Is Selective Neck Dissection Appropriate for the Treatment of the N0 Neck

It is generally accepted today that a radical neck dissection is not indicated for surgical treatment of the N0 neck.30 However, the preference between selective neck dissection and the modified radical neck dissection type III (MRND-III), in which the sternocleidomastoid muscle, the internal jugular vein, and the spinal accessory nerve are preserved, remains controversial. Some surgeons, particularly in Europe, advocate this type of MRND as the treatment of choice for the NO neck, whereas in North America many surgeons prefer a selective neck dissection. It is fitting to point out the differences between the two operations and to compare pertinent outcomes as they exist in the literature today.


One of the outcomes relevant to elective neck dissection is postoperative shoulder function. In this regard, two prospective nonrandomized studies have shown the superiority of the selective neck dissection in terms of objective measures of shoulder range of motion and postoperative electromyography (EMG) of the trapezius muscle.33'34

The results of these studies are almost intuitive as dissection of the spinal accessory nerve is less extensive in a selective neck dissection. Interestingly, however, both studies showed abnormalities of trapezius function, albeit temporary and variably severe, in almost one-fourth (22%) of patients undergoing selective neck dissection and in almost two-thirds (65%) of patients undergoing modified radical neck dissection. These observations should remind us that shoulder function must be evaluated postoperatively in every patient who undergoes neck dissection. Early detection of functional deficits and prompt institution of measures to rehabilitate the shoulder will prevent further deterioration of function, particularly in elderly patients.

It was recently suggested that dissection of the "supraspinal recess" may not be necessary because the occurrence of metastases in this area of the neck is low.35,36 This is the dissection of the fibrofatty tissues located medial to the upper end of the ster-nocleidomastoid muscle and above and behind the segment of the spinal accessory nerve located between the skull base and its entrance into the sternocleidomastoid muscle. Decreased morbidity in terms of function of the spinal accessory nerve has been cited as a reason to avoid dissection of this area. Future studies of shoulder function may be necessary to support such a contention.


To date, only one prospective study has compared the efficacy of a selective neck dissection and the MRND-III. This prospective multi-institutional study, conducted by the Brazilian Head and Neck Cancer Study Group, has shown no significant difference in neck recurrence or survival between two groups of patients with cancer of the oral cavity stage N0. One group in the study received a "classic" modified radical neck dissection and the other a supraomohyoid neck dissection.37 Unfortunately, the study was not randomized, and the investigators did not exclude patients with simultaneous recurrence at the primary site in calculating the recurrence rates in the neck. Furthermore, the results were not stratified according to the pathologic staging of the neck codes.

The effectiveness of the selective neck dissection for the treatment of the NO neck is best analyzed according to the histopathologic staging of the treated side of the neck. This is because the false-positive and false-negative rates or the clinical examination of the neck is variable but generally as high as 20%.38 It is also important to include a minimum follow-up evaluation of 2 years and to control the primary tumor in order to eliminate the possibility of reseeding of the neck by a recurrent tumor. The results can then be analyzed in the following categories.

Clinically N0/Pathologically N0 (cN0/pN0)

In a prospective analysis of our practice, the rate of recurrence in the neck in the category defined as clinically NO/pathologically NO (cNO/pNO), at 2 years of follow-up, with the primary controlled, was 0% among 66 patients treated with surgery alone and 4.2% among 48 patients who also received postoperative radiation therapy. The latter treatment was given because of several characteristics of the primary tumor: close margins, perineural invasion, and advanced stage.39 Almost identical recurrence rates for this category have been reported by other institutions that treat a large number of patients. These results are outlined in Table 2-1. Comparable results have been reported with the MRND-III, as shown in Table 2-2.

Clinically N0/Pathologically N +

In the category defined as clinically NO/pathologically N+, we must distinguish between those instances in which the histopathology of the surgical specimen shows a single positive


Selective Neck Dissection, Pathologic Stage N0, Recurrence in the Neck

Type of Dissection No. of Dissections Follow-up (mo) Recurrence Rate (%)


Selective Neck Dissection, Pathologic Stage N0, Recurrence in the Neck

Pitman et al. (1997)49




Ambrosch et al. (1996)5°




Kowalski et al. (1993)53




Pellitteri et al. (1997)54




Davidson et al. (1997)55




Kerrebijn et al. (1999)56



Modified Radical Neck Dissection Type III, Pathologic Stage N0, Recurrence in the Neck


Modified Radical Neck Dissection Type III, Pathologic Stage N0, Recurrence in the Neck

Type of Dissection

No. of Dissections

Follow-up (mo)

Recurrence Rate (%)

Calearo and Teatini (1983)44




Molinari et al. (198O)43




Suarez et al. (1993)57




Bocca et al. (1984)58



Gavilan and Gavilan (1995)59




Lingeman et al. (1977)6°


Modified Radical Neck Dissection Type III, Pathologic Stage N+, Recurrence in the Neck

Type of Dissection No. of Dissections Follow-up (mo) Recurrence Rate (%)

Calearo and Teatini (1983)44 63 36 3.17

Molinari et al. (1980)43 22 36 4.5

node and those in which either multiple positive nodes or extracapsular spread of tumor are found. Byers,40 Houck and Medina,39 and Spiro et al.41 and others have analyzed the results in this fashion. The recurrence rate among patients with either multiple positive nodes or extracapsular spread of tumor, with the primary controlled, at 2 years, varies between 9.5% and 15%. Treatment of these patients included selective neck dissection and postoperative radiation therapy. Leemans et al.42 reported a recurrence rate of 11.3% among patients treated with a comprehensive MRND with preservation of the spinal accessory nerve and the SCMM, but with removal of the internal jugular vein, who received postoperative radiation and were followed up for a minimum of 2 years.

