Advertisement

Comparison of right ventricle–pulmonary artery shunt position in the Single Ventricle Reconstruction trial

Open ArchivePublished:February 10, 2017DOI:https://doi.org/10.1016/j.jtcvs.2016.10.104

      Abstract

      Objective

      Placement of a right ventricle–pulmonary artery shunt to the left or right of the neoaorta may influence reinterventions, pulmonary artery development, and survival after the Norwood procedure because of differences in shunt and pulmonary artery geometry and blood flow.

      Methods

      We analyzed the Pediatric Heart Network Single Ventricle Reconstruction Trial public use dataset. Comparisons were made between patients who received a left- or right-sided right ventricle–pulmonary artery shunt during the Norwood procedure in both the overall (n = 274) and the propensity score–matched (67 pairs) patient cohorts.

      Results

      A left-sided shunt was placed in 168 patients (61%), and a right-sided shunt was placed in 106 patients (39%). At the 12-month follow-up, there were no differences in pulmonary artery measurements, hemodynamic measurements, or pulmonary artery reinterventions between shunt groups. However, the right-sided shunt was associated with fewer surgical shunt revisions in both the overall (8.3 vs 1.9 events per 100 infants, P = .05) and the propensity score–matched (17.9 vs 0 events per 100 infants, P < .001) patient cohorts. In the propensity score–matched cohort only, right-sided shunts were further associated with fewer serious adverse events (84 vs 46 events per 100 infants, P = .01) and improved transplantation-free survival at 3 years follow-up (61% [95% confidence interval, 48-72] vs 80% [95% confidence interval, 69-88], P = .04).

      Conclusions

      In the Single Ventricle Reconstruction trial, right ventricle–pulmonary artery shunt placement to the right of the neoaorta was associated with fewer shunt revisions and may contribute to improved outcomes in select patients.

      Key Words

      Abbreviations and Acronyms:

      CPB (cardiopulmonary bypass), L-RVPA (left-sided right ventricle–pulmonary artery), MBT (modified Blalock–Taussig), PA (pulmonary artery), R-RVPA (right-sided right ventricle–pulmonary artery), RVPA (right ventricle–pulmonary artery), SVR (Single Ventricle Reconstruction)
      Figure thumbnail fx1
      RVPA shunt positions in the Norwood procedure.
      RVPA shunt placement to the right of the neoaorta was associated with fewer surgical shunt revisions in the SVR trial.
      In the SVR trial, RVPA shunts led to improved transplantation-free survival but more unintended interventions and worse branch PA growth. This secondary analysis of the SVR trial found that RVPA shunt reinterventions may be reduced by placement of the RVPA shunt to the right of the neoaorta.
      See Editorial Commentary page 1501.
      The outcomes of stage I reconstruction (Norwood procedure) for patients with hypoplastic left heart syndrome have gradually improved over time, in part because of incremental advances in surgical technique.
      • Jaquiss R.D.
      Hypoplastic left heart syndrome: how to improve late survival and quality of life.
      The landmark Pediatric Heart Network Single Ventricle Reconstruction (SVR) trial evaluated the source of pulmonary blood flow in patients undergoing the Norwood procedure and demonstrated that use of right ventricle–pulmonary artery (RVPA) shunts led to improved early survival compared with modified Blalock–Taussig (MBT) shunts, but came at the expense of more unintended interventions and worse branch pulmonary artery (PA) growth.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      • Aiyagari R.
      • Rhodes J.F.
      • Shrader P.
      • Radtke W.A.
      • Bandisode V.M.
      • Bergersen L.
      • et al.
      Impact of pre-stage II hemodynamics and pulmonary artery anatomy on 12-month outcomes in the Pediatric Heart Network Single Ventricle Reconstruction trial.
      However, numerous technical aspects of RVPA shunt placement, such as shunt size, position, choice of conduit, and anastomotic technique, may further influence these results. One potentially important technical variable that was recorded, but not dictated, by the SVR trial protocol was placement of the RVPA shunt to the left or right of the neoaorta. Although the shunt was originally routed to the left of the neoaorta, several authors have advocated for rightward placement of the RVPA shunt.
      • Rumball E.M.
      • McGuirk S.P.
      • Stumper O.
      • Laker S.J.
      • de Giovanni J.V.
      • Wright J.G.
      • et al.
      The RV-PA conduit stimulates better growth of the pulmonary arteries in hypoplastic left heart syndrome.
      • Barron D.J.
      • Brooks A.
      • Stickley J.
      • Woolley S.M.
      • Stumper O.
      • Jones T.J.
      • et al.
      The Norwood procedure using a right ventricle-pulmonary artery conduit: comparison of the right-sided versus left-sided conduit position.
      • Kolcz J.
      • Skladzien T.
      • Kordon Z.
      • Rudzinski A.
      • Skalski J.
      Impact of right ventricle-pulmonary artery conduit placement on pulmonary artery development after the Norwood procedure in hypoplastic left heart syndrome.
      • Gist K.M.
      • Barrett C.S.
      • Graham D.A.
      • Crumback S.L.
      • Schuchardt E.L.
      • Erickson B.
      • et al.
      Pulmonary artery interventions after Norwood procedure: does type or position of shunt predict need for intervention?.
      Purported benefits of right-sided (R)-RVPA shunt placement include (1) reduced kinking of the shunt leading to fewer shunt reinterventions; (2) reduced tension and more central placement on the PAs leading to more symmetric PA blood flow, improved PA growth, and fewer PA reinterventions; and (3) greater ease of surgical access to the distal shunt and central PAs during the stage II procedure leading to improved stage II outcomes. Limited single-institution reports have supported some of these assertions and demonstrated improved oxygen saturations,
      • Gist K.M.
      • Barrett C.S.
      • Graham D.A.
      • Crumback S.L.
      • Schuchardt E.L.
      • Erickson B.
      • et al.
      Pulmonary artery interventions after Norwood procedure: does type or position of shunt predict need for intervention?.
      improved PA growth,
      • Kolcz J.
      • Skladzien T.
      • Kordon Z.
      • Rudzinski A.
      • Skalski J.
      Impact of right ventricle-pulmonary artery conduit placement on pulmonary artery development after the Norwood procedure in hypoplastic left heart syndrome.
      • Gist K.M.
      • Barrett C.S.
      • Graham D.A.
      • Crumback S.L.
      • Schuchardt E.L.
      • Erickson B.
      • et al.
      Pulmonary artery interventions after Norwood procedure: does type or position of shunt predict need for intervention?.
      decreased PA reinterventions,
      • Kolcz J.
      • Skladzien T.
      • Kordon Z.
      • Rudzinski A.
      • Skalski J.
      Impact of right ventricle-pulmonary artery conduit placement on pulmonary artery development after the Norwood procedure in hypoplastic left heart syndrome.
      decreased cardiopulmonary bypass (CPB) and ischemic times at stage II operation,
      • Barron D.J.
      • Brooks A.
      • Stickley J.
      • Woolley S.M.
      • Stumper O.
      • Jones T.J.
      • et al.
      The Norwood procedure using a right ventricle-pulmonary artery conduit: comparison of the right-sided versus left-sided conduit position.
      and even improved survival with R-RVPA shunts.
      • Barron D.J.
      • Brooks A.
      • Stickley J.
      • Woolley S.M.
      • Stumper O.
      • Jones T.J.
      • et al.
      The Norwood procedure using a right ventricle-pulmonary artery conduit: comparison of the right-sided versus left-sided conduit position.
      We aimed to validate these findings and compare outcomes between patients who received a left-sided (L)-RVPA or R-RVPA shunt in the SVR trial dataset, which represents the largest multicenter, prospective cohort of infants with longitudinal follow-up after the Norwood procedure. We hypothesized that R-RVPA shunt placement would be associated with superior outcomes.

