Original Cardiovascular

Total Anomalous Pulmonary Venous

Connection: A 40 years Experience Analysis
Anas Lemaire1,2 Sylvie DiFilippo2 Jean-Jacques Parienti3 Olivier Metton2 Julia Mitchell2
Roland Hnaine2 Jean Ninet2

1 Department of Cardiac Surgery, Hopital Europeen Georges Address for correspondence Anas Lemaire, MD, Department of
Pompidou, Paris, le-de-France, France Cardiac Surgery, European Hospital Georges Pompidou, 20 rue
2 Department of Pediatric Cardiac Surgery, Hopital Cardio-vasculaire Leblanc, 75015 Paris, France (e-mail: ana.lemaire@yahoo.fr).
et Pneumologique Louis Pradel, Lyon, Rhne-Alpes, France
3 Department of Research and Biostatistic, Centre Hospitalier
Universitaire de Caen, Caen, Basse-Normandie, France

Thorac Cardiovasc Surg

Abstract Background Total anomalous pulmonary venous connection is a rare cardiac malfor-

Downloaded by: Vanderbilt University. Copyrighted material.

mation associated with signicant morbidity and mortality rates. We report a large
surgical series study to evaluate mid-term and long-term results of conventional surgical
techniques.
Methods and Results We performed a retrospective analytic study of all patients
operated on for simple total anomalous pulmonary venous connection in the University
Hospital of Lyon, France, between January 1973 and June 2014. A total of 180 patients
were included (43% supracardiac, 27% intracardiac, 19% infracardiac, and 11% mixed
types). Mean cardiopulmonary bypass and aortic cross clamp times were respectively 66
and 39 minutes. Overall mortality was 27.1%, including 38 early deaths (21.1%) and 12
late deaths (6.1%). The percentage of early death greatly decreased over the eras, from
42.1% in the seventies to 7.4% after 2010. Besides the earlier era of intervention
(p < 0.0001), signicant risk factors for death in multivariate analysis were preoperative
pulmonary hypertension, acidosis, and cardiopulmonary bypass time. There were 24
reoperations, including 7 for pulmonary venous obstruction; 6 died. Factors directly and
independently associated with late complications were the anatomic type (mixed
forms, p 0.0023), and length of aortic cross clamp time (p 0.01). Long-term
Keywords results for survivals are excellent. We report 84.7% of asymptomatic patients with a
congenital cardiac mean follow-up of 10.8 years.
disease Conclusions The overall prognosis of total anomalous pulmonary venous connection
morbidity repair with conventional procedures has greatly improved over the years with excellent
mortality long-term results. A thorough evaluation of all preoperative characteristics is imperative
survival to achieve the best outcome.

