Fetal Lung Growth After Tracheal Ligation Is Not Solely

a Pressure Phenomenon
By Konstantinos Papadakis, Franqois I. Luks, Monique E. De Paepe, George J. Piasecki,
and Conrad W. Wesselhoeft, Jr
Providence, Rhode Island

l Fetal tracheal ligation increases lung growth in utero, utero tracheal obstruction can counteract the pulmonary
making it potentially applicable for antenatal treatment of hypoplasia associatedwith CDH,5.6making it a potential
diaphragmatic hernia. This phenomenon has been ascribed
modality to treat CDH antenatally.5,6This approach may
to increased intratracheal pressure, which activates as yet
unidentified pulmonary stretch receptors. The purpose of lend itself particularly well to minimally invasive fetal
this study was to determine whether the composition of surgery,7-9or even to fetal tracheoscopy.‘O.”
lung fluid has any effect on fetal lung development after Lung growth after tracheal obstruction, once believed
tracheal obstruction. Six sets of fetal lamb twins underwent to result in overdistended, hypertrophic lungs,‘?.i3seems
tracheal ligation with placement of intratracheal catheters at to representtrue lung hyperplasia5 Very little is known
122 days’ gestation (term, 145 days). In group 1 (n = 6).
about the exact mechanismsof lung growth after tracheal
tracheal fluid was aspirated daily, measured, and replaced
with equal volumes of saline. Their respective twins (group obstruction. A pressurephenomenonhasbeen postulated,
2, n = 6) had daily reinfusion of their own tracheal aspirates. through as yet undefined pulmonary stretch receptors.6,14
lntratracheal pressure was recorded daily in both groups. There is at least some in vitro evidence that humoral
Unobstructed fetal lambs (n = 7) were used as negative factors presentin lung fluid play a role in this exaggerated
controls. Animals were killed on postoperative day 14 (136 lung growth.‘” To test the importance of tracheal fluid
days). Lungs were weighed, perfusion fixed at 25 cm H20,
composition, we created an in vivo model of tracheal
and processed for standard morphometric analysis. Intratra-
cheal pressure remained between 3 and 5 torr in both obstruction whereby fetal lung fluid was replaced by
experimental groups throughout the entire postoperative saline.
period. In all 12 experimental fetuses, tracheal ligation
resulted in an almost threefold increase in lung fluid volume MATERIALS AND METHODS
by day 1; a slight decrease at a mean of 2.4 days; and a Ttme-dated pregnant ewes t 122 days, term, 135 days) wtth twm
second surge from day 4 on. Lung fluid volume was signifi- gestations underwent mtdlme laparotomy under general halothane
cantly higher in group 2 than in group 1 at all measured time anesthesta (0.52.0% in 100% 01) after intravenous mductton wtth
points (P < .05, Wilcoxon rank sum test) except on days 3,4, ketamine ( 1 g) A small hysterotomy was made transversely near the
and 8 (P= .06). Lung weight per body weight (LW/BW) at uterme horn with exteriortzatton of the fetal head and neck. The uterine
delivery was 0.045 2 0.008 in group 1, not significantly wall was secured around the fetal neck wtth Alhs clamps to mmtmize
different from unobstructed controls (0.038 ? 0.006). LW/BW amniotic flutd loss. The fetal trachea was then exposed and atraumatt-
in group 2 was 0.055 2 0.010, significantly larger than either tally clamped to avotd egress of tracheal fltnd. A tracheotomy was
group 1 or control (P < .05, single factor analysis of vari- made. and a 19.gauge inner dtameter CID) flanged tygon catheter was
ance). Air space fraction was comparable between the three Inserted mto the dtstal trachea and secured wtth 4-O polypropylene
groups. Alveolar numerical density was significantly lower (Prolene. Ethicon, Johnson & Johnson, Somerville, NJ). Trachea1
in groups 1 and 2 than in unobstructed controls (P < 0.05). ltgatton was performed wtth #2 sdh cephalad to the tracheotomy. A
Replacement of tracheal fluid with saline inhibits the lung 17.gauge ID ammottc catheter. fitted wtth a cage. was attached to the
hypertrophy seen after tracheal ligation. This phenomenon neck at the level of the trachea1 catheter. The fetal head and neck were
therefore appears more dependent on tracheal fluid growth returned to the uterus. Antibiotics (ampicillm. 700 mg and chloramphem-
factors than on increased intratracheal pressure after obstruc- col. 250 mg) were added to the amniottc fluid. The hysterotomy was
tion. The immediate decrease in net lung fluid production closed using a two-layer continuous closure and the catheters were
after saline exchange suggests that these humoral factors “Wttzeled.” The abdominal wall was closed after ensurmg adequate
play an important role in the initiation of lung cell prolifera- slack on the catheters. The catheters were extertortzed mto a pouch on
tion. the rtght flank of the ewe.
Copyright o 1997 by W.B. Saunders Company

