Can Blood Gas Values Predict Pulmonary Hypoplasia in Antenatally

7/25/2019 Can Blood Gas Values Predict Pulmonary Hypoplasia in Antenatally

1/6

Can Blood Gas Values Predict Pulmonary Hypoplasia in Antenatally

Diagnosed Congenital Diaphragmatic Hernia?

By J.F . Germain, C. Farnoux, D. Pinquier, A. Cortez, J.F. Hartmann, 0. Sibony, P. de Lagausie, and F. Beaufils

Paris, France

0 The prognosis of antenatally diagnosed congenital dia-

phragmatic hernias (CDH) is clearly related to the degree of

pulmonary hypoplasia (PH). After birth, controversies remain

regarding the implem entation of various therapies, espe-

cially the use of extracorporeal membrane oxygenation

(ECMO). In the literature, the persistence of a Pa02 below 100

mm Hg and of Paco2 above 40 mm H g despite optim al

conventional therapy indicates poor prognosis. Therefore,

since 1992, published and personal experiences led the

authors to exlude CDH patients from ECMO when conven-

tional therapy (including high-frequency oscillatory ventila-

tion and nitric oxide) did not obtain Paos of above 80 mm Hg

and Pacop of below 60 mm Hg. The aim of this retrospective

study is to determine whether blood gas results correlate

with postmortem findings. Between July 1990 and July 1994,

32 cases of CDH were monitored antenatally and managed

postnatally at the authors institution. Six patients survived;

26 died, including one immedia tely at birth. Thirteen were

treated by ECMO. Seventeen had a best Pa o2 of above 80 mm

Hg, including the six survivors. Fourteen did not reach this

level, and none of them survived. Twenty-three infants

underwent postmortem examination. PH was assessed using

two criteria: (1) lung weight to body weight ratio (LW/BW)

and (2) radial alveolar count (RAC). Two patients did not have

hypoplasia (LW/BW > 0.018). Twenty-one patients had PH;

12 of them had an LW/BW ratio of less than .009; for 9, the

LW/BW ratio was between .009 and .018, and the RAC (~3.1)

confirmed PH. All infants with a best Pao2 of less than 80 mm

Hg had PH. Patients with a best Paoz of greater than 80 mm

Hg included two infants who died from complications with-

out PH. eight infants with demonstrated PH. and the six

survivors. In conclusion. (1) No infant with nonhypoplastic

lungs has been deprived of ECMO by the authors criteria. (2)

Adequate values of blood gases may not elimin ate PH.

Therefore, this probably justifiesstatting ECMO when conven-

tional therapy fails. (3) Conversely, permanent poor values of

Paoz allowed the prediction of PH in all cases. Such patients

probably can be excluded from ECMO treatment.

Copyright o 1996 by WA Saunders Company

INDEX WORDS: Congenital diaphragmatic hernia, extracor-

poreal membrane oxygenation, pulmonary hypoplasia, post-

mortem examination.

LTHOUGH diversely appreciated, the mortality

A

rate associated with antenatally diagnosed CDH

remains above 50 .1-3 Prenatally diagnosed CDH is

associated with variable degrees of pulmonary hypo-

From the Faculty of Medicine, HEpit al Robert De b&, Paris, France.

Address reprint requests to J.F. Germain, Servrce de Pkdiatne 2

(R&anim ation P+?diatrique Polyvalente), HZpital dEnfants, 10, Bd du

Markhal de-Lattre-de-Tassigny, 21034 Dijon Cedex, France.

Copyright o 1996 by W.B. Saunders Company

0022-3468/96/3112-0006$03.00/O

plasia (PH), which involves the ipsilateral and contro-

lateral lungs. The short-term prognosis of CDH is

determined by anatomic and functional abnormali -

ties.4 In addition to PH, there is a pulmonary vascular

hyperreflectivity associated with pulmonary vascular

hypoplasia, which results in increased pulmonary

vascular resistance. In these conditions, right-to-left

extrapulmonary shunting may lead to severe or refrac-

tory hypoxemia. Although data show that nitric oxide

(NO) can reduce pulmonary vascular resistances in

persistent pulmonary hypertension of the newborn5

its efficacy in CDH remains to be demonstrated.6

To what extent PH is responsible for the poor

outcome of antenatally diagnosed CDH is not clear.

