Gas-chamber mediastinoscopy for dissection of the upper esophagus

Introduction
Dissection of the proximal thoracic esophagus during transhiatal esophagectomy is a blind procedure which carries specific operative risks. In the upper mediastinum the esophagus has intimate contact with the membranous trachea, recurrent laryngeal nerves and the arch of the azygos vein, structures which are especially vulnerable during the combined transhiatal and cervical approach. Furthermore, the lack of visual control over the operative field makes lymphadenectomy impossible in this area.
Mediastinoscopy uses the developments of video-endoscopic techniques to visualize the upper mediastinum using exclusively a cervical access. Nowadays this procedure is used mostly for diagnosis or biopsy of mediastinal masses or for surgical staging in patients with lung cancer (1). There are also reports which have shown that medias-tinoscopy may play a role in dissection of the upper esophagus in patients subjected to esophageal resection (2-7). However, the experience with mediastinoscopic-assisted esophagectomy is restricted to small number of patients while the technique is not standardized yet.
The aim of the present study is to evaluate in detail in an experimental setting the feasibility of the gas-chamber mediastinoscopy technique, a variant approach to medias-tinoscopic dissection of the upper esophagus.

Materials and Method
The experiment was performed in nine German Landrace pigs with an average weight of 25.02 kg.
After premedication with an intramuscular injection of 2 mg/kg azaperon (Stressnil®) and 0.5 mg/kg midazolam (Dormicum®) narcosis was induced by intravenous administration of 1.5 mg/kg midazolam and 10 mg/kg ketamine (Ketanest®). Animals were intubated and anaesthesia was maintained using an isoflurane enriched O2/air mixture. Fentanyl (500 Og/h i.v.) was used for analgesia and pancuronium (0.25 mg/kg/h) for muscle relaxation. The common carotid artery and internal jugular vein were catheterized and connected to membranous pressure transducers for continuous measurement of mean arterial pressure and central venous pressure, respectively. Ringer lactate solution was infused continuously during the operation at a rate of 20 ml/kg/h. Heart rate and rhythm were continuously monitored by a surface electrocardiogram. Temperature was monitored with a rectal thermometer and maintained at 37ºC throughout the operation by laying the animals on a heated blanket. Blood samples were withdrawn repeatedly during the experiment and analyzed in a blood gas analyzer (ABL50; Radiometer, Copenhagen, Denmark). Variations in O2 and CO2 blood gas pressure and saturation directed the fine tuning of ventilatory parameters. At the end of the experiment, while animals were still in deep anesthesia, a central venous injection of 60 mmol potassium chloride was administered to produce cardiac arrest and death.
During the experiments, all animals received care in compliance with the European regulations for animal experiments (8) and the United States National Research Council’s criteria as outlined in "Guide for the Care and Use of Laboratory Animals" prepared by the National Institution of Health (NIH publication no. 86-23, revised 1985).
Technique of mediastinoscopy
The left neck and chest were prepared and draped. Through a 6-cm long longitudinal left cervical incision a virtual working space was created around the cervical esophagus using conventonal sharp and blunt digital dissection techniques.
Mediastinoscopy was performed using a Karl-Storz® laparoscopic surgical unit placed at the caudal extremity of the operating table and equipped with a 30º laparoscope. There was no special instrumentation used in this procedure, the 5-mm and 10-mm trocars and the 35-mm long operating instruments being identical with the ones used in conventional laparoscopic surgery. Two 5-mm trocars were inserted 2 cm lateral from the cervical incision. Introduction of trocars was performed with visual and tactile feed-back to avoid injuries to the vital anatomical structures situated in close proximity to the working area. After penetration of the skin and superficial muscle layers, the shaft of the trocars was laid in the already created periesophageal cavity. The 10-mm optic trocar was placed at the superior pole of the cervical incision and the latter was closed tight with a running 3.0 polypropylene suture. CO2 was insufflated at a pressure of 5-mmHg and the peri-esophageal virtual chamber became a real space in which operative procedures became possible. (Fig. 1)

The operation was performed thorough the two 5-mm working trocars using non-traumatic graspers, scissors and the hook monopolar cautery. Attachments of the esophagus to the longitudinal vertebral ligament posteriorly, membranous trachea anteriorly and both pleura laterally were divided under strict visual control. Attempts were made to remain inside the mediastinum and avoid entering the pleural cavities. To obtain adequate exposure of the dissection plane the esophagus was carefully manipulated without grasping it, only by lifting it up or pushing it down or sideways with a blunt instrument introduced through the caudal 5-mm trocar. Proper use of the 30º laparoscope improved further visualization of mediastinal structures. Dissection proceeded gradually along the esophagus until the level of the tracheal bifurcation was reached. (Fig. 2)
A series of parameters were recorded during the operation: feasibility of esophageal dissection, presence of intraoperative incidents, blood loss, time needed to perform the procedure and the success rate in identification of surrounding anatomical structures such as the tracheal bifurcation, periesophageal lymph nodes and the arch of the azygos vein.

