Cyclo-Oxygenase–2 Expression Is Associated With Vascular Endothelial Growth Factor C Expression and Lymph Node Metastasis in Oral Squamous Cell Carcinoma

PATHOLOGY

Cyclo-Oxygenase–2 Expression Is Associated With Vascular Endothelial Growth Factor C Expression and Lymph Node Metastasis in Oral Squamous Cell Carcinoma Michihide Kono, DDS,* Masato Watanabe, DDS, PhD,y Harutsugi Abukawa, DDS, PhD,z On Hasegawa, DDS,x Takafumi Satomi, DDS, PhD,k and Daichi Chikazu, DDS, PhD{ Purpose: Cervical lymph node metastasis in oral squamous cell carcinoma (OSCC) is recognized as a poor prognostic factor, although its mechanism remains unclear. Recently, cyclo-oxygenase–2 (COX-2) level has been found to correlate highly with vascular endothelial growth factor C (VEGF-C) and lymph node metastasis, as in other solid tumors. However, there has been no report of this correlation in OSCC. Therefore, the aim of this study was to investigate whether COX-2 immunohistochemical expression in OSCC was associated with VEGF-C expression, histopathologic parameters, and lymph node metastasis. Materials and Methods:

Lymphatic vessel density, VEGF-C, and COX-2 immunohistochemical expression were examined pathologically in 60 specimens of invasive OSCC. Relations of histopathologic parameters to lymph node metastasis were analyzed.

Results: Expression levels of VEGF-C and COX-2 and lymphatic vessel density in the lymph node metastatic group were significantly higher than in the nonmetastatic group (P < .01). A significant correlation was found between the expression levels of VEGF-C and COX-2 (r = 0.512; P < .001). COX-2 expression was significantly related to lymph node metastasis (P = .004) and VEGF-C expression (P = .005). Univariate analysis showed that survival time was impaired by higher COX-2 and VEGF-C expression levels. Multivariate survival analysis showed that COX-2 expression was an independent prognostic factor. Conclusion:

This study showed that VEGF-C expression was upregulated by COX-2 in OSCC. High VEGF-C expression appears to promote peritumoral lymphangiogenesis. These data indicated that lymph node metastasis is promoted by COX-2 and VEGF-C in OSCC. Ó 2013 American Association of Oral and Maxillofacial Surgeons J Oral Maxillofac Surg 71:1694-1702, 2013

Cervical lymph node metastasis in oral squamous cell carcinoma (OSCC) is recognized as a poor prognostic factor, and its mechanism remains unclear. A recent study has reported that cancer cells secrete vascular endothelial growth factors C and D (VEGF-C and VEGF-D) and, by binding to vascular endothelial growth factor

receptor-3 (VEGFR-3), VEGF-C and VEGF-D induce intratumoral and peritumoral lymphangiogenesis; in this way, cancer cells invade lymph vessels and promote lymph node metastasis.1 VEGF-C, VEGF-D, and VEGFR-3 signals and peritumoral lymphatic vessel density (LVD; number of peritumoral D2-40–positive

Received from the Department of Oral and Maxillofacial Surgery, Tokyo Medical University, Tokyo, Japan.

Address correspondence and reprint requests to Dr Kono: Department of Oral and Maxillofacial Surgery, Tokyo Medical Univer-

*Graduate Student.

sity, 6-7-1 Nishi-shinjuku, Shinjuku-ku, Tokyo 160-0023, Japan;

yAssistant Professor. zInstructor.

Ó 2013 American Association of Oral and Maxillofacial Surgeons

xInstructor.

0278-2391/13/00416-3$36.00/0

kAssociate Professor.

http://dx.doi.org/10.1016/j.joms.2013.04.015

{Professor and Chair.

e-mail: [email protected]

1694

1695

KONO ET AL

lymphatic vessels) have been identified in many solid tumors, reported as poor prognostic factors, and associated with lymph node metastasis.2-13 It has been reported that cyclo-oxygenase–2 (COX-2) is expressed intensively in different malignancies.14-16 It has been suggested that COX-2 overexpression correlates with tumor aggressiveness and poor prognosis.17 However, whether COX-2 contributes to the formation of new lymphatic vessels is unclear. There have been reports that COX-2 level is highly correlated with VEGF-C and associated with lymph node metastasis in gastric, breast, lung, and head and neck cancers.18-22 COX-2, VEGF-C, and LVD were investigated because they are useful predictive lymph node metastatic biomarkers of OSCC. The present study investigated whether COX-2 immunohistochemical expression is associated with VEGF-C expression, histopathologic parameters, and lymph node metastasis in OSCC.

