• early stage;
  • tongue cancer;
  • neck dissection;
  • supraomohyoid neck dissection;
  • CT scan;
  • MRI


  1. Top of page
  2. Abstract


The role of elective and therapeutic selective neck dissection in patients with early stage cancer of the oral tongue remains controversial. The purpose was to investigate the role of neck treatment in the management of this condition.


A total of 380 patients with cT1-2N0 oral tongue cancer were retrospectively reviewed. Patients were staged by means of computed tomography (CT) or magnetic resonance imaging (MRI) scans. A total of 324 patients received elective neck dissection (END), whereas 56 participants received observation (OBS). In all, 287 patients received supraomohyoid neck dissection (SOND), whereas 37 patients received modified radical neck dissection (MRND). Overall survival (OS) and neck control rates (NCR) were investigated according to the treatment modality.


In the END group the occult metastasis (OM) rates in cervical lymph nodes were 5.2% for cT1 lesions and 14.6% for cT2 lesion (P = .005). The 5-year OS (P = .029) and NCR rates (P = .001) were significantly better in the END group compared with the OBS group. Patients who received MRND had a better 5-year NCR compared with SOND, albeit not significantly (91.4% vs 85.3%, P = .415). Multivariate analysis showed that END and stage were independent predictors of both NCR and OS.


END should be performed routinely in patients with early-stage oral tongue cancer, even in the presence of negative neck by CT scans and MRI. Cancer 2008. © 2008 American Cancer Society

In Taiwan, an endemic betel quid-chewing area, oral cavity cancer ranks fifth in cancer incidence. The most common oral cavity cancer sites are the tongue and cheeks. Notably, squamous cell carcinoma (SCC) of the mouth is characterized by a high potentiality of local invasion and metastasis to neck lymph nodes. In addition, regional recurrence is the most common cause of failure after surgical treatment of early stage oral tongue carcinoma.1 In this malignancy, survival rates decrease considerably when regional metastases are present at primary presentation or later after primary treatment.2, 3

The treatment of neck metastases in oral cavity cancer patients has changed rapidly in the past decade. In this regard, previous studies have shown that the most common sites for metastases of oral tongue cancer were at levels I and II.4–7 Moreover, the concept of functional neck dissection was popularized by Lindberg,5 Byers,8 Shah et al.,9 and Ballantyne10 led to the acceptance of selective neck dissection as treatment for lymph node disease in various stages. The term supraomohyoid neck dissection (SOND) refers to the removal of lymph nodes contained in levels I-III. This technique has been frequently used in the management of the clinically lymph node-negative neck in patients with SCC of the oral tongue. In this regard, it has been previously demonstrated that SOND and modified neck dissection provide similar control rates in patients with N0 SCC of the oral cavity.11–13 It is worth noting that some authors have suggested the potential usefulness of SOND not only in patients with N0 disease but also for selected individuals with N1 disease.11, 13 Moreover, Byers et al.14 concluded that all patients with SCC of the oral tongue should have levels I–IV nodes removed if an elective neck dissection is part of the initial therapy. Nonetheless, the proper management of the clinically negative neck in primary SCC of the oral tongue remains controversial.

Unnecessary morbidity is commonly cited as a reason for avoiding routine elective neck treatment. Moreover, doubtful survival benefit after elective neck interventions has been reported in randomized controlled clinical trials.13, 15–17 In any case, it is posited that CT scans and MRI may facilitate the detection of occult neck lymph node metastases.18

Given the controversial role of elective and therapeutic selective neck dissection in patients with early-stage SCC of the oral tongue, we retrospectively reviewed 380 consecutive patients undergoing primary radical surgery from 1995 to 2002. We specifically focused on the role of neck treatment in the management of this condition.


