 (10%), restricted in 23 (40%), and unrestricted in 29 (50%). Aortic arch obstruction was present in 17 patients (29%), and systemic outflow tract obstruction was present in 11 patients (19%). Ventriculoarterial discordance was present in 51 patients (88%), and the
The Glenn shunt was performed in 54 patients, either subsequent to first-stage palliation (n ¼ 40) or as a primary procedure in patients who had a balanced circulation with mild restriction to their pulmonary blood flow (n ¼ 14). Concomitant surgery at time of Glenn shunt (n ¼ 19) was performed in 14 of 54 patients (26%) and included pulmonary artery augmentation (n ¼ 6), Damus-KayeStansel connection (n ¼ 6), atrioventricular valve repair (n ¼ 2), permanent pacemaker implantation (n ¼ 2), residual arch obstruction repair (n ¼ 1), partial anomalous pulmonary venous connection repair (n ¼ 1), and enlargement of the bulboventricular foramen (n ¼ 1). After the Glenn shunt, there was no operative mortality. One patient (2%) was not considered to be a candidate for subsequent Fontan due to poor ventricular function and several cerebrovascular accidents and died at 11 years of age; 2 patients (4%) received heart transplants, 1 for progressive ventricular dysfunction and atrioventricular valve regurgitation requiring repair at time of Glenn, and 1 for ventricular dysfunction due to pacemaker-
CONGENITAL HEART
Between 2002 and 2015, 58 consecutive infants with DILV underwent their first palliative surgery at Children’s Healthcare of Atlanta, Emory University. Patients were identified using our institutional surgical database. Demographic, morphologic, clinical, operative, and hospital details were abstracted from the 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.
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
2066
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
Ann Thorac Surg 2017;104:2064–71
Table 1. Characteristics and Anatomic Details of Patient Cohort
CONGENITAL HEART
Characteristics
Overall
Weight, kg Weight 2.5 kg Prematurity Female Genetic syndrome or extracardiac anomalies Aortic arch obstruction Systemic outflow obstruction Pulmonary stenosis VA relationship Concordance Discordance, D loop Discordance, L loop
3.4 7 11 28 3 17 11 25
(3.0–4.6) (12) (19) (48) (5) (29) (19) (43)
7 (12) 23 (40) 28 (48)
Glenn (n ¼ 14) 6.5 0 2 7 0 0 0 10
(5.5–7.0) (0) (15) (50) (0) (0) (0) (71)
3 (21) 5 (36) 6 (43)
BT Shunt (n ¼ 15) 2.7 4 5 8 2 0 0 15
(2.5–3.5) (27) (33) (53) (13) (0) (0) (100)
2 (13) 6 (40) 7 (47)
PA Band (n ¼ 14) 3.4 0 2 6 1 2 0 0
(3.3–3.7) (0) (14) (43) (7) (14) (0) (0)
2 (14) 5 (36) 7 (50)
Norwood (n ¼ 15) 3.3 3 2 7 0 15 11 0
(2.8–3.9) (20) (13) (47) (0) (100) (73) (0)
0 (0) 7 (47) 8 (53)
Values are median (interquartile range) or n (%). BT ¼ Blalock-Taussig;
PA ¼ pulmonary artery;
VA ¼ ventriculoarterial.
induced cardiomyopathy (required pacemaker implantation for heart block complicating enlargement of the bulboventricular foramen at time of primary Glenn). The remaining patients either progressed to receive the Fontan operation (n ¼ 35, 65%) or are alive and considered proper candidates to receive future Fontan operation (n ¼ 16, 30%). Of note, concomitant surgery at time of Fontan operation (n ¼ 13) was performed in 11 of 35 patients (31%) and included pulmonary artery augmentation (n ¼ 7), permanent pacemaker implantation (n ¼ 3), Fig 1. Progression of events after initial palliative surgery in patients with double inlet left ventricle (DILV).
atrioventricular valve repair (n ¼ 1), Maze operation (n ¼ 1), and enlargement of the bulboventricular foramen (n ¼ 1; Fig 1). Subsequent to the Fontan operation, none of the patients in our series died or received heart transplants.
Overall Survival and Risk Factors Overall survival after the initial palliative surgery for the entire cohort was 96% and 94% at 1 and 10 years, respectively (Fig 2). The initial palliative surgery type did not significantly affect 10-year survival: 100% after band,
Ann Thorac Surg 2017;104:2064–71
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
2067
Fig 3. Overall freedom from death or heart transplantation after initial palliative surgery in patients with double inlet left ventricle. Blue-shaded area represents 95% equal precision band.
