Robot-assisted surgery in thyroid procedures
More than ever, our patient’s new expectations direct our surgical practice. Thus, the growing place of esthetic considerations in our society has contributed to the development of minimally invasive techniques. In thyroid surgery, several techniques have been developed to avoid cervical scars.
Initially, endoscopic procedures reduced scar size in a few selected cases. Subsequently, the scar could be hidden away in the axillary area thanks to a new robot-assisted procedure.
The advantages of this technique, successfully performed in more than 6,000 patients in South Korea between 2007 and 2011 [ , ], are now well known to thyroid surgeons. Although the learning curve is long and difficult and reproducibility remains controversial, this technique is still in permanent evolution and its future perspectives are broad, as seen from the number of publications and training programs on robot-assisted thyroid procedures [ ].
Yet, despite its high media profile, the number of centers actually performing this procedure is still small, and it is essential to continue assessment, especially in matters of safety for the patient and carcinologic success.
This paper reviews the available data on robot-assisted surgery procedures.
Robot-assisted surgery uses the da Vinci® System (Intuitive Surgical, Mountain View, CA, USA). The reference technique was described by Kang et al. in 2007.
The procedure is performed under general anesthesia, with the patient in dorsal decubitus, and the neck held in hyperextension by a silicon roll placed below the shoulders [ ].
Then the ipsilateral arm is placed in abduction, with elbow flexion of 90° in the frontal plane, and the wrist in pronation so as to position the ipsilateral hand above the patient’s head. The operative site includes the axillary, prepectoral and anterior cervical regions. Incision is performed along the medial axillary line (6 to 8 cm). The procedure is executed in three steps: creation of the work space and access to the thyroid lodge; adjusting the settings of the da Vinci® device; and thyroidectomy itself.
The first step begins with dissection between subcutaneous tissue and pectoral muscle, then between the clavicular and sternal head of the sternocleidomastoid muscle, below the hyoid muscles and down to the thyroid gland.
The gland is then released down to the sternal manubrium. A retractor is inserted to keep the work space open.
In the second step, the four arms of the robot are positioned:
three through the axillary incision (two ultrasonic scissors, one 30° endoscope, and one Maryland® forceps);
one through a second incision performed into the mammary furrow (ProGrasp® forceps). This instrument can also be introduced through the axillary incision, but robot arm mobility is then more limited.
Thyroidectomy is the final step. The surgeon takes position near the console and the assistant near the patient to manage aspiration of blood and smoke, retract tissue and provide compression if necessary.
Dissection is performed, as in open surgery, as close as possible to the thyroid parenchyma in order to conserve parathyroid vascularization. We begin with the superior pole, then the lateral part of the gland (middle thyroid vein), and end with the inferior part of the lobe. Then the lobe is reclined medially to facilitate dissection and identification of the inferior laryngeal nerve before isthmectomy.
Hemostasis is carefully checked, with abundant serum irrigation.
A drain is fitted, and closure performed in two planes.
Indications and contraindications
Before any surgery, it is essential to schedule explorations, including meticulous clinical examination and paraclinical tests. These explorations form the classical work-up in nodular or cancerous pathology (TSH and calcitonin assay, thyroid and neck nodules, ultrasound, fine-needle aspiration).
Selection of patients is a key element in robot-assisted thyroid surgery. It seems indispensable to choose the most suitable procedure for each patient, each pathology and also each surgeon.
Priority must be given to safety and efficacy. The literature data show that this technique is primarily indicated for lobo-isthmectomy and neck dissection (mediastinum and ipsilateral recurrent nerve area) [ , ].
Reference centers have described total thyroidectomy procedures and lateral neck dissection.
Nevertheless, the safety and reproducibility of these more complex surgeries are not clearly established, and these approaches are presently reserved to clinical trials.
There are many contraindications to robot-assisted surgery: history of cervical surgery or neck radiotherapy, thyrotoxicosis, thyroiditis, cancerous nodes or lesions in the posterior parenchyma (especially in the esotracheal angle, with high risk of trachea, esophagus or inferior laryngeal nerve lesions), suspicion of adjacent structure invasion (jugular vein, nerve trunk), and metastatic lesions.
