U.K. Controlled Trial of Intrapleural Streptokinase for Pleural Infection
http://www.100md.com
《新英格兰医药杂志》
ABSTRACT
Background Intrapleural fibrinolytic agents are used in the drainage of infected pleural-fluid collections. This use is based on small trials that did not have the statistical power to evaluate accurately important clinical outcomes, including safety. We conducted a trial to clarify the therapeutic role of intrapleural streptokinase.
Methods In this double-blind trial, 454 patients with pleural infection (defined by the presence of purulent pleural fluid or pleural fluid with a pH below 7.2 with signs of infection or by proven bacterial invasion of the pleural space) were randomly assigned to receive either intrapleural streptokinase (250,000 IU twice daily for three days) or placebo. Patients received antibiotics and underwent chest-tube drainage, surgery, and other treatment as part of routine care. The number of patients in the two groups who had died or needed surgical drainage at three months was compared (the primary end point); secondary end points were the rates of death and of surgery (analyzed separately), the radiographic outcome, and the length of the hospital stay.
Results The groups were well matched at baseline. Among the 427 patients who received streptokinase or placebo, there was no significant difference between the groups in the proportion of patients who died or needed surgery (with streptokinase: 64 of 206 patients ; with placebo: 60 of 221 ; relative risk, 1.14 ; P=0.08).
Conclusions The intrapleural administration of streptokinase does not improve mortality, the rate of surgery, or the length of the hospital stay among patients with pleural infection.
Pleural infection develops in about 65,000 patients each year in the United States and the United Kingdom.1 Approximately 15 percent of patients die,2 which is similar to the death rate among patients hospitalized with pneumonia,3,4 and 15 to 40 percent require surgical drainage of the infected pleural space.2,5 The median duration of inpatient care is 15 days, with 20 percent of patients remaining in the hospital for a month or longer.2
Apart from antibiotic therapy, treatment in patients with pleural infection consists mainly of drainage of the infected pleural fluid, and the intrapleural administration of fibrinolytic drugs is widely used in an attempt to reduce the need for surgery to achieve this drainage. Such therapy is intended to lyse the fibrinous septations within infected pleural-fluid collections and is supported by management guidelines.6,7 Small trials8,9,10,11,12 and case series13 have suggested that these agents improve drainage of pleural fluid, as quantified radiographically, and may also reduce the need for surgery,12 with few adverse effects. These studies have had low statistical power,14 however, and have therefore been unable to assess accurately whether these benefits translate into a reduction in mortality or in the frequency with which patients require surgical drainage.
We report the results of the First Multicenter Intrapleural Sepsis Trial (MIST1), which was funded by the U.K. Medical Research Council and supported by the British Thoracic Society. The aim of this trial was to define the role of intrapleural streptokinase in the treatment of pleural infection. The primary end point was the proportion of patients who died or required surgical drainage for pleural infection when treated with intrapleural streptokinase or matching placebo, given as an adjunct to normal care. Secondary end points included mortality and the frequency of surgery, analyzed separately, the length of the hospital stay, the proportion of patients with improvement on chest radiography, and dynamic lung volumes.
Methods
Study Design
This study was a double-blind, randomized trial performed in 52 centers in the United Kingdom that together make up the MIST1 Group. This group includes 27 teaching hospitals and 25 local hospitals; 18 of the centers had on-site facilities for thoracic surgery.
Inclusion and Exclusion Criteria
The inclusion criteria for the trial were the presence of pleural fluid that was macroscopically purulent, that was positive on culture for bacterial infection, that was positive for bacteria on Gram's staining, or that had a pH below 7.2 in a patient with clinical evidence of infection. Evidence of infection was assessed by the recruiting physician on the basis of clinical indicators such as fever, an elevated white-cell count, and an elevated serum level of C-reactive protein. Further details of the inclusion and exclusion criteria appear in the Supplementary Appendix, available with the full text of this article at www.nejm.org.
Chest-Tube Drainage and Antibiotic Therapy
All patients underwent chest-tube drainage and received intravenous antibiotics. Antibiotics were chosen by the managing clinician (usually the principal investigator at the medical center), in line with local microbiologic advice. Recommended regimens were listed in the protocol (see the Supplementary Appendix).
Randomization
The randomization code was prepared and held centrally by the trial statistician. After written informed consent was obtained, patients were randomly assigned to treatment groups by means of a telephone call to the study center. Subjects received either 250,000 IU of streptokinase (Streptase, Aventis) or a matching placebo, both administered in 30 ml of normal saline delivered into the pleural space through the chest tube every 12 hours for six doses (for details, see the Supplementary Appendix).
Primary End Point
The primary outcome studied was the number of patients who died or required surgical drainage of the infected pleural fluid during the three months after randomization. Referral for surgical drainage was made by the recruiting physician on the basis of a substantial residual pleural-fluid collection and evidence of persistent infection (e.g., fever or a persistent, marked elevation in the blood white-cell count or serum C-reactive protein level). Insertion, reinsertion, and repositioning of the chest tube were not classified as surgery when the end point was analyzed.
Secondary End Points
The secondary end points were the rate of death or surgical drainage 12 months after randomization; the rates of death and surgery, analyzed separately, at 3 and 12 months; the duration of the hospital stay; the severity of any residual abnormality on the chest radiograph 3 months after randomization; dynamic lung volumes 3 months after randomization; bleeding after surgery to drain empyema; and changes in the levels of antistreptokinase antibody from baseline to 3 months. Detailed definitions of these secondary end points are available in the Supplementary Appendix.
Statistical Analysis
A written analysis plan was prepared before the data were analyzed, and the analysis was performed in the Medical Research Council Clinical Trials Unit. It was decided before the data were analyzed that subjects who did not receive the assigned study drug because of death, surgery, or withdrawal of consent before the study drug arrived at their hospitals would be excluded from the primary analysis. The primary analysis was repeated with all subjects included. Outcomes were compared with use of relative risks (with Taylor series 95 percent confidence intervals), the Yates' corrected chi-square test, the unpaired t-test, the Mann–Whitney test, or analysis of variance, depending on whether the data were normally distributed (the data were analyzed with the use of Stata software, version 815). The power calculations assumed an event rate of 30 percent with placebo and a reduction of this rate by 40 percent, to 18 percent, with intrapleural streptokinase; 450 patients were required for analysis, with an alpha value of 0.05, a power of 80 percent, and a 5 percent rate of noncompliance. Further details of the data analysis and power calculation are available in the Supplementary Appendix.
