IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

INTERVENTIONAL CHEST RADIOLOGY

0033-8389/00 $15.00

+ .OO

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS Jeffrey S. Moulton, MD

Pleural fluid collections may be classified as uncomplicated or complicated.l2 Uncomplicated pleural fluid collections refer to those that resolve with conservative therapy. In general, this includes transudative effusions and small free-flowing exudative effusions. Complicated pleural fluid collections refer to those that do not resolve without drainage. This includes moderate to large unilocular or multilocular parapneumonic effusions; empyemas; malignant effusions; and hemothoraces (either sterile or infected). Drainage of complicated pleural fluid collections is necessary to control pleural sepsis, allow re-expansion of the underlying lung, and to prevent the long-term sequelae of pleural fibrosis and lung entrapment. Drainage of complicated pleural fluid collections has traditionally been accomplished by either closed-tube thoracostomy or by open surgical drainage. Closed-tube drainage is associated with a lower morbidity and mortality than open surgical drainage, but it may fail in a significant percentage of patients because of presence of loculations or the fibrinous nature of the fluid. Two techniques have been used in an effort to improve the efficacy of closed-tube drainage: (1)image-guided placement of drainage catheters and (2) intracavitary fibrinolytic therapy (ICFT). Image-guided placement of

thoracostomy tubes has been advocated to avoid the problem of tube malposition in relation to fluid loculations. Although this is effective in many patients, the drainage procedures tend to be prolonged and failure may still occur because of occlusion of the relatively small-bore catheters by fibrinous debris or clotted blood. Intracavitary fibrinolytic therapy has been advocated as a method to facilitate drainage of fibrinous pleural fluid through small catheters and to allow enzymatic dkbridement of the restrictive fibrin sheets covering the pleural surface. These two methods have significantly increased the effectiveness of closed-tube drainage and can obviate the need for surgery in most patients. This article reviews image-guided thoracostomy drainage and ICFT, including the rationale for their use, techniques of drainage, patient management, results, and an overview of their role in the larger spectrum of interventional pleural procedures. Emphasis is placed on the management of complicated parapneumonic effusions, because these are the most common entities encountered in clinical practice. RATIONALE BEHIND IMAGE-GUIDED DRAINAGE The temporal evolution of complicated exudative pleural effusions is classically divided

From the Department of Radiology, St. Anthony Hospital; and the Department of Radiology, University of Colorado School of Medicine, Denver, Colorado

RADIOLOGIC CLINICS OF NORTH AMERICA VOLUME 38 * NUMBER 2 * MARCH 2000

345

346

MOULTON

into three stages.12The first stage (exudative phase) is a thin free-flowing exudate. The cellular and proteinaceous content of the fluid is minimal and the underlying lung and pleura remain pliable. The second stage (fibrinopurulent phase) is characterized by fluid with increased cellularity and protein content. The protein in the fluid includes normal serum clotting factors that promote the formation of fibrin nets in the fluid, deposition of fibrin sheets on the pleural surfaces, and pleural loculations. There is atelectasis of the underlying lung caused by both the mass effect of the fluid and lung trapping by the relatively inelastic fibrin sheets. The third stage (organizing phase) is characterized by ingrowth of capillaries and fibroblasts into the pleural membranes. This eventually leads to the formation of a mature inelastic fibrous pleural peel. Progression from stage 1 to stage 2 can occur in as little as 1day, whereas progression from stage 2 to stage 3 requires several weeks. Early drainage of complicated effusions is necessary to prevent progression to the organized hase. In t e case of infected pleural collections (empyema) the immediate goal of drainage is to control pleural sepsis. An equally important goal of drainage of both infected and sterile collections is to allow re-expansion of the underlying collapsed lung. In the case of parapneumonic effusions re-expansion of the lung greatly facilitates transbronchial clearance of intrapulmonary secretions. It also promotes restoration of pleural elasticity, which is crucial in preventing the formation of a fibrous pleural peel. Regardless of the underlying origin, the basic surgical principles of drainage are the same: complete evacuation of the pleural fluid, sterilization and obliteration of the pleural space, and re-expansion of the adjacent lung with restoration of pulmonary function. The keys to accomplishing these goals are early dependent drainage, pleural dkbridement, and breakdown of loculations.

1:

Traditional Drainage Modalities

A wide spectrum of interventional procedures is available for drainage of complicated effusions including single or repeated thoracentesis, closed-tube thoracostomy drainage, thoracoscopy, and formal thoracotomy with either decortication or open drainage. The most common method of treatment is a moderately prolonged (7 to 10 days) trial of

closed-tube thoracostomy drainage followed by formal surgical drainage if unsuccessful.l2,25 The rationale behind this approach is that open surgical drainage is more invasive with a longer postoperative recovery and higher morbidity and mortality. Many surgeons, however, recommend a shorter trial (2 to 4 days) of closed drainage before proceeding to surgery.18,26 The rationale behind early open surgical intervention is that prolonged ineffective closed thoracostomy drainage leads to more technically difficult open drainage with higher associated morbidity and mortality.', 11, *l Some surgeons feel so strongly about this that they recommend open drainage as the initial therapeutic approach. Proponents of each approach support their stance with extensive statistics on the success rates, morbidity, and mortality of each therapeutic option. Some of these reports address the success of a single method of drainage in a given group of patients, whereas other reports attempt to compare the success of different methods of drainage. With few exceptions these reports are retrospective reviews with historical controls. The results of these studies are very difficult to compare because of the tremendous number of variables that affect outcome in individual cases. These variables include comorbid medical conditions, inclusion criteria for different studies, delays in diagnosis and institution of therapy, the stage of the effusion, virulence of the causative organism, host resistance, differences in antibiotic therapy, and aggressiveness of surgical intervention. Because of these variables, there is an inherent selection bias in all of the studies, which limits the usefulness of historical controls. With these limitations in mind, it is helpful to review the historical success of traditional closed-tube thoracostomy and to study the reasons why it is not always successful. Nonguided thoracostomy is accomplished by bedside placement of one or more large-caliber (22F to 34F catheter) thoracostomy tubes that are then placed to suction using a water-seal system. Closed drainage is considered successful if the effusion is completely drained without the need for open surgical intervention. Reported success rates of this method of drainage range from 6% to 78%, with most studies reporting less than 50% success.1,6, 9, The tremendous variability in success rates reflects the fact that the success of any given treatment modality depends on how early and on which patients it is used. When ana-

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

lyzed by stage, these studies demonstrate that nonguided, closed thoracostomy is effective in completely evacuating stage 1 effusions, but is less effective for stage 2 or 3 multiloculated effusions. If closed drainage could be made more effective for stage 2 loculated effusions, there would be little argument against its role as the initial therapeutic approach. In attempting to devise a better method of drainage it is very helpful to analyze the reasons why blind placement of large-caliber thoracostomy tubes often fails. The most common reason is that pleural adhesions and limiting fibrin membranes form loculations of fluid away from the drainage tube.lS,21 Even in skilled hands, it is difficult to accurately direct bedside placement of a thoracostomy tube into a specific loculation. Following partial evacuation of the pleural space, new adhesions and loculations form very quickly and may limit the effectiveness of a tube that was initially in satisfactory position (Fig. 1). Drains that are in satisfactory position may be ineffective if the fibrinous fluid assumes a gelatinous state or if the fibrinous debris and septations within a collection occlude the