received Georg Thieme Verlag KG DOI http://dx.doi.org/

May 23, 2016 Stuttgart New York 10.1055/s-0036-1588007.
accepted after revision ISSN 0171-6425.
July 15, 2016
Total Anomalous Pulmonary Venous Connection Lemaire et al.
Introduction
Total anomalous pulmonary venous connection is a rare
congenital cardiac malformation in which all the four pulmo-
nary veins are connected to the systemic venous system
instead of the morphologically left atrium.1,2 Four different
types are described, depending on the connection site: supra-
cardiac, intracardiac, infracardiac, and mixed forms.3 Despite
ongoing advances in the eld of pediatric cardiac surgery, this
malformation remains a serious pathology whose manage-
ment must be adapted case by case. The most distressing
complication is pulmonary vein stenosis, often fatal. The aim
of our large single institution surgical series study is to
evaluate mid- and long-term results of conventional surgery
for total anomalous pulmonary venous connection.
Material and Method
We performed a retrospective analytic study of all patients
operated on for total anomalous pulmonary venous connection
in the Pediatric Cardiac Surgery Department of Louis Pradel
Cardiologic Hospital in Lyon, France, between January 1973 and
June 2014. Medical records were selected using a computer
research of the hospital database software and the charts of 201
patients were identied and analyzed. All cases associated with
major congenital heart diseases (single ventricle, atrial isomer-
ism, hypoplastic left heart syndrome, double outlet right ventri-
cle, complete atrioventricular septal defect, pulmonary atresia
with ventricular septal defect, etc.) were excluded, leaving a nal
number of 180 patients. Data were extracted by review of
medical records. Patients demographics, medical preoperative
characteristics (symptoms, clinical data including signs of heart
failure, hemodynamic parameters, oxygen saturation, mechani-
cal respiratory support, nitric oxide utilization, drugs, and
medications), echocardiographic measurements (pulmonary
pressure, left heart measurements, anatomical type and site of
pulmonary veins drainage, presence and location of pulmonary
veins obstruction), computed tomography scan data when
available, were collected. Patients were classied into four
groups according to the total anomalous pulmonary venous
connection type dened using the Craig and Darling classica-
tion3: type 1, supracardiac; type 2, intracardiac; type 3, infra-
cardiac; and type 4, mixed. Pulmonary hypertension was
classied as grade 0 if the systolic pulmonary pressure was
normal (i.e., < 30% of systolic blood pressure), grade 1 if it
ranged 30 to 50% of the systolic blood pressure, grade 2 if it
ranged from 50% to 75% of the systolic blood pressure, and grade
3 if it was > 75% of the systolic blood pressure. Preoperative
characteristics and events, early postoperative course and long-
term outcome, morbidity, and mortality were assessed. Early
operative mortality was dened as death during the initial
hospitalization or within the rst 30 postoperative days. Length
of follow-up was divided into four decades. A complication was
dened as any signicant event requiring some kind of drug and/
or interventional or surgical therapy (heart failure, arrhythmia,
infection, renal failure, hypoxemia and respiratory events, bleed-
ing, thromboembolic events, stroke and seizures, neurological
impairment, phrenic palsy). An early complication was dened
Thoracic and Cardiovascular Surgeon
as an event occurring during the postoperative hospital staying
or within the rst 30 postoperative days, and late after.
All patients were operated on under standard cardiopulmo-
nary bypass and cold cardioplegia by conventional surgical
repair techniques depending on the anatomical type. Supra-
and infracardiac types underwent side-to-side anastomosis of
the pulmonary venous collector to the morphologically left
atrium lifting the heart to the right; intracardiac total anomalous
pulmonary venous connection were operated by unroong of
the coronary sinus and partitioning of the atria with pericardial
patch closure of the atrial septal defect to commit the pulmonary
veins to the left. Repair for mixed types depended on localization
of the connections.
Long-term follow-up was assessed on medical charts and
data collected at latest evaluation. Patients clinical status was
determined by analyzing the last clinical visit in the medical
records. If the time from the latest visit was more than 1 year,
more recent data on the patients clinical condition was
sought by means of a phone call to the patients general
practitioner, cardiologist, or pediatrician.
Downloaded by: Vanderbilt University. Copyrighted material.
Statistical Analysis
Data are given as frequency, median with minimum and maxi-
mum value, or mean  standard deviation as appropriate, with
the number of missing values indicated. Major outcome criteria
selected were death, early and late complications, and reopera-
tion for pulmonary vein stenosis. Multivariate risk factors
associated with those three outcome events were sought
through logistic regression. One model was performed for
preoperative factors including preoperative obstruction, level
of pulmonary arterial hypertension, anatomical type, preopera-
tive acidosis, and orotracheal intubation. A second model was
performed for intraoperative factors including year at repair
classied in decades, weight at intervention, aortic cross clamp
and extracorporeal circulation times, circulatory arrest time
(when used), and degree of hypothermia. Those variables
were selected by a stepwise logistic regression model, with a
p-value threshold of 0.1 to enter and stay in the model. The
patients survival was evaluated with KaplanMeier curves.
Univariate analyses were performed with 2 and the Student
t-test when appropriate. All data were analyzed with SAS
statistical software (SAS institute, Inc, Cary, NC).
Results
Preoperative and Peroperative Analysis
The patients characteristics are presented in Table 1. The
median age at diagnosis was 88 days, ranging from a few
hours of life up to 4 years of age. The average weight at
diagnosis was 3.9 kg ( 1.96). The anatomical types included
78 supracardiac forms (43%), 48 intracardiac forms (27%), 35
infracardiac (19%), and 19 mixed (11%). Pulmonary vein
obstruction was present in 45% of all patients and was located
at the site of the vertical vein in 40.7%, at the site of connection
to the right venous system in 21%, through a restrictive
foramen ovale in 14.8%, or via ductus venosus closure for
infracardiac types (14.8%). Associated cardiac lesions were
 Total Anomalous Pulmonary Venous Connection Lemaire et al.

