Surg Endosc DOI 10.1007/s00464-015-4228-4
and Other Interventional Techniques
DYNAMIC MANUSCRIPT
Transtracheal thoracic natural orifice transluminal endoscopic surgery (NOTES) in a swine model Mohamed Khereba1 • Vicky Thiffault1 • Eric Goudie1 • Mehdi Tahiri1 • Rachid Hadjeres2 • Maryam Razmpoosh2 • Pasquale Ferraro1 • Moishe Liberman1,3
Received: 9 November 2014 / Accepted: 1 May 2015 Ó Springer Science+Business Media New York 2015
Abstract Background Natural orifice transluminal endoscopic surgery (NOTES) has the potential to be the final frontier in minimally invasive procedures in thoracic surgery. In order for thoracic pleural NOTES to 1 day be ready for clinical trials, each step of the procedure must be independently evaluated for both safety and efficacy. The aim of this study was to evaluate the trachea as a portal of entry for thoracic NOTES. Methods Eight 40-kg swine underwent right thoracic pleuroscopy in a survival model. In order to avoid inadvertent injury to the superior vena cava, endobronchial ultrasound was employed to select the location of airway incision. A 7-mm linear incision was then performed at the chosen location using an endoscopic electrocautery needle knife through a therapeutic flexible videobronchoscope.
Presented at the Canadian Association of Thoracic Surgery Annual Meeting, Ottawa, Ontario, September 20, 2013.
Electronic supplementary material The online version of this article (doi:10.1007/s00464-015-4228-4) contains supplementary material, which is available to authorized users. & Moishe Liberman [email protected]
The mediastinal fat and parietal pleura were then dissected with electrocautery, and complete right pleuroscopy was performed. The tracheal and mediastinal portal of entry were then sealed with 1–2 cc of fibrin sealant. The pigs were kept alive for 21 days postoperatively. Postmortem diagnostic bronchoscopy was performed to assess tracheal healing. All tracheal specimens underwent histologic examination for healing and signs of mediastinal infection. Results Thoracic NOTES procedures on all eight pigs were successful. There were no intraoperative complications except for one minor bleeding episode within the mediastinal dissection site which stopped spontaneously. Two pigs died from severe laryngospasm in the early postoperative period. Six pigs survived for 21 days postprocedure and experienced uneventful postoperative courses. Postmortem examination demonstrated complete tracheal healing with appropriate scarring in all pigs. Conclusions The trachea appears to be a safe port of entry for thoracic NOTES procedures in a swine model. Smaller tracheal incisions followed by balloon dilatation are associated with less postoperative morbidity and mortality. Tracheal incisions sealed with fibrin sealant healed rapidly and without signs of mediastinal infection. This procedure represents a work in progress and is not yet ready for human trials.
1
Division of Thoracic Surgery, Department of Surgery, CHUM Endoscopic Tracheobronchial and Oesophageal Center (CETOC), University of Montre´al, Montreal, QC, Canada
Keywords Pulmonary (lungs) Ultrasonography Endoscopy Thoracoscopy
2
Department of Pathology, Centre Hospitalier de l’Universite´ de Montre´al, University of Montre´al, Montreal, QC, Canada
3
Division of Thoracic Surgery, Centre Hospitalier de l’Universite´ de Montre´al, 1560 rue Sherbrooke Est, 8e CD Pavillon Lachapelle, bureau D-8051, Montreal, QC H2L 4M1, Canada
In recent years, peritoneal NOTES has been studied in various animal and human models in order to perform various intraperitoneal and retroperitoneal procedures. Few animal studies and no human studies have investigated NOTES for thoracic pleural procedures [1, 2].
123
Surg Endosc
The esophageal route has been studied by some surgical research groups conducting animal studies evaluating thoracic NOTES procedures [3–5]; however, few studies have investigated the trachea as a portal of entry into the mediastinum and thoracic cavity in animal models [2, 6]. The trachea has been shown to be a reliable and relatively safe port of entry for pleural NOTES [2, 7]. However, no human studies evaluating thoracic pleural-based NOTES have been reported. Various methods of tracheal closure during transtracheal NOTES have been investigated by the same group in four sequential studies [2, 6, 8, 9]. Closure of the tracheal wound was performed either through stent insertion (Dumon; Novatech, Grasse, France), BioGlue, fibrin sealant (Tissucol; BAXTER AG, Baxter SL, Valencia, Spain), or combining glue with stenting. Although tracheal healing was observed on postmortem pathological analysis, air leaks constituted significant complications leading to increased mortality in most of the studies. Furthermore, death occurring due to major vascular injury was a problem in some animals upon entering the mediastinum blindly. In order to address these two important issues, we designed a study which: (1) uses endobronchial ultrasound (EBUS) to localize the safest site for tracheal incision into the thoracic cavity and (2) assesses the effect of the minimization of tracheal trauma with a technique of port site stretching (balloon dilation) in order to decrease tracheal fistulous complications.
subjects underwent transtracheal NOTES under general anesthesia. The primary outcome was to evaluate the feasibility of transtracheal natural orifice transluminal thoracic pleuroscopy. The secondary outcomes included: (1) a comparison between two techniques of tracheal incision (with or without balloon dilatation), (2) the evaluation of the safety of the tracheal port selection using EBUS, and (3) determination of short-term (3 weeks) tracheal healing with portal sealing using a fibrin sealant.
Anesthesia and analgesia Eight swine underwent thoracic NOTES procedures. Antibiotic prophylaxis was administered using enrofloxacin intramuscularly 1 h before surgery. Pre-anesthesia was performed with the administration of ketamine (25 mg/kg), atropine (0.04 mg/kg), and azaperone (4 mg/kg). The animal was placed in a dorsal position, and induction of anesthesia was performed through administration of propofol (1.66 mg/kg). Anesthesia was maintained with isoflurane inhalation 1–2.5 %. Analgesia was maintained during the procedure and postoperatively using a Fentanyl patch (100 lg) with intramuscular morphine injections and/or oral meloxicam if the animal demonstrated any signs of pain (Table 1).
