Selective Neck Dissection for Node-Positive Oral Cavity Squamous Cell Carcinoma: A Retrospective Cohort Study

Abstract

Introduction  Selective neck dissection in clinically node-negative neck is considered the standard of care for oral squamous cell carcinomas (SCCs). Controversy still prevails in node-positive disease regarding the extent of neck dissection. In our part of the world, comprehensive neck dissection is mostly considered to be the minimal optimal treatment for palpable neck disease.

Objective  To compare regional control and disease-specific survival between clinically node-positive and node-negative patients undergoing selective neck dissection for oral SCC.

Methods  This was a retrospective cohort study conducted in the department of ENT, Head and Neck surgery at a tertiary care hospital. All patients with biopsy-proven oral and lip SCC, with or without nodal disease, who underwent selective neck dissection between April 2006 and July 2015 were included in the study.

Results  During the study period, 111 patients with oral SCC underwent selective neck dissection, of whom 71 (62%) were clinically node-negative and 40 (38%) patients had clinically positive nodes in the neck. The mean follow-up was 16.62 months (standard deviation [SD]: 17.03). The overall regional control rates were 95 versus 96% for clinical negative versus positive nodes, respectively ( p  = 0.589). The disease-specific survival was 84.5% in the node negative group versus 82.5% in the node-positive group ( p  = 0.703).

Conclusion  Selective neck dissection in node-positive neck oral SCC has similar regional control rates when compared with node-negative neck SCC. The difference in disease-specific survival between the two groups is also not significant.

Introduction

Oral cancer is among the commonest cancers in some parts of South Asia. ^{1 2} Over 95% of oral cancers are squamous cell carcinomas (SCCs). ³ The estimated incidence of cervical metastasis in this group is of up to 40%. ⁴ Multiple studies have shown that the presence of neck metastasis is the most important prognostic factor in the management of oral cancers. ^{5 6} The presence of neck metastasis reduces the survival by up to 50%. ^{7 8} Therefore, neck dissection is considered a standard of care in the treatment of oral cancers. However, morbidity associated with neck dissection remains a challenge. ^{9 10} Currently, by consensus, supraomohyoid neck dissection (SND), as well for node negative oral SCC. However, the role of selective neck dissection in clinically palpable nodes is controversial, as many surgeons still recommend comprehensive neck dissection in patients with node-positive disease. ^{11 12}

In the past 2 decades, many clinicians have explored the role of selective neck dissection in node-positive SCC. In the beginning of the present century, Loree et al ¹³ showed that the overall regional control rates with selective neck dissection were 88 versus 71% for pathologically negative versus positive-node neck disease, respectively. Two recent review articles on the same subject have also concluded that SND can be an adequate substitute for comprehensive neck dissection (CND) without compromising oncological efficacy. ^{14 15} Interestingly, studies have also shown that < 50% of the clinically palpable nodes turn positive pathologically. ¹⁶ The reason for this lymphadenopathy can be superimposed infection at the cancer site or tuberculosis (TB) lymphadenitis, especially in the regions where TB is endemic. ¹⁷ Based on these observations, a recent article from the All India Institute of Medical Sciences has also proposed that comprehensive neck dissection in all clinically palpable oral cancers may not be warranted. ¹⁶

Despite the changing trends toward performing selective neck dissection internationally, the practice is still different in our part of the world. Most of our local surgeons still recommend comprehensive neck dissection as the standard of care for node-positive SCC, with specific reference to studies by Ehsan ul Haq et al. ¹⁸ and Shaikh et al. ¹⁹ There is severe scarcity of local published literature available on the subject of selective neck dissection in node-positive oral cancers. Therefore, the role of selective neck dissection in node-positive neck needs to be explored in our set of patients.

Based on this background, we intend to evaluate the role of selective neck dissection in clinically positive oral SCC, and to compare it with node-negative oral cancers in terms of regional control and disease-free survival.

Material and Methods

First, approval of the study was obtained from the Ethics Review Committee of the hospital (approval form no 36). A review of the prospectively collected oral cancer database of the Department of Otolaryngology Head and Neck Surgery at a tertiary care hospital was performed. All patients who underwent selective neck dissection for SCC involving the oral cavity and lip between April 2006 and July 2015 were reviewed. The medical records of these patients were also reviewed to confirm findings of the database. Only patients with primary SCC of the lip or of the oral cavity were included. Non-SCC cancers, patients who underwent resection of nonlymphatic structures such as the sternocleidomastoid muscle (SCM), the spinal accessory nerve (XI Nerve), the internal jugular vein (IJV), or the skin, and patients who underwent comprehensive neck dissection (that is, radical or modified radical neck dissection) and/or preoperative chemoradiation were excluded from the study.

