Correlation among 16 biological factors [p53, p21waf1, MIB-1 (Ki-67), p16INK4A, cyclin D1, E-cadherin, Bcl-2, TNF-α, NF-κB, TGF-β, MMP-7, COX-2, EGFR, HER2/neu, ER, and HIF-1α] and clinical outcomes following curative chemoradiation therapy in 10 patients with esophageal squamous cell carcinoma

The expression levels of 16 proteins were analyzed to identify prognostic correlations in esophageal squamous cell carcinoma (ESCC) treated with concurrent chemoradiation therapy (CCRT). The immunohistochemical expression levels of p53, p21waf1, molecular immunology borstel-1 (MIB-1, Ki-67), p16INK4A, cyclin D1, E-cadherin, Bcl-2, tumor necrosis factor (TNF)-α, nuclear factor (NF)-κB, transforming growth factor (TGF)-β, matrix metalloproteinase (MMP)-7, cyclooxygenase (COX)-2, epidermal growth factor receptor (EGFR), human EGFR type 2 (HER2/neu), estrogen receptor (ER) and hypoxia-inducible factor (HIF)-1α were studied in 10 cases of ESCC treated with CCRT. The patients underwent CCRT between 2000 and 2010. The mean patient age was 68.1 years (range, 46-80 years). The numbers of patients in stages I, II, III and IV of the disease were 2, 2, 3 and 3, respectively. Of the tumors, 8 were positive for p53, 6 for p21waf1, 7 for MIB-1 (Ki-67), 7 for p16INK4A, 7 for cyclin D1, 8 for E-cadherin, 3 for Bcl-2, 0 for TNF-α, 5 for NF-κB, 7 for TGF-β, 9 for MMP-7, 7 for COX-2, 5 for EGFR, 1 for HER2/neu, 1 for ER and 7 for HIF-1α. The 2-year overall survival rate of patients expressing high levels of MIB-1 was 71% (±17%) compared with 0% (P=0.019) for those expressing low levels. For NF-κB, the rate was 0% for patients with high levels compared with 100% (P<0.018) for those with low levels. The 2-year local control rates of HER2/neu were 0% in patients expressing high levels and 88% (±12%) in patients expressing low levels (P=0.027). The 2-year disease-free survival rates of HER2/neu and ER were 0% for patients expressing high levels compared with 56% (±17%) for those with low levels (P=0.027). There were no significant correlations between the expression levels of the other proteins and clinical outcomes. In the present study, high levels of MIB-1 and low levels of NF-κB, HER2 and ER were shown to be good prognostic factors following definitive CCRT for ESCC.


Introduction
Various investigators have studied the prognostic factors of esophageal squamous cell carcinoma (ESCC) treated with concurrent chemoradiation therapy (CCRT). In previous studies, it was considered that the expression levels of p53 (1-3), p21 waf1 (4), molecular immunology borstel-1 (MIB-1) (2,5), p16 INK4A (6), cyclin D1 (2,7,8), E-cadherin (8), Bcl-2 (1), tumor necrosis factor (TNF)-α (9), nuclear factor (NF)-κB (10), transforming growth factor (TGF)-β (11), matrix metalloproteinase (MMP)-7 (12), cyclooxygenase (COX)-2 (13,14), epidermal growth factor receptor (EGFR) (15) and hypoxia-inducible factor (HIF)-1α (16) may be used as prognostic factors for ESCC. However, these studies included patients treated with surgery or neoadjuvant chemoradiation therapy (CRT). In the case of definitive CRT, the prognostic factors may be different from those of surgery or neoadjuvant CRT. Unlike surgical resection alone, which is not reliant on the therapeutic response, and neoadjuvant CRT, which is able to achieve complete responses in the majority of cases by resecting (even for cases without marked responses), in definitive CRT it is important to be able to predict therapeutic responses before starting treatment. Identifying predictive parameters may aid the selection of primary therapies. The present study was conducted using biopsy specimens excised from patients prior to CCRT. The expression levels of 16 proteins were analyzed to identify prognostic correlations in ESCC treated with CCRT.

