- Email: [email protected]
First Use of Ventricular Septal Defect Occlusion Device for Endoscopic Closure of an Esophagorespiratory Fistula Using Bronchoscopy and Esophagoscopy
CHEST First Use of Ventricular Septal Defect Occlusion Device for Endoscopic Closure of an Esophagorespiratory Fistula Using Bronchoscopy and Esophagoscopy* Thomas Rabenstein, MD, PhD; Christoph Boosfeld, PhD; Rolf Henrich, MD; and Christian Ell, MD, PhD
A 70-year-old woman presented with a persistent, nonmalignant esophagorespiratory fistula. Since other treatment options failed or were denied, an experimental nonsurgical therapy was performed. A self-expanding ventricular septal defect (VSD) occlusion device (Amplatzer; AGA Medical Corporation; Golden Valley, MN) was bronchoscopically introduced to close the fistula. The double umbrella-like occlusion device is made from nitinol mesh and closes luminal contact between the esophageal and bronchial walls, with its waist filling out the fistula itself. The geometry of the occluder system can in theory be designed according to individual purposes and needs. The performed treatment was safe and successful, and the patient remained asymptomatic for 1 year after the first presentation. The treatment of chronic nonmalignant esophagorespiratory fistulas can be difficult. The self-expanding VSD occluder system described in this case might be useful in patients who are not surgical candidates. (CHEST 2006; 130:906 –909) Key words: Amplatzer; esophagorespiratory fistula; experimental treatment; interventional endoscopy; nonmalignant; nonsur*From the Department of Medicine II (Drs. Rabenstein, Henrich, and Ell), Dr. Horst-Schmidt Klinik Wiesbaden, Wiesbaden; and DRABO Medizintechnik–Amplatzer Support Europe (Dr. Boosfeld), Cologne, Germany. The authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Manuscript received October 3, 2005; revision accepted December 19, 2005. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Thomas Rabenstein, MD, PhD, Klinik Innere Medizin II, Dr.-Horst-Schmidt-Klinik Wiesbaden, Akademisches Lehrkrankenhaus der Johannes-Gutenberg-Universita¨t Mainz, Ludwig-Erhard-Str 100, 65199 Wiesbaden, Germany; e-mail: [email protected] DOI: 10.1378/chest.130.3.906 906
Selected Reports
gical treatment; therapeutic endoscopy; ventricular septal defect occluder Abbreviation: VSD ⫽ ventricular septal defect
picture of esophagorespiratory fistulas conT hesistsclinical of chronic cough, bouts of cough when swallow-
ing liquids, recurrent pulmonary infections, lung suppurations, and hemoptysis.1–3 In adults, communications between the esophagus and the tracheobronchial tree are mostly acquired.4 Most fistulas have a malignant origin. In these cases, implantation of covered stents either on the esophageal route or on the bronchial route is the treatment of choice because concomitant tumor stenosis prevents the migration of the self-expanding metal mesh prosthesis.5,6 Benign causes of such fistulas are iatrogenetic (ie, intubation, sclerotherapy, and surgery) and infections, inflammatory conditions, and corrosive ingestion. Developmental abnormalities are summarized as congenital fistulas; they are found three times more commonly to the right than to the left bronchial tree.7,8 A congenital fistula that remains asymptomatic until adulthood is rare. The recommended therapy of esophagorespiratory fistulas is surgery that consists of the division of the fistula and the excision of any permanently damaged lung segments.4,7,8 Alternatively, a simple stapling has been advocated for elderly patients. Endoscopic obliteration has been recommended for patients who decline to undergo or are unfit for thoracotomy.9 We report on the first use of a ventricular septal defect (VSD) occlusion device (Amplatzer; AGA Medical Corporation; Golden Valley, MN) for combined bronchoscopic and esophagoscopic closure of a benign esophagorespiratory fistula.