Materials and Methods PATIENTS

Paraffin-embedded archival specimens were obtained from 60 patients with invasive OSCC; all patients provided written informed consent. This study was approved by the Tokyo Medical University Hospital institutional review board. The patients’ conditions were diagnosed and treated at the Department of Oral and Maxillofacial Surgery, Tokyo Medical University Hospital from 2000 through 2005. The median age of the 60 patients with OSCC was 64.11 years (range, 29 to 91 yr). Thirty-seven patients were men and 23 were women. Thirty-four tumors were located in the tongue, 9 in the mandibular gingiva, 8 in the maxillary gingiva, 5 in the buccal mucosa, and 4 in the oral floor. Nine patients had T1 tumor, 32 had T2, 2 had T3, and 17 had T4; 7 patients were in stage I, 22 in stage II, 7 in stage III, and 24 in stage IV, according to the TNM classification. Thirty-nine patients presented with cervical lymph node metastases; 17 of these 39 patients presented with delayed cervical lymph node metastases (average, 10.6 months after surgery; range, 2 to 41 months).

HISTOPATHOLOGIC PARAMETERS

Histopathologic factors were evaluated by the deep invasive cell grading system of Bryne et al.23 According to this system, 4 parameters (degree of keratinization, nuclear pleomorphism, pattern of invasion, and lymphoplasmacytic infiltration) were measured in the deepest invasive margins, but not in the entire thickness of the tumor. The 4 parameters from each patient were scored from 1 to 4 (Table 1). Obtained scores

were categorized as grade 1 (score, 4 to 8), grade 2 (score, 9 to 12), and grade 3 (score, 13 to 16). IMMUNOHISTOCHEMICAL STAINING

Sections (4 mm) of paraffin-embedded tissue blocks were rehydrated by sequential immersion in xylene, graded ethanol, and water. Then, they were incubated in 0.3% hydrogen peroxide in methanol for 5 minutes followed by placement in a microwave oven for antigen retrieval. After washing in phosphate buffered saline, the slides were exposed to 10% normal goat serum for 10 minutes to decrease nonspecific binding, and this was followed by an overnight incubation at 4 C in a humidified chamber with monoclonal mouse antihuman D2-40 antibody (Abcam, Cambridge, UK) at 1:7 dilution, polyclonal rabbit antihuman VEGF-C antibody (Abcam) at 1:100 dilution, or polyclonal rabbit antihuman COX-2 antibody (Abcam) at 1:400 dilution. Immunostaining was performed with the Envision system (DAKO, Carpinteria, CA) in accordance with the manufacturer’s instructions. Peroxidase activity was visualized by applying diaminobenzidine chromogen containing 0.05% hydrogen peroxidase. Then, the sections were counterstained with hematoxylin, dehydrated, cleared, and mounted. Negative control staining was carried out by substituting nonimmune goat serum for the primary antibodies. QUANTIFICATION OF IMMUNOSTAINING

Results of immunohistochemical staining were semiquantified as described below. The tumorinvaded edges of the primary tumors were observed under a microscope at 400 magnification. All specimens were recorded as digital images using a Zeiss Axiovison imaging system (Carl Zeiss, Oberkochen, Germany) and a Nikon Coolpix 4500 digital camera (Nikon Corp, Tokyo, Japan). A highly expressed area of each antibody in the tumor parenchyma was recorded and evaluated by image analysis software (Image J, National Institutes of Health, Bethesda, MD; Adobe Photoshop, Adobe, San Jose, CA). Then, the expression ratio was calculated by averaging 5 different fields. Ratios were classified into the following 4 groups: +++, expression ratio greater than 50%; ++, expression ratio 20% to 49%; +, expression ratio 5% to 19%; and , expression ratio less than 5%. The high expression groups were +++ and ++. Peritumoral LVD was assessed by light microscopic examination of D2-40–positive lymph vessels at the invasive edge of the primary tumors. D2-40–positive vessels were counted under a 200 field at the tumor cluster-free area (number of lymph vessels per tumor-free area = LVD). The mean LVD from 3 fields was calculated. Results were classified into 2 groups