Unfortunately, there are no comparable reports for the MRND-III. The studies by Molinari et al.43 and by Calearo and Teatini44 (Table 2-3) report their results in all cases with histopathologically positive node metastases. At first glance, the figures reported suggest that the MRND-III yields better results for the neck with "histologically positive nodes." Some investigators have gone as far as performing a statistical analysis that compares the recurrence rates in these studies with the results reported by Byers40 and Pellitteri et al.45 for the selective neck dissection and have concluded that the results with the MRND-III are significantly better. These results are not comparable statistically for various reasons, however. First, the study by Molinari et al.43 included only patients with cancer of the larynx, and they included patients with clinically NO and N+ neck. Calearo and Teatini44 studied 265 patients, 195 with cancer of the larynx and 30 with tumors of the thyroid, lower lip, salivary glands, and skin. Second, although these investigators report results for patients with histologically positive nodes, they do not stratify their analysis to indicate whether the metastases were single or multiple or whether there was extracapsular spread. By contrast, the selective neck dissection studies included patients with squamous cell carcinoma of different sites in the upper aerodigestive tract, and they report results for patients with single metastasis and for patients with multiple nodes or extracapsular spread. In any event, the fact that Moli-nari and colleagues observed a recurrence rate of 4.5% in a series of patients with histologically positive nodes treated with surgery only is remarkable.43 Particularly because the reported recurrence rate observed in a similar series of patients treated with a radical neck dissection was 22%.46

Also pertinent today is the question recently posed by Johnson as to whether selective neck dissection is adequate therapy for patients with limited occult metastases, such as a single positive node (stage pN1). In a recent review of our experience the recurrence in the neck at 2 years in this category was 21%. A similar analysis performed by Ambrosch et al.47 demonstrated a 16% recurrence rate, and data provided to us by Byers showed a 25% recurrence rate for this category in the MD Anderson experience.48 There is no published detailed analysis of such cases trying to determine why the recurrence rates are so high. We can only speculate about possible causes and possible alternatives to decrease the rate of recurrence in these patients.

Perhaps the selective neck dissection is not extensive enough. Could these patients do better with a MRND? If we think in terms of the extent of the operation and compare the selective neck dissection with the MRND-III, it becomes apparent that the main difference between these operations is in the dissection of the lymph nodes in level V and level IV (Table 2-4). Is dissection of these levels expected to prove beneficial? An answer could be inferred by analyzing the incidence of metastases encountered in level V and the pattern of recurrences following selective neck dissection. The incidence ofsub-clinical metastases in the lymph nodes of level V is low (Table 2-5). An analysis of the location of the neck recurrences among 142 selective neck dissections studied by Pitman et al.49 showed that all recurrences developed in the regions predicted by the lymphatic drainage of the primary tumor site; none was outside the dissected levels. None was in level V. One recurrence was in the submandibular nodes (level I) from an oral cavity primary. The other 4 were at levels II and III from 3 laryngeal tumors and 1 oral cavity tumor. Similarly, in a study of 163 selective


Selective versus Modified Radical Neck Dissection Type III

Node Groups Removed



Dissection of Level V NO Neck: Incidence of Subclinical Metastases


Dissection of Level V NO Neck: Incidence of Subclinical Metastases

Primary Site


Skolnik (1976)61

Oral cavity




Shah et al. (1990)62

Oral cavity








Moe et al. (1996)63



Yang et al. (1998)22

Larynx, glottic


neck dissections reported by Ambrosch et al.50 the recurrences in the neck, in the absence of recurrence at the primary site, developed in the dissected area of the ipsilateral neck (levels II and III). One patient with carcinoma of the uvula had a recurrence in the retropharyngeal nodes. Recurrences in level V were not observed.

In another recent study, Gui-yi Tu51 performed dissection and frozen-section examination of the lymph nodes at level II in 155 patients with supraglottic carcinomas. In 13 patients, in whom the frozen section indicated metastatic carcinoma in the lymph nodes, a modified radical neck dissection was performed. In the remaining 142 patients, no additional surgery was performed in the neck. When postoperative radiation therapy was given to these patients for reasons related to the primary tumor, the fields encompassed only the upper and mid-jugular regions of the neck. All patients were followed for 5 years. The location of the recurrences in the neck is shown in Table 2-6. No recurrences were found in the lymph nodes of level V.

Because the cited analyses of the site of recurrence in selective neck dissections, and our experience, show that most recurrences develop in the dissected areas of the neck and not in level V, it is reasonable to speculate that dissection of levels II and III, where most recurrences take place, is not adequate or "radical" enough. It may be necessary to remove the jugular vein in these patients, as suggested and practiced by Leemans et al.42 These investigators reported a recurrence rate of 9.1% in patients treated with a MRND in which the jugular vein and the sternocleidomastoid muscle were removed, whereas the spinal accessory nerve was preserved. Their patients did not receive postoperative radiation therapy when only one node was found to be positive. However, their report does not clarify whether some of these patients received radiation for reasons related to the primary tumor. An alternative has been suggested by Yuen et al.,52 who observed that recurrence in the neck developed in 20% of patients with oral tongue cancer whose neck was staged pN1 and did not receive postoperative radiation therapy. By contrast, there were no recurrences among pN1 patients who received postoperative radiation. The observed difference could not be tested statistically because of the small size of the sample. A properly designed prospective multi-institutional study is clearly needed to elucidate the role of selective or MRND with or without radiation in treatment of the pN1 neck.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


  • Fredrik
    Why do we do neck dissection in N0 neck?
    8 years ago

Post a comment