Purpose: Cyclooxygenase-2 (COX-2) is generally elevated in tumors compared with normal tissue and apparently has an important role in tumor development. A number of studies have found high expression of COX-2 to be an unfavorable prognostic factor for overall survival in several cancers. However, the influence of COX-2 expression levels on tumor response to chemotherapy has been relatively little studied. The purpose of this study was to ascertain if COX-2 gene expression is associated with tumor response in the clinical treatment of colorectal cancer with the fluoropyrimidine-based therapy S-1.

Experimental Design: Patients with advanced (stage IV) colorectal cancer were treated with S-1 twice daily based on the patient's body surface area (BSA; BSA < 1.25 m2, 80 mg/d; 1.25 m2 ≤ BSA < 1.5 m2, 100 mg/d; BSA ≥ 1.5 m2, 120 mg/d) for 28 days followed by a 2-week period rest. mRNA was isolated from paraffin-embedded pretreatment primary tumor specimens and expression levels of COX-2 relative to β-actin as the internal reference gene were measured using a quantitative reverse transcription-PCR (Taqman) system.

Results: The overall response rate in a group of 44 patients treated with S-1 was 40.9%. Sufficient tumor tissue was available from 40 of these patients for COX-2 mRNA quantitation. COX-2 gene expression was significantly lower in the responding tumors compared with the nonresponders (P = 0.012, Wilcoxon test). Patients with COX-2 values above the cutoff value of 3.28 × 10−3 had a significantly shorter survival than those with COX-2 gene expressions below the cutoff value (adjusted P = 0.031).

Conclusions: Intratumoral COX-2 gene expression is associated with likelihood of response to chemotherapy with S-1 and is a prognostic factor for survival of patients after the start of S-1 chemotherapy.

Cyclooxygenases, which exist in two isoforms designated as COX-1 and COX-2, are rate-limiting enzymes in the formation of prostaglandins from arachidonic acid. Whereas COX-1 is generally considered to be constitutively expressed, COX-2 is highly inducible by various factors and is associated with inflammation (1). Cyclooxygenases are targets for the class of drugs known as the nonsteroidal anti-inflammatory drugs. Epidemiologic studies have shown that patients who regularly took nonsteroidal anti-inflammatory drugs had 40% to 50% decreased incidence and lower mortality from colorectal cancer and other cancers than those who did not take nonsteroidal anti-inflammatory drugs (24). Overexpression of COX-2 (but not COX-1) has been reported in many human malignancies (59). Up-regulation and overexpression of COX-2 in vitro was shown to reduce apoptosis and to increase angiogenesis, invasiveness of malignant cells, and conversion of procarcinogens to carcinogens (1016). Hida et al. (17) reported that a marked increase in COX-2 immunoreactivity in non–small-cell lung cancer, specifically adenocarcinomas, was associated with tumor-invasive lesions and lymph node metastasis, suggesting that increased COX-2 expression may be associated with an invasive and more aggressive phenotype in this disease. Consistent with the implications of these findings, several studies have shown the unfavorable prognostic significance of elevated COX-2 expression. Achiwa et al. (8) showed that high COX-2 expression correlated with inferior survival in a subgroup of stage I lung adenocarcinomas. A previous study from this laboratory showed an association between high COX-2 mRNA expression levels and worse survival in patients with curatively resected non–small-cell lung cancer (18).

Such observations implicating COX-2 in many of the basic processes of tumor development have suggested that targeting COX-2 with specific inhibitors may be an effective strategy for cancer treatment (19, 20). In particular, there has been much interest recently in using COX-2 inhibitors along with conventional anticancer therapy (21) based on the idea that many of the COX-2–regulated genes that contribute to tumor progression may also be determinants of tumor chemo- or radiosensitivity. For example, the putative role of COX-2 in promoting antiapoptotic signaling (16, 22) suggests that tumors that highly express COX-2 should be less sensitive to agents that act by inducing apoptosis. This group includes many of the commonly used anticancer agents. The finding that COX-2 up-regulates the multidrug resistance gene MDR1/pgp70 (23) suggests another mechanism by which COX-2 might promote chemoresistance. A further reason for thinking that COX-2 inhibitors will improve conventional cancer treatment derives from the role of COX-2 in promoting angiogenesis, suggesting the prospect that COX-2 inhibitors will inhibit neoangiogenesis (21, 24).