1696

COX-2 EXPRESSION IN OSCC

Table 1. DEEP INVASIVE CELL GRADING SYSTEM ACCORDING TO BRYNE ET AL

23

Points Morphologic Parameter Degree of keratinization Nuclear pleomorphism

1

2

>50% cells keratinized

3

Pattern of invasion

pushing, welldelineated infiltrating borders

20%-50% cells keratinized moderately abundant nuclear pleomorphism (50%-75% mature cells) infiltrating, solid cords, bands, or strands

Lymphoplasmacytic infiltration

marked

moderate

sparse nuclear pleomorphism (>75% mature cells)

4

5%-20% cells keratinized abundant nuclear pleomorphism (25%-50% mature cells)

0%-5% cells keratinized extreme nuclear pleomorphism (0%-25% mature cells)

small groups or cords of infiltrating cells

marked and widespread cellular dissemination in small groups or in single cells none

slight

Note: Grade 1, score 4 to 8; grade 2, score 9 to 12; grade 3, score 13 to 16. Kono et al. COX-2 Expression in OSCC. J Oral Maxillofac Surg 2013.

according to whether there were up to 12 or more than 12 microvessels per field. STATISTICAL ANALYSIS

Statistical analyses were performed with SPSS 16.0 (SPSS, Inc, Chicago, IL). Correlations among the expression levels of COX-2 and VEGF-C, levels of LVD, and clinicopathological characteristics were calculated by the Student t test, c2 correlation test, and the Pearson correlation coefficient, as appropriate. Logistic regression analysis was performed to examine the significance of the predictive factors for lymph node metastasis. The Kaplan-Meier method was used to estimate survival as a function of time, and survival differences were analyzed with the log-rank test. A multivariable test was performed to determine the factors associated with survival length by Cox regression analysis. The statistical significance level was defined as a P value less than .05 (2-tailed).

Results COX-2, VEGF-C, AND D2-40 EXPRESSIONS IN OSCC

Increased diffuse cytoplasmic staining for COX-2 and VEGF-C was observed in 40 (66%; high expression group) and 33 (55%; high expression group) of the 60 tumor samples, respectively. The median COX-2 expression level was 25.7% (range, 9.6% to 69.4%). The median VEGF-C expression level was 22.6% (range, 0% to 50.1%). However, occasionally, normal epithelial cells and stromal components showed faint staining, particularly the adjacent stromal endothelial cells, for VEGF-C. D2-40 expression was restricted to thin-walled vessel-like structures. D2-40–positive

lymphatic vessels were almost exclusively found within the tumor stroma, at the tumor’s invasion front. Occasional invasion of carcinoma cells into lymph vessels was observed. The median LVD was 12.1 microvessels per field (range, 3 to 19.6 vessels; Fig 1). COX-2 AND VEGF-C EXPRESSIONS AND LVD IN CERVICAL LYMPH NODE METASTASIS

As shown in Figure 2, there were significant differences shown by the Student t test. COX-2 expression levels were 19.0% in the nonmetastatic group and 28.0% in the metastatic group (P = .001). VEGF-C expression levels were 11.5% in the nonmetastatic group and 27.9% in the metastatic group (P < .0001). LVDs were 9.1 in the nonmetastatic group and 13.6 in the metastatic group (P < .0001). VEGF-C and COX-2 expression levels and LVD in the lymph node metastatic group were significantly higher than in the nonmetastatic group. CORRELATION BETWEEN LVD AND COX-2 AND VEGF-C (PEARSON CORRELATION COEFFICIENT)

