Eur J Cardiothorac Surg 1998;14:117-122
© 1998 Elsevier Science NL
Is aggressive surgery in pleural empyema justified?1
H. Renner,
S. Gabor,
H. Pinter,
A. Maier,
G. Friehs,
F.M. Smolle-Juettner
Department of Thoracic and Hyperbaric Surgery, University Medical School of Graz, A-8036 Auenbruggerplatz 29, Graz, Austria
Received 29 September 1997;
received in revised form 14 April 1998;
accepted 13 May 1998.
Corresponding author. Tel.: +43 316 385–2803; fax: +43 316 385–2756.
 |
Abstract
|
|---|
Objective: High risk and a long hospitalization time are often quoted as negative aspects of aggressive surgery in pleural empyema. We did a retrospective analysis evaluating outcome and duration of hospitalization in patients treated according to an aggressive schedule. Methods: Since 1989 we have treated 101 patients with pleural empyema (72 males, 29 females; mean age 50.3 years, range 11–91 years; 77 metapneumonic empyema, 24 empyema following trauma or abdominal surgery). Sixty-nine patients had had unsuccessful conservative pre-treatment (antibiotics, thorcozentses, drainage/irrigation, VATS). Thirty-one were critically ill patients. In eight cases a seropurulent stage of empyema was present, 17 patients had fibrinous membranes, 30 an organizing stage with and 46 without well identifiable dissection plane. Eighty-five patients proceeded to thoracotomy. Pulmonary abscesses or indurative pneumonia necessitated wedge-resection, lobectomy, or pneumonectomy in 29 cases. In the presence of gross necroses or callosities not amenable to decortication primary open-window thoracostomy (n=22) was carried out. In six cases a secondary open-window thoracostomy was carried out because of persisting putrid secretion and sepsis persisting after decortication or after drainage. The thoracostomy was closed when clean granulative tissue developed. Sixteen patients underwent only drainage and irrigation because of an early stage or because of a general condition not permitting thoracotomy. Results: Three patients died due to severe sepsis not responding to treatment, one had fatal bleeding from a duodenal ulcer (mortality rate 3.9%). The others were able to resume their preoperative activities. The median duration of hospitalization was 14 days (mean 21.1 days; SEM 1.7 days). Conclusion: Aggressive surgery for pleural empyema beyond the seropurulent stage ensures rapid relief from sepsis at a low mortality rate even in very ill patients.
Key Words: Pleural empyema • Decortication • Hospitalization time
|
Introduction
|
|---|
Many physicians are favouring treatment of empyema by drainage and irrigation or instillation of fibrinolytic agents or video-assisted thoracoscopic surgery (VATS), often irrespective of its stage
[1]
[2]
[3]
[4]. High success rates are reported, though it is conceded that `some' patients have to proceed to thoracotomy.
Though it has been known that radical surgery for pleural empyema has a 95% rate of success
[5], the stress of thoracotomy, a prolonged duration of hospitalization and a high mortality rate are quoted by its opponents. We did a retrospective analysis evaluating outcome and duration of hospitalization in patients treated according to an aggressive schedule.
|
Materials and methods
|
|---|
Patients
Since 1989 we have treated 101 patients with pleural empyema (72 males, 29 females; mean age 50.3 years; range 11–91 years).
Seventy-seven patients had metapneumonic empyema, and 24 an empyema following trauma or abdominal surgery (see Table 1). Coexisting disease or otherwise adverse conditions were found in 58 patients (see Table 2). Diabetes mellitus (n=16), alcoholism (n=19), malignant disease (n=8), cardiovascular disease (n=7) or psychiatric disorders (n=6) were evident before operation. Six patients were war refugees from Bosnia and presented in a very poor nutritional condition. Thirty-one patients were critically ill presenting with symptoms of either hyper- (n=21) or hypodynamic (n=10) septic shock. Three of them had septic renal failure requiring hemodialysis prior to admission, 16 required artificial ventilation and 28 continuous administration of catecholamines. All others had clinical signs of sepsis with at least four of the following criteria present. Fever >38.9°C or hypothermia, respectively, tachycardia, tachypnea, leucocytosis >15 000 or leucopenia, respectively, hypotension <90 mmHg or thrombocyto-penia <100 000.
