Abstract
Bronchopleural fistula is a potentially fatal disease most often caused after pneumonectomy. Concomitant problems such as pulmonary infection and respiratory failure are typically the main contributors to patient mortality because of the improper contact between the bronchial and pleural cavity. Therefore, bronchopleural fistulas need immediate treatment, which requires the accurate location and timely closure of the fistula. Currently, bronchoscopic interventions, because of their flexibility and versatility, are reliable alternative therapies in patients for whom surgical intervention is unsuitable. Possible interventions include bronchoscopic placement of blocking agents, atrial septal defect (ASD)/ventricular septal defect (VSD) occluders, airway stents, endobronchial valves (EBVs) and endobronchial Watanabe spigots (EWSs). Recent developments in mesenchymal stem cells (MSCs) transplantation technology and three-dimensional (3D) printed stents have also contributed to the treatment of bronchopleural fistula, but more research is needed to investigate the long-term benefits. This review focuses on the effectiveness of various bronchoscopic measures for the treatment of bronchopleural fistula and the directions for future development.
Keywords
Introduction
Bronchopleural fistula (BPF) is defined as a communication between the bronchial trunk or segmental bronchus and the pleura due to various causes. The most common cause is post-pneumonectomy, with the frequency of occurrence estimated to be 4.5–20% after total pneumonectomy and 0.5–1% after lobectomy, 1 but the mortality rate for BPF after pneumonectomy is as high as 18–50%. 2 Patients who underwent right pneumonectomy and a right lower lobectomy had the highest incidence. 3 Other causes include necrotising infections such as pulmonary tuberculosis, pneumonia, empyema, chemotherapy or radiation therapy, and thoracic trauma.4,5
The pleural cavity is a closed latent space containing a small amount of body fluid that acts as a lubricant during respiratory movement. When it comes into contact with the outside environment, prolonged fluid and air retention allow for bacterial invasion, resulting in the development of infections such as thoracic abscesses and mediastinitis. In addition, as air partially enters the pleural cavity, the amount of air that can be effectively exchanged is reduced, leading to inadequate ventilation/blood perfusion and inevitable hypoxia or even respiratory failure.
Based on the time of onset, BPF after pneumonectomy can be classified according to the modified Le Brigand classification 6 as follows: (1) early: 1–7 days after surgery; (2) intermediate: 8–30 days after surgery; and (3) late: more than 30 days after surgery. In general, fistulas most commonly occur within 8–12 days after surgery. 1 Early-stage patients are usually caused by failure of surgical stump suturing or acute ischemic necrosis of the bronchial stump. These patients present with sudden onset of dyspnoea, decreased oxygen saturation and blood pressure, subcutaneous or mediastinal emphysema, cough, purulent sputum and possibly even life-threatening tension pneumothorax, which often requires urgent surgical repair. Patients with intermediate or advanced stages of the disease usually have fistulas due to infiltration of the tumour growth in the stump or chronic ischaemia. 6 At this time, patients often experience pneumothorax, often accompanied by severe infection, decreased immune function and malnutrition. A small number of patients have insidious manifestations, with weight loss, chronic cough and fever as the first symptoms. As such, this is often difficult to diagnose and can prevent the condition from being controlled in time.
Before the advent and development of bronchoscopic intervention, surgical intervention was the only effective treatment for BPF. 7 Accordingly, fistulas were treated through fistula repair, thoracoplasty and other means. However, patients with malignant BPF are usually undergoing antitumor therapy such as radiotherapy and chemotherapy and are not tolerant to surgery due to poor health. Even if surgery is performed, the rate of postoperative complications and the risk of death are greater. 8 In such cases, using bronchoscopic interventional procedures to treat BPF has great benefits. This review focuses on the bronchoscopic interventions for BPF and the prospects for development.