Limited cervical mobility (e.g. osteoarthritis) and high body mass index (BMI) are relative contraindications.
Robot-assisted surgery is, on the other hand, indicated in the diagnosis and treatment of small nodular dystrophy (< 4 cm), and differentiated follicular carcinoma with low risk of recurrence (T1-T2, NO, MO).
Oncologic aspects and safety in robot-assisted surgery
The complexity and sophistication of robot-assisted surgery probably explain the increasing number of publications [ , ].
The first robot-assisted thyroidectomy was performed in 2007, and more than 100 similar surgeries were executed by a single surgeon between 2007 and 2009 [ ]. Since the princeps series, the feasibility and intraoperative safety of the technique have been investigated in series comprising 200, 338 and 1,000 patients [ ]. The first multicenter trial (4 centers), in 2011, assessed 1,043 patients with-low grade differentiated thyroid carcinoma [ ], and found “major” postoperative complication rates of 0.8 to 1% [ ], close to those reported for classical open thyroidectomy in centers of excellence.
These initial trials suggested that complication rates were similar whatever the technique (open, robot-assisted or endoscopic), and that robot-assisted surgery was feasible and a relatively safe.
However, although many studies assessed major complications, only a few included common thyroidectomy complications such as hypoparathyroidism or recurrent nerve palsy [ , ], and none reported systematic vagus nerve monitoring, which is a useful means of anticipating postoperative paresis or palsy [ ].
Rates of usual complications were significantly lower than routinely reported with open surgery: 36% transient hypoparathyroidism, 0.1% permanent hypothyroidism, 3.9% transient recurrent nerve palsy and 0.5% permanent palsy); however, the question of possible under-diagnosis arises, as control laryngoscopy, which is the only objective means of confirming palsy, was implemented in less than 29% of robot-assisted procedures [ , ].
On the other hand, only one trial assessed the specific complications of the transaxillary approach, i.e. axillary hematoma and brachial plexus injury [ , ]. Intraoperative monitoring was systematic in only one trial [ ], and postoperative assessment of plexus function was exceptional.
Complications unknown in routine open thyroid surgery occur in robot-assisted surgery. Brachial plexus neuropathy can result directly from the specific positioning of the patient’s arm to facilitate the surgical approach, and appears to be permanent in 0.04% of cases, with possible medico-legal implications which would justify systematic monitoring and postoperative assessment [ , ]. Respiratory and digestive tract lesions (including tracheal lesions) are a major complication hardly ever encountered in open thyroid surgery but found in 0.2% of robot-assisted procedures [ , , ].
We thus probably underestimate complication rates in robot-assisted thyroidectomy and, even after 7 years’ experience and hindsight, we are still unable to tell if complication rates are comparable with those of open surgery.
From the oncological point of view, the largest multicenter trial of robot-assisted surgery included 2,014 thyroid cancer patients [ ] and reported excellent postoperative results, with a very low rate of complications, excellent oncologic control, and also ergonomic improvement for the surgeon, a point of view that had not previously been taken into account: questionnaires completed by the surgeon revealed that less musculoskeletal problems were experienced during and after robot-assisted surgery [ ].
Qualitative evaluation of oncologic thyroid surgery is based on total resection, directly related to prognosis.
Short-term surgical success can easily be estimated from thyroglobulin blood concentrations and iodine scintigraphy [ ]. Long-term efficacy, however, can only be proved on long-term follow-up, to exclude recurrence [ , , , , , , ].
It is too soon to judge oncological results based on short-term follow-up in this relatively unaggressive disease, but it is nevertheless noteworthy that resection quality was equivalent in numerous trials comparing open and robot-assisted thyroidectomy [ , , , , , ].
Functional issues and patient satisfaction
Functional results and satisfaction rates in robot-assisted surgery have only recently been evaluated. Several studies showed that patients reported more inconveniences with open than robot-assisted thyroid surgery, and these problems mainly concerned neck scars [ , ].
