106: Thoracoplasty for Tuberculosis

The specialty of thoracic surgery was born in the convergence of two worldwide plagues. These were tuberculosis, as old as humankind, and avian influenza, which struck a war-wearied world in the winter of 1917 and killed more people than the bubonic plagues of the Middle Ages. Today, with tuberculosis becoming resistant to antituberculous drugs, avian flu beginning to appear around the world, and the world weakened again by war and the new pestilence of AIDS, it seems entirely possible that thoracoplasty may become the once and future operation.

First coming to prominence in Europe in the late nineteenth century for the treatment of chronic infections in the chest with complicating space problems and bronchopleural fistulas, thoracoplasty began to receive greater attention in the United States at the time of the 1917 to 1918 influenza epidemic. The most lethal complication was empyema thoracis. The mortality in some of the 29 army camps surveyed was as high as 70%.¹ A pneumonia commission and subsequently an empyema commission were appointed to study the problem in the clinic and the laboratory. In the amazing period of 1 year, through the efforts of these two commissions, mortality was reduced to an average of 4.3%.² As reported by Graham in his insightful book, Empyema Thoracis, two early examples of productive research delineated the adverse effects of open pneumothorax on ventilation and the differing pathology between complicating infection with Streptococcus hemolyticus or Pneumococcus,³ the former being much more common. Associated with this pleural disease were a large number of patients with complicating space problems and bronchopleural fistula.

At about the same time, the other impetus for thoracoplasty was developing in the treatment of tuberculous residua when it was recognized that healing could be accelerated by collapse of residual cavities. Many techniques were employed for this, including pneumothorax, pneumoperitoneum, and even the very first videothoracoscopy, which was used to release apical adhesions.⁴ Not infrequently, these patients developed infected spaces requiring thoracoplastic procedures, and it was a short step from thoracoplasty for complicating spaces to thoracoplasty as the primary procedure. Enthusiasm for this procedure as an alternative to prolonged, sometimes lifelong commitment to a sanitorium can be imagined from a story told to me by a Greek colleague whose mother had an eleventh rib thoracoplasty performed in three stages by Professor Sauerbruck before World War II using local anesthesia!

The most enthusiastic American proponent of thoracoplasty was John Alexander⁵ of the University of Michigan, who applied and evaluated the procedure in a large group of patients. A contemporary tale from this period was shared with me by one of his residents, who, along with this fellow trainee, noted that, paradoxically, the application of thoracoplasty to nontuberculous pulmonary cavities was associated with an increased mortality, a concept he only reluctantly accepted.

Thoracoplasty can be considered in four broad categories:

1. Thoracoplasty with a closed pleural space.

2. Thoracoplasty with open pleural drainage.

3. Thoracoplasty with transposition of a muscle flap into the space either to fill the space or to close a bronchopleural fistula or both.

4. Thoracoplasty done as a preliminary to avoid postresectional space problems.

Contemplation of thoracoplasty in the modern age implies a failure of medical therapy. An open negative cavity or resistant organism with unilateral or bilateral cavitary disease, bilateral disease so extensive that resection is not feasible, and tailoring thoracoplasty to avoid space problems after resection are the probable indications. In these patients, all the precautions observed in the great age of thoracoplasty still apply. First, timing and judgment are more important than the surgery, and thorough understanding of the natural history and pathophysiology of tuberculosis is required for this judgment.⁶ This begins with the therapeutic value of rest, fresh air, a high-calorie diet, and patience. Time is on the side of the physician. Between 6 and 12 weeks of preoperative antituberculous medical therapy is favored.

As with any thoracic procedures, the patient should be evaluated for cardiac and pulmonary functions and, if marginal, further assessed with differential V./Q. studies or a V.o_2,max study to see if the intended procedure can be tolerated. Many of these patients will have had bilateral pulmonary disease and, in some cases, prior surgery. An important part of the preoperative care is the bacteriologic analysis and treatment. Many patients are now manifesting resistance to multiple antituberculous drugs, and close collaboration with infectious disease colleagues is mandatory. Another problem here is infection with atypical tuberculosis, especially in patients with HIV infection or otherwise compromised immunity. The presence of cavitary disease in the lung, even if acid-fast bacilli-negative, leads to frequent invasion of the cavity by fungal organisms, the most frequent being Aspergillus fumigatus. These organisms or tuberculosis itself can lead to development of mycotic aneurysms in the cavity wall. The rupture of such aneurysms can create one of the true catastrophes in thoracic surgery with massive pulmonary hemorrhage. This possibility alone is reason for an aggressive approach to pulmonary cavitary disease. Finally, in this spectrum are bacterial mixed infections that may have led to bronchopleural fistula or empyema as the presenting problem. Again, the emergence of drug resistance is an ongoing problem. All these areas of consideration are made infinitely more difficult in the presence of infection and HIV infection.

