Endoscopic sentinel lymphadenectomy as a new diagnostic approach in the N0 neck
- First Online:
- Cite this article as:
- Werner, J.A., Sapundzhiev, N.R., Teymoortash, A. et al. Eur Arch Otorhinolaryngol (2004) 261: 463. doi:10.1007/s00405-003-0706-8
- 53 Views
Sentinel lymphadenectomy was developed to reduce the extent of surgical interventions in cancer patients. The sentinel node (SN) concept was first established for melanoma and breast cancer; within some years, it also became increasingly popular for head and neck cancer. As soon as the required sensitivity of the method proves to be feasible in the daily clinical routine, the discussion about the best surgical approach to single or multiple SN(s) will arise. Different objectives may here compete with each other. The incision should render the best exposure of the operation site and should be expandable in case further lymph node regions have to be dissected. Finally, a good functional as well as a good cosmetic result is desirable. Endoscopic lymph node excisions were performed in patients suffering from squamous cell carcinoma of the upper aerodigestive tract located in different sites (1× uvula, 2× epiglottis, 1× glottis). In preoperatively performed ultrasonic imaging (B-mode-ultrasonography), N0 necks were assessed. In contrast to previously reported endoscopic techniques in humans, the presented method requires no insufflation of carbon dioxide or external retraction of the skin. Following laser surgical resection of the primary tumor, the SN and further lymph node(s) with accumulation of tracer substance were identified and resected endoscopically via an incision that was afterwards extended to a normal neck dissection incision. Reports of histopathologic examination of the sentinel node(s) were compared to the respective neck dissection specimens. The presented method may help to reduce the degree of invasiveness frequently attributed to sentinel lymphadenectomy once the method has been established for head and neck cancer.
KeywordsN0-neckSentinel nodeMinimally invasive surgeryEndoscopy
The approach to the N0 neck in patients who are treated surgically for carcinomas of the upper aerodigestive mucous membranes is still a matter of debate. The concept of no surgery coupled with regular clinical follow-up is opposed by the concept of performing neck dissection at an early stage. A circumscribed lymph node dissection, as has been advocated in the sentinel node concept, could be a feasible and less invasive surgical alternative than selective neck dissection.
As soon as sentinel lymphadenectomy is proven to be a sensitive method in the clinical routine for head and neck cancer, the debate about the best surgical approach to the sentinel node(s) will arise. It has to be made clear how many incisions over how many suspicious lymph nodes have to be placed, and at which point the concept of so-called minimally invasive surgery can no longer be justified. The common ground for this discussion is first of all the fact that carcinomas of the upper aerodigestive mucous membranes drain to several lymph node levels. Furthermore, it has to be considered that the incision should be able to be extended to a full scale neck dissection incision if lymph node metastases become evident. Multiple incisions on the neck should be avoided, unless they could be connected in the shape of a known incision for neck dissection like the “MacFee” approach.
Based on the above-mentioned considerations, alternative approaches should be given some thought. Alternative approaches are already in use successfully in other specialities. In the last 2 decades, the establishment of minimally invasive procedures like video-assisted endoscopic operation techniques have reduced the extent of surgical trauma significantly and could thus reduce therapy-associated morbidity. In the meantime, these therapeutic concepts have become established in gynaecology and (general) visceral surgery [1, 2, 3].
The exploration of the adipous neck tissues and the included lymph nodes in all six regions of the neck in humans became possible through the endoscopic approach as described by Cougard et al. , who developed the technique for surgery of the thyroid and the parathyroid (Fig. 1d). Cougard and coworkers abstained from carbon dioxide inflation and performed endoscopically assisted surgery via skin incisions and external traction of the skin.
Dulguerov et al.  reported on the first endoscopically performed neck dissections on five cadavers without carbon dioxide inflation in 2001. Dulguerov and coworkers placed two incisions in the midline for the instruments and an additional incision above the sternocleidomastoid muscle for the optical endoscopes (Fig. 1d). Experience with endoscopically performed excisions of lymph node regions in humans with the exception of cadaver studies have not been described so far. One reason may be the complicated anatomy of the region, which may promote accidental damage to nerves and vessels in a limited surgical field. In addition, the techniques that have been developed for surgery of the thyroid and the parathyroid do not seem to work for the dissection of several lymph node regions. Inflation of carbon dioxide for the resection of lymphatic tissue in level IV, on the other hand, bears the risk of an accidental pneumothorax in itself .
Based on this background, the current communication reports on minimally invasive, endoscopic sentinel lymphadenectomy via a single skin incision, which is either placed in a skin fold or in the area of the hockey stick incision (Fig. 1e). Through the use of a specially designed endoscope with spreadable blades, an additional inflation of carbon dioxide gas and retraction of the skin is not needed. Based on our initial results with this technique, the possibilities and limitations of this method shall be discussed.
