Selective Neck Dissection

8 Selective Neck Dissection

Marcelo F. Figari


In this chapter, we analyze the path followed since the beginning of the 20th century, when Crile conceived radical dissection to tackle regional spread of squamous carcinoma of the upper aerodigestive tract, until present, when selective neck dissections have become established, for elective treatment as well as for some curative indications in the neck. Based on the best available evidence, we analyze the indications and techniques of supraomohyoid, lateral, posterolateral, superselective, and extended dissections, pondering their advantages and disadvantages. We also address the role of the various sources of energy available to facilitate these procedures, as well as the potential for a progressive adoption of minimally invasive techniques today and in the near future.

Keywords: head and neck, squamous cell carcinoma, elective neck dissection, therapeutic neck dissection, selective neck dissection, minimally invasive surgery

8.1 Introduction

Neck dissection is a very valuable procedure in the battle against cancer of the upper aerodigestive tract. First performed in the beginning of the 20th century, during the 1960s, the technique evolved to less aggressive variants, which were more respectful of normal anatomy and physiology, while maintaining oncological radicality.

In this chapter, we will describe the historical evolution, the rationale of oncologic spread to preferred lymph nodes, indications and technique of selective dissections, the best available evidence, and potential future developments.

8.2 Brief History

Resection of cervical lymph nodes as part of the treatment of head and neck cancer began at the end of the 19th century and beginning of the 20th century, at a time when radiotherapy still had not become established as an ancillary or alternative treatment strategy. The original description of radical neck dissection, performed by Crile in 1906 and later made popular by Hayes Martin, required the functional sacrifice of soft tissues, vessels, and nerves, given the lack of options available at the time.1,2 However, it is worth mentioning that, even before the radical procedure was conceived, multiple procedures that aimed at resecting selected lymph node groups had been described.3 Such developments were pioneered by von Langenbeck, Kocher, von Volkmann, Butlin, and others. Butlin is also credited with the original idea of elective cervical dissection in necks without macroscopic involvement.3 It is also worth mentioning that, among the first 132 cervical procedures reported by Crile in 1906, only 36 were radical. He then reported that 3-year survival achieved with the radical procedure was 75%, compared to 19% in the case of selective procedures. Such data were key to the acceptance of the radical neck dissection procedure.3

At the beginning of the 1960s, a more conservative therapeutic attitude was adopted. The possibilities of performing function-preserving surgeries or medical treatments for laryngeal or breast cancer are good examples of such attitude. By then, many ENT (ear, nose, and throat) surgeons worldwide had addressed the concept of cervical micrometastases without clinical findings in head and neck cancer, to justify potential prophylactic radical dissections (Agra and del Sel, Buenos Aires, 1947; Ogura, Washington University, St. Louis, MO, 1952; Alonso, Montevideo, 1952). That led professor Osvaldo Suárez, an otolaryngologist from Córdoba, Argentina, to describe in 1963 the functional radical neck dissection and to coin that name for the procedure.3,4 This intervention combined the oncologic concept of complete resection of potentially involved lymph nodes with the preservation of certain structures. The proposal was to excise lymph nodes and lymphatic collectors, together with the aponeurosis and cellular tissue, while preserving the sternocleidomastoid muscle (SCM), the internal jugular vein, the accessory spinal nerve, and as many vascular and nervous structures as possible, without compromising the goal of a radical resection.

Although undoubtedly Osvaldo Suárez was the first to conceive this functional concept in radical neck dissection, his publications were in local journals and in Spanish. Other authors disseminated the technique in English-language publications, and hence they have been credited with its development. Interestingly, Bocca and Pignataro5 made popular the following statement: “Radicality must be directed against the cancer rather than against the neck.”

8.3 Classification of Neck Dissections

Based on the classification of cervical lymph node groups proposed by investigators at the Memorial Sloan Kettering Cancer Center (MSKCC),2 mentioned in another section of this chapter and depicted in image Table 8.1, in 1988 the Committee on Head and Neck Surgery and Oncology of the American Academy of Otolaryngology–Head and Neck Surgery met to establish a classification that would allow harmonizing definitions and nomenclature regarding neck dissections, thus improving communication among professionals and the reporting of strategies and results.1 The modification introduced in 2002 by the Committee for Neck Dissection Classification of the American Head and Neck Society (AHNS) is depicted in image Table 8.2.6

The first step was to define radical neck dissection as the standard procedure, considering the remaining procedures as modifications of the standard. The consensus decision was to call modified radical neck dissection those procedures that preserve some of the nonnodal structures that were excised during the classic procedure. Selective neck dissection (SND) was that in which one or more of the nodal groups that were excised during radical procedures were preserved.