INDEX WORDS: Fetal surgery, fetal lung growth, congenital From thr D~~wott of Prdrarrrc~ Surgey and the Department of
diaphragmatic hernia, tracheal obstruction. Pathology, Bro+r II Unrver.~~ School of Medicme. Husbro Children k
und Rhode Island Hosprtals. Provrdmce. RI
Presented ut the 27th Ann~ui Mretmg of the Amertcun Pediatrrc

I NFANTS BORN WITH congenital diaphragmatic

hernia (CDH) often die of causes secondary to
pulmonary hypoplasia. This hypoplasia is, in part, a
Sur,qtal
Supported
Foundatro,~.
Assocrcrtron.
m put
Sun Diego,
b? Grunt
Cul$wniu.
No. 5177
Mu 20-23.
from
1996.
the Rhode Island

Address reprrnt reyur3t.t to Frutywr 1. Luhs. MD, DllYsion of

consequenceof compression by the herniated viscera
Pedintrrc Surgery, Hubro Children i Hospital, 2 Dudley St, Surte 180.
during gestation, and the severity of the hypoplasia Provrdence. RI 02905.
appearsto be a function of the degree and duration of Cop>riyht 8~ 1997 by WE Surtnders C’ompun~
compression.‘-“ Recent experience has shown that in 0022-3468/97/3202-0040$03 00/O

