You are here

Role of chemotherapy for advanced/recurrent gastric cancer: An individual-patient-data meta-analysis

European Journal of Cancer, Volume 49, Issue 7, May 2013, Pages 1565 - 1577

Editorial comment from Professor Lordick:
This meta-analysis illustrates that a variety of different chemotherapy regimens leads to comparable survival outcomes in advanced gastric cancer. A plateau may have been reached. Even the addition of new chemotherapy drugs to established combinations does only marginally improve outcomes. The overall median survival remains below one year. Progress is needed. This will not be reached without a better understanding of gastric cancer tumor biology and without the implementation of effective biologically targeted drugs to specific patient subgroups.

Abstract

We conducted an individual-patient-data meta-analysis of the efficacy of chemotherapy on overall survival (OS) and progression-free survival (PFS) in advanced/recurrent gastric cancer (AGC).

Our primary research question was whether the experimental arms of the trials included in the meta-analysis showed a benefit as compared with their corresponding control arms. MEDLINE (up to 2010), Cochrane Central Register of Controlled Trials, National Institutes of Health (NIH) trial registry and proceedings of major oncologic and gastrointestinal cancer meetings were searched. Randomised controlled trials for AGC closed to patient accrual before the end of 2006 were eligible.

As of December 2010, individual patient data were available from 22 trials (4245 patients, representing 47% of the targeted data) of 55 eligible trials. The overall comparison of experimental arms with the corresponding control arms showed statistically significant differences in terms of both OS and PFS. Hazard ratio was 0.88 (95% confidence interval 0.82–0.94, P < 0.0001) for OS and 0.81 (0.76–0.88, P < 0.0001) for PFS. The results of the sub-analysis of adding a given chemotherapeutic agent to any chemotherapy confirm the results of the overall analysis, with a hazard reduction of 11% for OS (P < 0.01) and 26% for PFS (P < 0.0001).

This meta-analysis of individual patient data shows that the additions of experimental chemotherapeutic agents to pre-existing control or standard regimens have produced a modest improvement in OS and PFS. Median survival remained below 1 year for all investigated chemotherapy regimens and none emerged as a clear standard.

Keywords: Advanced gastric cancer, Recurrent gastric cancer, Chemotherapy, Individual patient data, Meta-analysis, Randomised trial.

1. Introduction

Gastric cancer is the second most common cause of cancer mortality worldwide and remains a major global public health problem, although the incidence of gastric cancer has decreased over the three last decades. 1 In the United States (US), as many as 75% of all patients have advanced or metastatic disease at the time of diagnosis, and relative 5-year survival rates for gastric cancer of all stages are about 25%. 2

After more than 30 years of clinical research the prognosis of advanced/recurrent gastric cancer (AGC) remains grim, with median overall survival (OS) barely exceeding 1 year until recently. Small randomised studies have concluded that combination chemotherapies can significantly prolong median OS compared to best supportive care.3 and 4 Several combination chemotherapies have been developed and assessed in randomised studies: FAM (5-fluorouracil (5-FU), doxorubicin and mitomycin C (MMC)), 5 FAMTX (5-FU, doxorubicin, methotrexate), 6 EAP (etoposide, doxorubicin, cisplatin), 7 CF (cisplatin, 5-FU), 8 ELF (etoposide, 5-FU, leucovorin) 4 and ECF (epirubicin, cisplatin, 5-FU) 9 and 10; and more recently several combinations with new cytotoxic drugs have been tested, including DCF (docetaxel, cisplatin and 5-FU), 11 IFL (irinotecan, leucovorin, 5-FU), 12 XP (capecitabine, cisplatin)13 and 14 and EOX (epirubicin, oxaliplatin and capecitabine). 15 Recent systematic reviews or meta-analyses were either based on published data or focused on specific regimens ( eTable 1 ). In the setting of AGC, however, there are many chemotherapeutic regimens to choose from, but the true benefit of chemotherapy remains unclear, with median OS often less than 1 year together with significant side effects.

GASTRIC (Global Advanced/Adjuvant Stomach Tumor Research International Collaboration) is an academic, worldwide project which conducts individual patient-based (IPD) meta-analyses of randomised trials of post-operative adjuvant chemotherapy for resectable gastric cancer, or chemotherapy for AGC. 15 Given the large number of investigated treatments, the absence of well-accepted control, as well as their modest sample sizes and lack of clear results, IPD meta-analysis affords a unique opportunity to cast light on this vexing issue. While waiting for more effective treatments, a better understanding of current results using an IPD meta-analysis should help clarify the overall impact of chemotherapy on AGC.

2. Patients and methods

2.1. Study selection and data extraction

This IPD-based meta-analysis was conducted to determine the efficacy of chemotherapy on OS and progression-free survival (PFS) in AGC. The main inclusion criterion was that the study had to be a randomised controlled trial for AGC that had been closed to patient accrual before the end of 2006. We specified 10 group comparisons among eligible trials, which investigated 5-FU, MMC, anthracyclines, platinum agents, irinotecan and taxanes. Trials investigating immunotherapy, neoadjuvant or adjuvant therapy were excluded. Trials with radiotherapy or intraperitoneal chemotherapy were outside the scope of our research and were excluded.

Data from all published randomised trials comparing chemotherapeutic effects for AGCs were sought electronically. The strategy filter for computerised bibliographic searches of MEDLINE (1970–2010) is described in eTable 2 . No restriction on language of publication was considered. The Cochrane Central Register of Controlled trials and the National Institutes of Health (NIH) trial registry ( clinicalTrials.gov ) and proceedings’ books from major oncologic and gastrointestinal cancer meetings were also examined for published results. We also used the search results of the well-conducted review based on aggregate data from Wagner et al.16 and 17 To ensure that all relevant trials were included, researchers with expertise in the area were queried for the existence of unpublished trials.

The following data were requested for all individual patients: centre, date of randomisation, date of birth, sex, performance status, disease status at baseline, number of organs involved at study entry, location of metastasis, histology, TNM stage, prior surgery, liver/peritoneum involvement, ascites, date of start of chemotherapy, date of last status, survival status, cause of death, date of progression, progression status, tumour response, date of tumour response and reason for treatment discontinuation. Updated survival status and date of last follow-up were also requested. Data for patients excluded from analysis after randomisation were obtained whenever possible.

2.2. Research questions

The first overall question we investigated was whether the experimental arm or arms of trials included in the meta-analysis showed a benefit compared with the corresponding control arm. To achieve clinical interpretability despite the heterogeneity in treatment regimens tested in the various trials, we defined as ‘experimental’ the treatment arm which contained the greater number of drugs (e.g. triple combinations versus double combinations, regardless of the drugs involved in the respective regimens). When two arms consisted of regimens containing the same number of drugs, we defined as ‘experimental’ the treatment arm that included the newer agent, i.e. methotrexate versus MMC, CPT11 versus etoposide, CPT11 versus 5-FU and S-1 versus 5-FU. These conventions allowed us to define a control arm and one or more experimental arms for all trials included in the meta-analysis, except randomised phase II trials that did not aim to show superiority of the experimental over the control arm. We grouped the different trials according to chemotherapeutic agents. In sub-analysis, we distinguished ‘additive comparisons’, in which chemotherapy was compared to the same chemotherapy plus an additional agent; ‘substitutive comparisons’, in which one of the agents in a regimen was substituted by a different agent; and a last category of ‘other comparisons’.