Table 8.1 Cervical lymph node levels: modified Memorial Sloan Kettering Cancer Center classification1,2,6


Clinical location

Surgical boundaries


Submental triangle

S: symphysis of mandible

I: hyoid bone

A (M): left anterior belly of digastric muscle

P (L): right anterior belly of digastric muscle


Submandibular triangle

S: body of mandible

I: posterior belly of digastric muscle

A (M): anterior belly of digastric muscle

P (L): stylohyoid muscle


Upper jugular

S: lower level of bony margin of jugular fossa

I: level of lower body of hyoid bone

A (M): stylohyoid muscle

P (L): vertical plane defined by accessory nerve


Upper jugular

S: lower level of bony margin of jugular fossa

I: level of lower body of hyoid bone

A (M): vertical plane defined by accessory nerve

P (L): posterior border of sternomastoid muscle


Mid jugular

S: level of lower body of hyoid bone

I: horizontal plane along inferior border of anterior cricoid arch

A (M): lateral border of sternohyoid muscle

P (L): posterior border of sternocleidomastoid muscle or sensory branches of the cervical plexus


Lower jugular

S: horizontal plane along inferior border of anterior cricoid arch

I: clavicle

A (M): lateral border of sternohyoid muscle

P (L): posterior border of sternocleidomastoid muscle or sensory branches of the cervical plexus


Posterior triangle

S: convergence of sternocleidomastoid and trapezius muscles

I: horizontal plane along inferior border of anterior cricoid arch

A (M): posterior border of sternocleidomastoid muscle or sensory branches of the cervical plexus

P (L): anterior border of trapezius muscle


Posterior triangle (supraclavicular)

S: horizontal plane along inferior border of anterior cricoid arch

I: clavicle

A (M): posterior border of sternocleidomastoid muscle or sensory branches of the cervical plexus

P (L): anterior border of trapezius muscle


Anterior compartment

S: hyoid bone

I: sternal notch

A (M): common carotid artery

P (L): common carotid artery


Superior mediastinum

S: sternal notch

I: innominate artery

A (M): common carotid artery

P (L): common carotid artery

Abbreviations: S, superior; I, inferior; A (M), anterior (medial); P (L), posterior (lateral).

Table 8.2 Neck dissection techniques classification1,6

Type of neck dissection


Radical neck dissection (RND)

Removal of levels I–V, accessory nerve, internal jugular vein, and sternomastoid muscle

Modified radical neck dissection

Removal of levels I–V; preservation of one or more of the accessory nerve, internal jugular vein, or sternomastoid muscle (types I, II, III, respectively)

Selective neck dissection

Preservation of one or more levels of lymph nodes

Extended radical neck dissection

Removal of one or more additional lymphatic and/or nonlymphatic structure (s) relative to an RND, e.g., level VII, retropharyngeal lymph nodes, and hypoglossal nerve

Finally, extended radical neck dissection was defined as that which involves at least one additional nodal group or a nonnodal structure, compared to the radical procedure. We will review and define the various types of SNDs.

8.4 Types of Selective Neck Dissections

Four types of selective cervical dissections have been defined1:

Supraomohyoid dissection (image Fig. 8.1). This entails resecting the following groups: submental, submandibular (group I, excising the submaxillary gland), high jugular (group II), and mid jugular (group III). The deep limit of resection consists of the branches of the cervical plexus and the posterior limit is the lateral border of the SCM. Since nodes are present medially and laterally to the accessory spinal nerve, node group II is divided into IIa (anterior and inferior to the nerve) and IIb (posterosuperior to the nerve).

Lateral dissection (image Fig. 8.2). It involves the excision of levels II, III, and IV (upper, middle, and lower jugular lymph nodes).

Posterolateral dissection (image Fig. 8.3). Mostly employed in skin diseases of the occipital region, it entails removing the suboccipital and retroauricular nodes together with node levels II, III, IV, and V.

Anterior dissection (image Fig. 8.4). Mostly employed in the management of thyroid cancer, it involves the prelaryngeal (Delphian node), pretracheal and paratracheal chains (recurrent laryngeal nodes), and perithyroidal nodes in general.

