Fax: 011 (81) 822567109
Preclinical evaluation of antisense bcl-2 as a chemosensitizer for patients with gastric carcinoma
Version of Record online: 8 OCT 2004
Copyright © 2004 American Cancer Society
Volume 101, Issue 10, pages 2177–2186, 15 November 2004
How to Cite
Kim, R., Emi, M., Tanabe, K. and Toge, T. (2004), Preclinical evaluation of antisense bcl-2 as a chemosensitizer for patients with gastric carcinoma. Cancer, 101: 2177–2186. doi: 10.1002/cncr.20636
- Issue online: 29 OCT 2004
- Version of Record online: 8 OCT 2004
- Manuscript Accepted: 2 AUG 2004
- Manuscript Revised: 29 JUN 2004
- Manuscript Received: 29 MAR 2004
- antisense bcl-2;
- gastric carcinoma
Because bcl-2 is a critical factor for anticancer drug-induced apoptosis, the authors conducted a preclinical evaluation of antisense (AS) bcl-2 as an enhancer of the chemotherapeutic effect in the treatment of patietns with gastric carcinoma.
AS bcl-2 was used with 18-mer phosphorothiated oligonucleotides in the MKN-45 gastric carcinoma cell line. Drug sensitivity in vitro was evaluated using the methyl-thiazoldiphenyl tetrazolium assay, and antitumor effects in vivo were evaluated using the nude mouse xenograft. Apoptosis was determined with the terminal deoxyuridine triphosphate nick-end labeling assay. AS bcl-2 in vitro was treated with lipofectin, whereas it was administered intraperitoneally for 6 consecutive days twice every 2 weeks in vivo. Anticancer drugs were administered intraperitoneally four times per week.
bcl-2 was down-regulated to 60% of its initial value after treatment with 1.0 μM AS bcl-2 compared with the controls of random and mismatched oligonucleotides. Drug sensitivity to doxorubicin, cisplatin, and paclitaxel (TXL) was increased 3–4-fold when used in combination with AS bcl-2, which was determined with 50% inhibitory concentration values, compared with the control group. Increased drug sensitivity was associated with apoptosis, which increased in Bax and poly-adenosine diphosphate (ADP-ribose) polymerase and decreased in phosphorylated Akt (pAkt). The antitumor effect of cisplatin and TXL in vivo was enhanced significantly in combination with AS bcl-2. Down-regulation of bcl-2 was observed on Day 4 after the treatment with AS bcl-2.
Combination treatment with AS bcl-2 and anticancer drugs, including cisplatin and TXL, may be a new strategy for enhancing chemotherapeutic effects in the treatment of gastric carcinoma. Cancer 2004. © 2004 American Cancer Society.
bcl-2 is an important antiapoptotic protein that inhibits anticancer drug-induced apoptosis.1 Anticancer drug-induced apoptosis is associated with death receptor-dependent and receptor-independent pathways, in which mitochondria act as a central gate that regulates the release of cytochrome c through the anion channel followed by activation of caspase cascades.2 Overexpression of bcl-2 has been associated with drug resistance in various human malignancies; the expression of bcl-2 was down-regulated in chemotherapy responders, whereas its expression was increased or was not changed in nonresponders.3, 4 In addition, the introduction of the bcl-2 gene in vitro was followed by decreased drug sensitivity in tumor cells.5, 6 These findings indicate that bcl-2 plays a critical role in determining responsiveness to chemotherapy in tumor cells. Because the overexpression of bcl-2 was observed not only in hematologic malignancies, including leukemia,7 lymphoma,8 and myeloma,9 and also in solid tumors, including nonsmall cell lung carcinoma,10 melanoma,11 hormone-resistant prostate carcinoma,12 breast carcinoma,13 colorectal carcinoma,14 and gastric carcinoma,15 it is likely that the overexpression of bcl-2 also may be involved in inducing drug resistance to chemotherapy in patients with gastric carcinoma.
