| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Interventional Radiology Case Conferences |
1 All authors: Division of Abdominal Imaging and Interventional Radiology, Harvard Medical School and Massachusetts General Hospital, 55 Fruit St., White 270, Boston, MA 02114.
Received August 7, 2001;
accepted after revision September 17, 2001.
Address correspondence to P.R. Mueller.
Case History
 Top Case History References |
|---|
A water-soluble contrast esophageal swallow examination revealed an anastomotic leak into the mediastinum (Fig. 1A). A contrast-enhanced CT scan obtained using oral contrast material showed a 4x4x4 cm fluid collection on the right posterior aspect of the mediastinum (Fig. 1B). CT-guided drainage of the perianastomotic mediastinal fluid collection was performed with the patient under conscious sedation in the right lateral decubitus position (Figs. 1C,1D,1E).
|
|
|
|
|
Initially, 30 mL of pus was drained from the collection, but the catheter output decreased over the subsequent few days. In addition, the patient remained febrile, and his WBC continued to be elevated. CT performed to assess catheter position and the size of the residual mediastinal fluid collection revealed interval development of a large cavity in the posterosuperior aspect of the right pleural cavity, which was immediately contiguous with the original mediastinal collection (Fig. 1F).
|
The patient underwent fluoroscopy for evaluation. A contrast injection through the mediastinal catheter showed that contrast material filled this large loculation (Fig. 1G). In addition, the intrathoracic stomach distal to the anastomosis opacified with contrast material, indicating that the anastomotic leak remained. It was, therefore, decided to replace the mediastinal catheter with a larger 16-French multipurpose drainage catheter (Cook, Bloomington, IN) and to reposition it with the catheter tip extending into the pleural collection.
|
Under fluoroscopic guidance, the existing catheter in the mediastinum was exchanged for a Kumpe catheter (Cook) over an Amplatz guidewire (Cook). The Kumpe catheter and a Terumo glidewire (Boston Scientific, Watertown, MA) were manipulated in a posterolateral direction using anteroposterior and oblique fluoroscopic projections until the optimal position in the cavity was achieved. The Terumo glidewire was then exchanged for the Amplatz guidewire, and the 16-French catheter was advanced into the right pleural collection (Fig. 1H). Both an injection of contrast material and a CT scan obtained the next day confirmed a satisfactory catheter position (Fig. 1I).
|
|
Despite the satisfactory position of the catheter, the referring surgeons were still concerned about the anastomotic leak from the esophagus. They believed that the indwelling catheter would not be sufficiently large for prolonged drainage, most of which would occur when the patient was an outpatient. Three days after the catheter manipulation, the patient underwent a rib resection and the placement of a chest tube. The surgeons used a modified 14-mm Montgomery salivary bypass tube (Boston Medical Products, Westborough, MA) that had been cut to a length of 9 cm and to which extra side holes had been added. The patient then was discharged home.
The patient remained afebrile at home but noted continuous tube output as his oral intake increased. The surgeons decided to withdraw the mediastinal catheter in 2- to 3-cm increments at 1- to 2-week intervals, hoping that the fistula would seal. One month later, an esophageal swallow examination performed with water-soluble contrast material revealed free flow of contrast material across the anastomosis with no evidence of an anastomotic leak.
Dr. Maher. What is the reported incidence of anastomotic leakage after esophagectomy? Is this complication more commonly seen after one type of operation than another, and what are the risk factors associated with anastomotic leakage?
Dr. Boland. Esophageal anastomotic leakage is one of the most serious complications of esophagogastric surgery with reported incidences ranging from 1% to 29% [1]. Cervical anastomoses consistently have higher leakage rates than intrathoracic anastomoses (10-25% vs less than 10%, respectively) [1]. When esophageal anastomotic leakages occur, only 21% are detected before the seventh postoperative day. The median duration to the development of esophageal leakage is 12 days, with a range of 2 to 28 days [2]. Therefore, in our patient, the time of development of an esophageal leak was typical. Intrathoracic esophageal anastomotic failure, as seen in our patient, is associated with a mortality rate reported to be as high as 60% [3].
Dr. Maher. What are the appropriate diagnostic modalities to use in the investigation of a febrile patient who has undergone an esophagogastectomy and is thought to have an esophageal leak?
