| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2006;82:1821-1827
© 2006 The Society of Thoracic Surgeons
a Department of Thoracic Surgery, Katholisches Klinikum Koblenz, Koblenz
b Department of Thoracic Surgery, Klinik Schillerhoehe, Stuttgart, Germany
Accepted for publication May 15, 2006.
* Address correspondence to Dr Witte, Department of Thoracic Surgery, Katholisches Klinikum Koblenz, Kardinal-Krementz-Str. 1-5, 56073 Koblenz, Germany (Email: b.witte{at}kk-koblenz.de).
| Dr Huertgen discloses that he has a financial relationship with Richard Wolf, Knittlingen, Germany.
|
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentSouthern Thoracic Surgical...AcknowledgmentsReferences |
|---|
METHODS: From 2000 to 2004, we documented and assessed 226 consecutive procedures in a prospective database.
RESULTS: A total of 144 VAMLAs for the staging of resectable bronchial carcinoma, and 82 less extensive procedures for other indications were performed, combined with extended mediastinoscopy in 72 patients and with mediastinoscopic sonography in 26. Mean operation time was 54.1 minutes for VAMLA and 36.6 minutes for less extensive procedures. We observed nine complications: five recurrent nerve paralyses, one arterial and two venous injuries, and one mediastinitis. The complication rate was 3.98%, which dropped from 5.3% to 2.6% with growing experience. VAMS detected mediastinal lymph node involvement in 61 (32.8%) of 186 patients with bronchial carcinoma (N2, 45; N3, 16). Mediastinal reassessment at open surgery was done in the 130 resected patients and showed for VAMLA a specificity of 93.75%, a sensitivity of 100%, and a false-negative rate of 0.9%.
CONCLUSIONS: In our institutional practice, VAMS has replaced conventional mediastinoscopy for reasons of extended surgical options, safety, precision, education, documentation, and enhanced accuracy of pretherapeutic mediastinal staging. Mediastinal staging of resectable bronchial carcinoma is done by VAMLA, because the accuracy is equal to open lymphadenectomy and the access to the left paratracheal and tracheobronchial lymph nodes is improved. No increase in the complication rate was observed. Prolonged operation time was due to more extended procedures not possible with conventional mediastinoscopy, like VAMLA.
The technical features of mediastinoscopy remain essentially unchanged since its introduction by Carlens in 1959 [1]. During the last decade, two significant technical modifications of the conventional mediastinoscope changed mediastinoscopy forever. The idea of using a videocamera in mediastinoscopy comparable with video-assisted thoracoscopic surgery (VATS) is credited to Toni Lerut in 1989 [2]. The realization of this idea improved imaging of the mediastinal structures dramatically and made mediastinoscopy more standardized, user-friendly, and accessible to trainees. However, it was the development of a twin-bladed expanding videomediastinoscope by Linder and Dahan in 1992 [3] that allowed increased exposure, bimanual dissection, and the development of new surgical techniques like video-assisted mediastinoscopic lymphadenectomy (VAMLA) [3] or videomediastinoscopic exstirpation of mediastinal cysts [4].
Thus, Carlens' "method for inspection and tissue biopsy in the superior mediastinum" [1] is on its way to developing from the level of diagnostic endoscopy to the level of complex surgical procedures. This shows some parallels with the development from laparoscopy to laparoscopic surgery, or from thoracoscopy to VATS, probably with a similar perspective for its future. Therefore, we suggest the term VAMS for video-assisted mediastinoscopic surgery.
After preliminary studies [3], we introduced twin-bladed videomediastinoscopes as part of our clinical routine. In this article we analyze our experience with this new technique and report 226 consecutive cases, with focus on indications, procedures, and clinical feasibility.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentSouthern Thoracic Surgical...AcknowledgmentsReferences |
|---|
In contrast with a conventional mediastinoscope, the Linder-Dahan videomediastinoscope consists of a twin-bladed speculum that can be spread open to create a wide operative field for bimanual preparation with various grasping forceps, dissectors, clip appliers, scissors, and dissection aspirators (Fig 1). VAMLA dissection technique includes en bloc resection of the subcarinal, the right paratracheal, right tracheobronchial, and pretracheal compartments, and dissection and sampling of the the left-sided tracheobronchial and paratracheal compartments (Fig 2). It is guided by anatomic landmarks, very similar to open dissection.
