Version May 2024
INTRODUCTION — An alveolopleural fistula (APF) is a pathologic communication between the pulmonary parenchyma distal to a segmental bronchus (alveoli) and the pleural space. It presents as a pneumothorax and if it persists beyond five days is labeled as a prolonged air leak (PAL). PAL is associated with significant morbidity and prolonged hospitalization.
The evaluation and management of APF with a PAL is discussed here. Diagnosis and management of tracheoesophageal and bronchopleural fistula are discussed separately. (See "Bronchopleural fistula in adults" and "Tracheo- and broncho-esophageal fistulas in adults".)
ETIOLOGY — APF and PALs are most commonly seen in patients following lung volume reduction surgery (up to 46 percent) and pulmonary resection or biopsy (3 percent following wedge resection) but can also be seen following spontaneous pneumothorax [1-3]. Less common causes include pulmonary infections (eg, necrotizing pneumonia), malignancies (eg, pulmonary metastasis), pleural drainage procedures, barotrauma due to mechanical ventilation, chest trauma, and iatrogenic etiologies that can occur after thoracentesis or chest tube insertion (ie, pleural drainage procedures). (See "Diagnosis, management, and prevention of pulmonary barotrauma during invasive mechanical ventilation in adults".)
RISK FACTORS — In the setting of lung resection (robotic- or video-assisted, open) or lung volume reduction surgery, several risk factors for APF with PAL have been reported [1,2,4,5]:
●Chronic obstructive pulmonary disease or other underlying lung disease
●Female sex
●Low forced expiratory volume in one second or diffusing capacity for carbon monoxide
●History of smoking
●Diabetes mellitus
●Chronic steroid use
●Marked pleural adhesions noted during surgery
●Upper lobe emphysema or diffuse emphysema
●Increased age
●Large bullae
Although likely, it is unknown whether similar risk factors promote APF with PAL in nonsurgical populations.
CLINICAL FEATURES — There are no specific signs or symptoms associated with PAL, but some patients have a persistent pneumothorax on chest imaging. Although APFs can theoretically cause hypercapnia due to loss of alveolar ventilation through the air leak, this is rare in clinical practice. This is probably because the leaked gas participates in gas exchange rather than being lost or wasted [6]. Intractable respiratory acidosis and other serious complications are more likely due to the lung disease than the APF.
EVALUATION AND MANAGEMENT — An APF should be suspected in a patient with risk factors who has a PAL seen on chest tube drainage system for greater than five days. Continued flow through the fistula to the pleural space delays healing and prevents lung expansion; the longer air flows through the fistula, the more likely it is to remain patent. Thus, prompt reduction in the flow of air is critical for effective management. APF-associated PALs are not fatal but are associated with prolonged hospital stays, higher rates of intensive care unit admission, and high morbidity (eg, pneumonia, empyema, venous thromboembolism) [1,2].
There is a paucity of data and no consensus or guidelines on how best to manage APF with PAL. Significant variation among clinicians exists, but practice is evolving as expertise in interventional pulmonary grows. The approach outlined here is influenced by our expertise in interventional pulmonology, and we recognize that this strategy may not always be universally applied, particularly when interventional expertise is not available.
Air leak can be classified into drainage-dependent (also known as pressure-dependent) and drainage-independent (also known as pressure-independent) air leaks [7,8]. This is typically determined with pleural manometry, which is a technique that can directly measure the pleural pressures and differentiate between these two types of air leaks. The majority of patients with PAL following partial lung resection have drainage-dependent air leak. These patients have shorter hospital length of stay and less need for reexploration surgery compared with patients with drainage-independent air leak [9].
●Drainage-dependent air leak typically occurs in patients with nonexpandable lung where air leak continues only during the chest tube drainage process and stops when drainage is discontinued. This is caused by shape mismatch between the lung and thoracic cavity due to visceral pleural restriction. During drainage, excessively negative pleural pressure leads to a significant distortion of the subpleural alveolar units, resulting in the formation of the transient pressure-dependent APF and thereby formation of pneumothorax [7,8].
●Drainage-independent air leak occurs due to direct visceral pleural injury, and the air leak persists even if the drainage process is discontinued. This is due to a continual rise in the pleural pressure leading to an increase in the size of the pneumothorax.
