Open Drainage of Massive Tuberculous Empyema With Progressive Reexpansion of the Lung: An Old Concept Revisited

Transcription

1 Open Drainage of Massive Tuberculous Empyema With Progressive Reexpansion of the Lung: An Old Concept Revisited Sohaila Mohsin Ali, MD, Abdul Aziz Siddiqui, MD, and Joseph S. McLaughlin, MD Department of Surgery, Aga Khan University, Karachi, Pakistan Background. This study examined the results of open drainage of massive tuberculous empyema. Methods. During a 7-year period 47 patients with primary mixed chronic tuberculous empyema with near or total lung collapse were treated. The initial procedure was chest tube suction drainage, which permitted evaluation of the pleural cavity and the lung parenchyma despite minimal if any reexpansion of the lung. All patients were treated with antibiotics and multidrug regimens of antituberculosis agents. A pleurocutaneous window was established by removing sections of two ribs one intercostal space above the base of the pleural cavity. Irrigation was performed daily with dilute povidone iodine solution. Results. Twenty-eight patients achieved complete reexpansion of the lung after 4 to 30 months of drainage I n 1935, Leo Eloesser [1] described an operation for drainage of tuberculous empyema in which a portion of rib was resected and a cutaneous flap was constructed. He noted that reexpansion of the lung occurred, but not in those lungs that had undergone cavitation and fibrosis. He wrongly attributed this reexpansion to the cutaneous flap, which he thought acted as a valve and produced negative pressure in the chest when the patient coughed or expired, but he noted the paradox of this concept. Subsequently there have been anecdotal reports of lung reexpansion with open drainage. This series confirms Eloesser's and later observers' findings, extends this concept to include patients with bronchopleural fistula, and provides an explanation for this phenomenon. In December 1987 a 26-year-old mother of six presented to the Liaquat National Hospital in Karachi, Pakistan, with massive mixed tuberculous empyema (Fig 1). She had been treated with antituberculosis therapy and multiple thoracocenteses in the tribal area of northern Pakistan for 2 years. Reportedly at least a liter of purulent material had been drained at 6-month intervals. She was acutely and chronically ill with a massive left pleural empyema, gross inflammation of the left second and third costochondral joints, and abdominal tuberculosis with ascites. The Mantoux test was positive. Cul- Presented at the Thirty-second Annual Meeting of The Society of Thoracic Surgeons, Orlando, FL, Jan 29-31, Address reprints requests to Dr Ali, Division of Thoracic and Cardiovascula~ Surgery, University of Maryland Medical Center, 22 S Greene St, Baltimore, MD and are cured. Eleven are in various stages of reexpansion and probably will be cured. Eight patients did not achieve reexpansion. Criteria were established retrospectively on an ongoing basis that indicate when pulmonary reexpansion is possible. Conclusions. These totally collapsed "entrapped" lungs expanded to fill the entire pleural space despite the presence of bronchopleural fistulas and an "open" pleura. Reexpansion was progressive, gradual, and dependent on improved compliance, clearing of bronchial inflammation and obstruction, and pleural cleansing. Criteria are established that identify those patients in whom complete reexpansion may take place and the disease may be cured. (Ann Thorac Surg 1996;62:218-24) tures of the chest fluid grew Pseudomonas and ascitic fluid was culture positive for Mycobacterium tuberculosis. A chest tube was inserted and she was treated with appropriate antibiotics and a four-drug regimen of antituberculosis agents. A chest roentgenographic examination showed the lung to be collapsed against the mediastinum. Within a month her condition improved: her fever subsided, the costochondral infection regressed, ascites and partial bowel obstruction resolved, and she began to eat and gain weight. She was considered fit for discharge; however, the lung had not expanded despite application of continuous suction to the chest tube. At 4 months the chest tube was cut off and converted to an "empyema tube" (Fig 1B). By June 1988 the patient was markedly improved and plans were made to perform a decortication and possible lung resection, but the patient refused, Accordingly a pleurocutaneous fistula was established (Fig 1C). The patient was instructed to irrigate the pleural space twice daily with dilute povidone iodine solution. She was seen on a monthly or bimonthly basis until June 1989, when she returned to the northern area with a mature, draining pleurocutaneous window. In October 1989 she returned to Karachi because of mild bleeding from the fistula site. A chest roentgenographic ray examination revealed complete reexpansion of the lung (Fig 1D). The lung surface seen through the chest wall opening was covered with granulation tissue; subsequently this area epithelialized. The foregoing experience led to the treatment by open drainage of 46 similarly affected patients who had mas by The Society of Thoracic Surgeons /$15.00 Published by Elsevier Science Inc PII S (96)

