Vous êtes sur la page 1sur 9

Lung Volume, Pulmonary Vasculature, and Factors Affecting Survival in

Congenital Diaphragmatic Hernia


Donald W. Thibeault and Barbara Haney
Pediatrics 1998;101;289-295
DOI: 10.1542/peds.101.2.289

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://www.pediatrics.org/cgi/content/full/101/2/289

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly


publication, it has been published continuously since 1948. PEDIATRICS is owned, published, and
trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove
Village, Illinois, 60007. Copyright 1998 by the American Academy of Pediatrics. All rights
reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from www.pediatrics.org at Indonesia:AAP Sponsored on August 13, 2008

Lung Volume, Pulmonary Vasculature, and Factors Affecting Survival in


Congenital Diaphragmatic Hernia
Donald W. Thibeault, MD, and Barbara Haney, RNC, MSN
ABSTRACT. Objectives. There is a wide variation in
published mortality from congenital diaphragmatic hernia (CDH). The prevailing opinion is that this variation is
related directly to the degree of pulmonary hypoplasia.
Our aim was to test the hypothesis that other factors are
important for outcome.
The specific objectives of this study were: 1) to quantitate the degree of lung hypoplasia and pulmonary arterial wall thickness in infants eligible for, and treated
with, extracorporeal membrane oxygenation (ECMO), using postmortem analysis of lung DNA, wet lung weight,
lung volume, and vessel morphometrics; 2) to correlate
the degree of lung hypoplasia and vascular changes with
functional tests of oxygenation and estimated right ventricular systolic pressures (RVSP); 3) to determine the
minimum lung volume necessary for survival; and 4) to
determine contributory clinical factors as potential
causes of death in ECMO-treated infants with CDH.
Methodology. We retrospectively analyzed all 90 infants with CDH admitted consecutively over a 9-year
period to a childrens hospital with an ECMO program.
Infants were categorized as lived or died, with or without
ECMO. Indication for ECMO was an evolving process;
however, in general, it was the therapy of last resort for
pulmonary insufficiency. Clinically, the single best oxygenation index before ECMO or CDH repair while on
conventional ventilation, and serial echocardiograms before, during, and after ECMO, were obtained. Twelve of
14 infants dying with ECMO and 6 of 12 without ECMO
had postmortem examinations. Lung volume, DNA content, wet weights, and arterial wall thickness at the level
of alveolar ducts were measured in both lungs. Postmortem morphometric findings were correlated with in vivo
tests of cardiopulmonary function and contributory clinical factors in mortality.
Results. Sixty-three percent of all infants with CDH
and 61% of ECMO-treated infants lived. All infants with
CDH requiring ECMO had elevated RVSP/systolic systemic blood pressure ratios before ECMO (0.98 6 0.24).
Eighty-eight percent of ECMO-treated infants with CDH
decreased this ratio to <0.5 within 14 days, regardless of
lung size. However, infants dying with normal ratios still
had increased arterial wall thickness and muscle in both
lungs. In infants whose lung volume, DNA, and weight
were >45% of values predicted for age-matched controls,
the oxygenation index ranged from 4 to 29, significantly
less than that in infants with values <45% of predicted
values (range, 25 to 133). We speculate that eight infants
with lung volumes >45% of that for controls died from
From the Department of Pediatrics, Childrens Mercy Hospital, University
of MissouriKansas City School of Medicine, Kansas City, Missouri.
Received for publication Dec 10, 1996; accepted Aug 1, 1997.
Reprint requests to (D.W.T.) Childrens Mercy Hospital, 2401 Gillham Rd,
Kansas City, MO 64108.
PEDIATRICS (ISSN 0031 4005). Copyright 1998 by the American Academy of Pediatrics.

potentially preventable surgical and medical complications.


Conclusion. A minimum lung volume of 45% of the
value predicted from age-matched controls is required
for survival in ECMO-treated infants. The RVSP/systolic
systemic blood pressure ratio can be reduced with ECMO
to <0.5 in the majority of infants, even with lung volumes inadequate for survival. We speculate that 9% of
infants with adequate lung volume were potentially survivable, but died of medical and surgical complications.
Pediatrics 1998;101:289 295; lung volume, vessels, diaphragmatic hernia.
ABBREVIATIONS. CDH, congenital diaphragmatic hernia;
ECMO, extracorporeal membrane oxygenation; RVSP, right ventricular systolic pressure; VV, venoveno; OI, oxygenation index;
SBP, systolic systemic blood pressure; FRC, functional residual
capacity.