347

side-holes of the tube.15,21 Less commonly, the tube may be placed in an inappropriate position, such as within a pleural fissure, within the lung parenchyma itself, or in the chest wall (Fig. 2). Lastly, the presence of an inelastic fibrous pleural peel or persistent bronchopleural fistula prevents obliteration of the pleural space even with a properly positioned tube in place. Image-Guided Drainage Image-guided placement of thoracostomy tubes has been advocated to address the problem of tube malposition in relation to fluid loculations. By using modern cross-sectional guidance techniques, primarily CT and ultrasound, it is relatively easy to place one or more drainage catheters into specific loculations of fluid regardless of the size or location of the collection. This is the single most significant advantage of guided over nonguided thoracostomy and its importance should not be underestimated. Reported success rates of guided thoracostomy range from 67% to 83%.13,20, 27 The drains used for guided

Figure 1. Parapneumonic empyema in a 56-year-old man initially treated with nonguided placement of a 36F chest tube. Fluid output from the chest tube stopped after 2 days of drainage. CT scan obtained at that time shows collapse and consolidation of the left lung, with the visceral and parietal pleural surfaces apposed around the laterally positioned chest tube. There is a large residual loculation of fluid in the posterior pleural space. The chest tube is no longer in a position that allows drainage of the residual fluid. (From Moulton JS: Imageguided drainage techniques. Semin Respir Infect 1459-72, 1999.)

348

MOULTON

Figure 2. Parapneumonic empyema in a 51-year-old man. A, Chest radiograph obtained after nonguided placement of a 36F chest tube shows the tube superimposed over the midlung, with persistent pleural fluid lateral to the tube. The exact position of the tube in relation to the pleural space cannot be determined on this single-view chest radiograph. B, CT scan obtained 1 day after the above chest radiograph shows that the chest tube (arrow) is within the lung parenchyma rather than in the pleural space. In this position, the tube is not able to effectively drain the anterolateral loculation of air and fluid. (From Moulton JS: Image-guided drainage techniques. Semin Respir Infect 14:5%72, 1999.)

thoracostomy tend to be smaller than for nonguided drainage, ranging from 8F to 16F catheter in size. Drains of this size are better tolerated by patients than are large-bore thoracostomy tubes, but they are more prone to obstruction by fibrinous debris. The drainage procedures tend to be prolonged with reported duration of drainage ranging from 7 to 26 days.13,16, 2o

The primary surgical objections to imageguided thoracostomy are the relatively small size of the drains (and the resultant tendency to occlude) and the somewhat prolonged course of drainage. Despite proper positioning and aggressive catheter management, image-guided thoracostomy may still fail if the catheter is repeatedly occluded by fibrinous debris. After a period of initially rewarding

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

drainage, separate loculations still tend to occur at sites distant from the drain, just as with nonguided tubes (Fig. 3). Lastly, guided thoracostomy is no more effective than nonguided thoracostomy in treating stage 3 empyemas with a pleural peel, which is to say it is not very successful at all (Fig. 4).

349

lntracavitary Fibrinolytic Therapy

The use of fibrinolytic agents to facilitate drainage of hemorrhagic and fibrinous pleural fluid collections dates back to the late 1940s. The theory behind intracavitary instillation of fibrinolytic enzymes is that they de-

Figure 3. Multiloculated stage 2 parapneumonic effusion in a 38-year-old man. A, CT scan 2 days after hospital admission shows left lung consolidation and volume loss with separate anterior and posterior loculations of pleural fluid. B, CT scan of the upper chest 2 days after guided placement of a single posterior 12F chest drain shows apparent evacuation of both fluid loculations, with re-expansion of the left lung. Illustration continued on following page

350

MOULTON

Figure 3 (Continued). C, CT image of the lower chest at the same time as 6 shows persistent anteromedial and posteromedial loculations of fluid. These two loculations were not effectively drained by the single chest tube because of intervening pleural adhesions. Additional drains were subsequently placed for complete evacuation of the remaining loculations. (From Moulton JS: Image-guided drainage techniques. Semin Respir Infect 1459-72, 1999.)

crease the viscosity of gelatinous pleural fluid; break down fibrinous septations and adhesions; and d6bride the pleura of fibrinous sheets, allowing re-expansion of the underlying lung. Although the technique has enjoyed sporadic support in the literature over the last 40 years, it has become more widely accepted only recently. In both theory and practice, ICFT allows evacuation of viscous fluid through relatively small drains by dissolving the gelatinous component and associated fibrinous debris l5 It also apthat can occlude the ~atheter.~, pears to at least partially dbbride the restrictive pleural fibrin sheets that prevent reexpansion of the lung. As the lung expands and the visceral and parietal pleural surfaces come in contact, however, adhesions form quickly and may lead to the formation of fluid loculations away from the drain. Pleural adhesions at these contact points are not cawed by fibrin deposition and, outside 04 a direct mochadsiol effect, are rtor frr theory susceptible to breakdawn using fCFT.i7 lrt clinical experience, it appears that ICFT greatly facilitates complete evacuation of the loculation in which the drain is placed but does not reliably break down pleural adhe73

sions and allow drainage of adjacent loculations by a single pleural drain (Figs. 5 and 6 ) . The use of ICFT by way of chest tubes placed without imaging guidance is less likely to be successful largely because of the frequency of tube malposition in relation to loculations. For these reasons, aggressive image-guided catheter management is the most important factor in success and ICFT, although extremely helpful, is adjunctive. It is crucial to always keep the drain or drains properly positioned within any given loculation, and this can be accomplished only with relatively frequent cross-sectional imaging and tube manipulations as needed. ICFT should not be used in an attempt to salvage success by way of a malpositioLed tube.

TECHNIQUE OF IMAGE-GUIDED THORACOSTOMY Radlographlc Imaging for Assessmmt and Guidance

Conventional radiography plays a very important role in both preprocedure evaluation

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

Figure 4. Chronic stage 3 postoperative exudative effusion in a 70-year-old man. A, CT scan shows a loculated posterior pleural fluid collection with adjacent left lower lobe atelectasis. The effusion was present for 6 months at the time of this CT scan. There is only minimal visible parietal pleural thickening. B, CT scan obtained 12 days after guided placement of a 12F chest tulle shows a persistent pleural cavity. Much of the fluid has been evacuated and replaced by a loculated pneumothorax, with no re-expansion of the adjacent lung. The patient subsequently underwent a decortication, which showed a thick restrictive fibrous pleural peel that prevented re-expansion of the lung. (From Moulton JS: Image-guideddrainage techniques. Semin Respir Infect 1459-72, 1999.)