      Materials and Methods

       Study Design

      The SVR trial was conducted between May 2005 and July 2008 at 15 North American centers and randomized 549 patients with hypoplastic left heart syndrome or related single right ventricle lesions to receive an RVPA shunt or MBT shunt at the time of the Norwood procedure. Details of the trial design and main results have been reported.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      • Ohye R.G.
      • Gaynor J.W.
      • Ghanayem N.S.
      • Goldberg C.S.
      • Laussen P.C.
      • Frommelt P.C.
      • et al.
      Design and rationale of a randomized trial comparing the Blalock-Taussig and right ventricle-pulmonary artery shunts in the Norwood procedure.
      The Institutional Review Board at each participating center approved the trial (ClinicalTrials.gov number, NCT00115934), and written informed consent was obtained from 1 or both parents. The National Institutes of Health/National Heart, Lung, and Blood Institute Pediatric Heart Network SVR Trial public use dataset was used exclusively for the preparation of this work. Data were downloaded from https://www.pediatricheartnetwork.com/pud_login.asp?study_id=SVR on November 23, 2015. Only patients who received an RVPA shunt at the conclusion of the Norwood procedure were included in the analysis.

       Definitions and Study Measurements

      For patients who received an RVPA shunt at the conclusion of the Norwood procedure, the position of the shunt relative to the neoaorta (left or right) was recorded on the Supplemental Data Form (variable code RVPAS_SIDE). All other baseline, operative, and outcomes variables were reported as described in prior publications.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      • Aiyagari R.
      • Rhodes J.F.
      • Shrader P.
      • Radtke W.A.
      • Bandisode V.M.
      • Bergersen L.
      • et al.
      Impact of pre-stage II hemodynamics and pulmonary artery anatomy on 12-month outcomes in the Pediatric Heart Network Single Ventricle Reconstruction trial.
      • Ohye R.G.
      • Gaynor J.W.
      • Ghanayem N.S.
      • Goldberg C.S.
      • Laussen P.C.
      • Frommelt P.C.
      • et al.
      Design and rationale of a randomized trial comparing the Blalock-Taussig and right ventricle-pulmonary artery shunts in the Norwood procedure.
      • Hill K.D.
      • Rhodes J.F.
      • Aiyagari R.
      • Baker G.H.
      • Bergersen L.
      • Chai P.J.
      • et al.
      Intervention for recoarctation in the single ventricle reconstruction trial: incidence, risk, and outcomes.
      Center and surgeon yearly Norwood procedure volume estimates were based on the number of screened patients.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      • Ohye R.G.
      • Gaynor J.W.
      • Ghanayem N.S.
      • Goldberg C.S.
      • Laussen P.C.
      • Frommelt P.C.
      • et al.
      Design and rationale of a randomized trial comparing the Blalock-Taussig and right ventricle-pulmonary artery shunts in the Norwood procedure.
      Serious adverse events were defined as death, acute shunt failure, cardiac arrest, extracorporeal membrane oxygenation, unplanned cardiovascular reoperation, or necrotizing enterocolitis.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      • Ohye R.G.
      • Gaynor J.W.
      • Ghanayem N.S.
      • Goldberg C.S.
      • Laussen P.C.
      • Frommelt P.C.
      • et al.
      Design and rationale of a randomized trial comparing the Blalock-Taussig and right ventricle-pulmonary artery shunts in the Norwood procedure.
      Complications included any postoperative cardiac, respiratory, neurologic, gastrointestinal, infectious, renal, hematologic, vascular, or other complications as defined on the “PHN Code List M – Complications” data form. Phrenic nerve injury/diaphragm paralysis was defined as newly elevated diaphragm on chest x-ray. Vocal cord injury was defined as vocal cord immobility on direct visualization. Hemodynamic and angiographic measurements from interpretable prestage II catheterizations were performed by the Angiography Core Laboratory, and no additional independent angiographic review was performed for this study. The Nakata index (also called the “PA index”) was used to estimate central PA growth from mid-main branch PA measurements and was calculated using the formula π × [mid-main right PA diameter (mm)2 + mid-main left PA diameter (mm)2]/[4 × body surface area (m2)]. The lower lobe index was used to estimate peripheral PA growth from the proximal lower lobe branch PA (after the takeoff of the upper lobe branch) and was calculated using the formula π × [right lower lobe branch PA diameter (mm)2 + left lower lobe branch PA diameter (mm)2]/[4 × body surface area (m2)].
      • Aiyagari R.
      • Rhodes J.F.
      • Shrader P.
      • Radtke W.A.
      • Bandisode V.M.
      • Bergersen L.
      • et al.
      Impact of pre-stage II hemodynamics and pulmonary artery anatomy on 12-month outcomes in the Pediatric Heart Network Single Ventricle Reconstruction trial.
      • Ovroutski S.
      • Ewert P.
      • Alexi-Meskishvili V.
      • Holscher K.
      • Miera O.
      • Peters B.
      • et al.
      Absence of pulmonary artery growth after Fontan operation and its possible impact on late outcome.
      • Nakata S.
      • Imai Y.
      • Takanashi Y.
      • Kurosawa H.
      • Tezuka K.
      • Nakazawa M.
      • et al.
      A new method for the quantitative standardization of cross-sectional areas of the pulmonary arteries in congenital heart diseases with decreased pulmonary blood flow.
      The severity of branch PA stenosis was quantified as none (<15%), mild (15%-35%), moderate (>35%-50%), or severe (>50%), as described by Aiyagari and colleagues.
      • Aiyagari R.
      • Rhodes J.F.
      • Shrader P.
      • Radtke W.A.
      • Bandisode V.M.
      • Bergersen L.
      • et al.
      Impact of pre-stage II hemodynamics and pulmonary artery anatomy on 12-month outcomes in the Pediatric Heart Network Single Ventricle Reconstruction trial.