he high mortality rate in infants with congenital diaphragmatic hernias (CDH) is ascribed to
lung hypoplasia and pulmonary hypertension,1 4 but the remarkably wide variation in published mortality rates suggest that other clinical variables may affect outcome.1,5 When conventional
therapies fail, extracorporeal membrane oxygenation
(ECMO) is the prevailing therapy.6 However, the
inability to quantitate clinically the degree of lung
hypoplasia and the lack of standardized therapies for
pulmonary hypertension have made ECMO treatment protocol comparisons difficult to interpret. If
the cause of death in ECMO-treated infants is predominately hypoplasia and/or pulmonary hypertension, it should be readily apparent from postmortem
lung quantitative analysis.
The objectives of this study were:
1. to quantitate the degree of lung hypoplasia and
pulmonary arterial wall thickness in infants eligible for, and treated with, ECMO, using postmortem analysis of lung DNA, wet lung weight, lung
volume, and vessel morphometrics;
2. to correlate the degree of lung hypoplasia and
vascular changes with functional tests of oxygenation and estimated right ventricular systolic pressures (RVSP);
3. to determine the minimum lung volume necessary for survival; and
4. to determine contributory factors as potential
causes of death in ECMO-treated infants with
CDH.

PEDIATRICS Vol. 101 No. 2 February 1998


Downloaded from www.pediatrics.org at Indonesia:AAP Sponsored on August 13, 2008

289

METHODS
A retrospective analysis of all CDH-diagnosed infants admitted
to the intensive care nursery since beginning an ECMO program
(1987 to 1996) was performed. The 90 infants were outborn from a
four-state midwestern region. They were categorized as lived with
no ECMO, lived with ECMO, died with no ECMO, died with
ECMO, and those with congenital anomalies in addition to CDH.
During this 9-year period, the indication for ECMO treatment was
CDH with inability to survive on conventional ventilation, including high-frequency oscillatory ventilation. In the early years of the
ECMO program, CDH repair occurred before, during, or after
ECMO, with no standardization. Infants who died without ECMO
were those that died before ECMO could be instituted, although
some had CDH repair at other institutions. In recent years, patients with severe CDH were managed initially with ECMO or
with CDH repair after decannulation, and in two, a second trial of
venoveno (VV) ECMO was instituted after hernia repair.7
The most recent 15 infants with CDH were treated with inhaled
nitric oxide; 11 preoperatively, before ECMO; 4 after ECMO and
CDH repair. Of 8 infants with multiple congenital anomalies,
only 1 was treated with ECMO. Two had trisomies, 3 had multiple anomalies with normal chromosomes, and 3 had aortic
coarctation.
Postmortem examinations were obtained in 12 of 14 infants
dying with ECMO therapy and 6 of 12 infants dying without
ECMO. Infants dying with anomalies were excluded from postmortem analysis. Control (n 5 13) postmortem lung analysis was
performed on infants dying with birth asphyxia or other neurologic problems. Their ages at death were 5.4 6 5 days, with a range
of 0.3 to 28 days. Their lungs were relatively normal except for
edema, congestion, and mild ventilator changes.
Clinically, the single best oxygenation index (OI), OI 5 Fio2 3
MAP 3 100/PaO2, was obtained before ECMO or CDH repair. The
PaO2 was always assessed while on conventional ventilation, via
an umbilical arterial catheter placed below the ductus arteriosus
or from the left radial artery. Echocardiograms were obtained at
admission on all infants and repeated serially while on ECMO.
Echocardiograms were performed sporadically after ECMO and
only in infants whose pulmonary status deteriorated. During
echocardiographic examination, while on ECMO, pump flows
were reduced to 0 to 30 mL/kg per minute. RVSP was estimated
from the maximum velocity of the tricuspid regurgitant jet by
echocardiographic Doppler sampling.8 The pressure was expressed as the ratio of RVSP to systolic systemic blood pressure
(SBP). If there was no regurgitant jet, the echocardiographic findings were not used. Serial echocardiograms were obtained using
the following protocol. If the contralateral lung was not well
expanded on chest radiograph or if the pump flows could not be
reduced to at least 30 mL/kg per minute, the echocardiogram was
not attempted, because there was no hope of successful removal
from ECMO. If the contralateral lung was well expanded, the Fio2
of the ventilator was increased from room air to 0.4, the ventilatory rate was increased from 20 to 35 breaths per minute, and the
ventilator pressure was increased to a maximum of 25 cm H2O
peak inspiratory pressure and to 5 cm H2O positive end expiratory
pressure. These are the approximate maximal settings we desired
after decannulation. The ECMO pump flows were then decreased
over a 1-hour period to the lowest flow (0 to 30 mL/kg), which
maintained a normal pump venous oxygen tension (PvO2) and
carbon dioxide tension (PvCO2). The venous saturation was monitored continuously. With VV ECMO, the cephalic PvO2 and
PvCO2 were monitored. At zero pump flow, the arterial pO2 and
pCO2 were used to assess lung function. This protocol produced a
reproducible trend in the estimation of the RVSP.
At postmortem examination, the lungs were separated and
weighed. Lung volumes were determined by inflating the collapsed lungs with air at room temperature during a 30-minute
period to a maximum pressure of 35 cm H2O.9 The volume was
read directly from a calibrated syringe, and pressure was measured with a water manometer. Judged by visual examination, the
contralateral lung generally expanded well with slow, 35 cm H2O
air inflation. The ipsilateral lung expanded poorly in at least 25%
of lungs, with hemorrhage often seen on the surface. These general
observations were confirmed on histologic examination (after formalin expansion). Because of the potential for underestimation of
lung volume and overestimation of lung wet weights,10 the lung
DNA, which is not influenced by expansion or edema, also has

