Intrapleural fibrinolytic treatment of multiloculated thoracic empyemas

Intrapleural Fibrinolytic Treatment of Multiloculated Thoracic Empyemas Lary A. Robinson, MD, Anthony L. Moulton, MD, William H. Fleming, MD, Anselmo Alonso, MD, and Timothy A. Galbraith, MD Section of Thoracic and Cardiovascular Surgery, University of Nebraska Medical Center, Omaha, Nebraska

Acute multiloculated thoracic empyemas incompletely drained by tube thoracostomy alone usually require operation. To avoid a thoracotomy yet treat this difficult problem, intrapleural fibrinolytic agents were employed. Between April 1,1990, and April 1,1993,13 consecutive patients presenting with a fibrinopurulent empyema were demonstrated to have incomplete drainage. To facilitate drainage, streptokinase, 250,000 units in 100 mL 0.9% saline solution (3 patients), or urokinase, 100,000 units in 100 mL 0.9% saline solution (10 patients), was instilled daily into the chest tube, and the tube was clamped for 6 to 12 hours followed by suction. This routine was continued daily for a mean of 6.8 2 3.7 days (range, 1 to 14 days) until resolution of the pleural fluid collection was demonstrated by computed chest tomography and clinical indications. This regimen was completely successful in 10 of 13 patients (77%), who had

resolution of the empyema, eventual withdrawal of chest tubes, and no recurrence. Two patients, both pediatric liver transplant patients, had an initial good response but eventually required decortication. One patient with a good radiographic response became increasingly febrile during streptokinase therapy and underwent a thoracotomy, but no significant undrained fluid was found. This patient’s continued fever was believed to be a streptokinase reaction. Urokinase was used subsequently. No treatment-related mortalities or complications occurred. Intrapleural fibrinolytic agents, especially urokinase, are safe, cost-effective means of facilitating complete chest tube drainage, thereby avoiding the morbidity of a major thoracotomy for 77% of a group of multiloculated empyema patients who traditionally would have required open surgical therapy. (Ann Thorac Surg 1994;57:803-14)

When empyema is treated either by the cautery or incision, if pure and white pus flow from the wound, the patients recover; but if mixed with blood, slimy and fetid, they die. Hippocrates [l]

decortication, extrathoracic muscle transposition, and even thoracoplasty [&8]. Early thoracotomy and decortication are currently advocated for multiloculated acute empyemas by most thoracic surgeons and more than half of pulmonologists recently surveyed [9]. As thoracic surgeons, we generally followed this traditional management until recently confronted with a massively obese diabetic with the Pickwickian syndrome presenting with a postpneumonic acute multiloculated empyema, who remained toxic after incomplete pleural drainage with a chest tube. Thoracotomy and decortication, which would generally be advocated, was thought to carry almost prohibitive risks for morbidity and mortality in this patient. We were then introduced by our pulmonary medicine colleagues to the use of intrapleural fibrinolytic agents to facilitate complete empyema drainage. This nonsurgical approach is an alternative that is rarely discussed in the recent thoracic surgical literature, and has only a cursory reference in the major thoracic surgical textbooks [lo, 111. After the development of streptokinase in 1944, the first clinical application of this drug was by Tillett and Sherry in 1949 using intrapleural instillation of this agent for acute fibrinous pleurisy, bacterial empyema, and hemothorax [12]. Over the years, many others have reported on the utility of intrapleural streptokinase in facilitating acute empyema drainage by a chest tube, thereby avoiding a major thoracotomy [13-181.

M

ore than two millennia have passed since Hippocrates [l] described drainage of pleural empyemas by making an opening in the chest at the tenth rib and inserting a piece of tent as a drain. Although some type of surgical drainage has been advocated by various authors over the years, the modern day management of this problem is based on the principles developed by Evarts Graham and the World War I Empyema Commission. Their change in therapy to routine, closed intercostal drainage resulted in a dramatic fall in mortality from 75% to 15% in streptococcal empyemas [2]. Despite prompt intercostal tube insertion as well as aggressive antibiotic therapy, some patients fail to recover with simple drainage alone, as well recognized long ago by Hippocrates [l].Most have progression of the acute empyema to the fibrinopurulent stage with multiple loculations and viscous pus [3]. Aggressive surgical therapy has been developed, ranging from rib resection and open flap drainage to thoracotomy with empyemectomy, Presented at the Fortieth Annual Meeting of the Southern Thoracic Surgical Association, Panama City Beach, FL, Nov 4 6 , 1993. Address reprint requests to Dr Robinson, Thoracic and Cardiovascular Surgery, University of Nebraska Medical Center, 600 South 42nd St, Omaha, NE 68198-2315. 0 1994 by The Society of Thoracic

Surgeons

0003-4975/94/$7.00

804

ROBINSONET AL FIBRINOLYTIC TREATMENT OF EMPYEMAS

Urokinase, the other major fibrinolytic agent that is nonantigenic and nonpyrogenic, was first used in 1987 by Vogelzang and associates to treat infected extravascular hematomas [19]. However, urokinase was later used by Moulton and associates [20], Lee and colleagues [21], and most recently Gerardi and associates [22] successfully to facilitate drainage of loculated pleural empyemas. Despite its popularity among surgeons in the past, no reports on the use of fibrinolytic agents for acute empyemas have appeared in the US thoracic surgery literature for the last 22 years. Studies advocating this method of treatment are found primarily in the medical and radiology literature. It is our purpose to review the literature and to report to thoracic surgeons on our experience using intrapleural fibrinolytic agents in a consecutive group of patients with multiloculated acute pleural empyemas who traditionally would have been candidates for thoracotomy and decortication.

Material and Methods

Ann Thorac Surg 1994;57803-14

effusion seen on the chest radiogram or chest CT scan prior to invasive chest procedures.

Exclusion Criteria Patients excluded were those with empyemas after resectional operations or open heart procedures or those with longstanding chronic draining empyemas with a bronchopleural fistula. Also excluded were patients with obvious stage 111 chronic organizing empyemas [24] with a thick pleural peel demonstrated by chest CT scan.

Chest Drainage Catheter Local anesthesia with 1%lidocaine was used for all adults during tube thoracostomy. Light general anesthesia was used on some pediatric patients for chest tube insertion. Sizes of chest tubes inserted were 28F to 32F for adults and 20F to 24F in the pediatric group. All tubes were maintained on 20 cm H,O suction. Drainage from chest tubes was recorded every 8 hours, although there was drainage around the clamped tube in some patients.

Patient Population

Fibrinolytic Agent Instillation

From April 1, 1990, to April 1, 1993, 13 consecutive patients with multiloculated acute empyemas were studied by retrospective review of their hospital records and radiographic studies. All patients had standard posteroanterior and lateral chest radiograms in addition to decubitus views. Computed chest tomographic (CT)scans were also performed on all patients during the early stages of therapy, and many had follow-up CT scans upon completion of therapy. At least one thoracentesis was performed in all patients before therapy, and most had complete pleural fluid analysis including Gram stain and cultures. After the diagnosis of pleural empyema was made, appropriate antibiotics and chest catheter drainage were implemented. In general, the decision to insert a chest tube followed the recommendations of Light [23]. We tended to be aggressive with early insertion of chest drainage catheters, especially in patients with borderline positive pleural fluid analysis in whom there was radiographic evidence of fluid loculations.

