Balloon tracheal occlusion for congenital diaphragmatic hernia: Experimental studies

Balloon Tracheal Occlusion for Congenital Diaphragmatic Hernia: Experimental Studies By Toshio Chiba, Craig T. Albanese, Diana L. Farmer, Christopher F. Dowd, Roy A. Filly, Geoffrey A. Machin, and Michael R. Harrison San Francisco, California

Purpose: Temporary tracheal occlusion is an effective strategy to enlarge fetal lungs, but the optimal technique to accomplish occlusion is unknown. External clips are effective when applied fetoscopically (Fetendo clip), but require a difficult fetal neck dissection. This study was undertaken to assess the feasibility of intratracheal balloon occlusion, revisiting the internal occlusion strategy. Methods: (1) The internal diameter (ID) of human fetal trachea (53 fetuses; 14 to 41 weeks’ gestation) was compared using a computer-assisted image analyzer and sonography, ex vivo. (2) Volume to diameter relationship of the balloon (balloon configuration curve) was defined using an image analyzing computer. (3) Using the trachea of fetal sheep, pressures that break balloon tracheal seal (seal pressure) were investigated.

A

SUBSET of patients with prenatally diagnosed severe congenital diaphragmatic hernia (CDH) have a predicted high postnatal mortality rate based on liver position, lung-to-head ratio, and early sonographic diagnosis.1 Experimentally and, subsequently, clinically, prenatal temporary tracheal occlusion enlarges the fetal lungs and is a promising therapy for severe cases.1-3 Our initial outcome of internal tracheal occlusion using a water-impermeable, expandable polymeric foam plug was unsatisfactory because of both incomplete occlusion and tracheomalacia.2-4 Thereafter, we reported that in utero external tracheal occlusion with metal clips promotes the growth of hypoplastic lungs and has improved outcome for these infants when accomplished by fetoscopic techniques (Fetendo clip).5 Fetendo clip, however, requires a difficult endoscopic neck dissection with potential injury to the recurrent nerves and the trachea. Experimentally, the feasibility of placing an occluding From The Fetal Treatment Center and the Department of Surgery and the Department of Radiology, University of California San Francisco, San Francisco, CA and The Department of Pathology, University of California at Davis, CA. Presented at the 33rd Annual Meeting of the Pacific Association of Pediatric Surgeons, Las Vegas, Nevada, May 15-19, 2000. Address reprint requests to Michael R. Harrison, MD, The Fetal Treatment Center, University of California San Francisco, 513 Parnassus Ave, HSW-1601, San Francisco, CA 94143-0570. Copyright © 2000 by W.B. Saunders Company 0022-3468/00/3511-0009$03.00/0 doi:10.1053/jpsu.2000.18311 1566

Results: (1) Between 16 and 41 weeks’ gestation, tracheal ID (range, 0.7 to 5.4 mm) correlates significantly with gestational age. (2) Balloon volume required to achieve tracheal seal could be determined based on the tracheal growth curve and the balloon configuration curve. (3) Tracheal seal breaking points varied depending on the tracheal specimen tested. Conclusion: Internal tracheal occlusion using a balloon is feasible with minimal tracheal damage if the balloon volume is adjusted to fetal tracheal growth. J Pediatr Surg 35:1566-1570. Copyright © 2000 by W.B. Saunders Company. INDEX WORDS: Congenital diaphragmatic hernia, Fetendo clip, fetal trachea, detachable silicone balloon.

device using fetal tracheoscopy has been reported with the outcome comparable to those after regular tracheal ligation.6-8 We have decided to revisit the strategy of fetal temporal tracheal occlusion using a detachable silicone balloon (DSB). The DSB to be used was developed originally for cerebral arterial occlusion.9 In this report, we investigated tracheal development in human fetuses using pathologic specimens obtained at autopsy and, based on this result, attempted to formulate an approach to occlude the human fetal trachea with a DSB. MATERIALS AND METHODS