Prospective studies also demonstrated the functional superiority of the robot-assisted technique in terms of neck pain and discomfort or paresthesia [ , ].
These findings are very surprising, considering the large dissection area involved in the transaxillary approach. Other studies reported that prepectoral dissection in robot-assisted surgery induced pain and resistant paresthesia, which should even be counted as a specific complication of the procedure. Trials comparing objective and subjective vocal impairment reported contradictory outcomes [ , ]. Prospective studies with adequate numbers of patients are needed to compare the vocal impact of the two techniques.
Operating time and learning curve
Many studies have compared the operating time required for the two techniques, and concluded that robot-assisted surgery was significantly longer than the classical procedure, although duration was the same as for endoscopic surgery [ ]:
robot-assisted lobectomy: 136.63 ± 38.12 minutes;
robot-assisted total thyroidectomy: 146.12 ± 25.72 minutes.
Robot-assistance requires three steps: creating the work space, installing the robot arm, and surgery itself [ , , , , ].
Unlike in digestive surgery, there is no preexisting work space in the neck, and a dissection step is required to reach the actual operative area. This is the step that mainly extends total surgery time. However, many multicenter studies reported that operating time decreases after 35 to 40 robot-assisted procedures [ ], as the surgeon’s familiarity with the technique increases.
Some teams pointed out that, during this learning period, the patient is exposed to greater risk of certain complications, particularly concerning vocal prognosis and bleeding. It has been shown that a minimum number of procedures per surgeon per year are required to maintain sufficient expertise to guarantee good functional outcome and patient safety [ ]. In view of the small number of indications, it is very likely that only teams who, like the pioneering South Korean teams, practice these procedures intensively, with at least 500 thyroidectomies per year, are able to implement this technique satisfactorily. This probably explains why robot-assisted thyroidectomy is not widespread, despite its high media profile.
Although encouraging results have been reported, no randomized multicenter studies have been conducted, and it is not possible to clearly assess the performance of robot-assisted surgery. It seems clear that the esthetic benefit is real, but this should not mask issues of patient safety or resection quality.
Oncologically, follow-up has been insufficient to explore long-term efficacy in this innovative technique.
Based on the available evidence, choice of technique should take account of the patient’s expectations, comorbidities, morphology and staging, and also the surgeon’s and team’s experience.
As pointed out by Pr. Henning Dralle [ ], this new procedure raises many ethical questions. Should we adopt a new technique just because it improves the esthetic aspect? Is it justified to invest so much money in a new procedure that has not proved its superiority to the present gold standard? Should we already implement this technique on patients with more aggressive disease, with no scientific proof of safety and efficacy? Are we willing to expose our patients to an increased rate of complications until our level of knowledge and experience catches up?
This new technology could nevertheless find new applications, for example, in surgery teaching programs, with modeling sets. Real-life situations could be set up, enabling robot-assistance to overcome its own failures, by developing training programs and enhancing the learning curve.
Nevertheless, an ethical debate remains: can a measured risk for the patient be accepted in the name of scientific progress?
Robot-assisted thyroid surgery using the da Vinci® system offers an alternative to conventional open surgery for both surgeon and patient. In view of the rate of progress in robot systems, with miniaturization and increased complexity, and cost reductions, it is highly probable that in the near future robot-assisted surgery will find its place in our daily practice.
Robot-assisted neck dissection in thyroid and head and neck squamous cell carcinoma
Since the advent of surgical robotics in the field of surgery, transoral robotic surgery (TORS) has been perceived and adopted by many surgeons worldwide as a promising means of resecting primary tumors of the head and neck region. Under strict and appropriate clinical indications, this so-called “minimally invasive surgery” is coming to be regarded as an alternative to conventional radical open surgery. There is a general consensus that postoperative morbidity has been dramatically reduced using TORS [ , ]. Not much has changed, however, in terms of the surgical method of neck dissection. When neck surgery is indicated, conventional neck dissection utilizing a transcervical incision is still the mainstay. The patient is left with a gross neck blemish, even when treated by “minimally invasive” TORS. The patient will not only be dissatisfied esthetically, but will also be more prone to specific postoperative complications such as cervical lymphedema, cervical fibrotic band formation or incision wound dehiscence. Therefore, it can be considered that this is not truly “minimally invasive” head and neck surgery.
