Annals of Surgical Oncology

, Volume 16, Issue 1, pp 159–170 | Cite as

Risk Stratification of Patients with Oral Cavity Squamous Cell Carcinoma and Contralateral Neck Recurrence Following Radical Surgery

  • Chun-Ta Liao
  • Shiang-Fu Huang
  • I-How Chen
  • Joseph Tung-Chieh Chang
  • Hung-Ming Wang
  • Shu-Hang Ng
  • Chuen Hsueh
  • Li-Yu Lee
  • Chih-Hung Lin
  • Ann-Joy Cheng
  • Tzu-Chen Yen
Head and Neck Oncology

Abstract

Clinical outcome of patients with oral cavity squamous cell carcinoma (OSCC) and contralateral neck recurrence (CLNR) remains poor. We sought to identify factors associated with CLNR and incorporate them into a risk stratification scheme. Between January 1996 and June 2006, a total of 913 consecutive OSCC patients treated by radical surgery were investigated. Postoperative adjuvant therapy was performed in the presence of pathological risk factors. The duration of follow-up was at least 24 months in all surviving patients. Outcome measures were the 5-year CLNR and overall survival rates. In the entire study cohort, the 5-year CLNR rate was 7% (55/913). Specifically, it was 18% (17/132) in patients with local recurrence (LR), and 5% (38/781) in those without (P = 0.0002). In multivariate analysis, extracapsular spread (ECS) was the only independent risk factor for CLNR in patients with LR. Tumor subsite, poor differentiation, and presence of pN + disease were significant predictors of CLNR in patients without LR. We identified two groups of patients with high CLNR rates. The first group consisted of patients with ECS at the initial diagnosis and LR. The second group consisted of subjects with tongue cancer without LR harboring at least two risk factors. We conclude that, in patients who achieved local control, postoperative contralateral neck treatment is recommended for subjects with tongue cancer and at least two risk factors. Once LR occurs, contralateral neck treatment is recommended in patients with ECS.

Contralateral neck metastases are uncommon in patients with oral cavity squamous cell carcinoma (OSCC) at their initial presentation. However, they should be carefully screened when the primary tumor crosses the midline or is located at particular subsites, such as the tongue or the floor of mouth.1, 2, 3, 4, 5 In general, elective contralateral neck dissection (ND) is recommended in OSCC patients when the tumor crosses the midline. On the other hand, the use of elective contralateral ND is still a matter of debate for patients with tongue or mouth floor cancers that do not cross the midline.

The outcome of OSCC patients with neck lymph node recurrence remains dismal, especially when recurrence occurs at the contralateral site.2,4,6, 7, 8 Unfortunately, even with the use of elective contralateral ND for ipsilateral tumors crossing the midline, approximately one-third of neck lymph node recurrences occur at the contralateral site.2, 3, 4,8, 9, 10 Interestingly, we have previously shown that patients with local recurrence (LR) have a higher incidence of contralateral neck recurrence (CLNR) than those without. These observations raise several important questions. Firstly, it is possible that the incidence of CLNR may be underestimated in OSCC patients at their initial diagnosis. Secondly, risk factors for CLNR in OSCC patients have not been yet exactly identified. Thirdly, it is unclear whether the use of elective contralateral ND may reduce incidence of CLNR in this patient group.

The aim of this retrospective investigation was to identify the significant predictors of CLNR in a large cohort of OSCC patients recruited in an endemic betel quid chewing area. All patients in this study were treated by radical surgery, either with or without radiotherapy (RT). This study can clarify whether contralateral neck metastases are underestimated in OSCC patients at their initial presentation. Moreover, it may clarify whether contralateral ND can improve the clinical outcome of OSCC patients by reducing CLNR. Finally, correct risk stratification in this patient group might enable the development of a treatment flowchart aiming to reduce the incidence of CLNR in subjects with OSCC.

Patients and Methods

Patients

The study was approved by the Institutional Review Board for Human Research of Chang Gung Memorial Hospital (Taoyuan, Taiwan). Common eligibility criteria were as follows: (1) histological diagnosis of OSCC; (2) previously untreated tumor scheduled for radical surgery and ND; (3) no other suspected distant metastatic lesions detected by imaging, including magnetic resonance imaging (MRI)/computed tomography (CT) and fluorodeoxyglucose positron emission tomography (FDG PET); and (4) willingness to receive a CT-guided biopsy or surgical exploration, if necessary. Exclusion criteria included presence of midline lesions, bilateral lesions, second metachronous malignancies, and/or refusal or inability to receive definitive treatment for the disease. Patients with second metachronous malignancies other than SCC treated by RT at outside of head and neck region and did not receive chemotherapy were included.

All study participants underwent an extensive presurgical evaluation. This evaluation included medical history and complete physical examination, flexible fiber-optic pharyngoscopy, complete blood count and routine blood biochemistry, CT or MRI scans of the head and neck, chest radiography, bone scan, and liver ultrasonography (US). Tumor staging was performed according to the 1997 American Joint Committee on Cancer (AJCC), 5th edition staging criteria.11 There were two main reasons for the use of the AJCC 1997 criteria, instead of the AJCC 2002 staging system. Firstly, some histopathological specimens collected before 2002 were not available for further review. Secondly, correct staging of pT4b disease as per the AJCC 2002 criteria may prove troublesome for OSCC analyses.12, 13, 14

Surgery and Adjuvant Therapy

The primary tumors were excised with ≥1 cm safety margins (both peripheral and deep margins). In general, classic radical or modified NDs (level I–V) were performed in patients with clinically positive lymph nodes, whereas supraomohyoid NDs (level I–III) were used in clinically node-negative patients. Patients were treated with bilateral NDs if the primary tumor reached or crossed the midline sagittal plane of the oral cavity.