Aventis UK provided the streptokinase and placebo for the trial. The United Kingdom Medical Research Council provided funding for the trial. The British Thoracic Society promoted the trial. Neither the company nor these organizations had any influence on the design or execution of the trial or on decisions relating to publication.
Results
Patients
The demographic and clinical characteristics of the patients as recorded at baseline are shown in Table 1. Figure 1 shows the enrollment, assignment, and follow-up of the patients in the study. Three patients (1 percent) were lost to follow-up at three months — a lower rate than the 10 percent originally estimated. Recruitment was therefore halted when the number of patients for whom data could be analyzed reached the required size of 454. Twenty-four patients died, required surgery, or withdrew consent to the trial before receiving the study drug. Thus, the main analysis included 430 subjects (208 of whom received streptokinase, and 222 placebo). Follow-up for the primary outcome analysis was complete for 427 of these patients (99 percent). Fifty-four patients did not receive all six doses of the study drug. Most of these patients ceased to receive the assigned treatment because of an adverse event or because the chest tube became displaced at a time when residual pleural fluid was minimal, so that further chest drainage was not necessary.
View this table:
Table 1. Baseline Characteristics of the Patients, According to Study Group.
Figure 1. Enrollment, Randomization, and Follow-up in the Trial.
Data on the duration of the hospital stay were available for all 430 subjects, and radiographic data were available for 373 patients (86.7 percent) at three months. Baseline data on dynamic lung volumes were available for only 70 patients (16 percent), since most patients were too ill for this assessment to be performed. It was thought that this small sample would not be representative, and these data were therefore not analyzed. Dynamic lung volumes were available for 249 patients (58 percent) at three months, and these results were analyzed.
The characteristics of the two study groups were similar, and the bacteriologic results were typical of pathogens isolated from patients with empyema in the United Kingdom (Table 1). These characteristics were also similar in the subgroups of patients for whom data on radiographic outcomes and lung function were available.
Study Drugs
Independent assays of three study-drug vials expected to contain 250,000 IU of streptokinase confirmed that they contained the expected amounts (263,000 IU, 256,000 IU, and 298,000 IU); three placebo vials contained no active drug. In vitro assays confirmed high levels of fibrinolytic activity in the study vials containing streptokinase and no fibrinolytic activity in the vials of placebo.
Adverse Events
There was an excess rate of serious adverse events in the streptokinase group. Fourteen patients receiving streptokinase (7 percent) had such events, as compared with six (3 percent) in the placebo group (relative risk, 2.49 ; P=0.08) (Table 2). No patients had more than one adverse event.
View this table:
Table 2. Adverse Events Possibly Attributable to Study Treatment, According to Severity and Type.
Primary End Point
There was no statistically or clinically significant difference between the groups in the proportion of patients who required surgical drainage or who died in the three months after randomization. In the streptokinase group, 64 of 206 patients (31 percent) had one of these outcomes, as did 60 of 221 patients (27 percent) in the placebo group (relative risk, 1.14 ; P=0.43) (Figure 2).
Figure 2. Proportion of Patients Surviving without Requiring Pleural-Drainage Surgery (Panel A) and the Proportion Remaining Hospitalized during Follow-up (Panel B).
Secondary End Points
Death and Surgical Drainage Combined
At 12 months after randomization, there remained no difference in the proportion of patients who required surgical drainage or died. In the streptokinase group, 79 of 198 patients (40 percent) had one of these outcomes, as did 73 of 216 patients (34 percent) in the placebo group (relative risk, 1.18 ; P=0.24) (Figure 2).
Death and Surgical Drainage Analyzed Separately
When death and the need for surgical drainage were analyzed separately, there were no differences between the groups at 3 or 12 months. At three months, 32 of 206 patients (16 percent) in the streptokinase group and 30 of 221 (14 percent) in the placebo group had died (relative risk, 1.14 ; P=0.66). Also at three months, 32 of 206 patients (16 percent) in the streptokinase group and 32 of 221 (14 percent) in the placebo group had required surgery (relative risk, 1.07 ; P=0.87). At 12 months, 45 of 198 patients in the streptokinase group (23 percent) and 44 of 216 in the placebo group (20 percent) had died (relative risk, 1.12 ; P=0.64), and 36 of 198 in the streptokinase group (18 percent) and 34 of 216 in the placebo group (16 percent) had needed surgery (relative risk, 1.16 ; P=0.60). Seven patients (two in the streptokinase group and five in the placebo group) died after surgery.
Duration of the Hospital Stay
There was no difference in the duration of the hospital stay between the two groups (streptokinase group: median stay, 13 days; range, 1 to 271; placebo group: median, 12 days; range, 2 to 152; P=0.16 by the Mann–Whitney test) (Figure 2).
Outcomes on Chest Radiography
The side on which the pleural infection occurred and the size of the pleural-fluid collection were similar in the two study groups at baseline (Table 1). There was no difference between the groups in the degree of pleural thickening at three months (Table 3). Also at three months, there was a small but significant difference in favor of placebo in the height of the thorax on the side with empyema. There was no difference between the groups in the degree of reduction in the size of the pleural opacity at three months (Table 3).
View this table:
Table 3. Outcomes on Chest Radiography at Three Months.
Among all patients there was a small but significant difference between baseline and three months in the mean (±SD) vertical height of the thorax on the side not affected by disease, suggesting a small change in radiographic projection or patient inspiratory effort (at baseline: right, 22.8±3.4 cm; left, 20.8±3.4; at three months: right, 24.1±2.9 cm; left, 22.2±3.5; difference: right, 1.3 cm ; P<0.001; left, 1.4 cm ; P<0.001).