Consequently, the authors have been actively engaged in the investigation of minimally invasive surgical methods of neck dissection. Robotic neck dissection was first reported, in Korea, by Kang et al. [ ], who presented their technique of robotic selective neck dissection of levels IIA, III, IV and VB in thyroid papillary carcinoma with lateral neck node metastasis. They used the same transaxillary (TA) port to perform the total thyroidectomy; however, with this approach, the superior cervical level I and posterior levels IIb and Va were difficult to access because of the limited axis of view and instrumentation restrictions. Comprehensive neck dissection is commonly required in head and neck carcinoma with positive cervical metastasis, and we therefore developed an alternative route, a novel retroauricular (RA) approach, providing easier access to the superior and posterior cervical levels. Though this is not “minimally invasive surgery” in the strict sense but rather “remote access surgery”, we strongly believe that many of the postoperative clinical problems mentioned above can be solved by this neck dissection approach, which can therefore be viewed as minimally invasive surgery in the broad sense. Since 2010, we have developed and performed robotic neck dissection for patients with cN + neck and cN0 neck in head and neck cancer [ ].
Advantages and disadvantages of various surgical methods of neck dissection
Conventional open neck dissection via a transcervical incision offers the surgeon the best surgical view, since a direct approach is possible. Thus, in case of overt extensive extracapsular invasion of metastatic nodes, the best option is open neck dissection. Although there are currently many types of neck dissection which have refined classical radical neck dissection to spare vital structures and certain neck levels, minimizing significant forms of surgical morbidity, the resulting neck scar remains unavoidable. Despite efforts to minimize scarring, some patients may still suffer from disfiguring hypertrophic or keloid scar, which can adversely affect quality of life. Fibrotic bands may develop on the surgically treated neck, sometimes with trifurcation incision wound and skin dehiscence due to marginal necrosis, which may lead to drastic complications such as carotid artery exposure. Lymphedema is also a frequent complaint after conventional neck dissection. The long transverse incision in the neck can disrupt normal lymphatic drainage from the head to the neck area, inducing facial or neck edema.
The authors have personal experience of robotic neck dissection via a transaxillary-retroauricular (TARA) approach, performed in our initial period of robotic neck dissection [ ]. However, we came to realize that, although cosmetically superior for the patient, since the surgical wounds are “hidden” in areas not normally visible, the surgical access gained from the TA port alone is restricted, especially in levels I, IIb and Va. Therefore, an extra remote-access RA port was considered vital to complete neck dissection.
Our accumulated experience of robotic neck dissection showed that the procedure could confidently be performed through the RA port alone: all levels of the ipsilateral neck could be reached and, by eliminating the TA approach, the dissection area was greatly reduced. Moreover, the dissection area required for the RA approach is much smaller than for the TA approach, as shown by Terris et al. [ , ]. Another strong advantage of the RA approach is that the surgical anatomy will be much more familiar for the head and neck surgeon and, unlike the TA approach, all risk of intraoperative complication due to surgical procedures related to the obstructing clavicle are completely eliminated. However, this cosmetically excellent procedure has some drawbacks. Firstly, overall neck-dissection time is longer than with the conventional method, mainly due to additional skin flap elevation and work-space creation. Secondly, the technique requires considerable technical proficiency, and is not easy for surgeons lacking experience, especially in endoscopic surgery [ ]. Finally, the total cost of this robotic surgery is greater than conventional neck dissection.