Second, we refined the analysis of trials that compared chemotherapy with the same plus an additional compound to investigate classes of agents. This analysis included randomised comparisons of any treatment, including for example best supportive care only versus the same treatment plus a chemotherapeutic agent or combination of agents. Last, we grouped trials that compared any chemotherapy to any other with and without one or more agents of interest, namely 5-FU, MMC, anthracycline, platinum agents, irinotecan or taxane.

2.3. Endpoint

The primary endpoint for this analysis was OS, defined from the date of randomisation to the date of death from any cause. Patients lost to follow-up were censored at the date of last assessment. PFS, defined from the date of randomisation to the date of progression or death from any cause, was also determined. Progression was as assessed by an independent review committee when available, or as reported by the clinician otherwise. Patients lost to follow-up were censored at the patient’s last visit date or the last date that the absence of progression was documented. In cases without information on censoring for PFS, the date of last assessment of survival was used as the date of censor. Trials which did not collect information on date of progression were excluded from PFS analyses.

2.4. Data management and quality assessment of the trials

All data were centrally checked with standardised programs. Each trial was reanalysed and compared to the published results. Differences were queried to the corresponding author. In addition to the identification of inconsistencies and missing data, particular attention was paid to the randomisation process and patient follow-up. Randomisation was examined to identify possible differences with a true random process. Balance in the distribution of the main covariates was also investigated. Likewise, distributions of days/months of randomisation were checked to detect anomalies. 18

2.5. Statistical analysis

Treatment effects on OS and PFS were analysed using a two-sided log rank stratified by study. We used a fixed-effects model and the inverse variance method in which the weight of each trial was proportional to the variance of the number of events. Multi-arm trials suitable for different research questions were analysed in several comparisons. To account for correlation between the different comparisons, the weight of each pair-wise comparison was corrected according to a function of the number of events. For instance, in a three-group trial with d1 and d2 events in the two investigational arms and dc in the control arm, the weight of the two comparisons was modified by a factor equal to (di + dc/2)/(di + dc), with i being equal to 1 or 2. Forest plots of hazard ratios (HR) were produced with one line per trial. Heterogeneity between trials and groups of trials (e.g. defined by different chemotherapy regimens) was tested using chi-squared statistics 19 and measured with the I2 statistics, which estimates the proportion of variability due to heterogeneity between studies rather than sampling error. 20 All analyses were carried out on all randomised patients according to their allocated treatment arm, irrespective of the received treatment. Patients whose individual survival data were not available were excluded. All analyses were performed with SAS 9.2 using macros from the European Organisation for Research and Treatment of Cancer.

3. Results

Fifty-five eligible studies that had randomised 9054 patients were identified by the end of 2010 ( Fig. 1 ). Data were available from 22 studies, representing 4245 patients (47% of the targeted patients).4, 5, 6, 8, 12, 13, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, and 36 Demographic and baseline characteristics of all trial data are displayed in Table 1 . OS and PFS data were available for 4214 and 4073 patients (20 trials) respectively. No study was excluded due to questionable randomisation. Data were not obtained for 4809 patients in 33 trials because of no reply despite repeated attempts or refusal to share by the principal investigators,7, 9, 10, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, and 53 or because data were lost or inaccessible.3, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, and 65 The trials for which data were available were more recent and had better performance status than those with unavailable data ( eTable 3 ). No other differences in baseline characteristics were found.

gr1

Fig. 1 Study flow chart.

Table 1 List of included trials.

Trial Chemotherapy schedule No. of patients Accrual period Disease status at entry Median (range) follow-up days
Experimental arm Control arm Exp Ctr
Douglass (1984) 5 FAMe: 5-FU 325 mg/m2 i.v., days 1 through 5; 5-FU 450 mg/m2 i.v., day 22; and methyl-CCNU 60 mg/m2 orally, days 1 and 22; and doxorubicin 40 mg/m2 i.v., days 1 and 22. Each course was repeated every 6 weeks

FAM: 5-FU 600 mg/m2 i.v., days 1, 8, 29, 36; doxorubicin 30 mg/m2 i.v., days 1 and 29; and mitomycin C (MMC) 10 mg/m2 i.v., day 1. Courses were repeated every 8 weeks
FMe: 5-FU 300 mg/m2 i.v., days 1 through 5; and methyl-CCNU 175 mg/m2 orally, day 1. Each course was repeated every 7 weeks 89 48 1977–1981 Not supplied 156 (5–1107)
Cullinan (1985) 21 FAM: 5-FU 600 mg/m2 i.v., days 1, 8, 29 and 36; doxorubicin 30 mg/m2, days 1 and 29; and MMC 10 mg/m2, day 1

FA: 5-FU 400 mg/m2 i.v.; and 40 mg of doxorubicin on day 1 every 4 weeks
F: 5-FU 500 mg/m2 i.v., 5 days first. Courses were repeated at weeks 4 and 8, and every 5 weeks thereafter. [check] 103 53 1977–1980 Regional: 37.7%

Metastatic: 62.3%
207 (1–2645)
Wils (1991) 6 FAMTX: 5-FU 1500 mg/m2 i.v. day 1; methotrexate 1500 mg/m2 i.v., day 1; leucovorin 15 mg/m2 orally, every 6 h for 48 h; and doxorubicin 30 mg/m2 i.v., day 15. Courses were repeated every 4 weeks FAM: 5-FU 600 mg/m2 i.v. days 1, 8, 29 and 36; doxorubicin 30 mg/m2 i.v. days 1 and 29; and MMC 10 mg/m2 i.v. day 1. Courses were repeated every 8 weeks 108 104 1985–1989 Locally adv: 10.9%

Metastatic: 68.4%

Locally rec: 4.3%
223.5 (1–1406)
Nio (1992) 22 CFEpi: Cisplatin 50 mg/body i.v., day 1; 5-FU 250 mg/body i.v., day 2 through 5; and epirubicin 30 mg/m2 i.v., day 2. Courses were repeated every 2 weeks CF: Cisplatin 50 mg/body i.v., day 1; and 5-FU 250 mg/body i.v., day 2 through 5. Courses were repeated every 2 weeks 31 27 1989–1991 Locally adv: 15.5%

Metastatic: 37.9%

Locally rec: 43.1%
82.5 (10–358)
Coombes (1994) 23 Epi: Epirubicin 100 mg/m2 i.v., day 1. Course was repeated every 3 weeks F: 5-FU 500 mg/m2 i.v., day 1–5. Course was repeated every 3 weeks 36 33 1985–1988 Loco-regional: 5.8%