Journal ofPedIatric Surgery, Vol 32, No 2 (February), 1997: pp 347-351 347

348 PAPADAKIS ET AL

Twelve fetal twins were divided mto two groups. One fetus of each statistically significant difference in pressurebetweenthe
pregnant ewe (group 1, n = 6) underwent intraoperative, followed by
two groups. When lung fluid production was corrected
daily aspiration of lung thud, which was exchanged milliliter for
millihter with room temperature salme. Their respecttve twins (group 2, for body weight at birth, tracheal ligation resulted in an
n = 6) underwent dally aspiration and reinfusion of their own lung almost threefold increasein lung fluid volume by day 1
fluid. The amount of the daily fluid asptrates was recorded. Both groups (from 11.7 ? 4.8 to 28.4 ? 9.3 mL/kg in group 1 and
had daily measurements of mtratracheal and ammotic pressures (Grass from 12.1 ? 4.1 to 36.8 + 8.5 mL/kg in group 2); a slight
PI-l pressure transducer and Dash IV, 4-channel preamplifier-recorder,
Astro-Med. West Warwick, RI) True intratracheal pressure was defined decreaseat a meanof 2.4 days (nadir 24.6 ? 4.8 in group
as amniotic minus tracheal pressure.th After the daily manipulatton, 250 1,3 1.4 _f 7.3 mL/kg in group 2); and a secondsurgefrom
mg of ampictllin and 250 mg of chloramphemcol were Infused into the days 4 to 10. Lung fluid volume was significantly higher
amniotic catheter. Seven nonoperated fetal lambs were used as controls. in group 2 than in group 1 at seven of 10 measuredtime
At 136 days’ gestabon (postoperattve day 14) the ewe underwent
Cesarean section under the same anesthetic conditions. The previous points (P < .05, Wilcoxon rank sumtest; Fig 1). Catheter
hysterotomy was used to deliver the fetuses. Fetal breathmg was malfunction in several animals toward the end of the
prevented by placing a surgical glove over the fetus’ head. The fetuses experimental period precludesmeaningful data interpreta-
were killed by injection of a euthanasia solution (3 mL of Beuthana-
tion beyond day 10 postligation.
sia-D Special, Shermg-Plough Ammal Health Corp., Kenilworth, NJ)
into an umbilical vein after clamping of the cord. The fetus’ body
weights were recorded, and the fetal lungs and trachea were removed en
bloc. Lung Weight
There was no statistical difference in lung weight to
Histology and Morphometric Analysis body weight ratio (LW/BW) between group 1 (ligated/
The wet weights of the lungs were determmed. Fixatton was saline) and nonligated controls (P > .13, single factor
accomplished by tracheal instillation of 10% buffered formaldehyde ANOVA). LW/BW was significantly higher in group 2
mamtained at a pressure of 25 cm Hz0 with the lungs immersed in a
beaker containing buffered fixative The lungs of respective twins were (ligated/unaltered lung fluid) than in either group 1
fixed in parallel to eliminate fixation variabihty. After at least 7 days’ (P < .05) or control (P < .005). Resultsare summarized
fixation at 2 1 “C the trachea was clamped and the volume of the lungs in Table 1.
was estimated by the volume displacement method. The lungs were cut
m standardized sections (3 to 5 mm thick) and random blocks from each
lobe were embedded in parafbn. Hematoxylin and eosin (H&E)-stained Histology and Morphometry
slides (4 m thick) were prepared from each paraffin block (four per
animal). Histologically, the lungs of age-matchedcontrol ani-
Lung tissue sections were analyzed morphometrtcally using a mals showed well-developed alveolar sacs and ducts,
computerized image analysis system: microscope interfaced vta a
charge-coupled device (CCD) video camera (KP-161, Hitachi USA,
lined by thin alveolar septa(Fig 2A). In both experimen-
Woodbury, NY) to a Power Macintosh 7 100/8OAV (Apple Computer, tal groups, the air spaces appeared dilated and the
Cupertino, CA) eqmpped with software for image analysis (Image 1.59 alveolar septa attenuated (Figs 2B and C). Air space
for Macintosh. National Institutes of Health. Bethesda, MD). The fraction did not differ significantly between the three
system was programmed to measure lung atrspace fraction. The number
of alveoli per unit area was estimated by counting alveolar profiles
groups (Table 1). Alveolar numerical density (number of
within the test area, whereby an alveolus was defined as an airspace alveoli per cubic centimeter of lung tissue) was signifi-
either entirely or partially enclosed by respiratory epithelmm. Forty cantly lower in both ligation groups than in controls
fields were counted for each lung at a magnification of 200X ( 10 fields (P < .05, single factor ANOVA).
for each of the right apical, right diaphragmatic, left apical. and left
diaphragmatic lobes). Data were corrected for fetal body weight where
appropriate to allow for comparison between groups.
Statistical analysts was performed using the Wilcoxon rank sum test
(paired compartson of groups 1 i* 2). or single-factor analysis of
variance (ANOVA) for compartson of three nonparametric groups.
Where appropriate, values are expressed as mean + standard deviation.
A P value of less than .05 was consrdered statistically significant. All
procedures and protocols were approved by the Brown Umverstty
Animal Care and Use Committee.

