Expression and prognostic significance of EFNB2 and EphB4 genes in patients with oesophageal squamous cell carcinoma

Digestive and Liver Disease 39 (2007) 725–732

Alimentary Tract

Expression and prognostic significance of EFNB2 and EphB4 genes in patients with oesophageal squamous cell carcinoma M. Tachibana ∗ , Y. Tonomoto, R. Hyakudomi, M. Hyakudomi, S. Hattori, S. Ueda, S. Kinugasa, H. Yoshimura Digestive and General Surgery, Department of Surgery, Faculty of Medicine, Shimane University, Izumo 693-8501, Shimane, Japan Received 29 November 2006; accepted 14 May 2007 Available online 3 July 2007

Abstract Objective. Tyrosine kinases and its receptors play important roles in growth, migration, and invasion of malignant cells. Among those, there are only few reports examining the expression pattern of Eph/ephrin signalling system in oesophageal carcinoma. The prognostic importance of ephrin-B2 ligand (EFNB2) and its receptor EphB4, and its correlation with clinicopathologic characteristics are yet to be delineated in patients with oesophageal carcinoma. Materials and methods. EFNB2 gene and EphB4 receptor gene were examined of mRNA specimens in 61 patients with oesophageal squamous cell carcinoma using reverse-transcriptase polymerase chain reaction. EFNB2 protein was selectively examined using an immunohistochemical analysis. Results. EFNB2 mRNA expression was detected in 38 (62.3%) and EphB4 expression was found in 44 (72.1%) out of 61 cancer tissues analysed. There was a statistically significant correlation between EFNB2 expression and number of lymph node metastasis (P < 0.05), and a trend toward statistical significance for correlation between EFNB2 expression and American Joint Committee on Cancer Classification Stage (P < 0.1), indicating that EFNB2 expression was up-regulated by advancement of the disease process. EFNB2 protein was strongly expressed in tumour with high mRNA EFNB2 expression and was weakly expressed in tumour with low mRNA expression in some representative tumours. The 5-year overall survival rate (23%) of patients with positive EFNB2 gene expression was significantly worse than 55% of negative expression (P < 0.05). The results of multivariate analysis of prognosticators for survival showed that positive EFNB2 gene expression (P < 0.01) and number of lymph node metastasis (P < 0.05) were identified as significant factors indicative of a poorer survival. Conclusions. EFNB2 gene expression may be a biological marker and a useful prognostic indicator in patients with oesophageal squamous cell carcinoma. © 2007 Published by Elsevier Ltd on behalf of Editrice Gastroenterologica Italiana S.r.l. Keywords: Ephrin; Eph receptor; Oesophageal cancer; Tyrosine kinase

1. Introduction Oesophageal carcinoma is one of the most malignant tumours with a dismal prognosis. Although surgery is a potentially curative treatment of choice for early stage oesophageal cancer, it has some limitation since most patients present with advanced diseases. Despite recent advances in diagnostic method and meticulous lymphadenectomy during radical ∗

Corresponding author. Tel.: +81 853 20 2232; fax: +81 853 20 2229. E-mail address: [email protected] (M. Tachibana).

oesophagectomy with or without adjuvant treatments, half of the patients die within 3 years and the overall 5-year survival rate is only 20–23.8% [1–3]. Tyrosine kinases and its receptor tyrosine kinases (RTK) play important roles in growth, migration, and invasion of various normal and malignant cells [4]. Physiologically relevant interaction between the receptor and its ligand probably requires cell–cell contact since ligand has been reported to be most effective in receptor activation when cell-surface bound [5]. The cytoplasmic domain of two transmembrane ligands has been shown to undergo phosphorylation on conserved

1590-8658/$30 © 2007 Published by Elsevier Ltd on behalf of Editrice Gastroenterologica Italiana S.r.l. doi:10.1016/j.dld.2007.05.013