There was a significant correlation between COX-2 and VEGF-C expression levels (r = 0.512; P < .001; Fig 3). COX-2 expression was not significantly correlated with LVD (r = 0.127; P = .329). VEGF-C expression level was significantly correlated with LVD (r = 0.328; P = .010). ASSOCIATION BETWEEN VEGF-C, COX-2, AND LVD AND HISTOPATHOLOGIC PARAMETERS

COX-2 expression level was significantly associated with T classification (P = .01), stage (P = .01), lymph

KONO ET AL

1697

FIGURE 1. Immunohistochemical staining of A, B, oral squamous cell carcinoma with D2-40 (magnification, 200), C, D, vascular endothelial growth factor C (magnification, 400), and E, F, cyclo-oxygenase–2 (magnification, 400). Increased diffuse cytoplasmic staining for cyclo-oxygenase–2 and vascular endothelial growth factor C was observed. D2-40 expression was restricted to thin-walled vessel-like structures. A, Low lymphatic vessel density group; B, high lymphatic vessel density group; C, low vascular endothelial growth factor C group; D, high vascular endothelial growth factor C group; E, low cyclo-oxygenase–2 group; F, high cyclo-oxygenase–2 group. Kono et al. COX-2 Expression in OSCC. J Oral Maxillofac Surg 2013.

node metastasis (P = .004), and VEGF-C expression (P = .005). VEGF-C expression was significantly associated with stage (P = .01), LVD (P = .04), and COX-2 expression (P = .005) and lymph node metastasis

(P < .0001). LVD was significantly associated with location (tongue and oral floor; P = .008), VEGF-C expression (P = .04), lymph node metastasis (P = .002). Lymph node metastasis was significantly associated

1698

COX-2 EXPRESSION IN OSCC

(P = .008), and COX-2 expression (P = .03) were significantly associated with lymph node metastasis. SURVIVAL ANALYSIS

FIGURE 2. Cyclo-oxygenase–2 and vascular endothelial growth factor C expression levels and lymphatic vessel density in cervical lymph node metastasis. Expression levels of vascular endothelial growth factor C and cyclo-oxygenase–2 and lymphatic vessel density in the lymph node metastatic group were significantly higher than in the nonmetastatic group. COX-2, cyclo-oxygenase–2; LN meta, lymph node metastasis; LVD, lymphatic vessel density; VEGF-C, vascular endothelial growth factor C. Kono et al. COX-2 Expression in OSCC. J Oral Maxillofac Surg 2013.

with stage (P = .0002), LVD (P = .002), VEGF-C expression (P < .0001), and COX-2 expression (P = .004; Table 2). LOGISTIC REGRESSION ANALYSIS OF PREDICTIVE FACTORS FOR LYMPH NODE METASTASIS

Logistic regression analysis was performed to examine the significance of the predictive factors for lymph node metastasis as listed in Table 3. Multivariate analysis showed that VEGF-C expression (P = .004), LVD

FIGURE 3. Correlation between cyclo-oxygenase–2 and vascular endothelial growth factor C expression levels. There was a significant correlation between cyclo-oxygenase–2 and vascular endothelial growth factor C expression levels (Pearson coefficient of correlation). COX-2, cyclo-oxygenase–2; VEGF-C, vascular endothelial growth factor C. Kono et al. COX-2 Expression in OSCC. J Oral Maxillofac Surg 2013.

Kaplan-Meier curves for 5-year disease-specific survival (DSS) are shown in Figure 4. The lymph node metastatic group showed a significantly shorter DSS (P = .004 by log-rank test; Fig 4A) than the nonmetastatic group. Patients with a high expression of VEGF-C showed a significantly shorter DSS (P = .005 by log-rank test; Fig 4B) than patients with a low expression. Patients with a high expression of COX-2 also were found to have a significantly shorter DSS (P = .001 by log-rank test; Fig 4C) than patients with a low expression. The expression and prognosis in the cervical lymph node metastatic group (n = 39) were examined separately. The group with a high expression of COX-2 showed a significantly shorter DSS (P = .04 by log-rank test; Fig 5A). VEGF-C expression showed no significant relation (Fig 5B). In the Cox regression for DSS (including patient age, gender, lymph node metastasis, Bryne grade, stage, COX-2 expression, VEGF-C expression, and LVD), only COX-2 expression was significant (hazard ratio, 9.2; P = .032) and remained an independent prognostic factor.