Patients and methods
SND II-IV ipsilateral
SND II-V ipsilateral
SND I-III bilateral
SND II-IV bilateral
At the beginning of the operation, 1.2 mCi 99m Tc-Nanocoll (Amershan, Braunschweig, Germany), dissolved in 0.2–0.35 ml NaCl, was injected into the perimeter of the tumor via four injections as has been described extensively in previous publications [16, 17]. After the tracer injection, 15 min was spent waiting, a period of time during which the operation site was prepared. This time frame is usually sufficient to achieve adequate labeling of the initially draining lymph node . The next step of the operation was the transoral resection of the primary tumor. In all cases, resection of the tumors could be achieved with histologically safe margins (R0 resection). If the tumor is resected initially, a reduction of scattered radiation from the primary injection site can be achieved. This has been shown to be beneficial in the head and neck area where the primary tumor and the initial lymph node are in close proximity. Additionally, exposure of the operating staff can thus be reduced.
After the skin incision, a rigid endoscope was introduced through a specially designed tube (Storz, Tuttlingen, Germany) and the operation site explored for all anatomically and functionally relevant structures. A portion of the adipous tissue, which contains the lymph nodes, had to be resected before the radioactively labeled lymph nodes that had been identified with the gamma probe could be excised. The identified and excised lymph nodes were sent for histological examination separately. Histopathologic work-up included H&E staining, and immunohistochemical staining with the pancytokeratine-antibody MNF 116 (Chemicon, Hofheim, Germany) was done. The excised neck dissection specimens were mounted on cork plates, and the respective lymph node regions were marked with colored pins before they were sent for histological examination. All lymph nodes of each neck level were investigated by H&E staining. The histological results of the neck dissection specimens were then compared to the findings of the intraoperatively identified and separately excised sentinel node(s).
The primary aim of this investigation was to develop a feasible method for endoscopic sentinel lymphadenectomy. Therefore, full-scale neck dissection was performed after completion of endoscopic lymphadenectomy in all patients. The extent of neck dissection depended on the size and location of the primary tumor; however, in all cases selective neck dissection (SND) seemed to be appropriate (Table 1).
Results of the histopathological examination
Number of endoscopically identified SN
Location of the SN
Histology of the SN
Histology of the remainder lymph nodes
Total number of investigated lymph nodes
Level II A
1× free of tumor
17× free of tumor
Level II A
1× micro metastasis
1× micro metastasis (level II A), 30× free of tumor
2× level II A ipsilaterally
2× free of tumor
39× free of tumor
Level II A ipsilaterally
1× free of tumor
32× free of tumor
Following endoscopic lymphadenectomy, all patients were treated by SND (n=6) for therapy of the regional lymphatic basins. The initial skin incision was extended without difficulties in all cases to perform the conventional neck dissection (3× apron incision, 1× hockey stick incision). The changes of the operation site through manipulations during endoscopic lymphadenectomy did not affect the dissection and topographic overview during the conventional neck dissection. No adverse effects were noted during the operation or in the postoperative recovery period. Especially no accidental damage to nerves, secondary hemorrhages or edema became evident. In particular, no impairment of function of the oral branch of the facial nerve, the hypoglossus nerve or the accessory nerve was noted.
In the histopathologic work-up, a total of 124 lymph nodes from six neck dissection specimens were examined (median: 20 lymph nodes/neck specimen). The endoscopically identified sentinel node did represent the actual stage of metastatic lymph node status in all four patients. In detail, an endoscopically identified sentinel node reflected the histopathologic result of the neck dissection specimen [pN0(sn)] in three cases. In one further case (T2 carcinoma of the epiglottis), the intraoperatively identified sentinel node, which was found in the cranial portion of level IIA, showed a micrometastasis <2 mm [pN1(mic)(sn)]. The histological work-up of the complete neck dissection specimen showed another micrometastasis <2 mm [(pN1(mic)(sn)] in level IIA.
The first promising results with the sentinel node concept caused some euphoria. Although it is tempting to readily adopt the sentinel node concept to head and neck cancer, this should not be done uncritically. Apart from the indispensable exact definition and evaluation of method-specific sources of error and limitations, it should be an aim to define the best surgical access to sentinel lymphadenectomy in the head and neck area. The surgical access should therefore be carefully contemplated in advance, otherwise a potentially very useful procedure could be rejected prematurely.
Possible future developments could aim at video-assisted endoscopic operation techniques, which have already been described for some organs of the neck, like the thyroid, the parathyroid and branchiogenic cysts [8, 9, 10, 11, 12, 13, 14, 15]. These techniques should be used for the treatment, or rather the invasive diagnostics, of the regional lymphatic drainage in carcinomas of the upper aerodigestive tract. One advantage of the endoscopic procedure would be that the skin incision could be extended for the successive neck dissection. Thus, several lymph node regions could be reached via one incision, which could possibly reduce undesired side effects of an operation, such as disturbed skin sensitivity, scar formation and lymph edema.