Surgical technique Methods and techniques This study consists of an interventional prospective animal survival study with short-term follow-up. The study was approved by the institutional committee for animal protection at the Centre de Recherche du Centre Hospitalier de l’Universite´ de Montre´al (CRCHUM). Eight live porcine
Endotracheal intubation of the animal was performed with an 8-mm endotracheal tube. This tube (working tube) was used as a port for entrance of the flexible bronchoscope and the EBUS scope. A flexible bronchoscope was then introduced through the endotracheal tube. Diagnostic bronchoscopy was performed to delineate the tracheobronchial anatomy, determining the presence of left or right tracheal upper lobe
Table 1 Procedural data and outcome data by subject WT (kg)
Procedure time (min)
Incision size (mm)
1
41
141
3
6
Yes
No
None
POD 4
Complete healing
2
41
95
6
9
No
No
None
POD 3
Complete healing
3
39
84
5
8
No
No
None
POD 3
Complete healing
4
40.5
58
5
8
No
No
None
POD 3
Autolyzed specimen
5
37
44
6
–
No
No
Laryngospasm
Died
Minimal inflammatory reaction
6
40
53
7
–
No
No
Fever
POD 3
Complete healing
7
41
72
7
–
No
No
None
POD 3
Complete healing
8
30
62
10
–
No
No
Severe distress
Died
Minimal inflammatory reaction
WT weight, POD postoperative day
123
Balloon dilatation diameter (mm)
Immediate air leak
Delayed air leak
Complications
Pigtail removal
Histological analysis
Surg Endosc
bronchus. Selective left main bronchial intubation was performed using a 5.5-mm endotracheal tube (ventilating tube) under bronchoscopic guidance. Due to the fact that the swine trachea is much longer than the human trachea, an extension of the 5.5 endotracheal tube was achieved using a second 7.5mm endotracheal tube. Through this technique, left single lung ventilation was achieved. Curvilinear EBUS (Olympus EBUS Puncture Scope, BF-UC180F-A, Olympus USA) was then utilized to select an appropriate site for puncture and creation of a port of entry into the right pleural cavity transtracheally. EBUS color flow Doppler ultrasound was utilized to select a safe incision site away from any vascular structures. The chosen incision site was then marked utilizing EBUS-guided fine needle puncture (Olympus ViziShot, NA-201SX-4021, Olympus USA) under echographic visualization. The EBUS scope was then removed and a flexible therapeutic bronchoscope was passed through the working tube, and an electrocautery needle knife (Olympus NeedleCut3, KD-V441M, Olympus USA) was used to make a tracheal incision at the pre-marked point (EBUS marked). The animals were divided into two groups of four. In Group A, a tracheal incision of 3–6 mm was performed followed by balloon dilatation utilizing a endoscopic pulmonary balloon dilatation catheter (CRETM Pulmonary Balloon Dilator, Boston Scientific, Natick, MA, USA) in order to achieve an incision of 6–9 mm (VIDEO 1). In Group B, a larger incision (8–9 mm) was performed using the needle knife without balloon dilatation. The mediastinal fat and parietal pleura were then dissected with electrocautery and hydro-dissection in order to achieve access to the pleural cavity. A flexible therapeutic bronchoscope was introduced through the tracheal incision into the pleural cavity. Complete diagnostic pleuroscopy was performed. Identification of all major pleural and mediastinal organs was performed (VIDEO 2). A 12 Fr. Pigtail intercostal catheter was inserted under endoscopic vision during pleuroscopy to prevent pneumatic events in the perioperative period. The bronchoscope was then removed, and the tracheal incision was sealed using bronchoscopic injection of 1–2 cc of fibrin sealant (TISSEEL, Baxter Healthcare, Illinois, USA). The fibrin sealant was first injected in the depth of the incision into the mediastinal tissue while gradually retracting the applicator to the surface of the incision to finally form a glue cap over the incision. The ventilation tube was removed, and twolung ventilation was achieved through the 8-mm working tube. The presence of immediate air leak was evaluated. The animal was then weaned from general anesthesia, and extubation was performed. In animals 6, 7, and 8, the infiltration of 2 % lidocaine around and through the pigtail was performed in order to decrease postoperative pain related to the drain and the pleuroscopy.
The animals were kept alive for 21 days post-procedure. Vital signs, oxygen saturation, the presence or absence of air leak, and air entry were evaluated every day for the first three postoperative days. The intercostal pigtail drains were removed on the third postoperative day. Any postoperative change in behavior or signs of pain or distress was monitored, evaluated, and recorded. The animal weight was recorded in order to evaluate weight gain postoperatively as an indicator of general health.
Necropsy Necropsy was performed at 21 days following the procedure. Intramuscular injection of xylazine (2 mg/kg) and ketamine (25 mg/kg) followed by intravenous injection of phenobarbital sodium (108 mg/kg) were administered. Postmortem diagnostic bronchoscopy was performed to evaluate the gross tracheal healing pattern. Access to the thoracic cavity was achieved via median sternotomy. Pleural cavity, mediastinum, lung parenchyma, pericardium/heart as well as the presence of gross signs of infection were evaluated during necropsy. After necropsy, the trachea with the surrounding mediastinal tissues and lung underwent pathological analysis to evaluate tracheal healing and any signs of infection or the presence microabscesses.