Clinical trials combining selective COX-2 inhibitors such as celecoxib and rofecoxib have been initiated based on the credible ab initio rationale discussed above and on in vitro and animal model data showing that the COX-2 inhibitors do increase the cytotoxic activity of some anticancer agents (25, 26). However, current information on the effect of COX-2 expression on the intrinsic chemosensitivity of tumors is still very limited. In one of the few clinical studies to address directly the relationship between COX-2 expression and tumor chemosensitivity, Ferrandina et al. (27) showed that increased COX-2 expression was associated with resistance to cisplatin-based chemotherapy in both ovarian and cervical cancer.

The aim of the present study was to determine whether intratumoral COX-2 expression is a predictive factor for tumor response to fluoropyrimidine-based chemotherapy. The hypothesis based on the above discussion would be that high COX-2 levels should render the tumors less responsive to the drugs. The subject population was a group of colorectal cancer patients treated with S-1, an oral drug regimen widely used in Japan to treat various solid tumors such as colorectal cancer, gastric, and breast cancer, which consists of the 5-fluorouracil prodrug tegafur and two modulators, 5-chloro-2,4-dihydroxypyridine and potassium oxonate (28).

Patient population. Eligible patients had (a) a diagnosis of recurrent colorectal cancer after surgical operation or disseminated colorectal cancer; (b) Eastern Cooperative Oncology Group performance status of 0 to 2 with adequate hematologic, hepatic, and renal function; (c) no treatment during the preceding 4 weeks; and (d) a lesion that was measurable by radiological examination.

Treatment. Patients were treated at the Tokyo Women's Medical University with S-1 twice daily for 28 days, followed by a 2-week period rest. The S-1 was given orally after breakfast and dinner. As in previous phase II studies of S-1 in advanced gastric cancer (29, 30), BSA was used to determine the dose of S-1 administered as follows: BSA < 1.25 m2, 80 mg/d; 1.25 m2 ≤ BSA < 1.50 m2, 100 mg/d; BSA ≥ 1.5 m2, 120 mg/d. These treatments were repeated until disease progression as determined by the treating physician or at the physician's discretion.

The protocol was reviewed and approved by an institutional review board and an ethics committee before study activation, and informed consent was obtained from every patient according to the institutional regulations.

Clinical evaluation and response criteria. After two cycles of treatment, measurable disease was reassessed. Response criteria were based on the standard definitions for bimeasurable disease (31). Response was assessed by computer-assisted tomography in liver, lymph node, and lung metastases, as well as in primary lesions. To be classified as a responder, a tumor had to have a 50% reduction in the sum of the products of the perpendicular diameters of the of the indicator lesion without growth of other disease or the appearance of new lesions (31). Overall survival was calculated as the period from the initiation of chemotherapy with S-1 to death. Patients who were still alive at the last follow-up were censored at that time.

Laboratory methods; microdissection. A representative formalin-fixed, paraffin-embedded pre-S1 treatment tumor specimen was selected by a pathologist after examination of the H&E-stained slides. Ten-micron-thick sections were stained with nuclear fast red to enable visualization of histology for laser capture microdissection (PALM Microlaser Technologies AG, Bernried, Germany), which was done to ensure that only tumor cells were studied.

RNA isolation and cDNA synthesis. RNA isolation from formalin-fixed paraffin embedded specimens was done according to a proprietary procedure of Response Genetics, Inc. (Los Angeles, CA) (US patent no. 6,248,535). Following RNA isolation, cDNA was prepared from each sample as described previously (32).

Reverse transcription-PCR. Relative cDNA quantitation for COX-2 and an internal reference gene (β-actin) was done using a fluorescence-based real-time detection method [ABI PRISM 7900 Sequence Detection System (TaqMan), Applied Biosystems, Foster City, CA], as described previously (3234).

The primers and probe sequences used are given in Table 1. The PCR reaction mixture consisted of 600 nmol/L of each primer, 200 nmol/L probe, 2.5 units AmpliTaq Gold Polymerase, 200 μmol/L each dATP, dCTP, dGTP, 400 μmol/L dUTP, 5.5 mmol/L MgCl2, and 1× Taqman Buffer A containing a reference dye to a final volume of 25 μL (Applied Biosystems). Cycling conditions were 50°C for 10 seconds, 95°C for 10 minutes, followed by 46 cycles at 95°C for 15 seconds and 60°C for 1 minute. Colon, liver, and lung RNAs (Stratagene, La Jolla, CA) were used as control calibrators on each plate.