Any variation different from these four types should be defined clarifying the node groups involved, for example, “selective dissection with resection of levels I and II,” as may occur in the treatment of parotid gland cancer. In these selective dissections, the internal jugular vein, the spinal accessory nerve, and SCM are routinely preserved. Should it be necessary to sacrifice one of them, the dissection must be defined by the node groups resected, listing the structure or structures that have been sacrificed.

In 1994, Spiro et al, from the group of the MSKCC, reported on their 10-year experience with 10,650 patients. Considering only the 1,143 dissections performed on 1,035 patients, not previously treated, 420 were radical dissections, 263 were modified dissections, and 460 were selective dissections. During those years, there was a clear change in practice, since from 1984 to 1988 a total of 44% of radical dissections were performed, whereas the rate for the same procedure was 27%, during the period from 1989 to 1993.2 The MSKCC group agrees in stating that thanks to Robbins et al the classification and nomenclature of dissections were established in an orderly manner. However, they suggest the name selective for those dissections involving less than four, but at least three groups of lymph nodes, reserving the name limited dissections to those involving only one or two groups of lymph nodes.2

In the United Kingdom National Multidisciplinary Guidelines addressing the management of lymph node metastases in head and neck cancer, the authors summarize the concept saying that there is a growing trend to divide dissections into two large groups: radical dissections (comprehensive, removing levels I–V) and selective dissections (excising < 5 node levels).6

8.5 Indications for Selective Neck Dissections and Reported Clinical Experiences

8.5.1 N0 Neck: Elective Dissection

Involvement of cervical lymph nodes is the single most ominous prognostic factor for carcinomas of the upper aerodigestive tract.2,6 For primary tumors with an N0 neck by clinical assessment and tomography, there are three possible therapeutic options: elective cervical dissection, elective neck radiotherapy, or clinical surveillance, leaving the surgical option for cases with development of clinical metastases.7

Patients with primary tumors that will be treated surgically and with a greater than 20% risk of having microscopic cervical node metastases are candidates for elective neck dissection. Such conduct was already standard since the 1970s, when modified radical dissection became general practice, but became fully accepted in the 1990s, when the concept of selective dissection was reinforced.7 Based on the fact that lymphatic drainage to the neck is predictable from the various primary sites, the concept of selective dissection implied not only removing the groups of lymph nodes with a higher risk of micrometastases, but also preserving others with a lower risk.3 The surgical group at MD Anderson Cancer Center was probably the first to report a positive experience with selective dissection in 1978; recurrence rates were similar in selective and radical procedures performed for cancers of the oral cavity.3

In 1997, Pitman et al published a comparative study of modified radical dissection versus selective dissection as a staging strategy and treatment in primary tumors of the upper aerodigestive tract with N0 neck. Although the study has methodological limitations because it is retrospective and assessed heterogeneous populations, it clearly provides support to selective dissection as a diagnostic and therapeutic resource. Patients included had primary tumors in the oral cavity, oropharynx, hypopharynx, and larynx, and an N0 neck staged clinically and, since 1980, by computed tomography (CT); nodes were considered metastatic when they were larger than 1 cm. A group of 288 patients treated between 1974 and 1989 (modified radical dissection) and another group of 99 patients treated between 1990 and 1994 (selective dissection) fulfilled the following eligibility criteria: no prior treatment of the primary tumor or the neck (surgery or radiation), minimum follow-up of 24 months, free margins in the resected specimen of the primary tumor, complete pathology report describing the nodes’ characteristics, surgical report describing in detail the dissected node levels, and complete report of adjuvant treatments. Conclusions of the study were that selective dissection was as effective as the radical procedure for cervical staging and that regional recurrence rates were similar for both procedures. Accuracy of selective dissection for the detection of occult metastases was 90% (superior to that of imaging techniques).7 The authors believed that further studies are required to determine the potential therapeutic value of selective dissection in N1 necks. Prospective and randomized studies that could provide an evidence base for selective procedures are scarce.

In 1998, the Brazilian Head and Neck Cancer Study Group published the first prospective randomized study on 148 patients with primary tumors of the oral cavity (T2–T4), located in the oral tongue (n = 62), floor of the mouth (n = 49), gums (n = 12), and retromolar trigone (n = 25). Patient groups were homogeneous, without previous treatment, and had negative necks. Lymph node levels most frequently involved were II and III. Local recurrence rate was similar, as was actuarial survival at 60 months (63% in the radical dissection group and 67% in the selective dissection group).8 One year later, that same group of investigators published a similar experience, but prospectively comparing modified radical dissection with lateral neck dissection (LND) for the initial treatment of supraglottic and transglottic carcinomas staged as T2–T4, N0 M0.