100% after primary Glenn, 88% after shunt, and 86% after Norwood (p ¼ 0.4939). Univariable analysis that examined multiple patient characteristics, anatomic, and surgical details was performed and is shown in Table 2. The only factor associated with mortality on univariable analysis was weight 2.5 kg or less at time of surgery (hazard ratio 20.1, 95% confidence interval: 1.8 to 227.9, p ¼ 0.016). Overall freedom from death or heart transplantation after the initial palliative surgery for the entire cohort was 94% and 87% at 1 and 10 years, respectively (Fig 3). The initial palliative surgery type did not significantly affect 10-year freedom from death or heart transplantation: 94% after band, 94% after primary Glenn, 88% after shunt, and 79% after Norwood (p ¼ 0.5770; Fig 4). Univariable
analysis that examined multiple patient characteristics, anatomic, and surgical details was performed and is shown in Table 3. The only factor associated with mortality or heart transplantation on univariable analysis was weight 2.5 kg or less at time of surgery (hazard ratio 10.3, 95% confidence interval: 2.1 to 51.7, p ¼ 0.005).
Table 2. Univariable Analysis of Overall Survival After Surgery of Infants With Double Inlet Left Ventricle Variables Weight 2.5 kg Initial palliative surgery Glenn BT shunt PA band Norwood (reference) Prematurity Female Extracardiac anomalies Aortic arch obstruction Systemic outflow obstruction Pulmonary obstruction VA concordance versus discordance D loop versus L loop ventricles a
HR
95% CI
p Value
20.05
1.77–227.85
0.016
0 0.42 0.44 . 1.73 3.24 4.94 8.05 6.70 0.42 0
Not estimateda 0.04–4.63 0.04–4.93 . 0.18–17.02 0.34–31.17 0.51–48.17 0.84–77.56 0.91–49.46 0.04–4.00 Not estimateda
0.996 0.477 0.509 . 0.637 0.309 0.169 0.071 0.062 0.448 1.00
0.74
0.10–5.31
Comment Despite the diversity in cardiac morphology, clinical presentation, and subsequent initial surgery type, palliation outcomes for DILV were generally good and were comparable for the various anatomic subtypes or initial palliative procedures. This finding is likely restricted by the small cohort size and low incidence of late events (death or heart transplantation) that limited the power of our statistical analysis. Those events might have been a
0.768
No estimate was provided, given the lack of deaths in that group.
BT ¼ Blalock-Taussig; CI ¼ confidence interval; HR ¼ hazard ratio; PA ¼ pulmonary artery; VA ¼ ventriculoarterial.
Fig 4. Freedom from death or heart transplantation after initial palliative surgery in patients with double inlet left ventricle stratified by initial palliative surgery type: modified Blalock-Taussig (BT) shunt (blue line); pulmonary artery (PA) band (green line); Norwood procedure (red line); or primary Glenn procedure (brown line).
CONGENITAL HEART
Fig 2. Overall survival after initial palliative surgery in patients with double inlet left ventricle. Blue-shaded area represents 95% equal precision band.
2068
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
Table 3. Univariable Analysis of Death or Heart Transplantation After Surgery in Infants With Double Inlet Left Ventricle
CONGENITAL HEART
Variables Weight 2.5 kg Initial palliative surgery Glenn BT shunt PA band Norwood (reference) Neonatal palliation Prematurity Female Extracardiac anomalies Aortic arch obstruction Systemic outflow obstruction Pulmonary obstruction VA concordance versus discordance D loop versus L loop ventricles
HR
95% CI
p Value
10.32
2.06–51.71
0.005
0.30 0.29 0.63 . 2.07 1.93 0.82 4.94 3.50 4.43 0.50 0
0.03–2.89 0.03–2.78 0.11–3.80 . 0.25–17.24 0.37–10.03 0.18–3.58 0.51–48.17 0.78–15.65 0.97–20.17 0.10–2.58 Not estimateda
0.298 0.282 0.617 . 0.437 0.798 0.169 0.103 0.054 0.407 0.995
0.54
0.12–2.42
0.420
a
No estimate was provided given the lack of deaths or transplantations in that group.