Fistula localisation
Early and accurate identification and localisation are important for the management of BPF. Imaging findings such as increased air in the pleural cavity, the appearance of new air-fluid planes, changes in pre-existing air-fluid planes, the development of tension pneumothorax and air-fluid planes exceeding ⩾2 cm are strong indicators of BPF. 1 In some patients, abnormal channels between the bronchi or lung parenchyma and pleural cavity can even be directly observed by computed tomography (CT). 9 In addition, bronchoscopy can detect most bronchial fistulas located in the main bronchus or the postpneumonectomy stump, while bronchial fistulas located distally or peripherally often require serial balloon occlusion to locate. 10 To observe the air leak, balloon occlusion is first recommended using a flexible bronchoscope at each bronchial opening on the affected side. If the leak does not stop, the entire access from the chest tube opening to the closed drain needs to be tested. This method is cumbersome, time-consuming, and requires a chest drainage bottle.
Injecting methylene blue into a thoracic drainage tube or directly into the trachea is another reliable method for locating bronchial fistulas. Under direct bronchoscopic view, methylene blue is instilled through the surgical stump or suspected target bronchial opening, and the diagnosis is aided by observing the change in fluid level and the presence of blue fluid spillage from the chest drainage device. Alternatively, retrograde instillation of methylene blue through the chest drain while under bronchoscopy observation has also proven to be a viable approach. 11 The diagnosis is confirmed when the dye is observed to enter the tracheobronchial tree from the diseased airway. This procedure is easy and inexpensive, making it a feasible alternative to serial balloon occlusion. However, as the dye will exit through the least resistant channel, this method may miss the diagnosis of multiple fistulas. Nevertheless, this drawback can then be overcome by sealing the initial diseased airway with a balloon and then repeating this technique in other airways. Yet, when the patient requires surgery, the presence of dye will compromise the ability to discriminate between mediastinal and thoracic tissues. 11
Treatment
Conservative treatment of BPF is simple, safe and non-invasive. Thoracic drainage is usually the first step of intervention. The placement of a chest tube not only drains excess gas and fluid but also serves as a delivery channel for sclerosing agents to facilitate the performance of pleural fixation. 4 In addition, prophylactic use of adequate antibiotics and mechanical ventilation are also essential parts of the process. However, one study 12 has been published claiming that mechanical ventilation is a risk factor for BPF in patients after pneumonectomy. Therefore, when mechanical ventilation is necessary for patients with BPF, mean airway pressure should be reduced as much as possible, and an intermittent command ventilation mode with low tidal volume, reduced respiratory rate and shortened inspiratory time is appropriate. 1 Although conservative treatment does not directly heal BPF, it allows for partial closure of small fistulae (>3 mm in diameter) by removing pus and maintaining fluid balance or by bronchoscopic assistance. 13 In recent years, a retrospective study found that the success rate of the bronchoscopic technique was 71–92% when the fistula was distal and small (<6 mm). 14 In addition, bronchoscopic treatment may serve not only as the first treatment but may also provide a bridge to subsequent related treatments, offering new treatment pathways for patients with BPF. The advantages and disadvantages of the various treatment methods and the scope of their application are shown in Table 1. The success and complication rates of their BPF treated by each device in relevant literature are also shown in Table 1. Some of the values can range widely and may depend on the patient’s conditions and the procedures used in the study. In summary, the treatment of BPF is individualised and adaptive.
The advantages, disadvantages, success rate, complication rate and applications of different devices.
ASD, atrial septal detect; BPF, bronchopleural fistula; EBVs, endobronchial valves; EWS, endobronchial Watanabe spigot; VSD, ventricular septal detect.
Blocking agent
A variety of blocking agents have been used in BPF through bronchoscopy, which works mainly by inducing or stimulating granulation tissue formation, leading to cell proliferation, scarring and finally fistula closure. As early as 1977, two articles32,33 reported the successful treatment of BPF using glue and lead shots for the first time, respectively. Since then, ethanol, fibrin glue, albumin-glutaraldehyde tissue adhesive, oxidised regenerated cellulose, ethyl-2-cyanoacrylate and silver nitrate have been reported successively.34–39 Although not commonly used, argon plasma coagulation can be an alternative treatment for small and uncomplicated BPF cases. 40 Recently, Alcaraz et al. 41 explored the surgical treatment of bronchial fistulas with ethanolamine oleate and showed complete closure in 75% (6/8) of patients. However, no randomised controlled studies have been conducted that compare the advantages, disadvantages and safety of various occlusion agents. Furthermore, most successful cases involve fistulas of the peripheral type; therefore, this technique may not be widely used for BPF closure. The choice of occlusion material still depends on the size of the fistula and the patient’s underlying condition.
