| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Report |
1
Department of Radiology, University of California San Francisco, 505 Parnassus
Ave., (M-372), San Francisco, CA 94143.
2
Present address: Department of Diagnostic Imaging, Children's Hospital
Oakland, 747 Fifty Second St., Oakland, CA 94609.
3
Present address: Department of Radiology, Mahidol University, 2 Prannok Rd.,
Bangkok, Thailand 10700.
4
Department of Surgery, University of California San Francisco, San Francisco,
CA 94143.
5
Department of Pediatrics, University of California San Francisco, San
Francisco, CA 94143.
Received August 28, 2000;
accepted after revision July 23, 2001.
Address correspondence to B. F. Kammen.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. The diaphanous diaphragm is a consequence of air trapped in the porous polytetrafluoroethylene graft that creates an intragraft radiolucency apparent on postoperative chest radiographs obtained within the first 24 hr. This radiolucency is transient and gradually disappears over the first few postoperative days as the air is replaced by granulation tissue. This sign should be recognized and not mistaken for a persistent pneumothorax after repair of a congenital diaphragmatic hernia.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Initial chest radiographs of the 19 neonates obtained in the 1st days of life showed 14 (74%) of the congenital diaphragmatic hernias to be leftsided and five (26%) to be right-sided. All congenital diaphragmatic hernias were successfully repaired, and postoperatively, a chest radiograph was obtained 12-24 hr after the procedure and then periodically until hospital discharge. On average, radiographs were obtained at least every other day for 43 days (range, 20-119 days; median ± SD, 39 ± 27 days).
All chest radiographs (796) obtained in the 19 neonates were retrospectively reviewed by three experienced pediatric radiologists who were unaware of the type of surgery that each patient had undergone. Review was used to determine the postsurgical radiographic appearance of the repaired hemidiaphragm. The decision as to whether the diaphanous diaphragm sign was present was reached by consensus. If a radiolucency was identified along the course of the repaired diaphragm, then its location, morphology (curvilinear or geometric), location along the diaphragm (medial, middle, lateral), and time from appearance to disappearance were recorded. The evolution of the ipsilateral postsurgical pneumothorax was recorded to compare its disappearance with the time course to resolution of the diaphanous diaphragm.
Surgical records for all 19 neonates were subsequently reviewed to determine if the type of repair was related to the postsurgical radiographic appearance. Sixteen of 19 neonates had undergone repair with a Gore-Tex graft, and the remaining three with smaller defects had primary closure of their diaphragms.
We undertook in vitro analysis to determine the etiology and location of the diaphanous diaphragm by creating two phantoms with the polytetrafluoroethylene graft. The Gore-Tex graft, made from polytetrafluoroethylene biomaterial, is a 1-mm sheet material with a dual surface that incorporates a microporous node and fibril structure with regularly spaced macropores. One surface of the polytetrafluoroethylene graft is smooth, and the other is coarse, with the coarse side directed toward the abdomen in congenital diaphragmatic hernia repair.
The first phantom design consisted of the polytetrafluoroethylene graft placed between a fluid-filled balloon simulating abdominal contents and a second balloon inflated with air, simulating a pneumothorax (1A). The coarse side of the polytetrafluoroethylene graft faced the water-filled balloon. The entire phantom was placed in a box, which caused the polytetrafluoroethylene graft to be compressed by both balloons. The flaw with this design was that the presence of intervening air between the graft and balloon could not be excluded. To ensure that the seal between the graft and materials placed on either side would be airtight, we created a second phantom design. This second design consisted of a rectangular strip of polytetrafluoroethylene graft (110 x 20 x 1 mm) placed vertically along the diameter of a shallow cylindric container with the thinnest side of the graft flush with the bottom of the container (Fig. 1B). A gelatin solution was then poured to fill the container along both sides of the polytetrafluoroethylene graft and allowed to congeal (Fig. 1B).
|
|
Radiographs of both phantoms were obtained using a radiographic technique similar to that used for portable neonatal chest films (Figs. 1C and 1D). Specifically, a Cronex (Sterling Medical, Cherry Hill, NJ) 400-speed film was placed directly under the phantom before exposure, with the X-ray beam tightly collimated (focal-film distance, 1 m; 50 kVp; 300 mA; 0.002 sec). The direction of the X-ray beam was tangential to the graft (Figs. 1C and 1D).
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
The diaphanous diaphragm was always a 1-mm-thick crescentic radiolucent tract along the repaired hemidiaphragm that gradually disappeared on average over 4 days (range, 2-7 days) (Fig. 2C). In 13 of 16 patients, the morphology of the diaphanous diaphragm was of a curvilinear lucency following the expected contour of the hemidiaphragm (Fig. 2A,2B,2C,2D). In three patients, the appearance of the radiolucency was more geometric (Fig. 3).
|
|
|
A pneumothorax or hydropneumothorax surrounding the hypoplastic lung was identified on all postsurgical chest radiographs. The diaphanous diaphragm was less obvious when adjacent to a pneumothorax than when there was an adjacent hydropneumothorax caused by decreased conspicuity between the two adjacent air-filled spaces. However, a fine radiodense line, the superior surface of the graft that is located inferior to the pleural space but superior to the air trapped in the graft, made identification of the diaphanous diaphragm always possible.
