Patient characteristics
A total of 1,547 patients with OSCC were treated at the Department of Oral and Maxillofacial Surgery of Tokyo Medical and Dental University. ND was performed on 931 neck sides in 794 patients. Of these, 544 neck sides in 477 patients (30.8%) demonstrated cervical lymph node metastases on pathological examination. Among these patients, we retrospectively reviewed 49 patients had underwent extended ND (35 men and 14 women; age range, 23–83 years; median age, 63.0 years). The median follow-up period was 63 months (range, 1–179 months). In the 49 extended ND patients, the primary tumor affected the tongue in 29 patients, buccal mucosa in 10 patients, lower gingiva in 6 patients, upper gingiva in 3 patients, and floor of the mouth in a patient. Clinical T stage classification was as follows: 13 patients for cT1, 21 patients for cT2, 9 patients for cT3, 5 patients for cT4a, and a patient for cT4b. With respect to the histological differentiation of primary tumor, we found well differentiation in 17 patients, moderate differentiation in 7 patients, poor differentiation in 12 patients, and differentiation was unknown in 13 patients. The initial treatment for the primary tumor was surgery in 38 patients and brachytherapy (Cs-137, Ir-192, or Au-198) in 11 patients. Clinicopathological variables among patients who underwent extended ND are described in Table 1.
Table 1
Univariate analyses of correlation between clinicopathological characteristics and regional recurrence in patients with extended ND
Clinicopathological characteristics | Without regional recurrence (n = 31) | With regional recurrence (n = 18) | P value |
Sex | | | |
| Male | 23 | 12 | 0.815* |
| Female | 8 | 6 |
Age | | | |
| < 60 years | 16 | 11 | 0.729* |
| ≥ 60 years | 15 | 7 |
Primary site | | | |
| Tongue | 16 | 13 | 0.230** |
| Others | 15 | 5 |
cT stage | | | |
| T1-2 | 22 | 12 | 1.000* |
| T3-4 | 9 | 6 |
Initial treatment for primary tumor | | | |
| Surgery | 24 | 14 | 1.000** |
| Brachytherapy | 7 | 4 |
Period of extended ND | | | |
| Initial | 8 | 3 | 0.724** |
| Subsequent | 23 | 15 |
pN stage | | | |
| N1-2 | 11 | 5 | 0.754** |
| N3b | 20 | 13 |
Number of positive nodes | | | |
| < 4 | 20 | 9 | 0.487* |
| ≥ 4 | 11 | 9 |
Contralateral node metastasis | | | |
| Absent | 29 | 13 | 0.084** |
| Present | 2 | 5 |
ENE | | | |
| Absent in pN | 5 | 1 | 0.393** |
| Present in pN | 26 | 17 |
Number of resected non-lymphatic structures | | | |
| Single | 19 | 8 | 0.398* |
| Multiple | 12 | 10 |
Types of resected non-lymphatic structures | | | |
| Without neurovascular structure | 26 | 9 | 0.028* |
| With Neurovascular structure | 5 | 9 |
Adjuvant therapy | | | |
| Not performed | 14 | 2 | 0.025** |
| Performed | 17 | 16 |
*Chi-square test. **Fisher’s exact test. |
Extended ND
Extended ND was performed in 11 patients at the initial surgery combined with a primary resection, in 38 patients at subsequent cervical lymph node metastasis. The ipsilateral extended ND was performed in 43 patients and the contralateral in 6 patients. Dissected nodal levels included a patient with level I, a patient with levels I and II, 3 patients with levels I–III, 5 patients with levels I–IV, and 39 patients with levels I–V. The average number of the dissected nodes from surgical specimen was 50.8 (range, 8–141) and a median number of 3 metastatic lymph nodes (range, 1–17). Pathological ENE was found in 43 patients (87.8%). Table 2 shows the resected non-lymphatic structures and regional recurrence rate. Most frequently sacrificed tissues were as follows; mandible, mylohyoid muscle, platysma, and external carotid artery, with the regional recurrence rates of 29.4%, 31.3%, 27.3%, and 71.4%, respectively. There was no significant difference in the regional recurrence rate, regardless of whether the non-lymphatic structure requiring resection was single or multiple (29.6% vs 45.4%, P = 0.398). The regional recurrence rate was significantly higher in patients who required resection of neurovascular structures (neurovascular resection group) than in patients who did not require resection of these structures (neurovascular non-resection group) (64.3% vs 25.7%, P = 0.028) (Table 1).
Table 2
Resected non-lymphatic structures and regional recurrence rate
Non-lymphatic structure | No. of patients (n = 49) | Patients with regional recurrence |
Mandible | 17 | 5 (29.4%) |
Mylohyoid muscle | 16 | 5 (31.3%) |
Platysma | 11 | 3 (27.3%) |
External carotid artery | 7 | 5 (71.4%) |
Skin | 7 | 3 (42.9%) |
Hypoglossal nerve | 6 | 2 (33.3%) |
Masseter muscle | 5 | 2 (40.0%) |
Lingual artery | 2 | 2 (100%) |
Lingual nerve | 2 | 1 (50.0%) |
Hyoglossus muscle | 2 | 1 (50.0%) |
Superior thyroid artery | 1 | 1 (100%) |
Glossopharyngeal nerve | 1 | 1 (100%) |
Vagus nerve | 1 | 1 (100%) |
Sternohyoid muscle | 1 | 1 (100%) |
Geniohyoid muscle | 1 | 1 (100%) |
Genioglossus muscle | 1 | 1 (100%) |
Medial pterygoid muscle | 1 | 0 (0%) |
Thyrohyoid muscle | 1 | 0 (0%) |
Hyoid bone | 1 | 0 (0%) |
Complications of extended ND
Complications after extended ND included recurrent nerve paralysis in three patients, facial nerve paralysis in two patients, and chyle leak, fracture of mandible, and urgent tracheostomy due to larynx edema in a patient, respectively. At last follow-up, a patient had persistent gastrostomy and tracheostomy tube.