Atrial septal defect/ventricular septal defect occluders
Atrial septal defect (ASD)/ventricular septal defect (VSD) occluders were first used to treat ASD/VSD of the heart. Kramer et al. 42 described the effective treatment of two patients with BPF after pneumonectomy using Amplatzer ASD devices for the first time, heralding a broad increase in prospects for the treatment of BPF with ASD/VSD occluders. The device has a double-disc mesh structure made of woven nickel-titanium alloy wire with self-expanding properties. It can contract within the catheter and be delivered via catheter into the fistula site for subsequent release to seal the fistula by the waist, with the double discs located on each side of the fistula. The waist size varies from 4 to 40 mm, so the device can be matched to the size of the fistula and is equally suitable for patients with fistula diameters ⩾8 mm. While there is no standard for the selection of the waist diameter of the occluders, Motus et al. 43 recommended that the waist should be 30% wider than the diameter of the fistula. In congenital heart diseases, the size of occluders chosen should be comparable to or slightly more than the diameter of the fistula (within 2 mm). 44 In a study of 31 patients with BPF, Fruchter et al. 16 found that 96% of patients had immediate symptomatic relief after treatment with Amplatzer occluders (19 cases of ASD occluders and 12 cases of vascular plugs) and performed well at subsequent follow-up times. The device induced local granulation tissue formation, which enhanced their fistula occlusion without affecting airway patency. Zhang et al. 18 also invented a ‘sheath-free method’ for the placement of the ASD occluder, making the placement more convenient and efficient in clinical use. However, there are some reports that the device carries the risk of displacement and infection.13,17 At present, the treatment of BPF with cardiac occluders is still ‘off-label’ and is contingent on the specifics of the condition and the wishes of the patient and their family. Thus, there is a need for continuous improvement of the device in the future, to make it more adaptable to the environment of the airway to provide treatment protection for otherwise inoperable patients.
Airway stent
Airway stenting is more widely used in malignant BPF. It provides immediate symptomatic relief and significantly improves the patient’s quality of life. Airway stents include metal and silicone variants. There are advantages and disadvantages to various stents, but the needs of the individual patient, as well as the condition of the fistula, must be taken into account when deciding which stent to use.
Silicone stent
The Dumon silicone stent is the most commonly used stent invented by Dumon and was first used in patients with external compressional stenosis of the main airway. 45 Silicone stents can be classified as straight stents or Y-stents. Straight stents are mainly used for lesions involving the upper or middle trachea or mainstem bronchi, while Y-stents are most suitable for lesions involving the lower trachea, tracheal carina, mainstem bronchi and secondary carina. 46 In BPF treatment, silicone stents are rarely used because of their limited type and poor adaptability. However, silicone stents can overcome this disadvantage through manual modifications on-site, such as cutting, suturing and snapping. Moreover, silicone stents have the advantages of durability, ease of removal and low cost compared with metal stents. Zeng et al. 19 retrospectively analysed the clinical data of 17 patients with BPF who underwent bronchoscopic placement of a modified Y-shaped silicone stent, and 16 patients were successfully stented (initial success rate: 94.1%). All patients had improved respiratory symptoms, with a clinical success rate of 76.5%. For refractory BPF, placement of the modified silicone stent could be an effective and safe option. However, the stent needs to be completed under a rigid bronchoscope, which is more demanding for the operator. In addition, silicone stents have been used in conjunction with covered metal stents in patients with serious malignant airway stenosis and fistulas. 47 This hybrid stent approach involves a Y-shaped silicone stent to maintain central airway ventilation and a metal-covered stent placed distal to the Y-shaped silicone stent to reconstruct the airway from the main ramus to the distal main bronchus and to seal the fistula. The results of that study showed immediate relief of dyspnoea and improvement in various scores after placement of the hybrid stent. Furthermore, there were no serious complications associated with the hybrid stent within 6 months of placement. However, that study lacked control for single-stent placement, and the sample size was small. Thus, extensive studies exploring the prolonged benefits and safety of hybrid stents are warranted.