Case Report A 70-year-old woman presented with aggravated chronic cough and consecutive worsening of urinary incontinence. The cough was particularly tormenting at night; also, after drinking, the patient had bouts of cough, even though she had never had pneumonia or lung suppuration and did not show an increased body temperature. Laboratory findings including blood cell counts and C-reactive protein were completely normal. Functional pulmonary tests and blood gas analysis also showed no abnormalities. The clinical picture was attributed to an esophagorespiratory fistula of unknown origin, which had been diagnosed 20 years previously, after the accidental swallowing of a fish bone. Since the patient was only afflicted with tussive irritation, no treatment was performed. Radiologic evaluation (CT and barium study) confirmed an esophagorespiratory fistula to the right pulmonary segment 6, but during routine esophagoscopy, the GI end of the fistula could not be found. Bronchoscopy showed retention of fluid in the distal pulmonary segments of the right side and a bronchial edema of segment 6. After esophageal application of methylene blue, the blue fluid was immediately visible in right segment 6, but the Selected Reports
Figure 1. Contrast application during esophagoscopy resulting in opacification of the right bronchial tree via esophagorespiratory fistula.
guidewire. The course of this wire was now from stomach through the esophagus, then the fistula and trachea, and finally out of the patient through a side port of the anesthesia mask. Over this wire, we introduced a regular Amplatzer delivery sheath (8F/60 cm/45° tip) into the esophagus and advanced a 6-mm Amplatzer muscular VSD occluder (model 9-VSD-MUSC006) under fluoroscopic and gastroscopic (visual) control. The distal umbrella of the occluder pushed in the esophagus. Under tension of the whole assembly, the proximal umbrella unfolded in the bronchus of segment 6, while the waist filled the fistula itself (Fig 3, 4). Procedure time was 65 min, with radiation time of 10 min. A single fever peak of 38.9°C developed 2 h after the procedure, which was treated with IV applied imipenem for 3 days. No further side effects or complications were detected. On day 1 after the procedure, endoscopic evaluation showed correct position of occluder and closure of the fistula. The patient could thus be discharged without any medical complaints. Ten weeks after placement of the occluder, the patient returned for a routine control visit. She reported a symptom-free period of 3 weeks. In week 4 after insertion of the occluder, the chronic cough had recurred; in week 9, halitosis, expectorations,
exact localization of the bronchial end of the fistula was not possible. Bronchial application of methylene blue into the right segment 6 finally allowed esophagoscopic identification of the GI end of the fistula and consecutive guide wire and catheter cannulation. After the fistula had been brushed with a cytology brush, 2 mL of fibrin glue were applied in order to induce inflammation-related closure. Five months after fibrin glue injection, the patient returned. Esophagoscopy and bronchoscopy showed an unchanged picture. The fistula was still open, and swallowed fluids flowed directly into the bronchial system (Fig 1). The patient now agreed to an experimental treatment option by using the self-expanding VSD occluder (Fig 2) for closing the symptomatic fistula under general anesthesia. Due to the unfortunate caudocranial orientation of the fistula, a so-called guidewire loop had to be performed. A guidewire (Terumo 0.035; Terumo Corporation; Tokyo, Japan; 150 cm) was introduced into the esophagus and then extracted via the fistula and the bronchial pathway using a biopsy forceps. Subsequently, the wire was exchanged for a 260-cm Amplatzer super-stiff
Figure 2. The completely expanded VSD occluder for treatment of muscular VSD defects. www.chestjournal.org
Figure 3. Top, A: the VSD occluder through the bronchial route; the proximal position is controlled endoscopically by esophagoscopy. Bottom, B: the VSD occluder during release; the esophageal end is completely expanded. CHEST / 130 / 3 / SEPTEMBER, 2006
907
Figure 4. The VSD occluder in its final position; both ends are completely expanded.
and minor hemoptysis developed. Clinically, she had no fever, and C-reactive protein as well as leukocytes and hemoglobin levels were normal. Chest radiography showed that the occluder was slightly dislocated toward the bronchial system, and the esophageal end was only incompletely expanded. CT ruled out pulmonary abscess and infiltration. During esophagoscopy, the esophageal end of the fistula was clearly visible; fluoroscopically, the occluder appeared retracted into the mediastinum and was incompletely expanded. However, esophageally applied fluid did not reach the bronchial system. Bronchoscopy revealed increased mucosal vulnerability in segment 6 to be the origin of hemoptysis, but the bronchial end of the occluder could not be reached for extraction. The patient recovered under antibiotic therapy (piperacillin and combactam, three 4.5-g doses per day IV for 3 days and then mofloxacin, 400 mg po qd for 10 days) and could be discharged 5 days after hospitalization. Six months after the procedure, another follow-up examination was performed. The patient felt good, and she had neither fever nor cough or hemoptysis. Bronchoscopically, the occluder could not be reached because of the peripheral position; however, there were no signs of mucosal injury or hemorrhage. Esophagoscopy displayed a small fistula opening, but fluoroscopically the occluder was now completely expanded closely below the esophageal wall. Contrast studies proved complete closure of the fistula by the implanted VSD occluder (Fig 5). One year after closure of the fistula, the patient reported one further period of expectorations and minor hemoptysis, which were treated successfully with oral antibiotics for 8 days (mefloxacin, 400 mg/d). The initial chronic cough and the aspirations at night did not recur. Endoscopic and radiologic evaluation showed no changes: the fistula was closed by the completely expanded occluder, although the esophageal umbrella was lying in the mediastinum below the esophageal wall.