Table 1.

Sequences of PCR primers and probes

Cyclooxygenase-2 (COX-2)  
    Forward primer (COX2-633F) 5′-GCTCAAACATGATGTTTGCATTC-3′ 
    Reverse primer (COX2-711R) 5′-GCTGGCCTCGCTTATGA-3′ 
    Probe (COX2-658T) 5′-(FAM)TGCCCAGCACTTCACGCATCAGTT(TAMRA)-3′ 
β-Actin  
    Forward primer (b-actin-592F) 5′-TGAGCGGGCTACAGCTT-3′ 
    Reverse primer (b-actin-651R) 5′-TCCTTAATGTCACGCACGATTT-3′ 
    Probe (b-actin-611T) 5′-(FAM)ACCACCACGGCCGAGCGG(TAMRA)-3′ 
Cyclooxygenase-2 (COX-2)  
    Forward primer (COX2-633F) 5′-GCTCAAACATGATGTTTGCATTC-3′ 
    Reverse primer (COX2-711R) 5′-GCTGGCCTCGCTTATGA-3′ 
    Probe (COX2-658T) 5′-(FAM)TGCCCAGCACTTCACGCATCAGTT(TAMRA)-3′ 
β-Actin  
    Forward primer (b-actin-592F) 5′-TGAGCGGGCTACAGCTT-3′ 
    Reverse primer (b-actin-651R) 5′-TCCTTAATGTCACGCACGATTT-3′ 
    Probe (b-actin-611T) 5′-(FAM)ACCACCACGGCCGAGCGG(TAMRA)-3′ 

Statistical analysis. The COX-2 gene expression value was expressed as the ratio between two absolute measurements (gene of interest divided by an internal reference gene, β-actin). COX-2 gene expression values in the responding and nonresponding groups of patients were analyzed using the Wilcoxon test. To compare COX-2 gene expression values relative to demographic and clinicopathologic characteristics, the Wilcoxon test was used for two groups and the Kruskal-Wallis test was used for three or more groups. Medians and ranges were used to summarize COX-2 gene expression values within groups of patients. The associations between demographic and clinicopathologic variables and response were examined using Fisher's exact tests.

To evaluate the association of COX-2 gene expression with overall survival, COX-2 value was categorized into a low and a high value. The optimal cut point was determined by adapting the maximally selected χ2 method of Miller and Siegmund (35) and Halpern (36). To determine a P value that would be interpreted as a measure of the strength of the association, 2,000 bootstrap-like simulations were done to estimate the distribution of the maximal χ2 statistics. The association of each of the demographic and clinicopathologic variables with survival was tested using the log-rank test. A Cox proportional hazards model was used to investigate the effect of COX-2 gene expression on overall survival considering potentially important prognostic factors in colorectal cancer. All reported P values are two sided.

A total of 44 primary pre-S1 treatment colorectal cancer specimens from 44 patients, all of whom had stage IV disease, were studied. Eighteen (40.9%) of the patients were classified as responders and 26 patients (59.1%) were nonresponders. COX-2 mRNA level values were successfully obtained for 40 (90.9%) of the 44 patients. The remaining four tumor specimens had levels of β-actin (the internal reference gene) below our accepted cutoff for the assay and were therefore excluded. Nineteen of the 40 patients (47.5%) included in the study were male and the median age for all patients was 64 (range 39-74). Fifteen and 25 patients received S-1 chemotherapy administered at 100 and 120 mg/d, respectively, based on their BSA. The median cycle of S-1 chemotherapy patients received was 6 (range 2-14).

Eighteen patients (45%) for whom COX-2 expressions were measured responded to S-1 chemotherapy and 22 (55%) were nonresponders. The median overall survival for the 40 patients was 15.3 months (95% confidence interval, 12.1-19.8) with a median follow-up period of 8.7 months (15 patients were still alive at the time of the analysis of the data).

Clinicopathologic characteristics and COX-2 mRNA expression levels.COX-2 mRNA levels by demographic and clinicopathologic characteristics were analyzed (Table 2). No significant differences in COX-2 mRNA expression levels for these variables were found (Table 2).

Table 2.