Although gastric carcinoma is a common type of malignancy in Japan, to our knowledge standard chemotherapy or adjuvant regimens for patients with advanced disease have not been established to date. In fact, the traditional combination regimen, which is 5-fluorouracil (5-FU) based or cisplatin based, has been used extensively both in Western countries and in Japan.16, 17 Furthermore, the development of new anticancer drugs, such as CPT-11,18 S119 (a new fluorouracil derivative), and taxanes.20, 21 have succeeded in increasing the response rates and improving the median survival of patients with advanced gastric carcinoma. Currently, new combination treatments with S1 plus cisplatin, S1 plus docetaxel, or S1 plus CPT-11 are being evaluated with expectations of high response rates, which may be associated with the prolongation of overall survival. Nevertheless, the incorporation of new molecular targeting drugs in the therapeutic armamentarium, including antisense (AS) agents, monoclonal antibodies, and antiangiogenic agents, still is behind the therapeutic developments for the treatment of other solid tumors. Advances of new molecular targeting drugs will provide improvements in the survival of patients with advanced and recurrent gastric carcinoma. Recently, randomized Phase III trials of AS bcl-2 in combination with standard chemotherapy is being evaluated in patients with metastatic or recurrent melanoma and nonsmall cell lung carcinoma,22 and a Phase I/II trial for breast carcinoma is ongoing.23 In the current study, we evaluated the preclinical effect of AS bcl-2 on chemotherapy for patients with gastric carcinoma to assess its potential use as a chemosensitizer that may enhance the therapeutic effect of chemotherapeutic drugs.
MATERIALS AND METHODS
Chemicals, Drugs, and Supplies
Chemicals and supplies were purchased from the following suppliers. RPMI 1640 and fetal bovine serum were from GIBCO BRL (Tokyo, Japan); AS bcl-2 phosphorothioate oligonucleotide, mismatch of AS bcl-2, and random 18-mer oligonucleotide were from Biologica Company, Ltd. (Tokyo, Japan); Transfectam was from Biosera (Marlborough, France); Antibodies for specific immune blotting, including those against bcl-2, bcl-xL, Bax, phosphorylated Akt (pAkt), poly-adenosine diphosphate (ADP ribose) polymerase (PARP), and actin, were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA); peroxidase-conjugated antiimmunoglobulin G was from Upstate Biotechnology (Lake Placid, NY); the enhanced chemiluminescence (ECL) Western blot detection system was from Amersham Pharmacia Biotech, Ltd. (Buckinghamshire, U.K.); the in situ cell death detection kit was from Boehringer Mannheim (Indianapolis, IN); doxorubicin (DOX), mitomycin (MMC), cisplatin, and 5-FU were from Kyowa Hakko Company, Ltd (Tokyo, Japan); and paclitaxel (TXL) was from Bristol-Myers K.K. (Tokyo, Japan). DOX, cisplatin, and 5-FU were prepared with saline solution, and TXL was dissolved with dimethyl sulfoxide (DMSO).
A human gastric carcinoma cell line, MKN-45, was used.24 Cells were cultured in RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum plus 1% penicillin/streptomycin and 1% glutamine. Cultures were maintained in a 5% CO2-humidified incubator at 37 °C, and experiments were performed with exponentially growing cells.
The Methyl-Thiazoldiphenyl Tetrazolium Assay
The methyl-thiazoldiphenyl tetrazolium (MTT) assay was performed according to a previously described method.25, 26 Briefly, cells (2 × 103) were treated with anticancer drugs for 48 hours. The tetrazolium agent was then added to each well, and this was followed by a 4-hour incubation. Then, the culture supernatant fluid was removed and the cells were dissolved and mixed with DMSO. After thorough formazan solubilization, the absorbance of each well was measured with a microculture plate reader at 540 nm. Growth inhibition was expressed as the ratio of the mean absorbance of treated cells to that of control cells. Each experiment was performed in triplicate, and the growth-inhibition rate was calculated as a 50% inhibitory concentration (IC50)value.
The Terminal Deoxyuridine Triphosphate Nick-End Labeling Assay
The terminal deoxyuridine triphosphate (dUTP) nick-end labeling assay for assessment of apoptotic cell death was performed according to the manufacturer's protocol, as described previously.27 In brief, the cells were treated with drugs and then incubated with terminal deoxynucleotidyl transferase (TdT) and TdT buffer (biotin-16-dUTP), the reaction was made visible with nitroblue tetrazolium, and the cells were photographed. Apoptotic cell death was evaluated by the appearance of brightly labeled nuclei and apoptotic bodies. The induction of apoptosis was represented as the mean ± standard deviation upon examination of four random microscopic fields and compared with untreated cells.