       Statistical Methods

      All comparisons were made between patients who received an L-RVPA or R-RVPA shunt at the conclusion of the Norwood procedure regardless of the assigned treatment group, representing a non–intention-to-treat analysis. Although baseline characteristics were similar between shunt groups, several important operative parameters differed between groups. As a result, patients were propensity matched to produce 2 patient cohorts with a similar propensity for receiving an L-RVPA or R-RVPA shunt. Propensity scores for receiving an L-RVPA or R-RVPA shunt were calculated by multivariable conditional logistic regression including the following covariates: sex, birth weight, gestational age, prenatal diagnosis, any associated anatomic diagnosis, age at Norwood procedure, year of operation, low-volume center (≤15 cases per year), shunt diameter, CPB time, crossclamp time, and perfusion strategy. Patients were matched on propensity scores within a 0.1 margin using the greedy match algorithm in a 1:1 ratio. The model produced 2 groups with a similar propensity for receiving an L-RVPA or R-RVPA shunt (Figure E1). Matching covariates and standardized difference of the mean are shown in Table E1.
      Summary data are presented as median with interquartile range or frequency counts with percentages. Reinterventions and adverse events are reported as the event rate per 100 enrolled infants, consistent with prior SVR trial publications.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      Continuous and categoric variables were compared using the Mann–Whitney rank-sum test or the Fisher exact test, respectively. Reintervention and adverse event rates were compared using the Poisson means test. Overall transplantation-free survival was compared between groups using Kaplan–Meier curves and the Gehan–Wilcoxon test. Calculations were performed using STATA 11.1 (StataCorp LP, College Station, Tex) and SAS 9.2 (SAS Institute Inc, Cary, NC).

      Results

       Baseline Characteristics

      A total of 281 patients in the SVR trial underwent RVPA shunt placement during the Norwood procedure; the position of the shunt relative to the neoaorta was recorded for 274 patients (98%) who comprised the final study cohort. A left-sided shunt was placed in 168 patients (61%), and a right-sided shunt was placed in 106 patients (39%). The proportion of R-RVPA shunts remained nearly constant from year to year (35% in 2005, 43% in 2006, 36% in 2007, 41% in 2008), suggesting no change in surgeon preference for shunt position over the course of the study. Although RVPA shunt position was not randomized, baseline characteristics remained well balanced, with no significant differences between groups (Table 1).
      Table 1Baseline characteristics
      VariableTotal cohortPropensity score–matched cohort
      L-RVPAS

      (n = 168)
      R-RVPAS

      (n = 106)
      P valueL-RVPAS

      (n = 67)
      R-RVPAS

      (n = 67)
      P value
      Male sex108 (64%)61 (58%).2641 (61%)39 (58%).73
      Birth weight (kg)3.1 [2.7-3.4]3.1 [2.8-3.4].883.3 [2.9-3.7]3.0 [2.7-3.4].02
      Birth weight <2.5 kg24 (14%)11 (10%).465 (7%)7 (10%).76
      Gestational age (wk)38 [37-39]38 [38-39].4839 [38-39]38 [38-39].15
      Prenatal diagnosis134 (80%)79 (75%).3150 (75%)51 (76%).84
      Anatomic diagnosis.67.82
      Hypoplastic left heart syndrome144 (86%)95 (90%)57 (85%)59 (88%)
      Critical aortic stenosis1 (1%)1 (1%)1 (1%)1 (1%)
      Single RV with systemic outflow obstruction4 (3%)4 (4%)3 (4%)3 (4%)
      Right dominant AV canal with systemic outflow obstruction11 (7%)4 (4%)2 (3%)3 (4%)
      Straddling MV with LV hypoplasia and outflow obstruction1 (1%)01 (1%)0
      Other7 (4%)2 (2%)3 (4%)1 (1%)
      Any associated anatomic diagnosis42 (25%)22 (21%).4219 (28%)10 (15%).06
      Aortic atresia110 (65%)64 (60%).3947 (70%)38 (57%).11
      Obstructed pulmonary venous return2 (1%)4 (4%).1600
      Data presented as number (%) or median [interquartile range]. L-RVPAS, Left-sided right ventricle–pulmonary artery shunt; R-RVPAS, right-sided right ventricle–pulmonary artery shunt; RV, right ventricle; AV, atrioventricular; MV, mitral valve; LV, left ventricle.

       Operative Parameters

      Operative variables for patients who underwent RVPA shunt placement during the stage 1 procedure (n = 274), as well as those who progressed to undergo the stage II procedure (n = 219), are shown in Table 2. Differences in the conduct of surgery were appreciated between groups. L-RVPA shunts were used more commonly by high-volume centers and surgeons, and were associated with shorter CPB and crossclamp times and more frequent use of deep hypothermic circulatory arrest (DHCA) as the perfusion strategy. Shunt diameters also differed between groups (L-RVPA: 68% 5 mm, 26% 6 mm; R-RVPA: 46% 5 mm, 52% 6 mm), with 6-mm shunts used more often in the R-RVPA group.
      Table 2Operative variables of stage I and II procedures
      Stage I procedureTotal cohortPropensity score–matched cohort
      L-RVPAS