Tumor margins were cryosectioned. If a margin was positive, additional tissue was excised and cryosectioned in order to ensure that the margin was free of tumor. The surgical defects were repaired with primary closure or reconstructed immediately by plastic surgeons using free or local flaps. Postoperative RT was performed in patients with pT4 tumors, positive lymph nodes, or close margins (≤4 mm). RT was scheduled within 4–8 weeks after operation. The prescribed dose was 1.8–2 Gy per fraction per day, given 5 days per week. The total radiation dose was 66 Gy for patients with multiple positive neck lymph nodes and/or extracapsular spread (ECS), and 60 Gy in the remaining patients. In principle, contralateral RT was not routinely performed unless in the presence of contralateral pathological nodal metastases or when the tumor reached or crossed the midline.15 Concomitant chemoradiotherapy (CCRT) with cisplatin-based agents (30 mg/m2 weekly or a biweekly cisplatin/tegafur/leucovorin regimen) was administered to patients with ECS or pathological multiple nodal metastases.16,17

Data Analysis

Follow-ups were continued until June 2008. All patients received follow-up examinations for at least 24 months after surgical treatment or until death. Descriptive statistics are summarized using frequencies, percentages, means, medians, and ranges. The Kaplan–Meier method was used to provide estimates of the 5-year CLNR and overall survival (OS) rates. Statistical significance was tested by log-rank test. Univariate and multivariate analyses (UVA and MVA) were used to identify the independent predictors of CLNR. Independent prognosticators were identified by multivariate Cox regression analysis using the forward selection method. All tests were two-sided, and P values less than 0.05 were considered to be statistically significant.

Results

Patients

Between January 1996 and June 2006, a total of 1,064 consecutive patients with previously untreated OSCC were scheduled for radical surgery in our hospital. One hundred fifty-one patients were excluded due to the presence of midline lesions, bilateral lesions, or second metachronous malignancies. A total of 913 patients were available for the final analysis. There were 852 (93%) males and 61 (7%) females. Median age of study participants was 49 years (range 25–82 years).

Of the 913 patients, 842 (92%) received NDs. Seven hundred forty-eight patients (89%) had ipsilateral ND, and 94 (11%) received bilateral ND. Of the 842 ipsilateral ND, 465 (55%) had elective level I–III ND, 69 (8%) had elective level I–V ND, 181 (22%) had therapeutic level I–III ND, and 121 (14%) had therapeutic I–V ND. Of the 94 contralateral ND, 80 (85%) had elective level I–III ND, 3 (3%) had elective level I–V ND, 10 (11%) had therapeutic level I–III ND, and 1 (1%) had therapeutic I–V ND. Of the 913 patients, 455 (50%) had surgery alone, 304 (33%) had surgery plus RT, and 154 (17%) had surgery plus CCRT. With regard to tumor differentiation, 381 patients (42%) had well-differentiated tumors, 464 (51%) had moderately differentiated tumors, and 68 (7%) had poorly differentiated tumors. The pathological tumor status was as follows: pT1 (n = 163, 18%), pT2 (n = 375, 41%), pT3 (n = 178, 20%), and pT4 (n = 197, 22%). The pathological nodal status was as follows: pNx (no ND, n = 71, 8%), pN0 (n = 501, 55%), pN1 (n = 120, 13%), pN2a (n = 4, 0.4%), pN2b (n = 197, 22%), and pN2c (n = 20, 2%). The pathological stages were as follows: stage I (n = 146, 16%), stage II (n = 249, 27%), stage III (n = 172, 19%), and stage IV (n = 346, 38%).

At time of analyses (June 2008), 607 (67%) patients were alive and 306 (34%) were dead. One hundred thirty-two patients (15%) developed LR, 146 (16%) had neck recurrences, and 98 (11%) had distant metastases. Of the 146 patients with neck recurrence, 91 had ipsilateral neck recurrence, 47 had CLNR, and 8 had bilateral neck recurrence. Of the 55 CLNR, 1 patient had no ND (bilateral neck recurrence), 48 patients had ipsilateral ND (42 CLNR and 6 bilateral neck recurrence), and 6 had bilateral ND (5 CLNR and 1 bilateral neck recurrence).

In the entire study cohort, 5-year CLNR rate was 7% (55/913). Specifically, it was 18% (17/132) in patients with LR, and 5% (38/781) in those without (P = 0.0002) (Fig. 1a). In subjects with LR and CLNR (n = 17), two patients showed CLNR prior to tumor recurrence (4 months and 9 months, respectively). The time period between CLNR and local recurrence for the remaining 15 patients ranged from 0 to 10 months (mean 2.1 months, median 0 months). In subjects with LC and CLNR (n = 38), the time period between CLNR and primary surgery ranged from 1 to 41 months (mean 8.6 months, median 6 months). Specifically, 87% (33/38) of cases occurred within 12 months, and 95% (36/38) within 24 months. The 5-year OS was significantly higher in patients with LC (73%) than in those without (39%) (Fig. 1b) (P < 0.0001).
Fig. 1

a Contralateral neck recurrence (CLNR) and b overall survival (OS) rates in OSCC patients with local recurrence (LR) or local control (LC)

Univariate and Multivariate Analyses of 5-Year CLNR Rate in the LR Group (n = 132)