Treatment with AS bcl-2 Oligonucleotide
The AS bcl-2 phosphorothioate oligonucleotide sequence was as follows: 5′-TCT CCC AGC GTG CGC CAT-3′.28 The sequences of random (reverse) and mismatched oligonucleotide (ODN) were as follows: 5′-TAC CGC GTG CGA CCC TCT-3′; 5′-TCT CCC AGC ATG TGC CAT-3′. AS and control ODNs were added to the cells in the form of complexes with cationic lipopolyamine according to the manufacturer's protocol, as described previously.25, 26 Briefly, the medium was removed, and serum-free RPMI was added to the cells. The ODN-lipopolyamine complex was prepared and added drop-wise to the cells. After 12 hours of incubation, cells were washed with serum-free RPMI, and RPMI containing 10% fetal bovine serum was added to the cells. Cells were then treated with AS bcl-2 ODN twice for 12 hours per treatment. After the second treatment with AS bcl-2 ODN, cells were incubated with anticancer drugs for 48 hours, and cell viability was assessed by the MTT assay.
Western Blot Analysis
Western blot analysis was performed according to a previously described method.25–27 In brief, whole-cell lysates were extracted with lysis buffer (10 mM Tris-HCl, pH 8.0; 0.15 M NaCl; 1 mM ethylenediamine tetraacetic acid; 10 mM 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate [CHAPS]; 10 μg/mL aprotinin, and 0.02 mM phenyl methyl sulfonyl fluoride), and the protein concentration was determined in 10 μg of each sample. Proteins were separated on 10% sodium dodecyl sulfate-polyacrylamide gels and were transferred to polyvinyl difluoride membranes with a polyblot. Filters were blocked in 5% skim milk in phosphate buffered saline (PBS) for 1 hour. After incubation with primary antibody for 24–48 hours, filters were washed with PBS and then incubated with the secondary antibody in 5% skim milk in PBS for 1 hour. After washing the membranes with PBS for 30 minutes, bands were visualized with an ECL Western blot detection system. Films were exposed for 15–30 minutes.
In Vivo Experiments using the Nude Mouse Model
The antitumor effect in vivo was evaluated with the xenograft transplanted into nude mice according to the previously described method.25, 29 The MKN-45 gastric carcinoma cells were transplanted into the flank of nude mice. When the tumor size reached approximately 100 mm3, the mice were divided arbitrarily into 4 groups (n = 5 mice per group), according to no treatment provided, treatment with random ODNs, treatment with anticancer drug, and treatment with AS bcl-2 and anticancer drug, respectively. The AS bcl-2 was administered intraperitoneally at a dose of 5 mg/kg for 6 consecutive days twice in every 2 weeks, whereas the anticancer drug was administered intraperitoneally 4 times at one-third of a lethal dose to 50% of the animals each week, followed twice by AS bcl-2 (Fig. 1). The relative tumor volume was calculated as the following: tumor volume = length × width2/2. The toxicity of the treatment was evaluated by body weight loss in the treated mice.
Statistical significance was determined with the Student t test with analysis of variance (ANOVA) and with the Fisher least significant difference (LSD) test for multiple comparisons.
Down-regulation of bcl-2 by AS bcl-2
To assess the specific effect of AS bcl-2 on the expression of Bcl-2 protein in vitro, the effect of AS Bcl-2 was examined in a dose-escalation experiment. Figure 2A shows that the expression of Bcl-2 protein was suppressed by treatment with AS Bcl-2 but not by treatment with random ODNs, although mismatched ODN treatment suppressed the expression of Bcl-2 slightly. Therefore, it is likely that the down-regulation by AS Bcl-2 is a specific effect. According to the concentration-dependent inhibition of Bcl-2, it is likely that the maximal effect was reached at 1.0 μM (Fig. 2B), because toxic effects of both AS Bcl-2 and random ODNs were observed after treatment with > 1.0 μM (data not shown). These findings indicate that the appropriate concentration of AS Bcl-2 for specific inhibition of Bcl-2 is 1.0 μM in MKN-45 gastric carcinoma cells.