A number of these patients die of potentially revers-

ible causes7 and the heterogeneity of study popula-

tions makes comparisons difficult. Clinical uncertain-

ties regarding PH make ECMO decisions difficult.

Should ECMO be instituted in every case or should it

be limited to patients without severe PH? In fact, the

question is whether or not lethal PH can be predicted.

Until now, no antenatal or immediately postnatal

criterion appears to be clearly predictive of lethal PH.

Because of the lack of established clinical criteria for

PH, we performed a retrospective study in 32 infants

who had antenatally diagnosed CDH, to determine

whether blood gases resul ts correlated with post-

mortem findings.

MATERIALS AND METHODS

The records of 32 inborn infants who had CDH diagnosed

antenatally at our institution between June 1990 and June 1994

were reviewed. All were managed with the same protocol, includ-

ing immediate intubation, mechanical ventilation, volume loading,

and monitoring of preductal and postductal saturations. All pa-

tients were sedated (midazolam or fentanyl) and paralyzed (vecuro-

nium). After admission to the pediatric intensive care unit (PICU),

arterial and central venous lines were inserted, and blood gases

were noninvasively and continuously monitored using preductal

and postductal pulse oxymetry and combined transcutaneous

Poz/Pcoz. In addition, all infants underwent sequential echocardio-

graphic studies using a combined two-dimensiona l echo-Doppler

system (sono layer SSH-160 A, Toshiba, Tochigi-ken, Japan) in the

attempt to determine the degree of pulmonary hypertension and

left ventricle function. The studies were performed as soon as

possible after initi al stabilization, and were repeated as needed.

Ventilation Management

From June 1990 through the end of 1991, the patients were

ventilated mechanically with Servo 900C (Siemens, Germany).

1634

JournalofPediatm Surgery, Vol31, N o 12 (December), 1996: pp 1634.1639


2/6

PREDICTION OF PULMONARY HYPOPLASIA

Starting in 1992, all patients were then managed with high-

frequency oscillatory ventilati on* (HFOV; OHFl, D ufour, France).

Management

of

Persistent Pulmonary Hypertension

Respiratory acidosis was avoided, without seeking hypocapnia.

Prostacyclin was used until 1993, when NO9 was introduced. NO

and NO2 concentrations were monitored continuously using Poly-

tron (Drager-Industrie SA, Strasbourg, France). NO was started at

5 ppm and was increased subsequently, if needed, to 20 ppm.

ECMO Management

Indications for veno-arterial ECMO varied over time. Between

June 1990 and February 1992, ECMO was started in the absence of

the usual contraindications, for each patient who did not respond

to conventional therapy. In practice, the indication for ECMO was

determined by the persistence of an alveolo-arterial oxygen gradi-

ent (AaDo = [barometric pressure - 471

X I302

- Pacoz

[FIo? + (1 - FIOZ)/O.S] - Pao 2) higher than 610 mm Hg for more

than 8 hours,O and/or an oxygenation index (01 = mean airway

pressure x FIoz%/postductal Paoz) above 40 for more than 4 to 6

hours. Patients who experienced acute deterioration or untrac-

table hypercarbia also were treated with ECMO.

In February 1992, published and personal experiences showed

that poor blood gas values despite maximal conventional therapy

(best

Paoz < 100 mm Hg and best Pacoz > 40 mm Hg) were

associated with a poor prognosis,2 with or without ECMO.

Therefore, we believed that such patients should be excluded from

ECMO treatment. However, for ethical reasons, the value was

decreased to 80 mm Hg for the best preductal Paoz, and increased

to 60 mm Hg for the best Paco2. Patients not fulfil ling those criteria

were not offered ECMO.

Surgey

Patients underwent surgical repair only after stabilization had

been achieved, whether after conventional therapy or after success-

ful weaning from ECMO (F102 below .4 using conventional

therapy, with or without NO). No infant undetwent surgery during

ECMO therapy.

Parental Information

At the time of diagnosis and again at birth, parents were

informed of the protocol outline by a team that included a senior

intensivist and a senior pediatric surgeon. In addition, NO has

been used with informed parental consent.

