20 Elective Neck Dissection for Upper Aerodigestive Tract Cancers

Neck metastasis is one of the most important prognostic factors in head and neck cancers. Its presence even with the smallest of primary cancer upstages the disease to stage III, impacting survival and necessitating the addition of adjuvant therapy. Prognosis worsens with increasing size, number of lymph nodes, as well as the presence of extracapsular spread (ECS).¹ Early detection and appropriate treatment of the neck is therefore of paramount importance. When the primary modality of treatment is surgery, management of the neck could be elective or therapeutic. Elective neck dissection (END) is a nodal dissection procedure performed in a clinico-radiologically node-negative neck. Surgery is usually a selective procedure which entails removal of nodes at the highest risk of metastasis for the index cancer. Nodes harvested at surgery, if they contain metastases, are usually occult in nature. A therapeutic neck dissection (TND) in contrast is a procedure performed when nodal metastases are clinico-radiologically manifested. A TND entails a comprehensive dissection of the neck which usually is a modified neck dissection with removal of all levels of nodes (levels I–V) while preserving nonlymphatic structures (sternocleidomastoid, internal jugular vein, and spinal accessory nerve) unless involved by disease.

The role of END has been a contentious issue in the management of early oral cancers where surgery is the primary treatment modality. With an increasing role of transoral robotic surgery (TORS) and transoral laser microsurgery (TOLM) in oro- and laryngopharyngeal cancers, the role of END has gained attention in recent years.

20.2 Elective Neck Dissection in Oral Cancers

20.2.1 Background

The role of END for early oral cancer has been a debate that has plagued the head and neck community worldwide for over five decades. The primary cancer is usually treated by excision via per-oral route. Controversy surrounds the appropriate management of the neck in this clinical scenario. There are two schools of thought—one in favor of END and the other which proposes a wait and watch (WW) policy followed by TND in patients who develop a nodal relapse.

The proponents of the WW group cite a lack of robust evidence for disease control or survival in favor of performing an END. Furthermore, they state that up to 70% of patients who are true node negative undergo unnecessary surgery with associated morbidity and costs. They believe that intensive follow-up will ensure early detection of nodal relapse and timely salvage without compromise in disease control.

Proponents of END, on the other hand, cite a definite advantage in terms of locoregional control and survival. They allude to the fact that both the neck and the primary tumor are treated at a single stage. Moreover, neck dissection in this scenario is usually a selective procedure and not associated with significant morbidity or prolonged hospitalization. Being a staging procedure, nodal metastasis is identified at an occult stage in contrast to a WW approach, where recurrences are known to present at a higher N stage with possible ECS, necessitating larger surgery, and an increased need for adjuvant treatment with resultant worsening of overall prognosis. Given the fact that published studies had small sample sizes and were predominantly retrospective in nature, the management of the neck in this situation remained in the state of clinical equipoise. This in turn resulted in gross variability in practice across the globe.

Elective neck treatment was generally recommended in patients in whom the probability of occult metastasis was greater than 20%. This was based on the findings of Weiss et al² who using a decision tree analysis suggested 20% as the threshold for elective neck treatment. Clinicians therefore attempted to identify patients who were node positive using advances in imaging or at an increased propensity to nodal metastasis using adverse tumor factors as a surrogate. This guided treatment philosophy for the elective treatment of the neck.

20.2.2 Imaging to Predict Nodal Metastasis

Multiple imaging modalities in the form of ultrasonography (USG; with or without guided fine needle aspirate cytology [FNAC]), computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET) have been used in an attempt to accurately stage the neck prior to surgery. The diagnostic accuracy of these various imaging modalities was assessed in a meta-analysis by Liao et al³ who looked at 32 studies specifically in the node-negative situation. This meta-analysis showed no significant difference in sensitivity and specificity across the various imaging modalities, with the sensitivity ranging between 52 and 66% and the specificity between 78 and 83%. Although CT was seen to be more specific than USG, the overall diagnostic accuracy of the various imaging techniques was more or less similar and no single modality stood out as being the most reliable imaging tool for occult neck metastasis. Similarly, Kyzas et al⁴ in a meta-analysis looking at the role of 18 FDG PET concluded that the sensitivity was just 50% in the clinically node-negative neck. In this context, it is pertinent to note that in the recently published Sentinel European Node Trial (SENT)⁵ of 415 patients labeled as node negative following an extensive preoperative workup that included either CT and/or MRI with or without USG-guided FNAC, 23% of nodes still turned out to be positive when the sentinel node was examined. This was despite a rigorous workup in a trial setting which would be definitely more intense than in routine clinical practice. Given, therefore, the low sensitivity of imaging in the preoperative setting, this approach did not seem reliable in identifying patients with occult nodal metastasis.