Table 1 Preoperative characteristic of all 180 patients operated for total anomalous pulmonary venous connection in our series

Variable All patients Missing data Supracardiac Intracardiac Infracardiac Mixed

Number 180 78 (43%) 48 (27%) 35 (19%) 19 (11%)
Age at diagnosis, 88 (<11,619) 10 109 123 11 57
mean (range, d) (<11,619) (<12,232) (<161) (<1217)
Weight at diagnosis 3,9  1.96 19 4.22  2.24 4.28  2.31 3.06  0.53 3.69  0.95
 SD (kg)
Gender Female 75 0 30 18 17 10
Male 105 0 48 30 18 9
Age at surgery, 107 (<12,240) 0 136 133 13 97
mean (range, d) (11,702) (32,240) (163) (2272)
Preoperative obstruction 81 (45%) 4 39 (50%) 9 (18.7%) 24 (68.6%) 9 (47%)
Preoperative Grade 0 4 (2.2%) 4 2 0 1 1
pulmonary
Grade 1 58 (32.2%) 29 22 2 5
hypertension
Grade 2 61 (33 0.9%) 23 18 14 6
Grade 3 53 (29.5%) 23 7 18 5

Downloaded by: Vanderbilt University. Copyrighted material.

Note: Preoperative pulmonary hypertension are classied as; grade 0, no pulmonary hypertension; grade 1, infrasystemic pulmonary hypertension;
grade 2, isosystemic; and grade 3, suprasystemic pulmonary hypertension.

present in 33 cases (18.3%; 15 ventricular septal defects, 5
anomalous systemic venous return, 3 cor triatriatum, 2 aortic
coarctations, 2 coronary anomalies, and 6 others), and non-
cardiac associated lesions were present in 21 cases (11.7%),
including 8 syndromic forms (Cat Eye syndrome in 4 cases).
The median age at surgery decreased from 111 days in the
seventies to 18 days after 2010 (Fig. 1). The intervals
between rst symptoms and diagnosis and between diagno-
sis and surgery decreased over the decades and a graphical
depiction of this evolution using KaplanMeier curves is
shown in Fig. 2. Thus, the mean elapsed time between
symptoms and diagnosis is signicantly shorter in the era
after year 2000 (p 0.044), as is the interval between
diagnosis and surgery which decreased from 44 days in the
seventies to 8 days after 2000 (p 0.0049).
The mean cardiopulmonary bypass time was 66 minutes
(range, 19240) and the mean aortic cross clamp time was 39
minutes (1489). Surgical repair was performed under cardiac
arrest and deep hypothermia in 104 cases (57.8%). Mean
cardiac arrest time was 35 minutes (range, 1568 minute).
Mean minimum temperature reached was 23C (mean, 19C
for cardiac arrest procedure and 28. 5C for continuous car-
diopulmonary bypass). Mean aortic cross clamp and cardio-
pulmonary bypass times were signicantly related to the
anatomical type, in particular, aortic cross clamp time was
longer for mixed types than nonmixed (p < 0.0001), that is, an

Fig. 1 Evolution of median age at surgery (days) over the decades.