Dynamic Lung Volumes
The dynamic lung volumes were similar in the two groups at three months. The mean forced expiratory volume in one second was 2.22±0.78 liters in the streptokinase group and 2.18±0.90 liters in the placebo group (difference, –0.04 liter ; P=0.69). The mean forced vital capacity was 3.04±0.93 liters in the streptokinase group and 2.99±1.11 in the placebo group (difference, –0.05 liter ; P=0.70).
Bleeding after Drainage Surgery
Nine patients (5 in the streptokinase group and 4 in the placebo group) underwent video-assisted thoracoscopic drainage, 39 (17 in the streptokinase group and 22 in the placebo group) underwent thoracotomy with pleural decortication, and 9 (5 in the streptokinase group and 4 in the placebo group) underwent another form of open surgical drainage. The surgeon noted excessive perioperative bleeding in three patients (one in the streptokinase group and two in the placebo group). There was no difference in the patients' blood-transfusion requirements (data not shown).
Antistreptokinase Antibodies
The antistreptokinase-antibody level was markedly increased in the patients receiving streptokinase. Among 111 patients in the streptokinase group, the level was 2402±2818 arbitrary units at baseline and 10,707±4470 at three months (difference, 7953±4861 arbitrary units). Among 127 patients in the placebo group, the corresponding values were 2357±3031 arbitrary units at baseline and 2619±3298 arbitrary units at three months (difference, 357±2862 arbitrary units). The difference between the groups was 7956 arbitrary units (95 percent confidence interval, 6592 to 8600; P<0.001).
Subgroup Analyses
There was no evidence of a reduction in the need for surgical drainage or in mortality after the administration of streptokinase in any of the subgroups we analyzed, including those with and those without frankly purulent pleural fluid at baseline and those with and those without definite septation or loculation (details are available in the Supplementary Appendix).
Discussion
This controlled trial in patients with pleural infection demonstrates that intrapleural streptokinase does not reduce mortality, the need for drainage surgery, or the duration of the hospital stay and does not improve outcomes as measured radiographically or by tests of lung function. The results of our analysis exclude with 95 percent confidence the possibility of an improvement of 15 percent or more in the relative risk of death or the need for surgery as a result of the administration of streptokinase. This finding is unexpected, since it has been assumed on the basis of small trials that intrapleural fibrinolytic agents such as streptokinase are beneficial in patients with pleural infection, and since the administration of streptokinase is included in the management guidelines of major respiratory societies.6,7
The consensus in favor of the use of streptokinase arose because intrapleural fibrinolytic agents produced short-term improvements in radiographic outcomes and were apparently effective as rescue therapy after failed chest-tube drainage.8,9,10,11,13 Previous trials have not had adequate statistical power to assess whether these improvements in intermediate surrogate outcomes were associated with improvements in the major outcomes of clinical interest — mortality and the need for surgery. The results of the current trial show that they are not.
In view of the negative results, we confirmed that the trial drugs were active and that patients correctly received the assigned study treatment. The independent analysis of the vials of streptokinase and placebo demonstrated the expected levels of active drug, and the fibrinolytic efficacy of the streptokinase used in this trial was confirmed in vitro. The increase in antistreptokinase-antibody titers after the administration of streptokinase confirms that the trial groups correctly received the assigned treatments. It is unlikely that the negative results are due to an inadequate regimen of treatment with streptokinase. The six-dose regimen, in which a total of 1.5 million units of streptokinase is administered, represents more doses and a larger total dose of the drug than the regimens used in earlier studies examining intermediate surrogate end points, which had positive results.8,10,13 Finally, the overall combined rate of death or the need for surgery was similar to that in previous studies,2,5,16,17 suggesting that our study population was representative.
The subgroup analyses show that the negative results are not explained by the characteristics of the study sample. In particular, there is no evidence that efficacy was influenced by the presence of overtly purulent pleural fluid, the duration of the illness before randomization, the initial chest-tube size, or the presence or absence of pleural-fluid septation or loculation on the baseline chest radiograph.
An a priori decision was made to exclude from our analysis the 24 subjects who received no study drug because they died, required urgent surgery, or withdrew their consent before the delivery of the study drugs to their hospitals. Inclusion of these patients did not change the primary result of the trial (74 of 224 patients in the streptokinase group and 63 of 227 in the placebo group required surgical drainage or died) (relative risk, 1.19 ; P=0.26).
There was a substantial systemic antistreptokinase-antibody response in the group receiving streptokinase. Such a response might inhibit the efficiency of streptokinase given later for a myocardial infarction or venous thromboembolism. Therefore, patients who have received intrapleural streptokinase and later require systemic fibrinolysis should receive a different fibrinolytic agent.
The results of this trial make it possible to define the role of intrapleural fibrinolytic agents in pleural infection. Generally, the use of these drugs should be avoided, since they produce no advantage in long-term outcomes and have some adverse effects. However, studies conducted before this trial have established that fibrinolytic agents do lead to macroscopically effective in vivo lysis of intrapleural fibrin adhesions18 and reduce the volume of infected pleural-fluid collections.8,13 Our study was not designed to readdress this question. It was designed to assess whether streptokinase decreased mortality or the need for surgery. Thus, there may still be a role for fibrinolytic agents in treating the small subgroup of patients who have an exceptionally large, chest-tube–resistant collection of pleural fluid that causes substantial dyspnea, hypoxemia, or hypercapnia by the mechanical impairment of lung function.
It is interesting to consider why streptokinase did not improve long-term outcomes in this trial. The answer is probably that streptokinase alone did not produce sufficient clearance of pleural fluid — possibly because infected pleural fluid is viscous, lumpy, and resistant to tube drainage.19,20 Streptokinase does not improve these characteristics19,20; it simply breaches the barriers between pockets of pus. Therefore, despite the partial unification of the pleural collection, the fluid may still fail to drain adequately, because it cannot pass down the chest tube. This fact might explain the particular efficacy of pleural-fluid drainage with the aid of fibrinolytic agents in patients with loculated malignant pleural effusion,21 in whom the pleural fluid is thin and only the fibrin septations inhibit the removal of fluid. If this is true, then reducing the viscosity of pus may improve outcomes in pleural infection; DNase is a candidate for this role, since it reduces viscosity by fragmenting the free uncoiled DNA found in pus.19,20 This is the rationale for its use in airway clearance in cystic fibrosis. In pleural infection, data from both in vitro studies19,20 and case reports on clinical use22 suggest that DNase may also be helpful in draining empyema. Appropriate trials are needed to explore this possibility.