290

been reported. A portion of the left and right upper lobes was
analyzed for DNA content, and the total DNA of each lung was
calculated.11 The pulmonary arteries were infused with a barium
sulfate gelatin mixture.9,12 The main bronchi were perfused with
10% formalin at 24 cm H2O for 5 days. A radiograph was then
obtained to demonstrate the vascular pattern. Sections were taken
of the left and right upper lobes, the lingula, the left and right
lower lobes, and the right middle lobe, and from a longitudinal
hilar cut, which included the central hilar structures, surrounded
by parenchyma. Sections were immersed in 10% buffered formalin, and the area of the cut surface was measured with computerized image analysis (Optimas, Edmonds, WA). Sections then were
processed, embedded in paraffin, cut 5-mm thick, and stained with
hematoxylin and eosin and Millervan Gieson stains. The area of
the hematoxylin and eosin-stained sections was compared with
the area of the fixed tissue sections before processing to determine
the shrinkage correction factor. Arterial wall thickness and percentage of vessels with muscle were measured in barium-filled
arteries at the alveolar duct level. Twenty vessels were assessed in
both the left and the right lungs. The percent wall thickness was
calculated as (2 3 wall thickness 3 100)/external diameter.12 The
left and right lungs were assessed histologically for hemorrhage
using a scale of 0 to 5, where 0 5 no hemorrhage; 1 5 minimal
scattered interstitial hemorrhage; 2 5 mild diffuse interstitial hemorrhage; 3 5 severe diffuse interstitial hemorrhage; 4 5 severe
interstitial plus airway hemorrhage; and 5 5 severe hemorrhage to
the point of obliteration of the lung anatomy. A potentially survivable infant was defined as one with lung volume,13 lung
weight,10 and lung DNA $45% of predicted values; RVSP/SBP ,
.5; and OI , 30.3 We reviewed the history for possible contributory
factors for mortality in potentially survivable infants who died.

Statistical Analysis
Values in Table 1 are given as mean 6 SD. Statistical analyses
in Tables 1 and 2 and Fig 3 and text were performed with t tests
or the MannWhitney rank sum test, with P , .05 considered
significant. All statistics and graphs were generated with SigmaStat and SigmaPlot software (Jandel Scientific, San Rafael, CA).

RESULTS

The gestational age of the infants was 38 6 2 weeks


(range, 32 to 42 weeks). The overall CDH survival
was 63.3% (57/90). Of ECMO-treated infants, 61%
(22/36) survived. Seven of the infants with anomalies died. The one survivor was treated with ECMO.
The survival rate in infants without anomalies was
67%. The 5-minute Apgar scores of those infants not
given ECMO who died versus those who received
ECMO were 3.6 6 1.6 versus 6.4 6 2.1 (P , .001), and
mean airway pressure was 17.7 6 2.9 versus 14.2 6
2.9 (P , .005). Infants who died without ECMO as a
group appeared to be sicker at birth and needed
more support at the time of arrival at the ECMO
center. Table 1 relates the OI to outcome in the various treatment groups. Infants who lived and did not
require ECMO had a significantly lower mean OI
(4 6 3) than all other groups. Infants treated with
ECMO who survived had a relatively low mean OI
of 17 6 5, with a high of 35. However, their mean OI
was not significantly different from ECMO-treated
infants who died (21 6 13; range, 4 to 38). Infants
dying without ECMO had mean OI values of
53 6 29.
The lung weights, DNA, and lung volumes of the
right, left, or both lungs were standardized to body
weight as shown in Table 2. These parameters were
expressed additionally as a percentage of values predicted from controls. The variation of these parameters is considerable, but a pattern is discernible. In
infants who died before ECMO could be instituted

CONGENITAL DIAPHRAGMATIC HERNIA LUNG VOLUME


Downloaded from www.pediatrics.org at Indonesia:AAP Sponsored on August 13, 2008

TABLE 1.