Thoracic and Cardiovascular Surgeon

Total Anomalous Pulmonary Venous Connection Lemaire et al.
Fig. 2 Kaplan-Meier estimation curves about evolution of both symptoms diagnosis interval (a) and diagnosis surgery interval (b) over the
decades.
average of 19.3 minutes longer than intracardiac forms, 15.7
minutes longer than infracardiacs, and 11.3 minutes longer
than supracardiacs. Nineteen patients underwent concomitant
cardiac surgical procedure(s), including 9 pericardial enlarge-
ment of the superior or inferior caval vein, 4 ventricular septal
defect closure, 2 resection of an atrial membrane (cor triatrial),
1 pulmonary valvulotomy, 1 aortic coarctation correction, and
1 partial atrioventricular septal repair.
Early Outcomes
There were 38 early deaths (21.1%): 13 patients (7.2%) died in
the operative theater from global cardiac insufciency, major
pulmonary hypertension, and impossible weaning of cardio-
pulmonary bypass. Nine of these 13 peroperative deaths
(69%) occurred between 1972 and 1990. Among the 25
remaining early deaths, 15 were related to cardiac arrest
due to uncontrolled pulmonary hypertension crisis during
patient intensive care unit (ICU) stay, 4 occurred within the
postoperative hospital stay because of pulmonary vein ob-
struction (3 cases had major preoperative pulmonary vein
obstruction, and 1 patient operated on in 1979 developed a
signicant stenosis of the pulmonary drain vein anastomo-
sis). The percentage of early deaths greatly decreased over the
eras, from 42.1% in the seventies to 7.4% after 2010. Among
the 9 deaths since 2000, all presented with pulmonary
hypertension (6 suprasystemic [66.7%] and 3 isosystemic
[33.3%]; 7 patients [77.8%] had preoperative pulmonary
venous obstruction, 8 were preoperatively intubated
[88.9%] and 7 presented with preoperative acidosis [77.8%]).
Delayed sternal closure was required in 33 patients (6.7%).
The mean time from surgery to chest closure was 3.6 days
(range, 211 days). One patient with obstructive infracardiac
total anomalous pulmonary venous connection required
postoperative extracorporeal life support complicated with
6 reoperations because of bleeding and eventually died from
refractory cardiac arrest.
Peritoneal dialysis for renal function substitution was
performed in 19 patients (10.6%). The median postoperative
intubation time was 6 days (173), mean ICU and hospitali-
Thoracic and Cardiovascular Surgeon
zation stay were 10.1 (290) and 18.4 (498) days, respec-
Downloaded by: Vanderbilt University. Copyrighted material.
tively. There was no signicant difference over the decades
regarding intubation time (p 0.19), ICU stay (p 0.47), and
hospitalization stay (p 0.13). Nitric oxide (NO) was admin-
istered in 45 patients since its rst use in 1992. There was a
signicant difference over the decades in NO utilization
(39.3% patients from 1992 to 1999, 42.6% from 2000 to
2010, and 68.4% after 2010).
Overall, 131 early events occurred in 110 patients (61.1%).
These early complications included mainly arrhythmic events
(supraventricular tachycardia, transitory atrioventricular
block) and nonspecic post-cardiac surgery complications
such as cardiac arrests following pulmonary hypertensive
crisis in the ICU (19 cases [10.6%], 15 deaths), and phrenic
nerve palsy (17 cases [9.4%]). There was no statistical differ-
ence between patients with or without peroperative hypo-
thermia, regarding phrenic palsy occurrence (p 0.6995).
The systolic pulmonary artery pressure signicantly de-
creased in the early postoperative period (p < 0.001), from
an average over 65 mm Hg preoperatively to an average less
than 30 mm Hg in the rst postoperative day.
Long-Term Results
Survival
Among 180 patients, 38 died early postoperatively, 4 died
within late follow-up without reintervention, and 24 under-
went reoperations (of whom 7 died). Overall mortality is
27.1% (21.1% early and 6.1% late). The overall survival rates
were 81.1% at 1 month, 75% at 6 months, 75% at 10 years, and
73.9% at 20-year follow-up after surgery (Fig. 3A). Survival
was signicantly higher since the time of NO utilization, that
is, 1992 (p < 0.0001) (Fig. 3B). Survival rates were lower in
the group of patients with infracardiac total anomalous
pulmonary venous return (TAPVC; p 0.017) (Fig. 3C).
Late Complications and Reoperations
Among the 12 late deaths, 4 were due to noncardiac events
(1 central hypoxemia related to Goldenhar syndrome,
 Total Anomalous Pulmonary Venous Connection Lemaire et al.

Downloaded by: Vanderbilt University. Copyrighted material.

Fig. 3 Kaplan-Meier survival curves. 4A overall survival curve; 4B survival curves before and after the use of NO (1992). We note a signicant
improvement of survival probability when using NO, particularly in the early peri-operative period; 4C survival according to anatomical types. Type 1
(supracardiac), type 2 (intracardiac), type 3 (infracardiac), and type 4 (mixed). Early mortality is signicantly higher for infracardiac forms (p 0.017).