A total of 71 patients underwent radical dissections (13 bilateral) and 61 underwent selective lateral dissections (18 bilateral). Again, most node groups involved were levels II and III; there were four ipsilateral recurrences in the radical group and two recurrences in the selective group. Actuarial 5-year survival was 72.3% in the group with radical dissection and 62.4% in the group with lateral selective dissection.9

Given the very low current morbidity of selective dissections, today there is good reason to propose the elective treatment of the neck in patients whose risk of micrometastases is lower than 20%. Paleri et al propose it for primary tumors with a greater than 15% risk of occult metastases, which would encompass almost all primary tumors, except for glottic T1 and T2 tumors, and some selected T1 tumors of the oral cavity.6

In 2015, a group of investigators at the University of Belgrade, in Serbia, published a prospective, observational experience of cases and controls. They compared the historical experience of patients with supraglottic carcinomas treated surgically between 1988 and 1996 to patients with the same disorder treated between 1996 and 2005 (period in which they standardized performance of lateral elective dissections in N0 necks). Patients in the control group (n = 51) were operated for their primary tumor and subsequently watched and followed. Patients in the study group (n = 193) underwent surgery for their primary tumor and bilateral elective lateral dissection. Patients who recurred after selective dissection underwent a radical neck dissection. Patients with occult metastases received postoperative radiotherapy. In the study group, 18% had occult metastases (most of them at node level II) and 20% had extracapsular spread. The most relevant finding was that the control group had an 11.8% rate of lymph node recurrence, whereas that rate was 4.15% in the group that underwent selective dissection. Overall 5-year survival was not significantly different between both groups. The study concluded that elective and selective treatment of the neck was useful in primary supraglottic tumors to reduce regional recurrence rates.10

8.5.2 Selective Dissections in Melanoma

The role of selective dissection in the elective treatment of an N0 neck in cases of melanoma of intermediate thickness has progressively been replaced by sentinel node biopsy, in spite of the greater hurdles that this procedure entails in cases of head and neck cancer compared to other regions of the body. However, when the sentinel node biopsy cannot be performed or is nondiagnostic, melanomas of intermediate thickness, after excising the margins of the primary tumor, should undergo SND according to the following recommendations based on the type of drainage.11

Melanomas of the anterior region of the scalp, forehead, face, and pinna: selective dissection of levels I to IV. The actual value of resecting level IV has been questioned. Additionally, depending on the location, several authors have recommended performing a superficial parotidectomy.

Melanomas of the scalp posterior to the coronal line could be amenable to a posterolateral dissection, including the occipital nodes and cervical levels II to V.

Undoubtedly, selective dissections offer a great opportunity to detect occult metastases with a minimum of morbidity, thus allowing the surgeons to offer patients adjuvant therapeutic options at an early stage of treatment. Also, detection of extracapsular spread in these occult metastases is a good predictor in high-risk patients.7

8.5.3 N + Neck: Therapeutic Dissection

Although selective cervical dissection was devised with the rationale of minimizing morbidity to electively treat negative necks at risk, many groups have provided sufficient evidence showing that selective dissection may be curative in cases of positive necks that are not advanced.3 In 2002, with the goal of assessing the role of the sentinel node, Andersen, Warren, and Spiro evaluated the oncological adequacy of selective dissection in a group of 106 patients. Although 54.7% of patients were N1, 4.7% were N2a, 26.4% were N2b, 13.2% were N2c, and 0.9% were N3, 34% had extracapsular spread and 71.7% of patients required postoperative radiotherapy; regional control achieved with selective dissection was 94.3%. Clearly, extracapsular spread was the most negative factor with regard to recurrence and disease-related survival.3

In 2010, Weinstein et al, from Pennsylvania, published their experience with oropharyngeal carcinomas and a positive neck, treated with a combination of TORS (transoral robotic surgery) and SND, and subsequently followed and treated with radiotherapy or chemotherapy. The study was a prospective, phase I trial. The investigation was designed to assess the possibility of controlling cervical disease with selective dissection and adjuvant therapies, since

Cases were oropharyngeal carcinomas treated with TORS with free margins; hence, the potential indication of chemotherapy or radiotherapy would be due to cervical disease.

In the selective dissection, levels IIa, IIb, III, and IV were excised.

Since the study was prospective, with a single population arm, the decision of follow-up or adjuvant therapy was made at the beginning of treatment.