In the cohort of patients in whom the artifact was identified, the pneumothorax resolved before (n = 10), simultaneously with (n = 3), or after (n =3) the diaphanous diaphragm sign disappeared. In 10 (63%) of the 16 patients in whom the diaphanous diaphragm was identified, radiographically evident calcification developed along the course of the graft in the 1st week after the congenital diaphragmatic hernia repair (Fig. 2D). In the remaining six children, radiographically apparent calcification did not develop during the interval in which chest radiographs were obtained.
In Vitro Analysis
Radiographs of the two phantoms, with the direction of the X-ray beam
tangential to the graft, were obtained. A radiograph of the first phantom
revealed a subtle 1-mm-thick lucency (Fig.
1C) along the short segment of the graft tangential to the X-ray
beam. The X ray from the second phantom, in which the beam was tangential to
the graft along its entire length, revealed a 1-mm-thick radiolucency along
the entire course of the graft (Fig.
1D).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Etiologic considerations of the diaphanous diaphragm included a subpulmonic pneumothorax. However, the ipsilateral hydropneumothorax that is always present immediately following repair of a congenital diaphragmatic hernia [5, 6, 7] was separate from the diaphanous diaphragm. This distinction was more difficult in the presence of a pneumothorax. In most of the cases, as the hypoplastic lung grew, the surrounding pneumothorax resolved before the diaphragmatic radiolucency disappeared. When this sequence of events occurred, subsequent radiographic findings showing the diaphanous diaphragm could possibly be mistaken for a persistent subpulmonic pneumothorax.
Another possibility was that the diaphanous diaphragm represented a pneumoretroperitoneum. However, no logical explanation could account for the presence of air in the retroperitoneal space because the surgical procedure should not have breached the retroperitoneal boundary. Moreover, there is no retroperitoneal soft-tissue plane that coincides with the location, morphology, and orientation of the diaphanous diaphragm.
The possibility that the diaphanous diaphragm represented pneumoperitoneum was considered and dismissed because the radiolucency never crossed the midline as would be expected of free air in this compartment in a supine patient. Moreover, on chart review, there was no indication that any patient had an abdominal catastrophe that would have resulted in a pneumoperitoneum.
Given the aforementioned analysis of the potential locations of the diaphanous diaphragm, we hypothesized that this sign represented air trapped in the diaphragmatic graft because polytetrafluoroethylene is a porous air permeable substance that is 1 mm in width. To test our hypothesis, we obtained radiographs of two phantoms (Fig. 1A,1B,1C,1D), both of which showed a curvilinear radiolucency in the expected location of the graft. Because we could not exclude the possibility that the curvilinear air visualized on the radiograph of the first phantom was air trapped outside the graft between the graft and the balloon, we created a second phantom in which an airtight seal would be created on either side of the graft so that any radiolucency identified on a radiograph would have to be as a result of air trapped in the graft (Fig. 1A,1B,1C,1D). Subsequent radiographs of the second phantom confirmed that air was indeed trapped in the polytetrafluoroethylene graft.
Another finding seen in three patients that supported the intragraft location of the air was that the redundancy of some air contained in plicated patches created artifacts that were corrugated or more geometric in configuration (Fig. 3).
We attribute the gradual disappearance of the diaphanous diaphragm sign to infiltration of the graft by cellular tissue. Our hypothesis is corroborated by animal studies that have shown that fibroblastic tissue is gradually incorporated into the polytetrafluoroethylene graft [2]. Newman et al. [2] observed gross and histologic evidence of exuberant fibroblastic tissue infiltrating polytetrafluoroethylene grafts in dog models at 4 months after surgical repair with tissue incorporation into the remnant diaphragm completed by 7 months. We hypothesize that cellular infiltration begins immediately and that microscopic infiltration accounts for our observation of the gradual disappearance of the diaphanous diaphragm over a period of days. Calcification of the graft that was observed in our study has also been shown in animal experiments [2].
In conclusion, the diaphanous diaphragm is a new sign indicative of polytetrafluoroethylene patch repair of a diaphragmatic defect. If this sign is recognized, chest radiographs after repair of a diaphragmatic defect with a polytetrafluoroethylene graft can be accurately interpreted, and a mistaken diagnosis of a pneumothorax will be avoided.
Acknowledgments
We thank Jennifer Ogilvie for her help; Jim Buescher, John H. Conway, Alan
Nakagawa, and Roy Steele for their technical assistance; and Wendy Neale for
assistance with manuscript preparation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Marom, B. F. Kammen, and C. A. Gooding The "Diaphanous" Diaphragm: Previously Described Am. J. Roentgenol., October 1, 2002; 179(4): 1074 - 1074. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