Treatment outcome
Twenty-four patients developed recurrence, specifically regional recurrence in 12 patients, distant metastasis in 6 patients, and regional recurrence and distant metastasis in 6 patients. The median interval from surgery to regional recurrence was 4.1 (interquartile range, 51.5–190) months. As shown Fig. 1, the DSS rate was much lower for patients with regional recurrence (5.9% vs 79.6%, P < 0.001).
Independent predictive factors for regional recurrence
Univariate analysis was demonstrated that types of resected non-lymphatic structures (P = 0.028) and adjuvant therapy (P = 0.025) were significantly associated with regional recurrence. Then, by accomplishing multivariate Cox regression analysis, we identified that types of resected non-lymphatic structures (P = 0.048, HR = 2.59, 95% CI: 1.01–6.66) was independent predictive factor for regional recurrence, which was demonstrated in Table 3.
Table 3
Multivariate analysis using Cox proportional-hazard regression for regional recurrence
Clinicopathological factors | Hazard ratio (95% CI) | P value |
Types of resected non-lymphatic structures | | |
| Without neurovascular structure | 1.00 (Reference) | 0.048 |
| With Neurovascular structure | 2.59 (1.01–6.66) |
Adjuvant therapy | | |
| Not performed | 1.00 (Reference) | 0.053 |
| Performed | 4.38 (0.98–19.5) |
Independent predictive factors for DSS
Analyses of the clinicopathologic predictors of DSS were performed using a Cox proportion hazard model. Initially, univariate Cox regression was performed to identify variables significantly associated with DSS, demonstrating that number of positive nodes (P = 0.048, HR = 2.36, 95% CI: 1.01–5.54), contralateral node metastasis (P = 0.046, HR = 2.78, 95% CI: 1.02–7.59), types of resected non-lymphatic structures (P = 0.018, HR = 2.77, 95% CI: 1.19–6.43), and adjuvant therapy (P = 0.017, HR = 5.91, 95% CI: 1.38–25.4) were significantly associated with DSS (Table 4). Subsequently, those variables significantly associated with DSS mentioned above were included in multivariate Cox regression analysis to determine independent predictive factors for DSS, revealing that types of resected non-lymphatic structures (P = 0.038, HR = 2.58, 95% CI: 1.05–6.33) was independent predictive factors for DSS (Table 4). The 5-year DSS rate was significantly lower in neurovascular resection group than in neurovascular non-resection group (23.4% vs 63.1%, P = 0.014, Fig. 2).
Table 4
Cox proportional-hazard regression for DSS
Clinicopathologic factors | Univariate analysis | Multivariate analysis |
Hazard ratio (95% CI) | P value | Hazard ratio (95% CI) | P value |
Sex | | | | |
| Male | 1.00 (Reference) | 0.723 | | |
| Female | 1.18 (0.48–2.89) | | |
Age | | | | |
| < 60 years | 1.00 (Reference) | 0.929 | | |
| ≥ 60 years | 0.96 (0.41–2.25) | | |
Primary site | | | | |
| Tongue | 1.00 (Reference) | 0.320 | | |
| Others | 1.53 (0.66–3.56) | | |
cT stage | | | | |
| T1-2 | 1.00 (Reference) | 0.923 | | |
| T3-4 | 1.05 (0.43–2.57) | | |
Initial treatment for primary tumor | | | | |
| Surgery | 1.00 (Reference) | 0.179 | | |
| Brachytherapy | 0.43 (0.13–1.47) | | |
Period of extended ND | | | | |
| Initial | 1.00 (Reference) | 0.609 | | |
| Subsequent | 1.33 (0.45–3.91) | | |
pN stage | | | | |
| N1-2 | 1.00 (Reference) | 0.136 | | |
| N3b | 2.14 (0.79–5.81) | | |
Number of positive nodes | | | | |
| < 4 | 1.00 (Reference) | 0.048 | 1.00 (Reference) | 0.625 |
| ≥ 4 | 2.36 (1.01–5.54) | 1.27 (0.49–3.26) |
Contralateral node metastasis | | | | |
| Absent | 1.00 (Reference) | 0.046 | 1.00 (Reference) | 0.154 |
| Present | 2.78 (1.02–7.59) | 2.28 (0.73–7.09) |
ENE | | | | |
| Absent in pN | 1.00 (Reference) | 0.203 | | |
| Present in pN | 3.69 (0.49–27.5) | | |
Number of resected non-lymphatic structures | | | | |
| Single | 1.00 (Reference) | 0.296 | | |
| Multiple | 1.57 (0.68–3.62) | | |
Types of resected non-lymphatic structures | | | | |
| Without neurovascular structure | 1.00 (Reference) | 0.018 | 1.00 (Reference) | 0.038 |
| With neurovascular structure | 2.77 (1.19–6.43) | 2.58 (1.05–6.33) |
Adjuvant therapy | | | | |
| Not performed | 1.00 (Reference) | 0.017 | 1.00 (Reference) | 0.097 |
| Performed | 5.91 (1.38–25.4) | 3.72 (0.79–17.6) |