Discussion This is the first report on the flexible endoscopic closure of a chronic, nonmalignant esophagorespiratory fistula by the use of a self-expanding VSD occluder system.10 The reported fistula was probably acquired because a congenital fistula that remains asymptomatic until adulthood is rare. Increasing problems induced by chronic fluid aspiration developed after a long symptom-free period. For908
Figure 5. Contrast filling of the esophagus demonstrating long-term closure of the fistula 6 months after insertion.
tunately, no major infectious complications had occurred, but a therapeutic intervention to close the chronic fistula and thus avoid future complications seemed necessary. Esophagoscopic intervention by fibrin glue injection failed to close the fistula, and the patient did not agree with the surgical treatment. Finally an experimental method using the VSD occluder system was offered to close the fistula, and the patient agreed. The double umbrella-like occlusion device is made from nitinol mesh; it closes luminal contact between the esophageal and bronchial walls, with its waist filling out the fistula itself. The geometry of the occluder system can in theory be designed according to individual purposes and needs, but for this first experimental use we chose a standardized method.11 The prerequisites for placement of the system could be achieved by advanced endoscopic techniques for localization and guidewire cannulation of the fistula. Finally, it was only possible to introduce the system via the bronchial route, since an almost 180° angle between the fistula and esophagus thwarted a transesophageal approach. Correct positioning was achieved under fluoroscopic and gastroscopic guidance. The VSD occluder is able to be repositioned and retrieved,11 and we were prepared to extract and replace it with another occluder if the waist would not match with the length of the fistula. Fortunately, this was not necessary, as the fistula was closed completely after spontaneous retraction of the occluder through the esophageal wall and expansion in this position. Compared to vascular applications in which the VSD occluder is completely covered by an endothelial layer after a very short time, this cannot be expected in GI or bronchial use. But as long as the discs of the device are positioned closely to the organ walls and do not induce luminal obstruction, no long-term problems may be expected in this respect. Additionally, the device can be extracted if any significant problems occur. Another significant difference compared to vascular use is the possibility of combined radiologic and direct visible control during placement and release during endoscopic proceSelected Reports
dures. Thus, the use of VSD or atrial septal defect occluder systems is also attractive for endoscopic treatment of other acquired fistulas. For example, we used an atrial septal defect occluder for successful closure of a biliodigestive fistula between the bulbus duodeni and common bile duct, which had induced ascending food impaction, cholangitis, and bile sludge formation (sump syndrome). The treatment of chronic nonmalignant esophagorespiratory fistulas can be difficult. The self-expanding VSD occluder system described in this case might be useful in patients who are not surgical candidates. ACKNOWLEDGMENT: We thank Sabine Nunius for translation of this article.