Results Eight female Yorkshire swine underwent right thoracic pleuroscopy in a survival model. The weight of the animals at the time of procedure ranged from 30 to 41 kg. The time necessary to perform the procedure from intubation to extubation ranged from 44 to 141 min with a mean of 76 min. All sites of tracheal incision localized by EBUS examinations were performed safely without any vascular injuries. Vessels identified by EBUS included the azygous vein, superior vena cava (SVC), and the subclavian vessels. The safest site for right lateral tracheal incision in order to have access to the right pleural cavity was approximately 1 cm above the tracheal right upper lobe bronchus. The mean time from intubation to having full access to pleural cavity was 45.6 min (range 20–80 min). Hydro-dissection of the mediastinal tissue (fat, SVC, azygous vein, parietal pleura and lymph nodes) was effective and safe in all animals with no vascular injury. There was one minor venous bleeding episode during peritracheal mediastinal fat dissection. The bleeding was controlled with endoscopic pressure using the tip of the flexible bronchoscope and it stopped spontaneously. Multiple para-tracheal lymph nodes were identified during mediastinal dissection.
123
Surg Endosc
Fig. 1 NOTES pleuroscopy demonstrating intrathoracic organs
During pleuroscopy (Fig. 1), all right-sided pulmonary lobes, heart, diaphragm, mammary vessels, innominate vein, thymus gland, and phrenic nerve were identified and easily accessed with the flexible videobronchoscope. Time of pleuroscopy ranged from 4 to 25 min with a mean of 10.5 min. Tracheal incision closure was successfully achieved in all animals with the injection of 1–2 cc of fibrin sealant into the tracheal–mediastinal portal from the bottom of the tract to the tracheal luminal surface. There was only one immediate leak after incision closure. This leak stopped spontaneously after 4 min. Six animals survived for 21 days post-procedure. Two animals in the second group (no balloon dilatation) died on the first postoperative day. All of the surviving animals appropriately gained weight until the time of necropsy. Five animals showed no abnormal behavior with no coughing or any signs of distress for 21 days postoperatively. One animal in Group B showed signs of pain with fever on the fourth postoperative day and was treated with analgesia and amoxicillin/clavunate for 5 days. This animal returned to baseline following 2 days of treatment. During the first three postoperative days, all surviving animals had normal vital signs and normal oxygen saturations without any air leak from the pleural drain. Physical examinations revealed equal air entry with normal breath sound auscultation.
Postoperative mortality In the immediate postoperative period, in the fifth animal, there was excessive anxiety with seizure-like movements and hyperthermia (41.5 °C). Several attempts were made to
123
relax the animal with morphine, ketamine, midazolam, and lidocaine injection into the pleural pigtail. These maneuvers did not improve the clinical status of the animal. Oxygen saturation was maintained above 95 % with equal bilateral air entry and normal lung auscultation. Urgent retransfer back to the operating room was performed in order to control body temperature and anxiety under general anesthesia. Physical examination showed normal and equal air entry with a functioning pleural drain without air leak and 99 % oxygen saturation. An attempt at re-intubation in order to control aggressiveness and hyperthermia was performed. During re-intubation, the animal went into severe laryngospasm, was not intubatable, and died. The eighth animal showed severe aggressiveness and distress in the early postoperative period. Physical examination revealed no abnormalities with oxygen saturations above 95 %. Several attempts were made to control the distress with injection of ketamine and morphine, with no improvement. Finally, euthanasia was performed in this animal in order to relieve distress. Necropsy All animals underwent necropsy after either euthanasia (21 days) or death. Postmortem bronchoscopy revealed appropriate healing of the tracheal incisions in all surviving animals (Fig. 2). Necropsy showed postoperative flimsy adhesions within the ipsilateral pleural cavity. No signs of infection or abscess were identified. Postmortem bronchoscopy in the fifth and eighth animals (early mortality) showed intact fibrin sealant closure of tracheal incisions without any tracheal fistula (Fig. 3). Necropsy in both animals that died early postoperatively did not show any signs of lung or any other organ injuries. Necropsy revealed severe laryngospasm and laryngeal edema in the fifth animal.
Histopathology Histopathological analysis of the tracheal incision site was performed in seven animals using hematoxylin and eosin stains. One specimen was found to be completely autolyzed as a result of inadequate formalin fixation, and pathological microscopic analysis was therefore not possible. The two specimens of the non-surviving animals showed only signs of acute inflammation without healing. The other five specimens (surviving animals) showed appropriate tracheal healing with adequate formation of scar tissue and complete epithelialization (Fig. 4). No signs of micro-abscesses were found. There were no traces of fibrin sealant found at necropsy on gross or histological assessment of the trachea.