In the LR group, the following factors were found to be associated with the 5-year CLNR rate in UVA: female gender, tumor subsite, treatment with surgery plus RT/CCRT, tumor depth ≥ 5 mm, perineural invasion, vascular invasion, pathological lymph node metastases, pathological stage III–IV, level IV–V metastases, pN2c, and ECS (Table 1). In MVA, extracapsular spread was the only independent risk factor for 5-year CLNR (P = 0.002, hazard ratio [HR]: 4.957, 95% confidence interval [CI]: 1.763–13.934). The 5-year CLNR rate was significantly higher in patients with ECS (39%) than in those without (12%) (Fig. 2a) (P = 0.0001). Additionally, the 5-year OS was 48% in patients without ECS, whereas it dropped to 16% in those with ECS (P < 0.0001) (Fig. 2b). In patients with ECS, the presence of CLNR did not influence significantly the 5-year OS rate (P = 0.2014).
Table 1

Univariate analyses of risk factors for the 5-year contralateral neck recurrence rate in OSCC patients with either local control (LC group) or local recurrence (LR group)

Risk factors

LC group, total, n/LR group, total, n

LC (n = 781)

LR (n = 132)

Event n (5-year %)

P

Event n (5-year %)

P

Sex

  

0.6530

 

0.0382

    Male (852)

725/127

36 (6)

 

15 (18)

 

    Female (61)

56/5

2 (4)

 

2 (20)

 

Age, years

  

0.3633

 

0.0976

    ≤40 (189)

167/22

6 (4)

 

5 (26)

 

    >40 (724)

614/110

32 (6)

 

12 (18)

 

Subsites

  

0.2562

 

0.0003

    Tongue (358)

310/48

22(8)

 

7 (22)

 

    Mouth floor (28)

26/2

1 (6)

 

1 (50)

 

    Lip (19)

18/1

0 (0)

 

1 (100)

 

    Buccal (315)

274/41

9 (4)

 

6 (22)

 

    Alveolar ridge (112)a

94/18

5 (6)

 

0 (0)

 

    Hard palate (26)

16/10

1 (7)

 

0 (0)

 

    Retromolar (55)

43/12

0 (0)

 

2 (23)

 

Alcohol

  

0.3158

 

0.4443

    No (394)

345/49

14 (4)

 

8 (23)

 

    Yes (519)

436/83

24 (6)

 

9 (15)

 

Betel quid

  

0.2356

 

0.8306

    No (200)

185/15

6 (4)

 

2 (8)

 

    Yes (713)

596/117

32 (6)

 

15 (19)

 

Cigarette

  

0.2114

 

0.6067

    No (132)

115/17

3 (3)

 

3 (18)

 

    Yes (781)

666/115

35 (6)

 

14 (19)

 

Neck dissection

  

0.2282

 

0.3482

    No (71)

64/7

1 (2)

 

0 (0)

 

    Yes (842)

717/125

37 (6)

 

17 (19)

 

Contralateral neck dissection

  

0.7963

 

0.1952

    No (819)

705/114

35 (6)

 

14 (17)

 

    Yes (94)

76/18

3 (4)

 

3 (21)

 

Treatment modality

  

0.0174

 

0.0366

    S (455)

402/53

13 (3)

 

4 (10)

 

    S plus RT/CCRT (458)

379/79

25 (8)

 

13 (25)

 

    S plus RT (304)

248/56

10 (5)

0.0019

8 (22)

0.3122

    S plus CCRT (154)

131/23

15 (15)

 

5 (29)

 

Differentiation

  

<0.0001

 

0.9599

    Well/moderate (845)

722/123

30 (5)

 

16 (19)

 

    Poor (68)

59/9

8 (18)

 

1 (13)

 

Tumor depth (mm)b,d

  

0.0086

  

    <6 (267)

241/

5 (2)

   

    ≥6 (640)

534/

33 (7)

   

Tumor depth (mm)b,d

    

0.0248

    <5 (197)

/20

  

0 (0)

 

    ≥5 (710)

/112

  

17 (23)

 

Close margins (mm)b

  

0.1044

 

0.5489

    ≤4 (75)

58/17

5 (10)

 

1 (8)

 

    >4 (823)

723/100

33 (5)

 

12 (17)

 

Bone marrow invasion

  

0.2489

 

0.8174

    No (781)

679/102

31 (5)

 

14 (19)

 

    Yes (132)

102/30

7 (8)

 

3 (13)

 

Skin invasion

  

0.9067

 

0.2156

    No (854)

732/122

36 (6)

 

15 (17)

 

    Yes (59)

49/10

2 (5)

 

2 (55)

 

Perineural invasionb

  

0.0062

 

0.0213

    No (671)

584/87

22 (4)

 

9 (13)

 

    Yes (240)

196/44

16 (9)

 

8 (31)

 

Vascular invasionb

  

0.4702

 

0.0090

    No (894)

766/128

38 (6)

 

16 (18)

 

    Yes (17)

14/3

0 (0)

 

1 (33)

 

Lymph invasionb

  

0.7065

 

0.3092

    No (857)

736/121

37 (6)

 

15 (18)

 

    Yes (54)

44/10

1 (4)

 

2 (24)

 

Pathological T-status

  

0.2833

 

0.1317

    T1–2 (538)

478/60

21 (5)

 

7 (12)

 

    T3–4 (375)

303/72

17 (7)

 

10 (28)

 

Pathological N statusc

  

<0.0001

 

0.0092

    N0 (572)

501/71

13 (3)

 

6 (10)

 

    N1–2 (341)

280/61

25 (11)

 

11 (29)

 

Pathological stagec

  

0.0021

 

0.0836

    I–II (395)