References 1 Lim KH, Lim YC, Liam CK, Wong CM-M. A 52-year-old woman with recurrent hemoptysis. Chest 2001; 119:955–957 2 Kim IS, Lee GH, Jang MK, et al. Bronchial fistula. Gastrointest Endosc 2004; 59:696 – 697 3 Cook RJ, Anderson LD Jr, Adler DG, et al. Oesophagorespiratory fistula [letter]. Gastrointest Endosc 2003; 58:255 4 Csikos M, Horvath O, Petri A, et al. Surgical treatment of acquired, benign oesophago-respiratory fistulas [in German]. Zentralbl Chir 2004; 129:104 –107 5 May A, Ell C. Palliative treatment of malignant esophagorespiratory fistulas with Gianturco-Z stents: a prospective clinical trial and review of the literature on covered metal stents. Am J Gastroenterol 1998; 93:532–535 6 Ell C, May A. Self-expanding metal stents for palliation of stenosing tumors of the esophagus and cardia: a critical review. Endoscopy 1997; 29:392–398 7 Acosta JL, Battesby JS. Congenital tracheoesophageal fistula in the adult. Ann Thorac Surg 1974; 34:1751–1756 8 Azoulay D, Regnard JF, Magdeleinat P, et al. Congenital respiratory-oesophageal fistula in the adult: report on none cases and review of the literature. J Thorac Cardiovasc Surg 1992; 104:381–384 9 Ogunmola N, Wyllie R, McDowell K, et al. Endoscopic closure of oesophagobronchial fistula with fibrin glue. J Pediatr Gastroenterol Nutr 2004; 38:539 –541 10 Thanopoulos BD, Tsaousis GS, Konstadopoulou GN, et al. Transcatheter closure of muscular ventricular septal defects with the Amplatzer ventricular septal defect occluder: initial clinical applications in children. J Am Coll Cardiol 1999; 33:1395–1399 11 Parsi A, Bruch L, Szurawitzki G, et al. Transcatheter closure of muscular ventricular septal defects in two patients after myocardial infarction. J Interv Cardiol 2001; 14:219 –221
Management of Obstructing Pulmonary Broncholithiasis With Three-Dimensional Imaging and Holmium Laser Lithotripsy* J. Scott Ferguson, MD, FCCP; Jonathan M. Rippentrop, MD; Bernard Fallon, MD; Alan F. Ross, MD; and Geoffrey McLennan, MD, PhD www.chestjournal.org
Major airway obstruction due to broncholithiasis produces significant morbidity, and management is difficult. Many of the patients are elderly and are not good candidates for surgical removal. Bronchoscopic removal may be limited due to anatomic considerations, skill of the bronchoscopist, and exposure of the patient to additional procedural risks. Preprocedural planning with three-dimensional (3D) multidetector CT (MDCT) imaging enhances the bronchoscopist’s knowledge of the relationships of the target lesions with critical structures, and improves the efficiency of the application of specific endobronchial therapies. Here we report our experience treating obstructing broncholithiasis in two patients utilizing pretreatment planning with 3D MDCT imaging, followed by bronchoscopically delivered holmium laser fragmentation of the stones. (CHEST 2006; 130:909 –912) Key words: airway obstruction; bronchoscopy; lasers; lithiasis; lithotripsy; lymph nodes Abbreviations: 2D ⫽ two dimensional; 3D ⫽ three dimensional; Ho-YAG ⫽ holmium-yttrium aluminum garnet; MDCT ⫽ multidetector CT
airway obstruction due to broncholithiasis, alM ajor though uncommon, produces significant morbidity
and is difficult to treat using the surgical and bronchoscopic methods available. Removal of broncholiths via bronchoscopy is often extremely difficult due to embedding of the stone into surrounding structures, large size, or difficulty with crushing the stone using bronchoscopic forceps.1 Surgical interventions are often required; however, many patients are poor surgical candidates, and removal of the stone may require lobectomy or pneumonectomy.2 For these reasons, we have sought alternatives to traditional surgical and bronchoscopic approaches. The Nd-YAG and holmium-yttrium aluminum garnet (Ho-YAG) lasers are possible alternatives to surgery and mechanical crushing.3,4 However, these modalities are associated with rare but important complications,5 some of which may be lessened by the use of advanced imaging *From the Departments of Internal Medicine (Drs. Ferguson and McLennan), Urology (Drs. Rippentrop and Fallon), and Anesthesia (Dr. Ross), University of Iowa, Iowa City, IA. Drs. Ferguson and Rippentrop participated equally in the preparation of this article. This work was performed at the University of Iowa. Drs. Ferguson, Rippentrop, Fallon, and Ross have no conflicts of interest to declare. Dr. McLennan is part owner of VIDA Diagnostics (Iowa City, IA), which is a company that develops three-dimensional software imaging solutions. Manuscript received October 11, 2005; revision accepted February 21, 2006. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: J. Scott Ferguson, MD, FCCP, Department of Internal Medicine, Division of Pulmonary, Critical Care, and Occupational Medicine, 200 Hawkins Dr, C-33 GH, Iowa City, IA 52242-1089; e-mail: [email protected]. DOI: 10.1378/chest.130.3.909 CHEST / 130 / 3 / SEPTEMBER, 2006