General supportive care (drainage of air) — Most patients with PAL respond to conservative therapy (up to 80 percent in our experience). This involves continued chest tube thoracostomy with low wall suction to promote lung expansion and pleural apposition. In some cases with large fistulas (eg, secondary spontaneous pneumothorax in patients with chronic obstructive pulmonary disease secondary to rupture of large bullae), the addition of high suction and even insertion of a second chest tube may be warranted. In contrast, in patients with nonexpandable lung (trapped or entrapped lung (see "Diagnosis and management of pleural causes of nonexpandable lung")), suction should be avoided. Patients should also receive adequate nutrition and appropriate antibiotic therapy (if indicated), and therapy for comorbidities should be optimized. For patients who are mechanically ventilated, lowering the level of positive pressure and selective intubation of the healthy lung is appropriate, the details of which are discussed separately. (See "Management of persistent air leaks in patients on mechanical ventilation".)
It is controversial whether or not suction should be applied. On one hand, some experts, including us, apply low wall suction to increase adherence of the visceral and parietal membranes, thereby promoting healing and closure of the fistula. In contrast, other experts avoid suction believing that suction promotes continued patency. The latter practice is supported by studies of patients in whom routine suction is applied immediately postoperatively that have suggested a reduced incidence of APF with PAL in patients whose chest tubes were placed on water seal compared with suction [10-13]. However, none of these studies have been performed in patients with PAL from other etiologies. In addition, arguing against the practice of avoiding suction is our belief that healing of the visceral pleural membrane is considerably less likely to occur if not apposed with the parietal membrane. Thus, we apply suction but minimize it when possible, especially in patients with nonexpandable lungs (trapped or entrapped lung) with PALs.
There is no optimal follow-up time for patients who undergo conservative therapy. Nonetheless, in general, smaller PALs that are improving with time (at minimum five days and occasionally up to two weeks) are more likely to spontaneously undergo closure with conservative management than larger air leaks that have been present for a prolonged period of time and are not improving despite conservative therapy; the latter are unlikely to resolve spontaneously and require intervention.
Although the exact proportion is unknown, in our experience, approximately 20 percent of patients will fail conservative therapy and require some form of intervention. The presence of a severe air leak (see 'Quantifying the air leak' below), incomplete lung expansion on chest computed tomography (CT), and/or underlying lung disease (such as emphysema, interstitial lung disease, cancer) can probably predict the need of additional intervention for fistula closure.
Patients who fail supportive care — Options for fistula closure in patients who fail conservative therapy include ambulatory nonendoscopic devices, bronchoscopic methods, pleural procedures, or surgery. Choosing among these depends upon the available expertise and individual preferences (algorithm 1). With the exception of patients with spontaneous pneumothorax, who in general should undergo video-assisted thoracoscopic surgical (VATS) blebectomy and mechanical pleurodesis, we prefer a systematic approach that measures the size, location, and integrity of the interlobar fissure of the affected lobe with the APF. Our preference is based upon the rationale that it identifies those suitable for bronchoscopic closure, thereby allowing optimal selection of therapeutic options. For example, leaks associated with minimal collateral ventilation between the target lobe and adjacent lobes are better suited to bronchoscopic therapy, while large leaks associated with significant collateral ventilation might be suited for surgical repair or pleural procedures, depending on clinical status of the patient.
Quantifying the air leak — The quantification of air leak is important to assess whether the leak will spontaneously resolve with additional supportive care or need more aggressive intervention. In general, small air leaks resolve spontaneously, whereas larger leaks require intervention, although what constitutes small versus large is not clearly defined. Air leaks can be quantified using several methods. We prefer digital drainage system devices since they may be more accurate with less interobserver variability and are easy to use.