RESULTS
Intratracheal Pressureand Lung Fluid Volume
Intratracheal pressure (when corrected for amniotic
pressure)increasedfrom 1.9 -C 1.Otorr before ligation to
3.7 -+ 0.8 torr by the secondpostoperative day. Thereaf-
Fig 1. Average net lung fluid volume after tracheal ligation. Group
ter, intratracheal pressuresremained stable, ranging from 1, ligated/saline; group 2,ligated/unaltered fluid. l P c .05, Wilcoxon
3 to 5 torr throughout the study period. There was no rank sum test.
ROLE OF LUNG FLUID IN FETAL TRACHEAL LIGATION 349

Table 1. Lung Weight and Morphometric Data

Control Group 1 Group 2 PValue

Lung weight/body weight 0.038 5 0.006 0.045 t 0.008 0.055 + 0.010* <.05
Air space fraction (%) 74 8 + 6.6 74.4 t 5.1 75.4 + 2.5 Not significant
Alveolar numerical density (r106/cm3) 82.45 5 26.14t 57.01 2 11.46 44.74 -c 11.22 c.05

NOTE. Group 1, ligated/saline; group 2, ligated/unaltered fluid; control, unobstructed trachea

*Group 2 vgroup 1 and control.
tGroup 1 and group 2 vcontrol, single-factor ANOVA.

DISCUSSION
The exact mechanism by which tracheal ligation
promotes lung growth in the developing fetus is not
known. The present study evaluates the importance of
tracheal fluid composition in this process.Here. tracheal
ligation was performed in fetal twins under identical
pressureconditions. When tracheal fluid was preserved,
tracheal obstruction resulted in significantly increased
lung weight, as expected from previous reports.5%6.“.‘6
Replacement of tracheal fluid with saline resulted in
significant inhibition of this lung growth phenomenon.
Indeed, lung weight after ligation and salinereplacement
was not significantly different from that of unobstructed
controls. This indicates that tracheal fluid composition,
rather than intratracheal pressure,is critical in promoting
pulmonary growth. Air spacefraction was similar in all
three groups, confirming that the rise in lung weight and
lung volume after tracheal ligation is caused by tissue
growth, not just overdistension by an elevated intratra-
cheal pressure.6
DiFiore and Wilson’s have previously shown that lung
fluid after tracheal ligation is mitogenic to pneumocytes
in vitro. Although the actual cell proliferation and in-
crease in lung fluid production is indeed likely to be
mediated by one or more growth factors,‘7.‘ntheir study
did not establishwhether humoral factors are primary or
secondary effecters. Elevated intratracheal pressureand
pulmonary stretch could still have initiated this increase
in growth factor concentration. In the presentexperiment,
blunting of the lung fluid volume curve in group 1
(ligation/saline) was noted as early as day 1. despite
intracheal pressuressimilar to those in group 2 (ligation/
unaltered tracheal fluid). Thus. fluid composition seems
to play a role not only in subsequentlung growth, but also
in the initial signal to cell proliferation after tracheal
ligation.
The increasedintratracheal pressuremay still have an
effect on the ultimate lung architecture. When compared
with control lungs. the alveolar septain both experimen-
tal groups were shortened, causing coalescenceof the
Fig 2. Photomicrographs of random fields of respiratory tissue
alveolar spaces.These alveoli are poorly defined and from (A) control animals, (El group 1 (2-week tracheal ligation, daily
appearlarger. This would explain the observed reduction exchange of tracheal fluid with saline), and IC) group 2 (2-week
in alveolar numerical density (number of alveoli per mm3 tracheal ligation, no exchange of tracheal fluid). (A) In control lungs,
the alveoli are well demarcated by prominent septations. (B,Cl In
of lung tissue)after tracheal ligation, a finding that seems ligated lungs, the septa appear shortened and the alveoli less well
at odds with the concept of functional lung growth. The defined. (H&E, original magnification ~400).
350 PAPADAKIS ET AL

