See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/23483373

Congenital diaphragmatic hernia: A modern

day approach

Article in Seminars in Pediatric Surgery · December 2008

DOI: 10.1053/j.sempedsurg.2008.07.009 · Source: PubMed

CITATIONS READS

50 846

9 authors, including:

Katrin Zahn Regine Schaffelder

Universitätsmedizin Mannheim Universitätsmedizin Mannheim
32 PUBLICATIONS 220 CITATIONS 32 PUBLICATIONS 114 CITATIONS

SEE PROFILE SEE PROFILE

Thomas Schaible
Universitätsmedizin Mannheim
227 PUBLICATIONS 2,001 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Mechanical ventilation and neonatal ARDS View project

Neonatal Intensive Care Unit Structure View project

All content following this page was uploaded by Regine Schaffelder on 08 September 2017.

The user has requested enhancement of the downloaded file.

Seminars in Pediatric Surgery (2008) 17, 244-254

Congenital diaphragmatic hernia: a modern day approach

Karl-Ludwig Waag, MD,a Steffan Loff, MD,a Katrin Zahn, Mansour Ali, MD,a
Steffen Hien, MD,b Markus Kratz, MD,b Wolfgang Neff, MD,c
Regine Schaffelder, MD,d Thomas Schaible, MDb

From the Department of aPediatric Surgery, University Hospital, Mannheim, Germany;

b
Department of Pediatrics, University Hospital, Mannheim, Germany;
c
Department of Radiology, University Hospital, Mannheim, Germany; and the
d
Department of Obstetrics, University Hospital, Mannheim, Germany.

KEYWORDS Centralization of all complicated congenital diaphragmatic hernias (CDH) was organized in Germany from
Congenital 1998, collecting 325 consecutive patients with striking increasing survival rates. This series report 244
diaphragmatic hernia; patients from 2002 to 2007. Today, large defects are detected early in pregnancy by ultrasound and magnetic
CDH; resonance imaging (MRI). In extracorporeal membrane oxygenation (ECMO) patients, prenatal lung head
Prognostic factors; ratio (LHR) was 1.2 (median) at the 34th week of gestation or less than 25 ml lung tissue in MRI. This means
ECMO; that all patients below LHR of 1.4 should be transferred prenatally in a tertiary center. High risk group for
Treatment survival was defined as LHR below 0.9, ie, 10 ml in MRI planimetry. Inborn patients show better results than
outborns. In algorithm therapy, gentle ventilation plays an important role in preventing damage to the lung
tissue and avoiding long term ventilation. When PaCO2 was more than 75 mmHg, ventilation was changed
to high frequency oscillatory ventilation (HFOV). Indication for ECMO was seen in preductal PaO2 less than
50 mmHg over 2-4 h or less than 40 mmHg over 2 h. ECMO related risks included intracerebral bleeding
(9%), intrapulmonary bleeding (14%), and convulsions (16%). Surgically, a longitudinal midline incision for
exposure of the defect, the duodenal kinking, and probably for abdominal patching was perfect. A cone
formed goretex patch provided more abdominal space and reduced abundant intrathoracical cavity. No drain
was used. Postoperatve complications were described. Overall survival in 244 consecutive patients was
86.5% for all patients born alive. All those who needed ECMO survived in 71%, underlining ECMO as a
treatment of last choice. Follow-up for quality of life after CDH is described.
© 2008 Published by Elsevier Inc.

Embryologically, the diaphragm develops during the
8th and 12th weeks of gestation. A membrane (septum
transversum) separates the thoracic from the abdominal
cavity. Septum transversum growth starts anteriorly and
dorsally. The pleurocardial space is then divided from the
peritoneal area. These pleuroperitoneal folds consist pri-
marily of pleural membrane and peritoneum. Later, mus-
Address reprint requests and correspondence: Karl-Ludwig Waag,
MD, University Hospital, Department of Pediatric Surgery, Theodor
Kutzer Ufer 1-3, Mannheim 68167, Germany.
E-mail: karl-ludwig.waag@kch.ma.uni-heidelberg.de.
cle fibers migrate into this membrane from anterior to
posterior. Due to the late closure of this membrane on the
left side, left diaphragmatic hernias are more common
(87% left versus 11% right, 2% are bilateral) and are
predominantly posterior. The last to close is called the
pleuroperitoneal canal.
Dorsal defects are called Bochdalek hernia, and anterior
defects are named Morgagni hernia. Larrey described a rare
form of sternocostal defect.
Esophageal lengthening occurs at the same time as the
growth of the septum transversum. Any delay results in a
wide open hiatus and short esophagus.