  1. Top of page
  2. Abstract

We retrospectively reviewed 380 consecutive patients with early-stage SCC of the oral tongue undergoing primary radical surgery from January 1995 to August 2002. Preoperatively, all patients were staged as lymph node-negative by CT or MRI scans. Five patients with synchronous second primary lesions were excluded. Study participants underwent tumor excision either with or without neck dissection. The TNM staging followed the guidelines set by the American Joint Committee on Cancer (AJCC), 1997 edition.19

Patients in this series underwent an extensive presurgical evaluation. This evaluation included a medical history and complete physical examination, flexible fiberoptic pharyngoscopy, complete blood count and routine blood biochemistry, computed tomography (CT) or magnetic resonance imaging (MRI) scans of the head and neck, chest radiographs, bone scan, and liver ultrasound.

The primary tumors were excised with ≥ 1 cm safety margins (both peripheral and deep margins). Tumor margin tissue was cryosectioned. If a margin was positive, then additional tissue was excised and cryosectioned to ensure that the margin was free of tumor.

Supraomohyoid neck dissection (SOND) or modified radical neck dissection (MRND) were performed to ensure dissection of level IV/V lymph nodes, albeit preserving the spinal accessory nerve, the internal jugular, and the sternocleidomastoid muscle. Postoperative radiotherapy (RT) was performed on patients with 1 positive lymph nodes or close margins (≤4 mm). RT was scheduled within 4–8 weeks after the operation. The initial RT treatment field was to irradiate the whole tongue tumor bed area with 1–2 cm margins and the regional lymphatics with 6 MV x-ray beams via bilateral-opposed fields with either a matched anterior supraclavicular field or the 3D conformal RT technique based on postoperative CT simulator imaging. The prescribed dose was 1.8–2 Gy per fraction per day, given 5 days per week. The total radiation dose was 66 Gy to patients with multiple positive neck lymph nodes and/or extracapsular spreading (ECS) and 60 Gy for the rest of the patients. Concomitant chemoradiotherapy (CCRT) with cisplatin-based agents was administered to those with ECS, positive resection margins, or multiple neck lymph node metastases.

The major variables examined were as follows: presence or absence of neck dissection, clinical and pathological lymph node status, neck control rate (NCR), disease-free survival (DFS), disease-specific survival (DSS), and overall survival (OS). The primary study endpoint was the regional control rate. Follow-up time was calculated from the date of initial hospital stay for surgical treatment until the date of death or last contact. Differences between proportions of patients who developed disease recurrence were tested with the chi-square test. Survival curves were constructed using the Kaplan-Meier method and compared using the log-rank test. Statistical analyses were performed using SPSS v. 8 software (Chicago, Ill). P < .05 (2-tailed) was considered statistically significant.


  1. Top of page
  2. Abstract

Patient Characteristics and Treatment Modalities

Table 1 shows the clinicopathologic characteristics of the 380 study participants (325 males and 55 females). The median age of the sample was 48 years (range, 23–82 years). According to the AJCC 1997 tumor staging criteria, there were 195 (51.3%) T1 and 185 (48.7%) T2 lesions. The majority of patients had smoking, drinking, and/or betel quid-chewing habits. Forty-two (11.1%) patients had poorly differentiated carcinoma.

Table 1. Clinicopathologic Characteristics of Patients With Early Stage Tongue Cancer (N = 380)
CharacteristicsOBS group n = 56END group n = 324P
No. (%)No. (%)
  1. OBS indicates observation; END, elective neck dissection.

 Men46 (82.1)279 (86.1).436
 Women10 (17.9)45 (13.9) 
Age, y
 ≤409 (16.1)89 (27.5).072
 >4047 (83.9)235 (72.5) 
 Yes29 (51.8)210 (64.8).072
 No27 (48.2)114 (35.2) 
Betel quid
 Yes36 (64.3)227 (70.1).434
 No20 (35.7)97 (29.9) 
 Yes39 (69.6)257 (79.3).118
 No17 (30.4)67 (20.7) 
 Well/Moderate48 (85.7)290 (89.5).624
 Poor8 (14.3)34 (10.5) 
Clinical tumor status 
 T142 (75.0)153 (47.2)<.001
 T214 (25.0)171 (52.8) 

Of the 380 patients, 324 (85.7%) received elective neck dissections (ENDs), whereas 56 patients (14.7%) were observed (OBS). Two hundred eighty-seven patients received SOND (148 in cT1 and 139 in cT2), whereas 37 patients received MRND (5 in cT1 and 32 in cT2).