In 66 of the 101 patients (65.3%) with verified empyema (intrapleural fluid accumulation on chest roentgenograms and CT-scan in the presence of septic conditions and pus on thoracocentesis) attempts of conservative treatment prior to admission had been made. Eleven underwent only high-dose antibiotic treatment, 21 had multiple thoracocenteses with or without instillation of antibiotics, 18 had one, 11 two to four unsuccessful attempts of drainage with or without irrigation, and five had VATS intrapleural washing. The resulting delay ranged from 8 to 182 days (median 22 days). The remaining 35 patients had no pre-treatment and were admitted within 7 days from the beginning of symptoms (see Table 3).
Surgical treatment
The primary choice of therapy was based upon the findings on CT-scan and on the general condition of the patient. If the latter did not allow decortication, a tube thoracostomy was performed under local anaesthesia, irrespective of the stage of empyema. The pleural space within the range of the finger was palpated, ensuring a correct positioning of the tube and examining the consistency of the pleura. If necessary, a 2-cm fragment of a rib was resected in order to prevent external compression of the thoracostomy tube. Subsequently pleural irrigation using saline was carried out. It was continued, until the bacteriological specimens from the pleural fluid were sterile. If the septic condition did not subside within 48 h, in presence of callosities and a lung that failed to reexpand an open window thoracostomy, resecting two ribs to an extent of 8–10 cm was carried out. In these cases, the thoracostomy was left open to granulation.
In patients who would tolerate decortication, a drainage tube was only inserted if there was no evidence of pleural thickening or loculations on a recent CT-scan (see
Fig. 1
). If membranes or gross callosities were found on palpation during thoracostomy, or if the lung did not reexpand within 24 h after drainage the patient proceeded to thoracotomy. All other patients had primary lateral thoracotomy with debridement, decortication with or without resection and/or fenestration. Care was taken to remove the membranes as completely as possible. In the presence of well confined, thick callosities and a well identifiable dissection plane removal of the empyema sack in one piece was carried out (see
Fig. 2
).
View larger version (80K):
[in this window]
[in a new window]
|
Fig. 1. CT-scan of pleural empyema, recurring after three attempts of thoracocentesis. No loculations visible.
|
|
View larger version (127K):
[in this window]
[in a new window]
|
Fig. 2. Same patient as in
Fig. 1. Decortication was carried out 2 days after the CT-scan, following unsuccessful chest tube drainage. Surgical specimen. Empyema membrane removed as a whole. Multiple loculations.
|
|
In chronic organizing empyema without a residual dissection plane the therapeutic aim of providing a full reexpansion of the lung while removing all pyogenous membranes and callosities could not be achieved without creating peripheral parenchymal fistulae. These – often substantial – leakages remained untreated and were left to spontaneous closure. Two to three chest tubes were inserted before closing the pleural cavity.
Pulmonary abscesses draining into the pleural cavity were diagnosed macroscopically. By opening them totally, both their size and their anatomical relationships were determined. The extent and degree of parenchymal destruction were examined by intraoperative frozen section histology. In both cases primary resections were carried out, trying nevertheless to preserve as much parenchyma as possible. In the presence of cellulitis at the chest wall, of gangrene at the pleural surfaces or of thick callosities at the visceral pleura that could not be removed owing to a poor condition of the patient, primary open-window thoracostomy was carried out, resecting one to three ribs. A secondary open-window thoracostomy was performed in case of putrid secretion and severe sepsis persisting or recurring after decortication. After daily changes of dressings thoracostomies were closed either directly or by using the pedicled latissimus dorsi as soon as clean granulative developed. A flow chart of the surgical management is shown in
Fig. 3
.
Perioperative management
Patients in whom no bacterial culture was available preoperatively, or in whom the bacteriological evaluation had remained inconclusive had antibiotic treatment with broad-spectrum cephalosporines. All others had treatment according to the antibiograms.
If the lung showed chronic atelectases after decortication, or in patients who had open window thoracostomy without a stable callosity of the visceral pleura preventing collapse of the lung, positive pressure ventilation was kept up in the postoperative course. Weaning was carried out after closure of the chest wall or after disappearance of atelectases on the roentgenogram. Thoracic drainages were removed after subsidence of air loss and in the presence of a fluid volume not exceeding 100 ml/24 h.