Peritoneal: 18.8%

Metastatic: 69.6%
193.5 (3–1250)
Cullinan (1994) 24 FAMe: 5-FU 325 mg/m2 i.v., days 1 through 5; 5-FU 450 mg/m2 i.v., day 22; and doxorubicin 40 mg/m2 i.v., days 1 and 22. Courses were repeated every 5 weeks. Methyl-CCNU 110 mg/m2 orally, days 1, repeated every 10 weeks

FAMe plus TZT: 5-FU 325 mg/m2 i.v., days 1 through 5; 5-FU 450 mg/m2 i.v., day 22; doxorubicin 40 mg/m2 i.v., days 1 and 22; and TZT 250 mg/m2 i.v., daily for 3 days at 5 weeks. Methyl-CCNU 110 mg/m2 orally, days 1, repeated every 10 weeks. This full cycle was repeated every 8 weeks

FAC: 5-FU 300 mg/m2 i.v., daily for 5 consecutive days; doxorubicin 40 mg/m2 i.v., on day 1; and cisplatin 60 mg/m2 i.v., day 1. Courses were repeated every 5 weeks
F: 5-FU 500 mg/m2 i.v., daily for 5 consecutive days. Course was repeated every 5 weeks 184 72 1984–1992 Loco-regional: 43%

Metastatic: 57%
198 (2–2048)
Glimelius (1997) 4 ELF: Leucovorin 350 mg/m2, etoposide 120 mg/m2 and 5-FU 500 mg/m2 daily on 3 consecutive days. Courses were repeated every 3 weeks

Or

FL: 5-FU 500 mg/m2 i.v. and leucovorin 60 mg/m2 daily on 2 consecutive days. Courses were repeated every 2 weeks
Best supportive care 30 30 1991–1995 Loco-regional: 70%

Metastatic: 30%
206 (6–941)
Yamamura (1998) 25 FMTXPira: methotrexate 50 mg/m2 i.v.; 5-FU 650 mg/m2 i.v.; and pirarubicin 20 mg/m2 i.v. on day 1 and 15. Leucovorin 15 mg/body after 30 h. Courses were repeated every 2 weeks. F: 5-FU 650 mg/m2 day 1 and 15. Course was repeated every 2 weeks 39 35 1994–1996 Metastatic: 71.6%

Locally rec: 28.4%
156.5 (8–882)
Vanhoefer (2000) 26 FAMTX: 5-FU 1500 mg/m2 i.v. day 1; methotrexate 1500 mg/m2 i.v. followed after 1 h by 5-FU; leucovorin rescue 30 mg/m2 was started after 24 h orally every 6 h for 48 h; and doxorubicin 30 mg/m2 i.v., day 15. Courses were repeated every 4 weeks

ELF: folinic acid (leucovorin) 300 mg/m2 i.v.; etoposide 120 mg/m2; and bolus 5-FU 500 mg/m2 i.v., for 3 consecutive days. The cycles were repeated every 22 days
CF: 5-FU 1 g/m2 i.v. daily for 5 consecutive days; and cisplatin 100 mg/m2 i.v., on day 2. The cycles were repeated every 29 days 265 134 1991–1995 Locally adv: 5.0%

Metastatic: 16.3%

Locally rec: 78.7%
183 (2–2329)
Kim (2001) 27 EpiCF: Epirubicin 50 mg/m2 i.v. day 1; cisplatin 60 mg/m2 i.v. day 1; and 5-FU 1000 mg/m2 i.v., days 1 through 5. The cycles were repeated every 4 weeks CF: 5-FU 1000 mg/m2 i.v., day 1 through 5; and cisplatin 60 mg/m2 i.v., day 1. The cycles were repeated every 4 weeks 59 61 1997–2000 Locally adv: 5.0%

Metastatic: 95.0%
235 (34–235)
Ohtsu (2003) 8 CF: 5-FU 800 mg/m2 i.v. day 1 through 5; and cisplatin 20 mg/m2 i.v., day 1–5. The cycles were repeated every 4 weeks

UFTM: UFT (uracil/tegafur) 375 mg/m2 orally, daily; and MMC 5 mg/m2 i.v., day 1. The cycles were repeated every 4 weeks
F: 5-FU 800 mg/m2 i.v., days 1 through 5. The cycles were repeated every 4 weeks 175 105 1992–1997 Locally adv: 79.3%

Metastatic: 20.4%
204 (4–1565)
Pozzo (2004) 28 IFL: Irinotecan 80 mg/m2 i.v.; and leucovorin 500 mg/m2 i.v. and 5-FU 2000 mg/m2 i.v., on days 1, 8, 15, 22, 29 and 36. Cycles were repeated every 7 weeks IC: irinotecan 200 mg/m2 i.v., day 1; and cisplatin 60 mg/m2 i.v. day 1. Cycles were repeated every 3 weeks 75 73 1999–2000 Locally adv: 6.1%

Metastatic: 90.5%

Locally rec: 0.7%
257 (5–664)
Koizumi (2004) 29 CMF’: Cisplatin 70 mg/m2 i.v., on day 1; MMC 7 mg/m2 i.v., day 2; and 5’-DFUR 1200 mg/m2 orally, day 4–7, 11–14, 18–21 and 25–28. Cycles were repeated as long as possible CF’: Cisplatin 70 mg/m2 i.v., on day 1; and 5’-DFUR 1200 mg/m2 orally, days 4–7, 11–14, 18–21 and 25–28. Cycles were repeated as long as possible 33 29 1991–1996 Locally adv: 21.0%

Metastatic: 79.0%
190 (16–1848)
Bouche (2004) 30 FLC: Leucovorin 200 mg/m2 i.v.; and bolus 5-FU 400 mg/m2 i.v.; 5-FU 600 mg/m2 i.v.; cisplatin 50 mg/m2 i.v., on days 1 and 2. Cycles (15 days) were repeated every 14 days

IFL: Irinotecan 180 mg/m2 i.v.; 200 mg/m2 i.v.; and bolus 5-FU 400 mg/m2 i.v.; and 5-FU 600 mg/m2 i.v. on days 1 and 2. Cycles (15 days) were repeated every 14 days
FL: Leucovorin 200 mg/m2 i.v.; and bolus 5-FU 400 mg/m2 i.v.; and 5-FU 600 mg/m2 i.v. on days 1 and 2. Cycles (15 days) were repeated every 14 days 91 45 1999–2001 Metastatic: 98.5% 291 (17–1023)
Hawkins (2005) 31 DI: Docetaxel 60 mg/m2 i.v.; and irinotecan 250 mg/m2 day 1. Cycles were repeated every 3 weeks DF: Docetaxel 85 mg/m2 i.v., day 1; and 5-FU 750 mg/m2 i.v., day 1 through 5. Cycles were repeated every 3 weeks 43 43 1999–2000 Locally adv: 5.8%