726

M. Tachibana et al. / Digestive and Liver Disease 39 (2007) 725–732

intracellular tyrosine residues following receptor interaction [6]. This observation suggests that this receptor/ligand family can provoke bi-directional signalling and enables mutual cell–cell communication. Larger numbers of tyrosine kinases and RTKs have been correlated with carcinogenesis and cancer progression. Eph receptors were originally identified in 1987 in a human complementary DNA (cDNA) library screen for homologous sequences to the tyrosine kinase domain of the viral oncogene v-fps [7]. Eph receptors are divided into the A- or B-type based on their interactions with ephrin ligands. There are nine members of the A-type and four members of the B-type. On the contrary, ephrin ligands are divided into the A- or B-class based on their sequence conservation [4]. The important consequences of Eph/ephrin signalling are tumourigenesis and metastatic potentials in some cancers [4,8], angiogenesis and tumour vasculature [9], and neurological functions and development of central nervous systems. Among those Eph receptors and their ligands, EFNB2 cognates with EphB4 [10]. To date, few reports have examined concomitant expressions of EphB4 and its ligand EFNB2 in mammary gland [11], leukaemia–lymphoma cell lines [12], and endometrial cancer [13]. Moreover, there are only two reports examining the expression pattern of Eph/EFNA2 signalling system in oesophageal carcinoma [14,15]. However, there has been so far no study examining the expression of both EphB4/EFNB2 signalling system in oesophageal carcinoma. Therefore, we examined EphB4 and EFNB2 expression using reverse-transcriptase polymerase chain reaction (RTPCR) in 61 consecutive surgical specimens of oesophageal squamous cell carcinoma (SCC) patients, and the results were correlated with clinicopathologic features and patients’ survival.

2. Materials and methods 2.1. Patients and tissue collection Between 1986 and 2002, consecutive 61 patients who underwent curative resection for thoracic oesophageal SCC and had sufficient quality frozen tissue samples for conducting molecular analysis of ephrin-B2 and Eph B4 expression were included in this study. All 61 patients underwent uniform right transthoracic subtotal oesophagectomy with three-field lymph node dissection and received reconstruction with retrosternal gastric tube or colon. Immediately after resection, tumour tissue and corresponding normal mucosa tissue were separately collected and stored at −80 ◦ C for further analysis. The remaining specimen was submitted to the Central Pathology Department for characterization of the tumour by the pathologists. All frozen samples were preserved in the tissue bank at the university. The clinicopathologic characteristics of those patients were investigated based on the UICC-tumour-node-metastasis

(UICC-TNM) classification [16] of oesophageal SCC. Consent was obtained from all patients for use of tissue specimens for clinical research and the University Ethical Committee approved the research protocol. Moderately and poorly differentiated tumours were classified as being non-well differentiated tumours. The follow-up was complete in all patients. The data obtained at regular follow-up visits were stored in a database specially designed for patients with oesophageal SCC. 2.2. RNA isolation and cDNA synthesis The cancer tissue and corresponding normal oesophageal mucosa were separately obtained from the resected specimen immediately after resection and stored at −80 ◦ C for RNA extraction. Total RNA samples were extracted using the TRI reagent (Molecular Research Center, Cincinnati, OH) following the manufacturer’s instructions. Then, first-strand cDNA was synthesized using the Oligo (dT), SuperScript Reverse Transcriptase II (SS RTII; Invitrogen, Carlsbad, CA), 5× first-strand buffer, 2.5 ␮M dNTP mixture containing each deoxynucleotide triphosphate base (Takara Biochemicals, Ohtsu, Japan), and MilliQ water was added to give a total volume of 20 ␮l. The RT step was carried out at 42 ◦ C for 50 min followed by heating at 70 ◦ C for 15 min to inactivate the enzyme, and immediately cooled on ice. 2.3. Reverse-transcriptase polymerase chain reaction (RT-PCR) The cDNA samples were amplified by use of a thermal cycler (Program Temp. Control System PC-701, ASTEC, Fukuoka, Japan). These primers for each gene were derived from different exons and the sequences of the amplified products were confirmed to ensure that they represent the specific mRNA species and not genomic DNA. Negative controls (cDNA-free solutions) were included in each run. Each cDNA species was amplified for 30 cycles (ephrin-B2 and Eph B4) (95 ◦ C for 30 s, 55 ◦ C for 15 s, and 72 ◦ C for 30 s) and 23 cycles (␤-actin) (94 ◦ C for 35 s, 55 ◦ C for 35 s, and 72 ◦ C for 2 min), followed by a final extension at 72 ◦ C for 4 min. After defined cycles of PCR, 9 ␮l was electrophoresed on 1.5% agarose gel in 0.5M Tris–borate buffer from the 25 ␮l reaction volume, and amplified bands were detected by ethidium bromide staining. The intensity of ethidium bromide fluorescence was measured by a charged couple device system (Kodak Digital Science EDAS290 System, Invitrogen Japan K.K., Japan). For quantification of ephrin-B2 and Eph B4 mRNA, the samples were evaluated in comparison with the PCR product to ␤-actin. Tumours were divided into positive and negative for corresponding positive and negative expressions of EFNB2 and EphB4 mRNA, respectively. 2.4. Immunohistochemical analysis of EFNB2 Paraffin sections were selected and used for immunohistochemical analysis. Briefly, slides were deparaffinized,