Methods include:
●Digital chest drainage system devices – Digital chest drainage system devices can display the flow (mL/min) of air into the pleural space along with the pleural pressure difference in real time [14]. These devices are not always available but are being increasingly used by thoracic surgeons and interventional pulmonologists instead of standard commercially available drainage systems. The digital system works by maintaining the intrapleural pressure at a steady level within 0.1 cm H2O. The regulated suction adjusts according to the condition or need in the pleural cavity. The device will apply suction to keep the pleural cavity at the present level. If the patient does have an air leak with suction, the device will intermittently apply suction to restabilize the pleural space according to the degree of the air leak. The chest tube can be removed when there is no flow or air leak flow is less than 20 mL/min without large variation for at least six hours as measured by digital chest drainage system. Limited data suggest that for patients with an air leak from spontaneous pneumothorax, digital drainage systems were associated with a shorter length of stay and chest tube duration compared with analog systems [15].
●Air leak meter – The air leak meter on commercially available chest drainage systems can measure leak on a scale from 1 to 7, with the number representing the columns through which bubbling occurs (the higher the number, the greater the leak) (figure 1) [16].
●Observation using timing during the respiratory cycle – Some experts grade air leak from mild to severe when the leak occurs during forced expiration (grade 1), passive expiration only (grade 2), inspiratory only (grade 3), or inspiration and expiration (grade 4), with grade 1 being the least and grade 4 being the worst leak [10].
Localizing the air leak (sequential balloon occlusion) — Sequential balloon occlusion through flexible bronchoscopy is a useful method to locate the bronchial segment or subsegment that is supplying the APF [17]. A balloon (eg, a Fogarty balloon of appropriate size, typically 5 French) is placed endobronchially through the working channel and inflated to first occlude the main stem bronchus for up to two minutes; a significant reduction in air leak observed through the chest tube drainage system indicates that the bronchus selected leads to the defect. The operator subsequently repeats this step, moving distally through lobar, segmental, and subsegmental bronchi until the target lobe with the fistula is reached.
Evaluation of fissure completeness — While some lobes in the lung are fissurally contained, others communicate with adjacent lobes (also known as incomplete fissure or collateral ventilation), a phenomenon that is more common in patients with emphysema. The degree of collateral ventilation should be evaluated to determine whether bronchoscopic methods will be effective. While patients with evidence of an incomplete fissure may not be suitable for bronchoscopic methods, some patients in this category who undergo bronchoscopic treatment may be able to transition to water seal, allowing them to be discharged on a one-way valve (eg, Heimlich valve).
●Sequential balloon testing – Sequential balloon testing not only localizes the air leak (see 'Localizing the air leak (sequential balloon occlusion)' above), but once the target lobe is identified, it can also be used to assess collateral ventilation as an indicator of fissure completeness. On sequential balloon testing, a reduction in airflow of ≥50 percent (either estimated as bubbles per minute or measured with a digital chest drainage system) indicates that the APF may be responsive to bronchoscopic management [18]. If a reduction in airflow of <50 percent is observed, then significant collateral ventilation (ie, incomplete lung fissure) between the target and ipsilateral lobes is likely, and bronchoscopic treatment may not be effective [19].
●High-resolution CT of the chest – Fissure completeness between target and ipsilateral lobes can also be evaluated by direct visual assessment of high-resolution chest CT images of the lobar fissures (image 1) and/or quantified by chest CT image analysis software (figure 2) [19,20]. Fissure integrity >90 percent suggests minimal collateral ventilation and indicates that the APF may be responsive to bronchoscopic management [18]. If fissure integrity ≤90 percent is observed, then significant collateral ventilation (ie, incomplete lung fissure) between the target and ipsilateral lobes is likely, and bronchoscopic treatment may not be effective.
Minimal collateral ventilation and significant air leak reduction on bronchial occlusion — Air leaks with minimal collateral ventilation (eg, >90 percent complete (see 'Evaluation of fissure completeness' above)) that also demonstrate a ≥50 reduction in airflow with bronchial occlusion (see 'Localizing the air leak (sequential balloon occlusion)' above) are suitable for bronchoscopic methods of fistula closure.
Bronchoscopic interventions — Several options are available, none of which are proven to be superior over the other. Data to support these methods are generally derived from case reports or case series. Among the options, we prefer endobronchial valves because they are the best studied and easy to place and remove:
●Endobronchial valves (figure 3) – Unidirectional airway devices are available in different sizes (5, 6, 7, and 9 mm) that can be placed in lobar, segmental, or subsegmental bronchi using a flexible therapeutic bronchoscope with a 2.8 mm diameter working channel. Some commercial devices come with balloon kits to determine the optimal size. These airway valves limit airflow to portions of the lung distal to their placement while permitting mucous and air movement in the proximal direction, thereby reducing airflow through the fistula and allowing the defect to heal. Based upon studies, typically two to three valves are needed to control the air leak. Endobronchial valves have been approved by the US Food and Drug Administration to treat APFs through a humanitarian device exemption.