Association between COX-2 gene expression and demographic clinical and pathologic characteristics

COX-2 gene expression
VariablesnMedianRange (min-max)P*
Sex     
    Male 19 1.44 (0.39-7.20) 0.10 
    Female 21 3.28 (0.04-7.73)  
Age, y     
    <65 21 2.43 (0.41-7.73) 0.30 
    ≥65 19 1.80 (0.04-5.86)  
Relapse category     
    Liver metastasis 19 2.03 (0.52-5.65) 0.54 
    Peritoneal and lymph node metastasis 11 1.80 (0.39-7.20)  
    Lung metastasis 3.38 (0.39-5.86)  
    Local recurrence 5.18 (0.04-7.73)  
Histology     
    Well-differentiated adenocarcinoma 30 2.29 (0.39-7.73) 0.89 
    Moderately differentiated adenocarcinoma 1.80 (0.04-7.20)  
    Mucinous adenocarcinoma 1.44 (0.61-4.21)  
Performance status     
    0 5.40 (0.04-5.65) 0.82 
    1 22 2.37 (0.39-7.73)  
    2 15 1.80 (0.41-7.20)  
Dose level     
    100 mg/d 15 3.28 (0.04-7.20) 0.14 
    120 mg/d 25 1.80 (0.39-7.73)  
COX-2 gene expression
VariablesnMedianRange (min-max)P*
Sex     
    Male 19 1.44 (0.39-7.20) 0.10 
    Female 21 3.28 (0.04-7.73)  
Age, y     
    <65 21 2.43 (0.41-7.73) 0.30 
    ≥65 19 1.80 (0.04-5.86)  
Relapse category     
    Liver metastasis 19 2.03 (0.52-5.65) 0.54 
    Peritoneal and lymph node metastasis 11 1.80 (0.39-7.20)  
    Lung metastasis 3.38 (0.39-5.86)  
    Local recurrence 5.18 (0.04-7.73)  
Histology     
    Well-differentiated adenocarcinoma 30 2.29 (0.39-7.73) 0.89 
    Moderately differentiated adenocarcinoma 1.80 (0.04-7.20)  
    Mucinous adenocarcinoma 1.44 (0.61-4.21)  
Performance status     
    0 5.40 (0.04-5.65) 0.82 
    1 22 2.37 (0.39-7.73)  
    2 15 1.80 (0.41-7.20)  
Dose level     
    100 mg/d 15 3.28 (0.04-7.20) 0.14 
    120 mg/d 25 1.80 (0.39-7.73)  
*

Based on the Wilcoxon or Kruskal-Wallis test.

Response to S-1 chemotherapy and COX-2 mRNA expression levels. There was a significant difference in median COX-2 mRNA levels in responders versus nonresponders (P = 0.012, Wilcoxon test; Fig. 1). COX-2 gene expression levels were significantly lower in tumor tissue of responders (median: 1.22 × 10−3; range: 0.04 × 10−3 to 6.07 × 10−3) than those in the nonresponders (median: 3.32 × 10−3; range: 0.39 × 10−3 to 7.73 × 10−3). As can be seen in Fig. 1, 3 of the 15 patients with COX-2 mRNA expression level above 3 × 10−3 experienced a response to S-1 chemotherapy compared with 15 of the 25 patients with COX-2 mRNA expression ≤3 × 10−3. None of demographic and clinicopathologic variables (Table 2) were statistically significantly associated with tumor response to S-1 chemotherapy at the significance level of 0.10 based on Fisher's exact tests (data not shown).

Fig. 1.

COX-2 expression in colorectal tumors according to response (>50% tumor shrinkage) or nonresponse (<50% tumor shrinkage). COX-2 expression values in responding tumors are significantly lower than those in nonresponding tumors (Wilcoxon test, P = 0.012). Median values are indicated by horizontal bar.

Fig. 1.

COX-2 expression in colorectal tumors according to response (>50% tumor shrinkage) or nonresponse (<50% tumor shrinkage). COX-2 expression values in responding tumors are significantly lower than those in nonresponding tumors (Wilcoxon test, P = 0.012). Median values are indicated by horizontal bar.