M. Tachibana et al. / Digestive and Liver Disease 39 (2007) 725–732

rehydrated in graded alcohols and placed in PBS solution. Antigen retrieval was done by autoclaving the slides for 10 min in citrate buffer (pH 6.0) (Dako ChemMate, Dako, Denmark). Endogenous peroxidase activity was quenched by dipping in 3% H2 O2 for 30 min and nonspecific binding was blocked by treating slides with normal rabbit serum for 30 min. Rat anti-human EFNB2 monoclonal antibody (B8, Santa Cruz Biotechnology, California, 1:300 dilution) was used at 4 ◦ C for overnight incubation. The streptavidin–biotin peroxidase complex method was employed for the immunohistochemical steps and was done with a commercially available kit (Histofine SAB-PO kit, Nichirei, Tokyo, Japan). Colour development was done with the peroxidase substrate AEC (3-amino-9-ethylcarbazole). In negative control slide, the primary antibody was replaced by nonimmune serum. 2.5. Statistical analyses The standard Chi-square test with or without Yates’ correction was used for comparative analyses. The survival rates were estimated using the Kaplan–Meier method [17], and the statistical analysis was carried out using the log-rank test. In multivariate analysis, independent prognostic factors were determined by multiple stepwise regression analysis using the Cox model (StatView J4.5; Abacus Concepts, Inc., Berkeley, CA). The level of significance was set at P < 0.05.

3. Results 3.1. Patient and tumour characteristics Age ranged from 49 to 86 (mean, 65.2), and 55 were male and 6 were female (Table 1). Macroscopic tumour classifications were type 1 in 12 patients, type 2 in 18, and type 3 in 31. Tumours were located at the upper thoracic oesophagus in 5, middle thoracic oesophagus in 33, and lower thoracic oesophagus in 23. Pathological tumour stages (T) were T1 in 11, T2 in 18, T3 in 29, and T4 in 3 patients. Lymph node metastasis was found in 41 patients (65.6%) and the number of nodal metastasis counted 1 node in 8 patients, 2 nodes in 8, 3 nodes in 5, 4 nodes in 4, and >5 nodes in 16 patients. Pathological TNM stages were stage I in 8 patients, stage II in 1, stage III in 17, and stage IV in 19. 3.2. The amounts of EFNB2 and EphB4 genes in oesophageal tissues analysed by RT-PCR The sequences of each primer for RT-PCR are shown in Table 2. RT-PCR was performed to evaluate the amounts of EFNB2 and EphB4 mRNAs in oesophageal SCC samples (n = 61) and corresponding normal mucosa samples (n = 61). To standardize the amount of mRNA in each sample, ␤-actin mRNA was used as an internal control. EFNB2 mRNA level was 0.783 ± 0.819 in the cancer tissues and 0.767 ± 0.715 in the surrounding normal mucosa (not significant; P = 0.91).

727

Table 1 Patient and tumor characteristics Age (years); mean (range) Gender (M/F) Tumour length (cm); m ± S.D. (range)

65.2 (49–86) 55/6 4.9 ± 2.1 (0.8–10.2)

Tumour location Upper/middle/lower

5/33/23

Macroscopic classification I/II/III/IV

12/18/31

Histological type Well/moderately/poorly

32/22/7

T-status T1/T2/T3/T4

11/18/29/3

N-status N0/N1

20/41

Number of lymph node metastasis None/1/2/3/4/>5 m ± S.D.

20/8/8/5/4/16 2.9 ± 3.6

TNM stage I/IIA/IIB/III/IV

8/12/5/17/19

Lymphatic invasion Negative/positive

28/33

Vessel invasion Negative/positive

25/37

m ± S.D.: mean ± standard deviation.