Dr. Lucey. In the early postoperative period, before the initiation of fluid intake, all patients routinely undergo a water-soluble contrast esophageal examination to exclude the possibility of anastomotic leakage. Traditionally, patients are initially examined using water-soluble contrast material rather than barium sulphate for two reasons. First, in the presence of an esophageal leak, barium will extravasate into the mediastinum, and the presence of barium in the mediastinum has been reported to result in fibrosing mediastinitis [4]. In addition, the long-term presence of barium in the mediastinum makes interpretation of future mediastinal imaging difficult [5]. If no leak is revealed with water-soluble contrast material, barium sulphate can be administered. This protocol allows careful definition of the anastomotic region and confirms the absence of leakage because the denser barium is more likely to identify any tiny leak.
Contrast-enhanced CT should also be performed to determine the extent of inflammation in the mediastinum and to reveal the presence of focal mediastinal collections. Such information aids in assessing whether radiologic intervention is indicated.
Dr. Maher. Once anastomotic leakage is confirmed, what are the treatment options available for these patients?
Dr. Mueller. There is no universally accepted algorithm for the treatment of patients with intrathoracic leaks after esophageal resection [2]. Treatment varies depending on the surgical approach and the availability of subspecialists, such as critical care physicians and interventional radiologists.
Treatment of postoperative anastomotic leaks in the postesophagectomy patient also varies according to the site and severity of anastomotic leakage. All patients require nutritional support, preferably via the enteral route [1]. The treatment of fulminant early leaks is quite different from delayed leaks. Early fulminant leaks are usually due to gastric necrosis [6] with diffuse mediastinal infection. Thoracotomy and reoperation of the anastomosis are usually necessary. The patient is then left with a cervical esophagostomy and a closed-off stomach remnant that is returned to the abdomen. Despite this drastic treatment, mortality from fulminant early leaks approaches 90% [6]. Because of the diffuse nature of the inflammatory process in the mediastinum in this group of patients, interventional radiology does not play a role.
Most anastomotic leaks are usually well contained within the mediastinum [1]. Traditionally, these patients have been advised to undergo a surgical intervention that involves mediastinal drainage, debridement of infected and necrotic tissue, and placement of chest tubes to provide thorough drainage [1]. Typically, attempts to repair the anastomosis are made but are usually unsuccessful [1]. In essence, for many patients, this surgery is performed to render the mediastinum sterile so that reanastomosis can be successfully accomplished at a later stage.
Although little has been written in the literature advocating a role for interventional radiology in this setting, radiology has the potential theoretic advantage of permitting accurate placement of mediastinal catheters at the site of anastomotic breakdown, which allows control of the esophageal leak and drainage of single or multiple mediastinal collections. Drainage of mediastinal collections controls sepsis and helps prevent formation of fistulas in the adjacent pericardium, aorta, bronchi, or pulmonary vessels, which can result in fatal outcomes [6].
Dr. Maher. Once the decision to attempt percutaneous drainage of the mediastinum has been made, what are the critical technical points in performing the procedure?
Dr. Gervais. Treatment of intrathoracic esophageal anastomotic leakages requires that the catheters be placed in the mediastinal collection close to the site of anastomotic leakage to achieve adequate control of the leak. In our patient, the collection was located on the right side of the mediastinum. Therefore, catheter placement had the potential for transgression of the pleura, which could result in pneumothorax or in seeding the pleural cavity with infection, resulting in pleural empyema. Furthermore, care had to be taken to avoid the azygos vein and esophagus on the right side.
When draining abscesses appear inaccessible on axial CT, one can sometimes angulate the gantry to identify an access route in another plane, either through an intercostal space or other routes that avoid vital organs. Knowing the direction of angulation of the gantry, the operator can easily adjust the angle of catheter insertion [7]. In some patients, it may be helpful to inject saline into the pleural cavity to separate the visceral pleura from the parietal pleura, creating a larger space in which to place catheters and thus reducing the risk of pneumothorax [8].
Dr. Maher. What were the reasons for placing the patient in the right lateral decubitus position during the catheter procedure (Fig. 1C)?