351

352

MOULTON

Figure 5. Loculated infected hemothorax after a gunshot wound to the chest in a 56year-old man. The patient was initially treated by formal thoracotomy and large-bore chest tube placement. A, CT scan 8 days after surgery shows a large-bore chest tube entering laterally, with a separate loculated posterior pleural fluid collection. The patient showed signs of sepsis at the time of the CT scan. Guided placement of a 12F chest tube was performed immediately after the CT scan. B, CT scan obtained 4 days after placement of the 12F tube shows that the drain is in satisfactory position but is surrounded by residual fluid. There had been no significant drain output over the previous 24 hours. The patient was subsequently treated by intracavitary urokinase therapy through the 12F drain. Illustration continued on opposite page

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

353

Figure 5 (Continued). C,CT scan obtained 1 day after B shows complete evacuation of the cavity. The urokinase irrigation allowed effective drainage of relatively viscous fluid through the small chest tube. The key to effective drainage of this cavity was that the drain was properly positioned within the loculation before beginning urokinase therapy. Lytic therapy would most likely have been ineffective if administered through the larger lateral chest tube. (From Moulton JS: Image-guided drainage techniques. Semin Respir Infect 14:59-72, 1999.)

of the pleuroparenchymal disease process and in directly guiding the thoracostomy procedure. All patients with significant pneumonia initially undergo plain chest radiography. Depending on the degree of pulmonary consolidation, the presence of an associated pleural effusion may at least be suspected if not confirmed on routine films. Although decubitus chest radiographs are more accurate in confirming the presence of an effusion and give some information as to whether it is free flowing or loculated, the usefulness of these films has lessened with advances in crosssectional imaging. Cross-sectional modalities, primarily ultrasound and CT, have become the cornerstone in the imaging evaluation of pleural disease. Ultrasound is extremely accurate in confirming the presence of pleural fluid. It is widely available, inexpensive, can be performed portably on extremely ill patients, and does not use ionizing radiation. Ultrasound cannot image through bone or aerated lung, however, which limits its effectiveness in localizing small fluid loculations near the scapula, within a fissure, or in a paramediastinal location. Ultrasound is not useful in evaluating the status of the underlying lung and airways. The main advantage of ultrasound as a guidance modality is its flexibility. Pa-

tients can be imaged in any position, including an upright sitting position, which facilitates access to free-flowing effusions that are always in a dependent location. Sonography, therefore, is the most common modality used for image-guided thoracentesis. Access to loculated effusions that contact the chest wall can also be accomplished accurately. Ultrasound does not allow accurate continuous real-time monitoring of the course of a drainage catheter during placement, but this is not a significant practical disadvantage. CT is the most accurate modality available for complete evaluation of the pleural space, the underlying lung, the pulmonary hila, and the adjacent mediastinal structures. It can reliably establish the presence of pleural fluid, distinguish between fluid and drowned lung, differentiate loculated versus free-flowing effusions, and clearly map the size and location of all loculations regardless of location. CT is also capable of detecting the presence of other conditions that may have a profound impact on management decisions including central masses or other bronchial obstruction, lung abscess, or a bronchopleural fistula (BPF)(Fig. 7). As a guidance modality, the main limitation of CT is that it must be performed with the patient in a supine, prone, or decubitus position, which limits access to small free-

354

MOULTON

Figure 6. Stage 2 parapneumonic empyema in a 73-year-old woman. The patient presented with persistent fever and leukocytosis after 4 weeks of outpatient antibiotic therapy for pneumonia. A parapneumonic empyema was found, with the majority of the pleural fluid in the lower chest. This was treated with image-guided drainage. A, CT scan of the midchest after 3 days of drainage shows two adjacent undrained loculations of fluid with an intervening pleural adhesion. A 1OF drain was placed into the more lateral loculation and treated with intracavitary urokinase. B, CT scan 4 days after A shows the 1OF drain in place (arrow) within the lateral loculation, which has been completely drained. The pleural adhesion was not affected by the course of lytic therapy and the medial loculation remains undrained. An additional drain was subsequently placed into the medial loculation. Illustration continued on opposite page

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

Figure 6 (Continued). C,CT scan 2 days after 6 shows that the medial loculation has been completely drained by the second catheter.

Figure 7. A 78-year-old man, who presented with weight loss, cough, and fever. CT scan of the lower chest shows complete collapse and consolidation of the left lower lobe. There is an air- and fluid-filled cavity in the lung that has ruptured into the pleural space causing a pyopneumothorax. Open surgical exploration was performed and showed that the cavity was a necrotic obstructing squamous cell carcinoma with associated infected pleural carcinomatosis. This situation is not amenable to treatment by closed thoracostomy, because this form of treatment is unable to address the underlying obstructing tumor. (From Moulton JS: Imageguided drainage techniques. Semin Respir Infect 14:59-72, 1999.)

355

356

MOULTON

flowing effusions. The main advantage of CT is its ability to localize loculated effusions accarately, regardless of location. A safe and accurate access pathway to any given lowlation can be planned, including identification and localization of different fluid pockets in multiloculated effusions that may require separate catheters for complete drainage (Fig. 8). Most patients with loculated parapneumonic effusions in whom thoracostomy is contemplated should undergo a diagnostic CT examination before drainage. The drainage procedure itself can easily be accomplished immediately following the diagnostic examination. As with ultrasound, CT does not allow real-time continuous monitoring of the course of the catheter during placement, but with experience this is not a significant disadvantage. The recent advent of CTfluoroscopy may eliminate this limitation. Fluoroscopy is an additional guidance modality that is widely available but less frequently used. It is not as accurate or safe in guiding access to loculated effusions as crosssectional imaging and is limited to patients in a supine, prone, or decubitus position. It does, however, allow continuous monitoring of the course of the guidewire and catheter after needle access to the fluid has been accomplished. If catheter position is crucial, localization and needle access may be performed with cross-sectional imaging followed by fluoroscopic monitoring of the catheter placement procedure. Technique of Catheter Placement

The choice of a guidance modality is based on availability, the radiologist’s preference, the patient’s condition, and technical considerations regarding the size and position of loculations as outlined previously. The next decision to be made is whether one or more catheters are needed for complete drainage. Multiple loculations require multiple catheters (Fig. 9). It may also be beneficial to place separate catheters into different places within a very large unilocular effusion. As fluid is drained from a large effusion the lung reexpands and eventually contacts the chest wall. These chest wall contact sites quickly form adhesions that can isolate one or more pockets of fluid distant from the original catheter location. Placement of more than one catheter initially may avert the need for a second drainage procedure as lung re-expan-

sion progresses. For posterior effusions, it is helpful to have a drain low in the posterior costophrenic sulcus because this is a common site of persistent fluid loculation late in the course of the drainage procedure. The catheters used for pleural drainage have evolved significantly over the past decade. The materials used in catheter construction are much softer and more flexible than standard large-bore surgical thoracostomy tubes, therefore they cause considerably less chest wall pain and are easier to secure to the chest wall. The end of the catheters typically have a pigtail configuration, which is held in shape by an internal locking mechanism that provides some measure of protection against inadvertent catheter removal by an uncooperative patient. The most frequently used catheters range in diameter from 8F to 16F, although both smaller and larger drains are available. The choice of catheter size is based on the viscosity of the fluid to be drained. When performing a simple thoracentesis for a free-flowing effusion, an 8F or 10F catheter suffices. For drainage of loculated parapneumonic exudates, the author most commonly begins with one or more 12F catheter drains. If repeated catheter occlusion is a problem, then a 14F or 16F catheter drain may be placed during a subsequent catheter exchange. The site chosen for puncture should be at the edge of the inferior (dependent) part of the loculation so that the catheter may traverse a larger portion of the pleural space. Try to avoid puncture sites in the posterior chest wall, because this can lead to moderate discomfort for a supine patient lying on the tube entry site, although a paraspinous access route is occasionally necessary to drain a small posteromedial fluid loculation. Also, try to avoid placing a catheter just medial to the scapula because this leads to a greater risk of subsequent catheter dislodgement with scapular movement. As with all chest access procedures, great care must be taken to place the needle immediately cephalad to the closest rib to avoid the intercostal neurovascular bundles (which run immediately below each rib). This seems to be especially important when draining posterior apical collections, because the ribs are closer together in this region and the risk of hemorrhage from an intercostal vessel injury is greater. Most drainage procedures can be performed easily with the patient in a comfortable supine or contralateral posterior oblique position. A decubitus