Statistical evaluation
Categorical variables were compared with each other by means of the chi-square test, and multivariate analysis of dichtotomic variables was carried out by logistic regression analysis. The statistical methods were performed using the SPSS for Windows program package (Spectra Publishing, Sunnyvale, CA, USA).
|
Results
|
|---|
The definitive staging of empyema revealed a seropurulent stage in eight, a membranous stage in 17 and an organizing stage in 76 patients, 30 of whom had a well identifiable dissection plane and 46 of whom had not.
In sixteen patients drainage and irrigation was carried out because of an early stage (n=5) or because of a general condition not permitting decortication (n=11) were successful. In five operable patients drainage was converted into decortication and three patients who did not tolerate decortication underwent debridement and fenestration. Thus a total of 85 patients proceeded to thoracotomy and decortication and/or fenestration and debridement.
Twenty-nine patients underwent primary (n=25) or secondary (n=4) resection (three pneumonectomies, 17 lobectomies and nine wedge-resections). Primary open-window thoracostomy was carried out in 22 patients because of cellulitis of the chest wall (n=7) gangrene of the pleural surfaces (n=8) or of tight callosities at the visceral pleura (n=7). All seven patients with cellulitis of the chest wall had had multiple drainages or VATS pre-treatment. In six cases a secondary open-window thoracostomy was performed. Because diseased lung parenchyma which had been tentatively preserved during the first intervention was the source of persisting infection in all these cases, a secondary parenchymal resection was carried out in all of them. Twenty-five out of 66 patients who had ineffective pre-treatment required a lung resection, compared with only four out of 35 who had immediate adequate treatment (P<0.05).
The mean overall perioperative blood requirement was 3.9 units of packed red blood cells (0–17). For fenestration procedures it was 4.8 (2–8), for decortication 4.3 (0–17), for decortication and resection 4.1 (2–10) and for drainage alone 2.6 (0–5). Anaemia to a level of 10 mg/dl haemoglobin was tolerated depending on the general condition and on the age of the patient.
Bacteriological specimens remained sterile in 49 patients all of whom had undergone pre-treatment attempts. In the remaining 52 cases one (n=37) or more micro-organisms were found in the pleural fluid: Staphylococci (n=14), Streptococci (n=10), E. coli (n=9), Pseudomonas (n=8), Enterobacteriaceae (n=7), Klebsiella (n=5), Acinetobacter (n=4), Bacterioides (n=2) and Haemophilus influenza (n=1). Five patients were found to have tuberculosis.
In the 29 resection specimens one or multiple non-tuberculous pulmonary abscesses (n=14), organizing pneumonia with destruction (n=7), primarily necrotizing pleuropneumonia (n=5), and necrotizing tuberculosis with multiple abscesses (n=2) and one centrally necrotic bronchogenic carcinoma (T1N0), that had mimicked a pulmonary abscess were found.
In all but four patients the septic condition subsided within 48 h after the definitive surgical treatment. The type of micro-organism found or the fact whether the bacteriological specimen had remained sterile or not did not influence the clinical course. The mean duration of drainage was 5.4 days (3–32 days). In 20 out of the 28 patients with open window thoracostomy the chest wall was closed within 5–12 days (mean 5.7 days). Five patients declined the closure. They were discharged with the pleural space left to spontaneous healing.
In 43 patients positive pressure ventilation was continued postoperatively for an interval of 3–60 days (median 5 days). Weaning was carried out after closure of the chest wall (n=20) or after disappearance of atelectases on the roentgenogram (n=18). Due to respiratory insufficiency intubation and positive pressure ventilation had to be continued in five patients. Two were eventually weaned by using a tracheotomy, three had a fatal course.
Complications were seen in six patients: One haemorrhage requiring redo thoracotomy, and five cases of delirium tremens. Three patients died due to sepsis not responding to treatment (hepatic insufficiency, necrotzing pancreatitis, septic thrombocytopenia), one had fatal bleeding from a duodenal ulcer. The overall mortality rate is 3.9%. For drainage and irrigation alone it is 0%, and for decortication with resection it is also 0%. If fenestration becomes necessary the mortality rate rises to 10.7%. There is a significantly higher mortality rate in patients who had conservative pre-treatment attempts (6.0%) compared with those who had none (0%) (P>0.001; chi-square test). Two out of 16 patients with diabetes mellitus died (12.5%), compared with two out of 85 patients who had no evidence of diabetes (2.3%). A multivariate analysis confirmed the presence of diabetes mellitus as the single independent factor influencing mortality.