Metastatic: 94.2%
277.5 (7–931)
Moehler (2005) 12 IFL: Irinotecan 80 mg/m2 i.v.; leucovorin 500 mg/m2 i.v.; and 5-FU 2000 mg/m2 i.v., on days 1, 8, 15, 22, 29 and 36. Cycles were repeated every 7 weeks ELF: folinic acid (leucovorin) 300 mg/m2 i.v.; etoposide 120 mg/m2; and bolus 5-FU 500 mg/m2 i.v., for 3 consecutive days. The cycles were repeated every 22 days 58 62 2000–2003 Not supplied 271.5 (12–986)
Thuss-Patience (2005) 32 EpiCF: Epirubicin 50 mg/m2 i.v. day 1; cisplatin 60 mg/m2 i.v. day 1; and 5-FU 200 mg/m2 i.v., days 1 through 21. The cycles were repeated every 3 weeks DF: Docetaxel 75 mg/m2 i.v., day 1; and 5-FU 200 mg/m2 i.v., days 1 through 21. Cycles were repeated every 3 weeks 45 47 1999–2002 Locally adv: 1.1%

Metastatic: 93.6%

Locally rec: 1.1%
286 (19–1138)
Ajani (2005) 33 DCF: Docetaxel 60 mg/m2 i.v. day 1; cisplatin 75 mg/m2 i.v. day 1; and 5-FU 750 mg/m2 i.v., days 1–5. The cycles were repeated every 3 weeks DC: Docetaxel 85 mg/m2 i.v. day 1; and cisplatin 75 mg/m2 i.v. day 1. The cycles were repeated every 3 weeks 79 79 1998–1999 Locally adv: 3.2%

Metastatic: 94.9%

Locally rec: 1.9%
162 (6–690)
Van Cutsem (2006) 11 DCF: Docetaxel 60 mg/m2 i.v. day 1; cisplatin 75 mg/m2 i.v. day 1; and 5-FU 750 mg/m2 i.v., days 1 through 5. The cycles were repeated every 3 weeks CF: Cisplatin 100 mg/m2 i.v. day 1; and 5-FU 1000 mg/m2 i.v., days 1 through 5. The cycles were repeated every 4 weeks 227 230 1999–2003 Locally adv: 2.2%

Metastatic: 96.3%

Locally rec: 0.4%
231 (1–1202)
Roth (2007) 34 EpiCF: Epirubicin 50 mg/m2 i.v. day 1; cisplatin 60 mg/m2 i.v. day 1; and 5-FU 200 mg/m2 i.v., days 1 through 21

DCF: Docetaxel 85 mg/m2 i.v. day 1; cisplatin 75 mg/m2 i.v. day 1; and 5-FU 300 mg/m2 i.v., days 1–14
DC: Docetaxel 85 mg/m2 i.v. day 1; and cisplatin 75 mg/m2 i.v. day 1 82 39 1999–2003 Loco-regional: 33.6%

Metastatic: 66.4%
279.5 (7–2001)
Dank (2008) 35 IFL: Irinotecan 80 mg/m2 i.v.; and leucovorin 500 mg/m2 i.v. and 5-FU 2000 mg/m2 i.v., on days 1, 8, 15, 22, 29 and 36. Cycles were repeated every 7 weeks CF: Cisplatin 100 mg/m2 i.v. day 1; and 5-FU 1000 mg/m2 i.v., days 1 through 5. The cycles were repeated every 4 weeks 172 165 2000–2002 Locally adv: 1.8%

Metastatic: 94.7%

Locally rec: 0.6%
266 (9–1093)
Boku (2009) 36 IC: Irinotecan 70 mg/m2 i.v., on days 1 and 15; and cisplatin 80 mg/m2 i.v., day 1. Cycles were repeated every 4 weeks After six cycles, the same dose of irinotecan alone was continued every 2 weeks

S: S-1 40 mg/m2 orally, twice a day for 4 weeks, followed by a 2-week rest
F: 5-FU 800 mg/m2 i.v., daily for 5 days. Cycles were repeated every 4 weeks 470 234 2000–2006 Locally adv: 16.3%

Metastatic: 64.2%

Locally rec: 19.5%
354.5 (11–2621)
Total     2494 1748      
4245

Three of the patients were missing data for the allocated arm.

3.1. Experimental versus control grouped by type of comparison

When all trials included in the meta-analysis were considered for the overall comparison of experimental chemotherapy arms with the corresponding control arms, statistically significant differences were seen with regard to both OS and PFS ( Fig. 2 and eFig. 1 ). HR was 0.88 (95% confidence interval (95%CI) 0.82–0.94, P < 0.0001) for OS and 0.81 (95%CI 0.76–0.86, P < 0.0001) for PFS. These HRs translated into a median PFS difference of 4 weeks and a median OS difference of 3 weeks ( Fig. 3 ).

gr2

Fig. 2 Individual trial and overall hazard ratio for overall survival (OS) when comparing experimental arms versus control arms. I2 = 11%. The inverse of observed minus expected events measures the weight of each trial in the analysis. P values are from P-for-effect modification testing for heterogeneity within or across the groups of regimens. The sizes of data markers are proportional to the number of deaths in the trials. CI, confidence interval; HR, hazard ratio.

gr3

Fig. 3 Overall survival (OS) and progression-free survival (PFS) curves of investigational and control arms. Exp-OS, Experimental arm for OS; Ctrl-OS, Control arm for OS; Exp-PFS, Experimental arm for PFS; Ctrl-PFS, Control arm for PFS.

No heterogeneity in treatment effect on OS was detected between the three types of comparisons. It was of borderline significance for PFS (P = 0.04), with a larger benefit in trials comparing a regimen versus the same regimen plus a new compound. Further, statistically significant heterogeneity was seen among HRs of the various trials for PFS (P < 0.01, I2 = 44%) and to a lesser extent for OS (P = 0.31, I2 = 11%), which was mainly driven by trials from the ‘other’ type of comparison group. The trials in Fig. 2 and eFig. 1 are ordered according to the date of accrual termination, revealing the absence of any clear trend towards greater benefit in more recent years.

3.2. Any chemotherapy versus the same chemotherapy plus a given chemotherapeutic agent

This analysis also showed a statistically significant difference favouring the addition of a chemotherapeutic agent in terms of PFS in 1786 patients ( eFig. 2 ) and of OS in 1859 patients ( eFig. 3 ). HR was 0.74 (95%CI 0.67–0.81, P < 0.0001) for PFS and 0.89 (95%CI 0.80–0.98, P < 0.01) for OS. There was no evidence of heterogeneity among HRs of either PFS (P = 0.18) or OS (P = 0.57).

3.3. Contribution of individual agents

The last set of analyses looked at the randomised comparisons of any treatment regimen not including an agent of interest with the same or a different treatment regimen including the agent. This partially overlaps with the previous analysis. Forrest plots are displayed in eFig. 4 for PFS and eFig. 5 for OS.

For MMC, analysis of 839 patients from 5 trials5, 6, 8, 21, and 29 (31% of 2705 targeted patients) showed no difference in terms of PFS (HR = 1.06; 95%CI 0.91–1.24, P = 0.45) or OS (HR = 1.15; 95%CI 0.98–1.34, P = 0.084). There was, however, substantial heterogeneity between the outcomes of the various trials with regard to both PFS (P = 0.002) and OS (P = 0.003).