Case reports and case series consistently show a complete or near-complete resolution of air leak in the majority of patients with valve placement [21-26]. In the largest trial of 75 patients with APF and PAL who underwent valve implantation, air leak resolution occurred in 70 percent of patients with a median time to resolution of 16 days [26]. Additional subsequent procedures were needed in 20 percent of patients (eg, Heimlich valve, chemical pleurodesis). Two patients experienced complications, including empyema and contralateral pneumothorax. Another study has shown similar utility for valves in the closure of AFP with PAL associated with mechanical ventilation [24].
PAL treatment with endobronchial valves is more successful in patients without collateral ventilation, especially when complete lobar occlusion was achieved. One study examined the utility of assessing the integrity of the interlobar fissure to predict successful resolution of the PAL. This multicenter retrospective study of 26 patients included 16 patients without collateral ventilation and 10 patients with collateral ventilation. Of the 16 patients without collateral ventilation, 14 had complete resolution of the PAL, with a median time from endobronchial valve placement to air leak resolution of 4.5 days (three days with complete lobar occlusion versus 6.5 days with incomplete lobar occlusion), and 2 patients required additional procedures to achieve resolution of the PAL. Among the remaining 10 patients with collateral ventilation, 4 patients had complete resolution of the PAL, with a median time from endobronchial valve placement to air leak resolution of 17.5 days, and 6 patients required additional interventions [27].
During the coronavirus disease 2019 (COVID-19) pandemic, pneumothorax and PAL have occurred due to a combination of fibrotic parenchymal changes and prolonged high-pressure ventilation. Endobronchial valves have been reported in some case series to help support weaning patients with COVID-19 from mechanical ventilation without the need for more invasive intervention [28-30].
●Spigots – Bronchial occlusion can be achieved with a silicone spigot. Spigots are silicone plugs (not available in the United States) that are available in different sizes (5, 6, and 7 mm; sized according to balloon inflated in the airway segment of interest and confirming air leak decrease/cessation) and have studs on the lateral surface permitting anchorage within the target bronchus. In a case series of 60 patients with APF and PAL who underwent spigot placement (mean number of four spigots per patient), complete resolution of the air leak occurred in 40 percent, and a reduction of the leak occurred in an additional 38 percent [31]. The procedure was complicated by pneumonia (3 percent), dyspnea (3 percent), and fever (1 percent).
●Occlusive material – Anecdotal reports of successful fistula closure using fibrin glue and other sealants (including gelatin sponge and oxidized regenerated cellulose) have been reported [32-34].
●Airway stents – Although infrequently used, stents can be placed to occlude the affected lobar bronchus, thus diverting flow away from a target lobe. For example, if the left upper lobe is the target lobe, a fully covered self-expanding metallic stent can be a placed in the left main stem bronchus to the left lower lobe, thereby bypassing the left upper lobe. (See "Airway stents".)
Follow-up — During the immediate follow-up period, the clinician should assess for the degree of residual air leak as well as for symptoms of complications of the procedure. Complete success is considered one where the air leak resolves, allowing removal of the chest tube (usually over days), while a partial response is one where the air leak no longer requires suction and is controlled with a water seal. In cases of a partial response, an additional procedure, often the placement of an ambulatory drainage device (such as Heimlich valve, chest drain valve, mobile dry seal drain, or digital chest tube) is performed to achieve successful closure.
For those who undergo valve placement, follow-up depends on air leak cessation. If the leak resolves, then the chest tube is removed. However, if leak persists, patients are discharged on ambulatory drainage devices and reassessed in two weeks. All patients are reevaluated at six weeks post-air leak cessation, and valves are removed per manufacturer's recommendations. In addition, the clinician should look for symptoms of complications (eg, fever, pneumonia, displacement, and expectoration) during the period where valves are present. A similar protocol is used if plugs were deployed. Plugs also require removal at six weeks post-air leak cessation, whereas occlusive material dissolves with time (usually two weeks postplacement) and does not need removal. (See 'Ambulatory drainage devices' below.)