Selective neck dissection for SCC of the oral cavity or of the lip involves the systematic and comprehensive removal of lymph nodes and lymphatic-bearing tissue from neck levels I, II, III and/or IV. ^{20 21} Nonlymphatic structures such as the SCM, the XI nerve, and the IJV are preserved. The extent of SND in the present study corresponded to this definition.

All patients were divided into two groups, one with clinically node-negative and the other clinically node-positive, irrespective of the site and stage of the primary site. Both groups underwent selective neck dissection and postoperative radiotherapy, except for those in early T stage with pathologically N0 disease. All patients were regularly followed-up in the ear, nose, and throat (ENT) clinic. Data was prospectively collected by a dedicated research officer either face to face in the clinic, or by telephonic interviewing of the patients.

Categorical variables are analyzed as proportions and compared between groups with the chi-squared test. Continuous variables are analyzed as means with standard deviations (SDs). Both groups were analyzed separately in terms of disease-specific survival and regional recurrence/residual rates with Kaplan Meier curves.

Results

A total of 124 patients were retrieved from our database. On review of their medical records, 13 patients were excluded because of the exclusion criteria outlined above. A total of 111 patients were analyzed; 70 patients belonged to the clinically node-negative group and 40 patients belonged to the clinically node-positive group. Their mean age was 47 (SD: 11.97) years old. Eighty-five (76.6%) patients were male and 26 (23.4%) were females. Our mean follow-up period was 16.62 months (SD: 17.03), ranging from 1 to 78 months. All patients underwent selective neck dissection along with resection of the primary tumor. The mean total number of lymph nodes dissected in our patients was 34.75 (SD: 15.12). All patients beyond stage I disease were advised to undergo postoperative radiotherapy. Eight patients from the node-negative group and two from the node-positive group did not receive radiation. Distributions of different variables in both groups are shown in Table 1 .

Table 1. Demographics and other variables from node-negative and node-positive groups.

Variable	Node-negative	Node-positive
Total patients	71	40
Age (years old) *	49 (25–85)	43.70 (22–64)
Male	55 (77.5%)	30 (75%)
Female	16 (22.5%)	10 (25%)
Total number of nodes *	35	34.3
Follow-up (months) (SD) *	16.21 (14.48)	17.35 (21)
RTx Dose #	66 Gy	66 Gy

Abbreviations: RTx, Radiation therapy; SD, standard deviation.

^*mean, # Mode.

The most common primary site was the buccal mucosa (56.7%), followed by the tongue (25.2%). The different subsites and their association with node-negative and node-positive groups is depicted in Table 2 . In both groups, the buccal mucosa remains the most common site, followed by the tongue. Most of the patients in the present study had stage IV disease, mainly because of their T Stage. The distribution of T stage in our patients is shown in Fig. 1 . Both in the node-negative and in the node-positive group, T4 remains the most common stage, followed by T2 and T3, respectively. Further details are shown in Table 3 . Thus, the majority of our patients had locally advance disease.

Table 2. Oral cavity subsite distribution among node-positive and node-negative groups.

Node Status	Lip	Buccal mucosa	Oral tongue	Hard palate	Lower alveolus	Floor of mouth	Retromolar trigone	Total
Node-negative	3 (4.22%)	38 (53.5%)	21 (29.6%)	3 (4.22%)	2 (2.8%)	1 (1.4%)	3 (4.22%)	71 (100%)
Node-positive	1 (2.5%)	26 (65%)	6 (15%)	1 (2.5%)	4 (10%)	0 (0%)	2 (5%)	40 (100%)

Fig. 1.

Tumor (T) staging of the patients.

Table 3. Tumor stage distribution among node-positive and node-negative groups.