Results
All operations were successful, dissection of the upper esophagus being completed down to the level of tracheal bifurcation. Performed always under visual control, the procedure proved to be feasible and safe. The working instruments had enough freedom of movement to allow an intuitive and rather straightforward dissection. Manipulation of the esophagus and the side view offered by the 30º laparoscope offered adequate visualization on the operative field while individual coagulation of small esophageal arteries resulted in an almost bloodless operation (on average <20 ml). The plane of dissection could easily be developed around the esophagus and there were no injuries either to the esophageal wall or to the adjoining anatomical structures. The posterior membranous trachea, tracheal bifurcation and the origin of both right and left bronchi were accurately identified intra-operatively in all cases. The right and left pleura, representing in fact the lateral limits of dissection, were also visualized, an injury of the right pleura being observed during the operation in five of the nine cases (55,5%). Pleural lesions involved mainly the right pleura, as instruments came from the left side. Visualization of the arch of the azygos vein and para-esophageal lymph nodes was less successful, these structures being clearly identified in only two and three cases respectively.
There were no major intraoperative incidents. In two cases there was hemorrhage at the point of trocar insertion requiring surgical hemostasis while in another experiment tear of a small periesophageal artery caused minor bleeding arrested with monopolar electrocoagulation. In two of five cases with pleural injury there were alterations in blood oxygenation (hypercapnia, hypoxia and respiratory acidosis) without cardiovascular instability, the mean blood pressure and heart rate remaining within normal limits. When these complications occurred the operation was temporarily stopped, CO2 was released form the working chamber and ventilatory parameters were modified as follows: use of PEEP, increase of minute volume and hyperventilation. When SaO2, pO2 and pCO2 returned within normal range the operation was resumed and 3-5 mmHg PEEP (Positive End Expiratory Pressure) was added to the standard settings until the end of the procedure.
The overall mean operative time was 29.2 min of which 12.5 minutes were necessary for conventional dissection, 6.6 min were required to insert the trocars and close the cervical incision and the effective mediastinoscopic dissection was completed in 10 min. (Table 1)

Disscutions
The transhiatal approach to esophagectomy (THE) popularized by Orringer (9) was introduced in an attempt to reduce the high rates of morbidity and mortality accompanying trans-thoracic resections. Presence of intraoperative bilateral lung ventilation and avoidance of a thoracic incision are the important advantages of this technique reflected in a lower incidence of postoperative pulmonary complications compared with the transthoracic approach (10). However, as dissection is performed blindly in the upper mediastinum, new complications and potential dangers appeared. Injuries of the arch of the azygos vein or membraneous trachea are formidable complications which may jeopardize the life of the patient if not recognized and treated immediately (11). Lesions of the recurrent nerves have also been shown to occur more frequently after transhiatal esophagectomy (12). Besides hoarseness, their permanent damage is associated with poor quality of food intake, decreased ability to swallow liquids, and repetitive bouts of aspiration pneumonia due to the inability to close the glotis (13, 14). Furthemore, THE does not offer informations about lymph node status in the middle and upper mediastinum and thus accuracy of intraoperative and postoperative staging is altered while the risk of local recurrence is increased.
Cervical mediastinoscopy, a surgical procedure that offers visual and instrumental access to the upper mediastinum through a remote cervical incision may overcome some of the above mentioned incidents and limitations. Developed by Carlens in Sweden and popularized by Pearson in North-America (15,16), the classical technique develops a dissection plane anterior to the trachea and it is frequently used nowadays as a staging procedure to assess mediastinal lymph nodes in patients with lung cancer or for diagnosis of mediastinal masses (1).