909
A 70-year-old woman presented with a persistent, nonmalignant esophagorespiratory fistula. Since other treatment options failed or were denied, an experimental nonsurgical therapy was performed. A self-expanding ventricular septal defect (VSD) occlusion device (Amplatzer; AGA Medical Corporation; Golden Valley, MN) was bronchoscopically introduced to close the fistula. The double umbrella-like occlusion device is made from nitinol mesh and closes luminal contact between the esophageal and bronchial walls, with its waist filling out the fistula itself. The geometry of the occluder system can in theory be designed according to individual purposes and needs. The performed treatment was safe and successful, and the patient remained asymptomatic for 1 year after the first presentation. The treatment of chronic nonmalignant esophagorespiratory fistulas can be difficult. The self-expanding VSD occluder system described in this case might be useful in patients who are not surgical candidates. (CHEST 2006; 130:906 –909) Key words: Amplatzer; esophagorespiratory fistula; experimental treatment; interventional endoscopy; nonmalignant; nonsur*From the Department of Medicine II (Drs. Rabenstein, Henrich, and Ell), Dr. Horst-Schmidt Klinik Wiesbaden, Wiesbaden; and DRABO Medizintechnik–Amplatzer Support Europe (Dr. Boosfeld), Cologne, Germany. The authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Manuscript received October 3, 2005; revision accepted December 19, 2005. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Thomas Rabenstein, MD, PhD, Klinik Innere Medizin II, Dr.-Horst-Schmidt-Klinik Wiesbaden, Akademisches Lehrkrankenhaus der Johannes-Gutenberg-Universita¨t Mainz, Ludwig-Erhard-Str 100, 65199 Wiesbaden, Germany; e-mail: [email protected] DOI: 10.1378/chest.130.3.906 906
Selected Reports
gical treatment; therapeutic endoscopy; ventricular septal defect occluder Abbreviation: VSD ⫽ ventricular septal defect
picture of esophagorespiratory fistulas conT hesistsclinical of chronic cough, bouts of cough when swallow-
ing liquids, recurrent pulmonary infections, lung suppurations, and hemoptysis.1–3 In adults, communications between the esophagus and the tracheobronchial tree are mostly acquired.4 Most fistulas have a malignant origin. In these cases, implantation of covered stents either on the esophageal route or on the bronchial route is the treatment of choice because concomitant tumor stenosis prevents the migration of the self-expanding metal mesh prosthesis.5,6 Benign causes of such fistulas are iatrogenetic (ie, intubation, sclerotherapy, and surgery) and infections, inflammatory conditions, and corrosive ingestion. Developmental abnormalities are summarized as congenital fistulas; they are found three times more commonly to the right than to the left bronchial tree.7,8 A congenital fistula that remains asymptomatic until adulthood is rare. The recommended therapy of esophagorespiratory fistulas is surgery that consists of the division of the fistula and the excision of any permanently damaged lung segments.4,7,8 Alternatively, a simple stapling has been advocated for elderly patients. Endoscopic obliteration has been recommended for patients who decline to undergo or are unfit for thoracotomy.9 We report on the first use of a ventricular septal defect (VSD) occlusion device (Amplatzer; AGA Medical Corporation; Golden Valley, MN) for combined bronchoscopic and esophagoscopic closure of a benign esophagorespiratory fistula.
Case Report A 70-year-old woman presented with aggravated chronic cough and consecutive worsening of urinary incontinence. The cough was particularly tormenting at night; also, after drinking, the patient had bouts of cough, even though she had never had pneumonia or lung suppuration and did not show an increased body temperature. Laboratory findings including blood cell counts and C-reactive protein were completely normal. Functional pulmonary tests and blood gas analysis also showed no abnormalities. The clinical picture was attributed to an esophagorespiratory fistula of unknown origin, which had been diagnosed 20 years previously, after the accidental swallowing of a fish bone. Since the patient was only afflicted with tussive irritation, no treatment was performed. Radiologic evaluation (CT and barium study) confirmed an esophagorespiratory fistula to the right pulmonary segment 6, but during routine esophagoscopy, the GI end of the fistula could not be found. Bronchoscopy showed retention of fluid in the distal pulmonary segments of the right side and a bronchial edema of segment 6. After esophageal application of methylene blue, the blue fluid was immediately visible in right segment 6, but the Selected Reports
Figure 1. Contrast application during esophagoscopy resulting in opacification of the right bronchial tree via esophagorespiratory fistula.