2 Ann Thorac Surg ALI ET AL ;62: OPEN DRAINAGE OF TUBERCULOUS EMPYEMA A C D Fig 1. A 26-year-old woman with massive mixed tuberculous empyema. (A) Initial film reveals complete opacificafion of left lung field with shift of the mediastinum. (B) Four months later with persistently collapsed lung. Chest tube had been converted to empyema tube. (C) Lung showing signs of reexpansion 4 months after pleurocutaneous window. (D) Complete reexpansion at 16 months. sive mixed tuberculous empyema with collapsed and "entrapped" lung. Twenty-eight patients have experienced complete reexpansion of their lungs and are cured and 11 are in various stages of reexpansion and presumably will be cured. This experience is out of the ordinary, and the open elective treatment of mixed tuberculous empyema with total or near-total collapse and "entrapment" of the lung followed by complete reexpansion is presented together with a concept of the pathophysiology of this disorder and the response to therapy. Material and Methods Patients During a 7-year period from December 1987 through May 1995, forty-seven patients were evaluated and treated for chronic mixed tuberculous empyema, These patients ranged in age from 9 to 68 years. Twenty-six were male and 21 were female. Thirty-eight were being treated with antituberculosis drugs. Three had not been treated previously. The current drug status of the remaining 6 was unknown, although 3 of these had received therapy in the past. Culture Status Positive cultures for M tuberculosis were forthcoming in 24 patients. The 23 remaining patients had strong presumptive evidence of tuberculosis including at least three of the following: strong family history of tuberculosis, chronic low-grade fever, positive Mantoux test, high erythrocyte sedimentation rate, "typical" appearance of

3 220 ALI ET AL Ann Thorac Surg OPEN DRAINAGE OF TUBERCULOUS EMPYEMA 1996;62: A D Fig 2. Sixteen-year-old girl with mixed tuberculous empyema. (A, B) Empyema with air fluid level indicating bronchopleural fistula and complete collapse of the lung. (C) Chest tube is in place. The lung is collapsed against the mediasfinum. (D) After pleurocutaneous window the lung rim is visible. (E) Complete reexpansion of the lung 18 months later. C

4 Ann Thorac Surg ALI ET AL ;62: OPEN DRAINAGE OF TUBERCULOUS EMPYEMA A D B E C Fig 3. A 24-year-old cotton weaver. (A) Pleurocutaneous window: contralateral disease is evident, (B-F) Progressive and complete expansion over 25 months. F

5 222 ALI ET AL Ann Thorac Surg OPEN DRAINAGE OF TUBERCULOUS EMPYEMA 1996;62" endobronchial tuberculosis on bronchoscopy, and radiologic evidence of tuberculosis in the contralateral lung. In addition 28 had positive cultures of nontubercular organisms, most commonly Pseudomonas aeruginosa, Klebsiella species, Enterobacter species, and combinations of Pseudomonas and Staphylococcus and Enterococcus species. All patients with negative cultures had received or were receiving antibiotics at the time the cultures were obtained. Treatment After initial evaluation patients were placed on a fourdrug regimen of antituberculosis therapy. Four drugs were used routinely because of the significant incidence of resistance to INH and rifampicin in Pakistan. Chest tube drainage was carried out in those patients in whom this procedure had not been performed previously. Subsequently a modified Eloesser procedure was performed by removing 7.5-cm segments of two ribs anterior to the midaxillary line, one intercostal space above the most dependent part of the empyema cavity [1]. The skin was affixed to the thickened parietal pleura with interrupted absorbable sutures. These openings initially measured 5 5 cm but contracted to half of the original size as maturation took place. No window required revision. Irrigation was performed with 300 to 500 ml of dilute povidone iodine solution (5 ml of povidone iodine solution to 500 ml of normal saline solution or boiled water) twice daily. The patients were instructed to lie on their sides during this lavage unless a bronchopleural fistula was present, in which case the irrigation with a 50-mL syringe was performed sitting or standing up. Bronchopleural fistulas often were identified by the patient because of coughing and an iodine taste during the irrigation. Antituberculosis drug therapy was continued for a minimum of 9 months and for at least a month after complete reexpansion of the lung. Antibiotics were administered until drainage became minimal. It should be noted that none of these patients became completely sterile for nontuberculosis bacteria. If drainage increased cultures were again obtained and antibiotic administration reinstituted. Results Group I Twenty-eight patients experienced complete reexpansion of the lung from 4 to 30 months after open drainage. The majority of these patients presented with complete collapse of the lung against the mediastinum, in many despite the presence of functioning chest tube suction. The others had minimal reexpansion from previously placed chest tube suction. In those patients who were treated for the first time by closed suction drainage minimal if any reexpansion occurred. Recollapse of minimally reexpanded lungs was common when closed tube drainage was converted to open drainage. Group