Fitness to fly for children and adolescents after Fontan palliation

Abstract

Introduction
At cruising altitude, the cabin pressure of passenger aircraft needs to be adjusted and, therefore, the oxygen content is equivalent to ambient air at 2,500 masl, causing mild desaturation and a rising pulmonary vascular resistance (PVR) in healthy subjects. For Fontan patients with passive pulmonary perfusion, a rising PVR can cause serious medical problems. The purpose of this fitness to fly investigation (FTF) is to assess the risk of air travel for children and adolescents after Fontan palliation.

Methods
We investigated 21 Fontan patients [3–14y] in a normobaric hypoxic chamber at a simulated altitude of 2,500 m for 3 h. Oxygen saturation, heart rate, and regional tissue saturation in the forehead (NIRS) were measured continuously. Before entering the chamber, after 90 and 180 min in the hypoxic environment, blood gas analysis and echocardiography were performed.

Results
Heart rate and blood pressure did not show significant intraindividual changes. Capillary oxygen saturation (SaO 2 ) decreased significantly after 90 min by a mean of 5.6 ± 2.87% without further decline. Lactate, pH, base excess, and tissue saturation in the frontal brain did not reach any critical values. In the case of open fenestration between the tunnel and the atrium delta, P did not increase, indicating stable pulmonary artery pressure.

Conclusion
All 21 children finished the investigation successfully without any adverse events, so flying short distance seems to be safe for most Fontan patients with good current health status. As the baseline oxygen saturation does not allow prediction of the maximum extent of desaturation and adaption to a hypoxic environment takes up to 180 min, the so-called hypoxic challenge test is not sufficient for these patients. Performing an FTF examination over a period of 180 min allows for risk assessment and provides safety to the patients and their families, as well as the airline companies.

Full text

EDITED BY
Donald Hagler,
Mayo Clinic, United States
REVIEWED BY
David Alexander White,
Children’s Mercy Kansas City, United States
Diana Zannino,
Royal Children’s Hospital, Australia
*CORRESPONDENCE
N. Müller
nicole.mueller@ukbonn.de
RECEIVED 20 February 2023
ACCEPTED 02 June 2023
PUBLISHED 23 June 2023
CITATION
Müller N, Herberg U, Breuer J, Kratz T and
Härtel JA (2023) Fitness to fly for children and
adolescents after Fontan palliation.
Front. Cardiovasc. Med. 10:1170275.
doi: 10.3389/fcvm.2023.1170275
COPYRIGHT
© 2023 Müller, Herberg, Breuer, Kratz and
Härtel. This is an open-access article distributed
under the terms of the Creative Commons
Attribution License (CC BY). The use,
distribution or reproduction in other forums is
permitted, provided the original author(s) and
the copyright owner(s) are credited and that the
original publication in this journal is cited, in
accordance with accepted academic practice.
No use, distribution or reproduction is
permitted which does not comply with these
terms.
Fitness to fly for children and
adolescents after Fontan palliation
N. Müller1*, U. Herberg1,2, J. Breuer1,T.Kratz
1and J. A. Härtel1
1
Department for Pediatric Cardiology, Children’s Heart Center UK Bonn, University Hospital Bonn,
Bonn, Germany,
2
Department for Pediatric Cardiology, University Hospital Aachen, Aachen, Germany
Introduction: At cruising altitude, the cabin pressure of passenger aircraft needs to
be adjusted and, therefore, the oxygen content is equivalent to ambient air at
2,500 masl, causing mild desaturation and a rising pulmonary vascular resistance
(PVR) in healthy subjects. For Fontan patients with passive pulmonary perfusion,
a rising PVR can cause serious medical problems. The purpose of this fitness to
fly investigation (FTF) is to assess the risk of air travel for children and
adolescents after Fontan palliation.
Methods: We investigated 21 Fontan patients [3–14y] in a normobaric hypoxic
chamber at a simulated altitude of 2,500 m for 3 h. Oxygen saturation, heart
rate, and regional tissue saturation in the forehead (NIRS) were measured
continuously. Before entering the chamber, after 90 and 180 min in the hypoxic
environment, blood gas analysis and echocardiography were performed.
Results: Heart rate and blood pressure did not show significant intraindividual
changes. Capillary oxygen saturation (SaO
2
) decreased significantly after 90 min
by a mean of 5.6 ± 2.87% without further decline. Lactate, pH, base excess, and
tissue saturation in the frontal brain did not reach any critical values. In the case
of open fenestration between the tunnel and the atrium delta, P did not
increase, indicating stable pulmonary artery pressure.
Conclusion: All 21 children finished the investigation successfully without any
adverse events, so flying short distance seems to be safe for most Fontan
patients with good current health status. As the baseline oxygen saturation does
not allow prediction of the maximum extent of desaturation and adaption to a
hypoxic environment takes up to 180 min, the so-called hypoxic challenge test
is not sufficient for these patients. Performing an FTF examination over a period
of 180 min allows for risk assessment and provides safety to the patients and
their families, as well as the airline companies.
KEYWORDS
children and adolescents, Fontan circulation, fitness to fly, hypoxic challenge test,
normobaric hypoxia
1. Introduction
Caused by the unique physiology of passive pulmonary perfusion due to a lacking
subpulmonary ventricle, treating patients with Fontan palliation has several special issues
(1). This must be taken into account for the planning of everyday life and leisure
activities. Recently published data estimated a rising number of people living with Fontan
circulation from approximately 48,000 in 2020 to 60,000 in 2030 including 11 countries
across Europe, USA, Australia, and New Zeeland with an increasing number of adult
patients (55%–64%) (2).
Due to the increasingly better general condition of children after Fontan surgery, the
treating pediatric cardiologists have to take a stand on questions such as permission to
travel by air. In terms of the psychological aspect of living with Fontan circulation and its
TYPE Original Research
PUBLISHED 23 June 2023
|
DOI 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 01 frontiersin.org