|
|
Indications for videomediastinoscopy were the same as for conventional mediastinoscopy: bronchial carcinoma, lung metastases with suspected mediastinal involvement, mediastinal lymphomas of unknown etiology, and mediastinal tumors. VAMLA was indicated in cN0 or cN1 bronchial carcinoma staged by computed tomography (CT) scan to detect metastases in radiologically unsuspicious lymph nodes. Additional procedures were extended mediastinoscopy [6] to assess lymph nodes at station 5 and 6, and mediastinoscopic ultrasound [7] for suspected T4 with infiltration of superior vena cava, central pulmonary artery, and right atrium.
All patients had chest roentgenogram and CT scan before videomediastinoscopy.
A total of 226 consecutive videomediastinoscopic procedures were documented prospectively in a database, and assessed for patient data, indications, oncologic data, type and duration of the procedures, complications, and complication management. Accuracy of mediastinal lymph node staging was assessed by weighing the specimens of 144 VAMLA procedures and by reassessing the mediastinum in 130 resected cases.
Informed written consent of each patient was obtained. The interdisciplinary board of heads of departments in the institution evaluated and approved the study protocol. Because the protocol had impact only on data collection and did not change routine clinical proceedings, a decision was made not to apply for the votum of an external review board.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentSouthern Thoracic Surgical...AcknowledgmentsReferences |
|---|
The most frequent indication for VAMS was lung cancer in 186 patients (82.3%), and 144 presented with resectable bronchial carcinoma and received VAMLA for mediastinal staging. Forty-two patients had irresectable bronchial carcinoma and underwent minor VAMS procedures to determine histology or to confirm mediastinal nodal involvement. Other indications for minor VAMS procedures were mediastinal lymphoma in 34 patients (15%), enlarged mediastinal lymph nodes in 5 with lung metastases (2.2%), and suspected mediastinal tumor in 1.
Characteristics of the 186 bronchial carcinomas were 80 right-sided (43%) and 106 left-sided tumors (57%), 150 non-small-cell lung carcinoma (80.6%), 5 large-cell carcinoma (2.7%), 11 small-cell lung carcinoma (5.9%), 8 mixed (4.3%), and 12 with unknown histology (6.5%). The surgically determined nodal status of the 186 patients with bronchial carcinoma was N0 in 90 (48.4%), N1 in 30 (16.1%), N2 in 47 (25.3%), N3 in 16 (8.6%), and not specified in 3 (1.6%). All 16 cases of N3 disease and 45 of 47 cases of N2 disease were identified mediastinoscopically. In 2 patients, ipsilateral lymph node involvement was overlooked by mediastinoscopy and detected at thoracotomy. In both cases, involved nodes were detected by right-sided thoracotomy at the right paratracheal and subcarinal position.
In 144 patients, VAMLA was indicated for mediastinal staging of resectable bronchial carcinoma with unsuspicious lymph nodes on CT scan. The performing surgeon judged 125 VAMLA procedures (86.8%) to be complete, and 19 (13.2%) had to be left incomplete. The reasons for incomplete VAMLA were silicotic or calcified lymph nodes in 6 patients, extranodal tumor growth in 4, extensive scarring in 2, intraoperative complications in 3, contraction of cervical spine in 2, extreme adipositas in 1, and technical problems with the videomediastinoscope in 1 patient. The mean sample weight of the VAMLA specimens was 8.7 grams (range, 2 to 23.7 grams).
Additional mediastinoscopic investigations were performed in 98 patients (43.4%), consisting of 72 extended mediastinoscopies [6] and 26 mediastinal sonographies [7]. In 8 cases, frozen sections of specimens were done. Other procedures were performed in 83 patients (36.7%) during the same anesthesia, consisting mainly of rigid bronchoscopy (n = 50) and transesophageal echocardiography (n = 17).
Mean operation time was 54.1 minutes (range, 40 to 175 minutes) for VAMLA and 36.6 minutes (range, 20 to 120 minutes) for less extensive procedures.