BT ¼ Blalock-Taussig; CI ¼ confidence interval; HR ¼ hazard ratio; PA ¼ pulmonary artery; VA ¼ ventriculoarterial.
little more common in patients who have undergone the Norwood operation, which is naturally the most complicated procedure. The patients who needed the Norwood operation were those with ventriculoarterial discordance who had anatomic substrates for systemic outflow tract obstruction due to restricted bulboventricular foramen supplying the blood to the rudimentary right ventricle, occasionally associated with aortic valve hypoplasia and commonly associated with aortic arch obstruction. The procedure choice for patients with DILV, ventriculoarterial discordance, and aortic arch obstruction is either a Norwood operation or coarctation repair plus pulmonary artery band [10, 11, 14]. At our institution, patients with restrictive bulboventricular foramen, small aortic annulus, ascending aorta, or proximal arch diameters were more likely to undergo a Norwood operation whereas patients with aortic arch obstruction limited to coarctation and distal arch hypoplasia without significant restriction at the level of the bulboventricular foramen or significant hypoplasia of the aortic annulus, ascending aorta, or proximal aortic arch were considered for coarctation repair plus pulmonary artery band. In our current series, 15 patients with DILV underwent the Norwood operation, and only 4 received coarctation repair plus pulmonary artery band. That might have been related to the anatomic concerns about systemic outflow tract obstruction but it might also have been related to our institutional bias. Review of our institutional experience indicates, however, that coarctation repair plus pulmonary artery band strategy is a valid and potentially less complicated option for patients with DILV, unrestricted bulboventricular foramen, and arch obstruction
Ann Thorac Surg 2017;104:2064–71
that is limited to coarctation and distal arch hypoplasia [4, 14]. All those patients in our series underwent a Damus-Kaye-Stansel connection at time of Glenn shunt. Tricuspid atresia is a more common single ventricle anomaly that has been better studied in the literature. There are some similarities between DILV and tricuspid atresia that include the following: (1) left dominant ventricle morphology; (2) frequent ventriculoarterial discordance with the aorta originating from the rudimentary right ventricle; and (3) various clinical presentations that are related to the degree of associated pulmonary, systemic outflow, and arch obstruction [15]. The findings from the current series are in line with other studies of tricuspid atresia that have similarly shown that outcomes were comparable for various anatomic subtypes and palliation procedures, suggesting that the current advances in single ventricle management have mitigated the effects of surgical factors on palliation outcomes [15]. The good palliation outcomes of DILV in the current series are parallel to those in published reports from our center and other institutions. When compared with other single ventricle anomalies, DILV palliation outcomes seem to be superior after all palliative surgery types (shunt, band, Norwood, or Glenn) [4, 6, 8, 9, 11]. The superior outcomes in DILV patients might be related to the low incidence of genetic syndromes and extracardiac malformations, seen in only 5% of patients in our series. Genetic syndromes have been clearly linked to more complicated hospital course and higher operative mortality after neonatal cardiac surgery [16, 17]. In addition, the effect of genetic syndromes on survival after congenital cardiac surgery has been noted to extend beyond the operative period, with continuous attrition risk that persisted for at least 1 year after initial cardiac surgery [17]. We are currently studying the effects of genetic syndromes on palliation outcomes in neonates born with single ventricle anomalies. Our analysis indicates that genetic syndromes was the lowest among patients with DILV (compared with as high as 100% among patients with atrial isomerism and heterotaxy syndrome). We found that genetic syndromes and extracardiac malformations were associated with longer durations of mechanical ventilation, intensive care unit stay, and hospital stay after first-stage palliation, in addition to higher hospital mortality, lower progression to subsequent Glenn, and lower overall survival. Those inferior results are likely related to the morbidity associated with the extracardiac anomalies and the higher prevalence of additional risk factors such as low weight and prematurity [17]. The other potential explanation for the superior palliation outcomes among DILV patients is that those patients have left dominant ventricle morphology. There are several differences between the left and right heart structures involving the ventricles, atrioventricular valves, and the coronary supply that make the right ventricle and tricuspid valve less equipped to support the systemic circulation long term. The available reports from the literature vary, and although some have demonstrated
Ann Thorac Surg 2017;104:2064–71
Author Interview: The Author Interview can be viewed in the online version of this article [http://dx. doi.org/10.1016/j.athoracsur.2017.04.031] on http:// www.annalsthoracicsurgery.org.