Advances in endoscopic technology allowed development of an endoscopic variant of mediastinoscopy designed for esophageal resection called endoscopic microsurgical dissection of the oesophagus (EMDOE) (4). It performs dissection in a plane posterior to trachea and requires a special device which has a blunt tissue dilator at the tip designed to ride on the exterior wall of the esophagus and create a periesophageal space by mechanical dilatation. Separate openings for the fiberoptic bundle, working channel and flushing or suction devices offer the possibility to perform a visually controlled dissection. The feasibility and efficacy of the technique were proven by Buess and later on by Bumm in experimental and clinical studies (6,7). Bonavina (3) added successfully cervical video-mediastinoscopy to laparoscopic esophagectomy in a group of 12 patients and concluded that the procedure allows dissection of the upper thoracic esophagus under direct visual control and achieves a more accurate hemostasis compared to the conventional digital approach and thus has the potential to reduce morbidity in carefully selected patients. Despite these favorable results, the procedure is technically difficult. Structures to be dissected are only centimeters away from the mediastinoscope and orientation in the operative field, identification of anatomical structures and evaluation of bleeding may be distorted. Moreover, there is a single actuator instrument whose freedom of movement is limited to only one axis, continuous suction and permanent rinsing are mandatory to preserve a proper working operative field while lymphadenectomy is time consuming and is usually restricted only to node sampling (7).
Other technical variants were reported for medias-tinoscopic-assisted dissection of the upper esophagus. Tangoku (2) attached a retractor with a transparent flat tip to a 5-mm videoscope which was introduced in the peri-esophageal space through a left collar incision. As the working space was maintained open by the retractor, a visually controlled periesophageal dissection was performed using conventional non-laparoscopic instruments in 41 patients operated with a transhiatal esophagectomy approach. The thoracic duct, arch of the azygos vein, vagus and recurrent laryngeal nerves were identified and protected during the operation while visualization of adjacent periesophageal lymph nodes allowed surgical sampling.
Ikeda (5) described the technique of gas-chamber medias-tinoscopy for dissection of the esophagus from the cervical region down to the diaphragm in five patients. Although the technique seems promising it did not pass the experimental stage, an enquiry on Pubmed showing that it is reported in only two papers, belonging to the same author, in 2001 and 2002, on limited number of patients (5,17).
We have considered that the gas-chamber mediastinoscopy technique has certain advantages for dissection of the upper mediastinum and have designed this experimental animal model to further evaluate it. Unlike Ikeda, we have used a rigid laparoscope instead of a flexible endoscope and the esophagus was dissected only until the level of the tracheal bifurcation, covering the area least accessible for dissection during transhiatal esophagectomy. The rationale behind these modifications was to provide a feasible procedure which, coupled with a laparoscopic transhiatal approach for esophageal dissection, would offer an easy and safe totally endoscopic dissection of the entire esophagus.
The study has shown that the technique fulfils three conditions essential for a successful operation: a) easy creation and maintenance of a proper working space, b) bi-manual dissection and c) an appropriate freedom of movement for the operating instruments. The technique is simple and does not need to develop or purchase specially designed devices. The virtual cavity created by blunt finger dissection around the cervical esophagus can easily be maintained open using CO2 insufflation the way pneumo-peritoneum is used in laparoscopic surgery and offers a stable working space and appropriate view of the anatomical structures. The 5-mmHg insufflation pressure is sufficient for adequate distension of the virtual cavity but low enough to prevent troublesome side-effects of tension pneumothorax accompanying pleural injuries. Alterations of pulmonary parameters seen in cases with pleural injuries were not accompanied by cardiovascular abnormalities and responded well to adjustments of ventilatory parameters.
Manipulation of instruments is intuitive and, although there is a certain amount of difficulty due to the short distance between the working trocars, the degree of movements was satisfactory for a safe and, in the end, even straightforward dissection. Moreover, this technique reproduces the setting specific for minimally-invasive surgery allowing surgeons who have already developed appropriate skills in laparoscopic procedures to easily go through the learning curve. Proper use of the angled laparoscope and skilful manipulation of the grasper achieved satisfactory exposure of the dissecting plane for an incident-free procedure. One of the difficulties encountered during the procedure was to manipulate the esophagus without grasping it. One possible improvement in this respect could come from development of a variant “gold finger” which encircles the esophagus completely and could push it easier in any direction without the risk of slippage.