II Eleven patients are responding to therapy and their lungs are in varying stages of reexpansion after 3 to 30 months of open drainage. A 14-year-old gift with nearly complete reexpansion and resolved contralateral pulmonary infiltrates is awaiting epithelialization of the exposed lung at 20 months. Group III Eight patients did not achieve reexpansion of the lung. These patients were generally older, had had tuberculosis for longer periods, and had organisms that were resistant to drug therapy; lung destruction with fibrosis, bronchiectasis, and pleural adhesions was the rule. Comment Current therapy of chronic mixed tuberculous empyema includes pleural drainage and appropriate drug therapy, including antituberculous drugs and antibacterial agents. When total or near-total collapse is present and expansion does not take place definitive operation is indicated. Because these lungs usually defy decortication in the traditional manner, lung resection, often pneumonectomy, is carried out. Thoracoplasty and myoplasty may be required. The morbidity and mortality in these circumstances is high [2-6]. The reasons for this lack of reexpansion of the lung once the purulent material is evacuated probably relate to multiple factors. We believe the most important of these are tuberculous pneumonia and endobronchial disease and obstruction. Thickened pleura may play a role as may frank fibrosis and cavitation, but these factors, although important, were present in a relatively smaller number of patients. Commonly at operation decortication is not feasible. A traditional peel is not present, the lung surface is inflamed and friable, bronchopleural fistulas often are present, and positive pressure is not successful in reexpanding the lung. Operative maneuvers result in pulmonary laceration, the production of additional fistulas and failure. Pneumonectomy often becomes the only option [7]. The presented treatment scheme addresses both the bronchial and parenchymal disease and the secondary pleural involvement. Key is the eradication of active tuberculosis and secondary bacterial infection by appropriate drug therapy and open pleural drainage. The resolution of parenchymal and bronchial disease allows the lung to reexpand, but this is a gradual process that may take many months (Table 1). Pleural drainage and irrigation promotes elimination of the pleural infection and may prevent the formation of a thickened visceral pleura, which results in true entrapment. The reexpansion of these lungs initially was a fortuitous finding and was unexpected with an open pleural space. Once recognized the treatment plan became formalized on an ongoing basis. Presently, initial treatment consists of chest tube drainage and appropriate antibiotic therapy. Clinically most patients will be relieved of shortness of breath with chest tube drainage. Radiographic and clinical appraisal of the affected lung for potential reexpansion is performed at this time. Paradoxically the most common finding indicating potential reexpansion is total collapse of the lung against the mediastinum (Figs 2, 3). This indicates that the lung and

6 Ann Thorac Surg ALI ET AL ;62: OPEN DRAINAGE OF TUBERCULOUS EMPYEMA Table 1. Duration of Open Drainage Months Patients pleura have not undergone fibrosis. Chest tube drainage may produce minimal reexpansion, in which case the lung parenchyma may be evaluated. A clear rim of parenchyma showing no cavitation or bronchiectasis implies the potential for reexpansion. Cavitation and fibrosis, often with pleural adhesions, bode poorly for reexpansion. A fibrotic lung that does not totally collapse and on which thickened pleura can be identified probably will not respond. Those patients in whom a nonproductive cough persists on the basis of bronchial involvement will usually have reexpansion. In contrast, those patients in whom a productive cough persists on the basis of cavitary disease usually will not have reexpansion of their lung. In the earlier stages of this series, chest tube drainage was continued for several months. Presently open drainage is instituted at a much earlier stage, often within the first month after establishing the lack of expansion on suction drainage and assuring that mediastinal shift will not take place. Chest tube drainage is bothersome, is often painful, does not provide efficient drainage, and does not conveniently allow mechanical cleansing of the pleural space. Open drainage is well tolerated. Patients rapidly learn to irrigate their pleural cavity, and both the patient and the family feel they are important participants in the management of this condition. Although the treatment process is prolonged, the patient experiences progressive well being. This treatment scheme has been successful in 28 patients and probably will be successful in the 11 others who have demonstrated partial reexpansion as early as 3 months after open drainage. Eight patients did not achieve expansion of their lungs despite open drainage. Infection was controlled in all but 1, who died of uncontrolled drug-resistent tuberculosis within 1 month. Five of these patients were treated successfully by pneumonectomy and I awaits pneumonectomy. The last patient died of a brain tumor after 5 years of