1055-8586/$ -see front matter © 2008 Published by Elsevier Inc.

doi:10.1053/j.sempedsurg.2008.07.009
Waag et al. Congenital Diaphragmatic Hernia
Figure 1 Specimens of the lung in control and CDH, showing
rarification of all bronchi and hypoplasia. (Color version of figure
is available online.)
When there is a defect in muscle fiber growth and fibers
do not move into the pleuroperitoneal membrane suffi-
ciently, the result will be a true diaphragmatic hernia. In
these cases, there is a hernial sac, or in case of hypoplastic
muscle fibers, a congenital diaphragmatic eventration. The
timing of the development of the defect is important be-
cause of the duration of compressing the ipsi- and contralat-
eral lung. In large defects, the intrathoracic position of the
left liver lobe, stomach, spleen, and gut leads to early
compression of lung tissue and mediastinal shift. These are
the cases that are detected early in pregnancy.
Figure 2 Distribution of muscular fibers in the wall of pulmo-
nary vessels normally and in CDH. (Color version of figure is
available online.)
245
Figure 3 Different thickness of muscle fibers under normal, CDH,
and ECMO conditions.
Etiology
Most defects develop as isolated defects, ie, sporadically
(70%). There are other factors involved in about 20%
detected on experiments, such as in relation to medica-
tions like thalidomide or nitrofen and vitamin-A defi-
ciency. Diaphragmatic defects are seen as well in trisomy
13, 18, or 21 in 8%, such as in Beckwith Wiedemann and
Danny Drash syndrome. Familial conditions are rare
(2%). Here, chromosomal deviations are located on chro-
mosome 15q26.
Gene mutation is seen in deletion of 4p, 8q, 15q or as
duplication of 8p and as tetrasomy 12p.
Pathophysiology
The pathophysiology of this condition is thought to be as
follows:
Loss of lung function can be explained by loss of space
to develop. But other structural deficiencies are added, like
rarification of bronchi and lung vessels (Figure 1). Thick-
ening of alveolar membrane is verified histologically. Hy-
pertrophy of media can be found in periphery of lung
vessels resulting in smaller lumina, whereas these media
muscle fibers are regulating blood flow only in central, big
lung vessels (Figures 2 and 3). These alterations are found
not only in ipsilateral but also in contralateral lung. Thus,
mechanical reasons cannot explain the structural maldevel-
opment of contralateral lung.
All children suffer from persistent pulmonary hypertension,
which plays an important factor for treatment. In respect to
reduced lung function, the effect of additional surfactant treat-
ment is still controversial. It is possible that surfactant becomes
inactivated during mechanical ventilation.1,2
Calculating end diastolic and left ventricular volume
gives information that ejection fraction of the left ventri-
cle is reduced measuring arterial pulmonary flow at the
same time. Therefore, it may be helpful in these cases to
keep the ductus arteriosus open by administration of pros-
taglandins. Another strategy is lowering pulmonary vascu-
246
lar resistance by using vasoactive drugs like nitric oxide,
sildenafile, and milrinone. Because of transposition of liver,
spleen and gut into thorax early in pregnancy, the abdominal
cavity will not will grow as much. Thus, after surgical
closure of the diaphragm, it may be difficult to reduce all
organs inside the abdominal cavity without pressure on the
vena cava and diaphragm, and to affect the mesenteric blood
flow. Early diaphragmatic defect goes along with nonrota-
tion and nonfixation of mesentery. In a nonrotated position
of the gut, the duodenal passage may be blocked due to
kinking. This may not allow early regular feeding regime
after closure of diaphragm.
Similar to late closure of dorsal diaphragmatic area, the
dorsal rim of musculature is very weak and may be hidden.
In larger defects, it may be not be present at all.
Incidence
The incidence is 1:2500-3000 per live births. These figures
may be inaccurate, because of the unknown factor of the
hidden mortality in this malformation. Hidden mortality in this
respect means that only 40 of 100 fetuses with prenatal diag-
nosis will survive. In other words, there must be a high mor-
tality in utero. Even in advanced neonatal intensive care, the
survival rate remained at around 50% for decades. The normal
size of a growing lung is shown in Figure 4 as well as the lung
volume of those patients who did not survive.3 It should be
realized that today the prognosis is more and more related to
associated malformations, which are found in 40%.
Figure 5 Fetal MRI demonstrating normal lung volume and
position.
Seminars in Pediatric Surgery, Vol 17, No 4, November 2008
Figure 4 Fetal lung volume versus gestational age, and dead
patients due to CDH and controls.
Symptoms after birth
The size of the diaphragmatic defect and the timing of
transposition of intraabdominal structures intrathoracically
is an important factor in early presentation with symptoms
after birth. The abdomen seems to be empty and could even
be scaphoid. A newborn in respiratory distress is usually
treated with oxygen and ventilation primarily using a mask.
Typically, this measure leads to deterioration of oxygen-
ation by inflating the stomach and the gut, leading to more
compression of the lung and to increasing mediastinal shift.
Risk of producing pneumothorax is high. In any case of
sudden deterioration, a pneumothorax should be suspected.
Small or slit-like defects of the diaphragm may not be
apparent at birth, as long as there is no herniated gut. After
weeks or months— often during coughing— elevated intra-
abdominal pressure forces the gut through the defect, ac-
companied by pain and problems of passage. In these situ-
ations, mediastinal shifts are hardly ever found. In the x-ray,
a round, air-filled structure possibly with fluid may be seen.
The differential diagnosis from a lung cyst can be difficult.
Contrast studies may help.
Diagnosis
High-quality ultrasound (prenatal magnetic resonance even
more) can demonstrate congenital diaphragmatic hernia
(CDH) early in pregnancy. In ultrasound, identification of
liver or lung tissue may be difficult or unclear, especially in
cases of maternal obesity. It is of prognostic importance to
know whether left liver lobe is lying intrathoracically or not.