The primary outcome of the study was regional control and the secondary outcome was the assessment of complications. In 31 patients (29 men), 33 SNDs were performed (2 were bilateral). Primary oropharyngeal tumors treated with TORS: T1 in 29%, T2 in 48.4%, and T3 in 22.6% (as mentioned, all were resected with free margins). Results that attest to the usefulness of selective dissection show that staging increased by 33% in N0 necks and by 43% in N1 necks; extracapsular invasion could be detected in 70% of N2b necks and staging could be decreased to N0 in 4 of 14 N1 necks. Regarding adjuvant therapy, seven patients did not receive radiotherapy (four with N0, two with N1, and one with N2b). As to the remaining patients (24 patients), 50% received only adjuvant radiotherapy and another 50% received adjuvant chemoradiotherapy. Minimum follow-up was 3 years; only one patient had a contralateral recurrence and one patient had distant disease and was alive at the time of study closure. Investigators emphasized the high rate of local and regional control and the possibility of using selective dissection as an instrument to adapt and eventually decrease the intensity of adjuvant therapy.12

Another recent study of interest is that of Young Soo Rho et al, from Hallym University in South Korea. This was a prospective study in 44 patients with primary carcinomas of the oral cavity. Twenty-nine patients underwent SND (levels I–III: 12 patients; levels I–IV: 6 patients; and levels II–IV: 11 patients), and in 15 patients, due to the intraoperative finding of lymphatic metastases, the procedure was converted to a modified radical dissection (involving levels I–V). Extracapsular invasion was found in 8 cases of the selective group and in 6 cases of the radical group. In 20 cases of the selective group and in 11 cases of the radical group, adjuvant therapy was administered (chemoradiotherapy for stages T3 and T4, insufficient margins, extracapsular nodal or multiple nodal metastases, and radiotherapy only for early stages with limited extracapsular spread). There was no significant difference among the two groups in locoregional control, nor in overall or disease-related survival.13 These findings show that in selected cases a selective dissection may be enough to achieve effective regional control, as well as to help decide the type of adjuvant therapy needed to achieve such control.

8.5.4 SND in Patients with HPV + Squamous Oropharyngeal Carcinoma

A recent multicenter study (Washington University – Mayo Clinic), although retrospective, has shed light on the value of selective dissection for the treatment of N + necks in human papillomavirus–positive (HPV +) patients. All patients treated were included, according to the following criteria: carcinomas of the oropharynx, treated transorally, p16 +, neck staged as N1 to N3, and selective dissection (levels II–IV ± levels I and V, depending on the case). On occasions, the dissection had to be extended to nonnodal vital structures (spinal accessory nerve in 7% of cases, internal jugular vein in 13%, and SCM in 8% of cases). A total of 324 patients were followed for a mean of 49 months, and 83% of them (270 patients) received adjuvant radiotherapy. Regional recurrences occurred in 4% of patients and distant metastases in 6%, with a regional control rate after rescue of 98%. On univariate analysis, the absence of radiotherapy was associated with recurrence. On multivariate analysis, the use of adjuvant radiotherapy had a positive impact on disease-free survival (DFS), but was not significant in the case of overall survival (OS) or disease-specific survival (DSS). Five-year survival rates by Kaplan–Meier’s method for OS, DSS, and DFS were 88% (95% confidence interval [CI]: 84–92%), 93% (95% CI: 89–96%), and 83% (95% CI: 78–87%), respectively. These results support the notion that in cases of HPV + oropharyngeal carcinomas with clinical evidence of neck node metastases, selective dissection including levels II, III, and IV (occasionally extended to additional involved tissue), together with adjuvant radiotherapy, offers excellent regional control.4

8.5.5 Selective Dissection as Adjuvant Therapy

In addition to its role in the primary treatment of head and neck cancer, during this last decade its role has increasingly been explored as adjuvant therapy for tumors initially treated with nonsurgical therapies, mainly chemoradiotherapy.3 The classic approach in patients initially treated with chemoradiotherapy for head and neck cancer was to perform, after treatment, a scheduled dissection (either radical or modified radical) in all patients whose initial stage was N2 or N3. Robbins, back in 1990, already postulated the use of selective dissections in patients treated with organ-preserving protocols. In a population of 171 patients, of which 130 had an N2 or N3 neck before beginning treatment, Robbins analyzed 84 patients who underwent neck dissections after chemoradiotherapy. Of the 106 dissections, 92 were selective. Of note, the regional recurrence rate for radical procedures was 16%, whereas it was 4% for selective procedures. This could be partly attributed to a selection bias in this markedly heterogeneous population, but clearly, the low recurrence rate in selective procedures showed a change in therapeutic decision-making.