Human Fetal Tracheal Growth Preparation of the human fetal trachea. Tracheal segments (from the cricoid cartilage to the carina) were obtained at autopsy from 53 human fetuses ranged from 14 to 41 weeks’ gestation with no major anomalies. Immediately after clearing the lumen gently with normal saline, they were placed in 10% formalin solution. Before measuring tracheal diameter, the trachea was dissected free of adjacent soft tissues, and the tracheal length (from caudal end of the cricoid cartilage to the carina) was measured. The tracheal diameter was calculated based on the tracheal perimeter obtained with the use of computerized image analysis10 because the cross section of ex vivo tracheal specimens often is deformed and collapsed. Ultrasonographic measurement. Before computer-assisted tracheal morphometry was performed, the internal diameter (ID) of the midtrachea of randomly selected tracheal specimens (from the 12 tracheas of fetuses autopsied between 19 and 39 weeks’ gestation) were obtained. The tracheal specimen was placed underwater. A single observer performed all measurements using a 520-channel phased array real-time scanner (Sequoia, Acuson, Mt View, CA) with a 13-MHz

Journal of Pediatric Surgery, Vol 35, No 11 (November), 2000: pp 1566-1570

EXPERIMENTAL BALLOON TRACHEAL OCCLUSION

1567

seal pressure at each balloon volume over a given range of balloon inflation (Fig 1).

Statistical Analyses Regression analysis was used to compare the tracheal diameters and the gestational ages as determined by maternal gestational dates and also to compare the tracheal diameters obtained by computer image analysis and by sonographic investigations. Balloon diameter at each inflation volume measured by computer morphometry are presented as mean ⫾ SE. Statistical significance is assumed for a P value less than .01.

RESULTS

Fig 1. Method of measurement of tracheal seal pressure.

transducer. Unlike in vivo trachea, the cross section of these formalinfixed specimens often was deformed or collapsed. Sonographically, the mean of maximal and minimal tracheal ID was calculated as the sonographic tracheal ID (ex vivo). Correlation between 2 measurements, sonographic assessment, and computerized image analysis (described below), was evaluated. Computerized image analysis. Computer-assisted image analysis10 was used to quantitate the tracheal dimensions. The internal perimeter of a cross section of the midportion of each specimen was obtained, and its ID was calculated assuming that the cross section is a perfect circle. Computer-assisted image analysis was performed as follows. Color digital images were obtained with a videocamera (COHU Y/C 460 HTYL, 768 ⫻ 494 array; Leica) connected to a microscope (Stereo Zoom 6 photo; Leica, Northvale, NJ) using the Raster Ops Frame Grabber card and Frame Grabber 3.2 software (Raster Ops, Waco, TX). The inner surface of the transected tracheal lumen was outlined on each image, and the inner transverse perimeter was measured using the program NIH (National Institute of Health) Image 1.59.

The ID of the human fetal trachea increases linearly throughout gestation (Fig 2). The statistical correlation was considered significant. It ranges from 1.5 mm at 14 weeks to 4.5 mm at 41 weeks. The association between the sonographic (ex vivo) and computer-analyzed tracheal IDs was moderately strong and statistically significant (Fig 3). The balloon diameter increases, but not linearly, with inflation of the balloon. The slope of the curve is steep when the balloon volume is between 0.05 and 0.15 mL, and it becomes gradual with the volume of 0.15 mL (balloon diameter, 5.04 ⫾ 0.48 mm) or more (Fig 4). Based on the fetal tracheal growth and the balloon configuration curve, a nomogram was created as shown in Fig 5. This nomogram is expected to work as a template for an estimate of the contact volume for any given gestational age. The DSB loses its tracheal seal early by gradual balloon deflation. Tracheal seal capacity (mean seal pressure, 15 mm Hg at the maximum) is variable depending on the specimen tested as shown in Fig 6.

Computer-Assisted Measurements of Tracheal Balloon Size The DSB (maximal diameter, 7.8 mm; maximal inflation volume, 0.5 mL) was inflated and deflated 3 times in vitro with normal saline (under room temperature), and its diameter changes were recorded using the same computerized image analysis as described above.