Tuberculosis is best considered as a systemic disease, but when it presents in the lung, it begins as an exudative pneumonia favoring upper and posterior portions. It either goes on to resolution or to destruction of lung tissue with formation of scar or loss of vascular supply with resulting caseating necrosis. It is these foci that persist or excavate and may require surgical attention. Since the inception of antibiotic therapy, there has been a greater tendency to reepithelialization of the bronchocavitary junction with less air trapping. The cavity wall, surrounding parenchymal reaction, and pleural reaction are all less. This pleural reaction is very important because it is usually very vascular, and the adhesions formed are part of the body's attempt to prevent bronchopleural fistula.⁷ For this reason, mobilization should be done in the extraperiosteal plane unless resection is contemplated both to avoid bronchopleural fistula and because the systemic vessels are difficult to control and can cause massive operative blood loss. Another suggestion that pertains to adhesions is that although during the period of pneumothorax therapy it was always felt necessary to release apical adhesions when thoracoplasty was anticipated, it is best to preserve them. Dividing them lets the lung fall lower in the chest, thereby requiring resection of more ribs to obtain complete collapse.⁵ Before surgery, it is best to decide how many ribs will need resection and what the often compromised patient can tolerate. It may be necessary to stage the resections to reduce morbidity.

Thoracoplasty with Closed Pleural Space

The patient is best positioned in a lateral position, leaning slightly forward. There was some enthusiasm for operation in the prone position previously. Work that my team did on our service at the Albert Einstein College of Medicine in New York City demonstrated that this position, even in healthy volunteers, leads to an average reduction of cardiac output and blood pressure of approximately 25%, and we abandoned it. Use of a double-lumen endotracheal tube will lessen the secretions reaching the dependent lung. However, if the patient is being operated on for pulmonary hemorrhage, I prefer a large-diameter single-lumen tube, which allows for better removal of blood clots. In either case, it is highly advisable to do a bronchoscopic lavage at the end of surgery.

The incision should extend from the spine of the scapula down around its tip and anterior to the anterior axillary line (Fig. 106-1). After division of the trapezius, rhomboids, latissimus dorsi, and serratus anterior muscles, the scapula is elevated. This can be facilitated by placing one blade of a rib retractor on the point of the scapula and one on the chest wall and opening it to its greatest extent. This gives very good exposure of the upper ribs and frees the assistant for more productive activity.

Once the rib cage is exposed, it is necessary to divide the attachments of the serratus posterior, posterior and middle scalenus, serratus anterior, and pectoralis minor muscles to those ribs to be resected (Fig. 106-2). This is required for good collapse to occur. The periosteum is incised on the external surface of the ribs and stripped from them. This must be carried posteriorly to the costovertebral joint for total removal. Division of the costotransverse ligament facilitates this and removal of the transverse process, which is equally important if a posterior gutter is not to be left with incomplete collapse of the lung (Fig. 106-3).

At this point, a decision must be made about the ribs themselves. They may be resected, which often leads to late orthopedic and cosmetic changes.⁸ They may be left in place, or they can be rotated 180 degrees and fixed at the ends to act as a plombe. The insertion of various materials between the ribs and the collapsed chest wall to act as a plombe was the cause of many complications and has been abandoned. Once the collapse is complete, the wound is closed in anatomic layers. Sometimes with a six-rib thoracoplasty, the tip of the scapula impinges on the top of the seventh rib, forming a painful bursa. This led some surgeons to do a hemiscapulectomy. This is ill-advised because it leads to rotation of the scapula with added deformity of the shoulder and usually does not solve the problem. It is better to resect a portion of the underlying seventh rib (Fig. 106-4).