For those who fail bronchoscopic occlusion of APF (generally six to eight weeks), nonbronchoscopic methods or surgery are options. (See 'Surgical repair' below and 'Ambulatory drainage devices' below.)
Significant collateral ventilation or minimal reduction of air leak on bronchial occlusion — Large APF with PAL associated with significant collateral ventilation with other lobes (eg, ≤90 percent complete (see 'Evaluation of fissure completeness' above)) or a <50 reduction in airflow with balloon occlusion (see 'Quantifying the air leak' above) are not suitable for bronchoscopic valve treatment of fistula closure alone. Surgical repair and/or additional pleural intervention are typically necessary, depending upon the patient candidacy for surgery.
Surgical repair — Surgical intervention is usually considered for patients with spontaneous pneumothorax, those in whom bronchoscopic methods are not suitable or those refractory to bronchoscopic or other nonsurgical procedures. VATS may be used to achieve pleural adhesion with application of sclerosing agents under vision, pleural abrasion, or pleurectomy. Other surgical interventions include over stapling of parenchymal lesions and application of sealants, thoracotomy with muscle or omental flap reinforcement at the site of the leak, pleural tenting, and buttressing of staple lines to obliterate the fistula [35].
Refractory patients — For patients who fail or are not candidates for bronchoscopic or surgical approaches, ambulatory drainage devices and nonsurgical pleural procedures are options. They are also considered additive when surgical or bronchoscopic methods achieve a partial response only. Choosing among them should be individualized and depends upon fistula characteristics, local expertise, and patient preference. (See "Chemical pleurodesis for the prevention of recurrent pleural effusion".)
Ambulatory drainage devices — Chest drain valve (picture 1), Heimlich valve, mobile dry seal drain (picture 2 and picture 3), and digital chest tube drains are one-way small valves that are applied to a chest tube, allowing patients to ambulate and even facilitating earlier hospital discharge. Chest drain valves, Heimlich valves, and mobile dry seal drains can only be placed once the patient demonstrates no pneumothorax on water seal, while digital chest tube drains can be placed even when the patient requires continuous suction since the system has the capability to provide ambulatory suction (see 'Quantifying the air leak' above). Patients can be discharged home with these devices as long as they are asymptomatic without subcutaneous emphysema or enlarging pneumothorax size (table 1).
Nonsurgical pleural procedures
Chemical pleurodesis — Chemical pleurodesis (via medical thoracoscopy or chest tube) using sclerosants such as talc or doxycycline induces an inflammatory response that leads to a scar obliterating the pleural space and allowing for the pleural defect to seal [1]. Retrospective studies report success rates between 60 and 90 percent [1,36]. Successful pleurodesis requires direct apposition of the visceral and parietal pleura and thus should only be done if there is a small or no residual pneumothorax when the chest tube is on water seal; chemical pleurodesis performed with a large pneumothorax may result in failure of the lung to reexpand. Chemical pleurodesis and its complications are discussed separately. (See "Chemical pleurodesis for the prevention of recurrent pleural effusion" and "Medical thoracoscopy (pleuroscopy): Diagnostic and therapeutic applications", section on 'Blebectomy and/or pleurodesis for pneumothorax'.)
Autologous blood patch pleurodesis — Autologous blood patch pleurodesis (ABPP) is an alternative technique that has been used in patients with APF and PAL, with or without complete lung expansion. Studies have shown that ABPP is safe and well tolerated by most patients [37-43]. A review of 10 retrospective studies reported a success rate of 92 percent [37].
A sample of the patient's own blood (usually 100 mL) is directed to the pleural space via the chest tube followed by 20 mL of saline. The chest tube is then hung over an intravenous pole (for one to two hours) to avoid drainage of blood and blockage to the tube with thrombus, which could induce a tension pneumothorax, particularly in those who require suction for pleural apposition.