      n = 168
      R-RVPAS

      n = 106
      P valueL-RVPAS

      n = 67
      R-RVPAS

      n = 67
      P value
      Age at Norwood (d)6 [4-8]6 [5-8].086 [4-8]6 [5-8].60
      Norwood year2006 [2006-2007]2006 [2006-2007].992006 [2006-2007]2006 [2006-2007].94
      Center Norwood volume ≤15/y18 (11%)28 (26%).00118 (27%)15 (22%).69
      Surgeon Norwood volume ≤10/y49 (27%)56 (53%).000128 (42%)35 (52%).30
      RVPAS diameter (mm)5 [5-6]6 [5-6]<.00016 [5-6]5 [5-6].44
      Indexed RVPAS diameter (mm/kg)1.7 [1.5-1.9]1.8 [1.6-2.0].0081.7 [1.5-1.9]1.8 [1.6-2.0].09
      CPB time (min)128 [96-158]166 [130-185]<.0001145 [118-174]147 [118-177].77
      Crossclamp time (min)51 [41-61]66 [49-80]<.000155 [42-70]59 [45-73].43
      Perfusion strategy<.0001.98
       DHCA only118 (70%)33 (31%)27 (40%)28 (42%)
       RCP only32 (19%)43 (41%)24 (36%)24 (36%)
       DHCA and RCP18 (11%)30 (28%)16 (24%)15 (22%)
      Stage II proceduren = 132n = 87n = 48n = 59
      Age at stage II (d)157 [128-196]153 [124-183].51159 [132-202]158 [127-196].60
      CPB time (min)86 [61-127]93 [77-133].0290 [64-116]93 [77-116].32
      Data presented as number (%) or median [interquartile range]. L-RVPAS, Left-sided right ventricle–pulmonary artery shunt; R-RVPAS, right-sided right ventricle–pulmonary artery shunt; RVPAS, right ventricle–pulmonary artery shunt; CPB, cardiopulmonary bypass; DHCA, deep hypothermic circulatory arrest; RCP, regional cerebral perfusion.
      As a result of the differences in operative parameters, which could confound the comparison of outcomes between groups, a subgroup analysis was performed using propensity score matching to produce 2 patient cohorts (n = 67 patients per group) with a similar propensity for receiving an L- or R-RVPA shunt. In the propensity score–matched cohort, baseline characteristics and key operative parameters, including center volume, surgeon volume, shunt size, CPB time, crossclamp time, and perfusion strategy, were well balanced between groups. The only statistically significant difference between propensity score–matched patients was that patients in the R-RVPA shunt group weighed less at birth (Tables 1 and 2).

       Pulmonary Artery and Shunt Reinterventions

      Unintended PA and shunt interventions before 12 months of age are summarized in Table 3. There were no differences in the rate of PA interventions between L-RVPA and R-RVPA groups in the overall or propensity score–matched cohorts. The majority of PA interventions (72%, 79 of 110 PA interventions) entailed pulmonary arterioplasty at the time of stage II operation.
      Table 3Unintended pulmonary artery and shunt interventions before 12 months of age
      InterventionTotal cohortPropensity score–matched cohort
      L-RVPAS

      (n = 168)
      R-RVPAS

      (n = 106)
      P valueL-RVPAS

      (n = 67)
      R-RVPAS

      (n = 67)
      P value
      PA interventions
       Balloon dilation or stent13 (7.7)13 (12.3).336 (9.0)4 (6.0).75
       Surgical augmentation56 (33.3)28 (26.4).3723 (34.3)16 (23.9).34
      Any PA intervention69 (41.1)41 (38.7).8429 (43.3)20 (29.9).25
      Shunt interventions
       Balloon dilation or stent20 (11.9)9 (8.5).527 (10.4)7 (10.4).99
       Surgical revision14 (8.3)2 (1.9).0512 (17.9)0<.001
       Conversion to MBTS3 (1.8)3 (2.8).862 (3.0)2 (3.0).99
      Any shunt intervention37 (22.0)14 (13.2).1221 (31.3)9 (13.4).04
      Data presented as number of events (event rate/100 infants). L-RVPAS, Left-sided right ventricle–pulmonary artery shunt; R-RVPAS, right-sided right ventricle–pulmonary artery shunt; PA, pulmonary artery; MBTS, modified Blalock–Taussig shunt.
      The total rate of unintended shunt intervention was 67% higher with L-RVPA shunts (22 events vs 13.2 events per 100 infants), but this difference failed to reach statistical significance (P = .12). However, the rate of surgical shunt revision was 4.4-fold higher with L-RVPA shunts compared with R-RVPA shunts (8.3 vs 1.9 events per 100 infants, P = .05). In the overall cohort, 12 patients underwent 16 shunt revisions before stage II operation. Mortality for patients requiring surgical shunt revision was 42% (5 of 12 deaths), and 15 of the 16 (94%) surgical shunt revisions occurred before Norwood discharge. In the propensity score–matched cohort, similar results were obtained, and the rates of unintended shunt intervention and surgical shunt revision were significantly lower in the R-RVPA shunt group.

       Angiographic and Hemodynamic Measurements

      Angiographic and hemodynamic measurements from patients undergoing prestage II catheterization with interpretable angiograms (n = 194) are shown in Table 4. Isolated left and right PA diameters were similar between groups in both the total and the propensity score–matched cohorts. There were also no differences in indexed central or peripheral PA measurements, rates of severe branch PA or shunt stenosis, pulmonary-to-systemic flow ratios (Qp:Qs), pulmonary vascular resistance, or arterial oxygen saturations between shunt groups in either cohort.
      Table 4Prestage II catheterization pulmonary artery and hemodynamic measurements
      VariableTotal cohortPropensity score–matched cohort
      L-RVPAS