Thoracoplasty with Open Pleural Space

In cases where there has been pleural space infection with tuberculosis or mixed infection, there may have been a need for prior drainage. Pure tuberculous infection at times can be treated with aspiration and antituberculous drugs if caught early but has the potential for formation of a pleural peel within 3 to 4 weeks. In the past month, I had such a patient with no parenchymal changes. Early decortication and antituberculous drugs have given a surprisingly good result for him. Although I can find no reference for it, I remember when in the 1960s there was some enthusiasm for steroids in cases of tuberculous pleural and especially pericardial effusions coupled with aspiration and drug therapy.

When conservative measures have failed and early decortication is contraindicated by marginal cardiopulmonary reserve, open drainage is required. This can be accomplished with a small thoracotomy, removal of proteinaceous debris, and insertion of a large-bore mushroom catheter. For more definitive drainage, an Eloesser flap can be created.⁹ Using a U-shaped skin incision with the base uppermost and the tip of the flap over the lowest portion of the cavity, portions of the underlying ribs are excised, and the skin flap is turned in and sutured to the upper end of the pleural opening (Fig. 106-5). Total unroofing of the cavity as proposed by Schede¹⁰ has been abandoned. In these patients, closure of the space occurs by secondary intention from the periphery of the cavity. This process can be accelerated by a localized extraperiosteal thoracoplasty.

Thoracoplasty with Muscle Flap

The most challenging and debilitating disease occurs when either a primary or postresection space is caused by bronchopleural fistula. It should be noted that postresectional fistulas are encountered most commonly in patients operated on for tuberculous complications before adequate medical treatment has been achieved, such as for active pulmonary hemorrhage. In some of these patients, a combination of drainage, thoracoplasty, and antituberculous drugs will achieve a closure of the fistula. If this does not occur, it will be most helpful to bring in a vascularized muscle flap to apply to the bronchus or the fistula or to completely fill the cavity^11,12 (Fig. 106-6). It can be pointed out here that the vascularized chest wall of a thoracoplasty was the first such myoplasty. Here again, patience is a winning virtue because time must be allowed for maximum medical benefit. Muscle flaps should not be brought into areas of active tuberculous infection. The muscles available are the pectoralis major, serratus anterior, latissimus dorsi, rectus abdominis, and intercostals and should be based on an intact blood supply. These can be introduced directly into a drainage already established or through a small thoracotomy placed to achieve the most direct path through the chest wall to the fistula. They should be sutured around the parenchymal defect or leaking bronchus. Here again, the obliteration of space is a good basic principle and can be achieved by an appropriate thoracoplasty or by filling the space with muscle or both.

Tailoring a Thoracoplasty

In some patients who are candidates for resection of tuberculous residua, the adhesions and scar present in and around the lung that will remain are such that a postoperative space problem is predictable. In this case, a preliminary or synchronous thoracoplasty can be performed as described previously to tailor the thoracic space so that the lung will fill it. Temporal separation of the two procedures has the advantage of less impact on the patient. It also allows for evaluation of the result achieved so that, if needed, an additional rib or two can be conveniently resected at the time of pulmonary resection.

While this discussion has been limited to consideration of thoracoplasty in tuberculosis, it must be noted that the same forces that are leading to an upsurge in typical and atypical tuberculosis are also favoring development of various fungal diseases. The problem of residual space in the thorax is most often associated with aspergillosis, but especially in immune-compromised patients, often fungal disease may invade. In these patients, antifungal medical therapy is not always successful, and surgical intervention is required. All the considerations of maximal medical benefit, careful preoperative assessment, and strong nutritional support obtain. This topic is too broad for discussion here.¹³

Although the present need for thoracoplasty is infrequent, it seems that all thoracic surgeons should be familiar with its uses, variations, and timing as we move into a period of increasing drug-resistant tuberculosis and a parallel increase in opportunistic infections.

One use of thoracoplasty in the early- and mid-twentieth century reflected (1) the misguided impression that removal of ribs would prevent overdistention of the remaining lung after pneumonectomy, and (2) an aggressive surgical approach for limiting residual “space” in the chest and minimizing infections. Other surgical options for minimizing the residual thoracic space include complete mobilization of the lung to facilitate expansion into the space, an extrapleural “tent” to eliminate apical space, and maneuvers to facilitate cephalad mobilization of the diaphragm to minimize basilar space (phrenic nerve injection and pneumoperitoneum). In practice, thoracoplasty has been largely replaced by the use of extrathoracic tissue – primarily muscles or omentum – to eliminate intrathoracic space.