An alternative, and challenging, explanation for the results of this trial might be that reducing the volume of infected pleural fluid is less important than is widely believed in patients receiving modern antibiotics. This possibility would call into question the preeminent role assigned to the drainage of thoracic pus for 2500 years23 and thus should be considered only a speculative hypothesis requiring further research.
In summary, in this trial, intrapleural streptokinase had a modest adverse-event profile in patients with pleural infection but was ineffective in reducing mortality, the need for surgical drainage, or the length of the hospital stay. We conclude that the use of intrapleural streptokinase should generally be avoided in pleural infection.
Supported by a grant (G9721289) from the United Kingdom Medical Research Council. The trial drugs were supplied by Aventis UK.
* Participants in the MIST1 Group are listed in the Appendix.
Source Information
From the Oxford Centre for Respiratory Medicine, Churchill Hospital Site, Oxford Radcliffe Hospital, Headington, Oxford (N.A.M., E.L.H., F.V.G., G.L.R., T.E.A.P., D.J.L., R.J.O.D.); the Department of Respiratory Medicine, Royal Berkshire Hospital, Reading (C.W.H.D.); the Medical Research Council Clinical Trials Unit, London (A.J.N., R.G., J.H.D.); the Department of Population Sciences and Primary Care, University College, London (R.M.); and the Department of Respiratory Medicine, Manchester Royal Infirmary, Manchester (M.A.W.) — all in the United Kingdom.
Address reprint requests to Dr. Davies at the Oxford Centre for Respiratory Medicine, Oxford Radcliffe Hospital, Headington, Oxford OX3 7LJ, United Kingdom, or at robert.davies@ndm.ox.ac.uk.
References
Light RW, Girard WM, Jenkinson SG, George RB. Parapneumonic effusions. Am J Med 1980;69:507-512.
Davies CW, Kearney SE, Gleeson FV, Davies RJ. Predictors of outcome and long-term survival in patients with pleural infection. Am J Respir Crit Care Med 1999;160:1682-1687.
Fine MJ, Smith MA, Carson CA, et al. Prognosis and outcomes of patients with community-acquired pneumonia: a meta-analysis. JAMA 1996;275:134-141.
Lim WS, Macfarlane JT, Boswell TCJ, et al. Study of community acquired pneumonia aetiology (SCAPA) in adults admitted to hospital: implications for management guidelines. Thorax 2001;56:296-301.
Ferguson AD, Prescott RJ, Selkon JB, Watson D, Swinburn CR. The clinical course and management of thoracic empyema. QJM 1996;89:285-289.
Davies CWH, Gleeson FV, Davies RJO. BTS guidelines for the management of pleural infection. Thorax 2003;58:Suppl 2:18-28.
Colice GL, Curtis A, Deslauriers J, et al. Medical and surgical treatment of parapneumonic effusions: an evidence-based guideline. Chest 2000;118:1158-1171.
Davies RJO, Traill ZC, Gleeson FV. Randomised controlled trial of intrapleural streptokinase in community acquired pleural infection. Thorax 1997;52:416-421.
Bouros D, Schiza S, Tzanakis N, Chalkiadakis G, Drositis J, Siafakas N. Intrapleural urokinase versus normal saline in the treatment of complicated parapneumonic effusions and empyema: a randomized, double-blind study. Am J Respir Crit Care Med 1999;159:37-42.
Bouros D, Schiza S, Patsourakis G, Chalkkkiadakis G, Panagou P, Siafakas NM. Intrapleural streptokinase versus urokinase in the treatment of complicated parapneumonic effusions: a prospective, double-blind study. Am J Respir Crit Care Med 1997;155:291-295.
Thomson AH, Hull J, Kumar MR, Wallis C, Balfour Lynn IM. Randomised trial of intrapleural urokinase in the treatment of childhood empyema. Thorax 2002;57:343-347.
Diacon AH, Theron J, Schuurmans MM, Van der Wal BW, Bolliger CT. Intrapleural streptokinase for empyema and complicated parapneumonic effusions. Am J Respir Crit Care Med 2004;170:49-53.
Maskell NA, Davies RJO. Effusions from parapneumonic infection and empyema. In: Light RW, Lee YCG, eds. Textbook of pleural diseases. London: Arnold, 2003:310-28.
Lee YCG. Ongoing search for effective intrapleural therapy for empyema: is streptokinase the answer? Am J Respir Crit Care Med 2004;170:1-2.
StataCorp. Stata statistical software, release 8. College Station, Tex.: StataCorp L.P., 2003.
Chin NK, Lim TK. Controlled trial of intrapleural streptokinase in the treatment of pleural empyema and complicated parapneumonic effusions. Chest 1997;111:275-279.
Lim TK, Chin NK. Empirical treatment with fibrinolysis and early surgery reduces the duration of hospitalization in pleural sepsis. Eur Respir J 1999;13:514-518.
Maskell NA, Gleeson FV. Effect of intrapleural streptokinase on a loculated malignant pleural effusion. N Engl J Med 2003;348:e4-e4.
Light RW, Nguyen T, Mulligan ME, Sasse SA. The in vitro efficacy of varidase versus streptokinase or urokinase for liquefying thick purulent exudative material from loculated empyema. Lung 2000;178:13-18.
Simpson G, Roomes D, Heron M. Effects of streptokinase and deoxyribonuclease on viscosity of human surgical and empyema pus. Chest 2000;117:1728-1733.
Davies CW, Traill ZC, Gleeson FV, Davies RJ. Intrapleural streptokinase in the management of malignant multiloculated pleural effusions. Chest 1999;115:729-733.
Simpson G, Roomes D, Reeves B. Successful treatment of empyema thoracis with human recombinant deoxyribonuclease. Thorax 2003;58:365-366.
Hippocrates. Aphorism 44. In: Adams LB, ed. The genuine works of Hippocrates: the classics of surgery. Birmingham, Ala.: Gryphon Editions, 1985:768-71.(Nicholas A. Maskell, M.R.)