For anthracyclines, analysis of 1320 patients (8 trials5, 21, 22, 24, 26, 27, and 32) showed no difference with regard to PFS (HR = 0.92; 95%CI 0.82–1.04) or OS (HR = 0.97; 95%CI 0.87–1.09, 1461 patients from 10 trials5, 21, 22, 23, 24, 26, 27, 32, and 34; 52% of 2561 targeted patients). No statistically significant heterogeneity was detected.

For platinum agents, analysis of 2337 patients from eight trials8, 24, 26, 28, 30, 32, 35, and 36 (53% of 4369 targeted patients) showed a statistically significant difference in favour of the platinum-based regimen with regard to PFS (HR = 0.88; 95%CI 0.81–0.96) but not OS (HR = 0.96; 95%CI 0.88–1.05). There was, however, substantial heterogeneity between trials in terms of both PFS (P < 0.0001) and OS (P = 0.021). As a sensitivity analysis, we recomputed statistics without one study which had extreme results, 28 which showed a borderline benefit on OS (HR = 0.92, P = 0.08) while across-trial heterogeneity remained statistically significant for PFS (P = 0.01).

Analysis of taxanes from three trials32, 33, and 34 (n = 667, 69% of 971 targeted patients) showed no statistically significant difference with regard to either PFS (HR = 0.88; 95%CI 0.75–1.04) or OS (HR = 0.90; 95%CI 0.76–1.07). Some indication of heterogeneity was seen for PFS (P = 0.05) but not for OS (P = 0.25).

Finally, analysis of 1380 patients (5 trials;12, 30, 31, 35, and 36 81% of 1698 targeted patients) showed a statistically significant benefit of irinotecan-based regimes over their comparator in terms of PFS (HR = 0.81; 95%CI 0.72–0.90) and a marginally non-significant benefit in terms of OS (HR = 0.90; 95%CI 0.80–1.01). No statistically significant heterogeneity was detected.

4. Discussion

The results of this IPD meta-analysis show that OS and PFS have been improved to a certain extent by the addition of experimental chemotherapeutic agents to pre-existing control or standard regimens, or by the introduction of more active agents such as platinum, taxanes or irinotecan. The benefits of the experimental treatments over their corresponding controls are hazard reductions of 13% for OS (HR = 0.87, P < 0.0001, Fig. 2 ) and 21% for PFS (HR = 0.79, P < 0.0001, eFig. 1 ). Although these overall relative risk reductions translate into only modest absolute improvements in median OS and PFS, they nevertheless confirm the benefit of chemotherapy for AGC, and justify further efforts at improving current regimens.

This overall analysis used all data collected, and is accordingly the most reliable from a statistical point of view. As it combines different comparisons, however, it lacks the specificity to reveal the benefits of individual agents. The results of the sub-analysis of adding a given chemotherapy to any chemotherapy confirm the results of the overall analysis, with a hazard reduction of 11% for OS (P < 0.01) and 26% for PFS (P < 0.0001). Although this analysis isolates the contribution of an additional agent, it does not address the issue of which agent is the most effective. Looking at individual agents suggests potential benefits of the most recent generations, namely cisplatin and irinotecan, but not of MMC or docetaxel. Results on the addition of anthracyclines deserve comment as they are inconsistent with the literature-based meta-analysis of Wagner, who found an OS benefit compared to a 5-FU–cisplatin association, based on three studies. 17 Our results were not restricted to a 5-FU + cisplatin comparator but also included two American trials using 5-FU + MeCCNU 5 and 5-FU 21 as controls. Further, our meta-analysis did not include one trial 46 comparing ECF to MCF (n = 334) with positive results of borderline significance. However, pooling the summary statistics of this latter trial with the collected IPD would not have modified the overall conclusion regarding the benefit of anthracyclines on OS.

Several limitations of this meta-analysis warrant mention. First, gastric cancers are heterogeneous, and it was therefore impossible to analyse the data according to important known prognostic factors, such as performance status and tumour type. Although a recent study suggested the importance of HER2 as the first biomarker for AGC, 66 a more useful classification for gastric cancer in terms of prognosis and predictive ability might be important. Similarly, our meta-analysis consists of a mixture of patients with proximal and distal cancer, and of different types of extension from local but unresectable tumours to visceral metastases involving liver, peritoneum, distant lymph nodes and bones. 67 This does not invalidate the value of this analysis but does underline the importance of more precise analyses of all subcategories of gastric tumours, which may have different natural histories and responses to treatment. Second, as far as PFS is concerned, some studies (mainly commercially-sponsored) performed CT scans at regular intervals until disease progression, whereas in other studies imaging was performed at the time of symptomatic progression rather than at pre-determined time-points.

Finally, we were only able to include data for 47% of the total number of patients from eligible randomised trials of AGC to date. This remains sub-optimal, given that IPD meta-analyses should be comprehensive and include all the trials ever conducted, whether positive or negative, or published or unpublished. While at present access to individual patient data is still very difficult, a consensus is emerging that such data should be made available to address subsequent research questions. In the United States, the National Institutes of Health Data Sharing Policy states that clinical trial ‘data should be made as widely and freely available as possible while safeguarding the privacy of participants, and protecting confidential and proprietary data’. 68 Some journals have adopted a policy that trials of drugs and medical devices be considered for publication only if the authors commit to making the relevant anonymised patient level data available. 69 The pharmaceutical industry is also taking steps to making data from the trials they sponsor available for reasonable requests. 70 The GASTRIC collaboration favours any initiative to pool IPD since this provides the only reliable means to characterising the role of treatment specific regimens. Finally, as far as PFS is concerned, some studies (mainly commercially-sponsored) performed CT scans at regular intervals until disease progression, whereas in other studies imaging was performed at the time of symptomatic progression rather than at pre-determined time-points.

In conclusion, our IPD meta-analysis shows that OS and PFS have been modestly improved by the addition of experimental chemotherapeutic agents to pre-existing control or standard regimens. Nevertheless, median survival remains at less than 1 year for all investigated cytotoxic associations of chemotherapy, and none has emerged as a clear standard. We hope that molecularly targeted agents administered to subgroups of patients selected on the basis of biological characteristics may modify this disappointing situation.

Conflict of interest statement

This project was initiated under the auspices of the INCa, which served as a sponsor. The INCa did not participate in the design of the study. It participated in the conduct of the study at an earlier stage by centralising all the databases and by providing administrative and data management support. The sponsor had no role in the preparation, review or approval of the manuscript.

All members of the writing committee declare they have no conflict of interest with regard to this meta-analysis.

The corresponding author (Koji Oba) had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs. Oba and Paoletti contributed equally to this work.

Additional Contributions: The GASTRIC Group thanks all patients who took part in the trials and contributed to this research. The meta-analysis would not have been possible without their participation, or without the active participation of the collaborating institutions which provided their trial data (ECOG, EORTC, FFCD, GITSG, ICCG, ITMO, JCOG, NCCTG, SWOG, Hospital Clinic Villarroel of Barcelona, Metaxa Cancer Hospital of Pireus University, Jagiellonian Medical College of Cracow). We thank Caroline Tournoux-Facon for her help in project management in the early stage and Nicolas Thammavong for data management.