Patients were placed in the lateral decubitus position with the unaffected lung dependent during agent instillation, to be sure that all of this agent drained from the chest catheter into the treated pleural cavity. No premedication or analgesics were administered systemically or intrapleurally. The drainage catheter was clamped and streptokinase, 250,000 units in 100 mL of 0.9% saline solution, was instilled into the tube of 3 patients using a syringe and a 21-gauge needle. For the last 10 patients, urokinase, 100,000 units in 100 mL 0.9% saline solution, was instilled intrapleurally. Tubes remained clamped for a minimum of 6 hours, with a maximum of 12 hours. After unclamping, tubes were placed back on suction and drainage was recorded. Fibrinolytic agents were generally instilled in the evening. Patients remained in bed until the tube was unclamped, to minimize the amount of agent that might leak out around the tube thereby decreasing its effective dwell time in the pleural cavity. The number of consecutive days the agents were used depended on the clinical and radiographic response, although usually 5 to 10 days of daily treatments were recommended.

Inclusion Criteria Using the empyema staging characteristics defined by the Subcommittee on Surgery of the American Thoracic Society [24], only stage I acute exudative or stage I1 fibrinopurulent empyemas were included. Other criteria required for inclusion in this study were (1) exudative pleural fluid accompanying a pneumonia or subdiaphragmatic inflammation; (2) pleural fluid with at least one of the following: pH < 7.20, glucose < 40 mg/dL, lactate dehydrogenase > 1000 IUL, total protein > 5.0 g/dL, grossly purulent appearance or white blood cell count greater than 20,00O/pL, positive Gram stain or growth of bacteria on culture; (3) clinical toxicity with fever or leukocytosis; (4) multiloculation of fluid demonstrated on chest CT scans before or after chest catheter insertion; and (5) continued clinical toxicity after chest catheter insertion and radiographic evidence of incomplete empyema drainage. Multiloculation of pleural fluid is defined as obvious septations seen on chest CT scan or air fluid levels in the

Assessment of Response During fibrinolytic therapy, patients were assessed for a favorable response as demonstrated by (1) resolution of fever and systemic symptoms, (2) resolution of respiratory symptoms, (3) increased chest tube drainage, and (4) evidence of radiologic improvement by standard chest radiogram with or without repeat chest CT scan. An unfavorable response to fibrinolytic therapy was determined when there was (1) continued or worsening fever and systemic toxicity, and (2) no improvement or a worsening radiographic picture manifested by increasing pleural fluid or the new appearance of air-fluid levels. Clinical judgment by the attending staff surgeon dictated whether to continue fibrinolytic therapy or to proceed to thoracotomy and decortication. In patients successfully treated with fibrinolytic agents,

Ann Thorac Surg 1994;57803-14

chest tubes were removed acutely during the hospitalization in those with stage I exudative, culture-negative empyemas (complicated parapneumonic effusions [23]) after drainage was less than 50 mL in 8 hours. After a positive response with fibrinolytic agents and systemic antibiotics, chest tubes inserted for stage I1 fibrinopurulent empyemas with culture-positive, gross pus were managed long-term by conservative conversion of the chest tube to open drainage and gradual tube withdrawal. Patients were assessed carefully for evidence of drug complications including pain, allergic reaction, or hemodynamic changes. Fever was defined as a temperature greater than 38"C, or one degree elevation over baseline in previously febrile patients. Patients were also observed for evidence of bleeding. Many patients had various coagulation studies performed for clinical reasons unrelated to intrapleural fibrinolysis. However, no standardized coagulation monitoring was performed. All data summaries collected from this retrospective study are expressed as the mean t standard deviation.

Results Patient Demographics The patient population ranged in age from 6 to 87 years with a mean age of 40.9 ? 24.1 years (median, 42 years). There were 10 male and 3 female patients. Two of the patients were in the pediatric age group (6 and 9 years old, both boys).

Clinical Presentation The clinical characteristics and results of therapy are summarized in Table 1. The causes of the empyemas were postpneumonic in 9, contaminated sympathetic pleural effusions in 3, and postpneumonic with infected hemothorax in 1. One or more coexisting conditions were present in all patients, most commonly chronic obstructive lung disease in 5, severe obesity in 4, heavy ethanol intake in 5, and nonalcoholic liver disease in 4. Three patients were liver transplant recipients. Fever was present in all patients except two taking systemic steroids. Twelve patients had a leukocytosis; one liver transplant patient had leukopenia (<1,000 white blood cells/pL). All postpneumonic empyema patients presented with cough, purulent sputum production, and dyspnea.

Radiographic Findings The findings on chest radiograms were relatively similar, with large pleural effusions, atelectasis of the underlying lung, and for some patients, a pulmonary infiltrate. All effusions were categorized as large (reaching more than one third of chest height) based on the criteria of Himelman and Callen [3], and thereby were more likely to be loculated. Pleural air manifested by multiple air-fluid levels was seen in only 2 patients (patients 2 and 6). However, the CT scan findings in all patients were diagnostic, with each showing evidence of multiloculation, with septation and separated fluid collections. This pleural fluid loculation was verified clinically by incomplete drainage of fluid following chest tube insertion. In

ROBINSON ET AL FIBRINOLYTIC TREATMENT OF EMPYEMAS

805

addition, all patients continued to remain symptomatic and clinically toxic with only incomplete empyema drainage.

Pleural Fluid Analysis The pleural fluid chemistries and culture results from thoracenteses are shown in Table 2. Four patients did not have complete fluid analysis but cultures were obtained in all. Based on criteria described by Light [23], all effusions easily fit the category of an exudate. Based on pleural fluid analysis alone and using the classification system of the American Thoracic Society [24], 2 patients (patients 6 and 13) would be categorized as having stage I exudative empyemas (complicated parapneumonic effusions) and the other 11 as stage I1 fibrinopurulent empyemas. However, the radiographic findings of very large effusions with multiple loculations in both stage I patients and multiple air-fluid levels in one (patient 6) easily move these 2 patients into the stage I1 fibrinopurulent empyema category [24]. All patients had 3 to 14 days of antibiotic therapy prior to thoracentesis. Not surprisingly, only 2 patients (15%) had a positive Gram stain on their pleural fluid and 7 patients (54%) grew bacteria on pleural fluid culture. A variety of organisms were isolated from the empyemas. Most commonly found were gram-positive cocci but also included were gram-negative bacilli, a microbiologic spectrum similar to that described in recent prior series [6, 181.