      (n = 114)
      R-RVPAS

      (n = 80)
      P valueL-RVPAS

      (n = 45)
      R-RVPAS

      (n = 54)
      P value
      Body surface area (m2)0.31 [0.28-0.34]0.31 [0.30-0.34].130.31 [0.29-0.34]0.32 [0.30-0.34].39
      Central PA measurements
       Mid-main left PA diameter (mm)4.2 [3.4-5.2]4.8 [3.6-5.8].164.4 [3.6-5.4]4.8 [3.6-6.2].72
       Mid-main right PA diameter (mm)4.1 [3.2-5.1]4.4 [3.6-5.6].124.7 [3.6-5.9]4.5 [3.8-5.6].80
       PA (Nakata) index (mm2/m2)91 [59-135]118 [72-158].05106 [73-169]121 [73-162].73
      Peripheral PA measurements
       Proximal left lower lobe diameter (mm)4.9 [4.2-6.3]4.8 [4.2-5.7].294.5 [4.2-6.2]5.0 [4.0-5.8].87
       Proximal right lower lobe diameter (mm)4.8 [4.1-5.8]5.2 [4.2-5.9].285.1 [4.5-6.0]5.2 [4.5-5.9].75
       Total lower lobe index (mm2/m2)135 [100-193]134 [100-164].41150 [99-178]136 [99-179].57
      Stenoses
       Severe left PA stenosis (>50%)8 (7%)5 (6%).831 (2%)4 (7%).24
       Severe right PA stenosis (>50%)13 (11%)9 (11%).976 (13%)5 (9%).52
       Shunt stenosis31 (28%)23 (29%).8710 (23%)20 (38%).11
      Hemodynamic measurements
       Aortic oxygen saturation (%)74 [69-78]75 [71-78].2976 [68-78]75 [71-79].51
       Pulmonary-to-systemic flow ratio (Qp:Qs)1 [0.8-1.4]1 [0.7-1.1].051.2 [0.9-1.5]1.0 [0.7-1.3].09
       PVR (Wood units)1.9 [1.3-2.9]2.0 [1.6-2.8].522.3 [1.3-2.9]2.0 [1.7-3.0].83
      Data presented as number (%) or median [interquartile range]. L-RVPAS, Left-sided right ventricle–pulmonary artery shunt; R-RVPAS, right-sided right ventricle–pulmonary artery shunt; PA, pulmonary artery; PVR, pulmonary vascular resistance.

       Perioperative Outcomes and Transplantation-Free Survival

      Perioperative and longitudinal outcomes are shown in Table 5. In the total cohort, there were no significant differences in complications, serious adverse events, death, or cardiac transplantation after operation or at 12 months of follow-up between groups. The rates of phrenic nerve injury (3.0% vs 2.3%, P = .99) and vocal cord injury (1.5% vs 0%, P = .52) after stage II operation were equivalent between L-RVPA and R-RVPA shunt groups. Kaplan–Meier analysis similarly demonstrated no difference in transplantation-free survival between L-RVPA and R-RVPA shunt groups at 3 years of follow-up (Figure 1, A).
      Table 5Perioperative and 12-month outcomes
      Stage I outcomesTotal cohortPropensity score–matched cohort
      L-RVPAS

      (n = 168)
      R-RVPAS

      (n = 106)
      P valueL-RVPAS

      (n = 67)
      R-RVPAS

      (n = 67)
      P value
      Complications, No. (rate/100 infants)474 (282)267 (252).15212 (316)166 (248).02
      Discharged alive, No. (%)146 (87%)91 (86%).8055 (82%)61 (91%).13
      Stage II outcomesn = 132n = 87n = 48n = 59
      Complications, No. (rate/100 infants)148 (112)119 (137).1237 (77)75 (127).01
      Discharged alive, No. (%)125 (95%)83 (95%).8247 (98%)56 (95%).42
      Stage 1 operation to 12 mon = 168n = 106n = 67n = 67
      Complications, No. (rate/100 infants)877 (522)531 (501).47344 (513)330 (493).62
      Serious adverse events, No. (rate/100 infants)101 (60)52 (49).2756 (84)31 (46).01
      Death or cardiac transplantation44 (26%)24 (23%).5118 (27%)12 (18%).21
      L-RVPAS, Left-sided right ventricle–pulmonary artery shunt; R-RVPAS, right-sided right ventricle–pulmonary artery shunt.
      Figure thumbnail gr1
      Figure 1Transplantation-free survival stratified by shunt position in the (A) total cohort and (B) propensity score–matched cohort. RVPAS, Right ventricle–pulmonary artery shunt.
      In the propensity score–matched cohort, patients in the R-RVPA shunt group experienced fewer serious adverse events at 12 months of follow-up and improved transplantation-free survival at 36 months of follow-up (L-RVPA: 61% [95% confidence interval, 48-72], R-RVPA: 80% [95% confidence interval, 69-88]) (Figure 1, B).

      Discussion

      The SVR trial identified a short-term survival advantage with use of RVPA shunts during the Norwood procedure but also revealed several limitations of “first-generation” RVPA shunts, including increased shunt and PA interventions and worse PA growth.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      Since then, several technical modifications have been conceived in an attempt to mitigate the shortcomings of RVPA shunts while maintaining the hemodynamic advantages.
      • Jaquiss R.D.
      Hypoplastic left heart syndrome: how to improve late survival and quality of life.
      We examined RVPA shunt position relative to the neoaorta in the SVR trial public use dataset and found that RVPA shunt placement to the right of the neoaorta was associated with fewer surgical shunt revisions compared with L-RVPA shunt placement and may lead to improved survival in select patients. However, other key end points, including PA size, PA reinterventions, and hemodynamic parameters, appeared unaffected by RVPA shunt position.

       Shunt Interventions

      Construction of an RVPA shunt is technically challenging and has been the topic of considerable inquiry and debate. Technical considerations that require careful attention include (1) creation of the ventriculotomy and proximal anastomoses to prevent dynamic obstruction from ventricular muscle; (2) attention to the length, lie, and course of the shunt to prevent bending or kinking of the mid-portion or neoaorta/coronary compression; and (3) avoidance of tension or kinking at the site of distal PA insertion to prevent distal shunt stenosis or “bowtie” PA deformity. Presumably as a result of these technical challenges, the SVR trial and other single-institution series have reported an increased need for shunt intervention with use of RVPA shunts.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      • Pruetz J.D.
      • Badran S.
      • Dorey F.
      • Starnes V.A.
      • Lewis A.B.
      Differential branch pulmonary artery growth after the Norwood procedure with right ventricle-pulmonary artery conduit versus modified Blalock-Taussig shunt in hypoplastic left heart syndrome.
      We found a reduced rate of surgical shunt revision and overall shunt interventions with R-RVPA shunts when compared with L-RVPA shunts. Although the SVR trial data preclude a more granular analysis of the cause and nature of these shunt interventions, it should be noted that 94% of the surgical shunt revisions were performed during the Norwood hospitalization, whereas the rate of angiographic shunt stenosis observed later at prestage II catheterization was equivalent between groups (28% vs 29%). Thus, the data suggest there were likely fewer technical issues with R-RVPA shunt placement requiring early surgical revision, perhaps in part due to an improved lie of the shunt with less kinking or distortion at the mid-portion or proximal and distal insertion sites.
      • Rumball E.M.
      • McGuirk S.P.
      • Stumper O.
      • Laker S.J.
      • de Giovanni J.V.
      • Wright J.G.
      • et al.
      The RV-PA conduit stimulates better growth of the pulmonary arteries in hypoplastic left heart syndrome.
      Although it remains possible that other technical aspects of RVPA shunt creation not assessed in this dataset may partially explain the improved results in the R-RVPA shunt group, it should be noted that the first published reports advocating the use of ring-reinforced conduits
      • Bentham J.R.
      • Baird C.W.
      • Porras D.P.
      • Rathod R.H.
      • Marshall A.C.
      A reinforced right-ventricle-to-pulmonary-artery conduit for the stage-1 Norwood procedure improves pulmonary artery growth.
      or the proximal
      • Tweddell J.S.
      • Mitchell M.E.
      • Woods R.K.
      • Spray T.L.
      • Quintessenza J.A.
      Construction of the right ventricle-to-pulmonary artery conduit in the Norwood: the “dunk” technique.
      or distal
      • Mascio C.E.
      • Spray T.L.
      Distal dunk for right ventricle to pulmonary artery shunt in stage 1 palliation.
      dunk technique appeared well after the time period of the SVR trial, suggesting that these additional technical modifications may have been less relevant during the SVR trial period.