association with elevated symptoms of depression as a negative
predictor for quality of life, restrictions should be reduced to a
minimum (3).
The barometric pressure within the cabin of a passenger
aircraft is equivalent to a maximum of 2,438 m above sea level
(masl) (8,000 ft). This is regulated by the European and North
American authorities for normal operating conditions (4). The
barometric pressure at a cruising altitude leads to a lower partial
pressure of inspired oxygen equivalent to 15.2% ambient oxygen,
resulting in lower transcutaneous oxygen saturation (SpO
2
)in
healthy children and adults down to 94.4% (5,6) and 88%–94%
in term and preterm infants (7). Partial oxygen pressure (pO
2
)is
the major regulator of pulmonary vascular tone and a fall in
alveolar pO
2
is the main stimulus for hypoxic pulmonary
vasoconstriction (HPV). A rising pulmonary artery pressure
(PAP) is an early and inevitable consequence of ascent to high
altitude in humans with biventricular hearts (8). The level of
altitude has an inverse relation to arterial oxygen saturation
(SaO
2
) and a direct relationship to the PAP (9).
In Fontan circulation, where the blood flow is maintained
mainly by respiration and the central venous pressure (CVP), a
linear relationship between CVP and pulmonary vascular
resistance (PVR) exists (10). A mild increase in PVR might
already impair cardiac output (CO).
To assess the risk for patients with relevant underlying
diseases before they board an aircraft, hypoxic challenge tests
(HCT) have been established. By breathing oxygen-depleted air
(15.2%) for a period of 5–20 min (7,11–13), usually sitting in
an upright position in a body plethysmograph (11)or
breathing via a face mask (5), an approximate similar inspired
pO
2
to breathing air at cruising altitude can be simulated
under safe conditions.
The British Thoracic Society (BTS) already recommends HCT
for specific constellations to determine whether supplemental in-
flight oxygen is necessary or medical clearance can be given (4).
For Fontan patients, especially for those with a lower baseline
oxygen saturation due to fenestration of the Fontan tunnel or
pronounced collaterals, an evaluation of the pre-flight condition
appears to be beneficial. Spoorenberg et al. investigated children
and adolescents with congenital heart disease performing HCT,
integrating changing body positions and mild physical activity.
In this study, only two patients with Fontan circulation were
included (14). Morimoto et al. exclusively investigated 11
Fontan patients and eight volunteers with a mean age of
22 years under real flight conditions on two-hour flights
measuring percutaneous oxygen saturation (SpO
2
). He described
lower baseline SpO
2
in Fontan patients and a significant
reduction of SpO
2
after ascent and, more importantly, during
cruise after 1 h compared to healthy controls (15). This shows
that not only for children and adolescents with a Fontan
circulation but also ex-preterm babies with or without
bronchopulmonary dysplasia, adults with cyanotic congenital
heart disease, passengers with cardiovascular disease, or patients
with pulmonary hypertension, fitness to fly investigation is an
issue of great importance (5,7,12,16–18) and may be superior
to HCT testing.
The aim of our study was to investigate Fontan patients for
fitness to fly under safe conditions at a simulated altitude in a
normobaric altitude chamber with SpO
2
, capillary blood gas
analysis (BGA), regional brain oximetry on the forehead (rSO
2
),
and echocardiography in order to provide a better assessment of
altitude-related changes over a longer period of time.
2. Materials and methods
2.1. Participants
We investigated 21 children and adolescents after the Fontan
procedure with different types of underlying anatomical heart
diseases (Table 1). Participants were selected from the hospital’s
database and asked whether they were interested in participating
in the study. The patients and their legal guardians received
information material with the specific study design and gave
their written informed consent before participation.
The sample size of 21 participants was chosen as the number of
at least 20 subjects ensures that events (problems and risks) that
occur in at least 10% of cases can be observed at least once in
the study population with a probability of 85%.
Patients >3 years with a minimum interval to the Fontan
operation of 6 months and no contraindications for a stay at
altitude (e.g., severe pulmonary hypertension) were included.
Exclusion criteria were defined as baseline saturation below 85%,
TABLE 1 Patients characteristics, * = age as mean and [min; max].
Fontan
(n = 21)
Age Years 8.7 [3.2; 14.7]*
Gender f/m 7/14
Height cm 131.12 ± 20.29
Body mass kg 29.79 ± 11.83
Body-mass-index kg/m
2
16.68 ± 2.35
Body surface area (Mosteller) m
2
1.03 ± 0.28
Age at Fontan completion Month 36.8 ± 9.7
Years since Fontan surgery Years 5.7 ± 3.2
Patients with implanted
pacemakers
2
Patients with fenestration 7
Underlying cardiac defect I. Functional left ventricle
Tricuspid atresia 3
Double Inlet Left Ventricle
(DILV)
2
Pulmonary atresia 1
II. Functional right ventricle
Hypoplastic left heart syndrome 7
Double Outlet Right Ventricle
(DORV)
4
Pulmonary atresia with TGA 2
III. Indeterminate
Atrioventricular canal defect 1
Shone’s complex 1
Medication I. PDE-5-inhibitors (Sildenafil) 3
II. Cardioselective β-Blocker 6
III. ACE-inhibitors 11
Müller et al. 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 02 frontiersin.org