In right-sided CDH, the exact position of the gall bladder
may show liver site in ultrasound.
Prenatal lung– head ratio (LHR) as an important factor of
prognosis is widely accepted. Whenever this ratio is less than
Waag et al. Congenital Diaphragmatic Hernia
Figure 6 Fetal MRI in a CDH patient showing the small left
lung and the small gut in the left thorax. (Color version of figure
is available online.)
1.2 at 28 weeks of gestation (WG) in ultrasound, MRI is more
accurate (Figures 5-7). After the 20th WG, it shows accurate
measurements of lung volume, giving time to plan a delivery in
a tertiary referral center with extracorporeal membrane oxy-
genation (ECMO) possibilities, whenever in doubt.
Prenatal prognostic factors
High-risk group
As mentioned before, diagnosis of CDH before 25th WG
means a large defect and a worse prognosis is related to the
Figure 7 Lung volume calculation by planimetry.
247
presence of stomach and left liver above the diaphragm.
Statistically, 93% of the affected newborns are surviving
whenever liver is “down.” If left liver lobe is “up,” the
survival declines to 43%.4
LHR in accordance to age is well defined. At 28th WG,
this LHR 1.2 ⫾ 0.38 means borderline to high-risk group. In
MRI, lung volume of the affected side should not be less
than 10 ml.5 LHR in 34th WG is 1.8-3.0 normally. Patients
who do not need ECMO and have liver down show an LHR
of more than 1.5 (Figure 8).
LHR and lung volume
measurements (34th WG)
According to literature, mortality is believed to be the same
whether CDH is right- or left-sided, but some authors see
more mortality in right-sided defects.6
Measuring fetal hemodynamics and reaction to oxygen sup-
port to the mother by influencing fetal pulmonary lung perfu-
sion may become a prognostic factor. Dextroposition of fetal
heart is found often at the same time as left liver lobe “up.” In
these patients, ductus venosus joins the right atrium atypically
leading to volume load of pulmonary artery and to insufficient
filling of the left heart. Left ventricle stays too small consec-
utively with low-ejection fraction.
So, in respect to prognosis, small left ventricle and rela-
tionship of aorta to pulmonary artery are valuable.
Low birth weight, polyhydramnios, or hydrops are further
discussed as bad factors for prognosis, but the presence of
other associated malformations reduced survival rates.4,6-8
248
Figure 8 Correspondence of “liver up” and “liver down” to the
necessity of ECMO therapy and the need of patch in CDH patients.
“In born” patients suffering from CDH show better out-
come than “out born.” Logistic advantages are obvious.
Transport of the baby in utero is much safer as well as
delivery in a tertiary center providing ECMO. Prenatal
referral is advised in all features showing LHR below 1.4.
LHR below 1.2 makes ECMO necessity very likely, so
does a lung volume below 25 ml in MRI.
Our own results proved that all newborns with lung
volumes of 31.8 ml in average did not need ECMO. High-
risk group depending on ECMO showed LHR of 0.92, ie,
10.38 ml lung volume.
Proper evaluation of prognostic factors will reflect re-
sults (Figure 9). Our survival rate overall is 86.5% of all
newborns admitted; 71% of all patients needing ECMO
survived in our series.
Postnatal diagnosis
There are no breath sounds over the affected thorax, but
bowel sounds may be heard instead. Because of mediastinal
shift, the apex of the heart is displaced. A plain x-ray makes
the diagnosis of CDH as well as indicating whether the
stomach and other organs are lying in the chest. A naso-
gastric tube shows mediastinal shift and defines the po-
sition of stomach. Compression of contralateral lung is
demonstrable.
CDH is associated with cardiac malformations in 30-40%,
mainly arterial and ventricular septum defects and pulmonary
stenosis. Cerebral alterations are mostly found in patients with
chromosomal aberrations, like in syndromes such as Pallister–
Kilian, Beckwith–Wiedemann, and Denys–Drash. Nonrotation
induces duodenal kinks, and relevant clinical signs are realized
in 20-30%. Small hiatus muscles in big defects explain gas-
troesophageal reflux. Additional malformations can be de-
scribed in our series of 177 left-sided defects: lung sequestra-
tion of the 5 patients, accessory spleen in 5, and duplication of
Seminars in Pediatric Surgery, Vol 17, No 4, November 2008
gut in 3 patients were found as well as polycystic kidney and
hydronephrosis.9
Fetal surgery
Harrison and Soper have been the pioneers in developing
techniques for fetal correction and conservation of preg-
nancy despite surgery. Tocolysis at the end was the out-
standing problem.10-12
Experimentally, occlusion of trachea in fetal lambs
blocked outflow of pulmonary secretion, which is pro-
duced inside the lung. This accumulation of secretion
leads to expansion of the lung. But it is mainly not a
structural growth of the number of acini.13,14 Tracheal
occlusion is done between 26th and 28th WG and in-
creases LHR from 0.7 to 1.8. Duration of the plug should
be 6 weeks in the opinion of the Eurofetus group. This
success may be diminished again when occlusion is ter-
minated 2 weeks before delivery.
Plugging trachea by a balloon endoscopically is of
clinical relevance. Deprest is one of the very few who has
experiences in 20 fetuses in positioning the balloon en-
doscopically through mothers abdominal wall and uterus
via fetal mouth into the fetal trachea (fetoscopic endolu-
minal tracheal occlusion, FETO).10,11,15 After sufficient
pulmonary growth, the intratracheal balloon can be un-
blocked via some route (Figure 10). Timing and duration
of this procedure is still under discussion. Taking the
balloon out at the time of delivery is called ex utero
intrapartum (EXIT) procedure. Further experiences have
to be collected before evaluating whether patients with
LHR or lung volumes of high-risk group will profit from
this highly sophisticated treatment.11
Therapy algorithm
For prenatal planning, estimation of lung volume at the 34th
WG is important. Whenever lung volume is more than 20
ml and no factor of bad prognosis like “liver up” is present,
spontaneous delivery should be the mode of treatment.
Figure 9 Significance is shown between lung volumes in MRI
needing ECMO or not and as well borderline volumes for survival.
Waag et al. Congenital Diaphragmatic Hernia 249