very definition of alveoli has been subject to contro-
versy. I9 and has rendered reproducibility of morphomet-
ric results difficult.6.“.20
The lung fluid volume curve after tracheal ligation may
offer insight into the time course of lung growth after
tracheal obstruction. Within 24 hours after tracheal
ligation, net lung fluid production increases threefold,
from 10 to 15 mL/kg to 30 to 45 mL/kg. If fluid
production were to continue at the predicted preligation
rate of 3.4 to 4.5 mL/kg/h21-‘3 however, tracheal fluid
volume would have risen from 10 to 15 mL/kg to 90 to
120 mL/kg within the first day-a six- to tenfold
increase. Thus, net lung fluid production actually de-
creases after tracheal ligation; an increase is not seen until
3 days later. This initial reduction in net fluid production
could be the result of reduced secretion or accelerated
reabsorption of lung fluid; both could partially be caused
by a catecholamine or cortisol surge after surgical stress,
similar to what is observed intrapartum.‘4,25 This effect is
unlikely to continue beyond the immediate postoperative
period, however. Rather, the sustained fall in net fluid
production during the first 72 hours after tracheal ligation
appears to be the result of the obstruction itself. The noted
elevation in intratracheal pressure may partially be respon-
sible for this, because high intratracheal pressure is
known to impair fluid secretion and promote reabsorp-
tion.16
Why, then, does net fluid production rise after 3 days, if
hydrostatic pressures within the tracheobronchial tree
remain elevated throughout the entire 2-week period?
This could be a reflection of increased secretory cell mass
within the lung, suggesting that cell proliferation be-
comes apparent 2 to 3 days after tracheal obstruction.
Indeed. Hooper et a127 reported that two-thirds of the
increase in pulmonary DNA content occurred between 2
and 7 days postligation. Others have also commented on
the short time required for lung growth to occur after
tracheal obstructionh~“,**
The principle of lung growth after tracheal obstruction
is now well recognized, and preliminary reports in
humans have suggested its potential usefulness in the
antenatal treatment of CDH.zx.29 It is important that the
mechanisms involved in this process be better under-
stood, however, before clinical application becomes
widespread. Knowledge of the physiological and molecu-
lar basis of normal and accelerated pulmonary growth
may enable us, in the future, to upregulate, or turn off,
pulmonary growth more selectively after tracheal obstruc-
tion. It may become possible to avoid tracheal obstruction
altogether. by acting directly on the regulatory mecha-
nisms themselves.
ACKNOWLEDGMENT
The authors express their gratttude to Mohammed Traore for hts
technical assistance.

