Chapter 25: Bronchopleural Fistula

Pneumothorax, subcutaneous emphysema, pneumomediastinum, pneumopericardium, and other forms of extra-alveolar air are referred to as barotrauma. A bronchopleural fistula is a persistent leak from the lung into the pleural space, identified by either intermittent (during inspiration) or continuous chest tube air leak. Most barotrauma occurs in patients with trauma, acute respiratory distress syndrome (ARDS), chronic obstructive pulmonary disease (COPD), asthma, and postthoracic surgery. Properly treated extra-alveolar air and bronchopleural fistula are not usually life-threatening problems; however, they do complicate ventilator management.

Pathophysiology

Extra-alveolar air can develop with trauma, surgical procedures, tumors, and vascular line placement. During mechanical ventilation, extra-alveolar air forms as a result of alveolar rupture to allow gas to enter the adjacent bronchovascular sheath and dissect into the pleural space. Pulmonary disease, high pressure, and overdistention must be present for extra-alveolar gas to develop to a critical level. Extra-alveolar air develops most frequently in COPD and ARDS patients, particularly if complicated by necrotizing pneumonia. Maintaining peak alveolar pressure less than 30 cm H₂O and tidal volume 4 to 8 mL/kg ideal body weight avoids the setting where alveolar rupture is facilitated. Signs and symptoms of a pneumothorax during mechanical ventilation are listed in Table 25-1.

Chest Tubes

Pressure within the pleural space is normally subatmospheric. Once the thorax is entered, gas tends to move into the pleural space. To prevent the extension or development of a pneumothorax, a one-way valve must be attached to the chest tube to prevent air movement into the thorax. This is accomplished by use of an underwater seal (Figure 25-1). The chest tube is placed 2 cm under a column of water and, thus, gas may exit the pleural space when the pressure exceeds 2 cm H₂O. To accommodate fluid drainage, a second container is added to the drainage system. Fluid drains into the collection chamber without affecting the water seal. To facilitate fluid movement and to prevent loculated pockets of air from accumulating in the pleural space, a third chamber is frequently added to control the suction pressure applied to the thoracic space. The pressure applied to the pleural space is low (eg, –20 cm H₂O). In modern commercial systems, each of these chambers is incorporated into a singe device.

Techniques to Minimize Air Leak

Pneumothorax during mechanical ventilation is treated with chest tube drainage and suction. The combination of negative pleural pressure from the chest tube (–20 cm H₂O) and positive pressure from the ventilator establishes a pressure gradient across the lungs and may facilitate the development of a bronchopleural fistula. If a fistula develops, flow through the fistula is determined by the magnitude and duration of the pressure gradient across the lung. Ideally, the approach used to provide mechanical ventilation should minimize ventilating pressure, inspiratory time, and chest tube suction to avoid accumulation of pleural air. Some clinicians recommend independent lung ventilation or high frequency ventilation. Others have proposed manipulation of the chest tube suction system. Two specific approaches to modify chest tube suction are intermittent inspiratory chest tube occlusion and the application of intrapleural pressure equivalent to the level of positive end-expiratory pressure (PEEP). Experience with these maneuvers demonstrates a decrease in the air leak but collapse of lung units is common and neither technique has resulted in improved outcome. In general, a lung protective approach to ventilatory support with emphasis on minimizing airway pressures seems to work very well in the vast majority of patients.

Although air leak from a bronchopleural fistula should be avoided if possible, it is important to recognize that CO₂ elimination occurs through the fistula. The CO₂ concentration leaving the fistula may be similar to that exhaled from the endotracheal tube. In most cases, the fistula does not close until the underlying disease process has resolved. The presence of a bronchopleural fistula is an ominous sign. However, patients usually do not die from a bronchopleural fistula—they die with a bronchopleural fistula.

Indications

A bronchopleural fistula or other type of extra-alveolar air is not by itself an indication for mechanical ventilation. Its presence, however, increases the potential for problems with gas exchange. Indications for mechanical ventilation in this setting are apnea, acute ventilatory failure, impending acute ventilatory failure, or oxygenation deficit (Table 25-2).