       Pulmonary Artery Interventions and Size

      The SVR trial and other nonrandomized studies reported increased rates of PA interventions with use of RVPA shunts compared with MBT shunts, typically because of central PA stenosis at the site of RVPA shunt insertion (“bowtie” deformity) thought to result from tension or kinking at the distal anastomosis or increased shear stress and neointimal proliferation from pulsatile flow.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      • Barron D.J.
      • Brooks A.
      • Stickley J.
      • Woolley S.M.
      • Stumper O.
      • Jones T.J.
      • et al.
      The Norwood procedure using a right ventricle-pulmonary artery conduit: comparison of the right-sided versus left-sided conduit position.
      • Kolcz J.
      • Skladzien T.
      • Kordon Z.
      • Rudzinski A.
      • Skalski J.
      Impact of right ventricle-pulmonary artery conduit placement on pulmonary artery development after the Norwood procedure in hypoplastic left heart syndrome.
      • Gist K.M.
      • Barrett C.S.
      • Graham D.A.
      • Crumback S.L.
      • Schuchardt E.L.
      • Erickson B.
      • et al.
      Pulmonary artery interventions after Norwood procedure: does type or position of shunt predict need for intervention?.
      • Pruetz J.D.
      • Badran S.
      • Dorey F.
      • Starnes V.A.
      • Lewis A.B.
      Differential branch pulmonary artery growth after the Norwood procedure with right ventricle-pulmonary artery conduit versus modified Blalock-Taussig shunt in hypoplastic left heart syndrome.
      As a result, central PA augmentation is required in as much as 80% of RVPA shunt cases at the time of stage II operation.
      • Barron D.J.
      • Brooks A.
      • Stickley J.
      • Woolley S.M.
      • Stumper O.
      • Jones T.J.
      • et al.
      The Norwood procedure using a right ventricle-pulmonary artery conduit: comparison of the right-sided versus left-sided conduit position.
      However, data regarding the impact of RVPA shunts on branch PA growth are mixed. Some centers have reported improved branch PA growth with RVPA shunts,
      • Rumball E.M.
      • McGuirk S.P.
      • Stumper O.
      • Laker S.J.
      • de Giovanni J.V.
      • Wright J.G.
      • et al.
      The RV-PA conduit stimulates better growth of the pulmonary arteries in hypoplastic left heart syndrome.
      • Gist K.M.
      • Barrett C.S.
      • Graham D.A.
      • Crumback S.L.
      • Schuchardt E.L.
      • Erickson B.
      • et al.
      Pulmonary artery interventions after Norwood procedure: does type or position of shunt predict need for intervention?.
      • Pruetz J.D.
      • Badran S.
      • Dorey F.
      • Starnes V.A.
      • Lewis A.B.
      Differential branch pulmonary artery growth after the Norwood procedure with right ventricle-pulmonary artery conduit versus modified Blalock-Taussig shunt in hypoplastic left heart syndrome.
      presumably because of pulsatile flow and the more equal distribution of pulmonary blood flow compared with the predominantly right-sided MBT shunt placement. In contrast, the SVR trial identified worse central PA growth with RVPA shunts, as evidenced by a lower Nakata index and a higher rate of severe left branch PA stenosis.
      • Ohye R.G.
      • Sleeper L.A.
      • Mahony L.
      • Newburger J.W.
      • Pearson G.D.
      • Lu M.
      • et al.
      Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
      • Aiyagari R.
      • Rhodes J.F.
      • Shrader P.
      • Radtke W.A.
      • Bandisode V.M.
      • Bergersen L.
      • et al.
      Impact of pre-stage II hemodynamics and pulmonary artery anatomy on 12-month outcomes in the Pediatric Heart Network Single Ventricle Reconstruction trial.
      Previous comparisons of PA interventions and growth between L- and R-RVPA shunts in small single-institution studies also have produced conflicting results.
      • Barron D.J.
      • Brooks A.
      • Stickley J.
      • Woolley S.M.
      • Stumper O.
      • Jones T.J.
      • et al.
      The Norwood procedure using a right ventricle-pulmonary artery conduit: comparison of the right-sided versus left-sided conduit position.
      • Kolcz J.
      • Skladzien T.
      • Kordon Z.
      • Rudzinski A.
      • Skalski J.
      Impact of right ventricle-pulmonary artery conduit placement on pulmonary artery development after the Norwood procedure in hypoplastic left heart syndrome.
      • Gist K.M.
      • Barrett C.S.
      • Graham D.A.
      • Crumback S.L.
      • Schuchardt E.L.
      • Erickson B.
      • et al.
      Pulmonary artery interventions after Norwood procedure: does type or position of shunt predict need for intervention?.
      In our analysis of the SVR trial, we found no difference in the overall rate of PA intervention, angiographic PA size, or branch PA stenosis at prestage II catheterization between R-RVPA and L-RVPA shunt cases, suggesting that RVPA shunt position had little impact on the development of central or branch PA stenosis. These data suggest that other technical modifications aside from shunt position are needed to reduce the rate of PA stenosis with RVPA shunts.
      • Mascio C.E.
      • Spray T.L.
      Distal dunk for right ventricle to pulmonary artery shunt in stage 1 palliation.