The median follow-up period was 37.8 months. Twenty-four patients died of noncancer-related causes during follow-up, whereas 22 patients died of the disease. The occult metastasis (OM) rate was 10.1% (33 patients) in the END group.

Twenty-five (75.8%) patients had 1 metastatic lymph node (N1), whereas 8 (24.2%) patients had more than 1 lymph node metastasis in the ipsilateral neck (N2b). The OM rate was 5.2% (8 of 153 patients) in patients with cT1 lesions versus 14.6% (25 of 171) in patients with cT2 lesions (Table 2, P = .005).

Table 2. Relations Between Clinicopathologic Parameters, Regional Metastasis, and Occult Metastasis
CharacteristicsRegional recurrenceOccult metastasis
No. (%)PNo. (%)P
 Men45/325 (13.8).23429/279 (10.4).999
 Women11/55 (20.0) 4/45 (8.9) 
Age, y
 ≤408/98 (8.2).0339/89 (10.1).979
 > 4048/282 (17.0) 24/235 (10.2) 
 Yes36/239 (15.1).81525/210 (11.9).165
 No20/141 (14.2) 8/114 (7.0) 
Betel quid
 Yes37/263 (14.1).58223/227 (10.1).961
 No19/117 (16.2) 10/97 (10.3) 
 Yes41/296 (13.9).36130/257 (11.7).083
 No15/84 (17.9) 3/67 (4.5) 
 Well/Moderate51/338 (15.1).58332/290 (11.0).227
 Poor5/42 (11.9) 1/34 (2.9) 
Clinical tumor status 
 T128/195 (14.4).8318/153 (5.2).005
 T228/185 (15.1) 25/171 (14.6) 

The distribution of occult neck metastasis by level was as follows: 39.4% in level I, 51.5% in level II, 9.1% in level III, and 6.1% in level IV. No level V occult neck metastasis was observed (Table 3). Among the 37 patients who received MRND, separate skip metastasis in level IV neck lymph nodes was observed in 2.7% (1 case). Twenty-two patients received postoperative RT or CCRT.

Table 3. Distribution by Level of Occult Metastasis (n = 33) in END Patients (n = 324)
LevelT1, n = 8T2, n = 25Total
No. (%)No. (%)No. (%)
  1. END indicates elective neck dissection.

I5 (62.5)8 (32.0)13 (39.4)
II2 (25.0)15 (60.0)17 (51.5)
III0 (0.0)3 (12.0)3 (9.1)
IV1 (12.5)1 (4.0)2 (6.1)
V0 (0.0)0 (0.0)0 (0.0)

Patterns of Neck Recurrence

Overall, neck lymph node recurrence developed in 56 patients. Of these, 40 were in the END group and 16 in the OBS group. In the END group the distribution of neck recurrence was as follows: in the 16 patients with T1 lesions, there were 10 (62.5%) ipsilateral, 5 contralateral (31.3%), and 1 (6.3%) bilateral recurrences; in the 24 patients with T2 lesions there were 11 (45.8%) ipsilateral, 10 (41.7%) contralateral, and 3 (12.5%) bilateral recurrences. The distribution of neck recurrence by level is shown in Table 4.

Table 4. Distribution by Level of Neck Recurrence in the END Group (n = 40)
LevelT1 n = 16T2 n = 24Total n = 40
No. (%)No. (%)No. (%)No. (%)No. (%)No. (%)
  1. END indicates elective neck dissection.