353/42

9 (3)

 

4 (8)

 

    III–IV (518)

428/90

29 (8)

 

13 (26)

 

Level IV or V metastases

  

0.0197

 

0.0363

    No (888)

760/128

36 (5)

 

16 (18)

 

    Yes (25)

21/4

2 (20)

 

1 (100)

 

pN2c

  

0.4568

 

<0.0001

    No (893)

765/128

38 (6)

 

15 (17)

 

    Yes (20)

16/4

0 (0)

 

2 (67)

 

Extracapsular spread b

  

<0.0001

 

0.0001

    No (709)

612/97

22 (4)

 

8 (12)

 

    Yes (202)

167/35

16 (14)

 

9 (39)

 

    RT group

40/12

2 (2)

0.2950

3 (38)

0.5291

    CCRT group

112/19

13 (13)

 

5 (35)

 

Synchronous ipsilateral second OSCC

  

0.2282

 

0.4965

    No (883)

755/128

38 (6)

 

17 (19)

 

    Yes (30)

26/4

0 (0)

 

0 (0)

 

Tumor crossing the midline

  

0.6492

 

0.0946

    No (846)

728/118

35 (5)

 

14 (17)

 

    Yes (67)

53/14

3 (8)

 

3 (27)

 

S surgery; RT radiotherapy; CCRT concurrent chemoradiation; OSCC oral cavity squamous cell carcinoma

aLower and upper alveolar ridge were combined for the purpose of analysis

bUnavailable data: tumor depth (n = 6), close margin (n = 15), perineural invasion (n = 2), vascular invasion (n = 2), lymph invasion (n = 2), extracapsular spread (n = 2)

cPatients who had no neck dissection (n = 71) were classified as pN0

dBest cutoff value of tumor depth. Tumor depth was defined as the measured thickness from the surface of the normal mucosa to the deepest portion of the tumor

Fig. 2

Contralateral neck recurrence rates in LR patients with and without ECS (a). Overall survival rates in patients with and without ECS (b)

Univariate and Multivariate Analyses of 5-Year CLNR Rate in the LC Group (n = 781)

In the LC group, the following factors were found to be associated with the 5-year CLNR rate in UVA: treatment with surgery plus RT/CCRT, poor differentiation, tumor depth ≥6 mm, perineural invasion, pathological lymph node metastases, pathological stage III–IV, level IV–V metastases, and ECS (Table 1). In MVA, the independent risk factors for the 5-year CLNR rate were tumor subsite (P = 0.026, HR: 1.233, 95% CI: 1.026–1.484), presence of poor differentiation (P = 0.005, HR: 1.053, 95% CI: 1.016–1.092), and pathological lymph node metastases (P < 0.001, HR: 3.817, 95% CI: 1.927–7.557).

Subsite Analysis in the LC Group

The 5-year CLNR in patients with tongue, buccal, and alveolar ridge cancer were 8%, 4%, and 6%, respectively. Table 2 depicts the 5-year CLNR rate according to the tumor subsites (tongue, buccal, and alveolar ridge). In patients with tongue cancer, the following factors were found to be associated with the 5-year CLNR rate in UVA: poor differentiation, tumor depth ≥10 mm, perineural invasion, pathological lymph node metastases, level IV–V metastases, and ECS. In MVA, the independent risk factors for the 5-year CLNR rate were poor differentiation, perineural invasion, and level IV/V lymph node metastases (Table 3).
Table 2

Subsite analyses for the 5-year contralateral neck recurrence rate in OSCC patients with local control

Risk factors

Tongue (n = 310)

Buccal (n = 274)

Alveolar ridge (n = 94)

5-year

P

5-year

P

5-year

P

Sex

 

0.5808

 

0.6510

 

0.5213

    Male

8

 

4

 

6

 

    Female

6

 

0

 

0

 

Age (years)

 

0.5323

 

0.1050

 

0.2192

    ≤40

6

 

0

 

12

 

    >40

8

 

5

 

4

 

Alcohol

 

0.7999

 

0.2525

 

0.4829

    No

7

 

2

 

4

 

    Yes

8

 

5

 

8

 

Betel quid

 

0.2696

 

0.2422

 

0.1949

    No

5

 

0

 

0

 

    Yes

9

 

4

 

8

 

Cigarette

 

0.4271

 

0.3252

 

0.2644

    No

6

 

0

 

0

 

    Yes

8

 

4

 

7

 

Neck dissection

 

0.8938

 

0.3355

 

0.5728

    No

5

 

0

 

0

 

    Yes

8

 

4

 

6

 

Contralateral neck dissection

 

0.3819

 

0.0911

 

0.6016

    No

8

 

3

 

6

 

    Yes

4

 

14

 

0

 

Treatment modality

 

0.0521

 

0.0953

 

0.9386

    S

5

 

2

 

6

 

    S plus RT/CCRT

12

 

6

 

6

 

    S plus RT

7

0.0170

5

0.6503

0

0.0025

    S plus CCRT

22

 

6

 

23

 

Differentiation

 

0.0001

 

0.0034

 

0.3986

    Well/moderate

6

 

3

 

7

 

    Poor

30

 

14

 

0

 

Tumor depth (mm)a

 

0.0243

 

0.0256

  

    <10

4

 

0

   

    ≥10

12

 

6

   

Tumor depth (mm)a

     

0.4707

    <6

    

4

 

    ≥6

    

7

 

Close margins (mm)

 

0.3490

 

0.3903

 

0.5973

    ≤4

16

 

8

 

8

 

    >4

7

 

4

 

5

 

Bone invasion

 