Tumor stage	Node-negative	Node-positive
T1	13 (18.3%)	4 (10%)
T2	21 (29.6%)	5 (12.5%)
T3	9 (12.7%)	9 (22.5%)
T4	28 (39.4%)	22 (55%)
Total	71 (100%)	40 (100%)

All patients, irrespective of the primary site, were treated with resection of the primary tumor with grossly negative margins, selective neck dissection, and reconstruction of the defect with microvascular free flap, if required. A total of 101 out of 111 (91%) patients received adjuvant postoperative radiation.

The statistics related to the clinical and pathologic nodal staging of our patients is shown in Table 4 . A graphical presentation of the statistics is shown in Fig. 2 . One very important finding is the conversion rates of clinical versus pathological staging. Twenty-one out of the 71 (29.5%) patients of the clinically node-negative group turned out to be positive for cervical metastasis on histopathology. Most of the clinically missed nodal metastasis staging was N2b. Fifteen out of 40 clinically node-positive patients were pathologically N0. This conversion is of high statistical significance, with p value< 0.001, as shown in Table 4 .

Table 4. Nodal staging and conversion of nodal staging patients.

N Stage	Clinical	Pathological
N0	72(64.9%)	65(58.6%)
N1	28 (25.2%)	15 (13.5%)
N2a	06 (5.4%)	02 (1.8%)
N2b	05 (4.5%)	27 (24.3%)
N2c	0	02 (1.8%)
N3	0	0

Fig. 2.

Nodal (N) staging of the patients.

All patients had no evidence of distant metastasis prior to treatment. The main outcome measures of our patients were disease-specific survival and regional control rates. Additionally,, distant survival rates were also calculated. Disease-specific survival in the node-negative group was 84.5 compared with 82.5% in the node-positive group ( p  = 0.703). Seven out of 40 patients in the node-positive group expired because of the disease, 2 of whom were pathologically N0, and rest were N2. The earliest dip in the graph is at 2 months, because of 2 patients, both belonging to N1 stage clinically. This is shown as a Kaplan-Myer depiction of survival in Fig. 3 .

Fig. 3.

Comparison of disease specific survival in node positive and node negative group.

The regional control rate was of 94.4% in the node-negative group versus 95% in the node-positive group. In total, 6 patients developed neck disease; 3 of them presented with residual disease (that is, disease coming back within 6 months of surgery), while 3 presented with recurrence (that is, after 6 months of surgery). Interestingly, 4 of them were from the node-negative group and all presented residual disease, except for one who presented with recurrence at 1 year postoperatively. The cases in the node-positive group presented with recurrences at 7 and 57 months (postoperatively), respectively. The second case also had second primary SCC. The statistical difference of both groups in this regard is also insignificant ( p  = 0.47 not significant [NS]). The comparison of outcomes in both groups along with distant metastasis is shown in Table 5 . Further description as Kaplan Myer graphs are shown in Figs. 3 and 4 .

Table 5. Outcome measures of the two groups.

Clinical node status	Disease-specific survival (%)	Regional control (%)	Distant metastasis (%)
Node-negative	84.50	94.40	2.80
Node-positive	82.50	95	10

Fig. 4.

Comparison of regional recurrence in node positive and node negative group.

Discussion

In summary, disease-specific survival and regional control rates were statistically similar both in the node-positive and in the node-negative groups. Another important finding of the present study is the significant conversion rates of clinical nodal staging versus pathological staging. This shows that clinical staging alone is not sufficient to decide the type of neck dissection. Therefore, instead of exposing the patients to radical surgeries, we should look into other aspects of cancer biology to optimize the quality of life (QOL) of our patients, without negatively impacting the oncological outcome.

Neck dissection still remains the most important part of oral cancer treatment. Since its introduction in the literature by Crile, ²² the extent of neck dissection has been the subject of significant debate and discussion. Over the past 3 decades, there has been a significant increase in the amount of literature supporting conservative or selective neck dissection. The concept of selective neck dissection first came from the studies done by Shah et al. when the mapping of lymphatic spread was shown from different subsites of the head and neck. Shah et al. demonstrated in their remarkable study of 501 oral SCC patients that level IV metastasis only occurs in ∼ 15% of patients with oral SCC and in < 4% of patients on level V. ²¹ He also concluded that level V metastasis is almost always associated with metastasis at other levels. Other researchers have also supported the finding that that skip metastasis to level IV and V is rare. ^{16 23} In general, it is recommended to include all those levels in nodal dissection, in which the chances of occult metastasis are > 20%. ^{24 25} Therefore, elective neck dissection in clinically node-negative disease requires minimum clearance from level I to III.