The timing for introduction of trocars was modified from the conventional laparoscopic technique. In the first experiment, the lateral trocars were introduced after creation of pneumomediastinum. However, unlike laparos-copic surgery, the mediastinal working chamber is small in size and surrounded by vital anatomical structures. Therefore introduction of trocars for the rest of the study was done prior to closure of the cervical incision under strict visual and tactile control to avoid iatrogenic injuries.
The technique provided constant visual control over the operative field. Important anatomic structures such as trachea, carina, the emerging left and right bronchus and the arch of the azygos vein were identified and protected. Although lymph nodes were less visible during our experiment probably due to their small size, absence of a significant layer of periesophageal fatty tissue and lack of anthracotic impregnation, they were still clearly identified in 33.3% of the cases. This figure is still not satisfactory and may be improved with the use of techniques that increase the rate of lymph node detection, such as peritumoral injection of vital dyes which can be coupled with sentinel node harvesting. Data from the present experiment does not support the presumption that a complete lymphadenectomy can be performed using the mediastinoscopic approach. However, it shows that lymph nodes can be visualized and, if desired, sampled to offer additional information for intraoperative operative staging and individual tailoring of the treatment strategy. The importance of lymph node sampling in the upper mediastinum is outlined by reports which have shown that the lymph nodes along the recurrent nerve have the highest incidence of metastasis in cases with thoracic esophageal cancer (18,19). As they connect the thoracic and cervical lymph node groups, they also seem to act as sentinel nodes for the cervical lymph node stations predicting existence of metastases and thus the need for further cervical lymphadenectomy (20). For some authors, identification of metastases at this level represented a decisive factor to change the operative strategy and switch to a more radical procedure with extended lymph node dissection (20,2).
The most important intraoperative incident was represented by alterations of blood gas parameters observed in four of six cases with pleural injury which developed hypoxemia with hypercapnia and respiratory acidosis. There seems to be a particular sensitivity of the pigs, which have a unique pleural cavity, to tension pneumothorax, similar findings being reported in experimental studies that evaluated the effects of positive pressure intrapleural CO2 in pigs (21,22,23). This complication had no effect on the success of the procedures, temporary removal of the CO2 and adjustments of ventilatory parameters restoring the pulmonary parameters within normal range in all cases.
In conclusion, the technique of gas-chamber CO2 mediastinoscopy for dissection of the proximal esophagus is feasible and safe. It offers a fair amount of freedom of movement for the working instruments and adequate view over the operative field for a controlled and accurate dissection. Furthermore, it visualizes periesophageal lymph nodes leading to improved intraoperative staging and possible individualization of the treatment strategy. Further experimental and clinical studies are required to validate the technique and compare it with other variants of mediastinoscopy.

References
1. De Leyn, P., Lerut, T. - Multimedia Manual of Cardiothoracic Surgery, European Association for Cardio-thoracic Surgery, 2005, 1-10.
2. Tangoku, A., Yoshino, S., Abe, T., Hayashi, H., Satou, T., Ueno, T., Oka, M. - Mediastinoscope-assisted transhiatal esophagectomy for esophageal cancer. Surg. Endosc., 2004, 18:383.
3. Bonavina, L., Incarbone, R., Bona, D., Peracchia, A. - Esophagectomy via laparoscopy and transmediastinal endodissection. J. Laparoendosc. Adv. Surg. Tech. A., 2004, 14:13.
4. Buess, G., Becker, H., Lenz, G. - Perivisceral endoscopic oesophagectomy. În “Operative manual of endoscopic surgery”, sub redactia Cuschieri A., Buess G., Perissat J., Springer-Verlag (Berlin Heidelberg) 1992, pag. 149-165.
5. Ikeda, Y., Niimi, M., Kan, S., Sasaki, Y., Shatari, T., Takami, H., Kodaira, S. - Mediastinoscopic esophagectomy using carbon dioxide insufflation via the neck approach. Surgery, 2001, 129:504.
6. Buess, G., Kaiser, J., Manncke, K., Walter, D.H., Bessell, J.R., Becker, H.D. - Endoscopic microsurgical dissection of the esophagus. Int. Surg., 1997, 82:109.
7. Bumm, R., Holscher, A., Feussner, H., Tachibana, M., Bartels, H., Siewert, J. - Endodissection of the Thoracic Esophagus. Ann. Surg., 1993, 218:97.