A more recent study also supports the role of selective dissections after chemoradiotherapy to treat persistent nodal disease. A population of 62 patients treated with chemoradiotherapy underwent 69 selective dissections, a mean period of 10 weeks after completing therapy. Residual tumor was found in 46% of the specimens. Mean follow-up was 33 months, and 65% of patients were disease free. Among the 22 patients who recurred, only 4 had a regional recurrence and of them, 3 recurred in the contralateral neck. Therefore, using selective dissection as a complement to nonsurgical treatment, only 1 of 62 patients had an ipsilateral regional neck recurrence.14

In recent years, the use of positron emission tomography–computed tomography (PET-CT) has become an established tool for decision-making in potential residual neck disease, once chemoradiotherapy has been completed. According to the United Kingdom National Multidisciplinary Guidelines, if at 10 to 12 weeks after completing treatment there is no evidence of disease by PET-CT, the patient does not require elective treatment of the neck. Dissection should be considered in patients in whom the evidence of disease by PET-CT is inconclusive or questionable.6

8.5.6 Superselective Dissection

The definition of superselective neck dissection (SSND) involves the removal of lymph nodes of two adjacent neck levels. In the cases of dissection performed as part of a primary treatment, the most typical example is that of supraglottic carcinomas, which rarely involve microscopically levels other than IIa and III. In a large retrospective series published by Ambrosch et al, superselective dissection was performed in almost all laryngeal carcinomas treated with laser, electively and with very good local control rates. The finding of metastases in those levels generally results in the subsequent use of adjuvant radiotherapy.15 Superselective dissection also has a role in the treatment of residual node disease after chemoradiotherapy, since it is well known that in these cases persistence of nodes is usually limited to one or two levels.

In 2010, Goguen et al, from the Dana Farber Institute, proposed the use of superselective dissection in patients with residual nodes after chemoradiotherapy, when the disease was confined to a single node level. The authors reported that SSND was useful in 51 of 55 patients with disease limited to one node level and in 61 of 67 patients with disease limited to two levels.3 In 2012, Robbins et al published a retrospective series from two institutions, analyzing 35 SSNDs performed in 30 patients. In 23 cases, the SSND had been planned and in 12 cases it was performed as a salvage procedure. Mean follow-up was 33 months (range: 8–72 months); 3 patients had a recurrence of their primary tumor and 5 patients had distant metastases, but none had a neck recurrence. The group had a 60% 5-year DSS.16

To date, there are no prospective studies that can provide evidence for the use of superselective dissection in the case of residual nodes after implementing organ-preserving protocols. However, the advantage of a lower morbidity seems justified by the acceptable regional control rates.

8.5.7 Extended Selective Dissections

The concept of extended selective neck dissections (ESND) applies to cases in which, in addition to the nodes of one or two neck levels, the disease extends to other nonnodal contiguous structures (sternocleidomastoid or prethyroid muscles, spinal accessory nerve, internal jugular vein). In the study by Dhiwakar et al, 39 patients received 43 of these procedures; 18 were performed as part of the primary treatment and 25 as a surgical salvage following chemoradiotherapy. Even in this case of extended procedures, the regional recurrence rate observed in the first group was 0% and DFS at 5 years was 40%. In the second group, the result was quite different; extended selective dissection was performed as a rescue procedure, in which neck recurrence was 40% and DFS decreased to 30%.14

8.6 Contribution of the Sentinel Lymph Node Biopsy

The concept of sentinel node states that the tumor will spread from the primary site to one or several first echelon lymph nodes, of which the selective biopsy will allow predicting the status of the remaining nodes. The sentinel lymph node biopsy (SLNB) is well established for the treatment of melanoma and breast cancer; the concept is attractive in the case of head and neck cancer, for the possibility of assessing, with minimum morbidity, which patients may benefit with the elective treatment of the neck. Undoubtedly, the sentinel node biopsy is a reliable procedure.