Response to Therapy A summary of treatment and the responses to therapy is shown in Table 1. All patients initially underwent intercostal catheter drainage with a large-bore chest tube, except for patient 4, in whom an 8F pigtail catheter was used (the patient refused chest tube insertion). The time from the onset of symptoms in patients until initial intercostal catheter insertion was a mean of 17.2 -t 15.1 days (median, 12 days; range, 4 to 61 days). However, chest tube drainage plus parenteral antibiotics were inadequate treatment for all of the empyemas, as demonstrated by continued clinical toxicity and the radiographic findings of multiloculation of pleural fluid. All patients were then treated with daily intrapleural fibrinolytic therapy for a mean of 6.8 t 3.7 days (range, 1 to 14 days). Of the 3 patients treated with streptokinase, 2 (67%) responded completely with resolution of clinical and radiographic abnormalities. Of the 10 patients treated with urokinase, 8 (80%) responded completely and did not require operation. The number of days patients were treated with fibrinolytic agents varied and was determined by the attending staff surgeon based on the clinical picture, amount of chest tube drainage and the chest radiographs. Two patients required only a single day of therapy to elicit complete empyema drainage and radiographic clearing. The mean daily chest tube drainage after beginning intrapleural fibrinolysis varied greatly, as shown in Table 1, from a high of 600.5 2 290.5 mUday (patient 5) to a low of 107.0 -1- 36.0 mL/day (patient 13). The lower values of total daily drainage in some patients are misleading because there was some leakage of pleural fluid around the

806

ROBINSON ET AL FIBRINOLYTIC TREATMENT OF EMPYEMAS

Ann Thorac Surg 1994;57803-14

Table 1. Summary of Clinical Presentation and Therapy

Patient Age No. (y) Sex 1

36

2

19

3

4 2

4

4 4

5

4 4

6

34

7

87

8

58

9

72

10

6

11

9

12

60

13

21

Medical Problems

Radiographic Findings

Presentation

Severe obesity, diabetes mellitus, 5 d dyspnea, fever, chest RLL infiltrate, loc hypertension, sleep apnea, pain, bloody sputum effusion smoker M Liver transplant; thoracotomy 4 Right chest 10 d fever, dyspnea, density with weeks earlier to control bleeding productive cough from a liver biopsy atelectasis, gas present in effusion M Epidural hematoma, ethanol LOCeffusion, RLL Delirium tremens 2 d abuse, smoker after craniotomy. Then abscesshnfiltrate fever, green sputum M Regional enteritis, enterocutaneous Intraabdominal LOCeffusion left base, left fistulas abscesses, fever, subphrenic Staphyloccal sepsis abscess M Chronic hepatitis B, liver failure, 35 d cough, fever, chest Bilateral effusions leukopenia pain, green sputum, with loc left respiratory failure effusion F Toxic goiter, chronic bronchitis, 8 d high fever, RLL infiltrate, large effusion smoker productive cough, chest pain with air-fluid levels M COPD, Parkinson's disease, 7 d fatigue, fever, Right lower lung ethanol abuse abdominal pain field loc effusions M Obesity, ethanol abuse, smoker 4 d fever, chest pain, LLL consolidation productive cough with loc effusion M COPD (steroid-dependent), 21 d malaise, cough with RLL consolidation, ethanol abuse, smoker green sputum loc effusion M Acute liver failure, Sympathetic right pleural Large loc right encephalopathy, auxiliary effusion, fever effusion orthotopic liver transplant F

Admission Only for Empyema

Time from Symptoms Until Tube Inserted (days)

Yes

7

No

21

No

4

No

15

(8F pigtail catheter)

No

61

Yes

17

Yes

12

Yes

9

Yes

30

No

11

M Acute liver failure, aplastic anemia, Sympathetic right pleural Large loc right auxiliary orthotopic liver effusion, fever effusion transplant

No

9

M Obesity, ethanol abuse, smoker, COPD, CAD, hypertension F Obesity, mental retardation, seizures, sleep apnea

Yes

22

Yes

5

14 d dyspnea, fever,

cough, green sputum Dyspnea, cough, fever hypoxia, chest pain

CAD = coronary artery disease; COPD = chronic obstructive pulmonary disease; = right lower lobe; Strep = streptokinase; Urok = urokinase.

chest tubes onto the dressing a n d bed, especially while the tubes were clamped. In 4 patients successfully treated with fibrinolytic therapy alone, the chest catheter was removed within 1 to 4 days following completion of the intrapleural infusions. Six patients with grossly purulent, culture-positive empyemas underwent conversion of their chest tube to open drainage, a s was the preference of the attending surgeon, with gradual withdrawal of the tubes a s outpatients. In 5 of these patients, chest tubes were completely removed after a mean of 12.2 2 7.4 weeks (range, 6 to 24 weeks). One patient with open tube drainage died of progressive chronic liver failure 2 weeks after successful complete drainage of his empyema with fibrinolytic therapy and resolution of his toxicity from the empyema.

LLL infiltrate, loc effusion Massive right loc effusion

d = days;

LLL = left lower lobe;

loc = loculated;

RLL

Case lllustration A typical example of the salutary effects of intrapleural fibrinolytic therapy in multiloculated empyemas is found in patient 6. On admission, this 34-year-old smoker presented with 8 days of a productive cough a n d pleuritic

chest pain, accompanied by a 39.8"C temperature a n d white blood cell count of 16,700 cells/pL. Figure 1 shows the admission posteroanterior chest radiogram, before thoracentesis, easily demonstrating multiple air-fluid levels in the large effusion. Figure 2 shows a representative chest CT image also documenting the multiloculated nature of the empyema. Figure 3 shows a chest radiogram 1 day after chest tube insertion, demonstrating that the multiloculated empyema is incompletely drained. The

ROBINSON ET AL FIBMNOLYTIC TREATMENT OF EMPYEMAS

Ann Thorac Surg 1994;5780b14

807

Table 1. Continued Mean Daily Lytic Agent Chest Drainage and Days Complications After Agent Used From Agent Started (mL)

Days in Hospital After Agent Started

When Chest Tube Removed (weeks after insertion)

Success

17

16

Complete recovery

Result During Hospitalization

Long-Term Result

Strep 14 d

None

234

f. 122

Strep 10 d

None

224

f

71

Success

25

24

Complete recovery

Strep 3d

Fever 39.8"C; 208 ?Allergic

f

64

15

2

Complete recovery

Urok 9d

None

344 f 72

Failure? Thoracotomy performed because of high fever and suspected inadequate drainage Success

69

8

Complete recovery

Urok 10 d

None

601 f 291

Success for empyema; died of progressive liver failure

Urok 7d

None

115 f 84

Success

8

1

Complete recovery

Urok I d

None

1,065 (1 d)

Success

7

8

Complete recovery

Urok

None

225

Success

9

12

Complete recovery

Urok I d Urok 7d

None

Success

6

6

Complete recovery

None

56 (1d) 113 f 89

54

2

Complete recovery

Urok 6d

None

151 f 96

70

1

Urok 7d Urok 5d

None

160

* 42

Success

7

8

Died unexpectedly in hospital 2 mo after thoracotomy of unrelated problem Complete recovery

None

107 f. 36

Success

10

1.5

Complete recovery

8d

f

137

Failure; initially improved; thoracotomy and decortication; reoperation for bleeding Failure; initially improved; thoracotomy and decortication

patient remained febrile at this time. Figure 4 is a chest CT image at the same anatomic level as Figure 2, taken after a 7-day course of daily intrapleural urokinase. The patient had become afebrile 2 days after starting the urokinase. The patient was discharged home 8 days after receiving the first dose of intrapleural fibrinolytic therapy. Figure 5 is a chest radiogram taken in the outpatient clinic 6 weeks later and is essentially the same as a chest radiogram taken before the patient's illness.