I6 (37.5)5 (31.3)5 (20.8)10 (41.7)11 (27.5)15 (37.5)
II4 (25.0)3 (18.8)2 (8.33)3 (12.5)6 (15.0)6 (15.0)
III4 (25.0)0 (0.0)6 (25.0)2 (8.3)10 (25.0)2 (5.0)
IV3 (25.0)0 (0.0)4 (16.7)1 (4.2)7 (17.5)1 (2.5)
V0 (0.0)1 (6.3)1 (4.2)0 (0.0)1 (2.5)1 (2.5)
Parotid lymph nodes1 (6.3)2 (8.3)3 (7.5) 

In the OBS group the distribution of neck recurrence was as follows; in the 12 patients with T1 lesions there were 11 (91.8%) ipsilateral and 1 (8.3%) bilateral recurrences; in the 4 patients with T2 lesions there were 3 (75.0%) ipsilateral and 1 (25.0%) bilateral recurrences. Isolated contralateral neck metastases were not evident in either groups. The distribution of neck recurrence by level in OBS group is presented in Table 5.

Table 5. Distribution by Level of Neck Recurrence in the OBS Group (n = 16)
LevelT1 n = 12T2 n = 4Total n = 16
No. (%)No. (%)No. (%)No. (%)No. (%)No. (%)
  1. OBS indicates observation.

I7 (58.3)1 (8.3)4 (100.0)1 (25.0)11 (68.8)2 (12.5)
II4 (33.3)0 (0.0)0 (0.0)0 (0.0)4 (25.0)0 (0.0)
III5 (41.7)0 (0.0)2 (50.0)1 (25.0)7 (43.8)1 (6.3)
IV1 (8.3)0 (0.0)0 (0.0)0 (0.0)1 (6.3)0 (0.0)
V1 (8.3)0 (0.0)0 (0.0)0 (0.0)1 (6.3)0 (0.0)

In the OBS group, the neck recurrence rate was similar in patients with T1 lesions compared with patients with T2 lesion (28.6% vs 28.6%, respectively, P = .999). As shown in Figure 1, patients who received OBS were associated with significantly poorer neck control (5-year neck control rate: 69.3%) compared with patients who received END (5-year neck control rate: 86.1%, P < .001). Patients were further divided into 3 groups according to their neck treatment modalities: OBS, SOND, and MRND.

thumbnail image

Figure 1. (A) Neck control and (B) overall survival rates according to the neck dissection.

Download figure to PowerPoint

As shown in Figure 2, patients who received either MRND or SOND had better neck control rates as compared with the OBS group. However, no statistically significant difference was evident for patients treated with MRND compared with SOND. After allowance for potential confounders, multivariate analysis showed that END and tumor stage were independent predictors of NCR (Table 6). Salvage therapy was carried out in 52 patients with neck recurrence. Radical neck dissections were performed in 47 patients, including 10 patients who received postoperative radiotherapy and 25 patients who received postoperative CCRT. Five patients received salvage CCRT, whereas 5 patients received palliative chemotherapy alone.

thumbnail image

Figure 2. Neck control rates according to different neck treatments.

Download figure to PowerPoint

Table 6. Multivariate Analyses of Variables Influencing Neck Control Rate, Disease-free Survival, and Overall Survival
VariableHazard ratio95% Confidence intervalP
  1. SOND indicates supraomohyoid neck dissection; MRND, modified radical neck dissection; OBS, observation.

Neck control rate
 T classification1.4360.827–2.492.199
 Neck treatment  .002
Disease-free survival 
 T classification1.7781.135–2.787.012
 Neck treatment  .001
Overall survival
 T classification2.1221.133–3.971.019
 Neck treatment   

Disease-Free Survival and Overall Survival

Forty patients had local recurrence. The 5-year local control rate in the entire study cohort was 75.7% (78.4% in T1 lesions and 73.0% in T2 lesions). Among patients with local recurrence, 28 patients received additional salvage surgery. Three patients had secondary primary lesions at either the buccal mucosa (n = 2) or the gum (n = 1). Sixteen (4.2%) patients developed distant metastases (8 in the lungs, 6 in the skin, and 2 in the bone). Patients with local recurrence had a significantly higher risk of distant metastases (DM) compared with those without (15.0% vs 2.9%, respectively, P = .003). Moreover, patients with neck recurrence had a significantly higher risk of DM than those without (19.6% vs 1.5%, respectively, P = .0001).