Close modal

Survival in relation to COX-2 expression. Using 3.28 × 10−3 as a cutoff value, 26 patients (65%) had low COX-2 expression and 14 (35%) had high COX-2 expression. The overall survival by COX-2 expression (low versus high) is shown in Fig. 2. Survival for colorectal cancer patients with COX-2 values below the cutoff value was significantly longer than for those patients with COX-2 above this value (simulated P = 0.031; Fig. 2). The univariate analysis indicated that none of demographic or clinicopathologic factors (Table 2) were statistically significantly associated with survival at the significance level of 0.10 on the basis of the log-rank test (data not shown). Nonetheless, for completeness of the analysis, the association between COX-2, using the 3.28 cut point, and survival was reexamined stratifying for each of the clinicopathologic factors. In these reanalyses, the degree of the association (i.e., observed hazard ratio) was basically unchanged when adjusting each of demographic or clinicopathologic factors (data not shown).

Fig. 2.

Survival (Kaplan-Meier) plots indicating probability of survival for patients with COX-2 expressions above or below the cutoff COX-2 expression value for all patients (COX-2: β-actin, 3.28 × 10−3).

Fig. 2.

Survival (Kaplan-Meier) plots indicating probability of survival for patients with COX-2 expressions above or below the cutoff COX-2 expression value for all patients (COX-2: β-actin, 3.28 × 10−3).

Close modal

In this study, we found a significant association between high COX-2 expression in colorectal tumors and less favorable clinical outcome for patients treated with S-1 chemotherapy, both in terms of tumor response and patients' survival. Most previous clinical correlative studies have concentrated on the prognostic value of COX-2 overexpression for overall survival and only a few have directly addressed the effect of COX-2 expression on tumor response to chemotherapy. The most definitive of these, which involved treatment of ovarian and cervical tumors with platin-based therapy (27), showed that nonresponding tumors had a significantly higher percentage of COX-2–positive tumors than did the responders, whereas almost all progressing tumors were COX-2 positive. The present study shows that COX-2 expression is also associated with tumor response for fluoropyrimidine therapy in colorectal cancer, suggesting that the influence of COX-2 on tumor chemosensitivity may be a general effect acting on a variety of agents with different mechanisms of action in various tumor types.

The mechanism responsible for COX-2–induced desensitization of tumors to chemotherapy (i.e., whether it is due to apoptosis modulation by Bcl-2 induction or to effects on some other process) will require further study. It is also possible that COX-2 may directly influence the expression levels of target enzymes for some drugs, although for the present case, we did not see any correlation between COX-2 and any of the putative response markers for fluoropyrimidines such as TS or DPD gene expressions. Regardless of the mechanism, however, the finding that intratumoral COX-2 overexpression seems to lower the intrinsic chemosensitivity of tumors provides a more substantial experimental underpinning for the use of COX-2 inhibitors with conventional anticancer agents to increase the efficacy of the chemotherapy. Assessment of COX-2 expression status before chemotherapy could be used to identify patients that have a lower probability of response to chemo- or radiotherapy, who may therefore require a different treatment than those with low COX-2 for optimal benefit.

The critical question for the strategy of combining COX-2 inhibitors with chemotherapy is whether administration of a COX-2 inhibitor can reverse the effects of COX-2 overexpression and cause the tumor to acquire the more responsive low-COX-2 phenotype. The indications from both in vivo and in vitro studies thus far are encouraging. For example, whereas artificially increasing the levels of prostaglandin E2 in human colon cancer HCA-7 cells caused up-regulation of the expression of Bcl-2 (12), a COX-2 inhibitor, NS398, down-regulated Bcl-2 expression and concomitantly increased the amount of apoptosis in LNCaP human prostate cancer cells (37). Similarly, inhibition of COX-2 was shown to up-regulate the proapoptotic molecules par-4 (21) and caspase-3 (38) and to enhance drug-induced apoptosis in vitro (39). Such an effect on Bcl-2 expression should in theory increase the sensitivity of the cells to a variety of agents and, indeed, Hida et al. (40) found that a COX-2 inhibitor (nimesulide) increased the potency (decreased the IC50) of a number of anticancer drugs of various different mechanistic classes against non–small-cell lung cancer cells in vitro. In animal models and cell culture (4143), COX-2 inhibitors were found to enhance the effect of radiotherapy. The mechanism of this effect still remains unclear, however, because at least in one case studied (44), the COX-2 inhibitor, although enhancing the radiation effect, did not seem to change the apoptotic fraction of cells. Interestingly, some studies have found that selective COX-2 inhibitors seem to produce anticarcinogenic effects by COX-2–independent pathways as well (45). A recent clinical study found that celecoxib inhibited an induction of prostaglandin E production in lung cancer patients treated with paclitaxel/carboplatin (46). Comparison with historical response rates suggested that the addition of celecoxib may have increased the response rate as well (46). Because of the role of COX-2 in promoting angiogenesis, COX-2 inhibitors might also enhance the effectiveness of angiogenesis-directed therapy in a manner similar to that of epidermal growth factor receptor–targeted agents (47).