Figure 8. Multiloculated parapneumonic empyema in a 41-year-old woman. The patient initially presented with a large unilocular effusion that was treated with a large volume therapeutic thoracentesis. A chest radiograph (not shown) revealed multiple recurrent fluid loculations. A-C, Three images from a CT scan of the chest show multiple bilateral pleural fluid loculations. The air within several of the loculations was introduced by the initial thoracentesis. The presence of multiple separate loculations requires placement of multiple chest tubes for effective drainage. Five small-caliber chest drains were placed into five separate fluid loculations immediatelyafter the diagnostic CT examination. Each of the loculations was subsequently treated with intracavitary fibrinolytic therapy using tissue plasminogen activator. Illustration continued on following page

357

358

MOULTON

Figure 8 (Continued). 0,A digital scout image from a CT scan 2 days after the drainage procedure shows five chest drains in place. There has been a significant decrease in the amount of pleural fluid and near complete reexpansion of the right lung. The drains were sequentially removed over the subsequent 3 days and the patient recovered completely.

position is occasionally necessary to gain access to a posteromedial fluid loculation. In patients with a unilateral pneumonia and pleural collection, this positioning usually entails placing the patient with the healthy lung down. This can lead to hypoventilation in the normal lung and potentially dangerous hypoxemia. Close physiologic monitoring including ulse, blood pressure, and pulse oximetry s ould be performed throughout the drainage procedure. Catheter introduction may be performed using either the trocar technique or the Seldinger technique. The author uses the Seldinger technique exclusively because it carries less risk of inadvertent lung puncture or intrapulmonary catheter placement. It is also somewhat easier to place the relatively soft catheters over a wire after tract dilation with a stiff dilator has been performed. Other practitioners favor a direct trocar technique, and this is also safe in experienced hands. There are catheters available with relatively stiff catheter tips to avoid buckling in the chest wall during placement with either technique. Local anesthesia almost always suffices for the catheter placement procedure. The use of intravenous sedation is rarely necessary but can be helpful when treating particularly anxious patients. A small skin incision and limited blunt dissection of the subcutaneous fascia are used to facilitate passage of the

K

relatively soft catheter. This limited dissection is in distinct contrast to the complete dissection to the pleura necessary for surgical thoracostomy tube placement and helps to account for improved patient comfort. Needle access to the pleural space usually is performed with an 18-gauge single part needle followed by careful guidewire placement, fascia1 dilation, and catheter placement. After catheter introduction, a fluid sample should always be obtained and sent to the laboratory for appropriate analysis. The catheter is then secured to the skin and more complete fluid aspiration is performed. Some authors have recommended limiting the initial aspirated volume to 1.5 L to avoid causing re-expansion pulmonary edema. This is a real consideration when evacuating a chronic pneumothorax or a large transudative effusion because the underlying lung remains relatively normal and pliable. Rapid complete re-expansion of a previously collapsed lung can lead to the development of pulmonary edema. With parapneumonic effusions, the underlying lung usually is consolidated and both the lung and pleura are less compliant than normal. This decreased compliance limits initial lung re-expansion regardless of the fluid volume drained and there is little risk of inducing pulmonary edema. In addition, the increased resistance to re-expansion of consolidated lung often leads to moderate pa-

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

Figure 9. Multiloculated parapneumonic empyema in a 33-year-old man, who presented in septic shock and respiratory failure. A, CT scan of the lower chest at the time of hospital admission shows bilateral lower lobe consolidation with multiple bilateral loculated pleural fluid collections. The presence of multiple loculations requires placement of multiple chest tubes for effective drainage. B, Chest radiograph 1 day after hospital admission shows two drainage catheters on the left and two on the right. There I8 some residual pleural fluid on each Side, but there ha8 been slgnlflcant re-expanslonOf the lungs. Both the sepsis and reeplratory compmml6e Improved drametlcally after the pleural dralnage pmcedlrre. The patient subeoquently required two additlPna1 pleural dralnage proceduresto completely evacuate all of the fluid loculations. Chest drains were in place for a total of 6 days. (From Moulton JS: Image-guideddrainage techniques. Semin Respir Infect 1459-72, 1999.)

359

360

MOULTON

tient discomfort during large-volume fluid aspiration, and this also limits the volume that can be comfortably drained at the time of catheter placement. As atelectatic lung is reexpanded, it is typical for patients to develop a limited but uncomfortable coughing reflex. After the maximum comfortable volume of fluid has been drained, immediate follow-up imaging should be performed. If the drainage was performed under fluoroscopic or ultrasound guidance this is usually a repeat chest radiograph. If the drainage was performed with CT then limited repeat scans are performed. This is an additional benefit of CT guidance in that the follow-up scans are the most accurate method of assessing the size and location of any residual fluid collections in relation to the catheter. It is crucial that if significant residual fluid or separate undrained loculations remain then one or more additional catheters should be placed at this time (Fig. 10). Separate loculations have to be drained at some point and it is both easier on the patient and more cost-effective to do this at the time of the initial drainage. More aggressive initial catheter drainage often obviates the need for repeat drainage procedures at a later date.

Postdrainage Catheter Management Following the initial drainage procedure, the catheter should be placed to 20-cm water suction by way of a closed underwater seal system and daily tube output should be recorded. Catheter irrigation should be performed one to four times per day with sterile saline in an attempt to prevent catheter plugging with fibrinous debris. Incentive spirometry should be encouraged throughout the drainage procedure. A daily chest radiograph should be performed to assess the progress of drainage and to monitor the position of the drain. Although a plain chest radiograph is helpful in assessing whether or not a catheter has become partially or completely dislodged, it is less helpful in determining the amount of residual pleural fluid. There is usually quite a bit of residual parenchymal lung consolidation, which can obscure pockets of fluid particularly if they are directly posterior, directly anterior, or low in a costophrenic sulcus. Repeat limited CT scans are more valuable and may be performed at 2- to 3-day intervals or when indicated by the chest radiograph findings. Follow-up CT scans are evaluated