symptoms of failing Fontan-like plastic bronchitis or protein-losing
enteropathy within the last 6 months, pulmonary hemorrhage
within the last 6 months, acute infection at the time of the
examination, or a withdrawn declaration of consent (by the
patient or their legal guardians). As the topic of fitness to flyis
equally interesting for all families, patients with an open
fenestration in the Fontan tunnel (n= 7) or patients under
therapy with PDE-5 inhibitors (n= 3) were deliberately not
excluded from the study.
Approval was obtained from the local ethics committee of the
University Hospital Bonn (application number 215/19) and the
study corresponded with the Declaration of Helsinki.
The study was performed at the University Children’s Hospital
Bonn, and data were collected from November 2019 to November
2021.
2.2. Study procedure
2.2.1. Study design
A three-hour period was chosen for investigation in terms of
getting stable conditions after adaptation to hypoxia.
Additionally, most European holiday regions can be reached
within this flight distance. All investigations started at nine am in
the morning.
The time points for the evaluation were defined as Normoxia
(outside the chamber): T0 = baseline; Hypoxia (inside the
chamber): T1 = 30 min., T2 = 60 min., T3 = 90 min., T4 =
120 min., T5 =150 min., and T6 = 180 min.; Normoxia (outside
the chamber): T7 = 15 min. after termination of hypoxia.
The patients were encouraged to drink at least one liter of
liquid during the experiment. If necessary, a mobile toilet could
be used in the chamber, so this did not interrupt the exposure to
hypoxia. During the examination, families were asked to behave
as they would if traveling by plane, playing games, reading, or
doing homework. The activity level of the children was tried to
be limited by the parents to a low level. All participants were
asked to avoid sleeping during the experiment as this is already
known to cause a change in breathing patterns in healthy
subjects and thus influences the saturation.
2.2.1.1. Equipment
Continuous SpO
2
and heart rate as well as intermittent blood
pressure were displayed by CARESCAPE V100(GE Healthcare,
Chicago/USA).
Continuous brain oximetry was measured by Near InfraRed
Spectroscopy (NIRS, SenSmart X-100 Universal Oximetry
System, Terumo Cardiovascular, Ann Arbor/USA) with
EQUANOX rSO
2
Optode on the forehead starting right after
entering the chamber.
The blood gases were analyzed with Siemens Healthineers
RAPIDLab
TM
1265 Blood Gas Analyzer and nonheparinized
capillary tubes.
The flight simulation was performed in a fully climatized
normobaric altitude chamber (Höhenbalance, Going/Austria) with
stable temperature conditions between 21 and 22.5°C.
The chamber had 12 m
2
, containing approximately 37.5 m
3
of air.
Hourly, 60 m
3
of air was refreshed. Ambient oxygen was reduced
by membrane processes to a level of 15.2% (±0.2%) and replaced
by nitrogen (84.8%) at constant air pressure (≈1,013 hPa),
simulating an altitude of 2,500 masl (corresponding to Boyle’s
law, estimating a barometric pressure of 760 hPa at 2,500 m).
When CO
2
increased over 0.2%, fresh air was automatically
ducted into the altitude chamber. The hypoxic chamber was
started approximately 2 h in advance to get a stable condition.
The oxygen content in the chamber was measured using an
electrochemical oxygen partial pressure sensor (GOX100, GMH
Messtechnik GmbH, Greisinger, Regenstauf/Germany). Since the
upper part of the chamber consists of glass elements all around,
the examiners can stay outside the chamber, document the
measured values, and observe the patient clinically. The chamber
can be entered through a regular door without relevant changes
in oxygen content inside so that medical staff can switch between
the chamber and the outside area for the necessary examinations.
Echocardiography at T0, T3, and T6 was performed and
analyzed with a Philips IE33 ultrasound system. Velocity time
integral (VTI) was calculated by multiplication with heart rate as
a surrogate of CO (CO = VTI*HR).
2.2.1.2. Baseline data (T0, normoxia)
The baseline data were collected in a separate room at a sufficient
distance from the altitude chamber in ambient air with 21%
oxygen. Heart rate, blood pressure, and SpO
2
were measured in
an upright sitting position followed by the initial
echocardiography in a supine or left lateral position. The sample
for the first capillary blood gas analysis was taken in a sitting or
lying position from the fingertip using safety lancets with a
penetration depth of 0.85 or 1 mm. The optode for rSO
2
(NIRS)
was placed on the forehead.
2.2.1.3. Flight investigation (T1–T6, hypoxia)
After the patient and one accompanying person (mostly parents)
entered the hypoxic chamber, the monitoring was set up and the
three-hour investigation time was started.
Heart rate, blood pressure, transcutaneous oxygen saturation
(SpO
2
), and regional oxygen saturation (rSO
2
) were constantly
measured and documented every 30 min in an upright sitting
position. After 90 (T3) and 180 min (T6) in hypoxia, an
echocardiography was performed inside the chamber, again lying
on the examination couch in supine or left lateral positions. The
measurements started 3 min after changing to a supine position.
Blood gases were drawn while lying or sitting on the
examination couch after the echocardiography was performed.
The investigation ended after 180 min (T6) in the hypoxic
environment by opening the door and stopping the supply of
oxygen-depleted air.
2.2.1.4. Post-hypoxia period (T7)
Another probe for BGA was taken 15 min after the termination of
hypoxia and the removal of all electrodes and measurement
equipment.
Müller et al. 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 03 frontiersin.org