The strategy of targeting COX-2 during conventional therapy is attractive from several points of view. First, because COX-2 overexpression is generally observed in tumor tissue but not normal tissue, any enhancement of drug efficacy due to COX-2 inhibition should occur preferentially in tumors, thereby resulting in a gain in therapeutic index. In addition, as discussed previously, COX-2 overexpression seems to be a common phenomenon in cancer, and thus the use of COX-2 inhibitors may provide a means to enhance the effectiveness of treatment of a variety of different tumor types with a variety of different therapies. So far, if the present study is included, we know that high COX-2 levels are associated with lower tumor sensitivity to both cisplatin- and fluoropyrimidine-based therapies. It has also been observed that the responsiveness of cancer cells to selective COX-2 inhibitors is proportional to the degree of COX-2 expression (48), suggesting that patients with low intratumoral COX-2 expression may not benefit from addition of a COX-2 inhibitor to their therapy. Quantitative measurement of COX-2 expression status such as that described here during clinical trials should eventually establish levels of COX-2 expression in tumors at which addition of COX-2 inhibitors can be expected to be of benefit, and thus identify before treatment those patients who would be the best candidates for treatment with COX-2 inhibitors along with their conventional therapy.

In summary, this study shows that COX-2 expression may be a determinant of drug efficacy for fluoropyrimidine therapy of colorectal cancer, thus bolstering the concept that coadministration of COX-2 inhibitors may improve the clinical outcome of conventional chemo/radiotherapy, but with the caveat that the beneficial effects of using COX-2 inhibitors probably vary according to the initial intratumoral COX-2 expression level. Independent studies are necessary to confirm the role of COX-2 expression in predicting tumor response to fluoropyrimidine-based chemotherapy in patients with colorectal cancer. Among further directions for research in this area would be to quantitatively determine the effects of COX-2 inhibitors on the antitumor efficacy of different classes of drugs and to elucidate the molecular mechanism of this enhancement.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Williams CS, DuBois RN. Prostaglandins endoperoxide synthase: why two isoforms?
Am J Physiol
1996
;
270
:
393
–400.
2
DuBois RN, Giardiello FM, Smalley WE. Non steroidal anti-inflammatory drugs, eicosanoids and colorectal cancer.
Gastroenterol Clin North Am
1996
;
25
:
773
–91.
3
Norrish AE, Jackson RT, McRae C. Non steroidal anti-inflammatory drugs and prostate cancer progression.
Int J Cancer
1998
;
77
:
511
–5.
4
Harris RE, Namboodiri KK, Farrar WB. Non steroidal anti-inflammatory drugs and breast cancer.
Epidemiology
1996
;
7
:
203
–5.
5
Tomozawa S, Tsuno NH, Sunami E, et al. Cycloxygenase-2 overexpression correlates with tumor recurrence, especially haematogenous metastasis of colorectal cancer.
Br J Cancer
2000
;
83
:
324
–8.
6
Madaan S, Abel PD, Chaudhary KS, et al. Cytoplasmic induction and overexpression of cyclooxygenase-2 in human prostate cancer: implications for prevention and treatment.
Br J Urol
2000
;
86
:
736
–41.
7
Hwang D, Scollard D, Byrne J, Levine E. Expression of cyclooxygenase 1 and cylooxygenase-2 in human breast cancer.
J Natl Cancer Inst
1998
;
90
:
455
–60.
8
Achiwa H, Yatabe Y, Hida T, et al. Prognostic significance of elevated cyclooxygenase-2 expression in primary resected lung adenocarcinomas.