Moreover, EphB4 mRNA level was not significantly different (P = 0.42) between the cancer tissue (0.895 ± 0.681) and the normal mucosa (0.845 ± 0.622). 3.3. EFNB2 gene and clinicopathologic parameters, and immunohistochemical expression EFNB2 mRNA was detected in 38 (62.3%) out of 61 oesophageal SCC tissues analysed and was not expressed in the remaining 23 tumours. EphB4 mRNA expression was also found in 44 (72.1%) and was not detected in the remaining 17 tumours (Fig. 1). Tumours were divided into positive and negative groups for corresponding positive and negative expressions of ephrin-B2 and EphB4 mRNA, respectively. Gender and age did not correlate with mRNA EFNB2 expression (Table 3). There was a statistically significant correlation between EFNB2 mRNA expression and number of Table 2 Primer and probe sequences for RT-PCR Primer

Sequences (5 –3 )

EFNB2 Sense Antisense

GCAAGTTCTGCTGGATCAAC AGGATGTTGTTCCCCGAATG

Eph4 Sense Antisense

GTCTGACTTTGGCCTTTCCC TGACATCACCTCCCACATCA

␤-actin Sense Antisense

TGACCCAGATCATGTTTGAGA ACTCCATGCCCAGGAAGGA

728

M. Tachibana et al. / Digestive and Liver Disease 39 (2007) 725–732 Table 3 EFNB2 expression and clinicopathologic features in patients with oesophageal squamous cell carcinoma (n = 61) Variables

Fig. 1. mRNA EFNB2 and EphB4 mRNA expressions by RT-PCR analyses in oesophageal squamous cell carcinoma and normal mucosa adjacent to tumour. Representative mRNA EFNB2 expressions of six patients were shown (top) and mRNA EphB4 expressions were shown (middle). ␤-actin was used as an internal control (bottom). Differences of each mRNA expression were found among each sample.

lymph node metastasis (P < 0.05), and a trend toward statistical significance for correlation between EFNB2 expression and American Joint Committee on Cancer Classification (AJCC) Stage (P < 0.1), indicating that EFNB2 expression was up-regulated by advancement of the disease process. EFNB2 expression did not correlate with pN status. Interestingly, among 25 patients with 1–4 lymph node metastasis, 10 out of 20 patients with EFNB2(+) had distant lymph node metastasis (M1lym), whereas none out of 5 patients with EFNB2(−) had M1lym (P ≥ 0.05 by Chi-square test, but not significant by Yates’ correction). On the contrary, among 16 patients with >5 LN metastasis, 4 out of 7 patients with EFNB2(+) had M1ly, whereas 5 out of 9 patients with EFNB2(−) had M1lym (not significant). Moreover, there were no significant correlations between EphB4 expression and clinicopathologic characteristics, furthermore there existed no correlations between combined EFNB2 and EphB4 expressions and clinicopathologic parameters (data not shown). Immunohistochemical expressions of EFNB2 are shown in Fig. 2. EFNB2 protein was strongly expressed in tumour with high mRNA EFNB2 expression (a, ×100) and was weakly expressed in tumour with low mRNA expression (b, ×100).

EFNB2 expression Positive (n = 38)

Negative (n = 23)

P-value

Age (years) ≤65 >66

17 21

13 10

NS (0.43)

Gender Male Female

33 5

22 1

NS (0.49)

Tumour size (cm) ≤3.0 >3.1

5 33

4 19

NS (0.71)

Macroscopic classification 0.1 6 2.3 32

6 17

NS (0.34)

Tumor location Upper Middle/lower

3 20

NS (0.35)

Histological differentiation Well 18 Non-well 20

14 9

NS (0.43)

Depth of invasion T1,2 T3,4

16 22

13 10

NS (0.30)

Lymph node metastasis N0 11 N1 27

9 14

NS (0.57)

2 36

Number of lymph node metastasis 0 11 1–4 20 ≥5 7

9 5 9

<0.05

12 26

13 10

<0.10

Lymph vessel invasion Negative 18 Positive 20

10 13

NS (0.80)

Blood vessel invasion Negative 17 Positive 21

7 16

NS (0.29)

Stage I, II III, IV

NS: not significant.