Discussion The present data showed that VEGF-C expression, COX-2 expression, and LVD were significantly higher in the lymph node metastatic group than in the nonmetastatic group (Fig 2). Semiquantitative analysis showed that high levels of LVD, VEGF-C expression, and COX-2 expression were associated with lymph node metastasis. Furthermore, there was a significant correlation between COX-2 and VEGF-C expression levels (Fig 3) and between VEGF-C and LVD. Examination of histopathologic factors among LVD, COX-2 expression, and VEGF-C expression showed that peritumoral LVD was significantly associated with tumor location (tongue and oral floor). However, the Bryne deep invasive cell grading system did not correlate with any protein expression in this study. There was no significant correlation between lymph node metastatic group and the Bryne grading system, although the data showed a trend toward many lymph node metastases in the high-grade group. These results suggested that VEGF-C expression was upregulated by COX-2 because COX-2 expression was correlated with VEGF-C expression and these were associated with lymph node metastasis. High VEGF-C expression possibly promotes peritumoral lymphangiogenesis. The authors assume that lymph node metastasis may be promoted by COX-2 and VEGF-C in OSCC. Bryne et al23 modified the grading system reported by Anneroth in 1992.24 Although the Bryne system has

1699

KONO ET AL

Table 2. ASSOCIATION BETWEEN LVD, VEGF-C, COX-2, AND LYMPH NODE METASTASIS AND HISTOPATHOLOGIC PARAMETERS

LVD

VEGF-C

COX-2

Lymph Node Metastasis

Low High P Value Low High P Value Low High P Value Negative Positive P Value Age (yr) <65 $65 Gender Male Female Location Tongue and floor Other T classification 1, 2 3, 4 Stage 1, 2 3, 4 Histologic differentiation W/D M/D, P/D Degree of keratinization 1, 2 3, 4 Nuclear pleomorphism 1, 2 3, 4 Pattern of invasion 1, 2 3, 4 Lymphoplasmacytic infiltration 1, 2 3, 4 Bryne grade 1 2 3 LVD Low High VEGF-C Low High COX-2 Low High

.24 14 22

13 11

22 14

15 9

18 18

20 4

23 13

18 6

19 17

10 14

16 20

16 8

24 12

15 9

17 10

22 11

25 11

15 9

26 10

15 9

22 9 5

11 11 2

.13 15 12

12 21

19 8

18 15

16 11

12 21

21 6

20 13

18 9

11 22

13 14

19 14

17 10

22 11

17 10

22 11

16 10

24 11

19 8

22 11

16 7 4

17 13 3

20 7

16 17

.91

.85 10 10

17 23

11 9

26 14

13 7

25 15

18 2

23 17

14 6

15 25

12 8

20 20

13 7

26 14

13 7

26 14

14 6

26 14

15 5

26 14

13 5 2

20 15 5

14 6

22 18

14 6

13 27

.20

.008*

17 4

12 27

11 10

21 18

14 7

25 14

16 5

23 16

14 7

26 13

16 5

25 14

15 4 2

18 16 5

18 3

18 21

17 4

10 29

12 9

8 31

.12

.0002*

.91

.84

.18

.69

.75

.23

24 15

.61

.27

.23

17 4

.19

.61

.76

.57

27 12

.46

.76

.82

11 10

.59

.01*

.46

.74

25 14

.01*

.01*

.09

12 9 .84

.15

.39

16 23

.45

.07

.65

.39 11 10

.77

.43

.49

.33

.53

.04*

.15

.26

.002*

.005*

.0001*

.004

Abbreviations: COX-2, cyclo-oxygenase-2; LVD, lymphatic vessel density; M/D, moderate different; P/D, poorly different; VEGF-C, vascular endothelial growth factor C; W/D, well different. * P < .05 by c2 correlation. Kono et al. COX-2 Expression in OSCC. J Oral Maxillofac Surg 2013.

been a widely accepted grading system, only the limited area of the cells at the deep invasive margin of the tumor was graded. The Bryne system also omitted the stage of invasion and mitotic count from the Anneroth grading system; this omission increased the repro-

ducibility of the grading system.23 Other reports have shown a significant relation between the Bryne grading system and lymph node metastasis.25 However, the present data did not show any relation between the Bryne grading system and lymph node metastasis.