Indications and contraindications
Indications for RA robotic neck dissection are as follows:
biopsy-proven head and neck cancer requiring elective (cN0) or therapeutic neck dissection (cN +);
no previous treatment for head and neck cancer;
cN0 neck or cN + neck with minimal extracapsular spread of metastatic nodes on preoperative imaging.
refusal of the procedure;
primary definitive chemoradiation due to refusal of surgery;
salvage neck surgery;
inoperable state due to remote metastasis;
unresectable metastatic cervical nodes with overt extracapsular spread;
advanced nodal stage > N2;
cervical skin incision required to resect advanced primary tumor;
free-flap reconstruction required, although the authors have recently started to conduct successful simultaneous robotic neck dissection and free-flap reconstruction via a single RA approach, so that a need for free flap reconstruction is now only a relative contraindication [ ];
history of neck surgery of any kind;
short obese neck: the height and circumference of the neck are important factors for good exposure for robotic neck dissection; a long slender neck provides optimal exposure for work-space creation, although the procedure is equally feasible in shorter, thicker necks, and individual somatometry is not an absolute contraindication.
Retractors for skin flap elevation:
right-angle breast retractor;
self-retaining retractor (Sangdosa Inc., Seoul, Korea).
Instruments for dissection under direct vision:
monopolar cautery tips of variable length (spatula tip is preferred);
harmonic curved shears (Harmonic Ace 23E®; Johnson & Johnson Medical, Cincinnati, OH, USA);
DeBakey forceps or Russian forceps;
Yankauer suction catheter.
Robotic instruments (da Vinci® Robotic System, Intuitive Surgical Inc., Sunnyvale, CA, USA):
12 mm, 30° face-down endoscope (Intuitive Surgical Inc., Sunnyvale, CA, USA);
5 mm, Maryland® forceps (Intuitive Surgical Inc., Sunnyvale, CA, USA);
5 mm, Harmonic® curved shears (Intuitive Surgical Inc., Sunnyvale, CA, USA);
8 mm, ProGrasp® forceps (Intuitive Surgical Inc., Sunnyvale, CA).
Vessel ligation system:
Hem-o-lok® Ligation System (Teleflex Inc., Research Triangle Park, NC, USA).
Generally, the robotic neck dissection procedure can be divided into two main parts: gross dissection under direct vision, and robotic dissection. Most of the procedure can be conducted under direct vision and the robot merely aids the procedure in dissecting areas difficult to access under the naked eye. Thus, strictly speaking, the correct name for this procedure is “robot-assisted neck dissection” (RAND) rather than “robotic neck dissection”. The procedure differs depending on whether levels IV and V are included or not, and will therefore be discussed and illustrated according to two distinct types of neck dissection. For all types of RAND, the patient is positioned supine with the head rotated to the side contralateral to dissection. The neck is generally relaxed in its natural position. There is no need for neck extension with a shoulder roll.
Selective neck dissection (levels I–III)
Skin incision design
An RA or modified facelift (MFL) incision is used, depending on the type and extent of surgery. The RA incision is made at the RA sulcus and curved occipitally at its midpoint level to angulate smoothly down 0.5 cm inside and along the hairline ( Figure 8.1A ). The MFL incision is similar except that it has an additional preauricular limb. Using the MFL incision, dissection can be much easier, since the work-space created by skin-flap elevation is larger.
After skin incision, a subplatysmal skin flap is elevated just above the sternocleidomastoid (SCM) muscle, using a monopolar electrocautery under direct vision. The greater auricular nerve and external jugular vein are located superficially to the SCM muscle ( Figure 8.1B ). The skin flap is elevated until the anterior part reaches the midline of the anterior neck, the superior part the inferior border of the mandible and the inferior part the level of the omohyoid muscle. Skin flap elevation below the mandible should be performed carefully to minimize injury to the nearby marginal branch of the facial nerve. Normally, two assistant surgeons are required to comfortably lift the flap using an Army-Navy retractor or a right-angle breast retractor. After obtaining a sufficient work-space, a self-retaining retractor (Sangdosa, Seoul, South Korea) is applied through the space and is secured ( Figure 8.1B ).