3.4. Outcomes and overall survival rate Twenty-seven patients are alive and free of cancer at the time of this analysis. Thirteen patients died of unrelated causes to oesophageal cancer and the remaining 21 patients died of oesophageal diseases; 10 due to locoregional recurrences, 8 due to locoregional recurrences with distant metastases, and 3 due to distant metastases. One-, three-, and five-years overall survival rates of all 61 patients were 75%, 47% and 37%, and those of diseasespecific survival rates were 75%, 60% and 56%, respectively. Median survival time was 17.4 months of patients with positive EFNB2 mRNA expression and 35.5 months of negative

expression (P < 0.01). The long-term overall survival rate of patients with positive EFNB2 mRNA expression (n = 38) was significantly worse than that of negative expression (n = 23) (P = 0.023, Fig. 3). In stratification of node positive (n = 41) or advanced tumours (pT3,4, n = 32), the long-term survival rates were tended to be significantly different between positive and negative EFNB2 expressions (node positive; P = 0.081, T3,4; P = 0.078). On the contrary, there were no survival differences in terms of EphB4 expression (P = 0.28) and in terms of combined ephrin-B2 and EphB4 expressions (P = 0.988).

M. Tachibana et al. / Digestive and Liver Disease 39 (2007) 725–732

729

Fig. 2. Immunohistochemical EFNB2 expression. Representative EFNB2 immunohistochemical expressions were shown. EFNB2 protein was strongly expressed in tumour with high mRNA EFNB2 expression (a, ×100) and was weakly expressed in tumour with low mRNA expression (b, ×100).

Fig. 3. Cumulative survival curves according to EFNB2 expression. One-, three-, and five-years overall survival rates of 38 patients with positive EFNB2 expression were 68%, 40% and 23%, respectively, and those with negative expression were 83%, 55% and 55% (P < 0.05). Median survival time was 17.4 months of patients with positive EFNB2 mRNA expression and 35.5 months of negative expression (P < 0.01).

node metastasis (0 versus 1–4 versus 5≤), EFNB2 status (positive versus negative), age (65≥ versus 66≤), tumour size (3.0≥ versus 3.1≤), tumour location (upper versus middle/lower), macroscopic classification (type 0,1 versus type 2,3), T (T1,2 versus T3,4), histological differentiation (well versus non-well), blood vessel invasion (positive versus negative), and lymph vessel invasion (positive versus negative). The results of multivariate analysis of prognosticators for survival showed that positive EFNB2 expression (P = 0.006, relative risk 3.88) and number of lymph node metastasis (P = 0.023, relative risk 3.24) were identified as significant factors indicative of a poorer survival (Table 4). Additional multivariate analysis was done to evaluate the EphB4 expression in the multivariate analysis. The results of multivariate analysis showed that neither EphB4 expression nor combined EFNB2 and EphB4 status became independent prognosticators for survival (Table 4). 3.6. Pattern of recurrence and EFNB2 expression

3.5. Multivariate analyses for overall survival The following conventional prognostic parameters were entered into the multivariable analysis: number of lymph

Eight (57%) out of 14 patients who died of recurrences died mainly of distant diseases in positive EFNB2 expression, whereas only 3 (43%) out of 7 succumbed to distant recur-

Table 4 Multivariate analysis of prognostic factors for survival using Cox’s proportional hazard model Parameter

P-value

Hazard ratio

95% CI

EFNB2 Number of lymph node metastasis Depth of invasion (T) Age Macroscopic classification Tumor location Tumor size Histological differentiation Lymph vessel invasion Blood vessel invasion

0.006 0.023 0.696 0.099 0.452 0.292 0.976 0.185 0.199 0.933

3.88 3.24 0.82 0.50 0.52 0.44 1.03 1.74 1.28 0.95

1.47–10.28 1.17–8.93 0.31–2.20 0.22–1.14 0.09–2.85 0.09–2.02 0.20–5.20 0.77–3.93 0.43–3.84 0.29–3.13

CI: confidence interval.

730

M. Tachibana et al. / Digestive and Liver Disease 39 (2007) 725–732

rences in negative expression, but there was no significant difference.