Dr. Gervais. Generally, in performing biopsies or percutaneous drainage close to the midline, it can be useful to place the patient "ipsilateral side down" during the procedure. The dependent lung becomes deflated, which decreases the risk of transgressing the lung. Placing our patient in this position during this procedure caused the mediastinal collection to move into a more dependent position away from the midline and made it more accessible for percutaneous drainage. Also, placing the patient in this way made the lung fall into a more dependent position away from the proposed path of the drainage catheter, thus reducing the risk of pneumothorax and empyema.
Dr. Lucey. In what situation might you approach the patient with the contralateral side being dependent?
Dr. Boland. First, placing a catheter in the midline close to the esophagus requires a large angle between the catheter and the skin. The correct catheter angle may be difficult to create if the abscess is drained with the patient in the ipsilateral side—down position because the movement of the operator's hands is limited by the top of the CT table. Conversely, when the patient is placed contralateral side down, there is adequate room for the movement of the operator's hands, making accurate placement of the catheter easier.
In addition, the general pulmonary condition of the patient must be considered when positioning the patient. For example, if there is severe pulmonary dysfunction in one lung and the functioning lung is compromised by placing it in a "down" position, the patient may experience respiratory compromise.
Dr. Lucey. This procedure was performed using the trocar technique. What are the advantages of using the trocar technique rather than the Seldinger technique?
Dr. Mueller. In this institution, we perform most percutaneous drainages using the trocar technique, which is safe if performed correctly. After administering adequate local anesthesia (1% lidocaine) to the patient, a 10-cm, 19-gauge Turner needle (Cook) is advanced through the intercostal space and into the mediastinal collection under CT guidance (Fig. 1C). Once the position and course of the localizing needle are satisfactory, a skin nick is made adjacent to the entry point of the localizing needle with a number 11 blade, and the tract is enlarged by dissection with a metal forceps [7]. A hydrophilic 8- to 12- French catheter (Cook) is lubricated with sterile saline and is inserted parallel to the angle of the localizing needle using the tandem technique. When the catheter has reached the collection, the internal metal stiffener is detached, and the catheter is advanced into the collection (Fig. 1D). The catheter is secured with the pigtail locking mechanism. The contents of the collection are immediately aspirated. After axial CT scans confirm that the catheter is in a good position, the catheter is carefully fixed to the skin. The abscess is then aspirated until it is dry and irrigated with normal saline. CT scanning is repeated to assess abscess size and satisfactory catheter position in the abscess (Fig. 1E).
We use the trocar technique for several reasons. The trocar technique is the preferred technique for drainage of collections that are small and relatively superficial [7]. Unlike the Seldinger technique, the trocar technique requires no sequential dilatation [7]. The main disadvantage of the trocar technique is that repositioning the catheter may be difficult without reinsertion in cases in which the catheter is deployed outside the abscess cavity [7]. A second attempt at catheter placement would then be required. Therefore, the initial localizing needle should not be removed until the catheter has been successfully placed.
In situations requiring a long access route, as is frequently the case in pelvic or abdominal collections, the use of the trocar method is still possible and frequently appropriate, although it may be a little more difficult to use for this purpose than it is for drainage of superficial collections. The Seldinger technique is often perceived as being easier than the trocar technique, but this may not always be the case, especially when small collections are being drained or when CT is the sole imaging technique being used. In addition, without fluoroscopic guidance, kinking of the guidewire can occur, and occasionally, loss of access can occur when advancing the catheter [7].
Dr. Maher. Is there ever a role for simple aspiration of the collection in these patients, or should a drainage catheter always be left in place?
Dr. Boland. In most patients, it is advisable to insert a drainage catheter and to leave it secured for a number of days, particularly if the contents of the collection are suggestive of pus or infected material or if the patient shows signs of sepsis [7]. This precaution is especially advisable if there is a suspicion that the collection communicates with bowel, as was the case with our patient. Simple aspiration of collections connecting with bowel would be totally ineffective; immediate reaccumulation of the collection would occur once the needle or temporary catheter was removed.
Dr. Maher. The character of the patient's drainage did not change, but the amount of drainage tapered abruptly. What does this mean?