1700

COX-2 EXPRESSION IN OSCC

Table 3. LOGISTIC REGRESSION ANALYSIS OF PREDICTIVE FACTORS FOR LYMPH NODE METASTASIS

Characteristics LVD (low vs high) VEGF-C (low vs high) COX-2 (low vs high)

Parameter Estimate 2.352 2.186 1.939

SE

OR

0.883 10.507 0.762 8.901 0.894 6.954

P Value .008 .004 .03

Abbreviations: COX-2, cyclo-oxygenase–2; LVD, lymphatic vessel density; OR, odds ratio; SE, standard error; VEGF-C, vascular endothelial growth factor C. Kono et al. COX-2 Expression in OSCC. J Oral Maxillofac Surg 2013.

The lymphangiogenesis signaling system is involved in the VEGF-C, VEGF-D, and VEGFR-3 signaling pathways, which have been shown to play a central role in lymphangiogenesis in animal models.26-28 VEGF-C expression levels in primary tumors have been shown to correlate significantly with lymph node metastases in thyroid, prostate, gastric, colorectal, lung, and esophageal carcinomas.2-13 Much of the literature has reported an increased expression of VEGF-C or VEGF-D in OSCC. Overexpression of VEGF-C in OSCC has been correlated with lymphangiogenesis and lymph node metastasis,29-31 and these reports have associated VEGF-C expression with a significantly poor prognosis. COX-2 expression has been reported in various malignancies.14-16 It has been suggested that COX-2 overexpression correlates with tumor aggressiveness and poor prognosis.17 However, whether COX-2 contributes to the formation of new lymphatic vessels is unknown. Recent reports have shown that COX-2 expression is highly correlated with VEGF-C and associated with lymph node metastases in gastric, breast, lung, and head and neck cancers.18-22 There have been some reports on the COX-2 and VEGF-C signaling pathways. Iwata et al18 found that a COX-2 inhibitor decreased lymphatic metastasis. In their report, tumor lymphangiogenesis and lymph node metastasis were significantly decreased in etodolac-treated mice compared with control mice. Moreover, they reported that the major source of VEGF-C and VEGF-D was F4/80-positive macrophages as determined by immunohistochemical analysis, and that this was associated with tumor lymphangiogenesis. Conversely, Su et al21 found that COX-2 upregulated VEGF-C, which promoted lymphangiogenesis in human lung adenocarcinoma along the prostaglandin E receptor-1, c-Src tyrosine kinase, human epidermal growth factor receptor-2/Neu signaling pathway. At the mechanistic level, they found that COX-2 expression or prostaglandin E2 (PGE2) treatment could activate the human epidermal growth factor receptor-2/Neu tyrosine kinase

FIGURE 4. Kaplan-Meier survival analysis. Survival rate was estimated with the Kaplan-Meier method and analyzed using the logrank test. Five-year disease-specific survival curves are shown for the A, negative versus positive lymph node metastatic group, B, high versus low vascular endothelial growth factor C group, and C, high versus low cyclo-oxygenase–2 group. COX-2, cyclo-oxygenase–2; LN meta, lymph node metastasis; VEGF-C, vascular endothelial growth factor C. Kono et al. COX-2 Expression in OSCC. J Oral Maxillofac Surg 2013.

receptor through the prostaglandin E receptor-1– dependent pathway and that this activation was essential for VEGF-C induction.