Changes in Drug Sensitivity in Combination with AS Bcl-2
To assess whether the treatment with AS Bcl-2 increased drug sensitivity in vitro, the IC50 value of the AS Bcl-2 and anticancer drug combination was compared with that of the anticancer drug alone, in which the control was used with random ODNs. Table 1 shows that, although the IC50 value in the treatment with random ODNs alone was changed slightly compared with in the no-treatment controls, the IC50 values of DOX, CDDP, TXL in combination with AS Bcl-2 were decreased 3–4-fold compared with the values of the drugs alone. In contrast, the IC50 values of for MMC and 5-FU were decreased < 2-fold.
|Anticancer drug||Combination IC50 value (μM)|
|AS (−)||Random ODNs||AS bcl-2b|
Induction of Apoptotic Cell Death by AS Bcl-2
Because combination treatment with anticancer drugs and AS Bcl-2 increases drug sensitivity, in particular sensitivity to DOX, CDDP, and TXL in vitro, we examined whether the increased drug sensitivity was associated with induction of apoptotic cell death. Figure 3 shows that random ODNs alone increased apoptotic cell death slightly (4.1% ± 1.1%), but treatment with the combination of AS Bcl-2 and CDDP or TXL significantly increased apoptotic cell death compared with CDDP or TXL alone (20.1% ± 1.4% vs. 7.2% ± 2.3%, respectively; P < 0.05; and 24.2% ± 5.5% vs. 8.4% ± 3.3%, respectively; P < 0.05; Student t test). In contrast, combination treatment with random ODNs and these drugs did not increase apoptotic cell death. All of these results indicate that combination treatment with AS Bcl-2 increased anticancer drug-induced apoptotic cell death significantly, resulting in increased drug sensitivity.
To analyze the mechanism by which AS Bcl-2 enhances anticancer drug-induced apoptotic cell death, the expression of apoptosis-related proteins was evaluated. Figure 4 shows that, although treatment with AS Bcl-2 alone did not decrease pAkt, treatment with CDDP alone decreased pAkt, increased Bax, and cleaved PARP. Furthermore, the combination of AS Bcl-2 and CDDP induced a steeper decrease in pAkt and higher increases of Bax and cleaved PARP than CDDP alone. These findings indicate that AS Bcl-2-enhanced anticancer drug-induced apoptotic cell death is mediated by decreases in Bcl-2 and pAkt followed by the activation of caspase cascades.
Enhancement of Antitumor Effect of Anticancer Drugs by AS Bcl-2
To confirm the AS Bcl-2-induced increased drug sensitivity in vitro in association with the induction of apoptotic cell death, the antitumor effect was evaluated in vivo using the xenograft transplanted into nude mice. Figure 5 indicates that treatment with AS Bcl-2 alone showed a slight antitumor effect compared with that of random ODNs or no treatment. Despite the fact that an antitumor effect of TXL or CDDP was observed after a single treatment, the effect of combination treatment with AS Bcl-2 was enhanced significantly among the treatment groups (P < 0.05; ANOVA with Fisher LSD test). In contrast, the antitumor effect of 5-FU also was enhanced somewhat in combination with AS Bcl-2, although 5-FU per se showed an antitumor effect to the MKN-45 gastric carcinoma xenograft. These results indicate that treatment with AS Bcl-2 enhances the chemotherapeutic effect not only of a drug to which gastric carcinoma cells are sensitive but also of a drug to which gastric carcinoma cells are resistant. In addition, the toxicity of the combination treatment with AS Bcl-2 and anticancer drug was assessed by decreased weight loss in treated mice compared with mice in the no-treatment group; the observed weight loss in the treated mice was < 10% (data not shown).
To assess whether the chemosensitizing effect of AS Bcl-2 was due to down-regulation of Bcl-2 protein in vivo, the expression of Bcl-2 protein was evaluated after the intraperitoneal administration of AS Bcl-2. Figure 6 shows that the down-regulation of Bcl-2 was observed at Day 4 after treatment and continued until Day 15 after treatment with Bcl-2. In contrast, the expression of Bcl-xL protein was not changed during the treatment with AS Bcl-2. These findings suggest that the enhancement of the antitumor effect by the combination of AS Bcl-2 and anticancer drug is mediated by the down-regulation of Bcl-2, which is associated with activation of the drug-induced apoptosis pathway. Figure 6 also shows that there is no compensatory relation between Bcl-2 and Bcl-xL in the gastric carcinoma cells.