Surg Endosc Fig. 2 Tracheal incision during the procedure (left) and postmortem bronchoscopic view of the same incision completely healed (right)
Fig. 3 Postmortem bronchoscopic view of the eighth animal (early postoperative mortality) showing intact tracheal incision sealed with the fibrin sealant
Discussion With advances in flexible endoscopic technology, the idea of an almost noninvasive technology has emerged. Natural orifice transluminal endoscopic surgery (NOTES) is the ultimate step in the spectrum of minimally invasive surgery (MIS). Thoracic NOTES is still in its infancy, and few groups have evaluated thoracic NOTES in animal models [2–9]. Thoracic NOTES may 1 day allow for pericardial window, pleuroscopy with biopsy and pleurodesis, mediastinal and hilar lymph node dissection, resection of mediastinal cysts, and tumors as well as pulmonary resection. A transtracheal approach to thoracic NOTES may be able to offer a simple and safe entry point into the thoracic cavity. The transtracheal approach may be advantageous compared to the transesophageal approach due to decreased potential risk of post-procedural esophageal fistula and mediastinitis. The lack of food passage against the
Fig. 4 Complete fibrous healing of tracheal incision on histological analysis
repair, as compared to transesophageal NOTES, may decrease mediastinitis rates. Also, the trachea, being a more anterior structure, may allow for easier and more direct access to the pulmonary hilum, pericardium, and anterior mediastinal tumors. Fatal respiratory distress due to development of tension pneumothorax has been reported during transtracheal thoracic NOTES procedures [6, 9]. Selective single lung ventilation under bronchoscopic guidance prevented presence of air pressure in the trachea during the procedure eliminating the risk of air under tension within the pleural space. Tracheal isolation permitted the utilization of a conventional needle knife with cautery for tracheal incision without risk of airway fire. Major vascular structures surrounding the trachea may be hazardous during tracheal incision. Liu et al. [6] reported one vascular complication upon tracheal incision, and this was associated with massive hemorrhage and death. In order to circumvent this risk, we used EBUS to localize major mediastinal vessels which allowed for
123
Surg Endosc
appropriate and safe selection of the site of tracheal incision. This resulted in safe tracheal incision without any related morbidity in all cases. Tracheal incisions for thoracic NOTES have been mostly performed using homemade metallic knives [2, 6]. In order to perform the smallest possible incision, with the least airway trauma, we chose to use flexible endoscopic equipment to both incise the trachea and perform pleural and mediastinal exploration. Incisions performed with needle electrocautery knives were easily achieved without complication. Dissection into the mediastinal tissues with hydro-dissection and electrocautery was performed safely without significant tissue damage which could hinder tracheal healing. Histopathological analysis of incisions performed with the needle knife showed no signs of necrosis. In order to evaluate the best flexible endoscopic method for tracheal incision for transtracheal NOTES, we evaluated two techniques; a short tracheal incision with subsequent balloon dilatation (first group) and a longer tracheal incision without dilatation (second group). With no mortality and morbidity in the first group, it was obvious that smaller incisions followed by balloon dilatations were much safer, due to less direct tissue damage to the tracheal wall. This may be related to less cautery injury and smoke within the trachea and larynx associated with the first technique. The trachea is not a sterile organ, and therefore, perioperative antibiotics are typically administered in tracheal surgery. We administered a single prophylactic dose of antibiotics in all animals. There was only one animal that presented with low-grade fever on the fourth postoperative day and rapidly responded to oral antibiotics and analgesics. Blood cultures drawn during the febrile episode (before initiation of antibiotics) were negative. Necropsy and pathological analysis of lung and mediastinal tissue did not indicate the presence of mediastinal abscess. Tracheal stents and biological glues have been utilized by others to seal tracheal incisions during thoracic NOTES [2, 6, 8, 9]. Both methods have been proven to be effective; however, utilization of biological glue has been associated with less postoperative complications. Stent migration can be very dangerous as a distally migrating stent may occlude the airway [6]. In esophageal NOTES, the insertion of stents has been associated with delayed and impaired mucosal healing [5]. Tracheal incisions sealed with fibrin sealant healed rapidly without interference with normal tissue healing. In conclusion, the trachea appears to be a safe portal of entry for thoracic NOTES procedures in a swine model.
123
This approach offers a reliable port of entry to the pleural cavity. Utilization of EBUS for localization of major vascular structures in relation to the trachea before tracheal incision renders the transtracheal approach safe and prevents life-threatening vascular injuries. Tracheal incisions sealed with fibrin sealants heal rapidly with no signs of mediastinal infection. This procedure represents a work in progress and is not yet ready for human trials. Acknowledgments Research funded by the Canadian Foundation for Innovation and the Marcel and Rolande Gosselin Chair in Thoracic Surgical Oncology. In-kind support provided by Olympus Canada and Baxter Pharmaceuticals. Disclosures Dr. Mohamed Khereba MD, Ms. Vicky Thiffault RN, Dr. Eric Goudie, Dr. Mehdi Tahiri MD, Dr. Rachid Hadjeres MD, Dr. Maryam Razmpoosh MD, Dr. Pasquale Ferraro MD, and Dr. Moishe Liberman MD PhD have no conflict of interest or financial ties to disclose.
References 1. Trunzo J, Cavazzola L, Elmunzer B, Poulose B, McGee M, Marks J et al (2009) Facilitating gastrotomy closure during natural-orifice transluminal endoscopic surgery using tissue anchors. Endoscopy 41:487–492 2. Chen W, Chu Y, Wu Y, Liu C, Yuan H, Liu Y et al (2012) Endoscopic closure of a tracheal access site using bioglue after transtracheal thoracoscopy in a nonsurvival canine model. Eur Surg Res 48:26–33 3. Willingham F, Gee D, Lauwers G, Brugge W, Rattner D (2008) Natural orifice transesophageal mediastinoscopy and thoracoscopy. Surg Endosc 22:1042–1047 4. Gee D, Willingham F, Lauwers G, Brugge W, Rattner D (2008) Natural orifice transesophageal mediastinoscopy and thoracoscopy: a survival series in swine. Surg Endosc 22:2117–2122 5. Turner B, Kim M, Gee D, Dursun A, Mino-Kenudson M, Brugge W (2011) Prospective randomized trial of esophageal submucosal tunnel closure with a stent versus no closure to secure a transesophageal natural orifice transluminal endoscopic surgery access site. Gastrointest Endosc 73:785–790 6. Liu Y, Chu Y, Liu C, Liu HP, Wu Y, Liu H et al (2011) Feasibility of the transtracheal approach for the thoracic cavity in a large animal model. Surg Endosc 25:1652–1658 7. Lima E, Henriques-Coelho T, Rolanda C, Peˆgo JM, Silva D, Correia-Pinto J et al (2007) Transvesical thoracoscopy: a natural orifice translumenal endoscopic approach for thoracic surgery. Surg Endosc 21:854–858 8. Liu Y, Liu H, Wu Y, Ko P (2010) Feasibility of transtracheal surgical lung biopsy in a canine animal model. Eur J Cardiothorac Surg 37:1235–1236 9. Liu Y, Liu H, Wu Y, Ko P (2010) Feasibility of transtracheal thoracoscopy (natural orifice transluminal endoscopic surgery). J Thorac Cardiovasc Surg 139:1349–1350
and Other Interventional Techniques
DYNAMIC MANUSCRIPT
Transtracheal thoracic natural orifice transluminal endoscopic surgery (NOTES) in a swine model Mohamed Khereba1 • Vicky Thiffault1 • Eric Goudie1 • Mehdi Tahiri1 • Rachid Hadjeres2 • Maryam Razmpoosh2 • Pasquale Ferraro1 • Moishe Liberman1,3
Received: 9 November 2014 / Accepted: 1 May 2015 Ó Springer Science+Business Media New York 2015
Abstract Background Natural orifice transluminal endoscopic surgery (NOTES) has the potential to be the final frontier in minimally invasive procedures in thoracic surgery. In order for thoracic pleural NOTES to 1 day be ready for clinical trials, each step of the procedure must be independently evaluated for both safety and efficacy. The aim of this study was to evaluate the trachea as a portal of entry for thoracic NOTES. Methods Eight 40-kg swine underwent right thoracic pleuroscopy in a survival model. In order to avoid inadvertent injury to the superior vena cava, endobronchial ultrasound was employed to select the location of airway incision. A 7-mm linear incision was then performed at the chosen location using an endoscopic electrocautery needle knife through a therapeutic flexible videobronchoscope.