Clin Cancer Res
1999
;
5
:
1001
–5.
9
Sheehan KM, Sheahan K, O'Donoghue DP, et al. The relationship between cyclooxygenase-2 expression and colorectal cancer.
JAMA
1999
;
282
:
1254
–7.
10
Liu XH, Yao S, Kirschenbaum A, Levine AC. NS389, a selective cyclooxygenase-2 inhibitor, induces apoptosis and down-regulates bcl-2 expression in LNCaP cells.
Cancer Res
1998
;
58
:
4245
–9.
11
Elder D J, Halton DE, Hague A, Paraskeva C. Induction of apoptotic cell death in human colorectal carcinoma cell lines by a cyclooxygenase-2 (COX-2) selective nonsteroidal anti-inflammatory drug: independence from COX-2 protein expression.
Clin Cancer Res
1997
;
3
:
1679
–83.
12
Sheng H, Shao J, Morrow JD, Beauchamp RD, DuBois RN. Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells.
Cancer Res
1998
;
58
:
362
–6.
13
Tsujii M, Dubois RN. Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2.
Cell
1995
;
83
:
493
–501.
14
Tsujii M, Kawano S, Dubois RN. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential.
Proc Natl Acad Sci U S A
1997
;
94
:
3336
–40.
15
Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, Dubois RN. Cyclooxygenase regulates angiogenesis induced by colon cancer cells.
Cell
1998
;
93
:
705
–16.
16
Souza FR, Shewmake K, Beer GD, Cryer B, Spechler JS. Selective inhibition of cyclooxygenase-2 supresses growth and induces apoptosis in human esophageal adenocarcinoma cells.
Cancer Res
2000
;
60
:
5767
–72.
17
Hida T, Yatabe Y, Achiwa H, et al. Increased expression of cyclooxygenase 2 occurs frequently in human lung cancers, specifically in adenocarcinomas.
Cancer Res
1998
;
58
:
3761
–4.
18
Brabender J, Park J, Metzger R, et al. Prognostic significance of cyclooxygenase-2 mRNA expression in non-small cell lung cancer.
Ann Surg
2002
;
235
:
440
–3.
19
Mann M, Sheng H, Shao J, et al. Targeting cyclooxygenase 2 and HER-2/neu pathways inhibits carcinoma growth.
Gastroenterology
2001
;
120
:
1713
–9.
20
Dempke W, Rie C, Grothey A, Schmoll H-J. Cyclooxygenase-2: a novel target for chemotherapy?
J Cancer Res Clin Oncol
2001
;
127
:
411
–7.
21
Sweeney C. Why cyclooxygenase-2 inhibition plus chemotherapy?
Am J Clin Oncol
2003
;
26
:
S122
–5.
22
Ratnasinghe D, Daschner PJ, Anver MR, et al. Cyclooxygenase-2, P-glycoprotein-170 and drug resistance; is chemoprevention against multidrug resistance possible?
Anticancer Res
2001
;
21
:
2141
–7.
23
Zhang Z, DuBois RN. Par-4, a proapoptotic gene, is regulated by NSAIDs in human colon carcinoma cells.
Gastroenterology
2000
;
118
:
1012
–7.
24
Dicker AP, Williams TL, Grant DS. Targeting angiogenic processes by combination rofecoxib and ionizing radiation.
Am J Clin Oncol
2001
;
24
:
438
–42.
25
Liao Z, Milas L, Komaki R, Stevens C, Cox JD. Combination of a COX-2 inhibitor with radiotherapy or radiochemotherapy in the treatment of thoracic Cancer.
Am J Clin Oncol
2003
;
26
:
S85
–91.
26
Dicker AP. COX-2 inhibitors and cancer therapeutics: potential roles for inhibitors of COX-2 in combination with cytotoxic therapy: reports from a symposium held in conjunction with the Radiation Therapy Oncology Group June 2001 Meeting.
Am J Clin Oncol
2003
;
26
:
S46
–7.
27
Ferrandina G, Lauriola L, Distefano MG, et al. Increased cyclooxygenase-2 expression is associated with chemotherapy resistance and poor survival in cervical cancer patients.
J Clin Oncol
2002
;
20
:
973
–81.
28
Shirasaka T, Nakano K, Fukushima M, et al. Antitumor activity of 1M-tegaful-0.4 M 5-chloro-2, 4-dihydroxypyridine-1M potassium oxonate (S-1) against human colon carcinoma orthotopically implanted into nude rats.
Cancer Res
1996
;
56
:
2602
–6.
29