Treatment Groups and OI Index in 90 Consecutive Infants With CDH

Lived, no ECMO
Lived with ECMO
Died with ECMO
Died, no ECMO
Congenital anomalies (one lived with ECMO)
Total

% of Total

OI

OI Range

P Value*

35
21
14
12
8
90

38.9
23.3
15.6
13.3
8.9
100.0

463
17 6 5
21 6 13
53 6 29
29 6 16

0.824
3.035
4.038
25.0133
4.050

,.0001
,.0001
,.0001
,.0001

* In comparison with the group that lived, with no ECMO.


The OI was not significantly different in infants who lived or died with ECMO therapy.

TABLE 2.

Lung Weights, DNA, and Lung Volumes in Newborn Infants With or Without CDH
Lung DNA/BW (mg/kg)

Controls without CDH


(n 5 12)
CDH and ECMO
1
2
3
4
5
6
7
8
9
10
11
12
CDH, no ECMO
13
14
15
16
17
18

Lung Wet Weight/BW (g/g)

Lung Volume/BW (mL/kg)

Right

Left

Both

Right

Left

Both

Right

Left

Both

55 6 14

48 6 13

103 6 14

.013 6 .003

.011 6 .003

.024 6 .003

19.8 6 2.8

18 6 2.0

37.9 6 3.1

85 (154)*
65 (118)
59 (107)
45 (82)
45 (82)
44 (80)
43 (78)
40 (73)
34 (62)
25 (45)
32 (58)
16 (29)

51 (106)
7 (14)
16 (33)
12 (25)
12 (25)
9 (19)
18 (40)
13 (27)
8 (17)
11 (23)
8 (17)
10 (21)

136 (132)
72 (70)
75 (73)
57 (55)
57 (55)
54 (52)
61 (59)
53 (51)
42 (41)
36 (35)
40 (39)
26 (25)

.024 (185)
.014 (108)
.015 (115)
.021 (166)
.011 (85)
.011 (85)
.011 (85)
.010 (77)
.008 (61)
.004 (31)
.007 (54)
.006 (46)

.010 (91)
.002 (18)
.005 (45)
.004 (36)
.004 (36)
.003 (27)
.006 (54)
.004 (36)
.002 (18)
.002 (18)
.001 (9)
.004 (36)

.034 (141)
.016 (67)
.020 (83)
.025 (1.04)
.015 (62)
.014 (58)
.017 (71)
.014 (58)
.010 (42)
.006 (25)
.008 (33)
.010 (42)

16.0 (81)
16.5 (83)
16.5 (83)
16.1 (81)
15.7 (79)
15.5 (78)
15.0 (76)
15.0 (76)
12.8 (65)
9.2 (46)
11.3 (57)
8.0 (40)

12.0 (67)
2.0 (11)
3.5 (19)
2.3 (13)
2.8 (15)
3.0 (17)
5.9 (33)
2.0 (11)
1.5 (8)
2.0 (11)
1.0 (5)
1.5 (8)

28.0 (74)
18.5 (49)
20.0 (53)
18.4 (49)
18.5 (49)
18.5 (49)
20.9 (55)
17.0 (45)
14.3 (38)
11.2 (29)
12.3 (32)
9.5 (25)

25 (45)
19 (34)
14 (25)
11 (20)
7 (13)
6 (11)

7 (15)
21 (44)
5 (10)
3 (6)
14 (29)
2 (4)

32 (31)
40 (39)
19 (18)
14 (14)
21 (20)
8 (8)

.005 (38)
.004 (31)
.006 (46)
.003 (23)
.002 (15)
.002 (15)

.002 (18)
.004 (36)
.002 (18)
.001 (9)
.003 (27)
.001 (9)

.007 (29)
.008 (33)
.008 (33)
.004 (17)
.005 (21)
.003 (12)

7.0 (35)
5.8 (29)
9.0 (45)
4.0 (20)
3.7 (19)
3.0 (15)

2.0 (11)
4.8 (27)
3.1 (17)
0.5 (3)
3.4 (19)
0.5 (3)

9.0 (24)
10.6 (28)
12.1 (32)
4.5 (12)
7.1 (19)
3.5 (9)

* Percentage of value predicted from controls shown in parentheses.

(cases 13 through 18), the wet lung weight, DNA,


and lung volume were ,45% of values predicted
from age-matched controls. Infants 9 through 12 resembled those infants; in all, the three parameters
were ,45% of predicted values, whereas in infants 1
through 8, the three parameters were $45% of predicted values.
Table 3 summarizes the three lung size parameters, OI, possible contributory factors for mortality,
and an estimation of the potential of the adequacy of
lung tissue for survival. When lung volume, DNA,
and weight were $45% of predicted, the OI was
13.6 6 8.9 and ranged from 4 to 29. When the same
parameters were ,45% of predicted, the OI was
45.1 6 31.6 and ranged from 25 to 133 (P ,.01).
Possible contributory factors for mortality in patients
1 through 8 were uncontrollable hemorrhage (three
infants), bilateral bronchopneumonia (one infant),
pulmonary hypertension (three infants, two with supersystemic pressure), and respiratory insufficiency
with anasarca and renal failure after CDH repair
(one infant). These findings of greater lung volume,
DNA, weight, and relatively low OI values, in addition to potentially preventable surgical and medical
complication, suggest a cause of death other than
pulmonary hypoplasia.