References 1. Becker AE, Anderson RH. Double inlet ventricles. In: Pathology of congenital heart disease. London: Butterworths; 1981. 2. Jacobs ML, Anderson RH. Nomenclature of the functionally univentricular heart. Cardiol Young 2006;16(Suppl 1):3–8.
2069
3. Jacobs ML, Mayer JE. Congenital heart surgery nomenclature and database project: single ventricle. Ann Thorac Surg 2000;69(Suppl):197–204. 4. 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 2015;149:213–20. 5. Mosca RS, Hennein HA, Kulik TJ, et al. Modified Norwood operation for single left ventricle and ventriculoarterial discordance: an improved surgical technique. Ann Thorac Surg 1997;64:1126–32. 6. Alsoufi B, Slesnick T, McCracken C, et al. Current outcomes of the Norwood operation in patients with single-ventricle malformations other than hypoplastic left heart syndrome. World J Pediatr Congenit Heart Surg 2015;6:46–52. 7. Jacobs ML, Rychik J, Murphy JD, Nicolson SC, Steven JM, Norwood WI. Results of Norwood’s operation for lesions other than hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 1995;110:1555–61. 8. Alsoufi B, Gillespie S, Kogon B, et al. Results of palliation with an initial modified Blalock-Taussig shunt in neonates with single ventricle anomalies associated with restrictive pulmonary blood flow. Ann Thorac Surg 2015;99:1639–46. 9. 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. 10. Alsoufi B, Al-Wadai A, Khan M, et al. Outcomes of DamusKaye-Stansel anastomosis at time of cavopulmonary connection in single ventricle patients at risk of developing systemic ventricular outflow tract obstruction. Eur J Cardiothorac Surg 2014;45:77–82. 11. d’Udekem Y, Xu MY, Galati JC, et al. Predictors of survival after single-ventricle palliation: the impact of right ventricular dominance. J Am Coll Cardiol 2012;59:1178–85. 12. Alsoufi B, McCracken C, Ehrlich A, et al. Single ventricle palliation in low weight patients is associated with worse early and midterm outcomes. Ann Thorac Surg 2015;99: 668–76. 13. Tabbutt S, Ghanayem N, Ravishankar C, et al. Risk factors for hospital morbidity and mortality after the Norwood procedure: a report from the Pediatric Heart Network Single Ventricle Reconstruction trial. J Thorac Cardiovasc Surg 2012;144:882–95. 14. Alsoufi B, McCracken C, Shashidharan S, Kogon B, Border W, Kanter K. Palliation outcomes of neonates born with single-ventricle anomalies associated with aortic arch obstruction. Ann Thorac Surg 2017;103:637–44. 15. Alsoufi B, Schlosser B, Mori M, et al. Influence of morphology and initial surgical strategy on survival of infants with tricuspid atresia. Ann Thorac Surg 2015;99: 2117–22. 16. Patel A, Hickey E, Mavroudis C, et al. Impact of noncardiac congenital and genetic abnormalities on outcomes in hypoplastic left heart syndrome. Ann Thorac Surg 2010;89: 1805–13. 17. Alsoufi B, Gillespie S, Mahle W, et al. The impact of noncardiac and genetic abnormalities on outcomes following neonatal congenital heart surgery. Semin Thorac Cardiovasc Surg 2016;28:105–14. 18. Backer CL. The functionally univentricular heart: which is better—right or left ventricle? J Am Coll Cardiol 2012;59: 1186–7. 19. Julsrud PR, Weigel TJ, Van Son JA, et al. Influence of ventricular morphology on outcome after the Fontan procedure. Am J Cardiol 2000;86:319–23. 20. McGuirk SP, Winlaw DS, Langley SM, et al. The impact of ventricular morphology on midterm outcome following completion total cavopulmonary connection. Eur J Cardiothorac Surg 2003;24:37–46. 21. Alsoufi B, Gillespie S, Kim D, et al. The impact of dominant ventricle morphology on palliation outcomes of single ventricle anomalies. Ann Thorac Surg 2016;102:593–601.