The 5-year DFS in the entire study cohort was 75.7% (78.4% in T1 lesions and 73.0% in T2). The 5-year DFS rates in the OBS, SOND, and MRND groups were 55.6%, 78.5%, and 83.3%, respectively (Fig. 3). A statistically significant difference was observed in the DFS between the END and OBS groups (P = .0001), whereas the difference between the SOND and the MRND was not significant (P = .645). Despite the use of CCRT, the 5-year DFS was 71.4% in patients with ECS. Patients with pN2 disease had worse DFS than did those with pN1 or pN0 disease.

thumbnail image

Figure 3. Disease-free survival according to different neck treatments.

Download figure to PowerPoint

Multivariate analysis showed that END and tumor stage were independent prognosticators for the DFS. The 5-year OS rates for T1 and T2 lesions were 88.1% and 80.7%, respectively. The 5-year OS according to treatment modality in the OBS, SOND, and MRND groups were 75.1%, 87.2%, and 79.6%, respectively (Fig. 4). The difference was significant between the END and OBS groups (P = .029).

thumbnail image

Figure 4. Overall survival according to different neck treatments.

Download figure to PowerPoint


  1. Top of page
  2. Abstract

Patients with oral cavity cancer have a high incidence of occult cervical metastases,13, 15, 20–23 poor salvage rates,15, 20–22, 24 and an increased incidence of ECS in the presence of palpable lymph nodes.17 Notably, the presence of occult lymph node metastases represents a significant adverse prognostic factor in this patient group.25, 26 These observations provide the rationale for elective neck treatment in this clinical entity. Of interest, patients with occult metastases had a worse neck control and OS rates. In general, the major route of metastatic dissemination of this malignancy is through the lymphatic system. Specifically, levels I and II were the most frequently involved.25–27 The results of our present report are in keeping with previous findings. Accordingly, the occult metastatic rate was 10.2% in our series and the lymph nodes most frequently involved in OM were at levels I and II.

Although palpation is the most practical means of staging the neck, it has a false-negative rate of about 40%.18, 28 The use of CT may reduce the false-negative rate of the staging to 22.7%.18 The use of MRI or positron emission tomography (PET) scans can further improve detection rates for neck nodal metastases. In any case, it has been reported that the probabilities of occult neck metastasis after using MRI/PET were 6.7% in T1 tumors and 10.8% in T2 tumors.29 In our current study, we found that CT/MRI failed to detected neck lymph node metastasis in 5.2% of stage I patients and 14.6% of stage II patients.

A high incidence of neck recurrence has been reported in patients with T1-T2 cancer of the oral tongue treated by primary tumor excisions alone.3, 26 Specifically, cervical lymph node metastases developed subsequently in 38% to 43% of such patients.26, 30, 31 In this study, the neck recurrence rate in the OBS group (28.6%) was significantly higher compared with that observed in the END group (12.7%, P = .004).

Although contralateral regional metastases have been described in some series of patients with early-stage tumors of the oral cavity,1, 32 our data show that neck recurrence is mainly ipsilateral in patients treated with glossectomy alone. Our results are line with those obtained by Cunningham et al.3 in their series of 54 patients with stage I and stage II SCC of the oral cavity. Table 7 summarizes treatment results for different series with early stage tumors of the oral cavity.

Table 7. Literature Search for Elective and Therapeutic Neck Dissection in Patients With Early Stage Tongue Cancer
AuthorsYearStageNationsCase no.SiteCT/MRINeck dissection5-year neck recurrence rate5-year disease-free survival5-year overall survival
  • SOND indicates supraomohyoid neck dissection; MRND, modified radical neck dissection; OBS, observation; FOM, floor of mouth; CT, computed tomography; MRI, magnetic resonance imaging.

  • *

    Denotes a statistically significant difference.