In Vitro Measurement of Tracheal Seal Pressure Tracheal segments of approximately 5 cm in length were obtained from 3 sheep fetuses of 136 days gestation. Because the cross-section of these segments were oval, the ID (6 mm, 5.8 mm, and 7.5 mm) was expressed as the mean of 2 (maximal and minimal) diameters. One end of the trachea was closed completely by suture ligation leaving a short end of polyethylene catheter in the tracheal lumen. The catheter protruding outside was connected to an upright water column via a stopcock. A deflated DSB (maximal diameter, 9.0 mm; maximal inflation volume, 1.0 mL) with an infusion catheter was placed through the open end of the trachea, then, fully inflated with normal saline and left in place sealing the tracheal lumen. The balloon was deflated gradually until the tracheal seal was broken, measuring each tracheal

Fig 2. Tracheal growth in human fetal development. A scattergram of the tracheal ID plotted on the vertical (Y) axis and the gestational age plotted on the horizontal (X) axis. The correlation coefficient (r2) was determined 0.542 with the regression equation, Y ⴝ 0.11X ⴚ 0.05.

1568

CHIBA ET AL

Fig 3. Correlation between the sonographic and computer-analyzed tracheal IDs. A scattergram of 2 tracheal ID measurements (Y, sonographic ID; X, computer-analyzed ID). The correlation coefficient and the regression equation were 0.684 and Y ⴝ 0.689X ⴙ 0.608, respectively.

DISCUSSION

The Fetendo balloon procedure currently is the least invasive way to achieve fetal tracheal occlusion for fetuses with severe CDH. The goal of this study is to determine a clinically acceptable means to assure a complete seal of the tracheal lumen with the least mechanical or ischemic damage to the developing fetal trachea. We studied fetal autopsy specimens and investigated mechanical properties of the DSB. We learned important lessons in 3 areas. First, there is an expected correlation between the gestational age and the fetal tracheal ID, which can be expressed as a simple mathematical equation. This

Fig 4. Effect of balloon volume on balloon diameter. The Y axis represents the balloon diameter measured using the computer-assisted analysis at each inflation volume shown on the X axis. The diameter is likely to change more with tiny volume differences when the volume is less than 0.15 mL.

Fig 5. Nomogram for estimating contact volume of the balloon. This nomogram is composed of 2 scales; the tracheal growth scale (left) represents the gestational age and the tracheal ID, and the balloon configuration scale (right) represents the balloon diameter and the balloon inflation volume. The tracheal ID on the left scale is predicted from the gestational age using the equation shown in Fig 2. Balloon volume corresponding to each balloon diameter is estimated using the diagram in Fig 4. Based on the nomogram showing 4 variants (gestational age, predicted tracheal ID, balloon diameter, and balloon volume), 2 most important parameters (gestational age, contact volume of the balloon) are directly linked.

means that we can roughly estimate fetal tracheal ID based on gestational age. Second, assuming that the computer image analysis is the gold standard, sonographic evaluation of the tracheal diameter is reliable. Third, combining fetal tracheal growth curve and in vitro balloon diameter changes over a given range of volume enables us to estimate a minimally required balloon volume to make the balloon wall contact the tracheal inside (contact volume). In Fig 1, in vivo tracheal size is expected to be somewhat larger compared with that of formalin fixed specimens because, in our pilot study, fetal sheep trachea showed an approximate 10% reduction in ID when placed in formalin. Sonographic determination of tracheal ID has been shown by our study to be reliable because sonographic tracheal dimensions correlated well with the gold standard, computer-assisted measurements. In vivo sonographic results will likely exceed ex vivo results because the tracheal collapse or deformation after formalin fixation does not happen in vivo (Fig 3). According to Cooper et al,11 human fetal trachea can be well visualized sonographically over a wide range of

EXPERIMENTAL BALLOON TRACHEAL OCCLUSION

1569

Fig 6. Effect of balloon volume on tracheal seal pressure ex vivo. In the figure, only the first of a total of 3 consecutive test results is shown because the tracheal wall distension by repeat balloon inflation leads to lesser seal pressure with variable test-retest reliability. The seal pressure profile differs when the used tracheal segment differs. The seal pressure is determined by variable factors including tracheal ID and compliance. This experiment suggests that the use of bulky and somewhat floppy balloon is preferable.