In the current study, we showed the preclinical efficacy of AS bcl-2 in enhancing the chemotherapeutic effects of anticancer drugs in gastric carcinoma cells. In particular, the antitumor effects of CDDP and TXL were enhanced significantly in combination with AS bcl-2. With regard to 5-FU, the drug was effective, per se, and combination treatment with AS bcl-2 was somewhat more effective than treatment with the drug alone. A previous study also reported that combination treatment with AS Bcl-2 and CDDP enhanced the antitumor effect of CDDP in the severe combined immunodeficiency (SCID) mouse model.30 These findings imply that the therapeutic efficacy of the commonly used drugs CDDP and TXL for gastric carcinoma can be enhanced by AS Bcl-2. In addition, the mechanism by which AS Bcl-2 increases the anticancer drug-induced tumor responses is explained through activation of an apoptotic pathway involved in down-regulation of antiapoptotic proteins, which leads to the activation of caspase cascades. Indeed, down-regulation of the Bcl-2 protein is associated with a decrease in pAkt, which inhibits apoptosis by blocking the translocation of Bax into mitochondria for releasing cytochrome c, thereby activating caspase-3 and PARP, leading to apoptosis. It is likely that the inhibition of antiapoptotic signaling pathways, which are not only Bcl-2-dependent but also Akt-dependent, contributed to the increased chemotherapeutic effect in gastric carcinoma cells of anticancer drugs combined with AS Bcl-2. The other important observation regarding the difference in increased antitumor effects between CDDP, TXL, and 5-FU in combination with AS Bcl-2 may be due to the drug-induced cytotoxic action that is involved in Bcl-2. In turn, CDDP induces oxidative stress as a major cause of its effects, and Bcl-2 is an antioxidant in addition to its direct effects on maintaining mitochondrial membrane integrity.31 The effect of AS Bcl-2 may be to decrease resistance to oxidation. In fact, it has been shown that Bcl-2-overexpressing cells express relatively high levels of antioxidant enzymes and GSH,32, 33 and the down-regulation of Bcl-2 has been associated with GSH depletion, which blocks apoptosis.34 Similarly, Bcl-2 may be involved in microtubule function, although it interacts indirectly with taxanes. Paclitaxel decreases expression of Bcl-235 and phosphorylates Bcl-2,36 which facilitates dimerization with Bax to inhibit apoptosis. AS Bcl-2 may be effective for modulating these function. With regard to 5-FU, however, because Bcl-2 may not effect DNA synthesis directly, AS Bcl-2 did not interact with 5-FU; therefore, not much modulation of the antitumor effect was found. Thus, our preclinical evaluation of AS Bcl-2 suggests that, in targeting therapy through antiapoptotic proteins, the use of Bcl-2 is a new strategy for modulating the drug response in patients with gastric carcinoma.
Extensive clinical trials using AS Bcl-2 (G3139; Genasense, Oblimersen Sodium; Genata Inc.) are being conducted for hematologic malignancies and solid tumors.22, 23 Randomized Phase III trials are being evaluated for patients with refractory, chronic lymphocytic leukemia and refractory multiple myeloma comparing standard chemotherapy with AS Bcl-2 plus standard chemotherapy. Similarly, for solid tumors, randomized Phase III trials of standard chemotherapy with and without Bcl-2 are being conducted for patients with metastatic malignant melanoma and advanced nonsmall cell lung carcinoma who failed on initial therapy. In addition, a Phase I/II trial of AS Bcl-2 and DOX/docetaxel for patients with metastatic and locally advanced breast carcinoma is being evaluated. These extended clinical trials for various types of malignancies are expected to provide a potential survival benefit through the use of AS Bcl-2. Because the overexpression of Bcl-2 is observed in numerous types of carcinoma, including hematologic malignancies and solid tumors, and because the Bcl-2 protein has a critical role in inhibiting apoptosis through its mitochondrial function, an increased antitumor response induced by AS Bcl-2 may contribute to the prolongation of survival in patients with recurrent or refractory disease who fail to respond to initial chemotherapy.