       Perioperative and Long-Term Outcomes

      Prior authors have advocated for R-RVPA shunt placement because of greater ease of surgical access to the distal shunt and central PAs during the stage II procedure, avoiding the difficult dissection of the central PAs behind the neoaorta with the attendant risks of phrenic and recurrent laryngeal nerve injury and increased operative times.
      • Rumball E.M.
      • McGuirk S.P.
      • Stumper O.
      • Laker S.J.
      • de Giovanni J.V.
      • Wright J.G.
      • et al.
      The RV-PA conduit stimulates better growth of the pulmonary arteries in hypoplastic left heart syndrome.
      • Barron D.J.
      • Brooks A.
      • Stickley J.
      • Woolley S.M.
      • Stumper O.
      • Jones T.J.
      • et al.
      The Norwood procedure using a right ventricle-pulmonary artery conduit: comparison of the right-sided versus left-sided conduit position.
      • Gist K.M.
      • Barrett C.S.
      • Graham D.A.
      • Crumback S.L.
      • Schuchardt E.L.
      • Erickson B.
      • et al.
      Pulmonary artery interventions after Norwood procedure: does type or position of shunt predict need for intervention?.
      Conversely, concerns with R-RVPA shunt placement include the possibility of shunt compression of the coronary arteries or neoaortic root, as well as damage to the shunt during sternal reentry due to a retrosternal shunt trajectory that crosses the midline.
      • Kolcz J.
      • Skladzien T.
      • Kordon Z.
      • Rudzinski A.
      • Skalski J.
      Impact of right ventricle-pulmonary artery conduit placement on pulmonary artery development after the Norwood procedure in hypoplastic left heart syndrome.
      Last, long-term survival beyond the stage II procedure may be affected if shunt position leads to improved Fontan candidacy because of superior pulmonary blood flow and lung development.
      • Barron D.J.
      • Brooks A.
      • Stickley J.
      • Woolley S.M.
      • Stumper O.
      • Jones T.J.
      • et al.
      The Norwood procedure using a right ventricle-pulmonary artery conduit: comparison of the right-sided versus left-sided conduit position.
      In the overall patient cohort, our analysis of the SVR trial failed to identify any objective differences in perioperative outcomes associated with RVPA shunt position. Specifically, there were no differences in complications, serious adverse events, nerve injury, or discharge mortality after the stage I or II operations. Paradoxically, CBP and crossclamp times were longer with the R-RVPA shunt position. However, these results were heavily confounded by the differences in DHCA perfusion strategy, which essentially invalidates the comparison of perfusion times between groups. Transplantation-free survival at 3 years was further equivalent between RVPA shunt groups. However, these comparisons of patient outcomes in the overall cohort may have been partially confounded by differences in operative parameters, such as surgeon and center volume, which may have favored improved outcomes in the L-RVPA group
      • Tabbutt S.
      • Ghanayem N.
      • Ravishankar C.
      • Sleeper L.A.
      • Cooper D.S.
      • Frank D.U.
      • et al.
      Risk factors for hospital morbidity and mortality after the Norwood procedure: a report from the Pediatric Heart Network Single Ventricle Reconstruction trial.
      • Pasquali S.K.
      • Jacobs J.P.
      • He X.
      • Hornik C.P.
      • Jaquiss R.D.
      • Jacobs M.L.
      • et al.
      The complex relationship between center volume and outcome in patients undergoing the Norwood operation.
      and masked any benefit of R-RVPA shunt position.
      In the propensity score–matched cohort with balanced operative parameters, the results demonstrated fewer serious adverse events at 12 months and improved transplantation-free survival at 3 years in patients with R-RVPA shunts, despite the slightly lower birth weight of these patients. Surgical shunt revisions were associated with a 42% mortality rate and coincided with several serious adverse events, including death, acute shunt failure, cardiac arrest, extracorporeal membrane oxygenation, and unplanned cardiovascular reoperation. Thus, in the absence of other major differences in surgeon or center practices that could affect outcomes, it remains mechanistically feasible that the improved results in the propensity score–matched R-RVPA shunt group were related to fewer surgical shunt revisions. However, given that the propensity score match excluded more than 50% of patients, including the majority of patients who underwent surgery with DHCA, these results may not be generalizable to the overall Norwood population, and we view these survival results as speculative and hypothesis generating at this time.

       Study Limitations

      Our study contains several important limitations that should be acknowledged. RVPA shunt position was a nonrandomized surgical variable that was dictated by surgeon preference. Propensity score matching was used in an attempt to normalize the major differences in operative parameters between groups. However, the propensity score match excluded the majority of patients who underwent surgery with DHCA and yielded differences in birth weight between groups. Therefore, the results of the propensity score match subgroup analysis may not be generalizable to the population with circulatory arrest. The slightly lower birth weight of the patients in the R-RVPA group may have predisposed them to worse outcomes given that birth weight is a risk factor for morbidity and mortality after the Norwood operation, although there was no difference in the highest risk group of neonates with birth weights less than 2.5 kg.
      • Tabbutt S.
      • Ghanayem N.
      • Ravishankar C.
      • Sleeper L.A.
      • Cooper D.S.
      • Frank D.U.
      • et al.
      Risk factors for hospital morbidity and mortality after the Norwood procedure: a report from the Pediatric Heart Network Single Ventricle Reconstruction trial.
      • Curzon C.L.
      • Milford-Beland S.
      • Li J.S.
      • O’Brien S.M.
      • Jacobs J.P.
      • Jacobs M.L.
      • 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.
      In addition, shunt position may serve as a marker for several other surgeon preferences, center differences, or technical differences in shunt creation that were unrecorded in this dataset and unable to be controlled for by propensity matching. Further, other important late end points that may be influenced by pulmonary development, such as quality of life, exercise capacity, and functional capacity through the Fontan operation, were not assessed in this dataset. Finally, the surgical techniques of RVPA shunt creation and outcomes achieved during the study period from 2005 to 2008 may represent the learning curve for some surgeons and may not be reflective of contemporary practices.

      Conclusions

      In this unplanned secondary analysis of the SVR trial public use dataset, we found that RVPA shunt placement to the right of the neoaorta was associated with fewer shunt revisions and improved survival in a propensity score–matched cohort, suggesting that R-RVPA shunt placement may partially alleviate some of the previously recognized limitations of RVPA conduits. However, other key end points, including PA size, PA reinterventions, and hemodynamic parameters, appeared unaffected by RVPA shunt position.

       Conflict of Interest Statement

      Authors have nothing to disclose with regard to commercial support.
      The authors thank Sally Cai for assistance with statistical revisions.

      Appendix.