for tube position, the size and location of residual or new fluid loculations, and the status of the underlying lung parenchyma. Although there may be some concern about the cost of frequent repeat CT scans, this is greatly outweighed by their value in assessing progress and planning subsequent patient management. If a follow-up chest radiograph shows that the catheter is in good position and tube output continues, repeat CT scans are not necessary. If tube position is in question, if residual fluid is suspected, or if tube output has stopped or dramatically decreased, however, then a repeat CT scan is needed. Most hospitals are willing to assign a much lower charge to a limited noncontrast follow-up scan in order to maximize the costeffectiveness of this approach. Several management options are available if undrained loculations are identified on follow-up CT scans. Tubes may be manipulated into a more appropriate position under either CT or fluoroscopic guidance. CT is not an ideal guidance choice for tube manipulation because it does not allow real-time continuous monitoring of catheter position. Fluoroscopic guidance with the use of directional guidewires and catheters may be performed but this also has limitations. Such manipulation tends to be very uncomfortable for the patient and usually requires intravenous sedation. It is also technically difficult to manipulate a catheter into a separate loculation because of the physical resistance offered by intervening pleural adhesions. A more effective overall approach to separate or residual loculations is to place one or more new catheters into these loculations using the same cross-sectional guidance techniques as for the original drainage. If no fluid remains around the original catheter it can be removed at the same time. The single most important rule that must be followed is that the catheter must always be in good position in relation to the residual fluid for successful drainage. Intrapleural contrast injections have been advocated to identify separate loculations and to assess for the presence of a BPF. We have found that such injections are considerably less accurate than CT scans in assessing loculations. The presence or absence of an air leak is the most accurate way to identify a BPF. If contrast injections are performed then they should be done using iso-osmolar nonionic contrast to minimize the risk of pulmonary edema should a BPF be present. If tube output is minimal and follow-up

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

Figure 10. Stage 2 parapneumonic empyema in a 46-year-old man. A, CT scan on the day of hospital admission shows a posterolateral pleural fluid collection, with adjacent lung consolidation. There is an adhesion between the visceral and parietal pleura that divides the empyema into two separate loculations. A single posterior 12F chest drain was placed into the more medial of the two loculations. 6,CT scan performed after 3 days of drainage with adjunctive pleural urokinase therapy. The more medial of the two loculations has been completely evacuated. The more lateral of the two loculations was not effectively drained because the catheter was not in an appropriate position. The intervening pleural adhesion was not affected by fibrinolytic therapy. This situation required placement of a second drain into the lateral loculation. The second drainage procedure would not have been necessary if two catheters had been placed at the time of the initial drainage. (From Moulton JS: Image-guided drainage techniques. Semin Respir Infect 1459-72, 1999.)

361

362

MOULTON

imaging shows residual fluid with the catheter in good position then it can be assumed thatlluid viscosity or tube occlusion is preventing adequate drainage. In this situation there are two management options: (1) exchange for a larger patent catheter or (2) irrigation with intrapleural fibrinolytic agents. Exchange for a larger tube may be indicated if the original drain was 8F to 10F catheter in size. The exchange is relatively easy to perform over a guidewire with either crosssectional or fluoroscopic guidance. If the original tube was 12F catheter or larger in size then the chances of a slightly larger lumen being significantly more effective are low. In this situation, the use of intrapleural fibrinolytic agents is most helpful. Management of a BPF can be somewhat problematic. The diagnosis of a BPF is made by the demonstration of an air leak coming through a pleural catheter placed to suction. If an air leak is detected, the first step should be to inspect the integrity of all catheter and tube connections carefully and to verify the position of all catheter sideholes to be sure that the leak is not caused by a communication between the suction system and room air. If the leak is determined to be coming from the pleural space then the suction should be decreased or discontinued. Keeping a tube to suction in the presence of a BPF tends to keep the pleural rent open. Decreasing or discontinuing suction decreases the amount of air passing through the fistula while preventing the development of a tension pneumothorax. In most patients with stage 2 effusions, the fistula ultimately closes, particularly if the pleural cavity is collapsed and the visceral and parietal pleural surfaces are directly apposed. In stage 3 effusions, an inelastic pleural peel surrounds the pleural cavity. When a drain is placed to suction within such a cavity, the full negative pressure is applied directly to abnormal pleura. It has been said that nature abhors a vacuum and this situation is no exception. A pleural rent almost always develops in such a case and the resultant air leak prevents further collapse of the cavity (Fig. 11). A persistent BPF refractory to closed drainage usually requires formal surgical intervention. Procedure Termination

Three standard criteria are used to determine when to remove the chest drain. Clinical

improvement with resolution of fever and leukocytosis is desirable. If the pleural space is adequately drained and an alternate source of fever is apparent (i.e., persistent pneumonia), however, then it is not considered absolutely necessary that the white blood cell count be normal and the patient afebrile. The second criterion is that almost all of the pleural fluid must be drained by radiographic evaluation. If any significant fluid remains and the clinical response is incomplete, the drainage process must be continued. The author is more compulsive about removing all of the pleural fluid when the original drainage yielded culture-positive or grossly purulent fluid. Because of the presence of persistent parenchymal consolidation the final evaluation for the presence or absence of residual pleural fluid is best made by limited CT scanning. The third criterion is that there should be no more than 20 mL of net drain output over the 24 hours preceding tube removal. Any larger amount indicates that pleural fluid is being produced and tube removal will likely result in reaccumulation of fluid. TECHNIQUE OF INTRACAVITARY FlBRlNOLYTlC THERAPY

For many years, streptokinase was the only fibrinolytic agent available for intracavitary use. The use of streptokinase is occasionally complicated by febrile allergic reactions, particularly in patients with prior or current streptococcal infections. In addition, there can be antibody-mediated deactivation of streptokinase by antistreptococcal antibodies? This may explain why streptokinase is less effective than other fibrinolytic agents. In recent years urokinase has become the agent of choice for ICFT because it produces no antibody response. Febrile reactions to urokinase have not been reported with intracavitary use and its efficacy is greater than that of streptokinase, partly because there is no antibodymediated deactivation. The continued availability of urokinase has been a problem over the past several years because of potential problems with viral contamination. Recombinant tissue plasminogen activator has now become our agent of choice for ICFT. Like urokinase, tissue plasminogen activator is not subject to antibody-mediated deactivation. There are limited data in the medical literature regarding the intracavitary use of tissue

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

Figure 11. A 48-year-old woman who developed a chronic stage 3 hemothorax after central line placement. A, CT scan obtained 3 months after the original hemothorax shows a large loculated pleural fluid collection. There is only mild, visible thickening of the parietal pleura. The fluid collection was treated by guided placement of a single 12F chest tube. 6,CT scan after 6 days of drainage shows a loculated hydropneumothorax cavity identical in shape to the original, fluid-filled cavity. An air leak was present and there was no re-expansion of the adjacent atelectatic lung. The inability to collapse the cavity and expand the adjacent lung was caused by the presence of a chronic fibrous pleural peel. The patient subsequently underwent a successful surgical decortication. The presence of a fixed pneumothorax cavity after tube drainage of a loculated effusion is the most reliable indication that a stage 3 pleural peel is present and that further nonoperative treatment likely will fail.