The American College of Surgeons (ACS) conducted a prospective multicenter trial involving 25 institutions during a 3-year period. They studied a total of 140 patients with T1 and T2 tumors of the oral cavity (95: oral tongue; 26: floor of the mouth; and 19: other locations). Among the 106 SLNB that were clinically and pathologically (hematoxylin and eosin) negative, no other positive nodes were found; hence, negative predictive value (NPV) was 94%. With additional sections and immunohistochemistry techniques, NPV improved to 96%. For T1 lesions of the oral cavity, metastases were correctly identified in 100% of cases.17

In 2015, Schilling et al reported the results of the EORTC 24021 protocol (Sentinel European Node Trial [SENT]) on the study of the sentinel node in oral cancer. Fourteen European centers prospectively recruited 415 patients with cancer of the oral cavity staged as T1 and T2 N0. They found an average of 3.2 sentinel nodes per patient. If the sentinel node was positive, a cervical dissection was performed 3 weeks later. Duration of follow-up was 3 years. The sentinel node was located in 99.5% of cases, and 23% of them were found to be positive for metastases. There were 14% of false-negative cases, which subsequently required rescue. The procedure had a sensitivity of 86% and an NPV of 95%. DSS was 94%.18

The studies described, as well as many others, show that sentinel node biopsy is a reliable procedure for the staging of initial oral cancer. However, there is still controversy regarding the need to subject patients to a second procedure when a positive result is found. Perhaps, optimization of the procedure, with the addition of fast, safe, and standardized techniques that will allow analysis of the sentinel node in the operating room will contribute to its generalized acceptance for this group of patients in the future.17

8.7 Surgical Strategy According to Node Groups at Risk

The rationale of selective dissection is based on the fact that there are predictable patterns of lymphatic spread from the various mucosal sites of the upper aerodigestive tract. Many publications in the medical literature refer to this, although the first description was made by Lindberg, a radiotherapist, and later revisited by Byers, who in 1985 confirmed such patterns in the dissection specimens of 967 patients. The observation was relevant, but not perfect as a proof, since many of Byers’ dissections were selective and hence not all the node groups were available for analysis.3

In 1990, Shah et al published the findings of a pathology study of 192 radical dissections performed electively due to oral cavity cancers. They confirmed very relevant data: the groups that were usually affected when the primary tumor was in the oral tongue, floor of the mouth, gums, retromolar trigone, or the buccal mucosa were levels I, II, and III. Group V was never affected, except in 2% of gum carcinomas. Group IV was considered to be very low risk, since it was involved in 0% of primary tumors of the buccal mucosa and in 6% of cases of the retromolar trigone.19 Another important finding, in Byers’ as well as in Shah’s studies, is that primary tumors located in the visceral structures of the midline (oropharynx, hypopharynx, and larynx) generally do not involve level I.

In 2014, Pantvaidya et al,20 from the Tata Memorial Center in Mumbai, India, published a prospective study on the dissection specimens obtained from 470 patients with primary tumors of the oral cavity. At that institution, patients with stage N0 undergo SND (levels I–III or I–IV) and patients with N + undergo radical dissection (levels I–V), either the classic or modified procedure, depending on the presentation. Dissection specimens obtained from patients with stage N3 or patients who had received prior treatment of the neck were excluded from the study. The goal was to study the frequency of involvement of the various node levels, and to assess the involvement of levels IIb and V, aiming at their potential preservation. The authors verified that 95.7% of metastases were located in levels I to IV. With the exception of lip tumors, cases in which the greatest involvement was level Ia, for the remaining sites of the oral cavity, levels Ib and IIa were the most frequently involved.

Level IV was affected from 0% (lip ad hard palate) to 9% (floor of the mouth). On average, involvement of levels IIb and V was 3.8 and 3.3%, respectively. Multivariate analysis showed that involvement of level IIa is an independent predictive factor of the involvement of levels IIb and V; hence, it suggests the need for their excision. But the general conclusion is that in most cases, SND of levels I to IV may be adequate for treatment of the neck in oral cancer.20

Another study from India, a country where oral cancer is the most frequent in its population, was conducted by Chheda et al, from Ahmedabad, and published in 2017. In a population of 210 patients with oral cancer (more than 70% originated in the tongue), operated between 2010 and 2012, and excluding patients treated with neoadjuvant chemotherapy, they prospectively studied specimens from 120 modified radical dissections, 40 extended supraomohyoid dissections, and 50 supraomohyoid classic dissections. Of the 210 dissections, only 2 (0.95%) showed involvement of level IIb, and in such cases, level IIa was also involved. There was no evidence of isolated involvement of level IIb. The authors suggested performing frozen biopsy of level IIa to determine the need for excision of level IIb, since sensitivity of the clinical judgment is extremely low. The goal was to minimize the aggression of the spinal accessory nerve.21