Controversy prevails in the proper surgical management of the node-positive patient. As discussed above, many institutions, especially in our part of the world, are still performing comprehensive or modified versus radical neck dissection. This is despite the plethora of literature over the past few decades that have supported the less morbid selective neck dissection, which is still an oncologically sound operation. Mclean et al. published a systematic review on prophylactic level V dissection in node-positive mucosal head and neck SCC. They have shown that, irrespective of the site, the incidence of occult metastasis is not > 2.56%. On this basis, they have concluded that elective dissection of level V in absence of clinical involvement may be an extra morbidity for the patient. ²⁶

Recently, a study from the Shuakat Khanum memorial hospital has shown that SND can provide similar results in node-positive disease, if combined with postoperative radiotherapy. ²⁷ These results from South Asia match with the ones published by many authors in the European and American literature over the last 2 to 3 decades.

It is also debated in the literature whether the regional control of the disease in oral cancers after SND is due to radiation therapy or to chemoradiation therapy. ²⁸ However, in a systematic review published in 2018 Rodrigo et al have concluded that adjuvant chemoradiation is elemental in achieving regional control after SND. ¹⁵ As the National Comprehensive Cancer Network (NCCN) recommends radiation therapy in every node-positive oral cancer, ²⁹ this would also be an argument against performing a more radical neck dissection surgery. The main concern in performing a comprehensive neck dissection is the increased incidence of multiple complications and of sequelae, as described in the literature, for radical and modified radical neck dissection, such as damage to the spinal accessory nerve, to the phrenic nerve, to the brachial plexus, and to the thoracic duct. Comparatively, complication rates are much lower in selective neck dissection, as opposed to comprehensive neck dissection. Apart from complications, comprehensive neck dissections usually lead to neck and upper limb dysfunction despite sparing the spinal accessory nerve, mainly because of its excessive manipulation. ³⁰

The results of our study strongly support the notion that selective neck dissection can provide very good oncologic outcomes in most of the cases of oral SCC. Although our study is retrospective, all of our data bank was prospectively collected with a dedicated team continuously following patients from the initial evaluation to the most recent follow-up visit. We only evaluated lip and oral cancer patients undergoing selective neck dissection to minimize the bias, as different sites of the head and neck have different routes of spread. Another strong point of our study is that the results are from a single surgeon. The most significant limitations of the present study were that selective neck dissection was not compared with comprehensive neck dissection and that the mean follow-up was of < 2 years. Furthermore, around 5% of patients were lost to follow up and level of neck recurrence was not recorded in the data. The importance of the level of recurrence is obvious, as it can show the exact relation with the inclusion of different levels in neck dissection.

The results of our study are also consistent with the international literature. We have compared our outcomes – regional control and disease specific survival – in Tables 6 and 7 ^{31 32 33}

Table 6. Regional control rates after selective neck dissection in node-positive and node-negative groups.

Authors	Node-negative (%)	Node-positive (%)	Follow-up (months)
Loree et al. 13	88	71	24 (min)
Chepeha et al. 33	84.6	92.3	24.5 #
Iqbal et al. 27	91.1	86.9	29 >
Shimura et al. 11	——	94	42 (min)
Our Results	94.4	95	16.62 #

*Regional control in node-negative was not described in the present study, # mean follow-up, > median follow-up, (min) minimum follow-up.

Table 7. Disease-specific survival after SND in node-positive oral squamous cell carcinoma.

Author	DSS (%)	Follow-up period
Shah et al. 32	68.8 >	40.8*
Shin et al. 28	70	47 months*
Feng et al. 33	52.8	68 months #
Shimura et al. 11	61	42 (min)
Our Results	82.5	16.62*

Abbreviation: DSS, disease-specific survival.

^> actuarial survival, * mean follow up, # median follow up, (min) minimum follow up.

Conclusion

Selective neck dissection in neck node-positive oral SCC has similar regional control rates when compared with node-negative neck SCC. The difference in disease-specific survival between the two groups is also not significant. Furthermore, the study also reveals a high conversion rate of clinical node-negative to pathological node-positive. Therefore, we suggest that selective neck dissection can provide results comparable to those of node-negative patients in oral SCC. Therefore, we suggest that SND should be performed in all early node-positive patients, followed by postoperative radiation.