Figure 10 Endoscopic pictures of the obstructing balloon inside the fetal trachea. (Color version of figure is available online.)

But if the lung volume is below 20 ml and an additional 6. Surgery for defect closure after stabilization
bad factor is present, follow-up should be given to a tertiary 7. Surgery for defect closure after ECMO
center, expecting cesarean section in the 37th WG followed
In uncomplicated cases, there are no objections against spon-
by advanced treatment, including providing ECMO therapy
taneous delivery. Direct intubation avoids additional filling of
(Figure 11).
gut with air, increasing mediastinal shift and lung compression.
In uncomplicated cases after 38th WG spontaneous de-
Surfactant application seems to have an effect only in preterms.
livery (Figure 12), the following therapy treatment is sug-
Ventilation pressures above 25 cm H2O as well as positive
gested:
endexspiratoric pressure (PEEP) over 3-5 cm H2O induce
1. Primary intubation damage to lung structure. Ventilation frequencies of 80/min
2. Conventional gentle ventilation and inspiration/expiration ratio of 1:2 without auto-PEEP
3. NO inhalation if PaO2 ⬍80 mm Hg postductal should be sufficient. These are the rules of what is called
4. HFOV if PaCO2 ⬎75 mm Hg “gentle ventilation.”1,9,14-16
5. ECMO if PaO2 ⬍50 mm Hg preductal, PaO2 ⬍40 mm Whenever this ventilation is not sufficient, high-fre-
Hg postductal quency ventilation (HFOV) is the next step. Hypercapnia