2.2.2. Statistical analysis
GraphPad Prism (V. 9.4.0, GraphPad Software, San Diego, CA/
USA) was used for statistical analysis.
Continuous outcomes were assessed for normality using the
D’Agostino and Pearson test.
Variables fitting a normal distribution were tested on
differences using one-way repeated measures ANOVA and
Dunnett’s multiple comparison tests for post-hoc analysis.
Otherwise, the Friedman test was applied.
Participants with missing data were excluded from the analysis
but were shown in descriptive data. When data were missing, the
number of included values was reported in the tables. For blood
gas analysis, 5% of participants’data were missing at T0, T3, T6,
and 62% at T7.
Simple linear regression was performed for correlation analysis
between parameters, and Pearson’s correlation coefficient was
applied to measure correlation strength.
Quantitative data were presented in mean ± standard deviation
(SD) unless otherwise described.
p-values ≤0.05 (two-sided) were defined as statistically
significant and corresponding values in tables and figures are
presented as “*”.
3. Results
3.1. Clinical parameters
All participants were able to complete the examination. An
early termination due to clinical symptoms was not necessary. A
six-year-old boy was pre-syncopal twice after blood gas analysis
(before entering the chamber and at T3) but also finished the
examination since this was a known problem for him. In
addition, a 12-year-old girl complained of dizziness over a period
of 10 min after 1.5 h of hypoxia. Vital signs were stable at that
time, and the dizziness resolved spontaneously after distraction
and oral fluid intake. The remaining 19 patients did not feel any
changes and did not show any clinical signs or make complaints
of symptoms. At the end of the hypoxia period, the children
under 10 years of age, in particular, were exhausted and tired,
which was manifested by resentment, crying, and clinical
symptoms such as yawning and lack of movement. Due to the
young age of the participants and the large age range, an
objectification of this condition by means of a questionnaire or
similar was not reasonably possible but could be evaluated in
conversation with the parents.
All measured clinical parameters are listed in Table 2A. Heart
rate and blood pressure did not show significant intraindividual
changes. SpO
2
dropped significantly after entering the hypoxic
environment with a nadir 30 min after the start of altitude
exposure (mean delta T0-T1: 6 ± 3.8%). All patients showed an
increase in peripheral oxygen saturation during the course after
the initial drop without changes in the conditions (mean delta
T1–T6: 2 ± 2.4%) (Figure 1A).
The youngest patient with 7 months after Fontan surgery
(without fenestration but with a large IVC to atrium shunt) and
the oldest patient, 11 years post-Fontan completion (with
fenestration) showed the highest drop in oxygen saturation with
a loss of 15% from T0 to T1. The young one started with SpO
2
of 88% before entering the chamber then dropping to 73% at T2
and recovering after T3 with peripheral saturation values around
80%–82%. In the blood gas analysis, a SaO
2
of 81.2% was
measured at T3. The patient’s clinical condition was unimpaired
at all times. The oldest patient started with 96% SpO
2
,
desaturated to 81% at T1, again recovering after T3 with 86%–
87%. The SaO
2
at T3 showed significantly higher values of 88.9%.
In contrast, in all patients, a constant tissue saturation of the
brain could be derived via the NIRS optodes on the forehead
also, with a trend of improvement after 60 min. in hypoxia
(Figure 1B).
There was no correlation between SaO
2
and HR or rSO
2
.
3.2. Blood gas analysis
Repeated capillary blood gases from the finger were tolerated
even by the smaller patients. At T7, most children were desperate
TABLE 2 (A) clinical parameters; (B) blood gas analysis; p-value 1(T0–T3), p-value 2 (T0–T6), p-value 3 (T0–T7).
(A) T0 T1 T2 T3 T4 T5 T6
Heart rate [bpm] 86 ± 15 88 ± 13 95 ± 14 90 ± 14 97 ± 13 97 ± 13 94 ± 18
SpO
2
[%] 93 ± 4 86 ± 5 86 ± 4 87 ± 5 86 ± 4 87 ± 4 88 ± 4
rSO
2
[%] / 70±7 70±8 72±8 71±7 71±7 72±7
RRsys [mmHg] 108 ± 15 106 ± 16 109 ± 14 111 ± 13 110 ± 11 107 ± 11 110 ± 14
RRdia [mmHg] 63 ± 9 62 ± 11 62 ± 9 68 ± 11 63 ± 9 62 ± 12 65 ± 12
(B) T0 (n= 20) T3 (n= 20) T6 (n= 20) T7 (n=8) p-value 1 p-value 2 p-value 3
pH 7.44 ± 0.02 7.45 ± 0.02 7.45 ± 0.02 7.44 ± 0.03 ns 0.018* ns
pCO
2
[mmHg] 32.25 ± 6.8 29.13 ± 2.2 28.98 ± 2.51 29.2 ± 2.87 0.001* 0.002* ns
pO
2
[mmHg] 59.87 ± 7.99 49.98 ± 4.37 49.6 ± 4.59 59.59 ± 9.19 <0.001* <0.001* ns
HCO
3
act [mmol/l] 20.41 ± 1.13 19.87 ± 1.27 19.9 ± 1.31 19.29 ± 1.37 0.008* 0.016* 0.023*
HCO
3
std [mmol/l] 22.09± 0.71 21.92 ± 1.1 21.92 ± 0.89 21.38 ± 1.06 ns ns ns
Base Excess [mmol/l] −2.68 ± 0.83 −2.93 ± 1.1 −2.72 ± 1.07 −3.6 ± 1.25 ns ns ns
Lactate [mmol/l] 1.69 ± 0.37 2.18 ± 0.64 2.14 ± 0.55 2.3 ± 0.83 0.011* 0.029* ns
Hemoglobin [g/dl] 13.55 ± 1.13 13.68 ± 1.12 13.61 ± 1.25 13.43 ± 1.6 0.049* ns ns
SaO
2
[%] 90 ± 4 85 ± 4 85 ± 5 91 ± 4 <0.001* <0.001* ns
* indicates statistical significance.
Müller et al. 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 04 frontiersin.org

to finish the investigation and did not allow a fourth capillary blood
sampling. In order not to traumatize them and associate flying with
negative experiences, sampling was not encouraged (Table 2B).
In the blood gas analysis, capillary oxygen saturation after
90 min showed a significant decrease by a mean of 5.6 ± 2.87%
but did not drop further during the course in the majority of the
children (Figure 2). pO
2
, on the other hand, showed a further
decline from T3 to T6 with complete recovery at T7 after exiting
the hypoxic environment (Table 2B).
All over, the mean difference between the transcutaneous SpO
2
and the capillary SaO
2
was −2.6 ± 2.9%. A good correlation with
SpO
2
≤0.0001, R
2
= 0.6715 could be shown.
CO
2
decreased while base excess, bicarbonate, and pH
remained stable without significant changes. Lactate increased
significantly from T0 to T3, without an additional rise at T6 and
a further but non-significant increase up to T7 (here with a
greater range) (Table 2B).
3.3. Echocardiography
Table 3 highlights the parameters being influenced by hypoxia.
IVC Vmax, IVC MPG, Vmax at aortic arch, and E/É medial and
lateral were also recorded but not presented due to lack of
relevance to the research question or incomplete data sets caused
by the very different anatomical conditions.
Regarding the AV-valve inflow, significant changes could be
monitored with a reduction in AV E Vmax and a lacking
decrease of the a-wave leading to a reduction of AV E/A
indicating diastolic function.
The VTI as a marker for cardiac function showed a slight but
not significant decrease. CO did not change significantly from T0
to T6 (Table 3).
In seven patients with an open fenestration in the Fontan
tunnel, delta P between the venous system and the atrium could
be measured. Hypoxia did not lead to a relevant increase in delta
P as an indirect indicator of an acute increase in PAP. A clear
correlation between peripheral oxygen saturation and delta P
could not be shown (Figure 3).
4. Discussion
To our knowledge, this is the first study performing a complex
fitness-to-fly investigation in a normobaric hypoxia chamber
simulating flight conditions in children and adolescents after
completing the Fontan procedure with a duration of 3 h.
All patients finished the investigation in good clinical condition
and all measured parameters did not show any contraindication for
air traveling within a three-hour range for this group. To date, there
are no standard values for Fontan patients that can be used for a
safe stay on board a passenger aircraft.
Previous studies have already addressed this issue in terms of
hypoxic challenge tests (HCT) for different groups of patients.
Some were already summarized in reviews by Spoorenberg et al.
for children with congenital heart and lung disease in 2016 (19),
by Balfour-Lynn et al. for children with lung disease in 2016
(20), and by Herberg et al. for children with pulmonary
hypertension in 2020 (18). With a test duration of 5–20 min,
these are significantly shorter than the data presented here.
Considering the underlying physiology of altitude adaptation, the
selected time period seems too short in most cases.
Acclimatization to a hypoxic environment has at least two
FIGURE 1
(A) Peripheral oxygenation (finger); (B) tissue oxygenation (NIRS
forehead) at different time points (mean ± standard deviation).
FIGURE 2
Capillary SaO
2
,*=p< 0.05 in relation to T0.
Müller et al. 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 05 frontiersin.org