All patients who recovered were able to resume their normal life. Sequelae presenting as a marked pleural thickening persisted in 15 patients, six of whom had had only drainage and irrigation. The median overall duration of hospitalization was 14 days (mean 21.1 days; SEM 1.7 days). It was 10 days (mean 13.1 days; SEM 0.8 days; range 5–14 days) for decortication, 14 days (mean 16.9 days; SEM 2.3 days; range 3–35 days) for drainage and irrigation, and 24 days (mean 41.3 days; SEM 4.6 days; range 7–102 days) for open-window thoracostomy (see Table 4). With regard to the causes of empyema the shortest hospitalization times were found in metapneumonic empyema (median 14 days). In a multivariate analysis only the fact whether or not a fenestration was carried out had an independent influence on the hospitalization time, whereas the fact, whether definitive surgery had been delayed by pre-treatment did not exert any influence on the length of hospital stay (mean for pre-treated cases 22.9; no pre-treatment 17.8; P>0.05).
|
Discussion
|
|---|
In spite of their indisputable merits, drainage, irrigation and also video-assisted thoracoscopy have their strict limitations in the therapy of empyema. They are likely to be successful in early stages, whereas chronic empyema, multiple loculations, pulmonary abscesses or frankly gangrenous changes will hardly or only belatedly respond to conservative measures. Unfortunately, the latter are the very problems often escaping CT diagnosis and can only be detected if the chest is not opened for inspection and palpation
[6]
[7]. This is why we did not wait for longer than 48 h for the septic condition to subside before we proceeded to fenestration or decortication.
As most other thoracic surgeons we find ourselves confronted with an increasing number of seriously ill patients who have had ineffective pre-treatment of any kind
[7]
[8]
[9] causing prolonged courses of disease and necessitating staged interventions
[7]. The high overall rate of resections and fenestrations in our collective can be explained by the high proportion of advanced stages of empyema. Accordingly, the resection rate was significantly higher in patients who had unsuccessful pre-treatment by thoracocentesis, drainage, or VATS. We were unable, however, to confirm that the duration of hospital stay was positively correlated with the delay of surgery
[7]. This may be due to the fact, that we proceeded quickly to thoracotomy
[5]
[7]
[8]
[10]
[11] or to staged procedures.
Fenestration of the chest wall is a useful procedure in preventing complications deriving from a cellulitis of the chest wall or from widespread gangrene of the pleural surfaces. It can be lifesaving in moribund patients who would not tolerate decortication and in whom there is no indication for drainage
[12]
[13] as well as in postpneumonectomy empyema
[14]
[15]. Though 33 out of 36 patients recovered, the treatment carries some problems: The necessity of positive pressure ventilation in the presence of a collapsible lung and the fact that the procedure is time consuming, and a significantly longer hospitalization have to be considered.
A striking feature was the presence of a wide-spread cellulitis of the chest wall in all five patients who had undergone VATS. Chronic stages of empyema and pulmonary abscesses were found in these cases in whom a decortication in the true sense of speaking would not have been possible. The so-called VATS-decortication which is a matter of breaking loculations or of bluntly removing membranes followed by intrapleural washing is only feasible in the fibrinopurulent stage
[1]
[2]
[3].
Mortality rates of 10–15% have been reported after drainage and irrigation even in early stages
[1]
[16], whereas decortication in series including chronic cases yielded a 4–13% mortality
[7]
[8]
[17]. Considering these results the argument of a high operative risk of decortication cannot be supported. It has to be conceded, that our cases undergoing drainage alone had a 0% mortality rate, however, these patients were a highly selected, small group of early stages, whereas all those who did not quickly respond to drainage proceeded to fenestration or decortication.
In our collective the mortality for decortication with or without resection was 0% increasing to 10.7% only if fenestration became necessary. A 12.5% mortality rate was found in pre-existent diabetes mellitus which was confirmed as a significant predictor of fatal outcome by multivariate analysis, whereas other factors like age, coexisting malignancies, cardiorespiratory disease, psychiatric disorders, cirrhosis or poor nutritional condition did not influence prognosis. This underlines the benefits of a quick relief from sepsis as achieved by thoracotomy even in `high-risk' patients or in those requiring artificial ventilation and/or catecholamines at the time of admission.