Figure 11 Algorithm of CDH diagnosis and prenatal planning

in Mannheim. Figure 12 Algorithm of CDH therapy postnatally in Mannheim.
250
can be tolerated up to PaCO2 of 70 mm Hg and respiratory
acidosis until a pH of 7.2.
So far, there is no evidence for any therapy for lung
maturation later than 34th WG.
NO inhalation (10 ppm) is applied whenever PaO2 falls
postductally below 80 mm Hg. But there is no evidence-
based effect for NO in literature.9
Pre- and postductal oxygenation and arterial pressure
have to be monitored continuously by umbilical arterial
catheter placement. However, umbilical veins are not suit-
able, because the tip may easily be positioned in the hepatic
veins or ductus venosus. Drug administration, like catechol-
amine, may produce liver cell necrosis. The most efficient
catecholamine is suprarenin (dose up to maximally 0.2
␮g/kg/min), not increasing resistance of pulmonary vessels.
For continuous analog sedation, fentanyl (2-3 ␮g/kg/h) and
midazolam (0.05 ␮g/kg/h) reduce oxygen consumption. Re-
laxation of musculature (vecuronium 0.05 mg/kg/h) may
influence resistance of pulmonary vessels during days 1-3.
When patients reach PaO2 above 80 mm Hg and reduced
flow on the ductus arteriosus, they can be kept on this
regime until normo-capnia is reached; 24-48 hours later,
surgery can be started.
The honeymoon period is defined as the time after birth
when the newborn seems to be in a good status primarily
before (may be after or around 12 hours) oxygenation and
circulation is breaking down. An explanation for this is very
likely exhausting pumping force of left ventricle,16 which is
too small and not developed sufficiently.
ECMO indication
It is still under debate whether ECMO is the last option in
critical cases. There is no evidence-based study with com-
parable conditions.
Indication for the need of ECMO nowadays after 15
years of experience in our institution and after more than
240 patients on ECMO for CDH is defined as follows:
Figure 13 Principle of ECMO in details. (Color version of figure
is available online.)
Seminars in Pediatric Surgery, Vol 17, No 4, November 2008
Figure 14 ECMO in action on ICU. (Color version of figure is
available online.)
PaO2 postductally ⬍40 mm Hg over 2 hours
PaO2 preductally ⬍50 mm Hg over 2-4 hours
PaO2 postductally ⬍50 mm Hg over 12 hours
PaCO2 ⬎70 mm HG under HFOV
Indication for ECMO in our opinion is not only a question
of survival but also quality of life because of secondary
damage. Preductal PaO2 should always be above 50 mm Hg
and exceed at least 10 mm Hg postductal values. Differ-
ences of pre- and postductal figures allow an estimate of
blood flow through the ductus. Whenever both values of
pre- and postductal PaCO2 are not approaching a minimum
of 5 mm Hg, there is likely a need for ECMO. If PaCO2 is
increasing above 70 mm Hg despite HFOV, there is no other
chance than ECMO. Persistent PaCO2 of more than 100 mg
Hg or PaO2 less than 40 mm Hg does not allow survival
even on ECMO.
Risks under ECMO
Newborns younger than 34th WG or under 1800 g body
weight show significant problems getting cannulas big
enough for sufficient flow. V. cava, V. saphena, V. femo-
ralis, and the carotid artery may rupture and bleed.
Further risks, seen in our institution are: blood clots and
air bubbles induce embolization or infarctions of lung or
brain. Heparin in ECMO machines is still necessary but is
also the reason for cerebral bleed (9%) or bleeding into lung
tissue (14%). Convulsions are registered in 16%. Even after
reconstruction, the carotid artery may thrombose later.
Catheter-related local bleeding 7.5%
Intracerebral bleeding 9%
Intrapulmonary bleeding 14%
Convulsions 16%
Waag et al. Congenital Diaphragmatic Hernia
Figure 15 Plain x-ray of the thorax demonstrating the thick
venous canula in the v. cava superior, massively diverted by the
mediastinal shift.
ECMO technique (Figure 13)
ECMO machine is prefilled with blood and connected to
cannulas implanted into the carotid artery and through in-
ternal jugular and superior vena cava just in front of the right
atrium. Veno-venous technique by double lumen catheter is
abandoned because oxygen supply is not sufficient in CDH
patients and a good functioning right ventricle is necessary.
Heparinization is controlled as well as the number of throm-
bocytes, which should be above 80000-100000 ␮l.
Because of reduced ventilation and activation of cytokines,
lung will “white out” in x-ray. Capillary leak and general
edema are usual, needing additional volume and diuretics.
Normally after 3 days, recreation starts (Figure 14). As soon as
blood flow on ECMO can be reduced to 20% of patients
cardiac output and oxygen saturation on venous side (SvO2) is
falling below 70% for 24 hours (called “idling phase”), ECMO
can be terminated.
ECMO results came up to survival rates of around 70%
in recent years.9,17
Technique of ECMO cannulation
The right carotid artery is dissected as well as internal jugular
vein, taking care of vagal nerve and ansa cervicalis. Skin
incision is vertical along anterior sternocleido muscle, to have
better access for reconstruction of vessels after ECMO.
Dilation of vessel is precarious but necessary from time
to time to implant cannulas as big as possible (10-12 Ch for
the vein and 8-10 for the carotid artery). The tip of the
venous cannula has to be placed at right atrium and arterial
cannula should join aorta controlled by x-ray to provide
sufficient flow. It has to be taken into account that turning
the head back postoperatively, tips will be pushed forward
additionally (Figure 15).
251
Operative steps of closure
A longitudinal midline laparotomy is the best approach for
sufficient exposure of the diaphragmatic defect. A dorsal
muscular rim may be hidden primarily, but often can be
found except in large defects (Figure 16). Primary closure
needs incising rims for better healing. Nonabsorbable su-
tures 3 to 0 or 4 to 0 are used.
Whenever primary closure of the diaphragm is per-
formed under tension, the intrathoracic space is enlarged
and less volume exists intraabdominally as seen often on
postop radiograph (Figure 17, left in comparison to right
side). This is the opposite of what is needed; ie, tension on
the diaphragm increases as a result of raised intraabdominal
pressure. This is caused by a small peritoneal cavity trying
to accommodate all the abdominal viscera, which was oc-
cupying space in the chest.
Early patching is better than primary closure under ten-
sion, and cone-forming gives more intraabdominal space
than a flat patch (Figure 18). Fixating patch near aorta and
hiatus is fragile in big defects prone to recurrent hernia.
Therefore, the patch should be bigger than the defect and is
fixed with mattress sutures (Figure 19).
Our learning curve showed that problems of duodenal
kinking are not rare (16%). Whenever this is realized, du-
odenum is dissected free and may be splinted by a gastric
tube fed through. Analog in respect to likelihood of gastro-
esophageal reflux (20%), it is discussed to perform in big
defects a fundo-esophagopexy as a Thal procedure, for
example, prophylactically.9,17-20 Most experts of CDH are
no longer using drainage of pleura.
In a big defect with a big volume to be brought to the
abdomen, it is necessary to stretch the abdominal wall or
even to patch abdominal incision for adequate closure.
Too much intraabdominal pressure extends the dia-
phragm and compresses venous backflow through vena cava
to the heart.
Figure 16 This diaphragmatic defect shows a good anterior
diaphragmatic muscle and hiatus muscle and a small dorsal dia-
phragmatic rim; the lung is fully ventilated but hypoplastic. (Color
version of figure is available online.)
252 Seminars in Pediatric Surgery, Vol 17, No 4, November 2008