Nine complications were observed in 9 patients (3.98%): one right and four left recurrent laryngeal nerve palsies, two intraoperative lacerations of the azygos vein, one mediastinitis, and arterial bleeding from the aortic arch one. The latter was related to extended mediastinoscopy and not to videomediastinoscopy itself. In the first half of our sample (patients 1 to 113), six complications were noted, in the second half (patients 114 to 226) three complications occurred, dropping the complication rate from 5.3% to 2.6%.
Nine patients had received neoadjuvant chemotherapy or radiochemotherapy before videomediastinoscopy. The operation time was not prolonged in these patients. In two, complete mediastinal nodal dissection was impossible because of scarring and persistent bulky disease.
In 130 of 144 VAMLA patients, thoracotomy was performed for lung resection and mediastinal reexploration to assess scarring and remnants of lymphatic tissue, 19 of them after neoadjuvant therapy. Marked scarring of the tracheobronchial and subcarinal region with difficult reexploration was noted in 25 patients (19.2%). No scarring was observed in all 11 patients resected within 6 days after VAMLA. The mean period of time between videomediastinoscopy and resection was 12.0 days in patients with scarring and 12,7 days in patients without. Among 19 patients resected after neoadjuvant therapy and VAMLA, 6 had excessive and 11 had no scarring.
The only bronchial stump insufficiency (0.8%) occurred after a right-sided pneumonectomy. In this patient, the bronchial stump had not been covered, and resection was preceded by neoadjuvant chemoradiation therapy and repeat mediastinoscopy.
Lymph node remnants at videomediastinoscopically explored or dissected regions were identified in 24 (18.5%) of 130 thoracotomies. In 14 of those 24 patients, the finding of remaining lymph nodes was expected, as they had an incomplete VAMLA before. Lymph node remnants were located most often in the subcarinal station (n = 28), and were less frequent in the tracheobronchial, paratracheal, or pretracheal regions (n = 14). Positive lymph nodes overlooked by mediastinoscopy and detected at thoracotomy were found in 2 patients at the paratracheal and subcarinal station.
Figure 3 gives an overview over our heterogenous study collective.
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentSouthern Thoracic Surgical...AcknowledgmentsReferences |
|---|
In 144 patients with resectable bronchial carcinoma, systematic mediastinal lymph node dissection was done by VAMLA, a procedure impossible to perform by conventional mediastinoscopy or by videomediastinoscopy with tubular, nonexpanding mediastinoscopes. Owing to the more extensive procedure, the mean operation time of VAMLA was prolonged compared with VAMS (54.1 versus 36.6 min).
The performing surgeon judged 125 (86.8%) of 144 VAMLA procedures as complete, and 19 (13.2%) were considered incomplete dissections. Those 19 cases of intended, but not completed VAMLA procedures give an idea of the technical limitations of VAMS. Incompleteness of mediastinal dissection was most often due to preexistent morbidity as calcified or silicotic lymph nodes, extensive scarring, contracted cervical spine, or extreme obesity, in 11 cases. Other causes for incomplete dissections were extranodal tumor growth in 4 patients with not anticipated mediastinal involvement, intraoperative complications in 3, and technical problems with the equipment in 1.
Calcified or silicotic lymph nodes, adhesions, and scars are common mediastinoscopic findings. In those cases, mediastinal dissection is complicated by adhesions to adjacent mediastinal structures and lack of proper dissection planes. The value of a complete mediastinal dissection should be weighed carefully against the risk of airway, vascular, or esophageal injury. Sometimes, it is favorable to leave lymph node remnants or scarred tissue on vital mediastinal structures.
In most patients with contracted, scarcely reclinable cervical spines, dissection can be facilitated by gentle rotation or lateroversion of the head. Generally, a hypertrophic mediastinal fat pad impairs sight as it protrudes into the mediastinoscope and soils the endoscope lens with a greasy film. Large amounts of fatty tissue, especially in the pretracheal compartment, require dividing of the specimen.
Extranodal tumor growth into the mediastinal fatty tissue should only undergo biopsy. The attempt of complete dissection is likely to increase the risk of local complications and tumor propagation, and is senseless from the oncologic point of view.