satisfactory palliation with open drainage. Evarts Graham described the mechanism of lung reexpansion in patients with open drainage for bacterial empyema treated during the influenza pandemic of They attributed this reexpansion to the "gradual pull of adhesions between the lung and the chest wall" [8]. Such has not been the experience with tuberculous empyema, where lack of expansion has led to decortication, lung resection, and thoracoplasty. With modern antituberculous drug therapy the pulmonary and bronchial components are controlled and the lung becomes capable of reexpansion. The pathophysiology of this process is unusual. The exact mechanism is speculative, but certain factors seem clear. As the antituberculosis therapy eradicates the parenchymal disease, lung compliance is improved. Parenthetically, with the clearing of bronchial infection and obstruction, aeration becomes possible. The granulating lung surface adheres to the parietal pleura and gradually pulls the lung outward in a progressive manner. This continues until complete reexpansion of the lung takes place. This process occurs in the face of continued atmospheric pressure and has occurred in patients in whom unclosed bronchopleural fistulas persist, which close as they are sealed by granulation tissue against the chest wall. The final phase is epithelialization of the exposed lung and contraction of the chest wall defect. Of interest is the fact that these skin-lined fistulas heal completely. No patient has required revision of the skin-lined defect, which remains as a large epithelialized dimple. References 1. Eloesser L. An operation for tuberculous empyema. Surg Gynecol Obstet 1935;60: Langston HT, Barker WL, Graham AA. Pleural tuberculosis. J Thorac Cardiovasc Surg 1967;54: Samson PC. Empyema thoracis: essentials of present day management. Ann Thorac Surg 1971;11: Hankins JR, Miller JE, McLaughlin JS. The use of chest wall muscle flaps to close bronchopleural fistulas: experience with twenty-one cases. Ann Thorac Surg 1978;25: Ninvivo J, Brandolino M. Rare infections in thoracic surgery. In: Pearson FG, Deslauriers J, Ginsberg RJ, et al, eds. Thoracic surgery. New York: Churchill Livingston, 1995: Hood RM, Antmank Boyd A, et al. Pleural infection in surgical disease of the pleura and chest wall. Philadelphia: Saunders, 1986: Odell JA. Pleural tuberculosis. In: Deslauriers J, Lacquer LK, eds. International trends in general thoracic surgery. Vol 6. Surgical management of pleural disease. St Louis: Mosby, 1990: Graham E, Singer J, Ballon H. Surgical diseases of the chest. Philadelphia: Lea & Febiger, 1935: DISCUSSION DR RICARDO BEYRUTI (Silo Paulo, Brazil): I congratulate Dr Ali and associates for bringing attention to this subject. In Brazil we have to deal with a significant number of patients with empyema due to various causes including tuberculosis, and we have been using this same approach for several years, with similar good results. At the Silo Paulo University Medical School Hospital, we recently simplified the pleurocutaneous window surgical technique. Using computed tomographic scans, the lowest part of the pleural cavity is spotted and under general anesthesia a 3- to 4-cm incision is made at that level, over the posterior axilary line. A 3-cm segment of the adjacent rib is resected and a

7 224 ALI ET AL Ann Thorac Surg OPEN DRAINAGE OF TUBERCULOUS EMPYEMA 1996;62: specially developed silicone rubber stent is inserted. With this technique we were able to mantain a 2-cm opening as long as necessary. It is simpler and quicker, and it avoids all the tissue dissection and sutures commonly employed with the traditional technique. From April 1995 to date we have used this stent in 13 patients ranging in age from 18 to 87 years, with empyemas due to several causes. We could follow up these patients for up to 290 days. Two of them had complete recovery of their pleural disease, and the stent could be removed 67 and 115 days postoperatively. The pleurocutaneous tract in both of them closed spontaneously about I week after the stent removal, leaving just a small skin scar. The remaining patients continue to improve, and their evolution is absolutely comparable with what we used to see with the classic thoracostomy technique. DR DOUGLAS J. MATHISEN (Boston, MA): I thank Dr Ali and her co-authors for providing me with the manuscript to review before the presentation and also congratulate her on the presentation and the excellent results that they have achieved and sharing with us those results. I think her paper and the previous one have emphasized the need for all of us as thoracic surgeons once again to refamiliarize ourselves with the surgical management of tuberculosis. It certainly has enjoyed a rebirth and a resurgence in certain parts of the country and especially in certain groups. I have three questions that came as a result of being able to review the manuscript, and I hope Dr Ali will be able to clarify a couple of the points. If you identify the adverse factors for this approach, namely fibrosis, cavitation, and thickened pleura, what would you recommend doing from the outset? Would you choose this method or recommend going to a surgical resection if you identify those adverse factors? Second, were you