Patients and methods
Patients. A total of 10 patients who received CCRT for ESCC at the University of Tokyo Hospital (Japan) between June 2000 Correlation among 16 biological factors [p53, p21 waf1 , MIB-1 (Ki-67), p16 INK4A , cyclin D1, E-cadherin, Bcl-2, TNF-α, NF-κB, TGF-β, MMP-7, COX-2, EGFR, HER2/neu, ER, and HIF-1α] and clinical outcomes following curative chemoradiation therapy in 10 patients with esophageal squamous cell carcinoma and June 2010 were selected retrospectively. Only 10 patients (5 long-term survivors and 5 who had succumbed to cancer) were selected and examined, since this was a preliminary study to determine which type of immunostaining should be used in the following larger study. ESCC was confirmed histologically in all 10 patients. The patients consisted of 4 cases of good responses and 6 recurrent cases. The patients were staged according to the American Joint Committee for Cancer Staging and End Results Reporting 1997 staging system (17). The initial staging consisted of a patient history and medical examination, routine blood tests, chest X-rays, upper magnifying endoscopy, chest and upper abdomen computerized tomography (CT), barium contrast X-rays and pulmonary function tests. Bone scans and CT or magnetic resonance imaging of the brain were performed only in cases of clinical suspicion of metastases. Patients with technically unresectable cancer, patients who refused to undergo surgery or those considered medically unfit for surgery were eligible for definitive CCRT. The study was approved by the ethics committee of the University of Tokyo, Tokyo, Japan. Informed consent was obtained from all patients or the patient's families.
CRT method. The details of the treatment method have been previously reported (18)(19)(20). All patients received extended elective nodal irradiation and were treated with 50-50.4 Gy delivered at 1.8-2 Gy per fraction over 5-5.6 weeks. The clinical target volume (CTV) was defined as the whole thoracic esophagus (from the supra-clavicular fossae to the esophagogastric junction). The CTV comprised the M1a and regional lymph nodes (LNs), including positive LNs. The planning target volume was caluclated by adding margins of 5-10 mm to the respective CTVs. The treatment planning was entirely 3-dimensional. At least 4 fields were used (2 anterior-posterior opposed fields and 2 anterior-posterior oblique opposed fields to remove the spinal cord from the radiation fields) and 1 or 2 beams were added using the field-in-field technique if necessary. Treatment was delivered by linear accelerators with 6-10 MV photons.
All patients received chemotherapy (CTx) concurrently with irradiation. The CTx consisted of 2 cycles of 5-fluorouracil (800 mg/m 2 /day, days 1-4 and 29-32, continuous) combined with nedaplatin (80 mg/m 2 , days 1 and 29, bolus) and standard techniques were used for hydration and alkalization. The CTx began on the first day of irradiation. After CCRT, in the adjuvant setting, an additional 1 or 2 cycles of the same doses of CTx were administered to patients who had sufficient bone marrow function and performance status and did not refuse additional CTx.
Follow-up. Patients were followed up on a regular basis, with visits at 1 month following treatment, every 3 months thereafter during the first 2 years and every 6 months thereafter. Chest X-rays were performed at every visit, while chest and upper abdomen CT was performed every 6 months or more frequently at the suspicion of tumor progression. When tumor growth was identified by CT and/or PET, this was defined as local recurrence.
Immunohistochemistry (IHC). Specimens obtained from biopsies under endoscopy before the treatment were used.
Tumor samples were fixed with 10% formaldehyde in phosphate-buffered saline (PBS), embedded in paraffin and 4-µm thick sections were prepared. The sections were deparaffinized and pretreated with various methods, including microwaving, EDTA and heating, that are known to be effective at unmasking reactive sites for antibodies. The characteristics of the 16 primary antibodies used in IHC and their pretreatments are shown in Table I. Subsequently, IHC staining was performed with an automated IHC stainer. The slides were then rinsed briefly in water and counterstained with haematoxylin.
Statistical analysis. Statistical analyses of the correlations between the clinical results (alive vs. deceased, with vs. without local recurrence and with vs. without disease) and molecular markers were performed using Fisher's exact test. The association between each pair of proteins expressed was determined using the Pearson product-moment correlation coefficient.
The Kaplan-Meier product-limit method was used to estimate the probabilities of overall survival (OS), disease-free survival (DFS) and locoregional recurrence-free survival, while the log-rank test was used to estimate any differences. OS was calculated in months from the first day of CCRT to the date of mortality from any cause or to February 2012. Patients who remained alive in February 2012 were censored. P<0.05 was considered to indicate statistically significant differences.
In the present study, 10 patients were treated using a standard CCRT regimen. The 2-year OS, local control (LC) and DFS rates were all 50% and these results were comparable to previous reports (18)(19)(20). There was little difference in the clinical backgrounds between the 6 living and 4 deceased patients. In the cases of higher MIB-1 expression, the OS was significantly improved following definitive CCRT. In the cases with lower NF-κB expression, the OS and DFS rates were significantly improved following definitive CCRT. In the cases of lower HER2 and ER levels, the LC and DFS rates were significantly improved following definitive CCRT. TNF-α showed negative or extremely weak staining in all cases. There was no correlation between MIB-1 and NF-κB, which were independent prognostic factors. There was a strong correlation between HER2 and ER. All patients with high/low HER2 expression also expressed high/low ER. The results of the HER2 staining were entirely in accordance with those of ER.