The GASTRIC Investigators:

Secretariat: Marc Buyse, Stefan Michiels, Koji Oba, Xavier Paoletti, Philippe Rougier, Seiichiro Yamamoto, Kenichi Nakamura.

Steering Committee: Yung-Jue Bang (Seoul National University College of Medicine, Seoul, Korea); Harry Bleiberg (Brussels, Belgium); Tomasz Burzykowski (Hasselt University, Diepenbeek, Belgium); Marc Buyse (International Drug Development Institute, Louvain-la-Neuve, Belgium); Catherine Delbaldo (Hôpital Louis Mourier, Colombes, France); Stefan Michiels (Jules Bordet institute, Brussels, Belgium); Satoshi Morita (Yokohama City University, Kanagawa, Japan); Koji Oba (Hokkaido University, Hokkaido, Japan); Yasuo Ohashi (University of Tokyo, Tokyo, Japan); Xavier Paoletti (Institut Curie, Paris, France); Jean-Pierre Pignon (Institut Gustave-Roussy, Villejuif, France); Philippe Rougier (University Hospital Europeen Georges Pompidou [AP-HP], Boulogne, France); Junichi Sakamoto (Nagoya University, Aichi, Japan); Daniel Sargent (Mayo Clinic, Rochester, Minnesota, USA); Mitsuru Sasako (Hyogo College of Medicine, Hyogo, Japan); Eric Van Cutsem (Digestive Oncology Unit, University Hospital Gasthuisberf, Leuven, Belgium).

Collaborators: J Ajani, N Boku, C Barone, M Borner, O Bouche, R Bugat, C Coombes, S Cullinan, M Dank, H Douglass, N Fuse, B Glimelius, R Hawkins, S Kanzler, W Koizumi, TW Kim, F Lordick, M Moehler, Y Nio, A Ohtsu, C Pozzo, P Reichardt, A Roth, C Schumacher, K Shitara, P Thuss-Patience, A Tsuburaya, E Van Cutsem, U Vanhoefer, J Wils, Y Yamamura.

Acknowledgements

A first draft of this manuscript was developed at a meeting of investigators in Sapporo, Japan, September 24–28, 2010. The meeting was financially supported by unrestricted grants from GlaxoSmithKline. The GASTRIC project is partly supported by the Clinical Research Support Unit (J-CRSU), the Epidemiological and Clinical Research Information Network (ECRIN), the Institut National du Cancer (INCa, France) and the French Ministry of Health. The funders were not present at the meeting, were not involved in the analyses of the data, and did not comment on the present paper.

Appendix A. Writing committee

Koji Oba PhD (Translational Research and Clinical Trial Center, Hokkaido University Hospital, Hokkaido, Japan), Xavier Paoletti PhD (Institut Curie/ Inserm U900, Paris, France), Yung-Jue Bang MD PhD (Seoul National University College of Medicine, Seoul, Korea), Harry Bleiberg MD PhD (Jules Bordet, Brussels, Belgium), Tomasz Burzykowski PhD (Hasselt University, Diepenbeek, Belgium), Nozomu Fuse MD PhD (National Cancer Center Hospital East, Chiba, Japan), Stefan Michiels PhD (Jules Bordet institute, Brussels, Belgium), Satoshi Morita PhD (Yokohama City University, Kanagawa, Japan), Yasuo Ohashi PhD (University of Tokyo, Tokyo, Japan), Jean-Pierre Pignon MD, PhD (Institut Gustave-Roussy, Villejuif, France), Philippe Rougier MD, PhD (University Hospital Europeen Georges Pompidou (AP-HP), Paris, France), Junichi Sakamoto MD, PhD (Nagoya University, Aichi, Japan), Daniel Sargent PhD (Mayo Clinic, Minnesota, USA), Mitsuru Sasako MD, PhD (Hyogo College of Medicine, Hyogo, Japan), Kohei Shitara MD PhD (Aichi Cancer Center Hospital, Aichi, Japan), Akira Tsuburaya MD PhD (Kanagawa Cancer Center Hospital, Kanagawa, Japan), Eric Van Cutsem MD, PhD (Digestive Oncology Unit, University Hospital Gasthuisberf, Leuven, Belgium), Marc Buyse PhD (International Drug Development Institute, Louvain-la-Neuve, Belgium).

Appendix B. Supplementary data

 