Anatomic and Histological Studies

When informed parental consent was obtained, every patient

who died underwent a postmortem examination. After seeking

associated malformations, PH was assessed. The study was based

on two critena.lJ I4

1. The ratio of lung weight to body weight (LWIBW) was the

major criteria. I3 PH was considered as certain and major for cases

with an LW/BW below ,009, probable for those with an LW/BW

between ,009 and ,018, and absent for those with an LWiBW above

,018.

2. The radial alveolar count (RAC), as assessed by morphomet-

ric methods, was the average of the ipsilateral and controlateral

radial alveolar counts. This value was calculated from 60 to 100

counts, and from 10 histological sections. In practice, a line was

drawn from the center of each respiratory bronchiole to the nearest

connective tissue septum at right angles to the epitheli um, and the

number of alveoh included were counted. This was assessed

according to the Emery and Mithal 14 procedure, which can be

1635

related to the Azkenazi method13 by the followi ng equation: RAC

alveolar count = RAC alveolar septal count - 1.

Severe PH is easy to diagnose from the low LW/BW alone.13

Because lesser or borderline PH is more difficult to establish,13 and

because possible overestimation of the lung weight (pulmonary

edema, major pulmonary hemorrage) was taken into account for

each study. RAC was also used to contrast uncerta in cases, after

assessment of the LWIBW. Azkenazi et all3 took into account a

cutoff level of 4.1 for the RAC, but the number of alveolar septi was

considered. Because we counted the number of alveoli, severe PH

was asserted when RAC was less than 3.1.

Statistical Analysis

Data are presented as mean f standard deviation, or medians

and ranges. The features of the groups of patients were compared

using the Yates corrected x test or a nonparametric test (Mann-

Whitney U test). Pvalues of I .05 were considered significant.

RESULTS

Seventeen boys (53 ) and 15 gir ls (47 ) were

diagnosed as having CDH at our institution between

June 1990 and June 1994. Al l were inborn and

managed by the same team.

Clinical Resuits

Clin ical data and outcome are summarized in

Table 1. Antenatal diagnosis was made at an average

of 24 weeks gestation (range, 15 to 37 weeks). One

infant died despite resuscitation in the delivery room.

Table 1. Patient Data, Managem ent, and Outcome

Mean t ime of diagnosis (weeks of

gestation)

Polyhydrammos

Side o f CDH

Hermated organs

Birth asphyxia

Cesarean delivery

Term of birth (weeks of gesta-

t ion): range (median)

Birth weight (9): range (median)

Mean Apgar at I min:

range (median)

Mean Apgar at 5 min:

range (median)

Conventional therapy

ECMO

Surgery

Outcome

24 +- 5

n = 7 (22%)

left: n = 28 (87.5%)

right: n = 3 (9.4%)

bilateral, n = 1 (3.1%)

small bowel: n = 32 (100%)

stomach: n = 11 (37.9%)

Ilver: n = 1 (3.5%)

stomach and l iver:

n = 17 (58.6%)

spleen, n = 29 (91%)

n = 11 (34.5%)

n = 4 (12.5%)

35-41 (39)

1,510-3,800 (2,980)

o-9 (4.1)

O-IO (6.4)

MV: n = 10 (32.25%)

HFOV: n = 10 (32.25%)

HFOV and NO: n = 11 (35.5%)

n = 13 (40.6%)

n = 7 (21.9%)

Survivors, n = 6 (18.75%)

NonsurvIvors: n = 26 (81.25%)

Abbreviations: MV, mechanical venti lat ion; HFOV, high-frequency

oscillatoryv entllatlon; NO, mtric oxide.


3/6

1636

GERMAIN ET AL

Of the 31 remaining patients, one third were treated

with conventional ventilat ion (from June 1990 to the

end of 1991) one third with HFOV (1992), and one

third with HFOV combined with NO (since 1993).

Three patients responded to conventional therapy

alone and survived. In the other 28 cases, the alveolo-

arterial oxygen gradient and oxygenation index values

met the classical criteria for ECMO consider-

ation.lO,ll Thirteen of them were treated with ECMO;

four of these could be operated on, three of whom

survived. Nine patients died while receiving ECMO.

Fifteen patients were not treated with ECMO; four

of them had multiple lethal malformations, two expe-

rienced acute deterioration and died before ECMO

could be started, and nine had best blood gas values

that were considered too poor according to our

protocol (Paoz < 80 mm Hg or Pace, > 60 mm Hg).