Dr. Mueller. The character of the patient's drainage in this case was relatively clear despite the communication with the esophagus because the patient was not eating. Had the patient been fed, this material would have drained out of the catheter. The rapid change in the amount of drainage did not correlate with the patient's clinical signs of persistent fever and an elevated WBC. The decrease in drainage should be gradual, but in this patient, it was not. If there is an abrupt change in the amount of drainage and the patient is still unwell, the radiologist should ensure that the catheter is not clogged. Difficulty in irrigating the catheter is likely to mean that the decreased drainage is caused by catheter obstruction, which could explain why the patient remained febrile.
In our patient, a second loculation was not drained, causing the patient's continuing fevers as well as the lack of drainage from the first catheter. Injection of contrast material through the catheter showed a small communication between the two loculated areas. However, the isthmus between the two loculated collections was narrow, and it appeared that the initial catheter was not draining the second collection.
Dr. Maher. CT showed a second collection near the original collection. How do you decide between manipulating your catheter into the second collection or performing a second drainage?
Dr. Gervais. This decision is sometimes difficult to make. The location of the second collection and ease of access to this collection for a second percutaneous drainage will influence whether placement of a second catheter or manipulation of the existing catheter is contemplated. Injection of the catheter should always be performed to assess the feasibility of manipulation of the catheter. Communication of the original cavity with the second collection must be shown before embarking on remanipulation of the catheter. The injection can be performed using fluoroscopic guidance or immediately before CT scanning. If the injection is performed fluoroscopically, images should be acquired in the anteroposterior and lateral positions to ascertain the relative position of the two collections in orthogonal planes.
Dr. Maher. What directed you to obtain a follow-up CT so soon after drainage?
Dr. Gervais. Clinical follow-up is extremely important. The patient's clinical well-being, WBC, and the type and amount of drainage over time should be examined daily by the radiologist. Our patient was clinically improved but still maintained a low-grade fever and elevated WBC and, most important, had a dramatic change in catheter output. These findings suggested that the patient was not "cured" and that CT was necessary to determine if there was a residual mediastinal collection.
Dr. Maher. Why did you find the decrease in catheter output of concern? Such a decrease usually means the abscess is healing.
Dr. Boland. Evaluation of catheter output is valuable but is sometimes overrated as being clinically useful in determining successful catheter drainage. It is important to understand that the absolute amount of drainage is not as important as the day-to-day variation. For example, a small liver abscess may drain 20-30 mL per day, whereas a large subphrenic abscess may initially drain 100-200 mL per day. Therefore, what is important in monitoring catheter drainage is the character of the drainage and the change in drainage over time. Most abscess fluid is serosanguinous; however, if there is a fistulous communication, the character may change depending on the structure to which the fistula is connected. For example, bile, pancreatic juice, and urine all have different appearances.
Dr. Maher. How do you assess the impact of percutaneous drainage of mediastinal collections in this patient, and was the outcome achieved acceptable in this patient?
Dr. Mueller. As stated previously, the mortality associated with esophagoenteric anastomotic breakdown with intrathoracic leakage is widely reported as approaching 40-60% [3]. Some patients require additional surgery of the anastomosis. This patient did require minor surgical intervention subsequent to radiologic intervention, but overall the role of percutaneous drainage had a positive effect on ultimate patient outcome. In addition, the mediastinal drain—accurately placed in two stages under CT and fluoroscopic guidance by interventional radiologists—aided the surgeon in placing the definitive large-bore tube in the correct position.
Dr. Maher. In summary, we have presented a case of postoperative esophagoenteric anastomotic breakdown with intrathoracic leakage. The leak occurred on postoperative day 8 and was confirmed radiologically by a water-soluble contrast study. A mediastinal collection was revealed on CT and drained percutaneously, a procedure that was beneficial to patient outcome. The catheter outputs from the anastomotic leak were completely controlled and diverted from the mediastinum. This procedure aided in controlling mediastinal sepsis, which was vital to subsequent anastomotic healing.
|
TopCase HistoryReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. J. Brinster, S. Singhal, L. Lee, M. B. Marshall, L. R. Kaiser, and J. C. Kucharczuk Evolving options in the management of esophageal perforation Ann. Thorac. Surg., April 1, 2004; 77(4): 1475 - 1483. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