Download file

Supplementary data 1

Download file

Supplementary data 2

References

  • 1 J. Ferlay, H.R. Shin, F. Bray, et al. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893-2917 Crossref
  • 2 American Cancer Society. Global cancer facts & figures. 2nd ed. < http://www.cancer.org/Research/CancerFactsFigures/GlobalCancerFactsFigures/index >; 2011 [accessed 03.01.12].
  • 3 S. Pyrhönen, T. Kuitunen, P. Nyandoto, M. Kouri. Randomised comparison of fluorouracil, epidoxorubicin and methotrexate (FEMTX) plus supportive care with supportive care alone in patients with metastatic gastric cancer. Br J Cancer. 1995;71:587-591
  • 4 B. Glimelius, K. Ekstrom, K. Hoffman, et al. Randomized comparison between chemotherapy plus best supportive care with best supportive care in advanced gastric cancer. Ann Oncol. 1997;8:163-168 Crossref
  • 5 H.O. Douglass Jr., P.T. Lavin, A. Goudsmit, D.J. Klaassen, A.R. Paul. An Eastern Cooperative Oncology Group evaluation of combinations of methyl-CCNU, mitomycin C, Adriamycin, and 5-fluorouracil in advanced measurable gastric cancer (EST 2277). J Clin Oncol. 1984;2:1372-1381
  • 6 J.A. Wils, H.O. Klein, D.J. Wagener, et al. Sequential high-dose methotrexate and fluorouracil combined with doxorubicin–a step ahead in the treatment of advanced gastric cancer: a trial of the European Organization for Research and Treatment of Cancer Gastrointestinal Tract Cooperative Group. J Clin Oncol. 1991;9:827-831
  • 7 C. Barone, D.C. Corsi, C. Pozzo, et al. Treatment of patients with advanced gastric carcinoma with a 5-fluorouracil-based or a cisplatin-based regimen: two parallel randomized phase II studies. Cancer. 1998;82:1460-1467 Crossref
  • 8 A. Ohtsu, Y. Shimada, K. Shirao, et al. 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 Crossref
  • 9 J.S. Waters, A. Norman, D. Cunningham, et al. Long-term survival after epirubicin, cisplatin and fluorouracil for gastric cancer: results of a randomized trial. Br J Cancer. 1999;80:269-272 Crossref
  • 10 A. Roth, K. Kolaric, D. Zupanc, et al. High doses of 5-fluorouracil and epirubicin with or without cisplatin in advanced gastric cancer: a randomized study. Tumori. 1999;85:234-238
  • 11 E. Van Cutsem, V.M. Moiseyenko, S. Tjulandin. Et al.; V325 Study Group. Phase III study of docetaxel and cisplatin plus fluorouracil compared with cisplatin and fluorouracil as first-line therapy for advanced gastric cancer: a report of the V325 Study Group. J Clin Oncol. 2006;24:4991-4997 Crossref
  • 12 M. Moehler, A. Eimermacher, J. Siebler, et al. Randomised phase II evaluation of irinotecan plus high-dose 5-fluorouracil and leucovorin (ILF) vs 5-fluorouracil, leucovorin, and etoposide (ELF) in untreated metastatic gastric cancer. Br J Cancer. 2005;92:2122-2128 Crossref
  • 13 Y.K. Kang, W.K. Kang, D.B. Shin, et al. Capecitabine/cisplatin versus 5-fluorouracil/cisplatin as first-line therapy in patients with advanced gastric cancer: a randomised phase III noninferiority trial. Ann Oncol. 2009;20:666-673 Crossref
  • 14 D. Cunningham, N. Starling, S. Rao, et al. Capecitabine and oxaliplatin for advanced esophagogastric cancer. N Engl J Med. 2008;358:36-46 Crossref
  • 15 GASTRIC (Global Advanced/Adjuvant Stomach Tumor Research International Collaboration) Group. Benefit of adjuvant chemotherapy for resectable gastric cancer: a meta-analysis. JAMA. 2010;303:1729-1737
  • 16 A.D. Wagner, S. Unverzagt, W. Grothe, et al. Chemotherapy in advanced gastric cancer: a systematic review and meta-analysis based on aggregate data. J Clin Oncol. 2006;24:2903-2909 Crossref
  • 17 A.D. Wagner, W. Grothe, J. Haerting, et al. Chemotherapy for advanced gastric cancer. Cochrane Database Syst Rev. 2010;3:CD004064
  • 18 L.A. Stewart. On behalf of the Cochrane Working Group on meta-analyses using individual patient data. Practical methodology of meta-analyses (overviews) using updated individual patients data. Stat Med. 1995;14:2057-2079 Crossref
  • 19 W.G. Cochran. The combination of estimates from different experiments. Biometrics. 1954;10:101-129 Crossref
  • 20 J.P.T. Higgins, S.G. Thompson, J.J. Deeks, D.G. Altman. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557-560 Crossref
  • 21 S.A. Cullinan, C.G. Moertel, T.R. Fleming, et al. A comparison of three chemotherapeutic regimens in the treatment of advanced pancreatic and gastric carcinoma. Fluorouracil vs fluorouracil and doxorubicin vs fluorouracil, doxorubicin, and mitomycin. JAMA. 1985;253:2061-2067 Crossref
  • 22 Kyoto Research Group for Chemotherapy of Gastric Cancer (KRGCGC). A randomized, comparative study of combination chemotherapies in advanced gastric cancer: 5-fluorouracil and cisplatin (FP) versus 5-fluorouracil, cisplatin, and 4’-epirubicin (FPEPIR). Anticancer Res. 1992;12:1983-1988
  • 23 R.C. Coombes, C.E. Chilvers, D. Amadori, et al. Randomised trial of epirubicin versus fluorouracil in advanced gastric cancer. An International Collaborative Cancer Group (ICCG) study. Ann Oncol. 1994;5:33-36
  • 24 S.A. Cullinan, C.G. Moertel, H.S. Wieand, et al. Controlled evaluation of three drug combination regimens versus fluorouracil alone for the therapy of advanced gastric cancer. North Central Cancer Treatment Group. J Clin Oncol. 1994;12:412-416
  • 25 Y. Yamamura, I. Miyazaki, M. Ogawa, et al. A randomized controlled trial with methotrexate (MTX), 5-fluorouracil (5-FU) and pirarubicin (THP) vs 5-FU alone in advanced or recurrent gastric carcinoma. Tokai Hokuriku THP Study Group. Gan To Kagaku Ryoho. 1998;25:1543-1548
  • 26 U. Vanhoefer, P. Rougier, H. Wilke, et al. Final results of a randomized phase III trial of sequential high-dose methotrexate, fluorouracil, and doxorubicin versus etoposide, leucovorin, and fluorouracil versus infusional fluorouracil and cisplatin in advanced gastric cancer: A trial of the European Organization for Research and Treatment of Cancer Gastrointestinal Tract Cancer Cooperative Group. J Clin Oncol. 2000;18:2648-2657
  • 27 T.W. Kim, S.J. Choi, J.H. Ahn, et al. A prospective randomized phase III trial of 5-fluorouracil and cisplatin (FP) versus epirubicin, cisplatin, and 5-FU in the treatment of patients with previously untreated advanced gastric cancer (AGC). Eur J Cancer. 2001;37:S314 Crossref
  • 28 C. Pozzo, C. Barone, J. Szanto, et al. Irinotecan in combination with 5-fluorouracil and folinic acid or with cisplatin in patients with advanced gastric or esophageal-gastric junction adenocarcinoma: results of a randomized phase II study. Ann Oncol. 2004;15:1773-1781 Crossref
  • 29 W. Koizumi, Y. Fukuyama, T. Fukuda, et al. Randomized phase II study comparing mitomycin, cisplatin plus doxifluridine with cisplatin plus doxifluridine in advanced unresectable gastric cancer. Anticancer Res. 2004;24:2465-2470
  • 30 O. Bouche, J.L. Raoul, F. Bonnetain, et al. Randomized multicenter phase II trial of a biweekly regimen of fluorouracil and leucovorin (LV5FU2), LV5FU2 plus cisplatin, or LV5FU2 plus irinotecan in patients with previously untreated metastatic gastric cancer: a Federation Francophone de Cancerologie Digestive Group Study–FFCD 9803. J Clin Oncol. 2004;22:4319-4328 Crossref