The relationship of CDH and pulmonary hypertension in infants treated with ECMO was addressed
by plotting the changes in the ratio RVSP/SBP as a
function of days on ECMO. The ratio decreased in all
infants. In infants who survived ECMO (Fig 1), the
ratio before ECMO was 0.97 6 0.21. In 75% of those
infants (15/20), the ratio fell to #0.5 with 5 to 8 days
of ECMO. One infant required 11 days. The four
other infants were removed from ECMO when the
ratio was between 0.50 and 0.69, and these four infants had proven increases in the ratio after CDH
repair. One infant required a second course (8 days)
of VV ECMO to return the ratio to ,0.50. We were
not able to determine the RVSP/SBP in infants who
did well after CDH repair because of the limited
number of echocardiograms performed.
In infants who died after ECMO, the ratio of
RVSP/SBP immediately before ECMO was 1.08 6
0.16, not different from infants who survived with
ECMO (Fig 2). In 57% (8/14), the ratio fell below 0.50
after 4 to 14 days of ECMO. In 14% (2/14), after 12
days of ECMO, the ratio was still .0.50. These two
infants had lung volumes 39% and .45% of values
predicted from controls. Four infants were removed
from ECMO in ,4 days; all had ratios .0.50. The
ratio showed some decline while on ECMO in all

Downloaded from www.pediatrics.org at Indonesia:AAP Sponsored on August 13, 2008

ARTICLES

291

TABLE 3.

Relationship of Lung Size, Cardiopulmonary Function, and Cause of Death in Infants Dying With CDH
DNA/BW Lung Wet
Lung
OI Lived Days
(mg/kg) % Weight/BW
Volume/
(days)
on
Predicted
(g/g) % BW (mL/kg), %
ECMO
Predicted
Predicted

CDH and ECMO


1

132

141

74

6 49

70

67

49

29 15

15

73

83

53

6 14

55

104

49

13 32

18

55

62

49

13

52

58

49

4 11

59

71

55

24 13

51

58

45

14 19

14

41

46

38

33 12

11

10

35

25

29

37 11

10.1

11

39

33

32

36 10

10

12

25

42

25

38

5.0

CDH, no ECMO
13

31

29

24

31

1.2

14

39

33

28

34

1.0

15
16
17

18
14
20

33
17
21

32
12
19

25
133
32

0.8
0.17
2.0

18

12

52

0.29

infants. After ECMO, in 43% (6/14) of infants, the


ratio increased; however, not all infants underwent
echocardiographic retesting, thus, the percentage
may have been higher.
The newborn pulmonary intraacinar arteries normally have increased wall thickness in the first days
of life, in part attributable to thickened muscle.14
Figure 3 indicates this pattern in controls, in whom
the increased wall thickness fell to a constant level of
5.4 6 0.3% early in life (from 6 to 36 days of life).
After 6 days of age, contralateral lung alveolar ductassociated arteries in ECMO-treated infants with
CDH had a mean wall thickness of 12.1 6 2.7% (Fig
3) and in the ipsilateral lung, 22.8 6 5.0. This difference may be related to the profound hypoplasia of
the ipsilateral lungs. The arteriogram showed much
poorer filling of the ipsilateral lung (data not shown).
In all patients with CDH, after 6 days of age arterial
wall thickness in both lungs was significantly greater
than that in controls. After 6 days of age, 15.2 6 5.0%
of all control vessels had some muscle at the alveolar
ductal level in contrast to 45.6 6 19.5% of ECMOtreated infants with CDH. Five of 11 infants who
lived beyond 6 days had normal RVSP/SBP ratios
292

3.3

8.5

2.3

Possible Contributory
Clinical Factors for
Mortality

Adequate ECMO Before


Lung
CDH Repair
Tissue for
Survival

Pulmonary hypertension,
short ECMO run
Bleeding associated with
chest tube while on
ECMO
Pulmonary hypertension,
short ECMO run
Pulmonary failure and
pulmonary
hypertension after
abdominal surgery after
ECMO
Abdominal bleeding
while on ECMO
Bleeding associated with
chest tube while on
ECMO
Bronchopneumonia after
ECMO
Respiratory failure after
CDH repair
Respiratory failure after
CDH repair
Respiratory failure after
CDH repair
Cardiac tamponade after
CDH repair
Abdominal bleeding
while on ECMO
Pulmonary failure after
CDH repair
Pulmonary failure after
CDH repair
Pulmonary failure
Pulmonary failure
Pulmonary failure after
CDH repair
Pulmonary failure