Upper neck dissection under direct vision
Before the robotic dissection procedure, neck dissection is performed under direct vision using the conventional technique. First of all, the marginal branch of the facial nerve is identified, referring to the distal portion of facial artery and facial vein as landmarks, which are located along the inferior border of mandible. The perifacial lymph nodes surrounding the marginal nerve and the nearby lymphatic-adipose tissue are cautiously dissected, preserving the facial nerve. Next, the parotid tail is divided from its surrounding tissue and dissection is continued anteriorly, along the inferior border of the submandibular gland (SMG), revealing the posterior belly of the digastric muscle underneath. The anterior border of the SCM muscle is opened and followed inferiorly down to the omohyoid muscle. Dissecting the area between the inferior border of the digastric muscle and the anterior border of the SCM enables visualization of the internal jugular vein (IJV) contour. The transverse process of the atlas can be palpated in this region, and is a reliable landmark to locate the spinal accessory nerve (SAN). The SAN is identified and skeletonized, and level IIb is dissected ( Figure 8.2 ). The specimen is directed anteriorly toward the carotid sheath, and the lymphatic-adipose tissues of levels IIa and III are sequentially dissected.
Robot-assisted neck dissection (RAND) technique
The robotic arms are docked through the RA port. Three robotic arms are used; a 30° dual-channel endoscope (Intuitive Surgical) is placed at the center and Harmonic® curved shears (Intuitive Surgical) or a 5-mm monopolar spatula coagulator and 5-mm Maryland® dissector (Intuitive Surgical) are placed on either side of the central arm ( Figure 8.3 ). Additional aid from a fourth arm such as ProGrasp® forceps is not applicable using this approach, but endoscopic alligator forceps, held by a patient-side assistant surgeon, can be substituted.
Robotic dissection begins at the remaining undissected anterior portion of level I. Level I is dissected latero-medially. Confirming the posterior belly of the digastric muscle, the proximal facial artery is identified at the posterior border of the SMG, which is then sealed with Harmonic® curved shears or double-ligated with the Hem-o-lok® Ligation System by the assistant ( Figure 8.4 ). Handling the fibroadipose tissue with the Maryland® forceps, level I tissue containing the SMG is dissected away from its surrounding tissue. During dissection of SMG, the marginal mandibular branch of facial nerve, lingual nerve and hypoglossal nerve should be clearly identified and safely preserved. The mylohyoid muscle is revealed at the anterior border of the SMG, where the SMG ganglion and Wharton’s duct are visualized by retracting the specimen posteriorly. Each is then sealed by Harmonic® curved shears. The submental artery can be brought into view during this step and can also be ligated with the Harmonic® shears. Using the same instrument, the fibro-fatty level IA tissues between the two anterior bellies of digastric muscle on either side are dissected and detached. Following completion of level I dissection, the previously dissected levels II and III tissues are reflected medially using the Maryland® dissector and dissected from the carotid sheath using Harmonic® curved shears ( Figure 8.5 ). Dissection is then conducted inferiorly down to the superior belly of the omohyoid muscle and medially to the lateral border of the infrahyoid muscles. The superior thyroid artery and ansa cervicalis can be easily identified. After completion of dissection, the specimen is delivered through the RA port.
Modified radical neck dissection (levels I–V or II–V)
Just as in conventional modified radical neck dissection, the SCM muscle, the SAN or IJV can be resected, if necessary; however, the authors believe that resecting any of these vital structures fails to correspond to the notion of functionally and esthetically superior minimally invasive surgery, and we do all possible to preserve all structures, including the SCM muscle, SAN and IJV, in all cases ( Figures 8.6 and 8.7 ). Robot-assisted modified radical neck dissection should therefore be performed under strict indications, where there is no gross extensive metastatic lymph-node invasion.
Skin incision and flap elevation
If level I dissection is omitted, there is no need to elevate the skin flap as high as the inferior border of the mandible; however, the work-space must extend inferiorly down to the clavicle for levels IV and V to be accessible. Also, the subplatysmal skin flap should be raised more laterally with respect to the posterior border of the SCM, so that level V nodes are sufficiently addressed. This procedure can be greatly facilitated if the operator stands above the patient’s head and faces caudally.