Presented at the Canadian Association of Thoracic Surgery Annual Meeting, Ottawa, Ontario, September 20, 2013.
Electronic supplementary material The online version of this article (doi:10.1007/s00464-015-4228-4) contains supplementary material, which is available to authorized users. & Moishe Liberman [email protected]
The mediastinal fat and parietal pleura were then dissected with electrocautery, and complete right pleuroscopy was performed. The tracheal and mediastinal portal of entry were then sealed with 1–2 cc of fibrin sealant. The pigs were kept alive for 21 days postoperatively. Postmortem diagnostic bronchoscopy was performed to assess tracheal healing. All tracheal specimens underwent histologic examination for healing and signs of mediastinal infection. Results Thoracic NOTES procedures on all eight pigs were successful. There were no intraoperative complications except for one minor bleeding episode within the mediastinal dissection site which stopped spontaneously. Two pigs died from severe laryngospasm in the early postoperative period. Six pigs survived for 21 days postprocedure and experienced uneventful postoperative courses. Postmortem examination demonstrated complete tracheal healing with appropriate scarring in all pigs. Conclusions The trachea appears to be a safe port of entry for thoracic NOTES procedures in a swine model. Smaller tracheal incisions followed by balloon dilatation are associated with less postoperative morbidity and mortality. Tracheal incisions sealed with fibrin sealant healed rapidly and without signs of mediastinal infection. This procedure represents a work in progress and is not yet ready for human trials.
1
Division of Thoracic Surgery, Department of Surgery, CHUM Endoscopic Tracheobronchial and Oesophageal Center (CETOC), University of Montre´al, Montreal, QC, Canada
Keywords Pulmonary (lungs) Ultrasonography Endoscopy Thoracoscopy
2
Department of Pathology, Centre Hospitalier de l’Universite´ de Montre´al, University of Montre´al, Montreal, QC, Canada
3
Division of Thoracic Surgery, Centre Hospitalier de l’Universite´ de Montre´al, 1560 rue Sherbrooke Est, 8e CD Pavillon Lachapelle, bureau D-8051, Montreal, QC H2L 4M1, Canada
In recent years, peritoneal NOTES has been studied in various animal and human models in order to perform various intraperitoneal and retroperitoneal procedures. Few animal studies and no human studies have investigated NOTES for thoracic pleural procedures [1, 2].
123
Surg Endosc
The esophageal route has been studied by some surgical research groups conducting animal studies evaluating thoracic NOTES procedures [3–5]; however, few studies have investigated the trachea as a portal of entry into the mediastinum and thoracic cavity in animal models [2, 6]. The trachea has been shown to be a reliable and relatively safe port of entry for pleural NOTES [2, 7]. However, no human studies evaluating thoracic pleural-based NOTES have been reported. Various methods of tracheal closure during transtracheal NOTES have been investigated by the same group in four sequential studies [2, 6, 8, 9]. Closure of the tracheal wound was performed either through stent insertion (Dumon; Novatech, Grasse, France), BioGlue, fibrin sealant (Tissucol; BAXTER AG, Baxter SL, Valencia, Spain), or combining glue with stenting. Although tracheal healing was observed on postmortem pathological analysis, air leaks constituted significant complications leading to increased mortality in most of the studies. Furthermore, death occurring due to major vascular injury was a problem in some animals upon entering the mediastinum blindly. In order to address these two important issues, we designed a study which: (1) uses endobronchial ultrasound (EBUS) to localize the safest site for tracheal incision into the thoracic cavity and (2) assesses the effect of the minimization of tracheal trauma with a technique of port site stretching (balloon dilation) in order to decrease tracheal fistulous complications.
subjects underwent transtracheal NOTES under general anesthesia. The primary outcome was to evaluate the feasibility of transtracheal natural orifice transluminal thoracic pleuroscopy. The secondary outcomes included: (1) a comparison between two techniques of tracheal incision (with or without balloon dilatation), (2) the evaluation of the safety of the tracheal port selection using EBUS, and (3) determination of short-term (3 weeks) tracheal healing with portal sealing using a fibrin sealant.
Anesthesia and analgesia Eight swine underwent thoracic NOTES procedures. Antibiotic prophylaxis was administered using enrofloxacin intramuscularly 1 h before surgery. Pre-anesthesia was performed with the administration of ketamine (25 mg/kg), atropine (0.04 mg/kg), and azaperone (4 mg/kg). The animal was placed in a dorsal position, and induction of anesthesia was performed through administration of propofol (1.66 mg/kg). Anesthesia was maintained with isoflurane inhalation 1–2.5 %. Analgesia was maintained during the procedure and postoperatively using a Fentanyl patch (100 lg) with intramuscular morphine injections and/or oral meloxicam if the animal demonstrated any signs of pain (Table 1).