Background Intrapleural fibrinolytic agents are used in the drainage of infected pleural-fluid collections. This use is based on small trials that did not have the statistical power to evaluate accurately important clinical outcomes, including safety. We conducted a trial to clarify the therapeutic role of intrapleural streptokinase.
Methods In this double-blind trial, 454 patients with pleural infection (defined by the presence of purulent pleural fluid or pleural fluid with a pH below 7.2 with signs of infection or by proven bacterial invasion of the pleural space) were randomly assigned to receive either intrapleural streptokinase (250,000 IU twice daily for three days) or placebo. Patients received antibiotics and underwent chest-tube drainage, surgery, and other treatment as part of routine care. The number of patients in the two groups who had died or needed surgical drainage at three months was compared (the primary end point); secondary end points were the rates of death and of surgery (analyzed separately), the radiographic outcome, and the length of the hospital stay.
Results The groups were well matched at baseline. Among the 427 patients who received streptokinase or placebo, there was no significant difference between the groups in the proportion of patients who died or needed surgery (with streptokinase: 64 of 206 patients ; with placebo: 60 of 221 ; relative risk, 1.14 ; P=0.08).
Conclusions The intrapleural administration of streptokinase does not improve mortality, the rate of surgery, or the length of the hospital stay among patients with pleural infection.
Pleural infection develops in about 65,000 patients each year in the United States and the United Kingdom.1 Approximately 15 percent of patients die,2 which is similar to the death rate among patients hospitalized with pneumonia,3,4 and 15 to 40 percent require surgical drainage of the infected pleural space.2,5 The median duration of inpatient care is 15 days, with 20 percent of patients remaining in the hospital for a month or longer.2
Apart from antibiotic therapy, treatment in patients with pleural infection consists mainly of drainage of the infected pleural fluid, and the intrapleural administration of fibrinolytic drugs is widely used in an attempt to reduce the need for surgery to achieve this drainage. Such therapy is intended to lyse the fibrinous septations within infected pleural-fluid collections and is supported by management guidelines.6,7 Small trials8,9,10,11,12 and case series13 have suggested that these agents improve drainage of pleural fluid, as quantified radiographically, and may also reduce the need for surgery,12 with few adverse effects. These studies have had low statistical power,14 however, and have therefore been unable to assess accurately whether these benefits translate into a reduction in mortality or in the frequency with which patients require surgical drainage.
We report the results of the First Multicenter Intrapleural Sepsis Trial (MIST1), which was funded by the U.K. Medical Research Council and supported by the British Thoracic Society. The aim of this trial was to define the role of intrapleural streptokinase in the treatment of pleural infection. The primary end point was the proportion of patients who died or required surgical drainage for pleural infection when treated with intrapleural streptokinase or matching placebo, given as an adjunct to normal care. Secondary end points included mortality and the frequency of surgery, analyzed separately, the length of the hospital stay, the proportion of patients with improvement on chest radiography, and dynamic lung volumes.
Methods
Study Design
This study was a double-blind, randomized trial performed in 52 centers in the United Kingdom that together make up the MIST1 Group. This group includes 27 teaching hospitals and 25 local hospitals; 18 of the centers had on-site facilities for thoracic surgery.
Inclusion and Exclusion Criteria
The inclusion criteria for the trial were the presence of pleural fluid that was macroscopically purulent, that was positive on culture for bacterial infection, that was positive for bacteria on Gram's staining, or that had a pH below 7.2 in a patient with clinical evidence of infection. Evidence of infection was assessed by the recruiting physician on the basis of clinical indicators such as fever, an elevated white-cell count, and an elevated serum level of C-reactive protein. Further details of the inclusion and exclusion criteria appear in the Supplementary Appendix, available with the full text of this article at www.nejm.org.
Chest-Tube Drainage and Antibiotic Therapy
All patients underwent chest-tube drainage and received intravenous antibiotics. Antibiotics were chosen by the managing clinician (usually the principal investigator at the medical center), in line with local microbiologic advice. Recommended regimens were listed in the protocol (see the Supplementary Appendix).
Randomization
The randomization code was prepared and held centrally by the trial statistician. After written informed consent was obtained, patients were randomly assigned to treatment groups by means of a telephone call to the study center. Subjects received either 250,000 IU of streptokinase (Streptase, Aventis) or a matching placebo, both administered in 30 ml of normal saline delivered into the pleural space through the chest tube every 12 hours for six doses (for details, see the Supplementary Appendix).
Primary End Point
The primary outcome studied was the number of patients who died or required surgical drainage of the infected pleural fluid during the three months after randomization. Referral for surgical drainage was made by the recruiting physician on the basis of a substantial residual pleural-fluid collection and evidence of persistent infection (e.g., fever or a persistent, marked elevation in the blood white-cell count or serum C-reactive protein level). Insertion, reinsertion, and repositioning of the chest tube were not classified as surgery when the end point was analyzed.
Secondary End Points
The secondary end points were the rate of death or surgical drainage 12 months after randomization; the rates of death and surgery, analyzed separately, at 3 and 12 months; the duration of the hospital stay; the severity of any residual abnormality on the chest radiograph 3 months after randomization; dynamic lung volumes 3 months after randomization; bleeding after surgery to drain empyema; and changes in the levels of antistreptokinase antibody from baseline to 3 months. Detailed definitions of these secondary end points are available in the Supplementary Appendix.
Statistical Analysis
A written analysis plan was prepared before the data were analyzed, and the analysis was performed in the Medical Research Council Clinical Trials Unit. It was decided before the data were analyzed that subjects who did not receive the assigned study drug because of death, surgery, or withdrawal of consent before the study drug arrived at their hospitals would be excluded from the primary analysis. The primary analysis was repeated with all subjects included. Outcomes were compared with use of relative risks (with Taylor series 95 percent confidence intervals), the Yates' corrected chi-square test, the unpaired t-test, the Mann–Whitney test, or analysis of variance, depending on whether the data were normally distributed (the data were analyzed with the use of Stata software, version 815). The power calculations assumed an event rate of 30 percent with placebo and a reduction of this rate by 40 percent, to 18 percent, with intrapleural streptokinase; 450 patients were required for analysis, with an alpha value of 0.05, a power of 80 percent, and a 5 percent rate of noncompliance. Further details of the data analysis and power calculation are available in the Supplementary Appendix.