Figure 17 In x-ray, a tight direct closure or a plane patch as in the left picture demonstrates the difference to a cone-formed patch (right
picture) giving more abdominal space.

Access via thoracotomy sive access (Hirschl, personal communication). In literature,

minimal invasive surgery leads to higher rate of recur-
After open thoracotomy successful reduction of gut through rence.21
the diaphragmatic defect is possible as well, even when
patching is necessary. But gut cannot be deflated and con- (Mannheim series)
trolled intraabdominally. In case of high intraabdominal Postoperative complications n ⴝ 244 patients
pressure, reduction can be difficult or impossible because
Pneumothorax 30%
stretching and abdominal patching cannot be performed via Chylothorax 28%
thoracotomy. Gastro-oesophageal reflux (GER) 20%
Minimal invasive thoracoscopy closure is published in Intestinal problems 16%
case histories, but there exist the same restrictions as in open Recurrence 14%
approach. Additionally, it may be problematic to find the In any situation with sudden deterioration of the child on
defect to push gut intraabdominally under vision. Intratho- ventilation suffering from CDH, a pneumothorax has to be
racic positive pressure helps during reduction. considered. A chest radiograph confirms the diagnosis or, in
Defects with sternal (Larrey) clefts or Morgagni hernias an urgent situation, a needle puncture gives the hint for
are more suitable than Bochdalek hernia for minimal inva- intrapleural free air. Big diaphragmatic defect with suturing
medially near hiatus and esophagus may open up lymph
ducts and may result in chylothorax (28%) as in heart