Yes

Yes

No

Yes

Yes

No

Yes

No

No

Yes

No

Yes

No

Yes

No

No

No

No

No

No

No
No
No

No
No
No

No

No

but still had increased wall thickness. This indicates


that these thick-walled acinar vessels are reactive
and capable of relaxing and contracting under optimal or adverse conditions, respectively. When rated
0 to 5, lung hemorrhage in controls was 1.2 6 1.0,
with no difference between left and right lungs. The
contralateral lungs of infants with CDH, with or
without ECMO, were not different and were combined (1.3 6 1.1). The ipsilateral lungs also were
similar, with or without ECMO. However, hemorrhage was significantly increased in the ipsilateral,
compared with the contralateral, lung (2.6 6 1.3; P ,
.02), suggesting that the severely hypoplastic lung is
more susceptible to damage from assisted ventilation. Of the 15 infants treated with nitric oxide, all
either died or required ECMO.15
DISCUSSION
Degree of Lung Hypoplasia and Pulmonary Arterial
Wall Thickness

It is known that there is a spectrum of pulmonary


hypoplasia in infants with congenital diaphragmatic
hernia.15,16 Pulmonary hypoplasia, at postmortem

CONGENITAL DIAPHRAGMATIC HERNIA LUNG VOLUME


Downloaded from www.pediatrics.org at Indonesia:AAP Sponsored on August 13, 2008

Fig 1. Ratio of RVSP to the SBP 3 100 to give the RVSP as a


percentage of SBP as a function of days on ECMO in surviving
infants. Each large circle represents a single measurement. The
increases in the ratio over time occurred after ECMO. The one
infant with the recurrent ratio .100 underwent a second ECMO
course after CDH surgery with a consequent fall to normal range.

examination, can be quantitated easily by measuring


lung DNA, wet weight, and volume. Our data suggest that values of at least 45% of predicted values
from age-matched controls are required for survival.
We have shown that despite remarkable decreases
in estimated pulmonary arterial pressure during 2
weeks of ECMO, small acinar arteries remain thickwalled, with increased muscularization. Therefore,
ECMO may have resulted in vessel wall thinning
from remodeling or muscle relaxation. The precise
effect cannot be determined from our data; however,
these small arteries can contract under adverse conditions.16 The persistent, increased muscularization
may be secondary to the small vascular surface area
of hypoplastic lungs.16 That is, the decreased vascular bed has to accommodate a normal cardiac output.
Anatomic studies indicate that the ipsilateral lung
is significantly more hemorrhagic than the contralateral lung. This increased hemorrhage suggests that
the most hypoplastic lung is more sensitive to high
ventilator pressures and volumes than the contralateral lung. These findings suggest that early overaggressive ventilator management may affect later lung
recovery.
Correlation of Degree of Lung Hypoplasia and
Vascular Changes With Functional Tests of
Oxygenation and Estimated RVSP

The OI, as an indicator of outcome in CDH, has


been shown generally to be nonpredictive,1 with
some dissent.3 The OI, whether assessed from preductal or especially from postductal arterial oxygen

Fig 2. Ratio of RVSP to the SBP 3 100 to give the RVSP as a


percentage of SBP as a function of days on ECMO in infants who
died during or after ECMO. Large open circles represent a single
measurement. Small filled circles indicate the day of cessation of
ECMO.

pressures, is highly dependent on the type of ventilatory support, pulmonary vascular resistance, size
of the ductus arteriosus, intrapulmonary shunts, and
systemic arterial pressure. Nevertheless, the OI was
significantly lower in infants not requiring ECMO.
The best OI of those infants with lung volume $45%
of predicted was ,30. With one exception, those
with the greatest lung hypoplasia had OI values .30.
Therefore, although statistically the pre-ECMO OI
appears to be correlated with the degree of lung
hypoplasia, in individual cases additional information is needed to predict outcome.
Although lung hypoplasia clearly coexists with
pulmonary hypertension in infants with CDH,
ECMO can alter this relationship. With ECMO, the
RVSP decreases within 6 to 14 days, even in the
smallest hypoplastic lung. The rate of decrease varied from infant to infant, but reaching the minimum
rarely exceeded 14 days. Qualitatively similar pulmonary vascular responses occurred in ECMOtreated infants who lived and those who did not
survive, in that the RVSP/SBP ratio decreased with
ECMO treatment, but the time on ECMO was prolonged. Our data suggest that serial echocardiograms and calculation of the RVSP/SBP ratio is a
valuable additional tool in monitoring the course of
pulmonary hypertension in infants with CDH. A
RVSP/SBP ratio of ,0.5 correlates generally with the
ability to discontinue ECMO and is desirable before
decannulation or proceeding to surgery. Even when