Aventis UK provided the streptokinase and placebo for the trial. The United Kingdom Medical Research Council provided funding for the trial. The British Thoracic Society promoted the trial. Neither the company nor these organizations had any influence on the design or execution of the trial or on decisions relating to publication.
Results
Patients
The demographic and clinical characteristics of the patients as recorded at baseline are shown in Table 1. Figure 1 shows the enrollment, assignment, and follow-up of the patients in the study. Three patients (1 percent) were lost to follow-up at three months — a lower rate than the 10 percent originally estimated. Recruitment was therefore halted when the number of patients for whom data could be analyzed reached the required size of 454. Twenty-four patients died, required surgery, or withdrew consent to the trial before receiving the study drug. Thus, the main analysis included 430 subjects (208 of whom received streptokinase, and 222 placebo). Follow-up for the primary outcome analysis was complete for 427 of these patients (99 percent). Fifty-four patients did not receive all six doses of the study drug. Most of these patients ceased to receive the assigned treatment because of an adverse event or because the chest tube became displaced at a time when residual pleural fluid was minimal, so that further chest drainage was not necessary.
View this table:
Table 1. Baseline Characteristics of the Patients, According to Study Group.
Figure 1. Enrollment, Randomization, and Follow-up in the Trial.
Data on the duration of the hospital stay were available for all 430 subjects, and radiographic data were available for 373 patients (86.7 percent) at three months. Baseline data on dynamic lung volumes were available for only 70 patients (16 percent), since most patients were too ill for this assessment to be performed. It was thought that this small sample would not be representative, and these data were therefore not analyzed. Dynamic lung volumes were available for 249 patients (58 percent) at three months, and these results were analyzed.
The characteristics of the two study groups were similar, and the bacteriologic results were typical of pathogens isolated from patients with empyema in the United Kingdom (Table 1). These characteristics were also similar in the subgroups of patients for whom data on radiographic outcomes and lung function were available.
Study Drugs
Independent assays of three study-drug vials expected to contain 250,000 IU of streptokinase confirmed that they contained the expected amounts (263,000 IU, 256,000 IU, and 298,000 IU); three placebo vials contained no active drug. In vitro assays confirmed high levels of fibrinolytic activity in the study vials containing streptokinase and no fibrinolytic activity in the vials of placebo.
Adverse Events
There was an excess rate of serious adverse events in the streptokinase group. Fourteen patients receiving streptokinase (7 percent) had such events, as compared with six (3 percent) in the placebo group (relative risk, 2.49 ; P=0.08) (Table 2). No patients had more than one adverse event.
View this table:
Table 2. Adverse Events Possibly Attributable to Study Treatment, According to Severity and Type.
Primary End Point
There was no statistically or clinically significant difference between the groups in the proportion of patients who required surgical drainage or who died in the three months after randomization. In the streptokinase group, 64 of 206 patients (31 percent) had one of these outcomes, as did 60 of 221 patients (27 percent) in the placebo group (relative risk, 1.14 ; P=0.43) (Figure 2).
Figure 2. Proportion of Patients Surviving without Requiring Pleural-Drainage Surgery (Panel A) and the Proportion Remaining Hospitalized during Follow-up (Panel B).
Secondary End Points
Death and Surgical Drainage Combined
At 12 months after randomization, there remained no difference in the proportion of patients who required surgical drainage or died. In the streptokinase group, 79 of 198 patients (40 percent) had one of these outcomes, as did 73 of 216 patients (34 percent) in the placebo group (relative risk, 1.18 ; P=0.24) (Figure 2).
Death and Surgical Drainage Analyzed Separately
When death and the need for surgical drainage were analyzed separately, there were no differences between the groups at 3 or 12 months. At three months, 32 of 206 patients (16 percent) in the streptokinase group and 30 of 221 (14 percent) in the placebo group had died (relative risk, 1.14 ; P=0.66). Also at three months, 32 of 206 patients (16 percent) in the streptokinase group and 32 of 221 (14 percent) in the placebo group had required surgery (relative risk, 1.07 ; P=0.87). At 12 months, 45 of 198 patients in the streptokinase group (23 percent) and 44 of 216 in the placebo group (20 percent) had died (relative risk, 1.12 ; P=0.64), and 36 of 198 in the streptokinase group (18 percent) and 34 of 216 in the placebo group (16 percent) had needed surgery (relative risk, 1.16 ; P=0.60). Seven patients (two in the streptokinase group and five in the placebo group) died after surgery.
Duration of the Hospital Stay
There was no difference in the duration of the hospital stay between the two groups (streptokinase group: median stay, 13 days; range, 1 to 271; placebo group: median, 12 days; range, 2 to 152; P=0.16 by the Mann–Whitney test) (Figure 2).
Outcomes on Chest Radiography
The side on which the pleural infection occurred and the size of the pleural-fluid collection were similar in the two study groups at baseline (Table 1). There was no difference between the groups in the degree of pleural thickening at three months (Table 3). Also at three months, there was a small but significant difference in favor of placebo in the height of the thorax on the side with empyema. There was no difference between the groups in the degree of reduction in the size of the pleural opacity at three months (Table 3).
View this table:
Table 3. Outcomes on Chest Radiography at Three Months.
Among all patients there was a small but significant difference between baseline and three months in the mean (±SD) vertical height of the thorax on the side not affected by disease, suggesting a small change in radiographic projection or patient inspiratory effort (at baseline: right, 22.8±3.4 cm; left, 20.8±3.4; at three months: right, 24.1±2.9 cm; left, 22.2±3.5; difference: right, 1.3 cm ; P<0.001; left, 1.4 cm ; P<0.001).
Dynamic Lung Volumes
The dynamic lung volumes were similar in the two groups at three months. The mean forced expiratory volume in one second was 2.22±0.78 liters in the streptokinase group and 2.18±0.90 liters in the placebo group (difference, –0.04 liter ; P=0.69). The mean forced vital capacity was 3.04±0.93 liters in the streptokinase group and 2.99±1.11 in the placebo group (difference, –0.05 liter ; P=0.70).