phases. The initial constrictor phase takes several minutes and is
followed by a sustained phase. In total, it takes about 30–120 min
for the PVR to stabilize. Factors such as neurohumoral
mediators, red blood cells, and lung innervation may also
influence the response (8). For patients with Fontan physiology,
the PVR is one of the central mechanisms for functioning or
failing circulation as they lack the subpulmonary ventricle.
Accordingly, they have no way to compensate for increasing
pulmonary vascular resistance with mechanisms known from
healthy subjects (21). If the test duration is too short, the
adaptation processes are not completed, and, therefore, a reliable
statement on the airworthiness of Fontan patients using HCT
cannot be made. The direct measurement of PVR is only
possible with invasive methods like a central venous line or in
the catheter laboratory so that the interpretation of indirect
findings like SaO
2
, SpO
2
, rSO
2
, and lactate as well as
echocardiographic values are needed to assess the clinical
situation and a potential risk for the patients.
In most HCT investigations, patients are breathing oxygen-
reduced air (15.0–15.2%) via a face-mask (13,22) or sitting in a
sealed body plethysmograph (7,11,12). This setting indeed
seems not feasible for a test duration of more than 30 min.
In 2016, Spoorenberg et al. published fitness-to-fly data for
patients with congenital heart disease including changing body
positions (seated, lying supine, and standing) combined with
different activity levels (walking at 3 and 5 km/h). The test
duration here was 40 min. He involved two Fontan patients,
again breathing oxygen-depleted air via a face mask (13). Since
regular fluid intake is critical for good circulatory function,
especially for Fontan patients, a test duration of >30 min would
not be practical with such an experimental setup. At the current
time, there is no study that shows a comparable structure and
complexity over a three-hour investigation period.
The basis for this type of investigation is recommendations, i.e.,
from the BTS for a fitness-to-fly assessment to provide practical
advice for specialists dealing with patients with lung diseases and
the still existing uncertainty of families entrusted to their care
(4). For patients with congenital heart defects, we do not have
any existing guidelines.
Takken et al., Staempfli et al., Garcia et al., and Müller et al.
have investigated the effects of altitude on the Fontan circulation
but all in connection with physical exertion and thus set a
different focus. Nevertheless, these studies may help to
better frame the baseline data and are therefore included
here (23–26).
4.1. Basic monitoring
Since heart rate is influenced by many different factors, a slight
increase together with stable blood pressure in all of our patients
during the whole investigation time can be interpreted as a stable
condition of the patients. There were no signs of physical
exhaustion and a relevant need for increased cardiac output. The
broad distribution for heart rate is due to the wide age spectrum
(3.2–14.7y) and did not correlate with SaO
2
changes.
4.2. Oxygen saturation and blood gas
analysis
The average baseline saturation in the present young Fontan
cohort was 93 ± 4%. This is in line with data from Takken et al.
(94.6% ± 2.9), Staempfli et al. (90% ± 4) at 504 masl, Müller et al.
(92% ± 4) (23,24,26), and Morimoto et al. (93; 88%–96%), with
a similar range of baseline saturation (15). Compared to previous
fitness-to-fly data from patients with different underlying
diseases, the baseline saturation of Fontan patients is significantly
lower [Boosley: SpO
2
100% in 71 term and ex-preterm babies
(12); Vetter-Laracy: SpO
2
99% in 119 babies, term and preterm
with and without BPD; Oades: 97.1% in 22 children with cystic
fibrosis (27)]. The lower level of oxygen saturation in Fontan
TABLE 3 Echocardiographic parameters; p-value 1 (T0–T3), p-value 2 (T0–T6), * = p< 0.05.
T0 T3 T6 p-value 1 p-value 2
ΔP
fenestration
[mmHg] 5.4 ± 1.1 (7) 5.7 ± 1.2 (6) 5.7 ± 1.7 (7) ns ns
VTI
A0
23 ± 4.2 22.4 ± 4.3 21.6 ± 3.9 ns ns
CO [VTI × HR] 1,920 ± 392 2,016 ± 463 2,027 ± 557 ns ns
AV E Vmax [cm/s] 87.20 ± 17 78.10 ± 25.90 74.40 ± 16.10 ns 0.006*
AV A Vmax [cm/s] 57.40 ± 14.70 56.60 ± 12.5 55.40 ± 13.40 ns ns
AV E/A 1.6 ± 0.3 1.5 ± 0.3 1.4 ± 0.2 ns 0.043*
FIGURE 3
Relationship between SaO
2
and delta P (Fontan tunnel to atrium).
Müller et al. 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 06 frontiersin.org