All these patients died. Overall, six of the 32 patients

survived.

Blood Gas Results and Derived Severity Indexes

For 17 patients the best Paoz was above 80 mm Hg

(Table 2). The Al veolo-arterial oxygen gradient and

oxygenation index were significantly lower for these

patients compared with the 14 whose best Pao2 was

below 80 mm Hg (Z = .0004 and .OOOl, respectively).

Twenty-two patients had a best Pacoz below 60 mm

Hg.

Anatomic and Histological Results for Nonsurvivors

Twenty-three anatomic and histological studies

were performed; 22 were complete (necropsies) and

one included only multiple chest biopsies, from which

the RAC could be evaluated. For three patients,

postmortem examination was denied by the family;

one of these patients died immediately in the delivery

room, one had very poor blood gas values (best Paoz,

31 mm Hg; best Pacoz, 72 mm Hg) and died short ly

after birth, and one had better blood gas values (best

Table 2. Best Blood Gas Values and Best Derived Severity Indexes

Obtained Before ECMO Decision IHo2 = 1)

No. of Mea n Pm> Mean Pam, Mean Aam~

Patients

(mm Hg) (mm Hg)

Mean 01

(mm W

Best Pa0,

r80mmHg

17 288 f 142 31 2 14 12 IL 15 344-t 182

60mmHg 9 44 t 23 84k20 39k12

594k20

NOTE. Mean values were calculated for 31 patients in whom blood

gas samples could be obtained.

Abbreviations: AaDo,, alveolo-arterial oxygen gradient (FIo, = 1); 01,

oxygenation index

o ,0275 4 t

F

2 ,025 -

$ ,0225

-

8

3 ,02

& ,0175 -

s

P ,015 .

O

,uvxl 000 , , , t

1,5 2 2,5 3 3,5

4 4,5 5

RADIAL ALVEOLAR COUNT

Fig 1. Correlat ion of two methods of assessment of pulmonary

hypoplasia: lung weight to body weight rat io and radial alveolar

count. The equa tion for the regression line is y = .008x - .Ol. The

correlat ion is signif icant, with r = .8 and P = .OOOl. The horizontal l ine

represents the 3.1 l imit for the radial alveolar count. The vert ical l ine

represents the .009 l imit for the lung weight to body weight rat io.

Values below those l imits are indicative of severe pulmonary hypopla-

sia.

Paoz, 100; best Paco2, 24) but died despite ECMO

therapy.

Macroscopic postmortem examinations confirmed

CDH in all cases. Additional abnormalities were

hemorrhagic and/or thrombotic complications second-

ary to ECMO therapy (7 patients), multiple lethal

malformations (2 Frynz syndrome, 1 Cornelia De

Lange syndrome, and 1 Marfan syndrome), 4 urinary

tract malformations, 3 cardiac malformations (includ-

ing 2 left ventricle hypoplasias), and 1 congenital

cystic adenomatoid malformation of the lung. Only

seven patients had isolated CDH with PH.

In regard to PH, the mean LW/BW was .Ol f .006

(range, .003 to .025) and the mean RAC was 2.55 +

0.59 (range, 1.7 to 4.45). Lung weight was considered

to be overestimated for 10 patients, because of

pulmonary edema or pulmonary hemorrhage. Both

methods of PH assessment were correlated (r = .8,

P = .OOOl; Fig 1). PH was confirmed in all but two

patients; one of these died after nosocomial septice-

mia, and the other after arterial canula thrombosis

during ECMO.

Assessment of Predictive Criteria

of Pulmonary Hypoplasia

Comparison of patients with and without severe PH.

From outcome and postmortem data, two groups of

patients were distinguished. The pulmonary hypopla-

sia group (H) included 21 nonsurvivors; for 12 of

them the LW/BW was below .009; for the other 9 the

LWIBW was between .009 and .018, but the RAC was

below 3.1 (2.54 2 .34). The nonhypoplastic group


4/6

PREDICTION OF PULMONARY HYPOPLASIA 1637

(NH) consisted of two nonsurvivors devoid of PH at

postmortem study, and the six survivors for whom a

lack of marked PH was assumed in light of the

satisfactory outcome.