Sakata Y, Ohtsu A, Horikoshi N, Sugimachi K, Mitachi Y, Taguchi T. Late phase II study of novel oral fluoropyrimidine anticancer drug S-1 (1 M tegafur 0.4 M gimestat1 M otastat potassium) in advanced gastric cancer patients.
Eur J Cancer
1998
;
34
:
1715
–20.
30
Koizumi W, Kurihara M, Nakano S, Hasegawa K. Phase II study of S-1, a novel oral derivative of 5-fluorouracil, in advanced gastric cancer.
Oncology
2000
;
58
:
191
–7.
31
Green S, Weiss GR. Southwest Oncology Group standard response criteria, end point definitions, and toxicity criteria.
Investig New Drugs
1992
;
10
:
239
–53.
32
Lord RV, Salonga D, Danenberg KD, et al. Telomerase reverse transcriptase expression is increased early in the Barrett's metaplasia, dysplasia, adenocarcinoma sequence.
J Gastrointest Surg
2000
;
4
:
135
–42.
33
Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative PCR.
Genome Res
1009
;
6
:
986
–94.
34
Gibson UE, Heid CA, Williams PM. A novel method for real time quantitative RT-PCR.
Genome Res
1996
;
6
:
995
–1001.
35
Miller R, Siegmund D. Maximally selected chi-square statistics.
Biometrics
1982
;
38
:
1011
–6.
36
Halpern J. Maximally selected chi-square statistics for small samples.
Biometrics
1982
;
38
:
1017
–23.
37
Liu XH, Yao S, Kirschenbaum A, Levine AC. NS389, a selective cyclooxygenase-2 inhibitor, induces apoptosis and down-regulates bcl-2 expression in LNCaP cells.
Cancer Res
1998
;
58
:
4245
–9.
38
Konturek PC, Konturek SJ, Bielanski W, et al. Influence of COX-2 inhibition by rofecoxib on serum and tumor progastrin and gastrin levels and expression of PPARγ and apoptosis-related proteins in gastric cancer patients.
Dig Dis Sci
2003
;
48
:
2005
–17.
39
Hsueh CT, Chiu CF, Kelsen DP, Schwartz GK. Selective inhibition of cyclooxygenase-2 enhances mitomycin-C-induced apoptosis.
Cancer Chemother Pharmacol
2000
;
45
:
389
–96.
40
Hida T, Kozaki K, Muramatsu H, et al. Cyclooxygenase-2 inhibitor induces apoptosis and enhances cytotoxicity of various anticancer agents in non-small cell lung cancer cell lines.
Clin Cancer Res
2000
;
6
:
2006
–11.
41
Milas L, Kishi K, Hunter N, Mason K, Masferrer JL, Tofilon PJ. Enhancement of tumor response to γ-radiation by an inhibitor of cyclooxygenase-2 enzyme.
J Natl Cancer Inst
1999
;
91
:
1501
–4.
42
Kishi K, Petersen S, Petersen C, et al. Preferential enhancement of tumor radioresponse by cyclooxygenase-2 inhibitor.
Cancer Res
2000
;
60
:
1326
–31.
43
Petersen C, Petersen S, Milas L, Lang FF, Tofilon PJ. Enhancement of intrinsic tumor cell radiosensitivity induced by a selective cyclooxygenase-2 inhibitor.
Clin Cancer Res
2000
;
6
:
2513
–20.
44
Li M, Wu X, Xu X-C. Induction of apoptosis in colon cancer cells by cyclooxygenase-2 inhibitor NS398 through a cytochrome c-dependent pathway in esophageal cancer cells.
Int J Cancer
2001
;
93
:
218
–23.
45
Maier TJ, Schilling K, Schmidt R, Geisslinger G, Grosch S. Cyclooxygenase-2 (COX-2)-dependent and -independent anticarcinogenic effects of celecoxib in human colon carcinoma cells.
Biochem Pharmacol
2004
;
67
:
1469
–78.
46
Altorki NK, Keresztes RS, Port JL, et al. Celecoxib, a selective cyclo-oxygenase-2 inhibitor, enhances the response to preoperative paclitaxel and carboplatin in early-stage non-small-cell lung cancer.
J Clin Oncol
2003
;
21
:
2645
–50.
47
Gatzemeier U. Targeting the HER1/EGFR receptor to improve outcomes in non-small-cell lung cancer.
Oncology (Huntingt)
2003
;
17
:
7
–10.
48
Pyo H, Choy H, Amorino GP, et al. A selective cyclooxygenase-2 inhibitor, NS-398, enhances the effect of radiation in vitro and in vivo preferentially on the cells that express cyclooxygenase-2.
Clin Cancer Res
2002
;
7
:
2998
–3005.