Bleeding after Drainage Surgery
Nine patients (5 in the streptokinase group and 4 in the placebo group) underwent video-assisted thoracoscopic drainage, 39 (17 in the streptokinase group and 22 in the placebo group) underwent thoracotomy with pleural decortication, and 9 (5 in the streptokinase group and 4 in the placebo group) underwent another form of open surgical drainage. The surgeon noted excessive perioperative bleeding in three patients (one in the streptokinase group and two in the placebo group). There was no difference in the patients' blood-transfusion requirements (data not shown).
Antistreptokinase Antibodies
The antistreptokinase-antibody level was markedly increased in the patients receiving streptokinase. Among 111 patients in the streptokinase group, the level was 2402±2818 arbitrary units at baseline and 10,707±4470 at three months (difference, 7953±4861 arbitrary units). Among 127 patients in the placebo group, the corresponding values were 2357±3031 arbitrary units at baseline and 2619±3298 arbitrary units at three months (difference, 357±2862 arbitrary units). The difference between the groups was 7956 arbitrary units (95 percent confidence interval, 6592 to 8600; P<0.001).
Subgroup Analyses
There was no evidence of a reduction in the need for surgical drainage or in mortality after the administration of streptokinase in any of the subgroups we analyzed, including those with and those without frankly purulent pleural fluid at baseline and those with and those without definite septation or loculation (details are available in the Supplementary Appendix).
Discussion
This controlled trial in patients with pleural infection demonstrates that intrapleural streptokinase does not reduce mortality, the need for drainage surgery, or the duration of the hospital stay and does not improve outcomes as measured radiographically or by tests of lung function. The results of our analysis exclude with 95 percent confidence the possibility of an improvement of 15 percent or more in the relative risk of death or the need for surgery as a result of the administration of streptokinase. This finding is unexpected, since it has been assumed on the basis of small trials that intrapleural fibrinolytic agents such as streptokinase are beneficial in patients with pleural infection, and since the administration of streptokinase is included in the management guidelines of major respiratory societies.6,7
The consensus in favor of the use of streptokinase arose because intrapleural fibrinolytic agents produced short-term improvements in radiographic outcomes and were apparently effective as rescue therapy after failed chest-tube drainage.8,9,10,11,13 Previous trials have not had adequate statistical power to assess whether these improvements in intermediate surrogate outcomes were associated with improvements in the major outcomes of clinical interest — mortality and the need for surgery. The results of the current trial show that they are not.
In view of the negative results, we confirmed that the trial drugs were active and that patients correctly received the assigned study treatment. The independent analysis of the vials of streptokinase and placebo demonstrated the expected levels of active drug, and the fibrinolytic efficacy of the streptokinase used in this trial was confirmed in vitro. The increase in antistreptokinase-antibody titers after the administration of streptokinase confirms that the trial groups correctly received the assigned treatments. It is unlikely that the negative results are due to an inadequate regimen of treatment with streptokinase. The six-dose regimen, in which a total of 1.5 million units of streptokinase is administered, represents more doses and a larger total dose of the drug than the regimens used in earlier studies examining intermediate surrogate end points, which had positive results.8,10,13 Finally, the overall combined rate of death or the need for surgery was similar to that in previous studies,2,5,16,17 suggesting that our study population was representative.
The subgroup analyses show that the negative results are not explained by the characteristics of the study sample. In particular, there is no evidence that efficacy was influenced by the presence of overtly purulent pleural fluid, the duration of the illness before randomization, the initial chest-tube size, or the presence or absence of pleural-fluid septation or loculation on the baseline chest radiograph.
An a priori decision was made to exclude from our analysis the 24 subjects who received no study drug because they died, required urgent surgery, or withdrew their consent before the delivery of the study drugs to their hospitals. Inclusion of these patients did not change the primary result of the trial (74 of 224 patients in the streptokinase group and 63 of 227 in the placebo group required surgical drainage or died) (relative risk, 1.19 ; P=0.26).
There was a substantial systemic antistreptokinase-antibody response in the group receiving streptokinase. Such a response might inhibit the efficiency of streptokinase given later for a myocardial infarction or venous thromboembolism. Therefore, patients who have received intrapleural streptokinase and later require systemic fibrinolysis should receive a different fibrinolytic agent.
The results of this trial make it possible to define the role of intrapleural fibrinolytic agents in pleural infection. Generally, the use of these drugs should be avoided, since they produce no advantage in long-term outcomes and have some adverse effects. However, studies conducted before this trial have established that fibrinolytic agents do lead to macroscopically effective in vivo lysis of intrapleural fibrin adhesions18 and reduce the volume of infected pleural-fluid collections.8,13 Our study was not designed to readdress this question. It was designed to assess whether streptokinase decreased mortality or the need for surgery. Thus, there may still be a role for fibrinolytic agents in treating the small subgroup of patients who have an exceptionally large, chest-tube–resistant collection of pleural fluid that causes substantial dyspnea, hypoxemia, or hypercapnia by the mechanical impairment of lung function.
It is interesting to consider why streptokinase did not improve long-term outcomes in this trial. The answer is probably that streptokinase alone did not produce sufficient clearance of pleural fluid — possibly because infected pleural fluid is viscous, lumpy, and resistant to tube drainage.19,20 Streptokinase does not improve these characteristics19,20; it simply breaches the barriers between pockets of pus. Therefore, despite the partial unification of the pleural collection, the fluid may still fail to drain adequately, because it cannot pass down the chest tube. This fact might explain the particular efficacy of pleural-fluid drainage with the aid of fibrinolytic agents in patients with loculated malignant pleural effusion,21 in whom the pleural fluid is thin and only the fibrin septations inhibit the removal of fluid. If this is true, then reducing the viscosity of pus may improve outcomes in pleural infection; DNase is a candidate for this role, since it reduces viscosity by fragmenting the free uncoiled DNA found in pus.19,20 This is the rationale for its use in airway clearance in cystic fibrosis. In pleural infection, data from both in vitro studies19,20 and case reports on clinical use22 suggest that DNase may also be helpful in draining empyema. Appropriate trials are needed to explore this possibility.