Both groups are compared in Table 3. In summary,

when compared with group NH, group H had earlier

antenatal diagnosis (22 v 30 weeks gestation), a

poorer Apgar scores (1 minute: 3 v 6; 5 minute: 6 v 9),

more air leaks (67 v 25 ), and worse mean values

of blood gases and derived indexes.

Value of best blood gases during conventional therapy

for predicting PH.

One patient died at delivery

before any blood gas value could be obtained. For the

other 31 patients, the best Pao, during conventional

therapy is detailed in Table 4. Also, the best Pacoz,

the LW/BW, and the RAC are indicated. For 14

patients the best Paoz was below 80 mm Hg. Al l died.

Thirteen of them (93 ) underwent a histological and

anatomic study; all had PH.

Seventeen patients had a best Paoz above 80 mm

Hg, including 12 above 200 mm Hg. Six patients

survived. Among the 11 nonsurvivors, eight had

severe PH, two did not but died from an iatrogenic

cause, and postmortem study was denied in one.

Concerning Pace,, al1 patients with values above 60

mm Hg despite appropriate therapy were severely

hypoplastic.

Comparison of best blood gases according to pre-

ECMO management.

Best blood gases were com-

pared in

nonsurvivors with established PH

according to

the ini tial management (Fig 2). Paoz and Pace, were

significantly better in the eight hypoplastic patients

treated with the association of HFOV and NO (mean

best Paoz, 228 -+: 177 mm Hg; mean best Pacoz,

22 + 4 mm Hg) when compared with the seven

Table 3. Comparison of Patients with (Group H) and Without (Group

NH) Pulmonary Hypoplasia

Group H Group NH P

In = 21)

(n = 8)

Va lue

Time of antenatal diagnoses

(weeks of gestation)

Birth asphyxia

Term of birth (weeks of ges-

tation)

Birth weight(g)

Apgar a t 1 m in

Apgar a t 5 m in

Arr lakes

Best Paoz (mm Hg)

Best Pace, (mm Hg)

Best AaDo, (mm Hg)

Best 01

Survrvors

22.1 2 3.6

29.8 k 4.2

.0004

n = 8 (38.1%) n = 1 (12.5%) NS

39.1 + 1.5 38.9 f 1 NS

3,088 2 527 3,203 2 493 NS

3.3 t 2.2 6.4 + 1.9

,005

5.9 f 1.6 9.2 + 0.7 .OOOl

= 14 (66.6%) n = 2 (25%) .04

122.5 k 136.3

359.3 2 74.9

.0015

51.1 + 31.5 30.9 2 7.4 NS

521.8 k 1658 266.8 f 110.8 ,002

32.6 + 19.2 4.7 k 1

,001

0 6

.OOOl

Abbreviations: Aaoo,, oxygen alveolo-arterial gradient (Fi02 = 1); 01,

oxygenation Index: NS , not signif icant.

Table 4. Occurrence of Pulmonary Hypoplasia According to the Best

Pao, Value Obtained During Conventional Therapy

Pa02

P X Q

( m m H d ( m m H d

RAC LWIBW PH

Best Pao, 580 mm Hg

(n = 14)

Patient 3 30 92 2

Patient 5 31 72

Patient 6 43 121 2.9

Patient 8 44 64 2.6

Patient 10 37 105 2.65

Patrent 11 30 95 2.2

Patient 12 77 41 2.4

Patient 13 45 65 2.55

Patient 15 59 40 2.85

Patient 16 50 46 3

Patient 21 35 69 1.9

Patient 26 64 24 2.15

Patient 29 65 27 1.85

Patient 32 72 24 2.2

.0031

-

,008

.0042

.0033

.0097

.0116

,008

.0114

.0034

.0056

Mean values

49 + 16 63 2 31 2.4-t-.39 ,007 k.003

Best Pao, > 80 mm Hg

(n = 17)