patients can be explained by fenestration of the tunnel, severe veno-
venous collaterals, and the draining of the coronary sinus in the
common atrium, all leading to a desaturation under normoxia
and hypoxia (28).
Looking at the course of peripheral oxygen saturation, the
largest drop occurs within the first 30 min after hypoxia onset
(86% ± 4%). As the process continues, the saturation then
stabilizes at a somewhat higher level (Figure 1) but still below
baseline saturation. Morimoto reports SpO
2
values of 88% (78%–
95%) at the top of climbing, approximately 20 min after take-off,
and a minimum of 86% (78%–95%) after one hour under real
flight conditions (hypobaric hypoxia). With HCT investigation
times of a maximum of 20 min, these processes cannot be
captured and might lead to incorrect conclusions like erroneous
bans for patients.
As the drop of SpO
2
did not show a strong correlation with
baseline values, predictability of flight suitability exclusively based
on the SpO
2
at ground level is considerably difficult.
Looking at the capillary saturation measured at 0, 90, and
180 min in hypoxia, there is a difference of up to ±4.7 points
compared to SpO
2
values, probably due to a poor signal caused
by restricted peripheral circulation and, consequently, cold
hands in some patients and a technically induced deviation
between SaO2 and SpO2 from approximately 2%–3% (29).
Since this is a known problem, capillary SaO
2
is certainly the
more reliable method. With a correlation of R
2
= 0.672 between
both methods, SpO
2
still is a good parameter as long as the
signal is adequate.
The SaO
2
values cannot confirm the slight upward trend after
initial habituation detected by the continuous oximetry. It remains
stable in the follow-up after 180 min without a further decrease.
Regardless of all this, stable tissue saturation in the forehead
could be documented by NIRS in all patients at any time without
significant changes. As expected, there was no correlation to
SaO
2
, indicating safe conditions and stable hemodynamics with
the maintenance of perfusion of central organ systems even in
times of lower SpO
2
(R
2
= 0.035).
As a rule, it takes 20 min for a passenger aircraft to reach the
maximum cruising altitude. During the ascent there is a slow
decrease in pO
2
, so that adaptation to the hypoxic environment
can take place slowly. The presented study and the above-cited
HCT studies except those by Morimoto et al. expose the
participants to 15.0%–15.2% oxygen immediately without any
adaptation due to the study setup with normobaric simulated
altitude. This might make a difference in oxygen saturation,
especially within the first hour. Whether this leads to deviating
kinetics of the oxygen saturation must be clarified in a follow-up
study.
4.2.1. Lactate and acid-base balance
Rising lactate is an expression of anaerobic metabolic processes,
which occurs due to desaturation in the tissue. The assessment of
capillary lactate values in our cohort shows an initial increase at
T3. This seems to be an expression of acute hypoxia-induced
stress. Probably as a sign of stabilization of hemodynamics in the
course, lactate does not increase further until T6, comparable to
SaO
2
levels (Table 2B). At T7, no drop in lactate could be
observed (n= 8). This is most likely due to the short time
interval to the end of hypoxia and delayed leaching due to the
higher CVP and accumulation of lactate in the tissues over the
three-hour period even with only mild desaturation. Stable values
for pH, bicarbonate, base excess, and CO on the other hand
indicate balanced circulatory conditions at all times. The partially
statistically significant differences in Table 2B are not relevant to
the clinical condition of the patients.
Stable hemoglobin values in the majority of the patients
indicate sufficient drinking quantity during the examination. As
described by Zubac et al., the microclimate of airline cabins is
dry with a relative humidity of 10%–20% at cruising altitude, and
especially in patients with Fontan circulation, this must be
addressed before boarding an airplane, as dehydration can lead
to a deterioration in circulation and oxygen saturation causing
avoidable complications on board (30). This aspect is not
considered in the HCT due to the short examination time.
However, liquid intake would not be technically possible when
breathing through a facemask.
4.3. Echocardiography and pulmonary
vascular resistance
All patients underwent three echocardiograms during the
course and neither showed relevant impairment of systemic
ventricle function nor had indirect signs of increasing PVR like
rising lactate (Table 3). Whether this can also be applied to
older patients with Fontan circulation needs to be investigated in
further studies. Qunate et al. recently published the first
echocardiography data for older Fontan patients (20.5 years; ±5.4
years) under the same normobaric hypoxia with 15.2% oxygen,
comparing normoxia vs. hypoxia at rest and after peak and
continuous exercise (31). To our knowledge, no
echocardiography parameters have been published from Fontan
patients at that young age under hypoxic conditions over a
period of 3 h.
4.3.1. Cardiac output
Cardiac function was estimated via VTI and did not show a
significant decrease (Table 3). To describe cardiac output, VTI
can be multiplied by heart rate as done by Quante et al. They
described a decrease in CO by −12% at rest (30 min of adaption
sitting on a chair in the hypoxic chamber under the same
condition), compared to measurements in normoxia 2 weeks
before. In our present younger study group, CO was stable
without significant changes over the 3 h study period. Compared
to Quante (30 min), the first echocardiography in this study was
performed after 90 min in the hypoxic environment. Whether a
drop in HZV also occurred within the first 30 min in our cohort
and then a recovery until the first echocardiography was not
investigated here. Anyhow, this can also be used as an argument
for a longer examination duration, especially for Fontan
patients. Besides this, it must also be seen in the context of
acute exposure due to the study design with full altitude
Müller et al. 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 07 frontiersin.org