Complications of Therapy There were no complications of tube thoracostomy, both during initial insertion and in the long term. Twelve patients (92%) had no symptoms of any kind from the intrapleural fibrinolytic agents. Specifically there was no pain, dyspnea, or temperature elevation in these 12. Also

15

Not removed Died in hospital

there were no clinical signs of new or enhanced bleeding, changes in coagulation tests, signs of systemic absorption of the fibrinolytic agent, or other indications of drug toxicity. One patient (patient 3) early in the series manifested what we believe is evidence of drug (streptokinase) toxicity. In retrospect, it is thought that the sustained fever (393°C) in this patient that developed after the second dose of streptokinase was related to an allergic reaction, not due to toxicity from the empyema. This patient will be discussed in the next section. There were two deaths in this series, unrelated to the empyema and drug treatment. One patient (patient 5) died of progressive liver failure from chronic hepatitis B 2 weeks after empyema resolution. The other patient (patient ll), who was a pediatric liver transplant recipient,

808

ROBINSON ET AL FIBRINOLYTIC TREATMENT OF EMPYEMAS

Ann Thorac Surg

1994;57803-I4

Table 2. Pleural Fluid Analysis Patient No.

Appearance

Odor

pH

1 2 3

Cloudy yellow Purulent brown Hazy yellow

None 7.11 None . . .a Fetid 6.53

4

Purulent reddishbrown

Fetid

5 6 7 8 9

None None None None Fetid

11

Cloudy yellow Hazy yellow Cloudy yellow Cloudy yellow Purulent greybrown Cloudy yellow Cloudy yellow

None None

12 13

Cloudy yellow Cloudy yellow

Fetid 7.12 None 7.22

10

a

Glucose Protein (mg/dL) (g/dL)

White Blood Cell Count (cells/pL) 6,889

..

5.2

4,822

. . .a

. . .a

...

4

4.0

3,964

5,300

. .a

. . .a

...

. . .a

7.07 7.23 7.03 6.92 6.01

17 82 ,.. 66 42

2.8 3.7 3.4 4.7 2.8

650 1,381 1,326 1,407 59,600

20,556 1,278 1,800 10,000

. . .' . . .a

. .a . .a

. . .a

. . .a

. . .a ...

... . . .a

5.0 4.8

1,875 1,447

5,500 1,778

. . .a

18

LDH (IU/L)

<5 95

. . .a

Gram Stain No organisms No organisms No organisms Gram-positive cocci and Gramnegative rods No organisms No organisms No organisms No organisms Gram-positive cocci No organisms No organisms No organisms No organisms

Culture Microaerophilic Streptococcus No growth Alpha-hemoly tic Streptococcus Bacteroides, Klebsiella

Pseudomonas No growth No growth No growth Microaerophilic Streptococcus Staphylococcus a u reus Staphylococcus coagulase negative No growth No growth

Test was not performed.

LDH

=

lactate dehydrogenase.

Three patients, initially treated with intrapleural fibrinolytic agents for their empyema, ultimately required thoracotomy and decortication. Patient 3 was an alcoholic who aspirated during delirium tremens, which occurred 2 days

after operation for an epidural hematoma. A right lower lobe Klebsiella pneumonia and empyema subsequently developed. He was treated with parenteral antibiotics and chest tube insertion. It was elected to begin intrapleural fibrinolytic therapy in this patient soon after chest tube insertion because of the presence of multiple loculations on the chest CT scan, a low-grade fever, and a marked leukocytosis. Three hundred milliliters of pleural fluid drained out after the initial dose of streptokinase. Five hours after the second dose of streptokinase, however, he spiked a fever to 39.8"C and sustained this fever over the next few days despite a clearing chest roentgenogram,

Fig 1 . Posteroanterior chest radiogram on hospital admission in patient 6 demonstrates a large right pleural effusion and multiple airfluid levels. This radiogram was taken before thoracentesis.

Fig 2. Representative image from a chest computed tomographic scan performed on the same day as Figure I . This image, located 5 cm inferior to the carina, demonstrates multiple air-fluid levels in a large right pleural effusion, with a collapsed lung readily apparent. This is a classic example of a multiloculated pleural effusion.

died of metabolic abnormalities occurring with total parenteral nutrition 2 months after resolution of his empyema by thoracotomy. No deaths were related to the empyema, both in terms of intrapleural fibrinolysis or any delay to thoracotomy and decortication related to attempted intrapleural fibrinolysis.

Failures of Fibrinolysis

Ann Thorac Surg 1994;5780>14

Fig 3. Portable anteroposterior chest radiogram taken 1 day after chest tube insertion. The right pleural effusion is only partially drained and multiple air-fluid levels are still apparent, attesting to the loculated nature of the effusion.

increased chest tube drainage, and a slowly declining white blood cell count. After 3 days of streptokinase, the sustained high fever led us to believe that the empyema was incompletely drained and that a thoracotomy was necessary. In the next 2 days, with streptokinase administration stopped and while awaiting out-of-town relatives to arrive (insisted upon by the patient), his temperature returned to normal. However, all clinicians involved at that time thought that the patient had an inadequately drained empyema, and thoracotomy and decortication were carried out. At operation, only a minimal fibrous peel and fluid were found. We believe that in retrospect the fever was related to streptokinase and that the intrapleural fibrinolysis alone probably would have been successful if it had been continued. We subsequently switched from streptokinase to urokinase for intrapleural fibrinolysis in later patients, hoping to avoid allergic fevers.

ROBINSON ET AL FIBRINOLYTICTREATMENT OF EMPYEMAS

809

Fig 5. Posteroanterior chest radiogram taken on an outpatient basis 6 weeks after the initial hospital admission radiogram in Figure 1 . This last radiogram is essentially normal and is unchanged from a chest radiogram taken of this patient 1 year earlier.

The other two failures were pediatric liver failure patients who underwent auxiliary orthotopic liver transplants. Both had development of large right sympathetic pleural effusions, which are common after this operation. The effusions were drained numerous times with catheters and chest tubes, ultimately becoming loculated and infected. Initially, intrapleural fibrinolysis with urokinase was successful in both with fever resolution, increased pleural fluid drainage, and radiographic improvement. In 1 patient (patient lo), the chest tubes inadvertently dislodged and were reinserted, but ultimately this resulted in a recurrence of the empyema and the need for thoracotomy and decortication, with a long-term successful result. The other liver transplant patient (patient 11) had a persistent aplastic anemia (probably posthepatitis). Despite an initial good clinical response to intrapleural fibrinolysis, he required thoracotomy and decortication for recurrent loculated fluid. This patient also recovered completely from the empyema and thoracotomy, although he died unexpectedly 2 months later of an unrelated metabolic problem from total parenteral nutrition.

Long-Term Results The 10 patients successfully treated with intrapleural fibrinolysis uniformly had no recurrences and had near normal (or back to preempyema baseline) chest radiograms on follow-up in the outpatient clinic 2 to 6 months later. The 3 patients undergoing thoracotoniy and decortication all recovered completely from their empyema and operation. Their chest tubes were all removed before hospital discharge. Fig 4. Representative chest computed tomographic image from the same anatomic level as Figure 2 performed after a 7-day course of intrapleural urokinase therapy. Essentially all loculated fluid is gone and the lung is fully expanded.