363

364

MOULTON

plasminogen activator, but in our experience its efficacy appears to be comparable to that of urokinase and allergic reactions have not been encountered. On a per-dose basis tissue plasminogen activator is actually less expensive than urokinase and its use is more cost effective. ICFT should be instituted if there is persistent loculated pleural fluid despite good tube position. There are no hard and fast rules when to initiate ICFT in the course of pleural drainage. A trial of several days of drainage without ICFT may be warranted, particularly if there is steady clinical improvement and tube fluid output continues. If tube fluid output is low or stops despite the presence of residual fluid then a limited CT should be performed to evaluate tube position. If tube position is satisfactory then ICFT should be initiated. It often is advisable to begin ICFT immediately after the initial drainage procedure. As experience is gained with the technique it is often possible to predict which cases will not be drained completely without ICFT. Fluid with a high protein content usually forms fibrin nets that inhibit effective drainage. It is not necessary to wait for laboratory study results on the fluid to predict this occurrence because the fibrin debris may form visible clots in the syringe within minutes of fluid removal. There are also cases in which only a small amount of fluid can be drained during the initial procedure despite adequate tube position. In this situation it can be assumed that fibrin nets are present and are occluding the side holes of the catheter. Instituting ICFT earlier in the course of treatment may help prevent the formation of fluid loculations and can minimize the total duration of catheter drainage in days." l5 Intracavitary fibrinolytic therapy is relatively simple to perform. If urokinase is used, a total of 250,000 units is dissolved in 250 mL of normal saline and divided into 80 mL aliquots. Our standard dose for tissue plasminogen activator is 6 mg dissolved in 50 mL of normal saline. Solutions containing dextrose should not be used with tissue plasminogen activator because this can lead to drug precipitation. After instillation of 50 to 80 mL of the solution the chest drain is clamped for 2 to 4 hours. Fluid is then aspirated and the net output is recorded. The chest tube may then be placed back to suction or a second fibrinolytic instillation performed. From one to four instillations may be performed per day as time permits. A larger volume of fi-

brinolytic solution may be instilled for the treatment of large loculated effusions (>500 mL estimated volume). Multiple tubes may also be treated at the same time. For small pleural fluid loculations a smaller aliquot of fibrinolytic solution may be instilled. This procedure is repeated until all of the fluid has been drained. Hemorrhagic complications of ICFT are extremely rare. Several studies have shown that there is no systemic fibrinolytic effect after intracavitary instillation of fibrinolytic agents." When treating post-traumatic hemothoraces with ICFT it is common practice to wait 2 to 3 days after the initial trauma to begin ICFT.15,l7 This delay is instituted to minimize the theoretic risk that fibrinolytic agents cause active bleeding to recur. Active bleeding into the pleural space is considered an absolute contraindication to ICFT. The presence of a BPF is a relative contraindication to ICFT to prevent intrapulmonary lavage. It is possible for ICFT to open a recently closed BPF if the pleural hole is sealed by fibrinous debris. If this occurs then ICFT is withheld and the fistula usually closes again over a short period of time. It is common for the aspirated fluid to become grossly bloody toward the end of a course of ICFT. This is such a common occurrence that the presence of a bloody aspirate can be used as a sign that very little pleural fluid remains. The cause of this phenomenon is probably minimal bleeding from the inflamed pleural surfaces. It does not represent a significant amount of blood loss and other than indicating that most or all of the pleural fluid is gone it does not seem to have any clinical implications. SHORT-TERM RESULTS OF IMAGEGUIDED DRAINAGE

As mentioned previously it is difficult to compare directly the results of clinical trials that have studied the different treatment modalities available for empyema drainage. Selection bias and technique of drainage have such a profound impact on outcome that the use of historical controls is of limited use and only the most general comparisons can be made. Most series describing the results of nonguided placement of large-bore thoracostomy tubes report less than 50% success in obviating open surgical drainage.l>6, 9, 22 The majority of failures in these series were in

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

patients with multiloculated stage 2 empyemas or stage 3 pleural peels. Studies on the use of image-guided placement of smallercaliber chest tubes have reported success rates of 67% to 80% with the duration of chest tube drainage ranging from 7 to 26 days.16, 24 Direct comparison of these numbers with those for nonguided thoracostomy is misleading because many of the patients treated in the image-guided thoracostomy series had already failed treatment with large-caliber nonguided thoracostomy tubes. The resultant selection bias tends to underestimate the value of image-guided thoracostomy. Failures in the image-guided thoracostomy series were also in multiloculated stage 2 and stage 3 empyemas. Series that studied the use of ICFT in concert with nonguided thoracostomy have reported success rates of 44% to 93%.537, 14, l9 When used in concert with image-guided thoracostomy the reported success rates of ICFT lo,15,l7 Again, direct range from 63% to comparison of success rates between thoracostomy alone and thoracostomy with ICFT is misleading because most patients undergoing ICFT in these series had already failed closed thoracostomy alone. Technique and aggressiveness of drainage certainly have a profound impact on the ultimate outcome. Given proper technique, however, the efficacy of any form of therapy is most closely related to the age (and thus the stage) of the effusion at the time of drainage. Mitchell et all4reported a series in which they used ICFT by nonguided thoracostomy tubes with an overall 56% success rate. The mean effusion age was 11 days for successful cases and 6 weeks for those that ultimately required surgical decortication. Fraedrich et al,5 using similar technique, also found that closed drainage and ICFT were unsuccessful in obviating surgery when instituted during the third or organizing phase of pleural peel formation. Pollack and Passik17 used ICFT by the way of image-guided thoracostomy tubes with an overall success rate of 63%. The failures were all in patients in which the effusion was greater than 6 weeks in age at the time of thoracostomy. In the largest single series to date Moulton et all5used ICFT by the way of guided thoracostomy tubes in 118 patients and reported poor success in obviating surgery when drainage was instituted more than 6 weeks after the onset of the effusion. When drainage was instituted earlier in stage 2 disease the success rate was 96%. In a smaller

365

series by Lee et allothe success rate of guided thoracostomy and ICFT was 100% in stage 2 empyemas. These findings correlate well with the expected pathologic evolution of complicated pleural fluid collections. In stage 2 the pleural pathologic process is one of nonorganized adherent fibrinous debris, inflammatory thickening, and edema. The affected pleura has impaired elasticity causing a restrictive respiratory defect. One of the imaging manifestations of decreased pleural elasticity is peripheral lung ate1e~tasis.l~ The implication of this is that lung trapping can occur in the fibrinopurulent phase of evolution (Fig. 12). If the infected fluid and most of the debris are removed the remaining debris is resorbed and the inflammatory pleural reaction resolves allowing re-expansion of the lung.15,l9 The ability to collapse the pleural cavity completely is the best indicator that the process is still in stage 2 and that a fibrous pleural peel has not yet developed. The amount of visible pleural thickening is not a useful predictor of either stage or of the likelihood of successful thoracostomy drainage (Fig. 13). In stage 3 the pleural pathologic process is one of chronic inflammation and mature fibrous thickening that is associated with peripheral lung trapping and atelectasis (Fig. 14). The term pleural peel should be reserved for the chronic fibrous form of pleural thickening and should not be used to describe a stage 2 fibrinopurulent exudate with inflammatory pleural thickening, even if accompanied by lung trapping. The implications are not purely semantic in that the true fibrous thickening of the pleura in stage 3 is not affected by thoracostomy or fibrinolytic treatment and usually requires surgical decortication for effective treatment. It is crucial that closed drainage be instituted early in the course of the disease (before stage 3 ) to be successful. Although there are differing opinions among authors it seems that an effusion age of approximately 4 to 6 weeks is the cutoff between stage 2 disease (in which aggressive closed thoracostomy and possibly ICFT should be effective) and stage 3 disease (which may require dec~rtication).'~, 16, 23 LONG-TERM OUTCOME OF IMAGEGUIDED DRAINAGE

The long-term outcome of closed thoracostomy is similar to the immediate success

366

MOULTON

Figure 12. A 50-year-old woman with a stage 2 postoperative empyema 7 days after right middle lobectomy for necrotizing pneumonia. A rim of pleural fluid is visible just central to the chest wall and parietal pleura. The enhancing parietal pleural is only slightly thickened. Central to the fluid is an enhancing crescentic soft tissue density representing atelectatic peripheral lung (arrows). It is difficult to visualize the visceral pleural separate from the atelectatic lung. The rim of atelectatic subpleural lung is trapped by the adjacent noncompliant visceral pleura. This peripheral atelectasis completely resolved over 2 months after drainage of the empyema. (From Moulton JS: Image-guided drainage techniques. Semin Respir Infect 14:59-72, 1999.)