Downloaded from www.pediatrics.org at Indonesia:AAP Sponsored on August 13, 2008

ARTICLES

293

Contributory Clinical Factors for Mortality in


ECMO-treated Infants With CDH

The contributory factors for mortality in the eight


potentially survivable ECMO-treated infants included bleeding, early removal from ECMO without
achieving a normal RVSP, and perhaps, failure to
institute a second course of ECMO. If these eight
infants had survived, the highest overall survival
rate would have been 71%, (77%, if infants with
congenital anomalies are excluded). Although lung
hypoplasia and pulmonary hypertension are the
most important variables determining outcome, surgical timing and medical and surgical complications
are also significant variables.
Our results suggest that CDH treatment strategies
using ECMO should attempt to reduce the RVSP/
SBP ratio to ,0.50 before CDH repair. Infants with
lung volumes ,45% of predicted may have serious
problems with carbon dioxide exchange,20 which
may require longer ECMO use, or perhaps, a second
ECMO course.17,18 Lung volume as a unique bedside
predictor of outcome in CDH offers promise.13,2124
Finally, potentially preventable complications, not
related to severe pulmonary hypoplasia, are significant outcome variables.
Fig 3. Artery wall thickness at the alveolar duct level expressed
as a percentage of the diameter of the vessel as a function of age.
Open circles indicate individual measurements of controls. Solid
triangles represent infants who died before ECMO could be instituted. Solid circles represent infants who died during or after
ECMO. Values represent means 6 SD.

the estimated RVSP is normal, the acinar arteries are


not normal, and increased pulmonary vascular resistance can reoccur if the lung and vascular surface
area are too hypoplastic to support life or if adverse
conditions such as acidosis or hypoxia exist.16 A second course of ECMO should be considered if the
RVSP/SBP ratio increases and pulmonary deterioration occurs after CDH repair.17,18
Minimum Lung Volume Necessary for Survival

A lung volume of at least 45% of predicted values


appears to be the minimum needed for survival. Low
pump flows of 30 mL/min/kg may give adequate
cardiopulmonary support in infants with lung volumes ,45% of predicted values, even with low ventilator settings. However, discontinuing ECMO in an
infant with inadequate lung volume results in hypercarbia and the need for high ventilator support and a
return of pulmonary hypertension. Antunes et al13
have measured the functional residual capacity
(FRC) before CDH repair and showed, as we did,
that good outcome appears to correlate with the
largest lung volumes. Their CDH ECMO survivors
had an FRC of 12.3 mL/kg. Assuming that normal
FRC is 25 mL/kg,19 their survivor FRC was 49% of
predicted values, very close to our postmortem lung
volume assessments. It should be noted, however,
that our assessments were done at total lung volume
and included anatomic dead space.
294

ACKNOWLEDGMENT
This work was supported by the Katherine B. Richardson
Foundation.

REFERENCES
1. Lessin MS, Thompson IM, Deprez MF, Cullen ML, Whittlesey GC, Klein
MD. Congenital diaphragmatic hernia with or without extracorporeal
membrane oxygenation: are we making progress? Am Coll Surg. 1995;
181:6571
2. Geggel RL, Murphy JD, Langleben D, Crone RK, Vacanti JP, Reid LM.
Congenital diaphragmatic hernia: arterial structural changes and persistent pulmonary hypertension after surgical repair. J Pediatr. 1985;107:
457564
3. ORourke PP, Vacanti JP, Crone RK, Fellows K, Lillehei C, Hougen TJ.
Use of the post-ductal PaO2 as a predictor of pulmonary vascular
hypoplasia in infants with congenital diaphragmatic hernia. J Pediatr
Surg. 1988;23:904 907
4. Redmond D, Heaton J, Calix J, et al. A correlation of pulmonary hypoplasia, mean airway pressure, and survival in congenital diaphragmatic
hernia treated with extracorporeal membrane oxygenation. J Pediatr
Surg. 1987;22:11431149
5. Charlton AJ, Bruce J, Davenport M. Timing of surgery in congenital
diaphragmatic hernia. Anesthesia. 1991;46:820 823
6. Langham MR Jr, Krummel TM, Bartlett RH, et al. Mortality with extracorporeal membrane oxygenation following repair of congenital diaphragmatic hernia in 93 infants. J Pediatr Surg. 1987;22:1150 1154
7. Sigalet D, Tierney A, Adolph V, et al. Timing of repair of congenital
diaphragmatic hernia requiring extracorporeal membrane oxygenation
support. J Pediatr Surg. 1995;30:11831187
8. Berger M, Haimowitz A, VanTosh A, Berdoff RL, Goldberg E. Quantitative assessment of pulmonary hypertension in patients with tricuspid
regurgitation using continuous wave Doppler ultrasound. J Am Coll
Cardiol. 1985;6:359 365
9. Thibeault DW, Mabry S, Rezaiekhaligh M. Neonatal pulmonary oxygen
toxicity in the rat and lung changes with aging. Pediatr Pulmonol. 1990;
9:96 108
10. Thibeault DW, Beatty EC, Hall RT, Bowen SK, ONeill DH. Neonatal
pulmonary hypoplasia with premature rupture of fetal membranes and
oligohydramnios. J Pediatr. 1985;107:273277
11. Dische Z. Color reactions of nucleic acid components. In: Chagoff E,
Davidson J, eds. The Nucleic Acids. New York, NY: Academic Press,
1955:285305
12. Meyrick B, Reid L. Pulmonary arterial and alveolar development in
normal postnatal rat lung. Am Rev Respir Dis. 1982;125:468 473
13. Antunes MJ, Greenspan JS, Cullen JA, Holt WJ, Baumgart S, Spitzer AR.