guidewire. The course of this wire was now from stomach through the esophagus, then the fistula and trachea, and finally out of the patient through a side port of the anesthesia mask. Over this wire, we introduced a regular Amplatzer delivery sheath (8F/60 cm/45° tip) into the esophagus and advanced a 6-mm Amplatzer muscular VSD occluder (model 9-VSD-MUSC006) under fluoroscopic and gastroscopic (visual) control. The distal umbrella of the occluder pushed in the esophagus. Under tension of the whole assembly, the proximal umbrella unfolded in the bronchus of segment 6, while the waist filled the fistula itself (Fig 3, 4). Procedure time was 65 min, with radiation time of 10 min. A single fever peak of 38.9°C developed 2 h after the procedure, which was treated with IV applied imipenem for 3 days. No further side effects or complications were detected. On day 1 after the procedure, endoscopic evaluation showed correct position of occluder and closure of the fistula. The patient could thus be discharged without any medical complaints. Ten weeks after placement of the occluder, the patient returned for a routine control visit. She reported a symptom-free period of 3 weeks. In week 4 after insertion of the occluder, the chronic cough had recurred; in week 9, halitosis, expectorations,
exact localization of the bronchial end of the fistula was not possible. Bronchial application of methylene blue into the right segment 6 finally allowed esophagoscopic identification of the GI end of the fistula and consecutive guide wire and catheter cannulation. After the fistula had been brushed with a cytology brush, 2 mL of fibrin glue were applied in order to induce inflammation-related closure. Five months after fibrin glue injection, the patient returned. Esophagoscopy and bronchoscopy showed an unchanged picture. The fistula was still open, and swallowed fluids flowed directly into the bronchial system (Fig 1). The patient now agreed to an experimental treatment option by using the self-expanding VSD occluder (Fig 2) for closing the symptomatic fistula under general anesthesia. Due to the unfortunate caudocranial orientation of the fistula, a so-called guidewire loop had to be performed. A guidewire (Terumo 0.035; Terumo Corporation; Tokyo, Japan; 150 cm) was introduced into the esophagus and then extracted via the fistula and the bronchial pathway using a biopsy forceps. Subsequently, the wire was exchanged for a 260-cm Amplatzer super-stiff
Figure 2. The completely expanded VSD occluder for treatment of muscular VSD defects. www.chestjournal.org
Figure 3. Top, A: the VSD occluder through the bronchial route; the proximal position is controlled endoscopically by esophagoscopy. Bottom, B: the VSD occluder during release; the esophageal end is completely expanded. CHEST / 130 / 3 / SEPTEMBER, 2006
907
Figure 4. The VSD occluder in its final position; both ends are completely expanded.
and minor hemoptysis developed. Clinically, she had no fever, and C-reactive protein as well as leukocytes and hemoglobin levels were normal. Chest radiography showed that the occluder was slightly dislocated toward the bronchial system, and the esophageal end was only incompletely expanded. CT ruled out pulmonary abscess and infiltration. During esophagoscopy, the esophageal end of the fistula was clearly visible; fluoroscopically, the occluder appeared retracted into the mediastinum and was incompletely expanded. However, esophageally applied fluid did not reach the bronchial system. Bronchoscopy revealed increased mucosal vulnerability in segment 6 to be the origin of hemoptysis, but the bronchial end of the occluder could not be reached for extraction. The patient recovered under antibiotic therapy (piperacillin and combactam, three 4.5-g doses per day IV for 3 days and then mofloxacin, 400 mg po qd for 10 days) and could be discharged 5 days after hospitalization. Six months after the procedure, another follow-up examination was performed. The patient felt good, and she had neither fever nor cough or hemoptysis. Bronchoscopically, the occluder could not be reached because of the peripheral position; however, there were no signs of mucosal injury or hemorrhage. Esophagoscopy displayed a small fistula opening, but fluoroscopically the occluder was now completely expanded closely below the esophageal wall. Contrast studies proved complete closure of the fistula by the implanted VSD occluder (Fig 5). One year after closure of the fistula, the patient reported one further period of expectorations and minor hemoptysis, which were treated successfully with oral antibiotics for 8 days (mefloxacin, 400 mg/d). The initial chronic cough and the aspirations at night did not recur. Endoscopic and radiologic evaluation showed no changes: the fistula was closed by the completely expanded occluder, although the esophageal umbrella was lying in the mediastinum below the esophageal wall.