1 unknown, 1 neurological lesion, and 1 acute myeloid
leukemia) and occurred at 4 months, 12 years, 13 years, and
29 years respectively after surgery. The other 8 late deaths
were all related to pulmonary vein obstruction: 1 occurred
before reoperation and the 7 others were observed in the
early course after a surgical repair attempt. Patients who
underwent a reoperation for pulmonary vein obstruction
are presented in Table 3. Only one of them survived after
iterative surgery. Patients with pulmonary vein obstruction
presented with symptoms of heart failure and dyspnea. The
mean time between total anomalous pulmonary venous
connection repair and reoperation for pulmonary vein
obstruction was 3.1 months. Various materials such as
Dacron, Gore-Tex, or pericardium were used to achieve
surgical pulmonary valve (PV) plasty, except in one un-
treatable case diagnosed intraoperatively with major pul-
monary veins atresia. Death occurred within an average of
1.4 month after reoperation due to persistent pulmonary
vein stenosis (ranging from 1 day to 7 months).
The other late complications are displayed in Table 4 and
included: 5 persistent partial anomalous pulmonary venous
connection or atrial septal defects, 3 patients with neurologic
sequelae, 3 caval vein stenoses, and 3 cases with iatrogenic
right-to-left shunt due to postoperative deviation of the infe-
rior vena cava blood ow into the left atrium. These 3 patients
and the 5 cases with residual partial anomalous pulmonary
venous connection or atrial septal defect underwent reopera-
tions and had an event-free postoperative course. Among 3
patients diagnosed with superior vena cava postoperative
stenosis, only one underwent superior vena cava enlargement
plasty and two others remained asymptomatic.
Multivariate Analyses
Results of multivariate analysis are shown in Table 2. Signi-
cant factors for death were preoperative metabolic acidosis
(p 0.0157), preoperative pulmonary hypertension
(p 0.039), earlier era of surgical intervention (p < 0.0001),
lower weight at the time of surgery (p 0.0001), and longer
cardiopulmonary bypass time (p 0.08). Survival was also
signicantly improved by perioperative use of NO (p < 0.0001).
Signicant factors for the occurrence of early complica-
tions were preoperative pulmonary vein obstruction

Thoracic and Cardiovascular Surgeon

Total Anomalous Pulmonary Venous Connection Lemaire et al.

Table 2 Multivariate analysisa for preoperative risk factor for death, early complications, and late complications after surgical repair
of total anomalous pulmonary venous connection

Variable Data missing p-value Odds ratio 95% CI

Death
Preoperative data Acidosis 24 0.02 3.24 1.258.40
PHT grade 3 vs 2 24 0.04 1.57 0.584.25
PHT grade 0 or 1 vs 2 24 0.06 0.37 0.111.26
Peroperative data Decade of intervention 3 < 0, 0001 0.29 0.180.47
Weight at intervention 3 0,0001 0.33 0.2000.55
CBP time 3 0,08 1.01 0.991.02
Early complications
Preoperative data PVO 24 0.07 1.99 0.954.17
intubation 24 <0.0001 3.37 1.668.41
ECC length 3 0.009 1.02 1.011.03
Peroperative data Decade 3 0.009 0.65 0.460.90
Late complications

Downloaded by: Vanderbilt University. Copyrighted material.

Preoperative data Type 2 vs 1 24 0.15 0.63 0.182.17
Type 3 vs 1 24 0.24 0.70 0.202.41
Type 4 vs 1 24 0.002 4.81 1.4216.30
Intubation 24 0.09 2.30 0.876.04
Peroperative data ACC length 3 0.01 1.04 1.011.06

Abbreviations: ACC, aortic cross clamping; CBP, cardiopulmonary bypass; CI, condence interval; PHT, pulmonary hypertension; PVO, pulmonary
venous obstruction.
a
Variables included in the multivariable model had p < 0.1.