Time and Cost of Cure The use of intrapleural fibrinolysis to facilitate drainage of multiloculated empyemas appears to be efficient therapy

810

ROBINSON ET AL FIBRINOLYTIC TREATMENT OF EMPYEMAS

in terms of length of hospitalization. Of the 7 patients whose only reason for hospitalization was the empyema, the total hospital stay after beginning fibrinolytic therapy was a mean of 9.1 3.7 days (range, 6 to 18 days). This compares favorably with the longer hospitalizations of empyema patients undergoing traditional surgical therapy by early decortication [7, 81. The cost of these fibrinolytic agents should be considered in addition to hospital days. The current patient cost at our hospital (including markup) for one dose of streptokinase (250,000 units in 100 mL of 0.9% saline solution) is $122.46. For urokinase (100,000 units in 100 mL of 0.9% saline solution), the cost per dose is $212.65. A 7-day course of urokinase fibrinolytic therapy (the mean number of treatment days for this group) would cost $1489.00. The cost for the hospitalization days, which were a mean of 9.1 days in patients admitted for only the empyema at $270.00 per day for a semiprivate room, is $2,457.00. The surgeon’s fee for chest tube insertion and daily care (9.1 days) is $775.00. Finally, the combined average cost of intrapleural fibrinolytic therapy (excluding laboratory and radiology fees) is $4,721.00. For the 3 patients in our series undergoing operation, the mean patient charge for the operation alone (excluding laboratory, radiology and respiratory therapy charges as well as professional fees) was $6,515 k $963. This figure does not include the cost of reoperation for bleeding necessary on patient 10. To be added to this figure is the cost of at least one intensive care unit day postoperatively ($820.00 per day), plus further convalescent hospital days on the ward at $270.00 per day. The least number of total hospital days postoperatively in our surgical group was 11 days in patient 3, which then totals $3,520.00 (including the 1 intensive care unit day). The surgeon and anesthesiologist professional fees combined was a mean of $3,434 2 $411. Therefore the combined mean cost (excluding the significant expense of laboratory, radiology, and respiratory therapy fees) for thoracotomy and decortication is $13,468.00. Based on these calculations from this limited series from our hospital, it appears that conservative chest tube drainage with intrapleural fibrinolysis for a multiloculated empyema is less expensive than thoracotomy and decortication. Obviously the morbidity and potential mortality associated with operation is another major factor to consider.

*

Comment The extension of the infection in lung parenchyma into the pleural cavity may be a subtle clinical event, but it marks an important change in the morbidity and mortality from lower respiratory tract infections. The inflammatory response and consolidation with accompanying volume loss usually results in pleural fluid accumulation, a socalled parapneumonic effusion. When this fluid becomes contaminated, an empyema then develops. Recognition of the change to an empyema is often masked by the use of antibiotics for the primary lung infection. Drainage has long been considered the mainstay of treatment, as its importance was even stressed by Sir

Ann Thorac Surg

1994;5780>14

William Osler, who himself died in 1919 of a Huemophilus influenzue empyema. The selection of the proper therapy for the pleural infection depends on the stage of the empyema. To stress principles of management and review the proper therapy, the American Thoracic Society in 1962 described three separate stages of an empyema [24]. If inadequately treated, an empyema will progress through these three phases. Stage I, the exudative stage, has thin fluid with a low cellular content. The lung is readily reexpanded with closed drainage. This stage may last as briefly as 24 hours, depending on the infecting organism and its host. Stage 11, the fibrinopurulent stage, is characterized by a thicker exudate, often with frank, fetid pus containing great numbers of polymorphonuclear leukocytes and fibrin. There is a progressive tendency toward loculation, often with development of a limiting membrane covering the visceral pleura preventing full lung expansion. This stage averages 7 to 10 days. Stage 111, the organizing stage, results later, when fibroblasts grow into the exudate, covering the visceral and parietal pleurae, and changing the membrane to an inelastic “peel” completely preventing lung expansion. The exudate becomes thick and the lung becomes fixed in this chronic phase. For this stage, occurring in as little as 2 to 4 weeks from the onset of the pleural effusion, intercostal drainage alone will not reexpand the lung, and decortication or open drainage is required. Although chest tube drainage only may expand the lung in stage I exudative and in some stage I1 fibrinopurulent empyemas, the common development of loculations resulting in inadequate drainage and continued toxicity commonly calls for more aggressive therapy. Himelman and Callen [3] in their series of patients with loculated parapneumonic effusions documented an increase in mortality from sepsis and an increased need for operative decortication once loculations occurred. This early surgical approach of open thoracotomy for pleural drainage has been advocated by many authors over the last 46 years, since Samson and Burford in 1947 first recommended decortication within 2 to 3 weeks of the diagnosis of an empyema [PSI. A more contemporary variation on the theme of early surgical management is the use of thoracoscopy to debride and irrigate the pleural cavity [25]. However, when this route is chosen, at least 40% of patients require additional, more aggressive surgical procedures. Despite the recent interest in videoassisted thoracoscopic procedures, the technical difficulty caused by pleural adhesions and loculations present in these patients tends to relegate this approach to highly selected patients. Finally, the use of rib resection and drainage is advocated by other authors particularly for postoperative empyemas and those in immunocompromised or debilitated patients [6]. The use of intrapleural fibrinolytic agents is another viable alternative to facilitate drainage and lung expansion in stage I and I1 empyemas inadequately treated by a chest tube alone. This method, advocated by many in the 1950s and 1960s [12, 131 and reinforced by the American Thoracic Society in 1962 [24], has been largely overlooked by thoracic surgeons in the last several decades. One reason for this lack of enthusiasm may be the variable results and

Ann Thorac Surg 1994;5780>14

side effects of the fibrinolytic agent streptokinase used exclusively in the earlier studies, and the only drug meriting even a cursory mention in the major surgical textbooks [lo, 111. In addition, dissatisfaction often resulted when fibrinolytics were used as a last-ditch effort late in the course of the empyema, when there already had been pleural membrane organization with capillary and small vessel invasion [13]. However, realistic staging of the empyema makes the rational use of fibrinolytic agents feasible. Streptokinase, the first fibrinolyticagent to be described [16, 171, is a purified proteolytic enzyme derived from a bacterial protein of group C beta-hemolytic streptococci. Plasminogen is converted to the proteolytic enzyme plasmin by streptokinase. Plasmin then degrades fibrin clots and fibrinogen. A number of series over the years have shown surprisingly good efficacy in a variety of patient populations [12-18,261. The five most recent series report a combined 82.6% complete cure rate with chest tube drainage and intrapleural streptokinase alone in a total of 69 loculated empyema patients [14-17, 261. Despite the fact that intravenous streptokinase elicits a sustained systemic fibrinolytic state, intrapleural use of this agent caused no evidence of coagulation effects in a study of systemic fibrinolytic activity in 10 patients by Berglin and associates [27]. Nevertheless, Rosen and colleagues in their review [26] mentioned a single case report of major hemorrhage after intrapleural streptokinase that may have been related to systemic absorption of the agent. Other systemic side effects are common with intrapleural streptokinase, including fever as high as 40°C and pleural pain [13,26]. This pyrogenic reaction is not caused by the presence of a conventional bacterial pyrogen. Rather, the response has more the appearance of a delayed sensitivity-type reaction, probably from the presence of some contaminating streptococcal proteins as well as from the streptokinase itself. Further purification has failed to eliminate these reactions completely. Toxic responses such as arthralgias, nausea, malaise, and headache were also seen with this agent earlier [28], although they appear to be less frequent with more purified preparations of this agent [17]. Anaphylaxis, which may occur is as many as 3.3% of patients with intravascular administration of the agent [17], is a very uncommon event with intrapleural administration and has been described in only two case reports [29]. Specific immunoglobulin G antistreptokinase antibodies measurable up to 4 years later are formed with only one intravascular administration of streptokinase, as reported by Elliot and associates [30]. It has been found that repeated intrapleural streptokinase also results in formation of these neutralizing antibodies, but this is not surprising considering the systemic pyogenic reaction that commonly occurs with many patients [14, 281. This formation of antibodies is an important consideration in view of the older patient, who may subsequently require intravascular thrombolytic therapy later in life and may risk a severe allergic reaction to repeat administration of streptokinase. Urokinase is a naturally occurring direct activator of plasminogen produced by the kidney and excreted in the