Pleural empyema is essentially a benign disease with the very real outlook of a complete restitution of the patient, a condition seldom found in thoracic surgery. It seems unlogical to endanger survival or the chance of an optimal treatment and a rapid healing process by performing `little invasive' procedures at any rate or by `sparing the patient major surgery'. Patients who do not undergo an optimal treatment will deteriorate during the subacute phase due to sepsis and – if they survive – suffer from the sequelae of empyema throughout the rest of their lives
[18]
[19]. If a `conservative' treatment regimen including VATS does not succeed within a few days, the patient should proceed to thoracotomy.
|
Footnotes
|
|---|
Presented at the 11th Annual Meeting of the European Association for Cardio-thoracic Surgery, Copenhagen, Denmark, September 28 – October 1, 1997. 
|
References
|
|---|
- Wait M.A., Sharma S., Hohn J., Dal-Nogare A. A randomized trial of empyema therapy. Chest 1997;111:1548-1551.[Abstract/Free Full Text]
- Silen M.L., Naunheim K.S. Thoracoscopic approach to the management of empyema thoracis. Indications and results. Chest Surg Clin N Am 1996;6:491-499.[Medline]
- Landreneau L.R., Keenan R.J., Hazelrigg S.R., Mack M.J., Naunheim K.S. Thoracoscopy for empyema and hemothorax. Chest 1996;109:18-24.[Abstract/Free Full Text]
- Laisaar T., Puttsepp E., Laisaar V. Early administration of intrapleural streptokinase in the treatment of multiloculated pleural effusions and pleural empyemas. Thorac Cardiovasc Surg 1996;44:252-256.[Medline]
- LeMense C., Strange C., Sahn S.A. Empyema thoracis. Therapeutic management and outcome. Chest 1995;107:1532-1537.[Abstract/Free Full Text]
- Kirsch E., Guckel C., Kaim A., Steinbrich W. The findings and value of computed tomography in pleural empyema. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 1994;161:404-411.[Medline]
- Cham C.W., Haq S.M., Rahamim J. Empyema thoracis: a problem with late referral?. Thorax 1993;48:925-927.[Abstract]
- Galea J.L., De-Souza A., Beggs D., Spyt T. The surgical management of empyema thoracis. J R Coll Surg Edinburgh 1997;42:15-18.[Medline]
- Martella A.T., Santos G.H. Decortication for chronic postpneumonic empyema. J Am Coll Surg 1995;180:573-576.[Medline]
- Pothula V., Krellenstein D.J. Early aggressive surgical management of parapneumonic empyemas. Chest 1994;105:832-836.[Abstract/Free Full Text]
- Thurer R.J. Decortication in thoracic empyema. Indications and surgical technique. Chest Surg Clin N Am 1996;6:461-490.[Medline]
- Smolle-Juttner F., Beuster W., Pinter H., Pierer G., Pongratz M., Friehs G. Open-window thoracostomy in pleural empyema. Eur J Cardiothorac Surg 1992;6:635-638.[Abstract]
- Weissberg D., Refaely Y. Pleural empyema: 24-year experience. Ann Thorac Surg 1996;62:1026-1029.[Abstract/Free Full Text]
- Wong P.S., Goldstraw P. Post-pneumonectomy empyema. Eur J Cardiothorac Surg 1994;8:345-349.[Abstract]
- Deschamps C., Pairolero P.C., Allen M.S., Trastek V.F. Management of postpneumonectomy empyema and bronchopleural fistula. Chest Surg Clin N Am 1996;6:519-527.[Medline]
- Chin N.K., Lim T.K. Controlled trial of intrapleural streptokinase in the treatment of pleural empyema and complicated parapneumonic effusions. Chest 1997;111:275-279.[Abstract/Free Full Text]
- Ferguson A.D., Prescott R.J., Selkon J.B., Watson D., Swinburn C.R. The clinical course and management of thoracic empyema. Q J Med 1996;89:285-289.
- Zeidler D. Life expectancy and disability following pleural empyema. Versicherungsmedizin 1990;42(2):45-49.[Medline]
- Swoboda L., Laule K., Blattmann H., Hasse J. Decortication in chronic pleural empyema. Investigation of lung function based on perfusion scintigraphy. Thorac Cardiovasc Surg 1990;38:359-361.[Medline]
This article has been cited by other articles:
|
|
|
|
 
K. M. Al-Kattan
Management of tuberculous empyema
Eur. J. Cardiothorac. Surg.,
March 1, 2000;
17(3):
251 - 254.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