1701

KONO ET AL

In conclusion, the authors have presented new evidence that COX-2 expression correlates with VEGF-C expression. These 2 proteins act closely together, which results in the occurrence of lymph node metastasis in OSCC. Moreover, VEGF-C expression may be upregulated by COX-2 expression to promote lymphangiogenesis. COX-2 expression was associated with survival time and independent prognostic factors for OSCC. Further, these data support the concept that a COX-2 inhibitor might prevent lymph node metastasis. This approach might have the potential to improve the poor prognosis in OSCC.

References

FIGURE 5. Kaplan-Meier survival analysis. Survival rate was estimated with the Kaplan-Meier method and analyzed using the logrank test. Five-year disease-specific survival curves are shown for the lymph node metastatic group (n = 39): A, high versus low cyclo-oxygenase–2 group and B, high versus low vascular endothelial growth factor C group. COX-2, cyclo-oxygenase–2; VEGF-C, vascular endothelial growth factor C. Kono et al. COX-2 Expression in OSCC. J Oral Maxillofac Surg 2013.

Although this lymphangiogenesis pathway has been described, it is unknown whether this is the same pathway in OSCC. During angiogenesis in head and neck cancer, including oral cancer, COX-2 upregulates VEGF through the PGE2 pathway.32 Moreover, Gallo et al32 showed that COX-2 mRNA and protein expression was associated with an increase in PGE2 production from tumor tissue and significantly associated with lymph node metastasis. For the lymphangiogenesis pathway in OSCC, the authors infer that COX-2 upregulates VEGF-C by PGE2, similarly to the angiogenesis pathway, and in turn promotes lymph node metastases. Furthermore, tumor-associated macrophages in the tumor microenvironment may be related to tumor lymphangiogenesis and lymph node metastases in OSCC and other cancers. The authors plan to investigate this in future studies.

1. Stacker SA, Achen MG, Jussila L, et al: Lymphangiogenesis and cancer metastasis. Nat Rev Cancer 2:573, 2002 2. Kinoshita J, Kitamura K, Kabashima A, et al: Clinical significance of vascular endothelial growth factor-C (VEGF-C) in breast cancer. Breast Cancer Res Treat 66:159, 2001 3. Hirai M, Nakagawara A, Oosaki T, et al: Expression of vascular endothelial growth factors (VEGF-A/VEGF-1 and VEGF-C/VEGF-2) in postmenopausal uterine endometrial carcinoma. Gynecol Oncol 80:181, 2001 4. Hashimoto I, Kodama J, Seki N, et al: Vascular endothelial growth factor-C expression and its relationship to pelvic lymph node status in invasive cervical cancer. Br J Cancer 85:93, 2001 5. George ML, Tutton MG, Janssen F, et al: VEGF-A, VEGF-C, and VEGF-D in colorectal cancer progression. Neoplasia 3:420, 2001 6. Kitadai Y, Amioka T, Haruma K, et al: Clinicopathological significance of vascular endothelial growth factor (VEGF)-C in human esophageal squamous cell carcinomas. Int J Cancer 93:662, 2001 7. Yonemura Y, Endo Y, Fujita H, et al: Role of vascular endothelial growth factor C expression in the development of lymph node metastasis in gastric cancer. Clin Cancer Res 5:1823, 1999 8. O-Charoenrat P, Rhys-Evans P, Eccles SA: Expression of vascular endothelial growth factor family members in head and neck squamous cell carcinoma correlates with lymph node metastasis. Cancer 92:556, 2001 9. Niki T, Iba S, Tokunou M, et al: Expression of vascular endothelial growth factors A, B, C, and D and their relationships to lymph node status in lung adenocarcinoma. Clin Cancer Res 6:2431, 2000 10. Ohta Y, Shridhar V, Bright RK, et al: VEGF and VEGF type C play an important role in angiogenesis and lymphangiogenesis in human malignant mesothelioma tumours. Br J Cancer 81:54, 1999 11. Tang RF, Itakura J, Aikawa T, et al: Overexpression of lymphangiogenic growth factor VEGF-C in human pancreatic cancer. Pancreas 22:285, 2001 12. Tsurusaki T, Kanda S, Sakai H, et al: Vascular endothelial growth factor-C expression in human prostatic carcinoma and its relationship to lymph node metastasis. Br J Cancer 80:309, 1999 13. Bunone G, Vigneri P, Mariani L, et al: Expression of angiogenesis stimulators and inhibitors in human thyroid tumors and correlation with clinical pathological features. Am J Pathol 155:1967, 1999 14. Sano H, Kawahito Y, Wilder RL, et al: Expression of cyclooxygenase-1 and -2 in human colorectal cancer. Cancer Res 55:3785, 1995 15. Zimmermann KC, Sarbia M, Weber AA, et al: Cyclooxygenase-2 expression in human esophageal carcinoma. Cancer Res 59:198, 1999 16. Hida T, Yatabe Y, Achiwa H, et al: Increased expression of cyclooxygenase 2 occurs frequently in human lung cancers, specifically in adenocarcinomas. Cancer Res 58:3761, 1998 17. Tsujii M, Kawano S, DuBois RN: Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci U S A 94:3336, 1997