Patient 1 352

Patient 2 105

Patrent 7 400

Patient 9 221

Patient 14 100

Patrent 17 100

Patient 18 295

Patient 19 374

Patient 20 426

Patient 22 100

Patient 23 457

Patient 24 252

Patient 25 93

Patient 27 377

Patient 28 503

Patient 30 391

Patrent 31 349

47 3.55

73 2.7

28 4.45

36 -

33 2.65

49 2.6

30

29

26 -

24

25

22

1.7

28 2.3

20

2.5

14 2.7

21 2.15

26 -

.0226

.0146

.0248

.0129

,017

-

.Ol

.0073

.0128

.0067

Mean values 288 t 142 31 + 14 2.73 k .77 ,014 2.006

Yes

NA

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

No

Yes

No

Survrvor

Yes

Yes

Survivor

Survivor

Survivor

NA

Survivor

Yes

Yes

Yes

Yes

Yes

Abbreviations: RAC, radial alveolar count; LW/BW, lung werght to

body weight rat io; PH, pulmonary hypoplasia; NA, data not avai lable.

hypoplastic patients treated with HFOV alone (mean

best Paoz, 54 + 19 mm Hg; mean best Pacoz, 58 2 20

mm Hg) or with the six hypoplastic patients treated

with conventional ventilat ion alone (mean best Pao2,

59 it 33 mm Hg; mean best Paco2, 81 & 31 mm Hg).

DISCUSSION

Despite regular progress in respiratory manage-

ment, antenatally diagnosed CDH remains associated

with a mortal ity rate estimated at 40 to 80 1-3 and

significant morbidity. This poor prognosis is related to

several factors: associated lethal malformations, se-

vere PH, refractory hypoxemia induced by pulmonary


5/6

1638

GERMAIN ET AL

vivors. The anatomic and histological criteria used to

define PH in this study were LW/BW and RAC, as

derived from criteria of Askenasi et a1.13 Our data

confirmed the very strong correlation between both

variables, and allowed two major conclusions. First,

all patients unable to achieve a Pao, of 100 mm Hg

I

(the critica l level used in the literature12) or more

HFO

died, even when ECMO was used. All had severe PH.

AND O

Therefore, poor Paoz is predictive of severe PH.

Fig 2. Comparison of best blood gas values according to pre-

ECMO managem ent in patients with pulmonary hypoplasia. The

asterisk indic ates s ignificant differences betwe en the six patients

treated with mechanical venti lat ion (NIV) alone and the eight patients

treated with high-frequency oscillatory ventila tion (HFO) comb ined

with nitric oxide (NO). Dollar sign indicates significan t differences

between the seven patients treated with HFO alone and the eight

patients treated with HFO and NO.

vasoconstriction, and iatrogenic causes secondary to

various technical methods.7

Several attempts to define reliable prognostic fac-

tors have been made. Prenatally, echographic vari -

ables, such as left-to-right ventricular internal diam-

eter ratio, have been used to predict outcome,i5 and

ventricular disproportion identified before 24 weeks

gestation is associated with fatal outcome.r6 Thoracic-

to-abdominal transversal ratio is another echographic

measurement,17 which has been correlated with post-

natal outcome in cases of severe oligohydramnios, but

it could not be used in the study of the CDH lung.

Other attempts to define prenatally the postnatal

prognosis are based on fetal age at time of diagnosis,

the presence of polyhydramnios, or of mediastinal

shift, and the intrathoracic location of a dilated

stomach. Except for the latter, associated with a poor

prognosis,ls these data are of limited value,19-21 and

none can provide any help with respect to the

antenatal or postnatal decision for withdrawal.

Postnatally, the prognostic value of blood gases and

derived indexes has been highlighted by several

studies.12J2-25 However, few of them have involved

evaluating the degree of PH. Because ECMOlOJ1 and

N05,6 may allow a good prognosis when refractory

hypoxemia is related to acute pulmonary vasoconstric-

tion, but will only delay a fatal outcome when severe

PH is present, it appears necessary to identify PH

early in the postnatal course to avoid unnecessary and

costly procedures.

The degree of PH has been estimated from chest

radiographs, but it appears to be overestimated using

this approach. 26 In a study that included 66 patients,

Bohn et al showed the predictive value of Paco2.37

However, only five postmortem examinations were

performed in these patients. To better define criteria

for indisputable PH, we related poor blood gas values

to postmortem data obtained for most of our nonsur-

Concerning Pacoz, the same conclusion can be drawn

from our data; all patients unable to achieve a Paco2

below 60 mm Hg died and were severely hypoplastic.