4. Discussion The important consequences of Eph/ephrin signalling system are tumourigenesis and metastatic potentials in various cancers [4], angiogenesis, and tumour vasculature [8,9]. Recent data suggest that Eph/ephrin signalling system could play an important role in the development of novel inhibition strategies and cancer treatments to potentially target the receptor tyrosine kinase and/or its ligand [4]. In the present study, there were correlations between EFNB2 expression, and number of lymph node metastasis and AJCC Stage. The overall survival rate of patients with positive EFNB2 expression was significantly worse than negative expression. The results of multivariate analysis showed that positive EFNB2 expression and number of lymph node metastasis were identified as significant factors indicative of a poorer survival. Eph and/or ephrin expressions were common in certain cell lines and in numerous primary tumours such as gastric [18], colorectal [19,20], pancreatic [21], breast [22–24], endometrial [25,26], cervical [27], neuroblastoma [28,29], lung [30], melanoma [31,32], and Kaposi sarcoma [33]. In oesophageal SCC, mRNA EphA2 overexpression was positive in 40 (50%) of 80 patients [14]. In another study examining oesophageal SCC, EphA2 protein was positive in 140 (80.9%) of 173 patients and EFNA1 protein was positively found in 84.4% [15]. Moreover, variable amounts of EphA2 and EFNA1 mRNA overexpressions were observed in all 20 tumour samples analysed. These results indicate that some type of Eph/ephrin signalling system may play a major role in carcinogenesis of oesophageal SCC. However, there has been so far no study that investigated the expression pattern of EFNB2/EphB4 signalling system in oesophageal carcinoma. Among various biological functions of ephrin and Eph receptors in cancer, its roles of EFNB2/EphB4 signalling are thought to be angiogenesis [34], differentiation, and development [4]. Having these biological functions of EFNB2/EphB4 system, in the present study, EFNB2 overexpression was significantly correlated with number of lymph node metastasis and clinical stage. Moreover, Miyazaki et al. [14] reported that EphA2 expression was correlated with lymph node metastasis and tumour differentiation. Likewise, Xu et al. [15] mentioned that there was a significant correlation between EphA2 expression and lymph node metastasis. These results may suggest that ephrin/Eph signalling systems play a major role in oesophageal carcinogenesis through the biological functions such as angiogenesis, differentiation, and development. The long-term survival of patients with positive EFNB2 expression was significantly worse than negative expression and EFNB2 expression status became an independent prognosticator by multivariate analysis in the present

study. We also found that EFNB2 expression was correlated well with advancement of disease process. For those strong dependencies on favourable prognostic indicators (e.g. lymph node metastasis and advanced stage) in oesophageal carcinoma patients, this factor became an independent prognosticator for survival. This is the first, to our best knowledge, that show a prognostic significance of EFNB2 in patients with oesophageal carcinoma. Furthermore, Miyazaki et al. [14] reported that survival rates of EphA2-positive patients were poorer than negative ones by univariate analysis. Likewise, Xu et al. [15] mentioned that EphA2 was one of independent significances by multivariate analysis. Taken together, these data may suggest that ephrin/Eph signalling system provides a significant prognostic marker for patients’ survival. Although EFNB2 expression may represent an interesting observation, more definitive studies need to be required before we confirm these results. EFNB2 gene expression did not correlate with pN0/pN1 status instead of a significant correlation with number of lymph node metastasis. TNM classification for gastric cancer and colorectal cancer [16] were fully revised according to the number of involved lymph nodes, but this staging system by the number of positive nodes was not applied for oesophageal cancer. The number of metastatic lymph nodes is not an artificial classification and also gives precise information for advancement of the disease and prognosis without the problems related to nodal staging. Among the node-positive patients, the number of the metastatic lymph nodes clearly influenced the survival after curative oesophagectomy. Generally, patients with a large number of diseased nodes have a worse survival than those with a few metastatic nodes. This classification was significantly correlated with depth of invasion (pT) and lymph node metastasis (pN). Moreover, the number of diseased lymph nodes was correlated well with the Japanese nodal level and TNM stage. These findings support that the number of positive lymph nodes is useful for evaluating the tumour progression rather than pN category [35]. Expression status of EphB4 receptor neither shows a clear correlation with any clinicopathologic characteristics nor become a prognostic factor in this study. A special feature of ephrin and Eph receptor is the concept of bi-directional (reverse and forward) signalling [4,8,9]. An Eph receptor can act as a ligand in the same way that an ephrin ligand can act as a receptor. This is a possible reason why EphB4 receptor expression does not necessarily correlate with ephrin-B2 ligand status. Moreover, both EFNB2 and EphB4 mRNA levels were not significantly different between cancer tissues and corresponding normal mucosa tissue in this study. These discordant results might be explained by frequent dysplastic changes in the adjacent oesophageal mucosa to the cancer tissues [36]. In conclusion, our results indicate that EFNB2 gene is upregulated by advancement of the disease process and becomes a biological marker and may be a useful predictor of prognosis in patients with oesophageal SCC.