An alternative, and challenging, explanation for the results of this trial might be that reducing the volume of infected pleural fluid is less important than is widely believed in patients receiving modern antibiotics. This possibility would call into question the preeminent role assigned to the drainage of thoracic pus for 2500 years23 and thus should be considered only a speculative hypothesis requiring further research.
In summary, in this trial, intrapleural streptokinase had a modest adverse-event profile in patients with pleural infection but was ineffective in reducing mortality, the need for surgical drainage, or the length of the hospital stay. We conclude that the use of intrapleural streptokinase should generally be avoided in pleural infection.
Supported by a grant (G9721289) from the United Kingdom Medical Research Council. The trial drugs were supplied by Aventis UK.
* Participants in the MIST1 Group are listed in the Appendix.
Source Information
From the Oxford Centre for Respiratory Medicine, Churchill Hospital Site, Oxford Radcliffe Hospital, Headington, Oxford (N.A.M., E.L.H., F.V.G., G.L.R., T.E.A.P., D.J.L., R.J.O.D.); the Department of Respiratory Medicine, Royal Berkshire Hospital, Reading (C.W.H.D.); the Medical Research Council Clinical Trials Unit, London (A.J.N., R.G., J.H.D.); the Department of Population Sciences and Primary Care, University College, London (R.M.); and the Department of Respiratory Medicine, Manchester Royal Infirmary, Manchester (M.A.W.) — all in the United Kingdom.
Address reprint requests to Dr. Davies at the Oxford Centre for Respiratory Medicine, Oxford Radcliffe Hospital, Headington, Oxford OX3 7LJ, United Kingdom, or at robert.davies@ndm.ox.ac.uk.
References
Light RW, Girard WM, Jenkinson SG, George RB. Parapneumonic effusions. Am J Med 1980;69:507-512.
Davies CW, Kearney SE, Gleeson FV, Davies RJ. Predictors of outcome and long-term survival in patients with pleural infection. Am J Respir Crit Care Med 1999;160:1682-1687.
Fine MJ, Smith MA, Carson CA, et al. Prognosis and outcomes of patients with community-acquired pneumonia: a meta-analysis. JAMA 1996;275:134-141.
Lim WS, Macfarlane JT, Boswell TCJ, et al. Study of community acquired pneumonia aetiology (SCAPA) in adults admitted to hospital: implications for management guidelines. Thorax 2001;56:296-301.
Ferguson AD, Prescott RJ, Selkon JB, Watson D, Swinburn CR. The clinical course and management of thoracic empyema. QJM 1996;89:285-289.
Davies CWH, Gleeson FV, Davies RJO. BTS guidelines for the management of pleural infection. Thorax 2003;58:Suppl 2:18-28.
Colice GL, Curtis A, Deslauriers J, et al. Medical and surgical treatment of parapneumonic effusions: an evidence-based guideline. Chest 2000;118:1158-1171.
Davies RJO, Traill ZC, Gleeson FV. Randomised controlled trial of intrapleural streptokinase in community acquired pleural infection. Thorax 1997;52:416-421.
Bouros D, Schiza S, Tzanakis N, Chalkiadakis G, Drositis J, Siafakas N. Intrapleural urokinase versus normal saline in the treatment of complicated parapneumonic effusions and empyema: a randomized, double-blind study. Am J Respir Crit Care Med 1999;159:37-42.
Bouros D, Schiza S, Patsourakis G, Chalkkkiadakis G, Panagou P, Siafakas NM. Intrapleural streptokinase versus urokinase in the treatment of complicated parapneumonic effusions: a prospective, double-blind study. Am J Respir Crit Care Med 1997;155:291-295.
Thomson AH, Hull J, Kumar MR, Wallis C, Balfour Lynn IM. Randomised trial of intrapleural urokinase in the treatment of childhood empyema. Thorax 2002;57:343-347.
Diacon AH, Theron J, Schuurmans MM, Van der Wal BW, Bolliger CT. Intrapleural streptokinase for empyema and complicated parapneumonic effusions. Am J Respir Crit Care Med 2004;170:49-53.
Maskell NA, Davies RJO. Effusions from parapneumonic infection and empyema. In: Light RW, Lee YCG, eds. Textbook of pleural diseases. London: Arnold, 2003:310-28.
Lee YCG. Ongoing search for effective intrapleural therapy for empyema: is streptokinase the answer? Am J Respir Crit Care Med 2004;170:1-2.
StataCorp. Stata statistical software, release 8. College Station, Tex.: StataCorp L.P., 2003.
Chin NK, Lim TK. Controlled trial of intrapleural streptokinase in the treatment of pleural empyema and complicated parapneumonic effusions. Chest 1997;111:275-279.
Lim TK, Chin NK. Empirical treatment with fibrinolysis and early surgery reduces the duration of hospitalization in pleural sepsis. Eur Respir J 1999;13:514-518.
Maskell NA, Gleeson FV. Effect of intrapleural streptokinase on a loculated malignant pleural effusion. N Engl J Med 2003;348:e4-e4.
Light RW, Nguyen T, Mulligan ME, Sasse SA. The in vitro efficacy of varidase versus streptokinase or urokinase for liquefying thick purulent exudative material from loculated empyema. Lung 2000;178:13-18.
Simpson G, Roomes D, Heron M. Effects of streptokinase and deoxyribonuclease on viscosity of human surgical and empyema pus. Chest 2000;117:1728-1733.
Davies CW, Traill ZC, Gleeson FV, Davies RJ. Intrapleural streptokinase in the management of malignant multiloculated pleural effusions. Chest 1999;115:729-733.
Simpson G, Roomes D, Reeves B. Successful treatment of empyema thoracis with human recombinant deoxyribonuclease. Thorax 2003;58:365-366.
Hippocrates. Aphorism 44. In: Adams LB, ed. The genuine works of Hippocrates: the classics of surgery. Birmingham, Ala.: Gryphon Editions, 1985:768-71.(Nicholas A. Maskell, M.R.)