ROBINSON ET AL FIBRINOLYTIC TREATMENT OF EMPYEMAS

811

urine. This agent has the major advantages of being nonantigenic and nonpyrogenic, but still has a fibrinolytic effect very similar to streptokinase [31]. Despite its introduction into clinical medicine in 1965, it was not until 22 years later that Vogelzang and colleagues reported the successful use of urokinase outside of the vascular system for fibrinolysis of infected extravascular hematomas [19]. In the 3 patients treated in this initial series, there were no side effects, and specifically there was no evidence of systemic fibrinolysis with specific coagulation testing. This experience is similar to the lack of systemic fibrinolysis seen with intrapleural streptokinase reported by Berglin and associates [27]. Application of urokinase to loculated pleural effusions was first reported in 1989 by Moulton and colleagues [20]. They treated 5 patients with infected hemothoraces and 8 patients with empyemas initially with catheter drainage. When incomplete empyema drainage was recognized, urokinase was administered intrapleurally at 80,000 to 150,000 units per instillation, with a dwell time of 1 to 2 hours, for a mean 4.3 instillations per patient. This method was completely successful in resolving the empyema in 12 of 13 patients (92%), and there were no side effects or changes in coagulation studies. Lee and associates (211 in 1991 subsequently reported a prospective series of 10 patients with multiloculated empyemas. In their patients, an 8F catheter was inserted by fluoroscopic guidance, followed by daily urokinase instillations of 100,000 units in 100 mL of 5% dextrose in water. After a mean of 4 intrapleural instillations, 9 of 10 empyema patients (90%)were successfully treated with urokinase. No complications occurred nor were there changes in coagulation profiles. Analysis of the empyema fluid prior to treatment demonstrated plasminogen levels similar to plasma and elevated fibrin degradation products. These findings lend support to the presumed action of urokinase to accelerate the ongoing fibrinolytic process in the pleural cavity, thereby breaking down loculi of fluid and facilitating drainage. The most recent report in the literature was by Geradi and colleagues [22] in 1993 and dealt with the successful treatment of 4 of 6 (67%)loculated empyema patients with chest tube drainage and daily intrapleural urokinase. In the current series, 10 patients with loculated pleural empyemas were treated with intrapleural urokinase resulting in complete resolution in 8, with no complications or side effects. Although streptokinase was effective also in 3 other patients earlier in our experience, the high fever seen with this agent led to our change to urokinase. Unlike some prior urokinase series [20, 211, we chose to use larger, more conventional chest tubes for drainage, so that effective suction could be maintained between drug instillations to ensure the maximum lung reexpansion. In our experience, it is difficult to maintain tube patency and effective intrapleural suction with small 8F to 12F pigtail catheters in adult chests. Our two failures in the urokinase group were completely different from the other patients because they were not postpneumonic empyemas. Both failures were in pediatric liver failure patients who underwent auxiliary orthotopic liver transplants (patients 10 and 11). Persis-

812

Ann Thorac Surg

ROBINSON ET AL FIBRINOLYTIC TREATMENT OF EMPYEMAS

tent sympathetic right pleural effusions developed, commonly seen after this surgical procedure. The effusions eventually became contaminated, and finally loculated empyemas with staphylococcal species developed. The initial response in these patients to intrapleural urokinase was quite good, with improved drainage, but eventually they had recurrence and required open decortication. The failure in these 2 patients may have been related to their continuing pleural fluid accumulation stimulated by the subdiaphragmatic organ transplant. In addition, both were immunosuppressed by steroids and cyclosporine A, and 1 patient (patient 11) had a severe persistent leukopenia. Therefore, there were minimal inflammatory and immune mechanisms active to aid in resolving the empyema with intrapleural fibrinolytic therapy only. It is doubtful that young age played a role in the failure in these 2 patients, because Rosen and associates' recent series [26] had 100% success with intrapleural fibrinolysis in all 5 of their pediatric empyema patients. In any discussion of alternative clinical therapies in the current atmosphere of fiscal constraints and limited health care dollars, treatment costs must be factored into our decision making. For this reason, a simple analysis of patient costs was determined for intrapleural fibrinolytic therapy compared with traditional surgical decortication for loculated empyemas. Urokinase is not only clinically effective, but was quite efficient with a mean total hospital stay of only 9.1 5 3.7 days after the start of therapy in patients in whom the empyema was the sole reason for hospitalization. Although the total patient charge for a single dose of urokinase (100,000 units in 100 mL of 0.9% normal saline solution) seems high at $212.65, the total cost for the drug, hospital room, and professional fees is only one third that of the total costs for a patient undergoing thoracotomy and decortication ($4,721.00 versus $13,468.00, respectively). Clearly from a health care expenditure standpoint, intrapleural fibrinolysis with urokinase for loculated empyemas is cost-effective therapy. For this series, patients with loculated empyemas following chest operation, trauma, or bronchopleural fistulas were excluded in our review criteria, although during the study period there were no such patients at our institution. Nevertheless, our favorable experience with urokinase now would encourage us to try this "benign" intrapleural drug therapy cautiously in these previously excluded patients. Even patients with bronchopleural fistulas have been treated successfully with streptokinase with no adverse effects [15]. However, as in prior studies [20-221, we also excluded patients with an obvious stage I11 chronic organizing empyema having a very thick pleural peel, because fibrinolytic agents are quite unlikely to break down organized collagen containing capillary ingrowth. On the other hand, most early pleural peels seen with stage I1 fibrinopurulent empyemas will resolve completely with total drainage when followed up longterm by chest CT scan [32]. Our treatment protocol with streptokinase and later urokinase was developed empirically, based on the clinical studies in the literature [12-18, 20-221. The urokinase intrapleural dose of 100,000 units with dwell times of 6 to 8 hours daily appears safe without adverse effects found

1994;57:803-14

in our series, as well as in the other three urokinase clinical trials [20-221. However, the optimal doseresponse relationships, most effective timing, duration of treatment, or volume of administration are unknown at present. Only a prospective clinical trial will help to answer these questions. Managing the toxic empyema patient with incompletely drained pus has generally required a conventional open surgical approach, with its resultant substantial morbidity, mortality, and expense. A less aggressive means to facilitate complete pleural drainage using intrapleural fibrinolysis has been available and discussed in the literature since 1949. This approach has not been widely employed for a variety of reasons, however, especially by thoracic surgeons in the last few decades, despite the fact that in most cases the clinical outcome is the same or better than that in open surgical decortication. There is still no clear consensus on the role of any of these options in managing complicated empyemas, as indicated by the wide range of preferences expressed in a recent published survey of chest physicians [9]. The current series is only a preliminary study of empyema patients, and carries with it the weakness of not being prospective or randomized. Nevertheless, the excellent results with 13 patients strongly support this alternative clinical strategy as an adjunct to facilitate complete drainage of multiloculated empyemas. Although clinical judgment must be exercised in selecting therapy, we favor this safe, very cost-effective approach, which avoided the morbidity of a major thoracotomy for 77% of a group of empyema patients who traditionally would have required open surgical therapy.