M. Tachibana et al. / Digestive and Liver Disease 39 (2007) 725–732

Practice points • Tyrosine kinases and its receptors play important roles in growth, migration, and invasion of malignant cells. • The prognostic importance of ephrin-B2 ligand and its receptor EphB4, and its correlation with clinicopathologic characteristics are yet to be delineated in patients with oesophageal carcinoma. • Ephrin-B2 ligand gene expression is upregulated by advancement of the disease process and may be a useful prognostic indicator in patients with oesophageal SCC.

[11]

[12]

[13]

[14]

[15]

[16]

Research agenda

[17] [18]

• Although ephrin-B2 ligand gene expression may represent an interesting observation, more definitive studies need to be required before final confirmation.

[19]

[20]

Conflict of interest statement None declared.

[21]

[22]

References [1] Iizuka T, Isono K, Kakegawa T, Watanabe H. Parameters linked to ten-year survival in Japan of resected esophageal carcinoma. Chest 1989;96:1005–101. [2] Daly JM, Karnell LH, Menck HR. National cancer data base report on esophageal carcinoma. Cancer 1996;78:1820–8. [3] Tachibana M, Kinugasa S, Yoshimura H, Dipok KD, Nagasue N. Extended esophagectomy with 3-field lymph node dissection for esophageal cancer. Arch Surg 2003;138:1383–9. [4] Surawska H, Ma PC, Salgia R. The role of ephrins and Eph receptors in cancer. Cytokine Growth Factor Rev 2004;15:419–33. [5] Davis S, Gale NW, Aldrich TH, Maisonpeirre PC, Lothak V, Pawson T, et al. Ligands for Eph-related receptor tyrosine kinases that require membrane attachment or clustering for activity. Science 1994;266:816–9. [6] Bruckner K, Pasquale EB, Klein R. Tyrosine phosphorylation of transmembrane ligands for Eph receptors. Science 1997;275:1640–3. [7] Hirai H, Maru Y, Hagiwara K, Nishida J, Takaku F. A novel putative tyrosine kinase receptor encoded by the eph gene. Science 1987;238:1717–20. [8] Brantley-Sieders D, Schmidt S, Parker M, Chen J. Eph receptor tyrosine kinases in tumor and tumor microenvironment. Curr Pharm Des 2004;10:3431–42. [9] Augustin HG, Reiss Y. EphB receptors and ephrinB ligands: regulators of vascular assembly and homeostasis. Cell Tissue Res 2003;314:25–31. [10] Bennett BD, Zeigler FC, Gu Q, Fendly B, Goddard AD, Gillett N, et al. Molecular cloning of a ligand for the eph-related receptor