(p 0.0642), preoperative mechanical respiratory support
(p < 0.0001), longer cardiopulmonary bypass (0.0093), and
an earlier surgical era (p 0.0098). Factors directly and
independently correlated to late complications were the
anatomic type (p 0.0023 for the mixed type) and longer
aortic cross clamp time (p 0.01).
Clinical State at Latest Follow-Up
Overall, 131 patients (72.8%) survived throughout our long-term
follow-up after total anomalous pulmonary venous connection
repair. There were 4 patients who were lost to follow-up. The
mean follow-up was 10.8 years (17 days37.9 years). At their
latest evaluation, 84.7% were completely asymptomatic (New
York Heart Association [NYHA] class I), only one patient pre-
sented with NYHA class III symptoms, and 85.5% were taking no
medication. Echocardiographic evaluation showed normal pul-
monary pressures in 83.2% and grade I or II pulmonary hyper-
tension in 16.8%. There was no signicant difference between
systolic pulmonary pressure at time of hospital discharge and at
long-term follow-up latest assessment (p 0.35).
Discussion
The present study overall mortality (27.1%) is comparable to
other surgical series of total anomalous pulmonary venous
connection outcomes.47 In our series, mortality was associat-
ed with previously described risk factors, such as preoperative
pulmonary hypertension, lower weight at intervention, and a
longer length of cardiopulmonary bypass. The percentage of
early deaths greatly decreased over the eras, from 42.1% in the
seventies to 7.4% after 2010. In the present era, the worst case
scenario appears to be patients presenting with preoperative
pulmonary obstruction and suprasystemic pulmonary hyper-
tension with clinical preoperative instability, requiring preop-
erative intubation. Despite the convincing progress we have
witnessed in recent years in the eld of TAPVC surgery,
patients operated upon emergently with signicant pulmo-
nary vein obstruction and major pulmonary hypertension
remain burdened with the worst prognosis. Moreover, the
anatomical type and longer aortic cross clamp times were
signicant factors for early and late complications. The mixed
types presented the highest risk of late complications
(p 0.0023). This anatomical type was previously described
in the literature to be associated with a higher mortality rate.8
Of note, this form is associated with longer aortic cross clamp
times (51 vs 38 minutes in the whole series), which was found
to be the only signicant intraoperative risk factor for late
complications. It should be noted that, in this series, we only
included patients who survived until surgical repair. We did
not include the medical records of patients who died prior to
surgical intervention. The use of pre- and postoperative early
childhood longitudinal study in our center was comparatively
quite low, and perhaps should be used more often to further
improve management of the most severe patients.

Thoracic and Cardiovascular Surgeon

 Table 3 Summary of patients reoperated for pulmonary vein obstruction after total anomalous pulmonary venous connection repair in our series

Patient Year of Sex Type Preop-SPAP Age at Early complications SPAP at hospital Delay between first Surgical procedure at Death Delay between
surgery (mm Hg) surgery discharge intervention reintervention reoperation and
(d) (mm Hg) and reoperation death
(mo) (d)
1 1974 F Supra 40 185 None 48 4 Removal of obstructive No 
Dacron patch; atrial
repartitioning
2 1975 M Intra 45 144 PVO, transitory 30 2.1 PV enlargement plasty Yes 85
atrioventricular block with Dacron
3 1983 F Supra 42 12 Paroxystic Unknown 5.4 PV enlargement Yes 1
supraventricular plasty with Gore-Tex
tachycardia
4 1983 F Mixed 44 237 None 25 4.8 Cardiotomy nding Yes 7
almost atresia of the
PV. No repair possible.
5 1997 F Infra 85 2 PVO, 63 2.5 PV enlargement plasty Yes 54
neurological troubles with Gore-Tex
6 2007 F Supra 80 12 Neurological troubles 27 1.6 Pericardial enlargement Yes 36
plasty of the PV
7 2012 F Intra 70 12 None 45 1.2 Iterative Yes 215
1.8 complex surgerya

Abbreviations: F, female; M, male; PHT, pulmonary hypertension; PV, pulmonary veins; PVO, pulmonary venous obstruction; SPAP, systolic pulmonary arterial pressure.
a
Atrial repartitioning for intraatrial pulmonary venous obstruction 5 weeks after rst surgery, complicated by major pulmonary hypertension and bilateral pleurisy. Third intervention for pleural cleaning. Eight
months later, reappearance of suprasystemic pulmonary hypertension leading after evaluation to diagnosis of complete left portal vein thrombosis with almost absence of left lung vascularization. As her clinical
status worsened day-by-day, patient was scheduled for left pneumonectomy and sutureless plasty of the left pulmonary veins, but she died before surgery.
Total Anomalous Pulmonary Venous Connection

Thoracic and Cardiovascular Surgeon

Lemaire et al.

Downloaded by: Vanderbilt University. Copyrighted material.

Total Anomalous Pulmonary Venous Connection Lemaire et al.