All intraoperative vascular complications impaired the extent of lymph node dissection. In one case of arterial hemorrhage from the origin of the left carotid artery, immediate reconstruction through a sternotomy was mandatory, and nodal dissection had to be left incomplete. In 2 patients, the azygos vein was surrounded by enlarged tracheobronchial lymph nodes and started to tear during separation. Bleeding was controlled mediastinoscopically, and no further attempt was made to remove those lymph nodes completely.
At first sight, we were surprised that a technique with improved visibility and preparation had a complication rate of 3.98%. Our data, collected from the very beginning of the clinical introduction of videomediastinoscopy, suggest a relation between complications and extent of the procedure (Fig 4), and a learning curve, as the complication rate decreased from 5.3% to 2.6%. Other VAMS investigators reported complication rates of 4% in 219 patients who had lymph node biopsies for bronchial carcinoma [8] and 0.83% in 240 patients with biopsies for diverse indications [9]. The complication rate of conventional mediastinoscopy ranges from below 1% in sarcoidosis to 5% in bronchial carcinoma [10]. Thus, the frequency of complications depends more on the extent of the mediastinoscopic procedure and on the presence of malignancy than on technical features.
|
One case of arterial bleeding occurred during extended mediastinoscopy. The left carotid artery was lacerated at its origin from the aortic arch. This required open repair and was followed by a temporary hemiplegia.
Two small lacerations of the azygos vein occurred in bulky mediastinal disease. In both cases, the venous leak was compressed with a stick sponge and occluded with hemo-clips (Fig 5). In these patients, VAMS was beneficial in allowing a minimally invasive repair. One postoperative mediastinitis was thoracoscopically drained and healed uneventfully.
|
In 130 patients who underwent thoracotomy and lung resection after VAMLA, we observed that VAMLA was likely to cause adhesions and scars, and lung resection might be more difficult if postponed by more than 1 week. We did not notice any negative impact of VAMLA on bronchial stump healing, as the insufficiency rate was 0.8%.
This series of 226 videomediastinoscopies included 144 VAMLA procedures. The extent and thoroughness of mediastinal lymphadenectomy by VAMLA was assessed by weighing the VAMLA specimens and by mediastinal reexploration at open thoracotomy.
In a preliminary series of 46 VAMLA procedures [3], we had already investigated the number of removed lymph nodes. It was 20.7 lymph nodes (range, 5 to 60), a number equal to open lymphadenectomy [3], corresponding with a mean sample weight of 10.1 grams (range. 3.1 to 24 grams). In the larger series of 144 patients, we preferred weighing of VAMLA specimens to meticulous dissection with exact lymph node count for reasons of practicability in clinical routine, and found a mean sample weight of 8.7 grams (range, 2 to 23.7 grams), which is not different from the preliminary series. So, VAMLA was performed in the preliminaly study and in clinical routine with comparable radicality, equal to open lymphadenectomy.
Of the 144 VAMLA patients who underwent thoracotomy for tumor resection, 130 had mediastinal reexploration in the same procedure. Mediastinal involvement was diagnosed by VAMLA in 15 patients, and only 2 had mediastinal involvement diagnosed at thoracotomy that VAMLA had overlooked. The actual incidence of mediastinal involvement was 17 (13.1%) of 130. Thus, the sensitivity of all VAMLA procedures was 88.2%, the specificity was 100%, the false-negative rate was 1.7%, and the false-positive rate was 0%.
It should be noted that the resection group contained 14 VAMLA procedures (10.8%) judged to be incomplete by the surgeon at the time of VAMLA, representing 1 of 2 patients with overlooked positive lymph nodes. For complete VAMLA procedures, sensitivity is 93.75%, and the false-negative rate is 0.9% (Table 1). Comparison of the VAMLA sensitivity of 93.75% with sensitivities reported for other methods of mediastinal staging such as conventional mediastinoscopy, CT, or positron emission tomography scan is questionable, because VAMLA sensitivity refers to the ability of VAMLA to detect lymph node metastases in a mediastinum classified as uninvolved by CT scan and not to the detection of mediastinal involvement in lung cancer patients.
|
In a strategy of resection and adjuvant therapy, it is sufficient to detect minor mediastinal involvement at thoracotomy, and VAMLA seems to provide no further information. This is only true for right-sided tumors and their ipsilateral mediastinal compartment, however, where mediastinal lymphadenectomy at thoracotomy can be complete.