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

731

protein-tyrosine kinase Htk. Proc Natl Acad Sci USA 1995;92:1866– 70. Nikolova Z, Djonov V, Zuercher G, Andres AC, Ziemiecki A. Celltype specific and estrogen dependent expression of the receptor tyrosine kinase EphB4 and its ligand ephrin-B2 during mammary gland morphogenesis. J Cell Sci 1998;111:2741–51. Steube KG, Meyer C, Habig S, Uphoff CC, Drexler HG. Expression of receptor tyrosine kinase HTK (hepatoma transmembrane kinase) and HTK ligand by human leukemia-lymphoma cell lines. Leuk Lymphoma 1999;33:371–6. Takai N, Miyazaki T, Fujisawa K, Nasu K, Miyakawa I. Expression of receptor tyrosine kinase EphB4 and its ligand ephrin-B2 is associated with malignant potential in endometrial cancer. Oncol Rep 2001;8:567–73. Miyazaki T, Kato H, Fukuchi M, Nakajima M, Kuwano H. EphA2 overexpression correlates with poor prognosis in esophageal squamous cell carcinoma. Int J Cancer 2003;103:657–63. Xu F, Zhong W, Li J, Shanshen Z, Cui J, Nesland JM, et al. Predictive value of EphA2 and ephrin-A1 expression in oesophageal squamous cell carcinoma. Anticancer Res 2005;25:2943–50. Fleming ID, American Joint Committee on Cancer Classification (AJCC). AJCC cancer staging manual. Philadelphia: Lippincott; 1997. Kaplan EL, Meier P. Nonparametric estimation for incomplete observation. J Am Stat Assoc 1958;53:457–81. Nakamura R, Kataoka H, Sato N, Kanamori M, Ihara M, Igarashi H, et al. EphA2/EFNA1 expression in human gastric cancer. Cancer Sci 2005;96:42–7. Liu W, Ahmad SA, Jung YD, Reinmuth N, Fan F, Buncana CD, et al. Coexpression of ephrin-Bs and their receptors in colon carcinoma. Cancer 2002;94:934–9. Liu W, Jung YD, Ahmad SA, McCarty MF, Stoeltzing O, Reinmuth N, et al. Effects of overexpression of ephrin-B2 on tumour growth in human colorectal cancer. Br J Cancer 2004;90:1620–6. Duxbury MS, Ito H, Zinner MJ, Ashley SW, Whang EE. Ligation of EphA2 by ephrin A1-Fc inhibits pancreatic adenocarcinoma cellular invasiveness. Biochem Biophys Res Commun 2004;6:1096–102. Berclaz G, Flutsch B, Altermatt HJ, Rohrbach V, Djonov V, Ziemiecki A, et al. Loss of EphB4 receptor tyrosine kinase protein expression during carcinogenesis of the human breast. Oncol Rep 2002;9: 985–9. Noblitt LW, Bangari DS, Shukla S, Knapp DW, Mohammed S, Kinch MS, et al. Decreased tumorigenic potential of EphA2-overexpressing breast cancer cells following treatment with adenoviral vectors that express ephrinA1. Cancer Gene Ther 2004;11:757–66. Wu Q, Suo Z, Risberg B, Karlsson MG, Villman K, Nesland JM. Expression of EphB2 and EphB4 in breast carcinoma. Pathol Oncol Res 2004;10:26–33. Berclaz G, Karamitopoulou E, Mazzucchelli L, Rohrbach V, Dreher E, Ziemiechi A, et al. Activation of the receptor protein tyrosine kinase EphB4 in endometrial hyperplasia and endometrial carcinoma. Ann Oncol 2003;14:220–6. Takai N, Ueda T, Nishida M, Nasu K, Miyakawa I. The relationship between oncogene expression and clinical outcome in endometrial carcinoma. Curr Cancer Drug Targets 2004;4:511–20. Wu D, Suo Z, Kristensen GB, Li S, Troen G, Holm R, et al. Prognostic value of EphA2 and ephrinA-1 in squamous cell cervical carcinoma. Gynecol Oncol 2004;94:312–9. Tang XX, Evans AE, Zhao H, Cnaan A, London W, Cohn SL, et al. High-level expression of EPHB6, EFNB2, and EFNB3 is associated with low tumor stage and high TrkA expression in human neuroblastoma. Clin Cancer Res 1999;5:1491–6. Tang XX, Zhao H, Robinson ME, Cohen B, Cnaan A, London W, et al. Implications of EPHB6, EFNB2, and EFNB3 expressions in human neuroblastoma. Proc Natl Acad Sci USA 2000;97:10936–41. Kinch MS, Moore MB, Harpole DH. Predictive value of the EphA2 receptor tyrosine kinase in lung cancer recurrence and survival. Clin Cancer Res 2003;9:613–8.

732

M. Tachibana et al. / Digestive and Liver Disease 39 (2007) 725–732

[31] Vogt T, Stolz W, Welsh J, Jung B, Kerbel RS, Kobayashi H, et al. Overexpression of Lerk-5/Eplg5 messenger RNA: a novel marker for increased tumorigenicity and metastatic potential in human malignant melanomas. Clin Cancer Res 1998;4:791–7. [32] Easty DJ, Hill SP, Hsu MY, Fallowfield ME, Florenes VA, Herlyn M, et al. Up-regulation of ephrin-A1 during melanoma progression. Int J Cancer 1999;84:494–501. [33] Masood R, Xia G, Smith DL, Scalla P, Still JG, Tulpule A, et al. Ephrin B2 expression in Kaposi sarcoma is induced by human herpesvirus type 8: phenotype switch from venous to arterial endothelium. Blood 2005;105:1310–8.

[34] Heroult M, Schaffner F, Augustin HG. Eph receptor and ephrin ligandmediated interactions during angiogenesis and tumor progression. Exp Cell Res 2006;312:642–50. [35] Tachibana M, Yoshimura H, Kinugasa S, Dhar DK, Shibakita M, Ohno S, et al. Clinicophathologic factors correlated with number of metastatic lymph nodes in oesophageal cancer. Dig Liver Dis 2001;33: 354–8. [36] Sugimachi K, Sumiyoshi K, Nozoe T, Yasuda M, Watanabe M, Kitamura K, et al. Carcinogenesis and histogenesis of esophageal carcinoma. Cancer 1995;75:1440–5.