Surgical technique Methods and techniques This study consists of an interventional prospective animal survival study with short-term follow-up. The study was approved by the institutional committee for animal protection at the Centre de Recherche du Centre Hospitalier de l’Universite´ de Montre´al (CRCHUM). Eight live porcine
Endotracheal intubation of the animal was performed with an 8-mm endotracheal tube. This tube (working tube) was used as a port for entrance of the flexible bronchoscope and the EBUS scope. A flexible bronchoscope was then introduced through the endotracheal tube. Diagnostic bronchoscopy was performed to delineate the tracheobronchial anatomy, determining the presence of left or right tracheal upper lobe
Table 1 Procedural data and outcome data by subject WT (kg)
Procedure time (min)
Incision size (mm)
1
41
141
3
6
Yes
No
None
POD 4
Complete healing
2
41
95
6
9
No
No
None
POD 3
Complete healing
3
39
84
5
8
No
No
None
POD 3
Complete healing
4
40.5
58
5
8
No
No
None
POD 3
Autolyzed specimen
5
37
44
6
–
No
No
Laryngospasm
Died
Minimal inflammatory reaction
6
40
53
7
–
No
No
Fever
POD 3
Complete healing
7
41
72
7
–
No
No
None
POD 3
Complete healing
8
30
62
10
–
No
No
Severe distress
Died
Minimal inflammatory reaction
WT weight, POD postoperative day
123
Balloon dilatation diameter (mm)
Immediate air leak
Delayed air leak
Complications
Pigtail removal
Histological analysis
Surg Endosc
bronchus. Selective left main bronchial intubation was performed using a 5.5-mm endotracheal tube (ventilating tube) under bronchoscopic guidance. Due to the fact that the swine trachea is much longer than the human trachea, an extension of the 5.5 endotracheal tube was achieved using a second 7.5mm endotracheal tube. Through this technique, left single lung ventilation was achieved. Curvilinear EBUS (Olympus EBUS Puncture Scope, BF-UC180F-A, Olympus USA) was then utilized to select an appropriate site for puncture and creation of a port of entry into the right pleural cavity transtracheally. EBUS color flow Doppler ultrasound was utilized to select a safe incision site away from any vascular structures. The chosen incision site was then marked utilizing EBUS-guided fine needle puncture (Olympus ViziShot, NA-201SX-4021, Olympus USA) under echographic visualization. The EBUS scope was then removed and a flexible therapeutic bronchoscope was passed through the working tube, and an electrocautery needle knife (Olympus NeedleCut3, KD-V441M, Olympus USA) was used to make a tracheal incision at the pre-marked point (EBUS marked). The animals were divided into two groups of four. In Group A, a tracheal incision of 3–6 mm was performed followed by balloon dilatation utilizing a endoscopic pulmonary balloon dilatation catheter (CRETM Pulmonary Balloon Dilator, Boston Scientific, Natick, MA, USA) in order to achieve an incision of 6–9 mm (VIDEO 1). In Group B, a larger incision (8–9 mm) was performed using the needle knife without balloon dilatation. The mediastinal fat and parietal pleura were then dissected with electrocautery and hydro-dissection in order to achieve access to the pleural cavity. A flexible therapeutic bronchoscope was introduced through the tracheal incision into the pleural cavity. Complete diagnostic pleuroscopy was performed. Identification of all major pleural and mediastinal organs was performed (VIDEO 2). A 12 Fr. Pigtail intercostal catheter was inserted under endoscopic vision during pleuroscopy to prevent pneumatic events in the perioperative period. The bronchoscope was then removed, and the tracheal incision was sealed using bronchoscopic injection of 1–2 cc of fibrin sealant (TISSEEL, Baxter Healthcare, Illinois, USA). The fibrin sealant was first injected in the depth of the incision into the mediastinal tissue while gradually retracting the applicator to the surface of the incision to finally form a glue cap over the incision. The ventilation tube was removed, and twolung ventilation was achieved through the 8-mm working tube. The presence of immediate air leak was evaluated. The animal was then weaned from general anesthesia, and extubation was performed. In animals 6, 7, and 8, the infiltration of 2 % lidocaine around and through the pigtail was performed in order to decrease postoperative pain related to the drain and the pleuroscopy.
The animals were kept alive for 21 days post-procedure. Vital signs, oxygen saturation, the presence or absence of air leak, and air entry were evaluated every day for the first three postoperative days. The intercostal pigtail drains were removed on the third postoperative day. Any postoperative change in behavior or signs of pain or distress was monitored, evaluated, and recorded. The animal weight was recorded in order to evaluate weight gain postoperatively as an indicator of general health.
Necropsy Necropsy was performed at 21 days following the procedure. Intramuscular injection of xylazine (2 mg/kg) and ketamine (25 mg/kg) followed by intravenous injection of phenobarbital sodium (108 mg/kg) were administered. Postmortem diagnostic bronchoscopy was performed to evaluate the gross tracheal healing pattern. Access to the thoracic cavity was achieved via median sternotomy. Pleural cavity, mediastinum, lung parenchyma, pericardium/heart as well as the presence of gross signs of infection were evaluated during necropsy. After necropsy, the trachea with the surrounding mediastinal tissues and lung underwent pathological analysis to evaluate tracheal healing and any signs of infection or the presence microabscesses.