0.8610

 

0.2381

 

0.2594

    No

8

 

3

 

3

 

    Yes

0

 

7

 

8

 

Skin invasion

 

0.7760

 

0.4940

 

0.6789

    No

8

 

3

 

6

 

    Yes

0

 

5

 

0

 

Perineural invasion

 

<0.0001

 

0.2982

 

0.6378

    No

4

 

4

 

5

 

    Yes

19

 

2

 

8

 

Vascular invasion

 

0.9545

 

0.5932

 

0.8103

    No

8

 

4

 

6

 

    Yes

0

 

0

 

0

 

Lymph invasion

 

0.3305

 

0.5699

 

0.1673

    No

8

 

4

 

5

 

    Yes

0

 

0

 

25

 

Pathological T status

 

0.4752

 

0.0030

 

0.4112

    T1–2

8

 

1

 

3

 

    T3–4

6

 

8

 

7

 

Pathological N status

 

0.0169

 

0.0341

 

0.0010

    N0

5

 

2

 

0

 

    N1–2

14

 

7

 

18

 

Pathological stage

 

0.1195

 

0.0063

 

0.1468

    I–II

6

 

0

 

0

 

    III–IV

11

 

7

 

8

 

Level IV or V

 

0.0001

 

0.6835

 

0.7422

    No

7

 

4

 

6

 

    Yes

50

 

0

 

0

 

pN2c

 

0.4635

 

0.8457

 

0.7587

    No

8

 

4

 

6

 

    Yes

0

 

0

 

0

 

Extracapsular spread

 

0.0099

 

0.2654

 

0.0005

    No

6

 

3

 

2

 

    Yes

18

 

6

 

21

 

    RT group

18

0.7151

0

0.3523

0

0.2613

    CCRT group

19

 

7

 

25

 

Synchronous ipsilateral second OSCC

 

0.3601

 

0.6691

 

0.5513

    No

8

 

4

 

6

 

    Yes

0

 

0

 

0

 

Tumor crossing the midline

 

0.6807

 

0.8457

 

0.6919

    No

8

 

4

 

6

 

    Yes

11

 

0

 

0

 

S surgery; RT radiotherapy; CCRT concurrent chemoradiation; OSCC oral cavity squamous cell carcinoma

abest cutoff value of tumor depth

Table 3

Multivariate analyses of risk factors associated with the 5-year contralateral neck recurrence rate (patients with tongue cancer with local control, n = 310)

Characteristic (n)

P value

HR (95% CI)

Differentiation

0.012

 

    Well/moderate (284)

 

Reference category

    Poor (26)

 

1.058 (1.013–1.106)

Perineural invasion

0.001

 

    No (228)

 

Reference category

    Yes (82)

 

4.343 (1.781–10.594)

Level IV or V metastases

0.042

 

    No (301)

 

Reference category

    Yes (9)

 

1.036 (1.001–1.072)

HR hazard ratio; CI confidence interval

In patients with buccal cancer, the following factors were found to be associated with the 5-year CLNR rate in UVA: poor differentiation, tumor depth ≥ 10 mm, pathological T3–4 disease, pathological lymph node metastases, and pathological III–IV stage. In MVA, poor differentiation and the presence of pathological T3–4 disease were independent risk factors for the 5-year CLNR rate (Table 4).
Table 4

Multivariate analyses of risk factors associated with the 5-year contralateral neck recurrence rate (patients with buccal cancer with local control, n = 274)

Characteristic (n)

P value

HR (95% CI)

Differentiation

0.022

 

    Well/moderate (257)

 

Reference category

    Poor (17)

 

1.088 (1.012–1.169)

Pathological T status

0.024

 

    T1–2 (155)

 

Reference category

    T3–4 (119)

 

3.325 (1.175–9.414)

HR hazard ratio; CI confidence interval

In patients with alveolar ridge cancer, the presence of lymph node metastases and ECS were found to be associated with the 5-year CLNR rate in UVA. After allowance for potential confounders, no independent predictor of the 5-year CLNR rate was identified in multivariate analysis.

Prognostic Scoring of Risk Factors by MVA

Each factor identified as an independent prognosticator in MVA was given a score of 1. In patients with tongue cancer (n = 310), the 5-year CLNR rate differed significantly in patients with a score of 0 (3%) compared with those with a score of 1 (15%) or 2–3 (40%) (Fig. 3a). Five-year OS was 82% for patients with a score of 0, 62% for those with a score of 1, and 40% for those with a score of 2–3 (P < 0.0001) (Fig. 3b). In patients with scores of 2–3, presence of CLNR did not influence significantly the 5-year OS rate (P = 0.0891).
Fig. 3

a Contralateral neck recurrence (CLNR) and b overall survival (OS) rates in tongue cancer patients according to the prognostic scoring system

In patients with buccal cancer (n = 274), the 5-year CLNR rate differed significantly in patients with a score of 0 (0%) compared with those with a score of 1–2 (9%, P = 0.0005) (Fig. 4a). The 5-year OS rate was 84% in patients with a score of 0, and 58% in those with a score of 1–2 (P < 0.0001) (Fig. 4b). In patients with scores of 1–2, the presence of CLNR did not influence significantly the 5-year OS rate (P = 0.5575).
Fig. 4

a Contralateral neck recurrence (CLNR) and b overall survival (OS) rates in buccal cancer patients according to the prognostic scoring system

Discussion

Head and neck cancer comprises 4–5% of all malignancies in Taiwan. The oral habits of betel quid chewing, smoking, and drinking are prevalent in our OSCC patients. Oral cavity cancer is currently ranked fourth in cancer incidence and fifth in cancer mortality for male Taiwanese.18 Endemic use of betel quid chewing in Taiwan may account for the different subsite distribution of oral cancer observed in our study as compared with previous reports. Approximately 45–50% of oral cavity cancers in our series originated from areas classified as ICD-145 (buccal, retromolar, and hard palate).18 The incidence of tumors originating from these subsites is higher in our patients compared with that reported in Western countries. Of the tumor subsites in oral cavity, much lower incidence of cancer located at floor of mouth (3%) was noted in this hospital as compared with non-betel nut chewing countries, which was also compatible with the records at Department of Health, Taiwan.18 In the present study, we sought to identify factors associated with CLNR in OSCC patients enrolled in an endemic betel quid chewing area.