CONGENITAL HEART
superior single ventricle palliation outcomes in patients with left dominant ventricle morphology, others have not shown significant difference at midterm follow-up [11, 18–23]. At our institution, we have demonstrated that patients’ characteristics and underlying single ventricle anomaly were more associated with outcomes than dominant ventricle morphology. For example, pulmonary atresia and intact ventricular septum was associated with the lowest survival among single ventricle anomalies despite those patients having left dominant ventricle [21]. Moreover, palliation outcomes of patients with atrial isomerism and heterotaxy syndrome were related more to associated cardiac findings than dominant ventricle morphology [24]. Therefore, we believe that, in the current era, the advances in single ventricle management strategies might have neutralized the effects of dominant ventricle morphology on early and midterm outcomes, but those effects might become evident with longer follow-up, similar to patients with other cardiac anomalies such as corrected transposition of the great arteries. Despite being one of the largest series examining palliation results in DILV patients, our analysis was limited by the series size and the low number of events. The only risk factor that was associated with death or heart transplantation on univariable analysis was weight 2.5 kg or less at time of initial surgery. Low weight is an established risk factor for single ventricle palliation and has been found to be associated not only with higher hospital death, but also with higher interstage mortality, decreased proportion of patients progressing toward the Glenn shunt, and continuously higher death hazard for as long as a year after initial palliation [25]. The persistent increased risk in low weight patients despite the management advances indicates that a more vigilant followup might be necessary to improve outcomes in this high-risk group of patients. In summary, current outcomes of multistage palliation of DILV are relatively good compared with those reported for other single ventricle anomalies. Possible reasons include the low association with extracardiac anomalies or genetic syndromes and left dominant ventricle morphology. Survival is not greatly affected by cardiac morphology, clinical presentation, or type of initial palliative surgery. This information is helpful for decision making and family discussions.
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
2070
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS
22. Hirsch JC, Goldberg C, Bove EL, et al. Fontan operation in the current era. A 15-year single institution experience. Ann Surg 2008;248:402–10. 23. Tweddell JS, Nersesian M, Mussatto KA, et al. Fontan palliation in the modern era: factors impacting mortality and morbidity. Ann Thorac Surg 2009;88: 1291–9.
Ann Thorac Surg 2017;104:2064–71
24. Alsoufi B, Schlosser B, McCracken C, et al. Outcomes of multistage palliation of infants with functional single ventricle and heterotaxy syndrome. J Thorac Cardiovasc Surg 2016;151:1369–77. 25. Alsoufi B, Manlhiot C, Mahle WT, et al. Low-weight infants are at increased mortality risk after palliative or corrective cardiac surgery. J Thorac Cardiovasc Surg 2014;148:2508–14.
CONGENITAL HEART
DISCUSSION DR ROSS M. UNGERLEIDER (Corpus Christi, TX): Double inlet left ventricle (DILV) is, as you say, a form of single ventricle. So the first question I would have is, what was it that made you think to look at DILV as a standalone group? What was it about that group of patients that captured enough of your intrigue that you thought, well, let us just study them?
lowest association with genetic syndromes, about 5%, as compared with other single ventricle anomalies, for example, heterotaxy at 100%. The other possibility is the fact that the dominant ventricle in DILV is of left morphology; however, whether that affects outcomes in the short to mid term is not really proven.
DR ALSOUFI: Since I joined Emory a few years ago, I have performed few clinical studies that examined results of single ventricle palliation and the effect of some risk factors such as genetic syndromes on outcomes. That is when I noticed the superior survival in patients with DILV compared with those with other single ventricle anomalies. Given that my review of the literature did not identify recent studies focusing on DILV, I decided to collect more anatomic and surgical details and examine outcomes and associated factors specific to DILV patients.
DR UNGERLEIDER: Do you think that double inlet deserves its own chapter in textbooks?
DR UNGERLEIDER: Is The Society of Thoracic Surgeons (STS) Congenital Heart Surgery Database granular enough for you to compare your outcomes with DILV in the STS cohort? Is it possible for you to gather that kind of comparison? DR ALSOUFI: Naturally, the data are as good as the person entering them and there are always inherent limitations associated with the registry databases. However, the STS database is granular enough when it comes to the underlying single ventricle anomaly, so identification of patients with DILV is possible. DR UNGERLEIDER: So I guess what I am leading to as a question is, as an outcome of this, Bahaaldin, would you suggest to Dr Jacobs and his companions who work on the STS Congenital Heart Surgery Database how to add some granularity to the fields to track these entities going forward?