Mendelson331976T1-3N0USA307 (−43 N+ patients)TongueClinicalRND vs OBS20% for OBS group68.6% vs 77.4%
Whitehurst & Droulias341977T1-3N0 and N+USA137TongueClinicalND vs OBS28.6% vs 64.5%71.4% vs 36.4%
Vandenbrouck171980T1-3N0France75Tongue and FOMClinicalND vs OBD7.7% vs 47%46.2% vs 58.3%
Teichgraeber & Clairmont21984Stage I–IVUSA136Tongue and FOMClinicalND vs OBS75.0 % vs 26% tongue
73% vs 11% FOM
Cunningham31986T1-2N0USA54Tongue and FOMClinicalMRND vs OBS11.1% vs 41.9%3 yr 88% vs 77%
Fakih131989T1-2N0India70TongueClinicalRND vs OBS30% vs 57%63% vs 52%
Lydiatt351993T1-2USA156TongueClinical, CT not routineND vs OBS18.5% vs 16.5%65% vs 43%
Kligerman151994T1-2N0 I–IIBrazil67TongueClinicalSOND vs OBS24% vs 42% (locoregional control)3.5 yr 72% vs 49%*
Yuen11997T1-2N0Hong Kong63TongueClinical10 RND9% vs 47%*86% vs 55%*
Yii301999I–IIUK77TongueClinical13 ND (10 SOND, 3 MRND)2 yr 17% vs 43%75% vs 65%
14 RT to neck
Haddadin361999T1-2M0UK137TongueClinical47 ND vs 90 OBS10.6% vs 73.7%*80.5% vs 53.6%*
Dias372001pT1Brazil49Tongue 21Clinical23 SOND4% vs 14%*3 yr 97% vs 74%*
Keski-Säntti162006T1-2N0Finland80TongueCT 24SOND 4413% vs 36%*82% vs 77%63% vs 66%
Sono 24
MRI 35
Clinical 5
Present Study2006cT1-2N0Taiwan380TongueAll287 SOND; 37 MRND; 57 OBS12.3% vs 9% vs 31%*79% vs 83% vs 56 %*87% vs 80% vs 75%

Among patients treated with END, 12.7% developed a regional recurrence. Contralateral level I lymph nodes were the most frequent site of regional recurrence. This finding was in line with previous data.13 It is posited that this phenomenon may be due to an afferent communicating pathway that drains from the floor of the mouth into the contralateral lymph nodes.38 This may also occur silently before surgery. The second most common site of regional recurrence was ipsilateral level I nodes. These metastatic deposits might originate from the floor of mouth deep to the sublingual glands and superior to the lingual nerve. In 1938 Rouvierre4 first described the presence of lymph nodes in the floor of mouth and referred to them as the median and lateral lingual nodes. The median lingual nodes are located between the lingual septum and the medial side of the genioglossus muscle. The lateral lingual nodes are located on the lateral aspect of the genioglossus on the hyoglossus muscle, in direct relation to the sublingual glands.4, 39 Metastases to the lingual lymph nodes in tongue cancer have been previously described by Ozeki et al.40 In patients with early-stage tongue cancer, neck dissection and resection of primary tumors are usually performed separately due to the small tumor extent. It should be noted, however, that some metastatic foci in lingual lymph nodes may persist with this methodology. In this regard, our current data show that the composite resection of the neck alongside with primary tumor may reduce the presence of metastatic deposits in this area. Another possible source of ipsilateral level I recurrence may be the presence of occult metastasis to the perivascular submandibular lymph nodes. Albeit occurring rarely, regional recurrence at level I has been described.41 Surgeons could confidently omit resection of these lymph nodes due to the low metastatic rate in early-stage tongue cancer as well as the possible marginal mandibular nerve injury. In any case, meticulous lymph node dissection in this area is feasible and might lower the risk of recurrence.