Table 4 Incidence of early and late complications, and reinterventions in our series

Variable Value (percentage)

Early complications Rhythmic 21 (11.7%)
Cardiac arrest after major PHT 19 (10.6%)
Phrenic palsy 17 (9.4%)
Infectious 16 (8.9%)
Respiratory 15 (8.3%)
Acute renal failure 10 (5.6%)
Neurological 9 (5%)
Hemorrhagic 6 (3.3%)
Others 18 (10%)
Late complications PVO 8 (5 0.6%)
Persistent PAPVC or ASD 5 (3.5%)
Neurological 3 (2.1%)
Right-to-left shunt (iatrogenic) 3 (2.1%)
SVC or IVC stenosis 3 (2.1%)

Downloaded by: Vanderbilt University. Copyrighted material.

Others 6 (4.2%)
Reinterventions Pulmonary venous plasty for PVO 7 (4 0.9%)
Phrenoplicature 6 (4.2%)
Residual ASD closure 4 (2.8%)
Repartitioning of the right atrium for right-to-left shunt via the IVC 3 (2.1%)
SVC enlargement plasty 1 (0.7%)
Correction of a residual PAPVC 1 (0.7%)
Mediastinitis 1 (0.7%)
Pericardial effusion 1 (0.7%)

Abbreviations: ASD, atrial septal defect; IVC, inferior vena cava; PAPVC, partial anomalous pulmonary venous connection; PHT, pulmonary
hypertension; PVO, pulmonary venous obstruction; SVC, superior vena cava;

Phrenic nerve palsy was one of the more common and
important early postoperative surgery-related adverse events
and was evenly distributed over the decades. In the literature,
phrenic palsies after pediatric cardiac surgery (all pathology
combined) are reported in 0.3 to 12.8% of cases,912 generally
related to contusion, dissection, stretching, or thermal injury
of the phrenic nerve. Pediatric surgical procedures mostly
associated with this complication are tetralogy of Fallot (up to
31.5%),11 Fontan procedure (17.6%), BlalockTaussig shunt
(12.8%), and the arterial switch operation (10.8%).12 To our
knowledge, there is no report about phrenic nerve palsy rates
after total anomalous pulmonary venous connection repair.
The extensive posterolateral dissection of the parietal peri-
cardium to ensure correct exposure of the pulmonary veins
(a guarantee of quality repair) might account for the devel-
opment of phrenic nerve palsy. However, most of these cases
had event-free outcomes, which might explain the absence of
this reported complication in the literature.
Nevertheless, the present long-term single institution study
showed signicant improvement in outcomes over the past
decades. The utilization of NO to control pulmonary hyperten-
sive crisis in the early course after total anomalous pulmonary
venous connection surgery, is probably one of the main factors
accounting for these favorable results. The use of inhaled NO
dramatically changed perioperative prognosis by clearly
decreasing early mortality through better pulmonary hyperten-
sion management.13 However, mixed types and obstructed
forms are still burdened with high rate of complication and
mortality, which is also concordant with existing data.5,7,8 As
reported in the literature, the incidence of postoperative pulmo-
nary venous obstruction is correlated with preoperative pulmo-
nary veins morphology.14,15 The mortality after reoperation for
postrepair pulmonary vein obstruction and the percentage of
remaining pulmonary venous stenosis is known to be extremely
high.16 In our series, only one patient of the seven who under-
went reoperations for pulmonary vein obstruction survived for
the long term. Whatever the technique used to attempt pulmo-
nary venous plasty (Gore-Tex, Dacron, or autologous pericardial
patches), results are equally disappointing. The appearance and
development of pulmonary veno-occlusive disease is typically a
diffuse process, frequently involving all four pulmonary veins
progressively leading to brous cord-like and low compliant
vessels inaccessible to surgical venoplasty.
However, the long-term prognosis of the 72.8% patients
who survived in our series appeared to be excellent. Patients
operated during infancy for total anomalous pulmonary