Figure 18 A cone-formed patch reduces the abundant thoracical Figure 19 To prevent recurrency of a diaphragmatic hernia the
space and increases abdominal cavity to take all formerly intratho- patch is used bigger than the defect. (Color version of figure is
racical gut and organs. (Color version of figure is available online.) available online.)
Waag et al. Congenital Diaphragmatic Hernia
surgery.9,17 Often the leak closes with time, but in some
cases, capillary leak syndrome develops with a high mor-
tality. GER and intestinal problems of passage have been
discussed before. Postoperative adhesions leading to ileus
are rare.
Likelihood of developing recurrence depends mainly on
the size of the original defect. Suturing under tension or
tearing patch out of surrounding tissue produces a recurrent
hernia in about 14%. Infants grow mainly during their first
12-24 months. So in this period, secondary hernia may
happen especially in such cases when a small muscular rim
did not have any chance to grow. Location of recurrent
hernia is paraesophageal or posterior. For closure, an addi-
tional patch will cover the new defect leaving the first patch
in place. In general, increasing survival will increase mor-
bidity, needing outpatient control every 3 months during
their first years.
Liquid ventilation with perfluorocarbon
Liquid ventilation is by definition an assisted ventilation
with tracheal installation of perfluorocarbon (PFC) to
replace nitrogen as a carriage for oxygen and carbon
dioxide.
Biologically, PFC is chemically stable and nontoxic.
The aim of using PFC for liquid ventilation is to improve
gas exchange.
The effect of PFC is to influence lung compliance and
ventilation and gas transfer, allowing lower ventilatory
pressure avoiding barotrauma and high oxygen concentra-
tion. Liquid ventilation opens alveoli physically and reduces
microatelectasis as is seen with ECMO.
There are two forms of liquid ventilation. In total liquid
ventilation, the lung is filled completely with PFC, so there
is no free gas in the lung. In partial liquid ventilation, PFC
is combined with high frequency oxygen ventilation or
nitric oxide.
The first clinical trials using PFC in children occurred in
1984. Neonates with CDH and children suffering from
pertussis or Ebstein’s anomaly tolerated this procedure and
showed improvement of lung compliance and arterial oxy-
genation in these cases of severe respiratory distress syn-
drome.
Recently, there have been no further reports of PFC in
children, so that this procedure seems to be of no further
interest, at least at the moment.
Results
For decades, the survival rate of CDH patients rested at
about 50%. With more expertise in prenatal diagnosis and
postnatal intensive care, survival rates have increased re-
cently, inclusive using ECMO. In 2005, multicenter studies
of 50 centers in the US published survival of 83% in 2004
253
patients.13,22 The German Center for CDH showed, for 244
consecutive patients, a rate of survival of 86.5% of liveborn
cases.9,17
Most newborns that died did not reach surgical interven-
tion or died during or immediately after ECMO. This se-
lected group of cases with bad chances needing ECMO
survived in 71% in the German Center, a figure which
underlines ECMO as treatment of last choice.
Long term results
Overall survival 86.5%
Survival ECMO patients 71%
Neurological deficit 45%
Low body weight ⬍ p5 39%
Pectus excavatum 33%
Maldescended testes 26%
Weak abdominal wall 16%
Scoliosis 19%
Chronic lung disease: no ECMO 16%, after ECMO 54%
Late results are heavily influenced by accompanying con-
genital morbidity.5,17-19,23,24 Spirometry measurements
show reduced lung capacity ipsilaterally and obstruction as
well as restricted lung functions. Patients are prone to sco-
liosis and funnel chest. Neurological development was re-
duced in 45%. Children grow less and stay behind in body
weight (⬍ p5 39%) in the first 2 years.9 Multiple factors are
responsible for ongoing pulmonary hypertension: small
cross-section of pulmonary arteries, hyperplasia of media in
distal pulmonary vessels, missing vasodilatation to oxygen,
and an increased expression of endothelin A receptors.
New studies showed that sildenafil is useful as therapy in
acute as well as in chronic pulmonary hypertension. This
drug blocks phosphodiesterase type 5 (PDE5). PDE5 is a
key enzyme for NO, leading to pulmonary vasodilation.
Summary
New ways of ventilation have improved survival rates from
50% to 80-85%. Using gentle ventilation avoids destruction of
lung tissue and diminishes situations of long-term ventilated
patients. Good quality of life has to be the goal in future and