Addendum Since the original submission of the manuscript, we have used the method described herein of intrapleural urokinase as an adjunct to facilitate chest tube drainage and lung reexpansion in six more postpneumonic multiloculated stage I1 empyemas. This group included 1 adult liver transplant recipient and one 3-yearold child. All were successfully treated without complication, and did not require a surgical procedure.

References 1. Hippocrates. The genuine works of Hippocrates. Translated by Francis Adams. 1st ed. London: The Sydenham Society, 1849:768. 2. Graham EA, Bell RD. Open pneumothorax: Its relation to the treatment of acute empyema. Am J Med Sci 1918;156:839-71. 3. Himelman RB, Callen PW. The prognostic value of loculation in parapneumonic pleural effusions. Chest 1986;90:852-6. 4. Samson PC, Burford TH. Total pulmonary decortication: Its

5. 6. 7. 8.

evolution and present concepts of indications and operative technique. J Thorac Surg 1947;16:127-53. Samson PC. Empyema thoracis. Ann Thorac Surg 1971;ll: 210-21. Lemmer JH, Botham MJ, Orringer MB. Modern management of adult thoracic empyema. J Thorac Cardiovasc Surg 1985; 90~849-55. Van Way C, Narrod J, Hopeman A. The role of early limited thoracotomy in the treatment of empyemas. J Thorac Cardiovasc Surg 1988;96:43&9. Ashbaugh DG. Empyema thoracis: Factors influencing morbidity and mortality. Chest 1991;99:1162-5.

Ann Thorac Surg 1994;57803-14

9. Strange C, Sahn SA. The clinician’s perspective on parapneumonk effusions and empyema. Chest 1993;103:259-61. 10. DeMeester TR, Lafontaine E. The pleura. In: Sabiston DC, Spencer FC, eds. Surgery of the chest. 5th ed. Philadelphia: Saunders, 1990:467-76. 11. Deslauriers J, Beauchamp G, Desmeules M. Benign and malignant disorders of the pleura. In: Baue AE, Geha AS, Hammond GL, Laks H, Naunheim KS, eds. Glenn’s thoracic and cardiovascular surgery. 5th ed. Norwalk: Appleton and Lange, 1991:46&71. 12. Tillett WS, Sherry S. The effect in patients of streptococcal fibrinolysin (streptokinase) and streptococcal desoxyribonuclease on fibrinous, purulent and sanguinous pleural effusions. J Clin Invest 1949;28:17%90. 13. Geha AS. Pleural empyema: Changing etiologic, bacteriologic, and therapeutic aspects. J Thorac Cardiovasc Surg 1971;61:62W5. 14. Bergh NP, Ekroth R, Larsson S, Nagy P. Intrapleural streptokinase in the treatment of haemothorax and empyema. Scand J Thor Cardiovasc Surg 1977;11:2658. 15. Mitchell ME, Alberts WM, Chandler KW, Goldman AL. Intrapleural streptokinase in management of parapneumonic effusions: Report of series and review of literature. J Fla Med ASSOC1989;76:1019-22. 16. Lysy Y, Gavish A, Lieberson A, Werczberger A, Reifen R, Dudai M. Intrapleural instillation of streptokinase in the treatment of organizing empyema. Isr J Med Sci 1989;25: 284-7. 17. Aye RW, Froese DP, Hill LD. Use of purified streptokinase in empyema and hemothorax. Amer J Surg 1991;161:560-2. 18. Alfageme I, Mufioz F, Pefia N, Umbria S. Empyema of the thorax in adults: Etiology, microbiologic findings, and management. Chest 1993;103:83943. 19. Vogelzang RL, Tobin RS, Burstein S, Anschuetz SL, Marzano M, Kozlowski JM. Transcatheter intracavitary fibrinolysis of infected extravascular hematomas. Am J Roentg 1987;148: 378-80. 20. Moulton JS, Moore PT, Mencini RA. Treatment of loculated

ROBINSON ET AL FIBRINOLYTICTREATMENT OF EMPYEMAS

21.

22. 23. 24.

25. 26. 27. 28. 29. 30.

31. 32.

813

pleural effusions with transcatheter intracavitary urokinase. Am J Roentg 1989;153:941-5. Lee KS, Im J, Kim YH, Hwang SH, Bae WK, Lee BH. Treatment of thoracic multiloculated empyemas with intracavitary urokinase: A prospective study. Radiol 1991;179: 771-5. Gerardi DA, Schatz P, Lahiri B. Intrapleural urokinase in the treatment of empyema. Chest 1993;104(supp):161S. Light RW. Management of parapneumonic effusions. Chest 1991;100:892-3. Andrews NC, Parker EF, Shaw RR, Wilson NJ, Webb WR. Management of nontuberculous empyema: A statement of the subcommittee on surgery. Am Rev Respir Dis 1962;85: 9354. Ridley PD, Braimbridge MV. Thoracoscopic debridement and pleural irrigation in the management of empyema thoracis. Ann Thorac Surg 1991;51:461-4. Rosen H, Nadkami V, Theroux M, Padman R, Klein J. Intrapleural streptokinase as adjunctive treatment for persistent empyema in pediatric patients. Chest 1993;103:1190-3. Berglin E, Ekroth R, Teger-Nilsson AC, William-Olsson G. Intrapleural instillation of streptokinase. Effects on systemic fibrinolysis. Thorac Cardiovasc Surg 1981;29:124-6. Hubbard WN. The systemic toxic responses of patients to treatment with streptokinase-streptodornase. J Clin Invest 1951;30:11714. Goehring WO, Grant JJ. Allergic reaction to streptokinasestreptodornase solution given intrapleurally. JAMA 1953;152: 1429-30. Elliot JM, Cross DB, Cedarholm-Williams SA, White HD. Neutralizing antibodies to streptokinase four years after intravenous thrombolytic therapy. Amer J Cardiol 1993;71: 64c-5. Fletcher AP, Alkjaersig N, Sherry S, et al. The development of urokinase as a thrombolytic agent. J Lab Clin Med 1965; 65:713-31. Neff CC, vansonneberg E, Lawson DW, Patton AS. CT follow-up of empyemas: Pleural peels resolve after percutaneous catheter drainage. Radiol 1990;176:195-7.

DISCUSSION DR RONALD C. HILL (Morgantown, WV): This report discusses a nonsurgical approach to the treatment of empyemas. Thirteen patients with stage I and I1 empyemas were treated using thrombolytic agents. Patients excluded from operation were those having recent resectional operations or open-heart procedures, those with longstanding chronic draining empyemas and bronchopleural fistulas, and those with stage 111 chronic organizing empyemas. I found this to be a very interesting report and a fresh look at a thoracic problem that we have had for quite some time. I have three questions. Why were postresectional surgical patients excluded? Many times they are several days to weeks out from their resections before empyemas develop. I think that the thrombolytic agents may not have an effect on the suture lines at that point. What determined how long a tube was clamped once the thrombolytic agent was instilled in the pleural space? And although this was addressed in the report, have you had any experience with thoracoscopy drainage of empyemas either before or after thrombolytic treatment? This method in the early stage empyema may be faster and may also avoid thoracotomy. DR BRADLEY M. ROGERS (Charlottesville, VA): I enjoyed this report very much. This is what we are coming to expect, very practical reports from your group at this organization. I think this provides some very classic suggestions for us.