Upper neck dissection under direct vision
Great care should be taken to preserve the SCM muscle, SAN and IJV whenever robot-assisted modified radical neck dissection is performed. For comprehensive neck dissection, the marginal branch of the facial nerve is identified first, as described above for selective neck dissection of levels I–III. In levels II–V, the initial dissection is made along the inferior border of the SMG and the tail of parotid gland, beneath which the posterior belly of the digastric muscle is identified. Dissecting along the posterior belly of the digastric muscle and retracting it superiorly enables visualization of the IJV. The SAN is then identified near the IJV after palpating and checking the location of the transverse process of the atlas. Next, the fascia of the anterior border of the SCM is opened and the muscle is followed down to the medial border and dissection is continued inferiorly as far as possible. The entire course of the SAN is skeletonized, from its exit near the skull base to its entry at the trapezius muscle. The operator then opens the fascia along the posterior border of the SCM muscle, where it is then lifted using an Army-Navy retractor so that the fibro-fatty tissue inside and under the SCM muscle is dissected. The fibroadipose tissue as a whole is dissected off the SCM muscle during this procedure, taking care not to damage the SAN near its entry in the trapezius muscle. The SCM muscle is maintained in its retracted position using the self-retaining retractor, to dissect levels IIB, V, the lateral aspect to the carotid sheath of IIA, and upper III, under direct vision. A considerable portion of the upper neck can be dissected under direct vision, including levels IIA and III medial to the carotid sheath. The hypoglossal nerve is identified and preserved near the carotid bifurcation area, and the superior thyroid and lingual artery are identified and preserved. The final step consists primarily in removing the dissected specimen so that a clear surgical view is obtained for robotic dissection. In releasing the specimen for removal, care should be taken not to violate or transect any nearby lymph nodes. Making sure the SCM muscle is securely elevated, the robotic arms are now introduced.
Robot-assisted neck dissection (RAND) technique
If level I dissection is included in the operation, robotic surgery begins at level I. When docking the robotic arms, the axis should be parallel to the inferior border of the mandible. The specific procedures for level I dissection are identical to those described above for selective neck dissection (I–III). The specimen is removed before moving on to levels IV and V, for which the robotic arms must be realigned cephalo-caudally. The previously dissected portion of level III is robotically dissected supero-inferiorly down to levels IV and Vb. The lymphatic-fatty tissue is carefully dissected from the IJV, using Harmonic® curved shears. Any branches of the IJV can be identified during carotid sheath dissection and controlled in advance with Harmonic® curved shears or the Hem-o-lok® Ligation System to minimize unnecessary bleeding in the operating field. Level V is dissected postero-anteriorly. Generally, this procedure is fairly manageable, since the work-space has been secured after elevating the skin flap and SCM muscle. The fibro-fatty tissue of level VB is dissected latero-medially, and the omohyoid muscle, which is encountered during this process, is cut using Harmonic® curved shears.
Recently, since the introduction of the da Vinci® Xi model (Intuitive Surgical Inc., Sunnyvale, CA, USA), there have been some refinements to improve functional outcome. Instead of opening up the posterior border of the SCM as above, the robotic arms can also dissect level V from the medial side of the SCM underneath, with an assistant retracting the SCM. This refined RAND technique preserves the cervical plexus, with less trauma to the SAN, and improves postoperative neck sensation.
The transverse cervical artery and vein are identified and preserved during levels VB and IV dissection. Like in the conventional procedure, the level dissected must be above the facial carpet, so that structures such as the phrenic nerve or brachial plexus are safely preserved. The specimen is then retracted superiorly and medially so as to move on to level IV. The lymphatic-adipose tissue around the carotid sheath at level IV is carefully dissected using Harmonic® curved shears. The vagus nerve, carotid artery and IJV are all identified and preserved by carefully opening up the carotid sheath. When dissecting the inferior part of level IV, the lymphatic or thoracic duct should be anticipated around the IJV. Hemoclips or the Hem-o-lok® Ligation System should be used to control the lymphatic or thoracic duct, not only when there is evident injury but also when it is undisturbed, to prevent any future chyle leakage. Surgical management of postoperative chylous fistula is often difficult, since the relevant area is reached distally from the RA port. After completion of the neck dissection, the specimen is delivered through the RA incision port ( Figure 8.8 ).