exposure beginning immediately. Comparable with the
development of SpO
2
values, the actual conditions in the aircraft
with a slow climb to the final cruising altitude are probably
better compensated. However, in conjunction with the lactate
values, there is no evidence of an acute hemodynamic problem.
In addition, stable CO and lactate values indicate adequate fluid
intake during the study period. The broad distribution in
standard deviation can be explained by the varying age of the
patients and the different anatomical conditions of the ventricles
causing a varying CO.
4.3.2. Diastolic function
Quante et al. described a missing augmentation of atrial
contraction with a drop of e-wave and a lack of decreasing a-
wave. This could be confirmed in our younger cohort and must
be interpreted as an expression of hypoxia-related mild diastolic
dysfunction. If a long-term stay at a high altitude would lead to
an aggravation of ventricular dysfunction or an adaptation with
normalization of the values must be clarified. In healthy
individuals, it is known that on acute exposure to hypoxia,
ventricular diastolic dysfunction is usually prevented by an
augmentation of atrial contraction (32).
4.3.3. Pulmonary artery pressure
As already addressed, the PAP plays a key role that determines
the function or failure of the Fontan circuit. An estimation of the
PAP in healthy individuals with echocardiography is possible by
measuring the systolic right ventricular to the right atrial
pressure gradient. This was shown by Allemann et al., who
investigated 118 healthy children and adolescents 40 h after rapid
ascend to 3,450 m. They found a more than two-fold higher PAP
compared to a low altitude and an inverse relation to age,
resulting in a two-fold larger increase in 6–9 than in 14 to 16-
year-old participants (33).
The noninvasive echocardiography estimation of PAP in
Fontan patients only works if there is a fenestration between the
tunnel and the atrium. In that case, the transpulmonary gradient
can be measured via the delta P. As shown by Bouhout et al. in
a meta-analysis, a fenestration can effectively reduce pulmonary
pressure in the early postoperative period in Fontan patients
(34). Whether this could also be helpful in the further course,
e.g., for an acute, hypoxia-related change in PVR in connection
with high altitude or air travel is conceivable but needs further
investigation.
A significant increase in delta P could be excluded by
echocardiography in the seven patients with present fenestration
in our study. Further, there was no correlation between SaO
2
and
delta P (Figure 3). To what extent this is an indication of
worsening diastolic function with an increase in atrial pressure or
a lack of rising pulmonary resistance causing this remains an
interesting question that cannot be answered with certainty
without an invasive measurement of PAP. It is conceivable that
veno-venous collaterals also play a role in the sense of a pressure
relief of the venous system, especially in the case of above-
average desaturation under hypoxia.
The echocardiographic findings are very valuable, underlining
the good clinical conditions and subjective well-being of the
patients.
5. Conclusion
All 21 children finished the investigation without any adverse
events.
Afitness-to-fly investigation for patients with Fontan
physiology can help to categorize the risk profile and might
prevent emergency situations during air travel. The baseline
oxygen saturation is not a good marker for the maximum extent
of saturation decay in the hypoxic environment of a passenger
aircraft.
Regardless of the low oxygen saturation, the patients did not
complain of any subjective problems. Objective criteria such as
lactate, pH, and base excess as well as tissue saturation in the
frontal brain did not show any critical values. The altitude
equivalent of 2,500 masl does not lead to relevant impairment in
ventricular function. An adequate increase in cardiac output is
possible under these conditions. The PVR has a key role in the
Fontan circulation, and with regard to physiological adaptation
processes, a Hypoxic Challenge Test with a duration of less than
30 min hypoxic exposure might not be sufficient for this special
patient group.
From the results of the 21 children examined, it is currently
not possible to make any reliable general statement. The flight
fitness of Fontan patients, therefore, remains an individual
decision of the treating cardiac center together with the
patients and their families on the basis of their current state of
health. Establishing a sufficient fitness-to-flyprotocolfor
patients with Fontan circulation could help to provide safety
for patients, caregivers, and healthcare professionals as well as
Airline companies.
6. Limitations
Since the investigations took place in a normobaric hypoxia
chamber, all results are based on simulated altitude, which is
assumed to be equal to real altitude conditions for this special
issue. The small number of patients does not allow a
fundamental statement for all patients with Fontan circulation.
Due to logistic reasons, the children were exposed to a sudden
onset of hypoxia, induced by entering the chamber, with 15.2%
oxygen content. This does not correspond to the duration of the
climb of a passenger aircraft, which lasts on average up to
20 min and thus enables a slow adaptation to the falling effective
oxygen content.
Müller et al. 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 08 frontiersin.org