prenatal development. Thus, in vivo sonographic study of the fetal trachea is expected to be helpful for preoperative assessment before balloon tracheal occlusion. In the clinical setting, we must add a given volume of inflation to the predicted contact volume to achieve a full luminal occlusion (sealing volume or seal keeping volume). In this regard, osmotic pressure of the balloon fluid does not seem crucial because, in our pilot study, the balloon volume did not significantly change when half inflated with hyperosmotic (450 mOsm/L) contrast medium and placed in normal saline for 48 hours at 37°C. The question is how we can predict the volume to be added to the contact volume to achieve a tracheal seal. This estimation might be done retrospectively based on our clinical experience including changes in balloon diameter and length during long-term balloon placement.12 Our experience to date suggests that approximately 3 times the contact volume is likely to keep the tracheal seal without obvious tracheal damage. To predetermine the seal volume theoretically, we must consider 3 important factors in addition to tracheal dimension and balloon configuration: (1) tracheal wall compliance that may change depending on the stage of fetal development; (2) fluid pressure in the distal trachea, which is caused by tracheal mechanical obstruction; and (3) tracheal wall pressure and blood flow that may be compromised by the tracheal balloon. The fetal trachea is supposed to lose its compliance as fetal development progresses, but there have been no comprehensive data reported in the literature. According to Nardo et al,13 in sheep fetuses of 117 days gestation, the intratracheal pressure distal to the occlusion increased within 1 day of

obstruction reaching 4 to 6 mm Hg and there was no further significant increase for 10 days. Likewise, in other reports using sheep fetuses of 118 to 122 days of gestation,14,15 tracheal pressure rose soon after obstruction to 3 to 5 mm Hg with no significant increases during the experimental period of 2 to 4 weeks. Based on these reports, in our experimental model shown in Fig 1, balloon inflation with a volume of 70% to 100% of maximum inflation is required to seal the tracheal lumen (Fig 6). Tracheal wall damage predisposing to tracheomalacia must be avoided. In this regard, it is obvious that the lateral wall pressure exerted by the balloon should be kept below tracheal wall capillary pressure (25 to 35 mm Hg) to avoid tracheal wall ischemia. However, direct measurement of tracheal wall pressure is clinically impractical because insertion of a mechanical sensor between the balloon and the intact trachea is difficult, and even very slender sensors may give misleading values. An intracuff pressure is much easier to monitor, and, in human adults, it was recommended that the intracuff pressure of tracheostomy tube should be less than 25 mm Hg.16 In experiments using adult mongrel dogs,17,18 which have a mean arterial pressure of around 85 (systolic/diastolic; 110 to 120/70 to 80) mm Hg, safe tracheal wall pressure was reported to be less than 20 to 30 mm Hg for prolonged intubation because the average arteriolar pressure at the end of the capillary bed was estimated at 33 mm Hg. Because fetal aortic blood pressure is about 65 to 70/30 to 45 (systolic/diastolic) mm Hg,19 tracheal pressure exerted by the balloon at the tracheal wall should be lower than 20 mm Hg.

1570

CHIBA ET AL

Our experiments and the literature suggest that fetal tracheal wall pressure should be kept over 7 to 8 mm Hg to achieve adequate tracheal seal, and this pressure should be kept below 15 to 20 mm Hg to prevent tracheal ischemic damage. The internal tracheal occlusion is feasible using a DSB. The DSB can be used safely when the balloon inflation volume is properly adjusted according to the size of the developing trachea. The Fetendo balloon

procedure may replace the Fetendo clip procedure for infants with severe CDH. ACKNOWLEDGMENTS The authors thank Richard E. Sievers and Amanda E.M. Browne, The Department of Cardiology at UCSF for their excellent technical assistance. They also thank Dr Hirotaka Watada, The Hormone Research Institute at UCSF, for his suggestions regarding statistical data analyses.