Thoracic and Cardiovascular Surgeon

 Total Anomalous Pulmonary Venous Connection
venous connection with classical surgery who reached adult-
hood had excellent long-term outcomes, were totally asymp-
tomatic without any medication. Thus, patients who do not
have postrepair pulmonary vein obstruction may be consid-
ered as denitively cured.
Conclusion
Early diagnosis, rapid surgical repair, and optimal postopera-
tive resuscitation appear to be the best guarantee of favorable
outcome after total anomalous pulmonary venous connec-
tion repair. The overall prognosis has greatly improved over
the years and the surgical results of conventional techniques
in the last decades are satisfying. A thorough evaluation of all
preoperative characteristics is imperative to achieve best
medical and surgical outcomes. The long-term clinical results
for patients operated during infancy who have reached
adulthood are excellent.
Conicts of Interest
None.
Funding
This research received no specic grant from any funding
agency, commercial, or not-for-prot sectors.
References
1 Hlavacek AM, Shirali GS, Anderson RH. Pulmonary Venous Abnor-
malities. Paediatric Cardiology. 3rd ed. Philadelphia, PA: Churchill
Livingstone, Elsevier; 2010;24: 497522
2 Lewis FJ, Varco RL, Tauc M, Niazi SA. Direct vision repair of
triatrial heart and total anomalous pulmonary venous drainage.
Surg Gynecol Obstet 1956;102(6):713720
3 Craig JM, Darling RC, Rothney WB. Total pulmonary venous
drainage into the right side of the heart; report of 17 autopsied
cases not associated with other major cardiovascular anomalies.
Lab Invest 1957;6(1):4464
Lemaire et al.
4 Hancock Friesen CL, Zurakowski D, Thiagarajan RR, et al. Total
anomalous pulmonary venous connection: an analysis of current
management strategies in a single institution. Ann Thorac Surg
2005;79(2):596606, discussion 596606
5 Seale AN, Uemura H, Webber SA, et al; British Congenital Cardiac
Association. Total anomalous pulmonary venous connection:
morphology and outcome from an international population-based
study. Circulation 2010;122(25):27182726
6 Yong MS, dUdekem Y, Robertson T, et al. Outcomes of surgery for
simple total anomalous pulmonary venous drainage in neonates.
Ann Thorac Surg 2011;91(6):19211927
7 Karamlou T, Gurofsky R, Al Sukhni E, et al. Factors associated with
mortality and reoperation in 377 children with total anomalous
pulmonary venous connection. Circulation 2007;115(12):15911598
8 Kelle AM, Backer CL, Gossett JG, Kaushal S, Mavroudis C. Total
anomalous pulmonary venous connection: results of surgical
repair of 100 patients at a single institution. J Thorac Cardiovasc
Surg 2010;139(6):13871394.e3
9 Talwar S, Agarwala S, Mittal CM, Choudhary SK, Airan B. Diaphrag-
matic palsy after cardiac surgical procedures in patients with
congenital heart. Ann Pediatr Cardiol 2010;3(1):5057
10 Zhang YB, Wang X, Li SJ, Yang KM, Sheng XD, Yan J. Postoperative
diaphragmatic paralysis after cardiac surgery in children: inci-
Downloaded by: Vanderbilt University. Copyrighted material.
dence, diagnosis and surgical management. Chin Med J (Engl)
2013;126(21):40834087
11 Akay TH, Ozkan S, Gultekin B, et al. Diaphragmatic paralysis after
cardiac surgery in children: incidence, prognosis and surgical
management. Pediatr Surg Int 2006;22(4):341346
12 Joho-Arreola AL, Bauersfeld U, Stauffer UG, Baenziger O, Bernet V.
Incidence and treatment of diaphragmatic paralysis after cardiac
surgery in children. Eur J Cardiothorac Surg 2005;27(1):5357
13 Brunner N, de Jesus Perez VA, Richter A, et al. Perioperative
pharmacological management of pulmonary hypertensive crisis
during congenital heart surgery. Pulm Circ 2014;4(1):1024
14 Ando M, Takahashi Y, Kikuchi T. Total anomalous pulmonary
venous connection with dysmorphic pulmonary vein: a risk for
postoperative pulmonary venous obstruction. Interact Cardiovasc
Thorac Surg 2004;3(4):557561
15 Hyde JA, Stmper O, Barth MJ, et al. Total anomalous pulmonary
venous connection: outcome of surgical correction and manage-
ment of recurrent venous obstruction. Eur J Cardiothorac Surg
1999;15(6):735740, discussion 740741
16 Whight CM, Barratt-Boyes BG, Calder AL, Neutze JM, Brandt PW.
Total anomalous pulmonary venous connection. Long-term
results following repair in infancy. J Thorac Cardiovasc Surg
1978;75(1):5263
Thoracic and Cardiovascular Surgeon