able to distinguish any difference in your outcome in those patients who had bronchopleural fistulas or in the group who had pure tuberculosis versus a mixed bacterial and tuberculous infection? Finally, in the manuscript you emphasize this conservative approach and would hold off on the use of decortication initially, but does that approach hold for the entire course of these patients? As you see the lung expand and the patient's general condition improve, would you ever reconsider decortication at a later date to possibly speed up the process? DR ALl: Doctor Mathisen, thank you for your kind words. With respect to the adverse factors that we have identified in these patients, you will note that 8 patients did not have reexpansion. These were the patients in whom we went back and found that they originally had fibrosis and bronchiectasis and destroyed lungs, in addition to the empyemas. Now, when we recognize these patients, we treat them for I to 2 months to get them in optimal condition and proceed with resection. Therefore these patients are not subjected to very long-term open drainage and lavage. Your second question pertained to bronchopleural fistulas. Bronchopleural fistulas per se, in the absence of lung destruction, led to slower expansion but did not in and of themselves prevent full expansion. So the presence of fistulas alone would not deter me from continuing with this procedure. DR JOHN A. ODELL (Rochester, MN): I enjoyed this presentation, which reminded me very much of my experience in South Africa, where I managed many patients in a similar fashion. I would like to take issue with the term "tuberculous empyema," which has become ingrained into our vocabulary. In fact, what the patient has is a bacterial empyema in the presence of pulmonary tuberculosis. I would like to ask how you assess the underlying lung. I believe it is very important to assess this early on in the course of the disease once the pulmonary tuberculosis is under control. As an example, in a bronchogram of a patient managed in a similar fashion one can see the open drain at the bottom of the empyema cavity. The bronchogram demonstrates a bronchiectatic lung with cavities in the apical segment of the lower lobe and the apex of the upper lobe. In this type of patient it is better not to persist with the hope of lung reexpansion. It will never happen because this lung is abnormal. In contradistinction, with the patient on whom you do a bronchogram who has no bronchiectasis, it might be better to do a decortication. In these circumstances the patient is then unfettered (or removed) from prolonged drainage procedures and is then able to resume a normal life. I know it is difficult in the Third World to assess the underlying lung because Dionosil is no longer available for bronchograms, and in these circumstances computed tomographic scanning is often unavailable because of high costs. DR ALh Thank you, Dr Odell. Your point regarding terminology is well taken. In the Third World, because Dionosil is not available, we use a dilute barium solution, and although the results are not as clear, they still give a reasonable bronchogram. When there is total collapse of the lung against the mediastinum, it usually implies that major lung destruction is not present. In the example that you gave, the lung was further out towards the chest wall and was not completely collapsed. A clear rim of lung parenchyma showing no cavitation and bronchiectasis usually is taken to imply that the lung will eventually reexpand. The presence of a fibrotic lung on the chest roentgenograms or on the computed tomographic scans, which does not totally collapse, as in your patient with overlying thickened pleura, implies that lung expansion probably will not occur. Often after these patients have been drained of the pus, they continue to have a cough. If this cough is nonproductive, it usually implies that it is secondary to bronchial involvement and is not because of cavitary disease and bronchiectasis. Productive cough, because of cavitary disease with gross bronchiectasis, implies that there is significant lung destruction and the patient is better of having a lung resection, rather than continuing with a fruitless procedure that will not lead to cure. DR JOHN R. BENFIELD (Sacramento, CA): I want to make only two points. One is that those of us who spent time in tuberculosis sanitaria years ago learned what you have reiterated. Tuberculosis is a slow disease by and large and tincture of time is often a good way to make the best possible clinical decision for patients. The second point is that we know from studies of chronic atelectasis in animal models and in humans that the lung can resume its function after as much as two decades of atelectasis. Do you have any data at all about the function of the reexpanded lung after either spontaneous reexpansion or decortication? DR ALh Thank you very much, Dr Benfield. We have not done sophisticated studies to assess the function of the lung. However, the patients who came to us in significantly compromised ventilatory conditions have eventually gone back in the majority of instances to lead normal functioning lives. That is the only indicator that we have of return to normal lung function.