      Figure thumbnail fx2
      Figure E1Distribution of propensity scores for matched patients. RVPAS, Right ventricle–pulmonary artery shunt.
      Table E1Matching model results
      VariablePropensity score–matched cohort
      n = 67 per group
      Male sex.054
      Birth weight (kg).451
      Gestational age (wk).228
      Prenatal diagnosis.096
      Any associated anatomic diagnosis.118
      Age at Norwood (d).287
      Norwood year.155
      Center Norwood volume ≤15/y.250
      RVPAS diameter (mm).089
      CPB time (stage 1, min).071
      Crossclamp time (stage 1, min).239
      Perfusion strategy (stage 1).092
      Data are shown as standardized difference of the mean. RVPAS, Right ventricle–pulmonary artery shunt; CPB, cardiopulmonary bypass.

      References

        • Jaquiss R.D.
        Hypoplastic left heart syndrome: how to improve late survival and quality of life.
        J Thorac Cardiovasc Surg. 2015; 150: 1401-1403
        • Ohye R.G.
        • Sleeper L.A.
        • Mahony L.
        • Newburger J.W.
        • Pearson G.D.
        • Lu M.
        • et al.
        Comparison of shunt types in the Norwood procedure for single-ventricle lesions.
        N Engl J Med. 2010; 362: 1980-1992
        • Aiyagari R.
        • Rhodes J.F.
        • Shrader P.
        • Radtke W.A.
        • Bandisode V.M.
        • Bergersen L.
        • et al.
        Impact of pre-stage II hemodynamics and pulmonary artery anatomy on 12-month outcomes in the Pediatric Heart Network Single Ventricle Reconstruction trial.
        J Thorac Cardiovasc Surg. 2014; 148: 1467-1474
        • Rumball E.M.
        • McGuirk S.P.
        • Stumper O.
        • Laker S.J.
        • de Giovanni J.V.
        • Wright J.G.
        • et al.
        The RV-PA conduit stimulates better growth of the pulmonary arteries in hypoplastic left heart syndrome.
        Eur J Cardiothorac Surg. 2005; 27: 801-806
        • Barron D.J.
        • Brooks A.
        • Stickley J.
        • Woolley S.M.
        • Stumper O.
        • Jones T.J.
        • et al.
        The Norwood procedure using a right ventricle-pulmonary artery conduit: comparison of the right-sided versus left-sided conduit position.
        J Thorac Cardiovasc Surg. 2009; 138: 528-537
        • Kolcz J.
        • Skladzien T.
        • Kordon Z.
        • Rudzinski A.
        • Skalski J.
        Impact of right ventricle-pulmonary artery conduit placement on pulmonary artery development after the Norwood procedure in hypoplastic left heart syndrome.
        Eur J Cardiothorac Surg. 2012; 42: 218-224
        • Gist K.M.
        • Barrett C.S.
        • Graham D.A.
        • Crumback S.L.
        • Schuchardt E.L.
        • Erickson B.
        • et al.
        Pulmonary artery interventions after Norwood procedure: does type or position of shunt predict need for intervention?.
        J Thorac Cardiovasc Surg. 2013; 145: 1485-1492
        • Ohye R.G.
        • Gaynor J.W.
        • Ghanayem N.S.
        • Goldberg C.S.
        • Laussen P.C.
        • Frommelt P.C.
        • et al.
        Design and rationale of a randomized trial comparing the Blalock-Taussig and right ventricle-pulmonary artery shunts in the Norwood procedure.
        J Thorac Cardiovasc Surg. 2008; 136: 968-975
        • Hill K.D.
        • Rhodes J.F.
        • Aiyagari R.
        • Baker G.H.
        • Bergersen L.
        • Chai P.J.
        • et al.
        Intervention for recoarctation in the single ventricle reconstruction trial: incidence, risk, and outcomes.
        Circulation. 2013; 128: 954-961
        • Ovroutski S.
        • Ewert P.
        • Alexi-Meskishvili V.
        • Holscher K.
        • Miera O.
        • Peters B.
        • et al.
        Absence of pulmonary artery growth after Fontan operation and its possible impact on late outcome.
        Ann Thorac Surg. 2009; 87: 826-831
        • Nakata S.
        • Imai Y.
        • Takanashi Y.
        • Kurosawa H.
        • Tezuka K.
        • Nakazawa M.
        • et al.
        A new method for the quantitative standardization of cross-sectional areas of the pulmonary arteries in congenital heart diseases with decreased pulmonary blood flow.
        J Thorac Cardiovasc Surg. 1984; 88: 610-619
        • Pruetz J.D.
        • Badran S.
        • Dorey F.
        • Starnes V.A.
        • Lewis A.B.
        Differential branch pulmonary artery growth after the Norwood procedure with right ventricle-pulmonary artery conduit versus modified Blalock-Taussig shunt in hypoplastic left heart syndrome.
        J Thorac Cardiovasc Surg. 2009; 137: 1342-1348
        • Bentham J.R.
        • Baird C.W.
        • Porras D.P.
        • Rathod R.H.
        • Marshall A.C.
        A reinforced right-ventricle-to-pulmonary-artery conduit for the stage-1 Norwood procedure improves pulmonary artery growth.
        J Thorac Cardiovasc Surg. 2015; 149: 1502-1508.e1
        • Tweddell J.S.
        • Mitchell M.E.
        • Woods R.K.
        • Spray T.L.
        • Quintessenza J.A.
        Construction of the right ventricle-to-pulmonary artery conduit in the Norwood: the “dunk” technique.
        Oper Tech Thorac Cardiovasc Surg. 2012; 17: 81-98
        • Mascio C.E.
        • Spray T.L.
        Distal dunk for right ventricle to pulmonary artery shunt in stage 1 palliation.
        Ann Thorac Surg. 2015; 100: 2381-2382
        • Tabbutt S.
        • Ghanayem N.
        • Ravishankar C.
        • Sleeper L.A.
        • Cooper D.S.
        • Frank D.U.
        • 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: 882-895
        • Pasquali S.K.
        • Jacobs J.P.
        • He X.
        • Hornik C.P.
        • Jaquiss R.D.
        • Jacobs M.L.
        • et al.
        The complex relationship between center volume and outcome in patients undergoing the Norwood operation.
        Ann Thorac Surg. 2012; 93: 1556-1562
        • Curzon C.L.
        • Milford-Beland S.
        • Li J.S.
        • O’Brien S.M.
        • Jacobs J.P.
        • Jacobs M.L.
        • 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 Thorac Cardiovasc Surg. 2008; 135: 546-551