Controversy still exists about the role of elective treatment in the management of early SCC of the oral tongue and clinically negative neck.26, 42, 43 Unnecessary morbidity and doubtful survival benefit are commonly cited as the reasons for avoiding routine elective neck treatment. Nonetheless, the presence of cervical nodal micrometastases has been associated with a significant decrease in survival.3, 21 Kligerman et al.15 reported a lower rate of neck recurrence and better survival in patients treated by resection with SOND compared with those treated by resection alone. In our study, the 5-year NCR was much better for patients treated by END compared with OBS. In addition, the 5-year OS in the END group was superior compared with patients who had a subsequent therapeutic neck dissection. These data are in line with previous studies in early-stage tongue cancer.44 It is thus posited that END might improve both neck control and OS. Indeed, application of this technique might improve the chance of clearance of micrometastasis that cannot be detected by histology or imaging.

The possibility of level IV and level V (supraclavicular region) lymph node metastases or the so-called skip metastases might challenge the potential usefulness of SOND as an effective means by which locoregional control of oral cancer is obtained.14, 35 In a study of 277 previously untreated patients with SCC of the oral tongue, Byers et al.14 reported that 15.8% of patients had either level IV metastasis as the only manifestation of disease in the neck or the level III lymph node as the only lymph node present without disease in level I-II. In addition, Lydiatt et al.35 suggested that inclusion of the lower jugular chain of lymph nodes with the supraomohyoid neck dissection may increase the regional control rates by 20% to 24%. However, our data provide evidence that, in the group of patients treated by END, the incidence of skip metastasis to level IV in the absence of level I/II lesions is as low as 2.7% (1 case out of 37 patients). In the OBS group, the skip metastasis rate was 5.4% (3 of 56) in patients with regional recurrence who received salvage neck dissection. In our report the skip metastatic rate was lower compared with that reported in previous studies.9, 14, 45, 46 However, in our study all patients were staged with the use of CT/MR imaging. In the light of our data, routine dissection of level IV lymph node alongside SOND can provide little benefit to patients with early-stage tongue cancer. In our study, a total of 37 patients received MRND and 2 patients had level IV lymph node metastasis. Furthermore, NCR, DFS, and OS did not differ significantly between SOND and MRND groups. It is concluded that level IV nodes should not be routinely included in the neck dissection for patients with negative neck as assessed by CT/MRI scans.

Several articles have stated that tumor depth is an important factor contributing to neck lymph node metastasis.47–53 Other factors such as differentiation, DNA aneuploid, T stage, perineural invasion, infiltration pattern, and other molecular markers have also been proposed.35, 48–54 In general, these studies agree that the depth of tumor invasion more than 4 to 5 mm will have higher risk of neck lymph node metastasis. However, most of these studies were retrospective analyses that measured tumor invasion with a pathologic observation after formalin fixation but not direct measurement of tumor depth in situ. It has been reported that there may be a 20% to 47% shrinkage in tumor margins after a pathologic fixation procedure, especially for small tumors.55, 56 Therefore, it seems to not be practical to use the depth of tumor invasion or other pathologic parameters as a guideline to determine whether the patient should receive END or not. It will become 2 stages of surgery. Instead, it is proper to proceed with SOND at the time of neck operation. SOND has no major complication by experienced surgeons except for an operation scar. Based on the differential survival and disease control, we suggest SOND treatment for all early-stage oral tongue cancer patients except for those with definitely superficial tumor(s).

In summary, our study shows that level I/II lymph nodes are the most common sites of occult metastases in the END group, as well as the most common area of recurrence in the OBS group. Skip metastasis to level IV lymph nodes is rare in early-stage tongue cancer. According to our data, SOND is sufficient to remove the majority of lymph node metastases in this clinical entity. In patients with negative nodes according to CT/MRI staging, END with SOND may improve both the neck control and OS rates compared with observation alone. Hence, SOND should be performed routinely in patients with early-stage oral tongue cancer, even in the presence of negative neck by CT scans and MRI. It is unnecessary to extend dissection to level IV nodes for the patients with negative neck as assessed by CT/MRI scans.


  1. Top of page
  2. Abstract