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

Figure 13. Stage 2 parapneumonic empyema in an 18-year-old woman. A, CT scan obtained 1 day after hospital admission shows a collapsed right lower lobe with a large associated pleural effusion. No significant parietal pleural thickening is visible. This was treated by guided placement of two 12F chest tubes followed by ICFT. 13,CT scan 13 days after A shows near complete drainage of the empyema. There is extensive residual lung consolidation, with some volume loss. More marked parietal pleural thickening is now visible posteriorly. Illustration continued on following page

367

368

MOULTON

Figure 13 (Continued). C, CT scan obtained 2 months after discharge is normal. The pleural thickening seen in 13 was inflammatory rather than fibrotic in nature. Stage 2 inflammatory changes will resolve over several months following adequate pleural drainage.

a

rates. After successful draina e of the pleural space in stage 2 empyemas t ere is typically extensive residual pulmonary consolidation, peripheral lung atelectasis, and some residual inflammatory leural thickening. Routine chest radiograp y and CT scanning immediately after drain removal are far from normal. If the pleural space is effectively drained and the visceral and parietal pleural surfaces apposed, however, the residual pleural and parenchymal changes resolve over the subsequent 2 to 4 months (Fig. 15).7,15, 16, l9 Pulmonary function tests at that time may also be expected to return to normal or 23 Residual pleural and parenchyba~eline.~, mal changes during this recovery period should not be seen as evidence that decortication should have been performed. In stage 3 empyemas, the short-term and long-term success of closed drainage are very low. Closed drainage, whether guided or nonguided, is usually ineffective in collapsing the empyema cavity and is frequently complicated by the development of a BPF (Fig. 16). Even in those patients with stage 3 disease in whom pleural sepsis is controlled and the cavity collapsed by closed drainage there is some residual chronic pleural thickening and adjacent lung trapping. This does not imply that surgery is necessary for all stage 3 pleural peels. The decision to operate is based on the extent of the pleural peel and the general

K

health of the patient. In patients with focal pleural peels and multiple underlying medical problems who are relatively poor surgical candidates it may be preferable to be left with some degree of chronic atelectasis rather than be exposed to the risk of thoracotomy and 24 This argument assumes that decorticati~n.~~, pleural sepsis has been controlled by closed drainage and that respiratory compromise is minimal. The inability to control pleural sepsis is an indication for early open drainage, although we have found this to be a rare situation. Extensive peels that cause a greater compromise of pulmonary function, or peels surrounding a pleural cavity that cannot be collapsed, are better treated with decortication. Other potential indications for surgical intervention include central bronchial obstruction and necrotizing pneumonia, because these entities are not amenable to nonsurgical management (Fig. 17).', 11, 21 SUMMARY

Percutaneous image-guided catheter drainage with adjunctive ICFT has become the mainstay in the treatment of complicated pleural fluid collections. There are six basic principles of image-guided drainage and ICFT that must be understood to maximize the efficacy and safety of the procedure. 1. There must be a basic understanding of

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

Figure 14. Chronic stage 3 parapneumonic effusion in a 50-year-old man. A, CT scan shows a small posterior loculated pleural fluid collection with adjacent atelectasis of the right lower lobe of the lung. The fluid had been present for 2 months before the scan. The loculated effusion was treated by guided placement of a 12F drainage catheter and subsequent ICFT. B, CT scan after drainage and ICFT shows that the pleural effusion has been completely drained with apposition of the visceral and parietal pleural surfaces. The cavity was able to be collapsed because the pleural process was focal and there was enough adjacent pliable lung and pleura to allow partial re-expansion. A crescentic rim of peripheral atelctasis remains in the right lower lobe. The peripheral lung remained trapped because of underlying chronic pleural fibrosis, which was unaffected by drainage and ICFT. A chest radiograph 2 years later (not shown) showed stable peripheral right lower lobe atelectasis.

369

370

MOULTON

Figure 15. Stage 2 parapneumonic empyema in a 19-year-old woman. A, CT scan obtained 1 day after hospital admission shows a densely consolidated right lung, with an adjacent crescentic pleural fluid collection. A large caliber chest tube was in place but was not draining any fluid. The fluid loculation was treated by guided placement of two 12F chest tubes followed by ICFT. B, CT scan obtained after 4 days of drainage and ICFT through the 12F tubes shows near-complete drainage of the empyema, with extensive lung consolidation and more marked pleural thickening. An additional drain was placed for evacuation of the small residual posteromedial fluid loculation. Illustration continued on opposite page

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

371

Figure 15 (Continued). C, A chest radiograph obtained 3 months after discharge is normal. The residual pleural and parenchymal changes seen in 6 were inflammatory rather than fibrotic in nature. These stage 2 inflammatory changes resolve over several months after successful pleural drainage.

why traditional nonguided thoracostomy drainage fails in a significant percentage of patients. Tube malposition relative to fluid loculations, fluid debris and viscosity, and the presence of a stage 3 pleural peel are the primary reasons for failure. Image-guided placement of drains addresses the issue of tube malposition and ICFT greatly facilitates drainage of fibrinous fluid. 2. Proper use of cross-sectional imaging is one of the keys to ultimate success. CT and ultrasound allow very accurate assessment of the underlying pathologic process and are crucial in planning the drainage procedure, guiding the actual placement of drains, and following the course and outcome of treatment. The added costs of cross-sectional imaging are more than compensated by the increase in success of the drainage procedure. 3. Aggressive catheter management is the single most important factor in success. Multiple loculations require multiple catheters for adequate drainage. Pleural adhesions may form quickly as drainage progresses leading to the formation of undrained loculations. Frequent crosssectional imaging is needed to detect undrained loculations so that additional drainage catheters may be placed if needed. It is crucial that the drainage

catheter always be properly positioned in relation to fluid loculations. 4. Intracavitary fibrinolytic therapy is a very powerful adjunctive therapy to aid in complete evacuation of fluid collections that contain fibrin nets and debris. It can also partially debride the pleural surfaces of fibrinous debris and facilitate complete re-expansion of the underlying lung. Intracavitary fibrinolytic therapy should not be used in an attempt to salvage success by a malpositioned chest tube. 5. The ultimate success of closed drainage for complicated pleural fluid collections is closely related to the age of the effusion at the time of drainage. A very high rate of clinical success may be expected when these techniques are used in the treatment of stage 2 fibrinopurulent effusions. If drainage is delayed until the third stage (fibrous pleural peel formation) then closed drainage likely will fail and a formal thoracotomy and decortication will be necessary. Experience in the literature suggests that effusions up to 4 to 6 weeks in duration may be drained successfully but those older than 6 weeks likely will have an associated pleural peel. Effective pleural drainage must be instituted early in the course of the disease process. 6. There may be significant residual pleural

372

MOULTON

Figure 16. Stage 3 complicated parapneumonic effusion in a 68-year-old man. A, CT scan shows a loculated fluid collection, with adjacent pleural thickening and lung atelectasis. The effusion had been present for 6 months at the time of this CT scan. Marked parietal pleural thickening is visible. B, CT scan obtained 11 days after guided placement of a 12F chest tube shows a persistent thickwalled pleural cavity. Much of the fluid has been evacuated and replaced by a loculated pneumothorax. There has been partial re-expansion of the left lower lobe, but a dense band of peripheral atelectasis remains. The patient subsequently underwent a decortication that showed a thick fibrous pleural peel, which prevented re-expansion of the lung. Fibrotic pleural changes in stage 3 will not resolve with thoracostomy drainage alone.