Another important prospective study, also from India, is that of Agarwal et al, which was published in 2018. From 2011, they studied the dissection specimens of 231 patients with primary carcinomas of the oral cavity, and a neck staged clinically and with imaging (magnetic resonance and CT) as N0 in the preoperative assessment. The most frequent primary sites were the oral tongue and the buccal mucosa. Of the 231 cases, 71 showed microscopic metastases (30.73%). Again, levels affected most often were Ib and IIa. Level IIb was involved in 0.86% of cases and no patient had metastases in level IV. As in the study by Chheda et al, whenever there was involvement of level IIb, level IIa was also involved. The authors concluded that classic supraomohyoid dissection (levels I–III) is sufficient to treat patients with a primary tumor of the oral cavity and a neck staged as N0, and that level IIb should only be excised when there is proven involvement of level IIa.22

Another interesting study is the one published in 2016 by Agarwal et al, in the same population previously studied, but looking at the involvement of another node group, that of perifacial nodes. Their dissection and excision also implies the need to selectively dissect the marginal mandibular branch of the facial nerve, in order to preserve it. Agarwal et al found perifacial involvement in 8.22% of the dissections performed in the 231 patients previously mentioned. Primary sites most frequently affected (oral tongue and buccal mucosa) had similar involvement of the perifacial group (7.14 and 7.75%, respectively); hence, the authors recommend including them in N0 neck dissections for tumors in those locations, especially in cases of patients younger than 25 years, with an advanced T stage or infiltration greater than 5 mm.23

8.8 Technical Aspects of Selective Neck Dissection

In this section, we describe the removal of node compartments that are part of the following selective dissections:

Supraomohyoid neck dissection (SOHND): levels I, II, and III (image Fig. 8.1).

LND: levels II, III, and IV (image Fig. 8.2).

Posterolateral dissection: occipital level and levels II, III, IV, and V (image Fig. 8.3).

Indications are described in the appropriate sections. We will not discuss central neck dissection nor modified radical dissection, since both procedures are described elsewhere.

8.8.1 Perioperative Aspects

We shall not dwell on the general organizational aspects of head and neck surgical procedures, but we will underscore some issues relevant to SNDs. Preoperative informed consent should include the type of surgery, the functional preservation intent of the procedure, and also the potential functional sacrifices that could derive from changes in surgical conduct due to intraoperative findings. The preoperative history and physical examination should include information on past therapies that could influence certain steps of the dissection, such as neck biopsies, origin of neck scars, radiotherapy, etc. For the surgery, the patient is placed in supine position, with the thorax elevated approximately 30 degrees, or in inverted Trendelenburg position, to decrease venous congestion and bleeding. The type of airway control (orotracheal, nasotracheal, or tracheostomy) will depend mostly on the simultaneous treatment of the primary tumor. When the patient requires a tracheostomy, or if he or she already has one, we try not to include it in the incision planned for the dissection.

8.8.2 Skin Incisions

The incision employed should fulfill the following requirements:

Allow for the best possible exposure.

Include previous scars if they are located in the surgical path.

Try to preserve skin flaps’ blood supply, particularly in patients who have received previous radiotherapy.

Protect the trajectory of large vessels (a relative issue in case of selective dissections, where vessels do not lose the muscle protection, which does occur in classic radical dissections).

Facilitate the performance of reconstructive procedures, when necessary.

Offer patients the best esthetic result possible.

Although there are many possible incisions, our first recommendation is that of transverse incisions, which comply with most of the requirements mentioned. For supraomohyoid and lateral dissections, a transverse incision performed on a preexisting skin fold is usually adequate. (image Fig. 8.5). For supraomohyoid dissections, the incision is located one or two fingerbreadths below the mandibular border, and for the lateral dissection, it is located in the midpoint of the area of dissection. If the patient is tall and slim, or if the dissection needs to be extended to the posterior triangle, or to reach the suboccipital region, a MacFee type of incision can be very helpful. It consists of two transverse incisions located on preexisting skin folds, approximately two fingerbreadths below the mandibular border and two fingerbreadths above the clavicle, respectively (image Fig. 8.6). In cases of bilateral dissections, two separate or joined incisions are indicated, if that is helpful for the treatment of the primary tumor (image Fig. 8.5).

Feb 14, 2020 | Posted by in OTOLARYNGOLOGY | Comments Off on Selective Neck Dissection

Full access? Get Clinical Tree

Get Clinical Tree app for offline access