However, because proof of principle in combination treatment with AS Bcl-2 is very important for the rationale of Bcl-2 targeting therapy, a demonstration of down-regulation of Bcl-2 protein in target lesions will be required to explain an increased tumor response. Previous studies have shown that the down-regulation of Bcl-2 was observed in peripheral blood mononuclear cells37 and bone marrow stem cells,38 which may be surrogate markers for monitoring effectiveness of the targeting. However, the down-regulation of Bcl-2 in the actual targeted lesion was not documented fully in previous clinical studies, even in samples that were removed from the targeted lesions, which are difficult to obtain for analysis. A previous study has documented a sensitization to mitoxantrone after Bcl-2 down-regulation in a solid tumor-bearing murine model.39 Furthermore, another recent study on the treatment of imatinib-resistant, Philadelphia chromosome (Ph)-positive leukemia showed that treatment with AS Bcl-2 overcame the imatinib resistance of BCR-ABL-transformed cells transplanted into nude mice, suggesting a clinical implication of combination treatment with AS Bcl-2 and imatinib in patients with imatinib-resistant, Ph-positive leukemia.40 In that report, it is noteworthy that, after 10 consecutive days of intraperitoneal administration of 7 mg/kg AS Bcl-2, Bcl-2 showed a transient decrease at Day 5 but recovered to the basal level at Day 15 in the Ph-positive leukemia cells. In the current study, the down-regulation of Bcl-2 after intraperitoneal administration of 5 mg/kg AS Bcl-2 for 6 consecutive days was detected at Day 4 and continued up to Day 15 after the treatment. Although it is unclear whether the continuing down-regulation of Bcl-2 by the AS compound in the nude mouse model is due to differences in the duration of administration of AS Bcl-2 or the types of tumor cells, it is conceivable that the consecutive treatment with AS Bcl-2 is effective for down-regulating the Bcl-2 protein.
The modulation of signal-transduction pathways targeting Bcl-2 for apoptosis is a rationale for enhancing the chemotherapeutic effects of anticancer drugs. However, although the Bcl-2 mRNA was expressed in 75% of bone marrow stem cells from patients with acute leukemia, it was not observed in all responders to combination treatment with AS Bcl-2 and anticancer drugs.38 This finding not only suggests that AS Bcl-2 enhances the chemotherapeutic effect, but it also suggests the presence of other biologic factors that induce apoptosis, including antiapoptotic proteins such as Akt, nuclear factor κB, Bcl-xL, after treatment with AS Bcl-2. An immunostimulatory effect of AS Bcl-2 has been reported as another possibility.41, 42 The CpG islands contained in AS Bcl-2 ODNs can induce Th1 cytokines, such as interleukin 12 (IL-12), and splenomegaly in SCID mice.43 Despite the fact that the inhibition of the immunostimulatory effect by methylation of the CpG islands block the secretion of IL-12 and splenomegaly, the therapeutic effect of the methylated AS Bcl-2 was not different from that of unmethylated AS Bcl-2.43 In addition, a recent report indicates that the AS Bcl-2 (G3139, Oblimersen) may inhibit PC3 prostate carcinoma cell growth in part through a bis-CpG-dependent, non-AS mechanism.44 These findings do not eliminate the possibility of an immunostimulatory effect of AS Bcl-2, because it remains to be clarified whether AS Bcl-2 really can stimulate immune responses that include B cells, natural killer cells, and T cells and can promote an active tumor response in a tumor host. Clinical evaluation of a Phase II study will be necessary to assess the immune response elicited by AS Bcl-2.
The use of antisense Bcl-2 as a sensitizer for the chemotherapeutic effects of anticancer drugs in patient with gastric carcinoma may be a promising strategy for the enhancement of therapeutic efficacy. Several randomized and nonrandomized studies on solid tumors are ongoing, and the clinical evaluation results from those studies will be required for to document the enhanced tumor response to combinations of AS Bcl-2 and anticancer drugs. Molecular targeting therapy using AS Bcl-2 may be a good strategy for prolonging the survival of patients with advanced and metastatic gastric carcinoma by overcoming their poor prognosis.
- 9Expression of the bcl-2 family of pro- and anti-apoptotic genes in multiple myeloma and normal plasma cells: regulation during interleukin-6 (IL-6)-induced growth and survival. Eur J Haematol. 2002; 69: 76–89., , , , , .
- 16Japan Clinical Oncology Group Study (JCOG9205). Randomized Phase III trial of fluorouracil alone versus fluorouracil plus cisplatin versus uracil and tegafur plus mitomycin in patients with unresectable, advanced gastric cancer: The Japan Clinical Oncology Group Study (JCOG9205). J Clin Oncol. 2003; 21: 54–59., , , et al.,