MIB-1 monoclonal antibody recognizes proliferating cells in the G1, S, G2 and M phases of the cell cycle (22). The correlation between a high MIB-1 index and poor prognosis is well known in breast cancer (23). However, in the present study, in the cases of higher MIB-1 expression, the OS was significantly improved following definitive CCRT. Ressiot et al (5) concluded that the over-expression of MIB-1 was a significant factor for complete endoscopic response following CCRT in 56 patients with esophageal cancer, which is similar to the present results. The correlation between MIB-1 expression and a good response to CCRT is explained by the fact that MIB-1 expression is a marker of cellular proliferation.
NF-κB is a protein complex that controls the transcription of DNA. TNF-α and other transcription factors activate NF-κB. NF-κB activates interleukin (IL)-1 and others and is therefore involved in apoptosis or inflammation. Blocking NF-κB causes tumor cells to stop proliferating, die or become more sensitive to the action of antitumor agents. Thus, NF-κB is the subject of much active research among pharmaceutical companies as a target for anticancer therapy (24). Izzo et al (10) reported that activated NF-κB prior to therapy in 80 patients with esophageal cancer was associated with the lack of a complete pathological response, which supports the present result showing that patients with high expression levels of NF-κB had significantly poorer OS and DFS rates than those showing low levels. Table II  HER2 is a member of the EGFR family which is involved in the complex regulation of cell growth, proliferation and survival. HER2 protein has been implicated in the development of cancer. In addition to breast cancer, HER2 overexpression and gene amplification have also been reported in carcinomas of the colon, bladder, ovary, endometrium, lung, head and neck, esophagus and stomach (25). Akamatsu et al (26) reported that HER2 immunostaining was suitable for predicting resistance to CRT in 34 patients with ESCC. Additionally, according to Mimura et al (27), the survival rate of 66 patients with ESCC showing HER2 gene amplification was significantly worse than those without amplification.
The main function of ER is as a DNA-binding transcription factor regulating gene expression. Estrogen and ER have also been implicated in breast, ovarian, colon, prostate and endometrial cancer. Wang et al (28) suggested that ER expression may predict a better outcome for patients with ESCC.
These results concurred with the present results where patients showing high staining for HER2 and ER had significantly poorer LC and DRS rates than those with low staining. Estrogen promotes tumor proliferation through ER and in cells with overexpression of HER2, tumor proliferation is considered to be increased. The present result demonstrating that the prognosis of cases with overexpression of ER or HER2 was worse is therefore understandable. In the present study, p53, p21 waf1 , p16 INK4A , cyclin D1, E-cadherin, Bcl-2, TNF-α, TGF-β, MMP-7, COX-2, EGFR and HIF-1α were also investigated. The p53 gene is one of the well-known anti-oncogenes involved in apoptosis and the cell cycle (29)(30)(31). Also, the p21 and the p16 INK4A proteins are involved in the cell cycle and induce G1 arrest. Cyclin D1 is a member of the cyclin protein family involved in regulating cell cycle progression. Bcl-2 is a member of the anti-apoptotic family (32,33). E-cadherin is important in cell adhesion and loss of E-cadherin function or expression has been implicated in cancer progression and metastasis. TGF-β is important in stopping the proliferation of normal epithelial cells and promoting the invasion of cancer cells. TNF, being an endogenous pyrogen, is able to induce fever, apoptotic cell death, sepsis through IL-1 and 6 production, cachexia and inflammation, as wells as inhibit tumorigenesis and viral replication. The expression of COX-2 is induced by several stimuli, including growth factors, inflammation and cytokines. COX-2 is upregulated in numerous types of cancer (34). EGFR is the cell-surface receptor for epidermal growth factors and is a growth factor involved in the regulation of cell growth, proliferation and differentiation. Proteins of the MMP family are involved in the breakdown of the extracellular matrix in normal physiological processes, such as embryonic development, reproduction and tissue remodeling, as well as in disease processes, such as arthritis and metastasis (35). HIF-1α is a transcription factor that responds to changes in available oxygen in the cellular environment, particularly to decreases in oxygen or hypoxia (36). In the present study, these cancer growth-causing proteins did not appear to be prognostic factors for ESCC treated with CCRT.
However, the present study had a number of limitations. Firstly, the power may be poor since it involved only 10 patients in a retrospective setting. Secondly, the IHC method used was only able to examine protein expression levels and DNA mutation and amplification were not examined as with the polymerase chain reaction method.
We expect to continue these studies, thereby increasing the number of cases in a retrospective setting and confirming the conclusions of the present study.