REFERENCES
1. Areechon W, Retd L. Hypoplasia of lung with congenital diaphrag- obstruction with an expanding device m a fetal lamb model: Preliminary
matic hernia. Br Med J 1230-233, 1963 considerattons. Fetal Diagn Ther 11:67-7 1. 1996
2. Nguyen L. Guttman F, De Chadarevian JP. et al: The mortality of 12. Adztck NS, Harrison MR. Glick PL, et al: Experimental
congenital diaphragmatic hernia. Is total pulmonary mass inadequate, pulmonary hypoplasia and oligohydramnios: Relative contrtbutions of
no matter what? Ann Surg 198:766-770, 1983 lung fluid and fetal breathing movements. J Pediatr Surg l9:658-665,
3. Harrison MR. Jester JA, Ross NA: Correctton of congenital 1984
dtaphragmatic hemta in utero. I. The model: Intrathoracic balloon 13. Cannel JA, Friedman F. Addams FH: Fetal trachea1 ligation and
produces fatal pulmonary hypoplasia. Surgery 88:174-182, 1980 lung development. Am J Dis Child lO9:452-456. 1965
4. Harrison MR. Bressack MA. Churg AM, et al, Correctton of 14. LIU M, Skinner SJ. Xu J, et al: Stimulation of fetal rat lung cell
congenital dtaphragmatic hemta in utero. II. Simulated correctton proliferation m vitro by mechanical stretch. Am J Physiol 263:376-383.
permits fetal lung growth with survival at birth. Surgery 88:260-268, 1992
1980
15. DiFtore JW. Wdson JM: Lung liquid from fetal lambs with
5. Hedrick MH, Estes JM, Sullivan KM, et al, Plug the lung until it
tracheal occlusion ts mitogenic to pulmonary cells in vitro. Surg Forum
grows (PLUG): A new method to treat congenital dtaphragmatic hernia
45:666-668. 1994
in utero. J Pediatr Surg 29:612-617, 1994
16. Hashim E, Labege JM. Chen MF. et al: Reversible tracheal
6. DtFiore JW, Fauza DO, Slavin R, et al: Experimental fetal tracheal
obstruction m the fetal sheep: Effects on tracheal fluid pressure and lung
ligatton reverses the structural and physiologtcal effects of pulmonary
growth. J Pedtatr Surg 30: 1172- 1177, 1995
hypoplasta in congenital dtaphragmatic hernia. J Pediatr Surg 29:248-
17. Sundell HW. Gray ME. Serenius FS, et al, Effects of eptdermal
255,1994
7. Luks FL Deprest JA: Endoscopic fetal surgery: A new alternative? growth factor on lung maturation in fetal lambs. Am J Path01
Eur J Obstet Gynecol Reprod Btol52: l-3. 1993 100:707-736. 1980
8. Luks FI, Deprest JA. Vandenberghe K. et al: A model for fetal 18. Hooper SB. Han VKM, Hardmg R: Changes in lung expanston
surgery through intrauterine endoscopy. J Pedtatr Surg 29 lOO7- 1009, alter pulmonary DNA synthesis and IGF-II gene expression in fetal
1994 sheep. Am J Physiol265,L403-L-409. 1993
9. Vanderwall KJ. Bruch SW, Meuli M. et al: Fetal endoscoptc 19. Thurlbeck WM: Lung growth and alveolar multiplication. Patho-
(‘fetendo’) tracheal chp. J Pedtatr Surg 31: 1101-l 104, 1996 biol Annual 5:3-34, 1975
10. Luks FL Deprest JA, Vandenberghe K, et al: Fetoscopy-guided 20. Docimo SG, Crone RK, Davies P, et al: Pulmonary development
fetal endoscopy m a sheep model. J Am Co11 Surg 178:609-612, 1994 in the fetal lamb: Morphometric study of the alveolar phase. Anat Ret
11 Luks FI, Gilchrist BE Jackson BT, et al: Endoscopic trachea1 229:495-498, 199 1
ROLE OF LUNG FLUID IN FETAL TRACHEAL LIGATION 351

21. Lanman JT. Schaffer A, Llewellyn H, et al: Distensibility of the fluid volume in growth and maturation of the fetal sheep lung. J Clin
fetal lung with fluid m sheep. Pediatr Res 5:586-590. 1971 Invest 86: 1270-1277, 1990
22. Harding R. Sigger JN, Wickham PJ, et al: The regulation of flow 27. Hooper SB. Han VKM, Harding R: Changes m lung expansion
of pulmonary fluid in fetal sheep. Resplr Physlol57:47-49. 1984 alter pulmonary DNA synthesis and IGF-II gene expression in fetal
23. Beierle EA. Langham MR. Cassin S. In utero lung growth of sheep. Am J Physlol265:L403-409, 1993
fetal sheep with dlaphragmatlc hernia and tracheal stenosis J Pediatr 28. Bealer JE Vanderwall K, Adzick NS, et al: A new treatment
Surg 31.141-147, 1996 optlon for patients with congenital chaphragmatic hernia. Presented at
24. Silver M: Prenatal maturation. the timing of birth and how it may the 14th annual meeting of the Internatlonal Fetal Medicine and Surgery
be regulated in domestic ammals. Exp Physlol75:285-307. 1990 Society. Newport, RI, May 3-6, 1996
25. Liggms GC: The role of corhsol in preparing the fetus for buth. 29. Flake AW. Johnson MP, Treadwell M, et al. In utero treatment of
Reprod Fertd Develop 6: 141-150, 1994 nght-sided congenital dlaphragmatlc hernia (R-CDH) by prenatal
26. Moessmger AC. Harding R. Adamson TM, et al. Role of lung tracheal occlusion. Am J Obstet Gynecol 174 489. 1996 (abstract)