Endobronchial closure of bronchopleural fistulas with Amplatzer vascular plug

Abstract

OBJECTIVE

Bronchopulmonary fistula (BPF) is a severe complication following lobectomy or pneumonectomy and is associated with a high rate of morbidity and mortality. We have developed a novel minimally invasive method of central BPF closure using Amplatzer vascular plug (AVP) device that was originally designed for the transcatheter closure of vascular structures in patients with small BPF.

METHODS

Patients with BPFs were treated under conscious sedation by bronchoscopic closure of BPFs using AVP. After locating the fistula using bronchography, the self-expanding nitinol made AVP occluder to be delivered under direct bronchoscopic guidance over a loader wire into the fistula followed by bronchography to assure correct device positioning and sealing of the BPF.

RESULTS

Six AVPs were placed in five patients, four males and one female, with a mean age of 62.3 years (range: 51–82 years). The underlying disorders and etiologies for BPF development were lobectomy (two patients), pneumonectomy for lung cancer (one patient), lobectomy due to necrotizing pneumonia (one patient), and post-tracheostomy tracheo-pleural fistula (one patient). In all the patients, the bronchoscopic procedure was successful and symptoms related to BPF disappeared following closure by the AVP. The results were maintained over a median follow-up of 9 months (range: 5–34 months).

CONCLUSIONS

Endobronchial closure using the AVP is a safe and effective method for treatment of small postoperative BPF. The ease of their implantation by bronchoscopy under conscious sedation adds this novel technique to the armatorium of minimally invasive modalities for the treatment of small BPF.

INTRODUCTION

Central bronchopulmonary fistulas (BPFs) are associated with high morbidity and mortality [1, 2]. They occur as an uncommon, but often severe, complication of surgical procedures involving pulmonary resection, in particular pneumonectomy, with an estimated incidence ranging between 3% and 28% [3, 4]. The incidence of BPF following lobectomy is significantly lower (0.5%) [3]. The mortality rate associated with BPF ranges from 16% to 72% [3, 4]. BPFs are significantly more prevalent in patients with malignant, as opposed to benign, conditions and additional risk factors for their development include radiation and chemotherapy [1–4].

Treatment for central BPF includes a range of surgical and medical techniques: chest drains, Eloesser muscle flap, omental flap, transsternal bronchial closure, thoracoplasty, and prolonged antibiotic regimens [3–6]. Endoscopic treatment of BPF is based on the delivery of biological glue, coils, covered stents, and sealants to the BPF site [7–14]. The degree of success is variable, and depends on the underlying disease and the proximity and size of the fistulas, with larger fistulas having poor closure rates [8–12]. The use of Watanabe spigot has been described as a treatment option for closure of small BPF and alveolo-pleural fistulas [15].

In previous reports [16, 17], we described a novel method of large BPF closure in 12 patients using the Amplatzer septal and ductal occluder (AD) that is traditionally used for transcatheter closure of atrial septal defects and patent ductus arteriosum. This novel treatment is suitable only for patients with large BPF and long enough tract communicating between the bronchial system and the pleural cavity. For patients with small BPF that are sinus shaped, another mode of treatment is necessary. The Amplatzer vascular plug (AVP) is a new occluding device made of nitinol wires that is increasingly being used for transcatheter embolization in the peripheral vasculature and occlusion of abnormal vessel communications [18–20]. In the current study, we describe for the first time the use of AVP for bronchoscopic closure of postoperative BPF that are too small to be treated by AD.

METHODS

Amplatzer vascular plug (Fig. 1)

Figure 1:

AMPLATZER^® vascular plug used for bronchoscopic closure of bronchopleural fistulae.

The AVP (AGA Medical, Golden Valley, MN, USA) is a self-expandable cylindrical occluding device made of 144 nitinol mesh wires that allow the device to compress inside a catheter, and then when released from the delivering catheter, return to its shape to occlude the target vessel or in our case the airway lumen. The device has platinum markers on both ends. A stainless steel micro-screw is welded to one of the platinum marker bands, which allows attachment to the 135-cm-long delivery cable. The AVP is available in diameters ranging from 4 to 16 mm, in 2-mm increments. It is preloaded in a loader and delivered through currently available guiding catheters in sizes ranging from 5 to 8F. Since the AVP is a flexible nitinol wire mesh, it adjusts to the shape of airway and thus oversizing prevents device migration after deployment.