Data availability statement
The raw data supporting the conclusions of this article will be
made available by the authors, without undue reservation.
Ethics statement
The studies involving human participants were reviewed and
approved by Ethics committee of the University Hospital Bonn
(application number 215/19). Written informed consent to
participate in this study was provided by the participants’legal
guardian/next of kin.
Author contributions
Conceptualization NM, UH, JB, and JH. Formal analysis NM
and JH. Investigation NM, UH, TK, and JH. Writing-original
draft NM and JH. Writing-review and editing UH, JB, and TK.
All authors contributed to the article and approved the
submitted version.
Acknowledgments
We thank our study nurse Ute Baur for her help and
support.
Conflict of interest
The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could
be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their affiliated
organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article, or
claim that may be made by its manufacturer, is not guaranteed
or endorsed by the publisher.
References
1. Zentner D, Celermajer DS, Gentles T, d’Udekem Y, Ayer J, Blue GM, et al.
Management of people with a Fontan circulation: a cardiac society of Australia and
New Zealand position statement. Hear Lung Circ. (2020) 29:5–39. doi: 10.1016/j.hlc.
2019.09.010
2. Plappert L, Edwards S, Senatore A, De Martini A. The epidemiology of persons
living with Fontan in 2020 and projections for 2030: development of an
epidemiology model providing multinational estimates. Adv Ther. (2022)
39:1004–15. doi: 10.1007/s12325-021-02002-3
3. Rychik J, Atz AM, Celermajer DS, Deal BJ, Gatzoulis MA, Gewillig MH, et al.
Evaluation and management of the child and adult with Fontan circulation: a
scientific statement from the American heart association. Circulation. (2019) 140:
e234–84. doi: 10.1161/CIR.0000000000000696
4. Coker RK, Armstrong A, Church AC, Holmes S, Naylor J, Pike K, et al. BTS
clinical statement on air travel for passengers with respiratory disease. Thorax.
(2022) 77:329–50. doi: 10.1136/thoraxjnl-2021-218110
5. Withers A, Wilson AC, Hall GL. Air travel and the risks of hypoxia in children.
Paediatr Respir Rev. (2011) 12:271–6. doi: 10.1016/j.prrv.2011.02.002
6. Lee AP, Yamamoto LG, Relles NL. Commercial airline travel decreases oxygen
saturation in children. Pediatr Emerg Care. (2002) 18:78–80. doi: 10.1097/00006565-
200204000-00004
7. Vetter-Laracy S, Osona B, Peña-Zarza JA, Gil JA, Figuerola J. Hypoxia challenge
testing in neonates for fitness to fly. Pediatrics. (2016) 137:e20152915. doi: 10.1542/
peds.2015-2915
8. Wilkins MR, Ghofrani HA, Weissmann N, Aldashev A, Zhao L. Pathophysiology
and treatment of high-altitude pulmonary vascular disease. Circulation. (2015)
131:582–90. doi: 10.1161/CIRCULATIONAHA.114.006977
9. Penaloza D, Arias-Stella J. The heart and pulmonary circulation at high altitudes:
healthy highlanders and chronic mountain sickness. Circulation. (2007) 115:1132–46.
doi: 10.1161/CIRCULATIONAHA.106.624544
10. Becker K, Uebing A, Hansen JH. Pulmonary vascular disease in Fontan
circulation-is there a rationale for pulmonary vasodilator therapies? Cardiovasc
Diagn Ther. (2021) 11:1111–21. doi: 10.21037/cdt-20-431
11. Naqvi N, Doughty VL, Starling L, Franklin RC, Ward S, Daubeney PEF, et al.
Hypoxic challenge testing (fitness to fly) in children with complex congenital heart
disease. Heart. (2018) 104:1333–8. doi: 10.1136/heartjnl-2017-312753
12. Bossley CJ, Cramer D, Mason B, Hayward A, Smyth J, McKee A, et al.
Fitness to fly testing in term and ex-preterm babies without bronchopulmonary
dysplasia. Arch Dis Child Fetal Neonatal Ed. (2012) 97:199–203. doi: 10.1136/adc.
2011.212001
13. Spoorenberg ME, Hulzebos EHJ, Takken T. Feasibility of hypoxic challenge
testing in children and adolescents with congenital heart and lung disease. Aerosp
Med Hum Perform. (2016) 87:1004–9. doi: 10.3357/AMHP.4580.2016
14. Spoorenberg ME, Hulzebos EHJ, Takken T. Feasibility of hypoxic challenge
testing in children and adolescents with congenital heart and lung disease. Aerosp
Med Hum Perform. (2017) 87:1004–9. doi: 10.3357/amhp.4580.2016
15. Morimoto Y, Ohuchi H, Kurosaki K, Nakai M. Exercise-induced hypoxia
predicts hypobaric hypoxia during flight in patients after Fontan operation. Int
J Cardiol. (2021) 325:51–5. doi: 10.1016/j.ijcard.2020.09.069
16. Harinck E, Hutter PA, Hoorntje TM, Simons M, Benatar AA, Fischer JC, et al.
Air travel and adults with cyanotic congenital heart disease. Circulation. (1996)
93:272–6. doi: 10.1161/01.CIR.93.2.272
17. Smith D, Toff W, Joy M, Dowdall N, Johnston R, Clark L, et al. Fitness to fly for
passengers with cardiovascular disease. Heart. (2010) 96(Suppl 2):10. doi: 1136/hrt.
2010.203091
18. Herberg U, Knies R, Müller N, Breuer J. Altitude exposure in pediatric
pulmonary hypertension-are we ready for (flight) recommendations? Cardiovasc
Diagn Ther. (2021) 11:1122–36. doi: 10.21037/cdt-20-494
19. Spoorenberg ME, van den Oord MHAH, Meeuwsen T, Takken T. Fitness to fly
testing in patients with congenital heart and lung disease. Aerosp Med Hum Perform.
(2015) 87:54–60. doi: 10.3357/amhp.4408.2016
20. Balfour-Lynn IM. Hypoxic challenge test for airflight in children with respiratory
disease. Paediatr Respir Rev. (2017) 21:62–4. doi: 10.1016/j.prrv.2016.05.002
21. Netzer NC, Strohl KP, Högel J, Gatterer H, Schilz R. Right ventricle dimensions
and function in response to acute hypoxia in healthy human subjects. Acta Physiol.
(2017) 219:478–85. doi: 10.1111/apha.12740
22. Kobbernagel HE, Nielsen KG, Hanel B. Hypoxic challenge test applied to healthy
children: influence of body positions and exertion on pulse oximetric saturation. Arch
Dis Child. (2013) 98:602–6. doi: 10.1136/archdischild-2012-302763
23. Takken T, Evertse A, de Waard F, Spoorenburg M, Kuijpers M, Schroer C, et al.
Exercise responses in children and adults with a Fontan circulation at simulated
altitude. Congenit Heart Dis. (2019) 14:1005–12. doi: 10.1111/chd.12850
24. Staempfli R, Schmid JP, Schenker S, Eser P, Trachsel LD, Deluigi C, et al.
Cardiopulmonary adaptation to short-term high altitude exposure in adult Fontan
patients. Heart. (2016) 102:1296–301. doi: 10.1136/heartjnl-2016-309682
25. Garcia JA, McMinn SB, Zuckerman JH, Fixler DE, Levine BD. The role of the
right ventricle during hypobaric hypoxic exercise: insights from patients after the
Fontan operation. Med Sci Sports Exerc. (1999) 31:269–76. doi: 10.1097/00005768-
199902000-00011
Müller et al. 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 09 frontiersin.org

26. Müller N, Herberg U, Jung T, Breuer J, Härtel JA. Adequate Exercise
Response at Artificial Altitude in Fontan Patients. Front Pediatr. (2022) 10:1–13.
doi: 10.3389/fped.2022.947433
27. Oades P, Buchdahl R, Bush A. Predictions of hypoxaemia at high altitude in
children with cystic fibrosis. Br Med J. (1994) 308:15–8. doi: 10.1136/bmj.317.
7161.780
28. Gewillig M. The fontan circulation. Heart. (2005) 91:839–46. doi: 10.1136/hrt.
2004.051789
29. Jubran A. Pulse oximetry. Crit Care. (2015) 19:1–7. doi: 10.1186/s13054-015-
0984-8
30. Zubac D, Stella AB, Morrison SA. Up in the air: evidence of dehydration risk and
long-haul flight on athletic performance. Nutrients. (2020) 12:1–15. doi: 10.3390/
nu12092574
31. Quante H, Müller N, Härtel JA, Jung T, Manunzio U, Breuer J, et al. Systemic
ventricular function in Fontan patients at rest and after exercise at altitude. Front
Pediatr. (2023) 10:1–11. doi: 10.3389/fped.2022.1084468
32. Allemann Y, Rotter M, Hutter D, Lipp E, Sartori C, Scherrer U, et al. Impact of
acute hypoxic pulmonary hypertension on LV diastolic function in healthy
mountaineers at high altitude. Am J Physiol Hear Circ Physiol. (2004) 286:856–62.
doi: 10.1152/ajpheart.00518.2003
33. Allemann Y, Stuber T, de Marchi SF, Rexhaj E, Sartori C, Scherrer U, et al.
Pulmonary artery pressure and cardiac function in children and adolescents after
rapid ascent to 3,450 m. Am J Physiol Hear Circ Physiol. (2012) 302:2646–53.
doi: 10.1152/ajpheart.00053.2012
34. Bouhout I, Ben-Ali W, Khalaf D, Raboisson MJ, Poirier N. Effect of fenestration
on Fontan procedure outcomes: a meta-analysis and review. Ann Thorac Surg. (2020)
109:1467–74. doi: 10.1016/j.athoracsur.2019.12.020
Müller et al. 10.3389/fcvm.2023.1170275
Frontiers in Cardiovascular Medicine 10 frontiersin.org