Discussion This is the first report on the flexible endoscopic closure of a chronic, nonmalignant esophagorespiratory fistula by the use of a self-expanding VSD occluder system.10 The reported fistula was probably acquired because a congenital fistula that remains asymptomatic until adulthood is rare. Increasing problems induced by chronic fluid aspiration developed after a long symptom-free period. For908
Figure 5. Contrast filling of the esophagus demonstrating long-term closure of the fistula 6 months after insertion.
tunately, no major infectious complications had occurred, but a therapeutic intervention to close the chronic fistula and thus avoid future complications seemed necessary. Esophagoscopic intervention by fibrin glue injection failed to close the fistula, and the patient did not agree with the surgical treatment. Finally an experimental method using the VSD occluder system was offered to close the fistula, and the patient agreed. The double umbrella-like occlusion device is made from nitinol mesh; it closes luminal contact between the esophageal and bronchial walls, with its waist filling out the fistula itself. The geometry of the occluder system can in theory be designed according to individual purposes and needs, but for this first experimental use we chose a standardized method.11 The prerequisites for placement of the system could be achieved by advanced endoscopic techniques for localization and guidewire cannulation of the fistula. Finally, it was only possible to introduce the system via the bronchial route, since an almost 180° angle between the fistula and esophagus thwarted a transesophageal approach. Correct positioning was achieved under fluoroscopic and gastroscopic guidance. The VSD occluder is able to be repositioned and retrieved,11 and we were prepared to extract and replace it with another occluder if the waist would not match with the length of the fistula. Fortunately, this was not necessary, as the fistula was closed completely after spontaneous retraction of the occluder through the esophageal wall and expansion in this position. Compared to vascular applications in which the VSD occluder is completely covered by an endothelial layer after a very short time, this cannot be expected in GI or bronchial use. But as long as the discs of the device are positioned closely to the organ walls and do not induce luminal obstruction, no long-term problems may be expected in this respect. Additionally, the device can be extracted if any significant problems occur. Another significant difference compared to vascular use is the possibility of combined radiologic and direct visible control during placement and release during endoscopic proceSelected Reports
dures. Thus, the use of VSD or atrial septal defect occluder systems is also attractive for endoscopic treatment of other acquired fistulas. For example, we used an atrial septal defect occluder for successful closure of a biliodigestive fistula between the bulbus duodeni and common bile duct, which had induced ascending food impaction, cholangitis, and bile sludge formation (sump syndrome). The treatment of chronic nonmalignant esophagorespiratory fistulas can be difficult. The self-expanding VSD occluder system described in this case might be useful in patients who are not surgical candidates. ACKNOWLEDGMENT: We thank Sabine Nunius for translation of this article.