1702 18. Iwata C, Kano MR, Komuro A, et al: Inhibition of cyclooxygenase-2 suppresses lymph node metastasis via reduction of lymphangiogenesis. Cancer Res 67:10181, 2007 19. Bhattacharjee RN, Timoshenko AV, Cai J, Lala PK: Relationship between cyclooxygenase-2 and human epidermal growth factor receptor 2 in vascular endothelial growth factor C up-regulation and lymphangiogenesis in human breast cancer. Cancer Sci 101: 2026, 2010 20. Zhang XH, Huang DP, Guo GL, et al: Coexpression of VEGF-C and COX-2 and its association with lymphangiogenesis in human breast cancer. BMC Cancer 8:4, 2008 21. Su JL, Shih JY, Yen ML, et al: Cyclooxygenase-2 induces EP1- and HER-2/Neu-dependent vascular endothelial growth factor-C upregulation: A novel mechanism of lymphangiogenesis in lung adenocarcinoma. Cancer Res 64:554, 2004 22. Kyzas PA, Stefanou D, Agnantis NJ: COX-2 expression correlates with VEGF-C and lymph node metastases in patients with head and neck squamous cell carcinoma. Mod Pathol 18:153, 2005 23. Bryne M, Koppang HS, Lilleng R, et al: Malignancy grading of the deep invasive margins of oral squamous cell carcinomas has high prognostic value. J Pathol 166:375, 1992 24. Anneroth G, Batsakis J, Luna M: Review of the literature and a recommended system of malignancy grading in oral squamous cell carcinomas. Scand J Dent Res 95:229, 1987

COX-2 EXPRESSION IN OSCC 25. Neena PD, Siddharth AS, Keyuri BP, et al: Histological grading of oral cancer: A comparison of deifferent systems and their relation to lymph node metastasis. Natl J Community Med 2:136, 2011 26. Alitalo K, Tammela T, Petrova TV: Lymphangiogenesis in development and human disease. Nature 438:946, 2005 27. Achen MG, Stacker SA: Tumor lymphangiogenesis and metastatic spread-new players begin to emerge. Int J Cancer 119:1755, 2006 28. Pepper MS, Tille JC, Nisato R, et al: Lymphangiogenesis and tumor metastasis. Cell Tissue Res 314:167, 2003 29. Sugiura T, Inoue Y, Matsuki R, et al: VEGF-C and VEGF-D expression is correlated with lymphatic vessel density and lymph node metastasis in oral squamous cell carcinoma: Implications for use as a prognostic marker. Int J Oncol 34:673, 2009 30. Warburton G, Nikitakis NG, Roberson P, et al: Histopathological and lymphangiogenic parameters in relation to lymph node metastasis in early stage oral squamous cell carcinoma. J Oral Maxillofac Surg 65:475, 2007 31. Shintani S, Li C, Ishikawa T, et al: Expression of vascular endothelial growth factor A, B, C, and D in oral squamous cell carcinoma. Oral Oncol 40:13, 2004 32. Gallo O, Franchi A, Magnelli L, et al: Cyclooxygenase-2 pathway correlates with VEGF expression in head and neck cancer. Implications for tumor angiogenesis and metastasis. Neoplasia 3:53, 2001