Metal stent
In 1989, Simonds et al. 48 first placed a self-expanding metal stent into the trachea for bronchial obstruction. Subsequently, metal stents have gained widespread use in various respiratory diseases. Yet the use of tracheal stents to treat benign tracheal diseases has been warned against by the US Food and Drug Administration (FDA) because of the risk of stent-related complications and difficulties associated with their removal. 49 In BPF, metal stents are not routinely used, but in a few carefully selected patients, self-expanding covered stents may be used with caution. Furthermore, the stent is not left in the body for too long, to reduce complications if the condition permits. The metal mesh provides good support and is resistant to fracture, while the laminate surface is effective in sealing the fistula. Menna et al. 20 evaluated the efficacy of conical covered stents in the treatment of postoperative BPF and found that permanent closure of bronchial dehiscence was achieved in all patients without recurrence. In almost all patients (91.7%), the stent was successfully removed and not replaced due to the patient recovering from the fistula. Covered bullet stents, which have an occlusive bullet at one end, also have been shown to be a feasible and safe option for treating BPFs in several studies.50,51 Han et al. 21 described their experience with 148 patients undergoing postoperative BPF using customised, covered metallic stents. The stent shape can be adjusted according to the position and length of the bronchial stump. They achieved a success rate of 96.6% in patients with BPF who underwent first time airway occlusion stent implantation. This still underscores the importance of individualised treatment for BPF. However, stent placement in this article was done under fluoroscopy, which, if assisted by bronchoscope, may improve the success rate of sealing the bronchial fistula and reduce the risk of incorrect placement. The customisation of metal supports often takes time, and how to process them in the field is still the subject of constant investigation.
Stent management
Tracheal stents are correlated with improved overall survival and reduced mortality in several studies. 52 However, both metal and silicone stents are foreign bodies, and complications such as infection, granulation tissue formation, stent migration and mucus plugging are inevitable. Currently, some studies comparing silicone and metal stents in malignant airway disease showed no significant differences in symptom relief, safety, complication rates or survival between the two types of stents.53,54 In addition, stent placement into the trachea is not a one-time event and requires regular management and maintenance. Bronchoscopy is recommended 48–72 h after successful stent placement to confirm whether the stent is in place. 55 Then, bronchoscopy should be followed weekly to determine stent status and clear airway secretions. The follow-up time can be extended to once a month after the patient’s condition improves. The data suggest that bronchoscopy within 4–6 weeks after stent placement may help in the early detection of complications and their subsequent management. 56 This hints that patients should be actively followed up and reviewed to lower the probability of complications.
Endobronchial valves
The endobronchial valve (EBV) is a novel device for the management of BPF, further categorised as Zephyr endobronchial valve (ZEBV, Pulmonx Inc., Neuchatel, Switzerland) and intrabronchial valve (IBV, Spiration Inc., Redmond, WA, USA), originally designed for bronchoscopic lung decompression procedures. Following approval of the IBV by the FDA in 2008 under the Humanitarian Device Exemption Program for patients with persistent air leaks after parenchymal lung resection, the bronchial valve has been successfully used for postoperative persistent BPF occlusion.25,57,58 The ZEBV was also approved by the FDA in 2018 to treat breathing difficulties associated with severe emphysema. The device is a one-way valve structure that allows air and fluids out but keeps them from getting into the diseased part of the lung, thereby reducing air leaks and the incidence of obstructive pneumonia. It can also be placed through a flexible bronchoscope under general or local anaesthesia. A recent study 27 observed that EBVs therapy was effective in more than 70% of patients with BPF and prolonged air leaks, with no intraoperative or postoperative complications. An interesting finding is that a higher comorbidity burden appears to be a risk factor for delayed recovery, regardless of the EBVs implantation status. 59 To reduce the risk of infection, migration and granulation tissue formation, it is now recommended that the valve be removed 4–6 weeks after placement or upon resolution of the air leak. 60 Furthermore, the cost of these valves is a drawback that prevents their widespread use in clinical practice.