20.2.3 Tumor Characteristics

Various histological factors associated with the primary tumor were used to identify patients at a higher risk for metastasis. While increasing tumor size logically would be the most practical and easiest to use, this often does not hold true with a number of smaller tumors exhibiting biological aggressiveness and an increased risk of metastasis. Tumor grade, thickness, and aggressive histological factors (lymphovascular embolism, perineural invasion) have been described as predictors of occult metastasis in early oral cancers. Among them, the most extensively studied is tumor thickness. Spiro et al⁶ conducted a study on tongue and floor of the mouth cancers to look at the predictive value of the tumor thickness. The results showed that the risk of occult nodal metastasis in tumors ≤ 2 mm, 3 to 8 mm, and > 8 mm was 7, 26, and 41%, respectively. They also concluded that disease-related death increases with the increase in tumor thickness. Subsequent studies from different authors corroborated these findings but used different depths as cutoff. In a meta-analysis of 16 studies with 1,136 patients from the published literature, Huang et al⁷ looked at the nodal metastasis rates at sequential cutoff points of tumor thickness and found that there was a statistically significant increase in the risk of occult nodal metastasis between 4 and 5 mm (4.5 vs. 16.6%, p = 0.007). They thus concluded that the neck should be electively addressed at tumor thickness 4 mm and beyond. The association of PNI and grade with increased risk of nodal metastasis has also been extensively researched. While these pathological features were helpful in identifying patients at risk of metastasis, they were available only on postoperative histopathology, thus limiting their application in preoperative planning.

20.2.4 Previous Attempts to Address the Issue

There have been attempts to address this issue through numerous retrospective studies, as well as prospective randomized controlled trials (RCT). However, all these studies had serious limitations such as small number of patients and varying endpoints to draw meaningful conclusions. In an attempt to overcome these shortcomings, Fasunla et al⁸ conducted a meta-analysis on the four RCTs that had been published to date. The findings were in favor of END reducing the risk of disease-specific death. However, even this meta-analysis when critically analyzed seemed to have statistical limitations which included small numbers (283 patients) and wide variations with respect to time periods of accrual, analysis, end points, as well as gross heterogeneity of these studies. Moreover, the findings of one study seemed to be at variance with the others and could have thus influenced results. Given these limitations, therefore, there was still a need for generation of robust level I evidence to address the issue.

20.2.5 Current Evidence in the Management of the N0 Neck

Recently, results of a large single institutional RCT⁹ (NCT00193765) with adequate statistical power published by our group demonstrated the benefit of END. The trial initiated in 2004 was designed to demonstrate a 10% increase in overall survival (OS) by an END over the WW policy assuming a baseline 5-year OS of 60%. A total of 596 patients were randomized from trial initiation till June 2014 at which point the Data and Safety Monitoring Committee observing a difference in event rate in the two arms ordered an interim analysis. Out of the 596 patients, 500 who had undergone a minimum of 9 months of follow-up at that point in time were included in the final analysis. The arms were equally balanced for both stratification factors (site, sex, preoperative imaging, and T size) as well as grade, LVE and/or PNI, depth of invasion, and the receipt of adjuvant RT, all factors known to influence outcomes in these patients. The median depth of invasion of patients included was 6 mm indicative of the early nature of patients included in the trial. There was a highly significant difference in the OS (80 vs. 67.5%; p = 0.014) and in disease-free survival (DFS; 69.5 vs. 45.9%; p < 0.001) in favor of elective neck treatment. This translated into one life being saved for every eight patients (OS difference of 12.5%) and one less recurrence for every four patients (DFS difference of 23.6%) electively operated in this setting. After adjusting for covariates in a cox proportional hazard model for multivariate analysis, the study intervention of END showed a significantly improved OS.

The findings of our study were corroborated in an updated meta-analysis¹⁰ that included the findings of five RCTs of 779 patients. Adjusting for heterogeneity between the trials, this meta-analysis suggested that END at the time of resection of the primary tumor confers an improved DFS and OS in patients with clinically node-negative oral cancer. Thus, with strong level I evidence in favor of END, it should become the standard of care for early-stage node-negative oral cancers. A WW approach results in a statistically significant detriment in OS as well as DFS.

20.2.6 Is the Benefit Seen in all Groups of Patients?

While the OS benefit favors most subgroups, there was a suggestion⁹ that patients with thinner tumors less than 3 mm may not benefit from an END. While it is known that thinner tumors would have a lower propensity for neck node metastasis,⁶ and therefore fewer event rates, a study would need a very large sample size to conclusively confirm or disprove this issue. Moreover, at present there is no validated method of estimating depth of invasion of tumors preoperatively or at the time of primary surgery. Therefore, it would be safer to offer END to all patients irrespective of the tumor thickness. However, the only time it may seem logical against advocating an END is when a patient seeks opinion after surgery was performed elsewhere with tumor thickness on histopathology not more than 3 mm given the lack of definite evidence of support from published literature.

The other area of contention is whether the results of the benefit of END could be extrapolated to buccal and alveolar cancers, given that the majority of published literature is focused on tongue cancers. The findings of our RCT⁹

Stay updated, free articles. Join our Telegram channel

Join