REFERENCES 1. Adzick NS, Harrison MR: The fetus as patient, in Oldham KT, Colombani PM, Foglia RP (eds): Surgery of Infants and Children. Philadelphia, PA, Lippincott-Raven, 1997, pp 48-64 2. Harrison MR, Adzick NS, Flake AW, et al: Correction of congenital diaphragmatic hernia in utero VIII: Response of the hypoplastic lung to tracheal occlusion. J Pediatr Surg 31:1339-1348, 1996 3. Wilson JM, DiFiore JW, Peters CA: Experimental fetal tracheal ligation prevents the pulmonary hypoplasia associated with fetal nephrectomy: Possible application for congenital diaphragmatic hernia. J Pediatr Surg 28:1433-1440, 1993 4. Bealer JF, Skarsgard ED, Hedrick MH, et al: The ‘PLUG’ odyssey: Adventures in experimental fetal tracheal occlusion. J Pediatr Surg 30:361-365, 1995 5. Harrison MR, Mychaliska GB, Albanese CT, et al: Correction of congenital diaphragmatic hernia in utero IX: Fetuses with poor prognosis (liver herniation and low lung-to-head ratio) can be saved by fetoscopic temporary tracheal occlusion. J Pediatr Surg 33:1017-1023, 1998 6. Papadakis K, Luks FI, Deprest JA, et al: Single-port tracheoscopic surgery in the fetal lamb. J Pediatr Surg 33:918-920, 1998 7. Luks FI, Gilchrist BF, Jackson BT, et al: Endoscopic tracheal obstruction with an expanding device in a fetal lamb model: Preliminary considerations. Fetal Diagn Ther 11:67-71, 1996 8. Flageole H, Evrard VA, Vandenberghe K, et al: Tracheoscopic endotracheal occlusion in the ovine model: Technique and pulmonary effects. J Pediatr Surg 32:1328-1331, 1997 9. Higashida RT, Halbach VV, Dowd CF, et al: Intracranial aneurysms; interventional neurovascular treatment with detachable balloons-results in 215 cases. Radiology 178:663-670, 1991 10. Zhu BQ, Sun YP, Sudhir K, et al: Effects of second-hand smoke

and gender on infarct size of young rats exposed in utero and in the neonatal to adolescent period. J Am Coll Cardiol 30:1878-1885, 1997 11. Cooper C, Mahony BS, Bowie JD, et al: Ultrasound evaluation of the normal fetal upper airway and esophagus. J Ultrasound Med 4:343-346, 1985 12. Harrison MR, Albanese CT, Hawgood SB, et al: Correction of congenital diaphragmatic hernia in utero XI: Fetoscopic temporary tracheal occlusion by detachable balloon (FETENDO Balloon). (submitted) 13. Nardo L, Hooper SB, Harding R: Stimulation of lung growth by tracheal obstruction in fetal sheep: Relation to luminal pressure and lung liquid volume. Pediatr Res 43:184-190, 1998 14. Hashim E, Laberge JM, Chen MF, et al: Reversible tracheal obstruction in the fetal sheep: Effects on tracheal fluid pressure and lung growth. J Pediatr Surg 30:1172-1177, 1995 15. Papadakis K, Luks FI, De Paepe ME, et al: Fetal lung growth after tracheal ligation is not solely a pressure phenomenon. J Pediatr Surg 32:347-351, 1997 16. Ching NP, Ayres SM, Spina RC, et al: Endotracheal damage during continuous ventilatory support. Ann Surg 179:123-127, 1974 17. Joh S, Matsuura H, Kotani Y, et al: Changes in tracheal blood flow during endotracheal intubation. Acta Anaesthesiol Scand 31:300304, 1987 18. Homi J, Notcutt W, Jones JJ, et al: A method for comparing endotracheal cuffs. A controlled study of tracheal trauma in dogs. Br J Anaesth 50:435-444, 1978 19. Rudolph AM: Fetal circulation and cardiovascular adjustments after birth, in Rudolph AM, Hoffman JIE, Rudolph CD (eds): Rudolph’s Pediatrics. Stamford, CT, Appleton and Lange, 1996, pp 1409-1413