I rise to ask a few questions and to make a comment. We have approached empyemas in children in a slightly different fashion. We reported earlier this year the use of thoracoscopy in 10 children with unresponsive empyemas. We have used thoracoscopy to break up the locules in the chest, and have performed mechanical debridement of purulent material in the chest. We have used urokinase in 4 of those 10 patients when the fibrinous material we have left behind has clogged the chest tube in the first 24 to 36 hours. We have used a single dose of urokinase and usually that will reestablish drainage. We have been successful in all 10 of those patients in evacuating the empyema and have released the patients from the hospital in a mean of 6 days. I believe that you operated on 3 of your patients, and I wonder if on the basis of the findings at the time of open operation you think you could have done those with thoracoscopy to redebride the pleural space just as effectively. I notice that you use chest computed tomographic scan to evaluate the progress of the empyemas and to separate the patients who are in stage I1 and stage 111. We found this to be a little bit deceiving in the children. A lot of pleural edema may look like a thick pleural peel on chest computed tomographic scan, but is readily identifiable by ultrasound. I wonder if you have any comparison between using ultrasound and chest tomography in these patients. You mention in the manuscript that 1 of your patients had some bleeding through the chest tube. Did you obtain coagula-

814

ROBINSON ET AL

FIBRINOLYTIC TREATMENT OF EMI‘YEMAS

tion profiles in the patients who had rather prolonged treatments with the urokinase, and did you notice any abnormalities? The last question I have is whether you have any information on the appropriate dose of urokinase. You have chosen 100,000 units per dosage. We have used, purely arbitrarily, 50,000 units per dosage in the children, and have no idea how to really choose an appropriate dose. In your survey of the literature, have you seen any research that might provide us some information in that regard?

DR F. HAMMOND COLE, JR (Memphis, TN): I was curious about the cost of the urokinase versus the streptokinase, because I see you had 100% success with the streptokinase. In other words, when you finally explored the patient with possible failure, there was no fluid in the chest. I thoroughly enjoyed the report. This is one of the measures we have mentioned on rounds and seldom if ever actually do, so this will give me some encouragement in using streptokinase. DR NORMAN J. SNOW (Cleveland, OH): I wonder if you could comment on whether the endpoint for discharge was the empyema. There were quite long hospital stays. Was the empyema the primary reason for the continued hospitalization, or were these the liver transplant patients who had to stay in the hospital for other reasons? It seems like a long posttreatment stay relative to earlier treatment. DR WILLIAM A. COOK (North Andover, MA): It would seem to me that no discussion of this disease should go by without one caution. That is that these people are often very fragile and clinically treacherous. Any manipulation may very well send them into septic syndrome and start a chain of events that becomes irreversible. Would you comment on whether you think a period of antibiotic therapy, directed by culture or Gram stain, is important before you start your manipulations on these people? DR ROBINSON: I would like to thank the discussants for their comments. Doctor Hill, we chose not to use this treatment for patients who had recent chest operation. We decided not to take the chance that we might have bleeding problems with fibrinolytic agents in these patients, and therefore excluded them. I should mention that in the last 3 years, however, we have been fortunate and did not have any patients with a postpneumonectomy or postresectional empyema, so this is a moot point. The time in which the chest tube was clamped was based on the few articles in the literature on this topic. In one study from Korea in the radiology literature, urokinase was instilled into the pleural cavity via the chest tube, which was clamped overnight without any adverse reaction. In this particular series of 10 patients extensive coagulation profiles were performed on all patients, and no clotting abnormalities or other side effects were found with overnight chest tube clamping. Based on this study and others, we ultimately chose to use a 6- to 12-hour period for chest tube clamping. Finally, we did not perform thoracoscopy on any of these patients, so I cannot provide you with any information on what the pleural cavity looked like, or whether this surgical procedure would have helped. Doctor Rodgers, I have read your published series of pediatric patients who were treated by thoracoscopy for their empyemas and found the results quite interesting. We did not, however, use thoracoscopy on any patients in our series, and I doubt that it would have been helpful on the 3 patients requiring open thoracotomy in our series. One of these 3 was treated with streptokinase and underwent operation because of persistent

Ann Thorac Surg 1994;5780?-14

high fevers. Very little residual fluid or debris was found at the time of operation. The other 2 patients required an actual decortication. Our use of the chest computed tomographic scan as a tool to evaluate the empyemas before and after treatment was quite useful in helping to stage the patients. We also used information on the length of time from the onset of symptoms until hospital admission to better assess the chronicity of this problem. We did not, however, use ultrasound, and therefore I do not have any information comparing these two techniques. In the manuscript, none of our patients had bleeding complications from their chest tubes from the urokinase. I believe the discussion about bleeding involved 1 patient in whom a hemothorax developed that became infected. He subsequently was referred for chest tube drainage, and was successfully treated with intrapleural urokinase. We measured coagulation parameters in several patients who showed no abnormalities, but no standardized protocol was followed. As I mentioned previously, there has been one study in the radiology literature in which coagulation profiles were carefully evaluated after the use of intrapleural urokinase, and no abnormalities were found. The dose of 100,000 units of urokinase was based on several prior series from the literature. I cannot tell you whether this is the most appropriate dose for adults, but we have found it to be generally effective and free of side effects. Perhaps future studies on intrapleural urokinase should address the question of the appropriateness of this dosage. Doctor Cole, cost is an important consideration, and we did examine this issue. In my initial presentation, time did not permit me to mention the cost of these agents. The patient charge for streptokinase in our medical center is $122.00 per dose, and urokinase is $212.00 per dose. We also broke down the total cost for hospitalization, fibrinolytic drugs, and the surgeon’s fee for the 7 patients admitted just for empyema treatment. Excluding the cost of laboratory studies and roentgenograms, this averaged out to $4,721.00 per patient. If we roughly compared this with the cost of the 3 patients undergoing surgical decortication excluding laboratory and x-ray study costs, we found that the mean cost of operation, hospital room, and physician’s fees averaged $13,468.00per patient. A more detailed cost comparison is found in the full manuscript. We believe that these preliminary cost comparisons suggest that chest tube drainage with fibrinolytic therapy for loculated empyemas is much less expensive than a surgical approach. In addition, the nonantigenic urokinase is only minimally more expensive than streptokinase. And, Dr Cole, I hope that our experience with this method of conservative therapy for multiloculated empyemas will encourage you to try this on appropriate patients at your institution. Doctor Snow, the end point for discharge after empyema treatment was left up to the attending surgeon. The mean hospital stay for the 7 patients whose only reason for hospitalization was the empyema was 9.1 days. Some of the longer hospitalizations in the series were in patients successfully treated with urokinase who had other comorbid disease requiring therapy. We thought that this was quite a reasonably short length of time in the hospital considering that we were treating the difficult problem of complicated, multiloculated empyemas. Of the 3 patients undergoing decortication, the shortest hospital stay was 11 days. However, the overall cost for the surgical versus the urokinase treatment groups was far different, as I mentioned earlier. Doctor Cook, all of the patients in this series were receiving antibiotics for 3 to 14 days before urokinase treatment. The antibiotics used were chosen based on pleural fluid culture results whenever there was a positive culture. Consequently, we felt comfortable embarking on enzymatic debridement of the pleural space to facilitate better empyema drainage.