Results Eight female Yorkshire swine underwent right thoracic pleuroscopy in a survival model. The weight of the animals at the time of procedure ranged from 30 to 41 kg. The time necessary to perform the procedure from intubation to extubation ranged from 44 to 141 min with a mean of 76 min. All sites of tracheal incision localized by EBUS examinations were performed safely without any vascular injuries. Vessels identified by EBUS included the azygous vein, superior vena cava (SVC), and the subclavian vessels. The safest site for right lateral tracheal incision in order to have access to the right pleural cavity was approximately 1 cm above the tracheal right upper lobe bronchus. The mean time from intubation to having full access to pleural cavity was 45.6 min (range 20–80 min). Hydro-dissection of the mediastinal tissue (fat, SVC, azygous vein, parietal pleura and lymph nodes) was effective and safe in all animals with no vascular injury. There was one minor venous bleeding episode during peritracheal mediastinal fat dissection. The bleeding was controlled with endoscopic pressure using the tip of the flexible bronchoscope and it stopped spontaneously. Multiple para-tracheal lymph nodes were identified during mediastinal dissection.
123
Surg Endosc
Fig. 1 NOTES pleuroscopy demonstrating intrathoracic organs
During pleuroscopy (Fig. 1), all right-sided pulmonary lobes, heart, diaphragm, mammary vessels, innominate vein, thymus gland, and phrenic nerve were identified and easily accessed with the flexible videobronchoscope. Time of pleuroscopy ranged from 4 to 25 min with a mean of 10.5 min. Tracheal incision closure was successfully achieved in all animals with the injection of 1–2 cc of fibrin sealant into the tracheal–mediastinal portal from the bottom of the tract to the tracheal luminal surface. There was only one immediate leak after incision closure. This leak stopped spontaneously after 4 min. Six animals survived for 21 days post-procedure. Two animals in the second group (no balloon dilatation) died on the first postoperative day. All of the surviving animals appropriately gained weight until the time of necropsy. Five animals showed no abnormal behavior with no coughing or any signs of distress for 21 days postoperatively. One animal in Group B showed signs of pain with fever on the fourth postoperative day and was treated with analgesia and amoxicillin/clavunate for 5 days. This animal returned to baseline following 2 days of treatment. During the first three postoperative days, all surviving animals had normal vital signs and normal oxygen saturations without any air leak from the pleural drain. Physical examinations revealed equal air entry with normal breath sound auscultation.
Postoperative mortality In the immediate postoperative period, in the fifth animal, there was excessive anxiety with seizure-like movements and hyperthermia (41.5 °C). Several attempts were made to
123
relax the animal with morphine, ketamine, midazolam, and lidocaine injection into the pleural pigtail. These maneuvers did not improve the clinical status of the animal. Oxygen saturation was maintained above 95 % with equal bilateral air entry and normal lung auscultation. Urgent retransfer back to the operating room was performed in order to control body temperature and anxiety under general anesthesia. Physical examination showed normal and equal air entry with a functioning pleural drain without air leak and 99 % oxygen saturation. An attempt at re-intubation in order to control aggressiveness and hyperthermia was performed. During re-intubation, the animal went into severe laryngospasm, was not intubatable, and died. The eighth animal showed severe aggressiveness and distress in the early postoperative period. Physical examination revealed no abnormalities with oxygen saturations above 95 %. Several attempts were made to control the distress with injection of ketamine and morphine, with no improvement. Finally, euthanasia was performed in this animal in order to relieve distress. Necropsy All animals underwent necropsy after either euthanasia (21 days) or death. Postmortem bronchoscopy revealed appropriate healing of the tracheal incisions in all surviving animals (Fig. 2). Necropsy showed postoperative flimsy adhesions within the ipsilateral pleural cavity. No signs of infection or abscess were identified. Postmortem bronchoscopy in the fifth and eighth animals (early mortality) showed intact fibrin sealant closure of tracheal incisions without any tracheal fistula (Fig. 3). Necropsy in both animals that died early postoperatively did not show any signs of lung or any other organ injuries. Necropsy revealed severe laryngospasm and laryngeal edema in the fifth animal.
Histopathology Histopathological analysis of the tracheal incision site was performed in seven animals using hematoxylin and eosin stains. One specimen was found to be completely autolyzed as a result of inadequate formalin fixation, and pathological microscopic analysis was therefore not possible. The two specimens of the non-surviving animals showed only signs of acute inflammation without healing. The other five specimens (surviving animals) showed appropriate tracheal healing with adequate formation of scar tissue and complete epithelialization (Fig. 4). No signs of micro-abscesses were found. There were no traces of fibrin sealant found at necropsy on gross or histological assessment of the trachea.