IMAGE-GUIDED MANAGEMENT OF COMPLICATED PLEURAL FLUID COLLECTIONS

Figure 17. A 56-year-old man with a necrotizing left lower lobe pneumonia. Lung window (A) and mediastinal window (B) are shown from the same CT scan. The left lower lobe is consolidated and replaced by multiple necrotic air-filled cavities. There is a small adjacent empyema (arrow) containing some air, which indicates that a bronchopleuralfistula is present. Surgery is necessary for this patient because the necrotic lower lobe will not heal with antibiotics and pleural drainage. The empyema was treated with guided drainage to control pleural sepsis and the patient subsequently underwent a successful decortication and left lower lobectomy.

373

374

MOULTON

and parenchymal inflammatory changes after complete drainage of a stage 2 effusion. If the fluid in the pleural space has been adequately drained and the visceral and parietal pleural surfaces apposed, then the residual inflammatory pleural thickening and associated lung consolidation resolve over 2 to 4 months and pulmonary function returns to baseline. Imaging studies immediately after complete pleural drainage are not normal. These residual abnormalities should not be interpreted as evidence that open surgical drainage should have been performed. Effective closed drainage carries lower morbidity, mortality, and cost than does open surgical drainage. For radiologists and clinicians alike it does not suffice simply to place one or more thoracostomy tubes, round daily, and hope that the occasional use of fibrinolytic agents does the rest. Without a more aggressive approach to catheter position and management the efficacy is no greater than that historically seen with nonguided closed drainage and surgeons will continue to plead for earlier effective open drainage. References 1. Ali I, Unruh H: Management of empyema thoracis. Ann Thorac Surg 50355359, 1990 2. Berglin E, Ekroth R, Teger-Nilsson AC, et al: Intrapleural instillation of streptokinase: Effects on systemic fibrinolysis. Thorac Cardiovasc Surg 29:124126, 1981 3. Bouros NP, Schiza S, Panagou P, et al: Role of streptokinase in the treatment of acute loculated parapneumonic effusions and empyema. Thorax 49:852-855, 1994 4. Davies RJO, Trail1 ZC, Gleeson FV: Randomized controlled trial of intrapleural streptokinase in community acquired pleural infection. Thorax 52416-421, 1997 5. Fraedrich G, Hofmann P, Effenhauser P, et al: Instillation of fibrinolytic enzymes in the treatment of pleural empyema. Thorac Cardiovasc Surg 30:36-38, 1982 6. Grant DR, Finley RJ: Empyema: Analysis of treatment techniques. Can J Surg 28:449451,1985 7. Jerjes-Sanchez C, Ramirez-Rivera A, Elizalde JJ, et al: Intrapleural fibrinolysis with streptokinase as an adjunctive treatment in hemothorax and empyema: A multicenter trial. Chest 109:1514-1519, 1996

8. LaHorra JM, Haaga JR, Stellato T, et al: Safety of intracavitary urokinase with percutaneous abscess drainage. AJR Am J Roentgenol 160:171-174, 1993 9. LeBlanc KA, Tucker W Empyema of the thorax. Surg Gynecol Obstet 158:66-70, 1984 10. Lee KS, Im J-G, Kim YH, et al: Treatment of thoracic multiloculated empyema with intracavitary urokinase: A prospective study. Radiology 179:771-775, 1991 11. Lemmer JH, Botham MJ, Orringer MB: Modem management of adult thoracic empyema. J Thorac Cardiovasc Surg 90S49-855, 1985 12. Light RW Parapneumonic effusions and empyema. Clin Chest Med 6:55-62, 1985 13. Merriam MA, Cronan JJ, Dorfman GS, et al: Radiographically guided percutaneous catheter drainage of pleural fluid collections. AJR Am J Roentgenol 151:1113-1116, 1988 14. Mitchell ME, Alberts WM, Chandler KW, et al: Intrapleural streptokinase in management of parapneumonic effusions. J Fla Med Assoc 76:1019-1022, 1989 15. Moulton JS, Benkert RE, Weisiger KH, et al: Treatment of complicated pleural fluid collections with image-guided drainage and intracavitary urokinase. Chest 108:1252-1259, 1995 16. Neff CC, vansonnenberg E, Lawson DW, et al: CT follow-up of empyemas: Pleural peels resolve after percutanious c a t h e r drainage. Radiology 1 7 6 1 9 5 197, 1990 17. Pollack JS, Passik CS: Intrapleural urokinase in the treatment of loculated pleural effusions. Chest 105:868-873, 1994 18. Pothula V, Krellenstein DJ: Early aggressive surgical management of parapneumonic empyemas. Chest 105:832-836, 1994 19. Robinson LA, Moulton AL, Flemming WH, et al: Intrapleural fibrinolytic treatment of multiloculated thoracic empyemas. Ann Thorac Surg 57:803-814, 1994 20. Silverman SG, Mueller PR, Saini S, et al: Thoracic empyema: Management with image-guided catheter drainage. Radiology 1 6 9 5 9 , 1988 21. Smith JA, Mullerworth MH, Westlake GW, et al: Empyema thoracis: 14-year experience in a teaching center. Ann Thorac Surg 5L39-42, 1991 22. Storm HKR, Krasnik M, Bang K, et al: Treatment of pleural empyema secondary to pneumonia: Thoracentesis regimen vs tube drainage. Thorax 47821824, 1992 23. Toomes H, Vogt-Moykopf I, Ahrendt J: Decortication of the lung. Thorac Cardiovasc Surg 31:33%341, 1983 24. vanSonnenberg E, Nakamoto SK, Mueller PR, et al: CT- and ultrasound-guided catheter drainage of empyemas after chest tube failure. Radiology 151:349353, 1984 25. Vianna NJ: Nontuberculous bacterial empyema in patients with and without underlying disease. JAMA 215~69-75, 1971 26. Wehr CJ, Adkins RB: Empyema thoracis: A ten year experience. South Med J 79:171-176, 1986 27. Westcott J: Percutaneous catheter drainage of pleural effusion and empyema. AJR Am J Roentgenol 144:1189-1193, 1985

Address reprint requests to Jeffrey S. Moulton, MD Department of Radiology St. Anthony Central Hospital 4231 West 16th Avenue Denver, CO 80204