On the left side, complete dissection of the ipsilateral tracheobronchial and paratracheal nodes requires the division of the ligamentum arteriosus Botalli and is not routinely performed. Dissection of the contralaterel mediastinal compartments via left thoracotomy remains necessarily incomplete. Unfortunately, left-sided tumors, especially those of the lower lobe, tend to metastasize to the contralateral stations 2 and 4, and a considerable incidence of missed metastases has been reported. Hence, the "resection and adjuvant therapy" strategy applied to left-sided tumors runs a risk of overlooking ipsilateral and contralateral mediastinal involvement. As VAMLA dissection of the stations 2R and 4R can be easily achieved, left-sided tumors might be an indication for VAMLA with any therapeutic strategy.
In our opinion, neoadjuvant treatment might be beneficial for all patients with any—even minor—extent of mediastinal involvement; therefore, it is crucial to detect even minor mediastinal disease pretherapeutically and not only at thoracotomy. This has been an important rationale for the development of VAMLA and its introduction into clinical practice.
In our institutional practice, videoassisted mediastinal surgery with the expanding Linder-Dahan mediastinoscope has replaced conventional mediastinoscopy for reasons of extended surgical options, safety, precision, education, documentation, and enhanced accuracy of pretherapeutic mediastinal staging. No increase in complication rate was observed.
The possibility of bimanual dissection allows procedures not possible with conventional mediastinoscopy, such as safe management of minor bleeding and complete mediastinal lymphadenectomy (VAMLA). From the oncologic point of view, VAMLA may be indicated if neoadjuvant therapy before resection is considered for minor mediastinal involvement and pretherapeutic detection is essential. From the surgical standpoint, VAMLA can reach the accuracy of open surgery and is feasible to increase the extent of mediastinal lymphadenectomy especially in left-sided bronchial carcinoma.
|
Southern Thoracic Surgical Association: Fifty-Third Annual Meeting |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentSouthern Thoracic Surgical...AcknowledgmentsReferences |
|---|
Please visit the STSA (http://www.stsa.org) or CTSNet (http://www.ctsnet.org) websites for detailed information regarding the meeting.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentSouthern Thoracic Surgical...AcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentSouthern Thoracic Surgical...AcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. Bar, M. Papiashvilli, G. Fink, J. Sandbank, and D. Stav Cervical Mediastinoscopic Lymphadenectomy for Accurate Staging in Lung Cancer Asian Cardiovasc Thorac Ann, August 1, 2009; 17(4): 357 - 361. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Witte, A. Messerschmidt, H. Hillebrand, S. Gross, M. Wolf, E. Kriegel, W. Neumeister, and M. Hurtgen Combined videothoracoscopic and videomediastinoscopic approach improves radicality of minimally invasive mediastinal lymphadenectomy for early stage lung carcinoma Eur. J. Cardiothorac. Surg., February 1, 2009; 35(2): 343 - 347. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Karfis, E. Roustanis, J. Beis, and J. Kakadellis Video-assisted cervical mediastinoscopy: our seven-year experience Interactive CardioVascular and Thoracic Surgery, December 1, 2008; 7(6): 1015 - 1018. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Ernst, F. J. F. Herth, R. Eberhardt, and M. Krasnik Response Chest, September 1, 2008; 134(3): 672 - 673. [Full Text] [PDF]
|
|
|
|
|
|
B. Witte, W. Neumeister, and M. Huertgen Does endoesophageal ultrasound-guided fine-needle aspiration replace mediastinoscopy in mediastinal staging of thoracic malignancies? Eur. J. Cardiothorac. Surg., June 1, 2008; 33(6): 1124 - 1128. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Leschber, D. Sperling, W. Klemm, and J. Merk Does video-mediastinoscopy improve the results of conventional mediastinoscopy? Eur. J. Cardiothorac. Surg., February 1, 2008; 33(2): 289 - 293. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Witte and M. Hurtgen Video-assisted mediastinoscopic lymphadenectomy MMCTS, October 18, 2007; 2007(1018): 2576. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