CONGENITAL DIAPHRAGMATIC HERNIA LUNG VOLUME


Downloaded from www.pediatrics.org at Indonesia:AAP Sponsored on August 13, 2008

14.
15.

16.

17.

18.

Prognosis with preoperative pulmonary function and lung volume


assessment in infants with congenital diaphragmatic hernia. Pediatrics.
1995;96:11171122
Haworth SG, Hislop AA. Pulmonary vascular development: normal
values of peripheral vascular structure. Am J Cardiol. 1983;52:578 583
Kinsella JP, Truog WE, Walsh WF, et al. Randomized, multicenter trial
of inhaled nitric oxide and high frequency oscillatory ventilation in
severe persistent pulmonary hypertension of the newborn (PPHN).
Pediatr Res. 1996;39:222A
Vacanti JP, Crone RK, Murphy JD, et al. The pulmonary hemodynamic
response to perioperative anesthesia in the treatment of high-risk infants with congenital diaphragmatic hernia. J Pediatr Surg. 1984;19:
672 678
Dela Cruz TV, Stewart DL, Robinson TW, Bond SJ. The use of a second
course of extracorporeal membrane oxygenation in neonatal patients.
ASAIO J. 1996;42:230 232
Lally KP, Breaux CW Jr. A second course of extracorporeal membrane
oxygenation in the neonateis there a benefit? Surgery. 1995;117:

175178.
19. Berglund G, Karlberg P. Determination of the functional residual capacity in newborn infants. Acta Paediatr. 1956;45:541544
20. West JB. Causes of carbon dioxide retention in lung disease. N Engl
J Med. 1971;284:12321236
21. Dimitriou G, Greenough A, Laubscher B. Lung volume measurements
immediately after extubation by prediction of extubation failure in
premature infants. Pediatr Pulmonol. 1996;21:250 254
22. da Silva WJ, Abbasi S, Pereira G, Bhutani VK. Role of positive endexpiratory pressure changes on functional residual capacity in surfactant treated preterm infants. Pediatr Pulmonol. 1994;18:89 92
23. Antunes MJ, Cullen JA, Holt WJ, Gauthier TW, Baumgart S, Greenspan
JS. Continued pulmonary recovery observed after discontinuing extracorporeal membrane oxygenation. Pediatr Pulmonol. 1994;17:143148
24. Goldsmith LS, Greenspan JS, Rubenstein D, Wolfson MR, Shaffer TH.
Immediate improvement in lung volume after exogenous surfactant:
alveolar recruitment versus increased distention. J Pediatr. 1991;119:
424 428

Downloaded from www.pediatrics.org at Indonesia:AAP Sponsored on August 13, 2008

ARTICLES

295

Lung Volume, Pulmonary Vasculature, and Factors Affecting Survival in


Congenital Diaphragmatic Hernia
Donald W. Thibeault and Barbara Haney
Pediatrics 1998;101;289-295
DOI: 10.1542/peds.101.2.289
Updated Information
& Services

including high-resolution figures, can be found at:


http://www.pediatrics.org/cgi/content/full/101/2/289

References

This article cites 21 articles, 2 of which you can access for free
at:
http://www.pediatrics.org/cgi/content/full/101/2/289#BIBL

Citations

This article has been cited by 7 HighWire-hosted articles:


http://www.pediatrics.org/cgi/content/full/101/2/289#otherarticle
s

Subspecialty Collections

This article, along with others on similar topics, appears in the


following collection(s):
Respiratory Tract
http://www.pediatrics.org/cgi/collection/respiratory_tract

Permissions & Licensing

Information about reproducing this article in parts (figures,


tables) or in its entirety can be found online at:
http://www.pediatrics.org/misc/Permissions.shtml

Reprints

Information about ordering reprints can be found online:


http://www.pediatrics.org/misc/reprints.shtml

Downloaded from www.pediatrics.org at Indonesia:AAP Sponsored on August 13, 2008