not only survival rate with the risk of a high morbidity. Success
in difficult cases is optimized by referral to tertiary centers. It
is the responsibility of gynecologists, neonatologists, and pe-
diatric surgeons to communicate early and decide who of the
neonates suffering from severe forms of CDH needs to be
referred to a specialized tertiary center.
References
1. Greenspan JS, Shaffer TH. Ventilator-induced airway injury: a critical
consideration during mechanical ventilation of the infant. Neonatal
Netw 2006;25:159-66.
254
2. Migliazza L, Bellan C, Alberti D, et al. Retrosective study of 111 cases
of congenital diaphragmatic hernia treated with early high-frequency
oscillatory ventilation and presurgical stabilization. J Pediatr Surg
2007;42:1526-32.
3. Skari H, Bjornland K, Haugen G, et al. Congenital diaphragmatic hernia:
a meta-analysis of mortality factors. J Pediatr Surg 2000;35:1187-97.
4. Albanese CT, Lopoo J, Goldstein RB, et al. Fetal liver position and
perinatal outcome for congenital diaphragmatic hernia. Prenat Diagn
1998;18:1138-42.
5. Cortes RA, Keller RL, Townsend T, et al. Survival of severe congen-
ital diaphragmatic hernia has morbid consequences. J Pediatr Surg
2004;40:36-46.
6. Graham G, Devine P. Antenatal diagnosis of congenital diaphragmatic
hernia. Semin Perinatol 2005;29:69-76.
7. Fumino S, Shimotake T, Kume Y, et al. A clinical analysis of prog-
nostic parameters of survival in children with congenital diaphrag-
matic hernia. Eur J Pediatr Surg 2005;15:399-403.
8. Hedrick HL, Crombleholme TM, Flake AW, et al. Right congenital
diaphragmatic hernia: Prenatal assessment and outcome. J Pediatr Surg
2004;39:319-23.
9. Dahlheim M, Witsch M, Demirakca S, et al. Angeborene Zwerchfell-
hernie: Ergebnisse eines ECMO-Zentrums. Klin Pädiatr 2003;215:213-
22.
10. Harrison MR, Keller MD, Hawgood SB, et al. A randomized trial of
fetal endoscopic tracheal occlusion for severe fetal congenital dia-
phragmatic hernia. N Engl J Med 2003;349:20.
11. Kohl T, Gembruch U, Filsinger B, et al. Encouraging early clinical
experience with deliberately delayed temporary fetoscopic tracheal
occlusion for the prenatal treatment of life-threatening right and left
congenital diaphragmatic hernias. Fetal Diagn Ther 2006;21(3):
314-8.
12. Kunisaki SM, Chang RW, Androli S, et al. Hyperoncotic enhancement of
fetal pulmonary growth after tracheal occlusion: an alveolar capillary
morphometric analysis. J Pediatr Surg 2006;41:1214-8.
View publication stats
Seminars in Pediatric Surgery, Vol 17, No 4, November 2008
13. Cass DL. Fetal surgery for congenital diaphragmatic hernia: the North
American experience. Semin Perinatol 2005;29(2):104-11.
14. Baquero H, Soliz A, Neira F, et al. Oral sildenafil in infants with
persistent pulmonary hypertension of the newborn: a pilot randomized
blinded study. Pediatrics 2006;117:1077-83.
15. Boloker J, Bateman DA, Wung JT, et al. Congenital diaphragmatic hernia
in 120 infants treated consecutively with permissive hypercapnea/sponta-
neous respiration/elective repair. J Pediatr Surg 2002;37:357-66.
16. Inamura N, Kubota A, Nakajima T, et al. A proposal of new thera-
peutic strategy for antenatally diagnosed congenital diaphragmatic
hernia. J Pediatr Surg 2005;40:1315-9.
17. Loff S, Schaible T, Nützenadel W, et al. Symposium interdisciplinary
update on congenital diaphragmatic hernia. Monatsschrift Kinderhei-
lkd 2006;154:713-28.
18. Chen C, Jeruss S, Chapman J, et al. Long-term functional impact of
congenital diaphragmatic hernia repair on children. J Pediatr Surg
2007;42:657-65.
19. Moss RL, Chen CM, Harrison MR. Prosthetic patch durability in
congenital diaphragmatic hernia: a long-term follow-up study. J Pedi-
atr Surg 2001;40:1701-5.
20. Su W, Berry M, Puligandla PS, et al. Predictors of gastroesophageal
reflux in neonates with congenital diaphragmatic hernia. J Pediatr Surg
2007;42:1639-43.
21. Arca MJ, Barnhart DC, Lelli JL, et al. Early experience with minimal
invasive repair of congenital diaphragmatic hernia: results and lessons
learned. J Pediatr Surg 2003;38:1563-8.
22. Javid PJ, Jaksic T, Skarsgard ED, et al. Survival rate in congenital
diaphragmatic hernia: the experience of The Canadian Neonatal Net-
work. J Pediatr Surg 2004;39:657-60.
23. Heling KS, Bollmann R, Wauer RR, et al. Der Stellenwert der Lung-
to- head-Ratio in der Prognoseeinschätzung bei Feten mit isolierter
Zwerchfellhernie. Z Geburtsh Neonatal 2004;207:89-166 (suppl 2).
24. Stolar CJH, Levy JP, Dillon PW, et al. Anatomic and functional
abnormalities of the esophagus in infants surviving congenital dia-
phragmatic hernia. Am J Surg 1990;159:204-7.