Neck control remains an important issue in patients with head and neck cancer. In particular, salvage therapy at the previous surgical site remains a major challenge in the presence of neck recurrence. In general, CLNR is uncommon in head and neck cancer patients at their initial presentation. In this context, elective contralateral ND is generally recommended only when the tumor crosses the midline. However, we have repeatedly observed that CLNR occurs more frequently in patients with local relapse. It is currently unclear whether contralateral neck metastases are underestimated in OSCC patients at initial presentation. Therefore, correct identification of risk factors associated with CLNR is paramount to improve the clinical outcome of this patient group.

In this study, we examined prevalence and risk factors for CLNR in our historical cohort of OSCC patients. Patients were considered at high risk for CLNR when their risk to develop this recurrence was >20% within 2 years after radical surgery. We used this threshold based on previous data suggesting that elective ND should be performed only when the neck lymph node metastasis rate is higher than 20%.19 The correct identification of factors associated with CLNR might allow their incorporation into a risk stratification scheme aiming to improve the patient outcome.

To the best of our knowledge, no consensus has been achieved on the use of contralateral neck treatment in OSCC patients.2, 3, 4,8, 9, 10 It has been suggested that T3–4 tumor or tumor subsites should be carefully considered in the planning of contralateral ND, regardless of the presence of ipsilateral pathologic lymph nodes.1, 2, 3, 4, 5 In contrast, other authors have suggested that the presence of ipsilateral pathologic lymph nodes is of greater clinical significance than tumor subsite.2,4,8 Lymphatic drainage crossing the midline should be carefully considered for establishing the risk of CLNR. In this regard, Kowalski et al. have provided evidence that contralateral metastases occur more frequently when cancer extends within 1 cm of the oral cavity midline.3 Moreover, studies of lymphatic drainage from the oral cavity have shown that lymphatics readily cross the midline.20 Consequently, some institutions treat the contralateral neck when the tumor extends to within 1 cm of the midline. In this study, however, the treatment of contralateral ND was pursued when the tumor reached or crossed the midline rather than when it extended within 1 cm of the midline. In addition, elective contralateral neck RT was pursued in all patients who were scheduled for RT in the presence of tumors that reached or crossed the midline. For a tailor-made treatment plan, contralateral neck radiation was not performed routinely in our patients with buccal cancer.15 Previous studies on CLNR in OSCC patients identified tongue cancer and mouth floor cancer as risk factors for CLNR. However, patients with LR were excluded from previous investigations. In this study, we were able to show that both LR and tongue cancer are significantly associated with higher CLNR rates. We therefore examined whether specific risk factors may influence the risk of CLNR in patients with or without LR and tongue cancer.

Local recurrence was an independent risk factor for CLNR in our study cohort. We thus stratified our patients according to the presence or absence of LR. In patients with LR, only ECS was identified as an independent predictor of 5-year CLNR rate. Interestingly, patients with LR and ECS showed significantly higher 5-year CLNR (39% versus 12%, P = 0.0001) and OS rates compared with those without (48% versus 16%, P < 0.0001). Among patients with ECS at their initial presentation (n = 202), 25 patients (12%) had CLNR (16 patients were in the local control group, and 9 in the local recurrence group). It is our current policy to treat patients with ECS with ipsilateral ND plus adjuvant therapy. Patients in this group are followed up for LR by using neck US every 3 months until 24 months postoperatively. Once LR occurs, these patients are treated with contralateral ND plus adjuvant therapy.

Tumor subsite, poor differentiation, and presence of pN + disease were significant predictors of CLNR in patients with local control. Further analyses allowed the identification of prognosticators for CLNR in patients with tongue cancer. In this patient group, the independent risk factors for the 5-year CLNR rate were poor differentiation, perineural invasion, and level IV/V lymph node metastases. A prognostic scoring system was thus formulated by summing up the three significant factors identified by multivariate analysis. In order to reduce the incidence of CLNR, contralateral ND and adjuvant therapy are recommended in high-risk patients with tongue cancer (score 2–3, 5-year CLNR 40%). In the intermediate-risk group (score 1, 5-year CLNR 15%), neck US examinations are recommended every 3 months until 24 months postoperatively. Observation should be considered sufficient for low-risk patients (score 0, 5-year CLNR 3%).

Poor differentiation and the presence of pathological T3-4 disease were independent risk factors for the 5-year CLNR rate in patients with buccal cancer. However, the incidence of 5-year CLNR was low even in the high-risk group (9%), and it dropped to zero in low-risk patients. Neck US examinations every 3 months until 24 months postoperatively are thus recommended in the high-risk group, whereas observation should be sufficient in low-risk individuals.

The results of the Radiation Therapy Oncology Group (RTOG) (#9501) and European Organization for Research and Treatment of Cancer (EORTC) (#22931) trials have provided evidence that—in patients with head and neck cancer—surgery plus CCRT had a better impact on clinical outcome compared with surgery plus RT.21,22 The benefits of CCRT were especially evident in head and neck cancer patients with positive margins and ECS.23 In our study, a total of 202 patients (22%) had ECS, and 10 subjects (1.1%) had positive margins (≤1 mm). No positive impact of surgery plus CCRT compared with surgery plus RT was found for CLNR (Tables 1 and 2). There are two possible explanations for these results. The first reason can be the retrospective nature of our study. Secondly, it should be noted that only 25% of patients in the RTOG (#9501) and EORTC (#22931) trials had oral cavity cancer, compared with 100% in our study. Notably, 80% of our patients had a long history of betel quid chewing. Altogether, these differences between our study and previous trials [RTOG (#9501) and EORTC (#22931)] can explain at least in part the observed discrepancies.

Some limitations are inherent in the present study. Firstly, our results were obtained in a cohort of OSCC patients treated with radical surgery and enrolled in an endemic betel quid chewing area. Secondly, tumor subsites were not homogenously distributed in our cohort, with the majority of patients having tongue (n = 358), buccal (n = 315) or alveolar ridge (n = 112) cancer. We thus limited our analysis to these three main tumor sites. Thirdly, this research was based on a treatment philosophy that does not account for lymphatic drainage crossing the midline. We were thus unable to analyze the risk associated with the lesions located within 1 cm from the midline due to the retrospective nature of our investigation.

Based on our current findings, we have developed a treatment flowchart aiming to reduce the incidence of CLNR in OSCC patients (Fig. 5). Further studies are warranted to test the clinical effectiveness of our management strategy.
Fig. 5

Treatment flowchart recommendation, aiming to reduce the incidence of CLNR in OSCC patients

Notes

Acknowledgements

This study was supported by grants NMRPG160031, CMRPG361051, and CMRPG370061 from the Chang Gung Memorial Hospital.

Conflicts of interest statement

The authors indicated no potential conflicts of interest.

Authors’ contributions

Conception and design: C.T. Liao, T.C. Yen.

Funding support: NMRPG160031, CMRPG361051, and CMRPG370061.

Administrative support: T.C. Yen.

Provision of study materials or patients: C.T. Liao, S.F. Huang, I.H. Chen, J.T.C. Chang, H.M. Wang, S.H. Ng, C. Hsueh, L.Y. Lee, C.H. Lin, A.J. Cheng, T.C. Yen.

Collection and assembly of data: C.T. Liao, S.F. Huang, I.H. Chen, J.T.C. Chang, H.M. Wang, S.H. Ng, C. Hsueh, L.Y. Lee, C.H. Lin, I.H. Chen, S.F. Huang, A.J. Cheng, T.C. Yen.

Data analysis and interpretation: C.T. Liao, T.C. Yen.

Manuscript writing: C.T. Liao, T.C. Yen.

Final approval of manuscript: C.T. Liao, T.C. Yen.