Moreover, for ethical reasons, we lowered the limit of

100 mm Hg of Pao2 (taken from data in the litera-

ture12 ) to 80 mm Hg. Avoiding ECMO in these cases

of low Paoz and high Pace,, as we did for a number of

patients, would not deprive the patients of a poten-

tially beneficial procedure.

However, and this is our second major finding, in

this study, Pao2 of above 100 mm Hg during conven-

tional management does not allow any conclusion

regarding PH; among patients with Paoz in excess of

100 mm Hg were survivors, nonsurvivors with PH, and

nonsurvivors without PH. The same uncertainty con-

cerning outcome and PH is shown in our data when

Paco2 is below 60 mm Hg. Advances in conventional

management have been achieved; for patients whose

postmortem examination demonstrated severe PH,

Pao2 was significantly higher and Pacoz significantly

lower in those treated with HFOV and NO as

compared to those who had HFOV or conventional

therapy alone. Although the patients were not ran-

domized, (because comparisons between those groups

have been taken from historical data), these resul ts

support the assertion that high Pao2 values do not

always correlate with the absence of PH. Therefore,

offering ECMO to these patients remains mandatory

despite uncertainties regarding its benefit.

In conclusion, a poor best Pao2 (worse than 100

mm Hg) is always associated with severe PH and poor

outcome, and such patients probably can be excluded

from ECMO treatment, a decision that remains

ethical in light of our postmortem data. On the other

hand, a best Pao2 of above 100 mm Hg does not

always correlate with a good outcome or the absence

of PH. All such patients should be treated with

ECMO when this procedure is necessary. However,

we believe that this approach, as opposed to the

ECMO for all attitude,28-30 can be safely imple-

mented only if the conditions of this study, especially

birth, resuscitation, and postnatal management in the

same institution, and the use of modern therapies like

HFOV and NO, can be reproduced.


6/6

PR ED I CT I O N O F PU LM O N AR Y H YPO PLASI A

1639

REFERENCES

1. Adzick NS, Harr ison MR , G lick PL, et al: Diaphragmatic

hernia m the fetus : Prenatal diagnosis and outcome in 84 cases . J

Pediatr Surg 20:357-361,1985

2. Har rison MR , Adzick NC, Estes JM, et a l : A prospective

study of the outcome for fetuses with diaphragmatic hernia. JAMA

271:382-384,1994

3. Manni M, Heydanus R, den Hollander NS, et al: Prenatal

diagnosis of congenital diaphragmatic hernia: A retrospective

analysis of 28 cases. Prenat Diagn 14:187-190,1994

4. Simson JNL, Eckstein HB: Congenital diaphragmatic hernia:

A 20 year experience. Br J Surg 72:733-736,1985

5. Roberts J D, Polaner DM , Lang P, et al: Inhaled nitr ic oxide

in persistent pulmonary hypertension of the newborn. Lancet

340:818-819,1992

6. Karamanoukian HL, Glick PL, Zayek M, et al: Inhaled nitr ic

oxide in congenital hypoplasia of the lungs due to diaphragmatic

hernia or oligohydra mnios. Pedia trics 94:715-718,1994

7. Price MR , G alantowicz ME , Stolar CJH: Congenital diaphrag-

matic hernia, extracorporeal membrane oxygenation, and death: A

spectrum of etiologies. J Pediatr Surg 26:1023-1027, 1991

8. Clark R H, Yoder BA, Sell MS: Prospective, randomized

comparison of high-frequency oscil lat ion and conventional ventila-

t ion in candidates for extracorporeal membrane oxygenation. J

Pediatr 124:447-454, 1994

9. Geggel RL: Inhalational nitr ic oxide: A selective pulmonary

vasodilator for treatment of persistent pulmonary hypertension of

the newb orn. J Pediatr 123:76-79, 1993 (editorial)

10. Short BL, Pearson GD : Neonatal extracorporeal membrane

oxygenation: A review. J Int Care Med 1:47-54,1986

11. Bartlett RH, Gazzaniga AB, Toomasian AB: Extracorporeal

membrane oxygenation (ECMO) in neonatal respiratory failure.