The proposed mechanism of action is likely due to direct sealing of the air leak as well as the induction of pleural inflammation and eventual pleurodesis. The optimum blood required to achieve successful pleurodesis is unknown and ranges from 50 to 120 mL. However, studies have shown that patients who received 100 to 120 mL of blood achieved complete resolution of PAL faster and more successfully compared with lower blood volumes (50 to 60 mL) [37-40]. Thus, we prefer use of higher volumes of blood for this procedure.
Complications occur in <10 percent of cases and include pleuritis, empyema, and chest tube obstruction [40].
Special populations
Patients with spontaneous pneumothorax — Patients with primary spontaneous pneumothorax, particularly those with failed lung reexpansion who have PAL, are generally treated with VATS surgical pleurodesis, a practice that is consistent with most guidelines [44-46]. Patients with secondary pneumothorax also, typically, undergo surgical pleurodesis, although valves may be an option in those not fit for surgery [47-49]. (See "Treatment of primary spontaneous pneumothorax in adults" and "Treatment of secondary spontaneous pneumothorax in adults".)
SUMMARY AND RECOMMENDATIONS
●An alveolopleural fistula (APF) is a pathologic communication between the pulmonary parenchyma distal to a segmental bronchus (alveoli) and the pleural space. An air leak for greater than five days noticed on chest tube is called a prolonged air leak (PAL). (See 'Introduction' above.)
●APF with PALs are most commonly seen in patients following lung volume reduction surgery and pulmonary resection (typically wedge resection or biopsy) but can also be seen following spontaneous pneumothorax. Less common causes include pulmonary infections, malignancies, pleural drainage procedures, mechanical ventilation, and chest trauma. (See 'Etiology' above.)
●For most patients with PAL, we suggest conservative measures rather than aggressive interventions (algorithm 1) (Grade 2C). (See 'General supportive care (drainage of air)' above.)
•Conservative therapy involves continued chest tube thoracostomy and low wall suction in those with expandable lung; a second chest tube may be warranted in those with large leaks, while suction should be avoided in those with nonexpandable lung.
•Other therapies include optimization of nutrition and therapy for comorbidities and lowering the level of positive pressure and selective intubation of the healthy lung in those who are mechanically ventilated.
•The exception to this rule is patients with PAL due to spontaneous pneumothorax, in whom video-assisted thoracoscopic blebectomy and pleurodesis, rather than conservative measures, are preferred. (See 'Special populations' above and "Pneumothorax: Definitive management and prevention of recurrence".)
●For patients who fail conservative therapy, we prefer a systematic approach using quantitative or semiquantitative air leak, bronchial occlusion, and computed tomography techniques that measure the size and location of the APF and the integrity of the interlobar fissure adjacent to the APF (algorithm 1). (See 'Patients who fail supportive care' above and 'Quantifying the air leak' above and 'Localizing the air leak (sequential balloon occlusion)' above and 'Evaluation of fissure completeness' above.)
•For patients with APF associated with minimal collateral ventilation (eg, interlobar fissure >90 percent complete) and significant (eg, ≥50 percent) reduction of the leak on bronchial occlusion, we suggest bronchoscopic therapy rather than continued conservative therapy, surgery, ambulatory drainage devices, or pleural procedures (Grade 2C). No bronchoscopic method is superior, but we prefer endobronchial valve placement rather than spigots or occlusive materials based upon the rationale that they are easily placed/removed, better studied, and approved as a humanitarian use device. (See 'Minimal collateral ventilation and significant air leak reduction on bronchial occlusion' above.)
•For patients with significant collateral ventilation (eg, interlobar fissure ≤90 percent complete) or minimal reduction of the leak on bronchial occlusion (eg, <50 percent), we suggest surgical repair rather than continued conservative therapy, bronchoscopic interventions, ambulatory devices, or pleural procedures (Grade 2C). (See 'Significant collateral ventilation or minimal reduction of air leak on bronchial occlusion' above.)
●For patients who fail or are not candidates for bronchoscopic or surgical approaches, ambulatory drainage devices and nonsurgical pleural procedures are options (eg, blood patch or chemical pleurodesis). They are also considered additive when surgical or bronchoscopic methods achieve a partial response only. Choosing among them should be individualized and depends upon fistula size, local expertise, and patient preference. (See 'Refractory patients' above.)
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟
