Significant invasion depth of early oral tongue cancer originated from the lateral border to predict regional metastases and prognosis

Int. J. Oral Maxillofac. Surg. 2009; 38: 653–660 doi:10.1016/j.ijom.2009.01.004, available online at http://www.sciencedirect.com

Clinical Paper Head and Neck Oncology

Significant invasion depth of early oral tongue cancer originated from the lateral border to predict regional metastases and prognosis

J. Junga, N. H. Chob, J. Kimc, E. C. Choia, S. Y. Leed, H. K. Byeona, Y. M. Parka, W. S. Yanga, S.-H. Kima a Department of Otorhinolaryngology, Yonsei University College of Medicine, Seoul, Korea; b Department of Pathology, Yonsei University College of Medicine, Seoul, Korea; c Department of Radiology, Yonsei University College of Medicine, Seoul, Korea; d Department of Otorhinolaryngology, ChungAng University College of Medicine, Seoul, Korea

J. Jung, N. H. Cho, J. Kim, E. C. Choi, S. Y. Lee, H. K. Byeon, Y. M. Park, W. S. Yang, S. -H. Kim: Significant invasion depth of early oral tongue cancer originated from the lateral border to predict regional metastases and prognosis. Int. J. Oral Maxillofac. Surg. 2009; 38: 653–660. # 2009 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved. Abstract. In oral tongue cancer, tumor depth is crucial for cervical lymph node metastasis. There is no standardized method to predict tumor invasion or deciding who should undergo selective neck dissection. In this study, calculated MRI invasion depth was compared with histopathologic (HP) invasion depth to find a correlation, and determine a cutoff value of invasion depth that predicts occult neck node metastasis. 50 patients, diagnosed with T1 or T2 oral tongue cancer originating from the lateral border of the tongue, underwent MRI screening and received surgical excision as primary treatment. MRI and HP invasion depths were compared and the cutoff value determined. The invasion depth to determine the presence of nodal metastasis where summation of specificity and sensitivity was greatest was 8.5 mm HP, 10.5 mm in T1 weighted enhanced axial image, and 11.5 mm in T2 weighted MRI axial image. The relation coefficient of T2 weighted MRI invasion depth and HP depth was 0.851, and accuracy 84%, all of which showed higher correlation compared with T1 weighted enhanced axial image. HP depth was significantly correlated with survival rate. The measurement of invasion depth using MRI is a prerequisite for determining a surgical plan in early oral tongue cancer.

Apart from the lip, the most common site of cancer development in the oral cavity is the oral tongue. Oral tongue carcinomas exhibit neck node metastasis more than any other carcinomas3,5,21. It has been 0901-5027/060653 + 08 $36.00/0

reported that at the time of diagnosis, 40% of all tongue cancer patients have neck metastasis and 40% of stage 2 lesions show occult metastasis13. The cure rate decreases by 50% if there is neck node

Keywords: tongue cancer; invasion; depth; regional metastases; prognosis. Accepted for publication 19 January 2009 Available online 23 February 2009

metastasis and the single most important prognostic factor in oral tongue carcinoma is neck node metastasis12. For early tongue carcinoma at stages T1 or T2, the mainstay of treatment is surgical

# 2009 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.

654

Jung et al.

resection. This generally involves supraomohyoid neck dissection (SOND) as the elective neck dissection (END) in stage I or II carcinoma, mainly because there is a great chance of occult metastasis and END may improve neck control and the overall survival rate compared with observation alone6,9. There is a strong relationship between depth and lymph node metastasis1,2,14,15,19,25. It has been reported, that of the factors influencing lymph node metastasis in oral tongue carcinoma, tumor depth is more important than tumor size1,16. YUEN et al.25 reported that of the factors such as tumor width, diameter, area, volume and depth, which affect lymph node metastasis, local recurrence and survival rate, only depth was significant. Controversies remain about the parameters of histopathologic (HP) depth in determining the existence of occult metastasis and the necessity of performing selective neck dissection4,10.22,23. A preoperative imaging method for tumor depth that can indirectly show the possibility of occult metastasis is required for oral tongue carcinoma. MRI is useful for measuring tumor depth. According to IWAI et al.7, there is a strong correlation between MRI depth and the HP depth of oral tongue carcinoma and it is possible to predict nodal metastasis before surgery on the basis of tumor depth. LAM et al.11 reported that T1-weighted MRI imaging shows a strong correlation with HP depth. There is no uniform method to assess the invasion depth of tongue cancer on MRI. The object of this study is to assess the cutoff value of the MRI and HP invasion depth that indicates the existence of lymph node metastasis of tongue cancer and to calculate the disease-free survival and overall survival rate according to depth. The clinical significance of tumor invasion depth on the MRI was investigated by evaluating the correlation and the accuracy between MRI and HP depth of oral tongue carcinoma. Materials and methods

Retrospective analysis was carried out on 50 patients who were diagnosed as having T1 or T2 early oral tongue squamous cell carcinoma originating from the lateral border of the tongue and who consequently received surgery as primary treatment after having undergone preoperative MRI from January 2002 to December 2005. Patients with ventral or floor of the mouth tongue cancer were excluded from the study. Patients whose MRI images were not retrievable on the picture

Fig. 1. MRI depth was measured from deepest tumor invasion to the presumed original surface level ignoring exophytic growth.

archiving and communication system were excluded as were those at stage cM1. The male to female ratio was 26:24 and the mean age was 52  12 years. In terms of TNM staging, 13 patients were in stage I, 19 in stage II, 7 in stage III, and 11 in stage IV. In 16 patients, the cancer was identified as pathologic stage T1 (pT1) and in 34 as pathologic T2 (pT2). There were 18 patients with positive pathologic node (pN+) and 32 with negative pathologic node (pN0). Of the 18 patients with pN+, 7 were pN1, 10 were pN2b and 1 was pN2c. 46 patients with negative clinical node (cN0) or cN1 received partial glossectomy with unilateral or bilateral SOND, and 4 patients who were cN2b received partial glossectomy with radical neck dissection or modified radical neck dissection with contralateral SOND followed by postoperative radiotherapy. Patients who received total glossectomy due to midline extension of the tumor were also excluded. The mean period of follow up was 24 months. All MRI examinations were performed using the 1.5-T system (Intera: Philips Medical Systems, Best, the Netherlands) with Synergy Head/Neck coils. Gadolinium-enhanced T1 weighted images were achieved with settings of TR (repetition time) 500 ms, TE (echo time) 12 ms at the coronal plane, section thickness 3 mm, field of view (FOV) 23 cm and acquisition

matrix 256  256. Invasion depth on MRI was defined as tumor thickness, measured as follows. The reference line was determined as a horizontal line connecting the mucosal junction of the tumor and the length perpendicular to this line towards the deepest point of tumor infiltration was measured (Fig. 1). If the tumor was ulcerative, the reference line was determined in the same way to be considered as the presumed original surface level, exophytic lesions were ignored and length measurement was simplified to represent invasion ability. Invasion depth was assessed by the same radiologist on gadolinium-enhanced T1 weighted image and T2 weighted image. The surgical specimen was fixed in formalin, embedded in paraffin, stained with hematoxylin-eosin, a 3-mm slice was then made into a slide. Using a X12.5 microscope, maximal tumor depth was measured in the same way as for MRI. The cutoff value of invasion depth that can best indicate the existence of nodal metastasis was determined as the highest summation value of sensitivity and specificity on the basis of a receiver operation characteristic curve. With this value, tumor invasion depth was divided into two groups and correlation with nodal metastasis was evaluated using X2. Correlation between HP depth and MRI depth was calculated by Pearson correlation test and applying the standards given by YUEN

Significant invasion depth of early oral tongue cancer originated

655

et al.25 the values of HP depth were categorized into three groups and the accuracy of MRI depth measurement was evaluated. To investigate factors affecting disease-specific recurrence and death, disease-free survival rate and overall survival rate according to T stage, TNM stage, node metastasis, MRI depth and HP depth were calculated using the Kaplan–Meier method. Results

Of the 50 patients, 42 were in negative clinical node (cN0). In 11 of the 42 patients, positive metastatic nodes were reported, the occult metastasis rate was 26%. Of the 8 patients with a positive clinical node (cN+), 1 had a negative pathologic node. Invasion depth and nodal metastasis

The mean values of HP depth were significantly different between the 32 pN0 patients and the 18 pN+ patients; 8.0  4.5 mm and 12.4  3.9 mm, respectively (p = 0.001) (Fig. 2A). Gadolinium-enhanced T1 weighted MRI depths (T1WGd MRI depth) were significantly different between pN0 and pN+ groups, the mean values of which were 7.9  5.2 mm and 12.2  4.0 mm, respectively (p = 0.005) (Fig. 2B). The mean values of T2 weighted MRI depth (T2W MRI depth) were also significantly different between the two groups; 8.2  5.3 mm and 12.5  3.9 mm, respectively (p = 0.005) (Fig. 2C). Cutoff values for HP depth and MRI depth

Fig. 2. Tumor invasion depth according to nodal metastasis. (A) HP depth was significantly different according to nodal metastasis (p = 0.001) (B), (C) T1WGd MRI depth and T2W MRI depth were significantly different according to nodal metastasis (p = 0.005).

The cutoff value of HP depth that could determine the existence of nodal metastasis was 8.5 mm. The cutoff values for T1WGd MRI depth and T2W MRI depth were 10.5 mm and 11.5 mm, respectively. With the HP depth cutoff value of 8.5 mm as a standard, groups were subdivided into those >9 mm and those <9 mm; the nodal metastasis rates for each group were 53% and 10%, respectively (p = 0.002). As for the T1WGd MRI depth with a cutoff value of 10.5 mm, the nodal metastasis rate in the group with values >11 mm was 52%, and for those <11 mm was 24%, both of which were significantly different (p = 0.040). Similar results were observed for the T2W MRI depth with a cutoff value of 11.5 mm, the nodal metastasis rate in the groups >12 mm and <12 mm were 52% and 24%, respectively (p = 0.040). There was no significant difference in nodal metastasis rate between the cT1 and cT2 groups, the values were

656

Jung et al.

Table 1. Nodal metastasis rate according to invasion depth. Results according to HP depth with cutoff 9 mm HP depth <9 mm

HP depth 9 mm

N0 18 N+ 2 * x2 test: p = 0.002 - HP depth <9 mm: nodal metastasis rate = 10% - HP depth 9 mm: nodal metastasis rate = 53% Results according to T1WGd MRI depth with cutoff 11 mm T1WGd depth <11 mm N0 22 N+ 7 * x2 test: p = 0.040 - T1WGd depth <11 mm: nodal metastasis rate = 24% - T1WGd depth 11 mm: nodal metastasis rate = 52% Results according to T2W MRI depth with cutoff 12 mm T2W depth <12 mm N0 22 N+ 7 * x2 test: p = 0.040  T2W depth <12 mm: nodal metastasis rate = 24%  T2W depth 12 mm: nodal metastasis rate = 52%

14 16 *

T1WGd depth 11 mm 10 11 *

T2W depth 12 mm 10 11 *

Results according to T stage N0 N+

T1

T2

13 4

19 14 *

* x2 test: p = 0.187.  T1: nodal metastasis rate = 24%.  T2: nodal metastasis rate = 42%.

24% and 42%, respectively (p = 0.187) (Table 1). Correlation between HP depth and MRI depth

Pearson correlation coefficient of HP depth and T1WGd MRI depth was 0.851 (p < 0.001) and the Pearson correlation coefficient of HP depth and T2W MRI depth was 0.813 (p < 0.001) suggesting that HP depth shows stronger correlation with T2W MRI depth than with T1WGd MRI depth (Figs. 3, 4). The least

squares regression equation was Y = 0.917X + 0.696 for T1WGd MRI depth and Y = 0.883X + 1.304 for T2W MRI depth. HP depth was categorized into three groups and MRI measurement accuracy was examined for each, the results were 84% for T1WGd image and 80% for T2W, showing that T1WGd was more accurate than T2W (Table 2). Survival rate

Disease-free survival rate and overall survival rate were calculated according to T

stage, TNM stage, node metastasis, HP depth and MRI depth. Applying the Kaplan–Meier method, disease-free survival and overall survival were calculated with N = 50. There were no significant differences of disease-free survival for T stage (p = 0.228), node metastasis (p = 0.769), T1WGd MRI depth with cutoff value 11 mm (p = 0.635) and T2W MRI depth with cutoff value 12 mm (p = 0.233). Disease-free survival rate was significant only for HP depth with cutoff value 9 mm, where the group <9 mm was 82% and the group >9 mm was 59% (p = 0.042) (Fig. 5). For the overall survival rate, there were no significant differences for T stage (p = 0.105), node metastasis (p = 0.170), T1WGd MRI depth with cutoff value 11 cm (p = 0.645) and T2W MRI depth with cutoff value 12 cm (p = 0.080) but for HP depth the overall survival rate was 93% in the group <9 mm and 60% in the group >9 mm (p = 0.023) (Fig. 6). Nodal staging considering invasion depth

Occult node metastasis according to the clinical N staging system that takes invasion depth into account was evaluated (Table 3). Clinical N0 patients were divided into two groups; N0a: T1WGd depth <11 mm, N0b: T1WGd depth 11 mm. N0a patients were determined to be a clinically low risk group and N0b patients as a high risk group for nodal metastasis. Similar to the clinical N stage, pathologic N stage was further classified according to the HP depth cutoff value 9 mm; pN0a: HP depth <9 mm, pN0b: HP depth 9 mm. There was no difference between N0 and N+ in overall survival rate according to the original pathologic N

Fig. 3. Scatterplot shows tumor depth as determined from pathologic specimens compared with tumor depth as determined from contrastenhanced T1-weighted images (T1WGd images).

Significant invasion depth of early oral tongue cancer originated

657

Fig. 4. Scatterplot shows tumor depth as determined from pathologic specimens compared with tumor depth as determined from T2-weighted images (T2W images).

Table 2. MRI invasion depth accuracy. Pathologic tumor Invasion depth

T1WGd 3 mm (n = 10) >3 mm, but 9 mm (n = 14) >9 mm (n = 26) Overall accuracy (%)

staging system. In overall survival rate a significant difference between pN0a and pN0b plus pN was noted, of 92% versus 62% (Fig. 7).

MRI invasion depth T2W

No.

%

No.

%

7 11 24

70 79 92

7 10 23

70 71 88

84

80

HP depth was divided into three groups and MRI measurement accuracy was examined for each, the results were 84% for T1WGd image and 80% for T2W, showing that T1WGd was more accurate than T2W.

Fig. 5. Tongue cancer disease-free survival rate related to HP depth with cutoff value = 9 mm. Disease survival rate was 82% in HP depth <9 mm group and 59% in HP depth 9 mm group (p = 0.042).

Discussion

Nodal metastasis is the most important prognostic factor in oral tongue carcinoma7,12. The most common cause of surgical treatment failure in oral tongue carcinoma is nodal recurrence25. In this study, of the 14 patients with recurrence, ipsilateral node recurrence was noted in 10 patients (71%). The most relevant factor in nodal metastasis is the invasion depth of tongue cancer, but there is controversy about the standard value of depth that distinguishes nodal metastasis. There is no verified preoperative study to measure tumor invasion depth exactly, therefore it is difficult to apply tumor depth in the AJCC (American Joint Committee on Cancer) TNM staging system. Tumor invasion depth is excluded in the TNM staging system despite its importance in disease prognosis, therefore a new revised staging system has been proposed in the present study. LAM et al.11 reported that preoperative MRI is a good measurement for estimating the invasion depth of oral tongue carcinoma. In this study, the correlation between T1WGd MRI depth and HP depth was found to be strong with a Pearson correlation coefficient of 0.851. Classifying into three groups according to HP depth (<3, 3–9, >9 mm), MRI depth values showed accuracies over 80% in T1WGd depth and T2W MRI depth (Table 2). The correlation with HP depth was stronger and the values more accurate for T1WGd depth than for T2W MRI depth, probably because of the tendency

658

Jung et al.

Fig. 6. Tongue cancer overall survival rate related to HP depth with cutoff value = 9 mm. Overall survival rate was 93% in HP depth <9 mm group and 60% in HP depth 9 mm group (p = 0.023).

to overestimation in T2W images due to inflammation and edema.11 The methods of invasion depth measurement proposed by LAM et al.11 included exophytic tumor growth, this was excluded in the present study and only the invasion depth from the original surface was measured, mainly so that only the invasion potency of the tumor was considered. The reason for using the original

surface as the baseline for ulcerative tumors was to reflect the invasion potency accurately, starting from the epithelial cell lining of the tongue. Significant differences were noted according to the status of nodal metastasis at all HP depths, T1WGd MRI depth and T2W MRI depth (Fig. 2). This shows that preoperative MRI is a valuable tool for predicting nodal metastasis. The cutoff

values of invasion depth that can discern nodal metastasis were 9 mm for HP depth, 11 mm for T1WGd MRI depth and 12 mm for T2W depth. The fact that the cutoff value of T1WGd depth is 2 mm larger than the HP depth can be explained by the difference made due to the shrinkage and distortion of the surgical specimen8,20. Significant differences in nodal metastasis rate were observed according to each cutoff values, but the difference between T1 and T2 stages was not statistically significant. It can be speculated that the depth of tumor is more important than the size as a prognostic factor in nodal metastasis. Particularly in terms of HP depth, dividing the group under the cutoff value of 9 mm, the nodal metastasis rate was merely 10% in the group <9 mm compared with over 50% in the group >9 mm. The cutoff value of 9 mm in HP depth is higher than the values reported in previous studies. BYERS et al.3 reported that a muscular invasion depth exceeding 4 mm had more possibility to metastasize as a prognostic factor in oral tongue cancer, and SPIRO et al.23 set 2 mm as a cutoff value. O’BRIEN et al.17 reported that difference in the rate of survival and nodal metastasis was present with 4 mm. It is difficult to compare studies, because previous studies have included tumors originating from various anatomical lesions such as ventral and dorsal tongue, base of tongue and floor of mouth, the mean invasion depth in the group with nodal metastasis could be lower than the present study in which only tongue cancer of lateral tongue origin was included. The present study included only those patients in whom nodal metastasis

Table 3. Occult cervical lymph node metastasis.

Clinical N0 patients were divided into two groups; N0a: T1WGd depth <11 mm, N0b: T1WGd depth 11 mm. N0a patients were determined as a clinically low risk group and N0b patients as a high risk group for nodal metastasis.

Significant invasion depth of early oral tongue cancer originated

Fig. 7. Overall survival rate in revised pathologic N staging system. (A) Overall survival rate in original N staging system: no difference between N0 and N+. (B) Overall survival rate in revised pathologic N staging system: significant difference between pN0a and pN0b plus pN+ (p = 0.039).

was confirmed pathologically using surgical specimen obtained after partial glossectomy and selective neck dissection in cN0 patients, whereas the studies of O’BRIEN and SPIRO study included cases of nodal recurrence after initial surgical treatment, which can lead to a lower cutoff value compared with the present study. It

can be considered that an HP depth of 9 mm is a clinically important cutoff value in early oral tongue cancer originating from the lateral border. ODELL et al.18 reported that the pattern of invasion was the closest correlation with metastasis and recurrence in this type of carcinoma. The present study showed

659

that the group in which the cancer had a tentacular border (finger like invasion border) (N = 19) had a statistically significant rise in the nodal metastasis rate compared with a pushing border (N = 31), this was related to the invasion pattern (p = 0.026). Even if the invasion pattern is an important factor in determining survival rate and nodal metastasis rate, these pathologic factors can not be evaluated preoperatively in the same way as MRI depth measurements, and are not appropriate for use as factors in determining treatment. Investigating the possibility of occult nodal metastasis before surgery is important because it could be the cause of surgical failure. KUROKAWA et al.10 reported that if the depth of tumor invasion exceeds 5 mm, there would be a greater chance of occult cervical lymph node metastasis. The occult nodal metastasis rate of oral tongue cancer in this study was 26%. To decrease the occult nodal metastasis rate, nodal staging taking into account invasion depth was suggested in this study. Occult nodal metastasis according to the revised clinical N staging system decreased to 7% compared with 26% in the AJCC TNM staging system. SPIRO et al.23 claimed that in oral tongue carcinoma, the increase in tumor thickness is related more to treatment failure and low survival rate than the increase in tumor size, and that disease-related death is uncommon in patients with an invasion depth <2 mm. Calculating the diseasefree survival and the overall survival rate, the results revealed that only HP depth was statistically significant. MRI depth did not show a statistically significant relationship with the survival rate. When the authors divided the cases using a parameter of 11 mm in MRI T1Gd depth, the overall survival rate of the group <11 mm was 82%, compared with 70% observed in the other group, though there was no statistical significance (p = 0.170). Similar to the result of SPIRO et al.23, there was no significant difference in survival rate between T1 and T2 (disease free survival rate: 82% versus 66%, p = 0.228; overall survival rate 94% versus 72%, p = 0.105). There was no difference in survival rate according to the status of nodal metastasis, although nodal metastasis is one of the most important prognostic factors of tongue cancer (disease free survival: 72% versus 72%, p = 0.769; overall survival rate: 81% versus 67%, p = 0.170). As a result of a pathologic review of nodal metastasis of 18 pN+ patients, no extracapsular spread (ECS) was observed in any of the 18 cases. Therefore, when admitting that presence of ECS affects

660

Jung et al.

the survival rate more than the simple presence of nodal metastasis24, the fact that there was no statistical difference in survival rate between the pN+ group without ECS and the pN0 group, can be explained. In pathologic N staging which considers invasion depth, significant difference was shown in survival rate according to the N staging (Fig. 7). Classifying HP depth according to 9 mm cutoff value, diseasefree survival rates were 82% versus 59% and overall survival rates were 93% versus 60%. According to YUEN et al.25, diseasefree survival of tongue cancer when tumor invasion depth was larger than 9 mm was 60%, which was similar to the results of this study. SHAHA et al.22 reported that when the invasion depth of floor of mouth cancer was larger than 9 mm, the overall survival rate was 65% which was slightly higher than the overall survival rate of tongue cancer in this study. From these observations, it can be considered that invasion depth, which greatly affects the survival rate in oral tongue carcinoma, is a vital factor to be included in the TNM staging system. In conclusion, estimation of invasion depth using MRI as a preoperative study in oral tongue carcinoma is essential in planning surgical treatment strategies such as the extent of elective neck dissection. Invasion depth, which greatly affects occult node metastasis rate, disease-free survival rate and overall survival rate, must be included in the TNM staging of oral tongue carcinoma.

5.

6.

7.

8.

9.

10.

11.

12.

References 1. Asakage T, Yokose T, Mukai K, Tsugane S, Tsubono Y, Asai M, Ebihara S. Tumor thickness predicts cervical metastasis in patients with stage I/II carcinoma of the tongue. Cancer 1998: 82: 1443– 1448. 2. Brown B, Barnes L, Mazariegos J, Taylor F, Johnson J, Wagner RL. Prognostic factors in mobile tongue and floor of mouth carcinoma. Cancer 1989: 64: 1195–1202. 3. Byers RM, Weber RS, Andrews T, McGill D, Kare R, Wolf P. Frequency and therapeutic implications of ‘‘skip metastases’’ in the neck from squamous carcinoma of the oral tongue. Head Neck 1997: 19: 14–19. 4. Fukano H, Matsuura H, Hasegawa Y, Nakamura S. Depth of invasion as a predictive factor for cervical lymph node

13.

14.

15.

16.

metastasis in tongue carcinoma. Head Neck 1997: 19: 205–210. Ho CM, Lam KH, Wei WI, Lau SK, Lam LK. Occult lymph node metastasis in small oral tongue cancers. Head Neck 1992: 14: 359–363. Huang SF, Kang CJ, Lin CY, Fan KH, Yen TC, Wang HM, Chen IH, Liao CT, Cheng AJ, Chang JT. Neck treatment of patients with early stage oral tongue cancer: comparison between observation, supraomohyoid dissection, and extended dissection. Cancer 2008: 112: 1066– 1075. Iwai H, Kyomoto R, Ha-Kawa SK, Lee S, Yamashita T. Magnetic resonance determination of tumor thickness as predictive factor of cervical metastasis in oral tongue carcinoma. Laryngoscope 2002: 112: 457–461. Johnson RE, Sigman JD, Funk GF, Robinson RA, Hoffman HT. Quantification of surgical margin shrinkage in the oral cavity. Head Neck 1997: 19: 281–286. Kligerman J, Lima RA, Soares JR, Prado L, Dias FL, Freitas EQ, Olivatto LO. Supraomohyoid neck dissection in the treatment of T1/T2 squamous cell carcinoma of oral cavity. Am J Surg 1994: 168: 391–394. Kurokawa H, Yamashita Y, Takeda S, Zhang M, Fukuyama H, Takahashi T. Risk factors for late cervical lymph node metastases in patients with stage I or II carcinoma of the tongue. Head Neck 2002: 24: 731–736. Lam P, Au-Yeung KM, Cheng PW, Wei WI, Yuen AP, Trendell-Smith N, Li JH, Li R. Correlating MRI and histologic tumor thickness in the assessment of oral tongue cancer. AJR Am J Roentgenol 2004: 182: 803–808. Layland MK, Sessions DG, Lenox J. The influence of lymph node metastasis in the treatment of squamous cell carcinoma of the oral cavity, oropharynx, larynx, and hypopharynx: N0 versus N+. Laryngoscope 2005: 115: 629–639. Leipzig B, Cummings CW, Chung CT, Johnson JT, Sagerman RH. Carcinoma of the anterior tongue. Ann Otol Rhinol Laryngol 1982: 91: 94–97. Martinez-Gimeno C, Rodriguez EM, Vila CN, Varela CL. Squamous cell carcinoma of the oral cavity: a clinicopathologic scoring system for evaluating risk of cervical lymph node metastasis. Laryngoscope 1995: 105: 728–733. Morton RP, Ferguson CM, Lambie NK, Whitlock RML. Tumor thickness in early tongue cancer. Arch Otolaryngol Head Neck Surg 1994: 120: 717–720. Nathanson A, Agren K, Biorklund A, Lind MG, Andreason L, Anniko M, Freijd A, Lejdeborn L, Kinman S, Kumlien A. Evaluation of some prog-

17.

18.

19.

20.

21.

22.

23.

24.

25.

nostic factors in small squamous cell carcinoma of the mobile tongue: a multicenter study in Sweden. Head Neck 1989: 11: 387–392. O’Brien CJ, Lauer CS, Fredricks S, Clifford AR, McNeil EB, Bagia JS, Koulmandas C. Tumor thickness influences prognosis of T1 and T2 oral cavity cancer—but what thickness? Head Neck 2003: 25: 937–945. Odell EW, Jani P, Sherriff M, Ahluwalia SM, Hibbert J, Levison DA, Morgan PR. The prognostic value of individual histologic grading parameters in small lingual squamous cell carcinomas: The importance of the pattern of invasion. Cancer 1994: 74: 789–794. Rasgon SM, Cruz RM, Hilsinger RL, Sawicki JE. Relation of lymph node metastasis to histopathologic appearance in oral cavity and oropharyngeal carcinoma: a case series and literature review. Laryngoscope 1989: 99: 1103–1110. Sevens A. Fixation and fixatives. Theory and Practice of Histological Techniques. Edinburgh: Churchill Livingstone 1990 p. 28. Shah JP. Patterns of cervical lymph node metastasis from squamous carcinomas of the upper aerodigestive tract. Am J Surg 1990: 160: 405–409. Shaha AR, Spiro RH, Shah JP, Strong EW. Squamous carcinoma of the floor of the mouth. Am J Surg 1984: 148: 455– 459. Spiro RH, Huvos AG, Wong GY, Spiro JD, Gnecco CA, Strong EW. Predictive value of tumor thickness in squamous carcinoma confined to the tongue and floor of the mouth. Am J Surg 1986: 152: 345–350. Woolgar JA, Scott J. Prediction of cervical lymph node metastasis in squamous cell carcinoma of the tongue/floor of mouth. Head Neck 1995: 17: 463–472. Yuen AP, Lam KY, Wei WI, Lam KY, Ho CM, Chow TL, Yuen WF. A comparison of the prognostic significance of tumor diameter, length, width, thickness, area, volume, and clinicopathological features of oral tongue carcinoma. Am J Surg 2000: 180: 139–143.

Corresponding author Address: Se-Heon Kim Department of Otorhinolaryngology Yonsei University College of Medicine 134 Shinchon-Dong Seodaemun-Gu Seoul 120-752 Korea Tel.: +82 2 2228 3622 fax: +82 2 393 0580 E-mail: [email protected]