Ann Surg 204:236-245,1986

12. Wilson JM , Lund DP. Lil lehei CW , et al: Congenital

diaphragmatic hernia: P redictors of severity in the ECM O era. J

Pediatr Surg 26:1028-1034,199l

13. Askenasi SS, Perelman M: Pulmonary hypoplasia: Lung

weight and radial alveolar c ount as cr iteria of diagnosis. Arch Dis

Child 54:614-618,1979

14. Cooney TP, Thurlbeck WM : The radial alveolar count

method of Emery and Mithal: A reappraisal. Intrauter ine and early

postnatal lung growth. Thorax 37:580-583,1982

15. Karamanoukian HL, G lick PL: Cardiac function in fetuses

with congenital diaphragmatic hernia. JAMA 272:29, 1994

16. Sharland GK, Lockhart SM, Heward AJ, et al: Prognosis in

fetal diaphragmatic hernia. Am J Obstet Gynecol 166:9-13,1992

17. Johnson A , Callen NA , Bhutani VK, et al: Ultrasonic ratio

of fetal thoracic to abdominal circumference: An association with

fetal pulmonary hypoplasia. Am J Obstet Gy necol 157:764-769,

1987

18. Burge DM , Atwell JD, Freeman N V: Could the stomac h site

help predict outcome in babies with left sided congenital diaphrag-

matic hernia diagnosed antenatally? J Pediatr Surg 24:567-569,

1989

19. Glick PL, Leach C L, Besner GE, et al: Pathophysiology of

congenital diaphragmatic hernia III: Surfactant replacement for

the high-r isk neonate with congenital diaphragmatic hernia. J

Pediatr S urg 28:1-4,199 Z

20. Harr ison MR , Langer JC , Adzick NS, et al: Correction of

congen ital diap hragma tic hernia in utero. V. Initial c linical experi-

ence. J Pediatr Surg 25:47-57,199O

21. Harr ison MR , Adzick NS, Flake AW , et al: Correction of

congenital diaphragmatic hernia in utero. VI. Hard earned lessons.

J Pediatr Surg 28:1411-1418, 1993

22. Atkinson JB, Poon MW : ECM O and the management of

congenital diaphragmatic hernia with large diaphragmatic defect

requiring prosth etic patch . J Pediatr Su rg 27:754-756,1992

23. Bohn DJ, James I, Fil ler RM , e t al: The relationship

between Pacoz and ventilat ion parameters in predicting survival in

congen ital diap hragma tic hernia. J Pediatr Surg 19:666-671,1984

24. ORourke P, Vacanti J, Crone R, et al: Use o f postductal

Pao? as predictor of pulmonary vascular hypoplasia in infants w ith

congen ital dia phragm atic hernia. J Pediatr Surg 23:904-907 , 1988

25. Johnston P W, Liberman R. Gangitano E, et al: Ventilation

parameters and arterial blood gases as a predictor of hypoplasia in

congen ital dia phragm atic hernia. J Pediatr Surg 25:496-499 , 1990

26. Cloutier R, Allard V, Fournier L, et al: Estimation of lungs

hypoplasia on postoperative chest x-rays in congenital diaphrag-

ma tic hernia. J Pediatr Surg 28:1086-1089 , 1993

27. Bohn P, Tamura M, Perrin D, et al: Ventilatory predictors of

pulmon ary hypoplas ia in congen ital diaphragm atic hernia, con-

firmed by morphologic asses sme nt. J Pediatr 11:423-431,1987

28. Bailey PV, Connors RH , Tracy FT, et al: A cr it ical analysis

of extracorporeal membrane oxygenation for congenital diaphrag-

matic hernia. Surgery 106:611-616,1989

29. Newm an KD, Anderson KD, Van Meurs K, et al: Extracor-

poreal membrane oxygenation and congenital diaphragmatic her-

nia: Should any infant be excluded? J Pediatr Surg 26:1048-1053,

1990

30. VD Staak FHJM, De Haan AFJ, Geven WB , et a l : Improv-

ing survival for patients with high-r isk congenital diaphragmatic

hernia by using extracorporeal membrane oxygenation. J Pediatr

Surg 30:1463-1467, 1995

Can Blood Gas Values Predict Pulmonary Hypoplasia in Antenatally

Documents

Transcript of Can Blood Gas Values Predict Pulmonary Hypoplasia in Antenatally