References 1 Lim KH, Lim YC, Liam CK, Wong CM-M. A 52-year-old woman with recurrent hemoptysis. Chest 2001; 119:955–957 2 Kim IS, Lee GH, Jang MK, et al. Bronchial fistula. Gastrointest Endosc 2004; 59:696 – 697 3 Cook RJ, Anderson LD Jr, Adler DG, et al. Oesophagorespiratory fistula [letter]. Gastrointest Endosc 2003; 58:255 4 Csikos M, Horvath O, Petri A, et al. Surgical treatment of acquired, benign oesophago-respiratory fistulas [in German]. Zentralbl Chir 2004; 129:104 –107 5 May A, Ell C. Palliative treatment of malignant esophagorespiratory fistulas with Gianturco-Z stents: a prospective clinical trial and review of the literature on covered metal stents. Am J Gastroenterol 1998; 93:532–535 6 Ell C, May A. Self-expanding metal stents for palliation of stenosing tumors of the esophagus and cardia: a critical review. Endoscopy 1997; 29:392–398 7 Acosta JL, Battesby JS. Congenital tracheoesophageal fistula in the adult. Ann Thorac Surg 1974; 34:1751–1756 8 Azoulay D, Regnard JF, Magdeleinat P, et al. Congenital respiratory-oesophageal fistula in the adult: report on none cases and review of the literature. J Thorac Cardiovasc Surg 1992; 104:381–384 9 Ogunmola N, Wyllie R, McDowell K, et al. Endoscopic closure of oesophagobronchial fistula with fibrin glue. J Pediatr Gastroenterol Nutr 2004; 38:539 –541 10 Thanopoulos BD, Tsaousis GS, Konstadopoulou GN, et al. Transcatheter closure of muscular ventricular septal defects with the Amplatzer ventricular septal defect occluder: initial clinical applications in children. J Am Coll Cardiol 1999; 33:1395–1399 11 Parsi A, Bruch L, Szurawitzki G, et al. Transcatheter closure of muscular ventricular septal defects in two patients after myocardial infarction. J Interv Cardiol 2001; 14:219 –221
Management of Obstructing Pulmonary Broncholithiasis With Three-Dimensional Imaging and Holmium Laser Lithotripsy* J. Scott Ferguson, MD, FCCP; Jonathan M. Rippentrop, MD; Bernard Fallon, MD; Alan F. Ross, MD; and Geoffrey McLennan, MD, PhD www.chestjournal.org
Major airway obstruction due to broncholithiasis produces significant morbidity, and management is difficult. Many of the patients are elderly and are not good candidates for surgical removal. Bronchoscopic removal may be limited due to anatomic considerations, skill of the bronchoscopist, and exposure of the patient to additional procedural risks. Preprocedural planning with three-dimensional (3D) multidetector CT (MDCT) imaging enhances the bronchoscopist’s knowledge of the relationships of the target lesions with critical structures, and improves the efficiency of the application of specific endobronchial therapies. Here we report our experience treating obstructing broncholithiasis in two patients utilizing pretreatment planning with 3D MDCT imaging, followed by bronchoscopically delivered holmium laser fragmentation of the stones. (CHEST 2006; 130:909 –912) Key words: airway obstruction; bronchoscopy; lasers; lithiasis; lithotripsy; lymph nodes Abbreviations: 2D ⫽ two dimensional; 3D ⫽ three dimensional; Ho-YAG ⫽ holmium-yttrium aluminum garnet; MDCT ⫽ multidetector CT
airway obstruction due to broncholithiasis, alM ajor though uncommon, produces significant morbidity
and is difficult to treat using the surgical and bronchoscopic methods available. Removal of broncholiths via bronchoscopy is often extremely difficult due to embedding of the stone into surrounding structures, large size, or difficulty with crushing the stone using bronchoscopic forceps.1 Surgical interventions are often required; however, many patients are poor surgical candidates, and removal of the stone may require lobectomy or pneumonectomy.2 For these reasons, we have sought alternatives to traditional surgical and bronchoscopic approaches. The Nd-YAG and holmium-yttrium aluminum garnet (Ho-YAG) lasers are possible alternatives to surgery and mechanical crushing.3,4 However, these modalities are associated with rare but important complications,5 some of which may be lessened by the use of advanced imaging *From the Departments of Internal Medicine (Drs. Ferguson and McLennan), Urology (Drs. Rippentrop and Fallon), and Anesthesia (Dr. Ross), University of Iowa, Iowa City, IA. Drs. Ferguson and Rippentrop participated equally in the preparation of this article. This work was performed at the University of Iowa. Drs. Ferguson, Rippentrop, Fallon, and Ross have no conflicts of interest to declare. Dr. McLennan is part owner of VIDA Diagnostics (Iowa City, IA), which is a company that develops three-dimensional software imaging solutions. Manuscript received October 11, 2005; revision accepted February 21, 2006. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: J. Scott Ferguson, MD, FCCP, Department of Internal Medicine, Division of Pulmonary, Critical Care, and Occupational Medicine, 200 Hawkins Dr, C-33 GH, Iowa City, IA 52242-1089; e-mail: [email protected]. DOI: 10.1378/chest.130.3.909 CHEST / 130 / 3 / SEPTEMBER, 2006
909