8. Council of Europe - European convention for the protection of vertebrate animals used for experimental and other scientific purposes. European Treaty Series, 1986, no 123. Strasbourg.
9. Orringer, M.B., Sloan, H. - Esophagectomy without thoracotomy. J. Thorac. Cardiovasc. Surg., 1978, 76:643.
10. Hulscher, J.B., van Sandick, J.W., de Boer, A.G., Wijnhoven, B.P., Tijssen, J.G., Fockens, P., Stalmeier, P.F., ten Kate, F.J., van Dekken, H., Obertop, H., Tilanus, H.W., van Lanschot, J.J. - Extended transthoracic resection compared with limited trans-hiatal resection for adenocarcinoma of the esophagus. N. Engl. J. Med., 2002, 347:1662.
11. Orringer, M.B., Marshall, B., Iannettoni, M.D. Transhiatal Esophagectomy for Treatment of Benign and Malignant Esophageal Disease. World J. Surg., 2001, 25:196.
12. Bolton, J.S., Teng, S. - Transthoracic or transhiatal esophagectomy for cancer of the esophagus - does it matter? Surg. Oncol. Clin. N. Am., 2002, 11:365.
13. Baba, M., Natsugoe, S., Shimada, M., Nakano, S., Noguchi Y, Kawachi, K., Kusano, C., Aikou, T. - Does hoarseness of voice from recurrent nerve paralysis after esophagectomy for carcinoma influence patient quality of life? J. Am. Coll. Surg., 1999, 188:231.
14. Baba, M., Aikou, T., Natsugoe, S., Kusano, C., Shimada, M., Nakano, S., Fukumoto, T., Yoshinaka, H. - Quality of life following esophagectomy with three-field lymphadenectomy for carcinoma, focusing on its relationship to vocal cord palsy. Dis. Esophagus, 1998, 11:28.
15. Carlens, E. - Mediastinoscopy: a method for inspection and tissue biopsy in the superior mediastinum. Chest., 1959, 36:343.
16. Pearson, F.G. - Mediastinoscopy: a method of biopsy in the superior mediastinum. J. Thorac. Cardiovasc. Surg., 1965, 49:11.
17. Ikeda, Y., Niimi, M., Kan, S., Takami, H., Koidara, S. - Thoracoscopic Esophagectomy Combined With Mediastinoscopy via the Neck. Ann. Thorac. Surg., 2002, 73:1329.
18. Fujita, H., Sueyoshi, S., Tanaka, T., Fujii, T., Toh, U., Mine, T., Sasahara, H., Sudo, T., Matono, S., Yamana, H., Shirouzu, K. - Optimal Lymphadenectomy for Squamous Cell Carcinoma in the Thoracic Esophagus: Comparing the Short- and Long-term Outcome among the Four Types of Lymphadenectomy, World J. Surg., 2003, 27: 571.
19. Endo, M., Yoshino, K., Kawano, T., Nagai, K., Inoue, H. - Clinicopathologic analysis of lymph node metastasis in surgically resected superficial cancer of the thoracic esophagus. Dis. Esophagus., 2000, 13:125.
20. Shiozaki, H., Yano, M., Tsujinaka, T., Inoue, M., Tamura, S., Doki, Y., Yasuda, T., Fujiwara, Y., Monden, M. - Lymph node metastasis along the recurrent nerve chain is an indication for cervical lymph node dissection in thoracic esophageal cancer. Dis. Esophagus, 2001, 14:191.
21. Jones, J.L., Graeber, G.M., Tanguiling, G.G., Hobbs, G., Murray, G.F. - Effects of insuflation on hemodynamics during thoracocopy. Ann. Thorac. Surg., 1993, 55:1379.
22. Marcus, D.R., Lau, W.M., Swanstrom, L.L. - Carbon dioxide pneumothorax in laparoscopic surgery. Am. J. Surg., 1996, 171:464.
23. Sandbu, R., Birgisdottir, B., Arvidsson, D., Sjostrand, U., Ruberstsson, S. - Optimal positive end-expiratory pressure (PEEP) settings in differential lung ventilation during simultaneous unilateral pneumothorax and laparoscopy. Surg. Endosc., 2001, 15:1478.

• Timeviewed: 3968 times