Ventilator Settings

The goal of ventilator settings is to reduce the pressure gradient across the lung. Thus, the ventilating pressures and PEEP should be minimized (Table 25-3 and Figure 25-2) as much as clinically possible. A ventilatory pattern should be chosen that results in the least gas exiting the fistula, provided gas exchange targets are met. The use of pressure ventilation in this setting controls peak alveolar pressure. However, pressure-controlled ventilation may increase the leak through the fistula because it maintains a higher alveolar pressure throughout the inspiratory phase. The choice of pressure-controlled or volume-controlled ventilation should be determined by the mode that best minimizes air leak through the fistula.

Some of these patients require paralysis to establish the lowest air leak across the fistula and acceptable cardiopulmonary function. Whether spontaneous breathing should be allowed depends on the severity of the underlying disease process and the hemodynamics and gas exchange during spontaneous breathing. Pressure support ventilation should be used cautiously. With pressure support, inspiration terminates when flow decelerates to a predetermined level. If the leak across the fistula is greater than this level, the ventilator will not appropriately cycle from inspiration to exhalation during pressure support ventilation. Thus, careful setting of cycling criteria is important and these criteria may need to be frequently modified as ventilation continues. Moreover, suction applied to the chest tube may trigger the ventilator.

Permissive hypercapnia and the acceptance of low arterial oxygenation (Pao₂ > 50 mm Hg) are necessary for some of these patients. This is particularly true if the underlying disease state is ARDS, COPD, or trauma. Respiratory rate is set high enough to maximize CO₂ elimination but low enough to minimize fistula leak and air trapping. Depending on the underlying disease state, this may be a rate as low as 10/min or as high as 30/min or more. Tidal volume should also be as low as possible but normally in the 4 to 8 mL/kg ideal body weight range and inspiratory time should be short as possible, normally 0.5 to 0.8 second. All of these maneuvers are designed to minimize the air leak via the fistula. However, because all of these patients present with different levels of leak and pathophysiology, it is important to try various ventilator settings and determine the specific setting that results in the least air leak in the particular patient.

Management of oxygenation is difficult with a bronchopleural fistula, since PEEP used to improve oxygenation increases the leak. As a result, a high Fio₂ is needed. PEEP should be set at the minimal level necessary to maintain open unstable lung units. The goal is to minimize PEEP and mean airway pressure. However, particularly in ARDS and trauma, the oxygenation deficit may be severe and higher levels of PEEP required.

Independent Lung Ventilation

The use of a double lumen endotracheal tube with two ventilators (either synchronized or asynchronous) has been proposed for the management of severe bronchopleural fistula. This approach is only recommended when the fistula is the result of disruption of a large airway or where maintenance of an acceptable level of gas exchange is impossible and surgical intervention is planned. This should be considered a short-term solution. Of concern with independent lung ventilation is the potential damage to both the trachea and mainstem bronchi resulting from the use of a double lumen tube, the difficulty of maintaining proper position of the tube, the difficulty with suctioning and secretion clearance, and the technical issues due to the use of two ventilators. Settings on the two ventilators should be based on the pathology of the ventilated lung. Each lung may be ventilated in a similar manner but with lower pressures and volumes to the affected lung or with continuous positive airway pressure alone to the affected lung. The volume of the air leak, as well as hemodynamic and gas exchange stability, are the key variables used to determine the adequacy of ventilator settings.

High Frequency Ventilation

Little data other than case reports support improved outcome with high frequency ventilation. The use of high frequency ventilation is not recommended. Lack of accepted management protocols, high cost of the equipment, a limited number of patients requiring the technology, and lack of data indicating improved outcome all support this recommendation. Many centers that used high frequency ventilation in the setting of bronchopleural fistula in the past have abandoned its use. In addition, recent randomized controlled trials indicate that the use of high frequency oscillation in ARDS does not improve mortality.

Monitoring

Key concerns during monitoring of patients with a bronchopleural fistula (Table 25-4) are assurance of adequate gas exchange (pulse oximetry and arterial blood gases) and evaluation of the extent of the air leak. The volume of the air leak is quantified by measuring the difference between inhaled and exhaled V_T. Such estimates of air leak can be made using the monitoring and waveform capabilities of current generation ventilators and many indicate both inspiratory and expiratory tidal volumes.

Liberation

The specific approach used to liberate these patients from mechanical ventilation is not based on the presence of the fistula, but rather on the underlying disease. In general, as the underlying disease improves, the fistula begins to close. The presence of a fistula is not an indication to continue mechanical ventilation. The approach to liberation is not specific to the presence of a bronchopleural fistula.