Surg Endosc Fig. 2 Tracheal incision during the procedure (left) and postmortem bronchoscopic view of the same incision completely healed (right)
Fig. 3 Postmortem bronchoscopic view of the eighth animal (early postoperative mortality) showing intact tracheal incision sealed with the fibrin sealant
Discussion With advances in flexible endoscopic technology, the idea of an almost noninvasive technology has emerged. Natural orifice transluminal endoscopic surgery (NOTES) is the ultimate step in the spectrum of minimally invasive surgery (MIS). Thoracic NOTES is still in its infancy, and few groups have evaluated thoracic NOTES in animal models [2–9]. Thoracic NOTES may 1 day allow for pericardial window, pleuroscopy with biopsy and pleurodesis, mediastinal and hilar lymph node dissection, resection of mediastinal cysts, and tumors as well as pulmonary resection. A transtracheal approach to thoracic NOTES may be able to offer a simple and safe entry point into the thoracic cavity. The transtracheal approach may be advantageous compared to the transesophageal approach due to decreased potential risk of post-procedural esophageal fistula and mediastinitis. The lack of food passage against the
Fig. 4 Complete fibrous healing of tracheal incision on histological analysis
repair, as compared to transesophageal NOTES, may decrease mediastinitis rates. Also, the trachea, being a more anterior structure, may allow for easier and more direct access to the pulmonary hilum, pericardium, and anterior mediastinal tumors. Fatal respiratory distress due to development of tension pneumothorax has been reported during transtracheal thoracic NOTES procedures [6, 9]. Selective single lung ventilation under bronchoscopic guidance prevented presence of air pressure in the trachea during the procedure eliminating the risk of air under tension within the pleural space. Tracheal isolation permitted the utilization of a conventional needle knife with cautery for tracheal incision without risk of airway fire. Major vascular structures surrounding the trachea may be hazardous during tracheal incision. Liu et al. [6] reported one vascular complication upon tracheal incision, and this was associated with massive hemorrhage and death. In order to circumvent this risk, we used EBUS to localize major mediastinal vessels which allowed for
123
Surg Endosc
appropriate and safe selection of the site of tracheal incision. This resulted in safe tracheal incision without any related morbidity in all cases. Tracheal incisions for thoracic NOTES have been mostly performed using homemade metallic knives [2, 6]. In order to perform the smallest possible incision, with the least airway trauma, we chose to use flexible endoscopic equipment to both incise the trachea and perform pleural and mediastinal exploration. Incisions performed with needle electrocautery knives were easily achieved without complication. Dissection into the mediastinal tissues with hydro-dissection and electrocautery was performed safely without significant tissue damage which could hinder tracheal healing. Histopathological analysis of incisions performed with the needle knife showed no signs of necrosis. In order to evaluate the best flexible endoscopic method for tracheal incision for transtracheal NOTES, we evaluated two techniques; a short tracheal incision with subsequent balloon dilatation (first group) and a longer tracheal incision without dilatation (second group). With no mortality and morbidity in the first group, it was obvious that smaller incisions followed by balloon dilatations were much safer, due to less direct tissue damage to the tracheal wall. This may be related to less cautery injury and smoke within the trachea and larynx associated with the first technique. The trachea is not a sterile organ, and therefore, perioperative antibiotics are typically administered in tracheal surgery. We administered a single prophylactic dose of antibiotics in all animals. There was only one animal that presented with low-grade fever on the fourth postoperative day and rapidly responded to oral antibiotics and analgesics. Blood cultures drawn during the febrile episode (before initiation of antibiotics) were negative. Necropsy and pathological analysis of lung and mediastinal tissue did not indicate the presence of mediastinal abscess. Tracheal stents and biological glues have been utilized by others to seal tracheal incisions during thoracic NOTES [2, 6, 8, 9]. Both methods have been proven to be effective; however, utilization of biological glue has been associated with less postoperative complications. Stent migration can be very dangerous as a distally migrating stent may occlude the airway [6]. In esophageal NOTES, the insertion of stents has been associated with delayed and impaired mucosal healing [5]. Tracheal incisions sealed with fibrin sealant healed rapidly without interference with normal tissue healing. In conclusion, the trachea appears to be a safe portal of entry for thoracic NOTES procedures in a swine model.
123
This approach offers a reliable port of entry to the pleural cavity. Utilization of EBUS for localization of major vascular structures in relation to the trachea before tracheal incision renders the transtracheal approach safe and prevents life-threatening vascular injuries. Tracheal incisions sealed with fibrin sealants heal rapidly with no signs of mediastinal infection. This procedure represents a work in progress and is not yet ready for human trials. Acknowledgments Research funded by the Canadian Foundation for Innovation and the Marcel and Rolande Gosselin Chair in Thoracic Surgical Oncology. In-kind support provided by Olympus Canada and Baxter Pharmaceuticals. Disclosures Dr. Mohamed Khereba MD, Ms. Vicky Thiffault RN, Dr. Eric Goudie, Dr. Mehdi Tahiri MD, Dr. Rachid Hadjeres MD, Dr. Maryam Razmpoosh MD, Dr. Pasquale Ferraro MD, and Dr. Moishe Liberman MD PhD have no conflict of interest or financial ties to disclose.
References 1. Trunzo J, Cavazzola L, Elmunzer B, Poulose B, McGee M, Marks J et al (2009) Facilitating gastrotomy closure during natural-orifice transluminal endoscopic surgery using tissue anchors. Endoscopy 41:487–492 2. Chen W, Chu Y, Wu Y, Liu C, Yuan H, Liu Y et al (2012) Endoscopic closure of a tracheal access site using bioglue after transtracheal thoracoscopy in a nonsurvival canine model. Eur Surg Res 48:26–33 3. Willingham F, Gee D, Lauwers G, Brugge W, Rattner D (2008) Natural orifice transesophageal mediastinoscopy and thoracoscopy. Surg Endosc 22:1042–1047 4. Gee D, Willingham F, Lauwers G, Brugge W, Rattner D (2008) Natural orifice transesophageal mediastinoscopy and thoracoscopy: a survival series in swine. Surg Endosc 22:2117–2122 5. Turner B, Kim M, Gee D, Dursun A, Mino-Kenudson M, Brugge W (2011) Prospective randomized trial of esophageal submucosal tunnel closure with a stent versus no closure to secure a transesophageal natural orifice transluminal endoscopic surgery access site. Gastrointest Endosc 73:785–790 6. Liu Y, Chu Y, Liu C, Liu HP, Wu Y, Liu H et al (2011) Feasibility of the transtracheal approach for the thoracic cavity in a large animal model. Surg Endosc 25:1652–1658 7. Lima E, Henriques-Coelho T, Rolanda C, Peˆgo JM, Silva D, Correia-Pinto J et al (2007) Transvesical thoracoscopy: a natural orifice translumenal endoscopic approach for thoracic surgery. Surg Endosc 21:854–858 8. Liu Y, Liu H, Wu Y, Ko P (2010) Feasibility of transtracheal surgical lung biopsy in a canine animal model. Eur J Cardiothorac Surg 37:1235–1236 9. Liu Y, Liu H, Wu Y, Ko P (2010) Feasibility of transtracheal thoracoscopy (natural orifice transluminal endoscopic surgery). J Thorac Cardiovasc Surg 139:1349–1350