DR ALSOUFI: It depends how big your textbook is. DR UNGERLEIDER: I guess what I am saying is as we start getting granular, we start taking single ventricles and now we get into all these different types, at what point does this become an entity of enough distinction because of the outcomes or because of the approaches or because of something that deserves a standalone description? DR ALSOUFI: I believe that multiinstitutional studies can allow more sophisticated analysis to improve our understanding and knowledge of rare cardiac defects. There are always limitations related to data granularity in the registry databases and to followup, for example, in the STS database. I have previously discussed a project with Drs Jeff and Marshall Jacobs hoping to study Norwood outcomes for various single ventricle anomalies, and I hope that this project might see the light if we can figure out how to link STS data with other databases to allow follow-up beyond hospital discharge. DR UNGERLEIDER: One final question. The patients with the balanced circulation, I presume those are the ones with a moderate amount of pulmonary stenosis? DR ALSOUFI: Correct.
DR ALSOUFI: Again, the underlying single ventricle anomaly is already there. Adding more anatomic details to certain cardiac defects might be possible and might prove to be beneficial for future studies. I believe that the STS database is beginning to expand toward that direction and some additional anatomic data have been recently added. This is similar to the increased details that are already in place, for example, for genetic syndromes and extracardiac malformations. DR UNGERLEIDER: You mentioned in a study that you presented earlier today where non–DiGeorge genetic syndromes were a marker for poor outcome. And so it obviously begs another question, which is, do you think the good outcomes that you are having are related to that or is it related to double inlet on its own or a combination? DR ALSOUFI: It is hard to know. One possibility is the fact that there is little association between DILV and genetic syndromes. I have recently studied the effect of genetic syndromes on single ventricle palliation outcomes. What I found is that DILV has the
DR UNGERLEIDER: So when we see that, those are the ones we can begin to think could survive to the next stage of palliation without an intervention? DR ALSOUFI: Right. DR UNGERLEIDER: Thank you. DR JEFFREY JACOBS (St. Petersburg, FL): First of all, a great presentation. As usual, it is up to your typical standard. I just wanted to respond to the dialogue that Ross had with you about whether it is possible to replicate a study like this in the STS Congenital Heart Surgery Database. I think the majority of the fields and data elements are in the database, so from that point of view it is possible. The trick is the longitudinal follow-up, and follow-up in the STS database stops at hospital discharge in 30 days. It is not always easy to identify that a given patient who had stage I is the same patient who had stage II who is the same patient who had stage III. So the trick is, how do we transform
Ann Thorac Surg 2017;104:2064–71
our STS Congenital Heart Surgery Database into a platform for longitudinal follow-up. A lot of people are working on it but nobody has figured out how to do it.
2071
In your data, it looks to me like those with a Glenn or BlalockTaussig shunt or both had pulmonary outflow obstruction, presumably, and those with a pulmonary artery band or those with a pulmonary artery band and arch obstruction must have had systemic outflow obstruction. So if you made your categories larger in the ways I have suggested, then I wonder if some of these things that have marginal p values might in fact turn out to be significant. DR ALSOUFI: Thank you for your comments. In fact, that was exactly the purpose of our study and we wanted to examine the effect of anatomic and subsequently surgical variables on palliation outcomes. In my presentation, there were no representative figures for those anatomic variables as none of them was a significant factor on our analysis. However, we have specifically examined the effect of ventriculoarterial discordance, along with the presence of pulmonary, systemic, and arch obstruction on transplant-free survival. Those are demonstrated on the tables here, and as you can see, there might be a trend on univariate analysis with decreased transplant-free survival in patients with systemic outflow obstruction that did not reach statistical significance, likely owing to limitations in our study because of sample size and the small number of events.
CONGENITAL HEART
DR JOHN E. MAYER (Boston, MA): The way I thought about single ventricle patients like this is heavily influenced by where the aorta arises, which type of chamber. If it is arising from a morphologic right ventricle and the morphologic left ventricle is the dominant ventricle, then you have got single ventricle with transposition, which is a different group of patients than if you have patients with normally related great vessels. The second way that I thought about single ventricle patients is that most single ventricles have obstruction to either one circulation or the other: they either have obstruction to the pulmonary circulation or they have obstruction to the systemic circulation. Therefore, I wonder if in your analysis of your data you have looked at categories based on these two anatomic considerations? One example is single ventricle double inlet with normally related great vessels, single ventricle with transposition, so the aorta is arising from an infundibular chamber, or those with systemic outflow obstruction as opposed to pulmonary outflow obstruction. I think that this sort of categorization scheme would give us something that would be a little bit easier to understand and maybe would yield some more statistical significance.
ALSOUFI ET AL SINGLE VENTRICLE PALLIATION OF DILV PATIENTS