Up to now, there are two major trends in implicating endoscopy in the neck. The first one is to create an air-tight cavity in the dissected area that is maintained by insufflation of CO2 similar to the procedures in the body cavities. The instruments are inserted through special trocars. This technique is called endoscopic surgery [5, 20]. The second technique is characterized by larger skin incisions as compared to purely endoscopic surgery. In this approach, the cavity is maintained through external traction to the skin or by special retractors. The endoscope here is only a tool that enhances the view through a small opening . The established procedure of CO2 insufflation allows successful thyroidectomy , resection of submandibular glands , neck dissection  and carotid endarterectomy .
Major disadvantages of the published techniques in animals and humans include the prolonged duration of the interventions, the need for a special equipment (mainly for the purely endoscopic approach, and to a lesser extent for the endoscopically assisted approach) and the need for special training of the surgical team. The large number of complications, occasionally even fatal, described in the early literature on human and animal thyroid and parathyroid surgery [4, 22, 23, 24, 25, 26, 27] may be attributed to the use of gas for sustaining the operating pouch in the majority of cases. The reported incidents include arterial and venous injuries, embolism, pneumothorax, pneumomediastinum and subcutaneous emphysema [23, 25]. Common surgical complications such as injury of cranial nerves and wound infections have also been described [6, 20]. Previous interventions in the neck are regarded as specific relative contraindications for the endoscopic surgery, although successful reoperations with endoscopic approaches have also been described .
It was our aim to develop a minimally invasive, endoscopic staging procedure, which—should the necessity occur—could be able to be extended to a full scale neck dissection in cases of positive histology. These aims required an approach via skin incisions that can be positioned as necessary depending on the main metastatic routes of the tumor.
One of the technical difficulties in the development of endoscopic sentinel lymphadenectomy is the availability of flexible gamma probes that would allow easier manipulation in limited space. A flexible probe with a reduced size of the collimator as well as a flexible coating with Wolfram and rubber could contribute to the safe identification of sentinel node(s). Specially designed instruments like scissors and forceps for endoscopic procedures adapted to the anatomic peculiarities of the respective structures would undoubtedly contribute to easier dissection and resection of sentinel node(s).
Despite the mentioned difficulties, identification and resection of the sentinel node(s) was feasible in all patients. An average of 25 min (range: 20–40 min) was required for each operation. Histopathologic examination of the endoscopically identified and resected sentinel node(s) in comparison to the neck dissection specimen showed the predictive value of the sentinel node concept, similar to the previous results with an open neck [16, 17, 18]. The intraoperatively identified sentinel node that was disease-free on histological examination reflected the actual stage of metastatic disease in all patients. In 3/4 patients the results was pN0(sn). Even more interesting was the proof of a micrometastasis in an endoscopically identified and resected sentinel node and a further micrometastasis in the neck dissection specimen in a further patient [pN1(mic)(sn)].
In the two cases in which bilateral neck dissection was performed, the primary tumor in one case was a T1 carcinoma of the uvula, with surrounding areas of mucosa showing severe dysplasia. The second case was a left-sided supraglottic T2 carcinoma that protruded to the midline. During intraoperative readings with the gamma probe, in both patients the tracer was accumulated exclusively to one side of the neck (left) within a lymph node, although the tracer had been injected via four spots in the perimeter of the tumor. During transcutaneous readings on the right side of the neck with the gamma probe, no accumulation of the tracer substance was noticed in both cases; therefore, we abstained from performing endoscopic sentinel lymphondectomy. Rather than performing sentinel lymphadenectomy, it seemed to be more appropriate to stage the contralateral neck by selective neck dissection. Indication for bilateral selective neck dissection was the high rate of bilateral occult metastases that has been reported in literature repeatedly.
Neither during subsequent intraoperative readings nor during extracorporal control readings with the gamma probe could intranodal tracer accumulation be demonstrated in the contralateral neck dissection specimen. Future investigations will have to show whether the sentinel node biopsy is a suitable method for the depiction of lymphatic drainage of midline-based carcinomas (e.g., epiglottis).
The first results on endoscopically performed selective lymphadenectomy lead to the assumption that this method of lymph node dissection could achieve some significance in the therapeutic regime of the clinical N0 neck if based on the sentinel node concept. However, the technical modalities will have to be optimized. Further, prospectively collected data should definitely be analyzed as a prerequisite. Within such an investigation, it would make sense to send the excised lymph node to frozen section examination. Depending on the histopathologic result, a surgical resection of the lymphatic drainage in the form of a selective neck dissection could then be indicated. At present, it can be stated that the technical aspects and importance of endoscopic lymphadenectomy in the neck shows scientific and clinical potential. The question about the significance of the procedure, however, can currently not be answered definitely.