References

  1. 1.
    Woolgar JA. Histological distribution of cervical lymph node metastases from intraoral/oropharyngeal squamous cell carcinomas. Br J Oral Maxillofac Surg. 1999;37:175–80.CrossRefPubMedGoogle Scholar
  2. 2.
    Koo BS, Lim YC, Lee JS, et al. Management of contralateral N0 neck in oral cavity squamous cell carcinoma. Head Neck. 2006;28:896–901.CrossRefPubMedGoogle Scholar
  3. 3.
    Kowalski LP, Bagietto R, Lara JR, et al. Factors influencing contralateral lymph node metastasis from oral carcinoma. Head Neck. 1999;21:104–10.CrossRefPubMedGoogle Scholar
  4. 4.
    Kurita H, Koike T, Narikawa JN, et al. Clinical predictors for contralateral neck lymph node metastasis from unilateral squamous cell carcinoma in the oral cavity. Oral Oncol. 2004;40:898–903.CrossRefPubMedGoogle Scholar
  5. 5.
    Lim YC, Lee JS, Koo BS, et al. Treatment of contralateral N0 neck in early squamous cell carcinoma of the oral tongue: elective neck dissection versus observation. Laryngoscope. 2006;116:461–5.CrossRefPubMedGoogle Scholar
  6. 6.
    Spiro RH, Alfonso AE, Farr HW, et al. Cervical node metastasis from epidermoid carcinoma of the oral cavity and oropharynx. A critical assessment of current staging. Am J Surg. 1974;128:562–7.PubMedGoogle Scholar
  7. 7.
    Kowalski LP, Bagietto R, Lara JR, et al. Prognostic significance of the distribution of neck node metastasis from oral carcinoma. Head Neck. 2000;22:207–14.CrossRefPubMedGoogle Scholar
  8. 8.
    Chow TL, Chow TK, Chan TT, et al. Contralateral neck recurrence of squamous cell carcinoma of oral cavity and oropharynx. J Oral Maxillofac Surg. 2004;62:1225–8.CrossRefPubMedGoogle Scholar
  9. 9.
    González-García R, Naval-Gías L, Sastre-Pérez J, et al. Contralateral lymph neck node metastasis of primary squamous cell carcinoma of the tongue: a retrospective analytic study of 203 patients. Int J Oral Maxillofac Surg. 2007;36:507–13.CrossRefPubMedGoogle Scholar
  10. 10.
    González-García R, Naval-Gías L, Rodríguez-Campo FJ, et al. Contralateral lymph neck node metastasis of squamous cell carcinoma of the oral cavity: a retrospective analytic study in 315 patients. J Oral Maxillofac Surg. 2008;66:1390–8.CrossRefPubMedGoogle Scholar
  11. 11.
    Fleming ID, Cooper JS, Henson DE, Hutter RVP, Kennedy BJ, Murphy GP, et al., editors. AJCC cancer staging manual. 5th ed. Philadelphia: Lippincott-Raven; 1997.Google Scholar
  12. 12.
    Greene FL, Page DL, Fleming ID, Fritz AG, Balch CM, editors. AJCC cancer staging manual. 6th ed. New York: Springer-Verlag; 2002.Google Scholar
  13. 13.
    Liao CT, Chang JT, Wang HM, et al. Surgical outcome of T4a and resected T4b oral cavity cancer. Cancer. 2006;107:337–44.CrossRefPubMedGoogle Scholar
  14. 14.
    Liao CT, Ng SH, Chang JT, et al. T4b oral cavity cancer below the mandibular notch is resectable with a favorable outcome. Oral Oncol. 2007;43:570–9CrossRefPubMedGoogle Scholar
  15. 15.
    Lin CY, Lee LY, Huang SF, et al. Treatment outcome of combined modalities for buccal cancers. Int J Radiat Oncol Biol Phys. 2008;70:1373–81.CrossRefPubMedGoogle Scholar
  16. 16.
    Bachaud JM, Cohen-Jonathan E, Alzieu C, David JM, Serrano E, Daly-Schveitzer N. Combined postoperative radiotherapy and weekly cisplatin infusion for locally advanced head and neck carcinoma: final report of a randomized trial. Int J Radiat Oncol Biol Phys. 1996;36:999–1004.CrossRefPubMedGoogle Scholar
  17. 17.
    Wang HM, Wang CS, Chen JS, et al. Cisplatin, tegafur, and leucovorin: a moderately effective and minimally toxic outpatient neoadjuvant chemotherapy for locally advanced squamous cell carcinoma of the head and neck. Cancer. 2002;94:2989–95.CrossRefPubMedGoogle Scholar
  18. 18.
    Taiwan cancer registry, 2008 annual report. Available from: http://crs.cph.ntu.edu.tw/. Accessed October 27, 2008.
  19. 19.
    Weiss MH, Harrison LB, Isaacs RS. Use of decision analysis in planning a management strategy for the stage N0 neck. Arch Otolaryngol Head Neck Surg. 1994;120:699–702.CrossRefPubMedGoogle Scholar
  20. 20.
    Werner JA, Dunne AA, Myers JN. Functional anatomy of the lymphatic drainage system of the upper aerodigestive tract and its role in metastasis of squamous cell carcinoma. Head Neck. 2003;25:322–32.CrossRefPubMedGoogle Scholar
  21. 21.
    Cooper JS, Pajak TF, Forastiere AA, et al. Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head and neck. N Engl J Med. 2004;350:1937–44.CrossRefPubMedGoogle Scholar
  22. 22.
    Bernier J, Domenge C, Ozsahin M, et al. Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer. N Engl J Med. 2004;350:1945–52.CrossRefPubMedGoogle Scholar
  23. 23.
    Bernier J, Cooper JS, Pajak TF, et al. D efining risk levels in locally advanced head and neck cancers: a comparative analysis of concurrent postoperative radiation plus chemotherapy trials of the EORTC (#22931) and RTOG (# 9501). Head Neck. 2005;27:843–50.CrossRefPubMedGoogle Scholar

Copyright information

© Society of Surgical Oncology 2008

Authors and Affiliations

  • Chun-Ta Liao
    • 1
    • 2
  • Shiang-Fu Huang
    • 1
    • 2
  • I-How Chen
    • 1
    • 2
  • Joseph Tung-Chieh Chang
    • 2
    • 3
  • Hung-Ming Wang
    • 2
    • 4
  • Shu-Hang Ng
    • 2
    • 5
  • Chuen Hsueh
    • 2
    • 6
  • Li-Yu Lee
    • 2
    • 6
  • Chih-Hung Lin
    • 2
    • 7
  • Ann-Joy Cheng
    • 2
    • 8
  • Tzu-Chen Yen
    • 2
    • 9
  1. 1.Department of Otorhinolaryngology, Head and Neck SurgeryChang Gung Memorial Hospital and Chang Gung UniversityTaoyuanTaiwan, ROC
  2. 2.Department of Head and Neck Oncology GroupChang Gung Memorial Hospital and Chang Gung UniversityTaoyuanTaiwan, ROC
  3. 3.Department of Radiation OncologyChang Gung Memorial Hospital and Chang Gung UniversityTaoyuanTaiwan, ROC
  4. 4.Department of Hema-OncologyChang Gung Memorial Hospital and Chang Gung UniversityTaoyuanTaiwan, ROC
  5. 5.Department of Diagnostic RadiologyChang Gung Memorial Hospital and Chang Gung UniversityTaoyuanTaiwan, ROC
  6. 6.Department of PathologyChang Gung Memorial Hospital and Chang Gung UniversityTaoyuanTaiwan, ROC
  7. 7.Department of Plastic and Reconstructive SurgeryChang Gung Memorial Hospital and Chang Gung UniversityTaoyuanTaiwan, ROC
  8. 8.Department of Medical Biotechnology, Biostatistics Consulting Center/Department of Public HealthChang Gung Memorial Hospital and Chang Gung UniversityTaoyuanTaiwan, ROC
  9. 9.Nuclear Medicine and Molecular Imaging CenterChang Gung Memorial Hospital and Chang Gung UniversityTaoyuanTaiwan, ROC

Personalised recommendations