Endobronchial Watanabe spigot
The endobronchial Watanabe spigot (EWS), first invented by Watanabe and his team in 1991, 61 is a silicone bronchial plug in the form of a socket with a studded surface and gripping ends. It is available in three sizes: small, medium and large, which are 5, 6 and 7 mm in diameter, respectively, allowing adaptation to different bronchial diameters. EWS can be simply placed into the target bronchus, usually the peripheral bronchus, by the ‘forceps grasp’ 30 method, the ‘curette’ method 31 or the ‘traction’ method 62 under bronchoscopy, thereby preventing air from entering the diseased bronchus. Bronchial obstruction with the EWS is an appropriate treatment for BPF when the target bronchus is well-defined. Himeji et al. 29 found in a retrospective study that the application of the EWS had an 85.7% success rate for cases of pneumothorax, pyothorax with bronchial fistula and postoperative air leakage. The biggest drawback of the EWS is migration. To prevent migration, it is recommended that EWS should be inserted as far as possible, and appropriate medications should be used to reduce cough. 30 The specific timing and indications for their placement still require considerable research.
Outlook
Currently, mesenchymal stem cells (MSCs) are starting to be used for the treatment of BPF due to their strong differentiation capacity and weak immunogenicity. MSCs can be delivered surgically or bronchoscopically to the lesion site and left to divide and differentiate into specific cells to achieve fistula closure. In 2015, Petrella et al. 63 successfully treated a patient with BPF for the first time with bone marrow MSCs. In addition, the use of umbilical cord MSCs for BPF closure after pneumonectomy has also been reported. 64 Bottoni et al. 65 performed a study on the treatment of post-pneumonectomy BPF with bronchoscopic autologous fat fillers and observed a 100% success rate without rejection, with positive efficacy observed even in fistulas >8 mm in diameter. MSCs transplantation technology may be a new option for treating severe lung diseases with a broad developmental perspective. It is currently used only in cases where other interventions have failed, and large-scale clinical trials are still needed to investigate the effectiveness of MSCs in such settings.
The advent of three-dimensional (3D) printed stents has opened a new era of stent placement. Almost all stents are held in place by the friction created by the pressure applied to the airway wall. Hence, the stent must fit perfectly into airway wall to seal the fistula. As a result, it is important to perform stent placement without stretching the fistula to maximise the benefit. In large part, this can be done by cutting, suturing or ordering special forms of stents to simulate the human airway structure as closely as possible, which is usually time-consuming and subject to error. 52 3D-printed stents can be designed to fit an individual in a matter of weeks or even hours by deriving the appropriate model from CT or bronchoscopic images. 66 The 3D printing technology is widely used, except for metal stents 67 and silicone stents, 68 biodegradable stents 69 and drug-eluting stents 70 that have been reported for airway stenosis. In the future, these are expected to be used in the treatment of BPF. With the rapid advances in bronchoscopic treatment technology, treatment options for large fistulas can now be explored more thoroughly.
Conclusion
The treatment of BPFs often requires a multidisciplinary approach, including thoracic surgery and respiratory intervention. There are no uniform guidelines recommending standardised treatment of BPF, and bronchoscopic treatment is usually recommended for fistulas ⩽8 mm in diameter, but the closure of large fistulas remains a challenge. It is worth acknowledging that bronchoscopic treatment plays an integral role in BPF, and treatments should be tailored according to the patient’s individual condition.