TECHNIQUE

The bronchoscopic procedure was carried out under conscious sedation at the bronchoscopy suite; the patient is placed in the supine position and nasal oxygen is administered at 2–4 l min⁻¹. Patients receive topical anesthesia with lidocaine, IV alfentanil (0.5 mg), and intermittent boluses of IV midazolam (3–5 mg). No other sedating or hypnotic agents were used. The fiberoptic bronchoscope (BF-1T240 Video Bronchoscope; Olympus Corporation, Tokyo, Japan, or Pentax FB-18X Fiber Bronchoscope, Pentax, Tokyo, Japan) was introduced transnasally and advanced to view the BPF. Bronchography was performed to outline the anatomy of the fistulas, in particular its length. The size and type of AVP chosen were according to the diameter of the bronchial stump and the length of the fistula with the intent that the device would completely cover the fistulous connection and fit inside the stump. The delivery cable was introduced through the working channel of the bronchoscope, allowing direct visualization of the deployment. Once positioned by holding the delivery shaft steady and pushing the plug forward, the AVP was released by rotating the delivery cable counter clockwise under both direct vision and fluoroscopic guidance. Following its deployment, a dynamic bronchogram was performed by injecting contrast medium through a catheter to check device position in relation to the stump and fistula. Representative figures of study patients are presented in Figs. 2–5. The bronchoscope and sheath were removed from the airway, and the patients were transferred to the recovery room and were discharged following a chest X-ray.

Figure 3:

The RLL stump BPF after Amplatzer vascular plug placement.

Figure 4:

BPF in the anterior segment of the right upper lobe (RUL) following right bi-lobectomy.

Figure 2:

A stump bronchopleural fistula (BPF) following right lower lobe (RLL) lobectomy.

Figure 5:

The RUL BPF after Amplatzer vascular plug closure.

RESULTS

Overall, five patients, four males and one female, with a mean age of 63.7 years (range: 51–82 years), were treated at our institution with bronchoscopic AVP implantation between January 2008 and October 2010. In one patient, two AVPs were implanted. Written informed consent was obtained from all the patients before the procedure, and all the patients were informed of the fact that an off-label use of the device is being employed. The underlying disorders and etiologies for BPF development were surgery for lung cancer in four patients, trachea-pleural fistula due to previous tracheostomy in one patient, and right bi-lobectomy due to necrotizing pneumonia in one patient. In the latter patient, two AVPs were required to seal two adjacent BPFs. The AVPs were implanted in stumps in the right lower lobe in two patients, in the left lower lobe in one, and in the anterior segment of the right upper lobe in one, and in the upper trachea in one patient. There were no immediate complications related to the procedure and all patients were discharged within 3 h. All the patients underwent follow-up bronchoscopy at 1 and 3 months that confirmed the proper placement of the device. In all the patients, the procedure was successful and symptoms related to BPF disappeared following closure by the AVP. The results were maintained over a median follow-up of 9 months (range: 5–34 months).

DISCUSSION

Various applications of the AVP have been previously reported, including treatment of Pulmonary arteriovenous malformations, anomalous venous connections (left-sided hepatic vein–coronary sinus, coronary fistula, coronary saphenous vein graft aneurysm, patent ductus arteriosus, patent ductus venosus, aorticopulmonary collaterals, azygous and hemiazygous veins, and left superior vena cava), and internal iliac artery (IIA) aneurysms with or without abdominal aortic aneurysm [17–19].

Endoscopic treatment of BPF is based on the delivery of biological glue, coils, covered stents, and sealants to the BPF site [6–13]. The degree of success is variable, and depends on the underlying disease and the proximity and size of the fistulas, with larger fistulae having poor closure rates [7–11]. In a previous report, Chae et al. [14] described their experience using endobronchial Watanabe spigot (EWS), for closure of alveolo-pleural fistulas. The EWS is a silicone-made bronchial filler with studs on the surface and graspable parts at both ends. It is available in three sizes, with diameters of 5, 6, and 7 mm and is radiopaque on chest radiography. The spigot is grasped by the forceps through the working channel of the bronchoscope. Once the spigot is loaded on the forceps, the bronchoscope is introduced into the patient and the spigot is inserted firmly into place as far as the subsegmental bronchi.

We present our initial experience with a novel technique for the endobronchial closure of BPF by the implantation of AVP. The technique described herein enables direct visualization of the fistula by flexible bronchoscope and fluoroscopy simultaneously, while allowing the operator to perform standard transcatheter procedures. The AVP is designed to provide effective closure of the defect, and, owing to the large range of device sizes, they can be appropriately matched to the diameter and length of the lesion. The occluders can be fully deployed, retrieved, and repositioned prior to detachment, which is an important advantage for accurate positioning and coverage of the fistula. In the bronchial tree, according to our experience with AD [15, 16], the device induces local granulation tissue formation that potentiates its occluding properties without compromising airway patency. The technique employed was well tolerated by the patients, with no side effects or complications.

The current report is the first case series published regarding the use of AVP for the closure of BPF. Our findings suggest that the technique of endobronchial AVP implantation may be suitable for BPFs that originate in the main bronchi and lobar bronchi that are too small for closure by AD. The ease of their implantation under conscious sedation at the bronchoscopy suite adds this novel technique to the armatorium of minimally invasive modalities for the treatment of BPF.

Funding

Conflict of interest: none declared.