  • 31 Hawkins R, Cunningham D, Soerbye H et al. Randomized phase II trial of docetaxel plus irinotecan versus docetaxel plus 5-fluorouracil (5FU) in patients with untreated advanced gastric adenocarcinoma (AGAC). In: 39th Annual meeting of American Society of Clinical Oncology. Chicago, USA; May 31–June 3, 2003.
  • 32 P.C. Thuss-Patience, A. Kretzschmar, M. Repp, et al. Docetaxel and continuous-infusion fluorouracil versus epirubicin, cisplatin, and fluorouracil for advanced gastric adenocarcinoma: a randomized phase II study. J Clin Oncol. 2005;23:494-501
  • 33 J.A. Ajani, M.B. Fodor, S.A. Tjulandin, et al. Phase II multi-institutional randomized trial of docetaxel plus cisplatin with or without fluorouracil in patients with untreated, advanced gastric, or gastroesophageal adenocarcinoma. J Clin Oncol. 2005;23:5660-5667 Crossref
  • 34 A.D. Roth, N. Fazio, R. Stupp. Swiss Group for Clinical Cancer Research. Docetaxel, cisplatin, and fluorouracil; docetaxel and cisplatin; and epirubicin, cisplatin, and fluorouracil as systemic treatment for advanced gastric carcinoma: a randomized phase II trial of the Swiss Group for Clinical Cancer Research. J Clin Oncol. 2007;25:3217-3223 Crossref
  • 35 M. Dank, J. Zaluski, C. Barone, et al. Randomized phase III study comparing irinotecan combined with 5-fluorouracil and folinic acid to cisplatin combined with 5-fluorouracil in chemotherapy naive patients with advanced adenocarcinoma of the stomach or esophagogastric junction. Ann Oncol. 2008;19:1450-1457 Crossref
  • 36 N. Boku, S. Yamamoto, H. Fukuda, et al. Gastrointestinal Oncology Study Group of the Japan Clinical Oncology Group. Fluorouracil versus combination of irinotecan plus cisplatin versus S-1 in metastatic gastric cancer: a randomised phase 3 study. Lancet Oncol. 2009;10:1063-1069 Crossref
  • 37 G. Cocconi, V. DeLisi, B. Di Blasio. Randomized comparison of 5-FU alone or combined with mitomycin and cytarabine (MFC) in the treatment of advanced gastric cancer. Cancer Treat Rep. 1982;66:1263-1266
  • 38 V. De Lisi, G. Cocconi, F. Angelini, et al. The combination of cisplatin, doxorubicin, and mitomycin (PAM) compared with the FAM regimen in treating advanced gastric carcinoma. A phase II randomized trial of the Italian Oncology Group for Clinical Research. Cancer. 1996;77:245-250 Crossref
  • 39 A. Villar, F. Asensio, M. Candel, et al. Chemotherapy of advanced gastric carcinoma (stage IV): a randomized study of FAM versus 5-FU plus BCNU. Chemioterapia. 1987;6:57-62
  • 40 D. Kelsen, O.T. Atiq, L. Saltz, et al. FAMTX versus etoposide, doxorubicin, and cisplatin: a random assignment trial in gastric cancer. J Clin Oncol. 1992;10:541-548
  • 41 P.J. Loehrer, D. Harry, R.T. Chlebowski. 5-Fluorouracil vs. epirubicin vs. 5-fIuorouracil plus epirubicin in advanced gastric carcinoma. Investigational New Drugs. 1994;12:57-63 Crossref
  • 42 G. Cocconi, M. Bella, S. Zironi, et al. Fluorouracil, doxorubicin, and mitomycin combination versus PELF chemotherapy in advanced gastric cancer: a prospective randomized trial of the Italian Oncology Group for Clinical Research. J Clin Oncol. 1994;12:2687-2693
  • 43 V. De Lisi, G. Cocconi, M. Tonato, et al. Randomized comparison of 5-FU alone or combined with carmustine, doxorubicin, and mitomycin (BAFMi) in the treatment of advanced gastric cancer: a phase III trial of the Italian Clinical Research Oncology Group (GOIRC). Cancer Treat Rep. 1986;70:481-485
  • 44 G. Cocconi, P. Carlini, A. Gamboni. Et al.; Italian Oncology Group for Clinical Research. Cisplatin, epirubicin, leucovorin and 5-fluorouracil (PELF) is more active than 5-fluorouracil, doxorubicin and methotrexate (FAMTX) in advanced gastric carcinoma. Ann Oncol. 2003;14:1258-1263 Crossref
  • 45 N.C. Tebbutt, A. Norman, D. Cunningham, et al. A multicentre, randomised phase III trial comparing protracted venous infusion (PVI) 5-fluorouracil (5-FU) with PVI 5-FU plus mitomycin C in patients with inoperable oesophago-gastric cancer. Ann Oncol. 2002;13:1568-1575 Crossref
  • 46 P. Ross, M. Nicolson, D. Cunningham, et al. Prospective randomized trial comparing mitomycin, cisplatin, and protracted venous-infusion fluorouracil (PVI 5-FU) With epirubicin, cisplatin, and PVI 5-FU in advanced esophagogastric cancer. J Clin Oncol. 2002;20:1996-2004 Crossref
  • 47 Elsaid AA, Elkerm Y. Final results of a randomized phase III trial of docetaxel, carboplatin and 5FU versus epirubicin, cisplatin and 5FU for locally advanced gastric cancer. In: 41th Annual meeting of American Society of Clinical Oncology. Orlando, USA; May 13–15, 2005.
  • 48 M.P. Lutz, H. Wilke, D.J. Wagener. European Organisation for Research and Treatment of Cancer Gastrointestinal Group; Arbeitsgemeinschaft Internistische Onkologie. Weekly infusional high-dose fluorouracil (HD-FU), HD-FU plus folinic acid (HD-FU/FA), or HD-FU/FA plus biweekly cisplatin in advanced gastric cancer: randomized phase II trial 40953 of the European Organisation for Research and Treatment of Cancer Gastrointestinal Group and the Arbeitsgemeinschaft Internistische Onkologie. J Clin Oncol. 2007;25:2580-2585 Crossref
  • 49 S.H. Park, E. Nam, J. Park, et al. Randomized phase II study of irinotecan, leucovorin and 5-fluorouracil (ILF) versus cisplatin plus ILF (PILF) combination chemotherapy for advanced gastric cancer. Ann Oncol. 2008;19:729-733
  • 50 Ridwelski K, Fahlke J, Kettner E et al. Docetaxel-cisplatin (DC) versus 5-fluorouracil-leucovorin-cisplatin (FLC) as first-line treatment for locally advanced or metastatic gastric cancer: Preliminary results of a phase III study. In: 44th Annual meeting of American Society of Clinical Oncology. Chicago, USA; May 30–June 3, 2008.
  • 51 Imamura H, Iiishi H, Tsuburaya A et al. Randomized phase III study of irinotecan plus S-1 (IRIS) versus S-1 alone as first-line treatment for advanced gastric cancer (GC0301/TOP-002). In: 2008 Gastrointestinal cancers symposium. Orland, USA; January 25–27, 2008.
  • 52 W. Koizumi, H. Narahara, T. Hara, et al. S-1 plus cisplatin versus S-1 alone for first-line treatment of advanced gastric cancer (SPIRITS trial): a phase III trial. Lancet Oncol. 2008;9:215-221 Crossref
  • 53 I.P. Popov, S.B. Jelić, Z.V. Krivokapić, et al. Bimonthly 24 h infusion of high-dose 5-fluorouracil vs EAP regimen in patients with advanced gastric cancer. A randomized phase II study. Med Oncol. 2008;25:73-80 Crossref
  • 54 The Gastrointestinal Tumor Study Group. A comparative clinical assessment of combination chemotherapy in the management of advanced gastric carcinoma: The Gastrointestinal Tumor study Group. Cancer. 1982;49:1362-1366
  • 55 S. Gupta. Treatment of advanced gastric cancer with 5-fluorouracil versus mitomycin C. J Surg Oncol. 1982;21:94-96 Crossref
  • 56 The Gastrointestinal Tumor Study Group. Randomized study of combination chemotherapy in unresectable gastric cancer. Cancer. 1984;53:13-17
  • 57 G. Schnitzler, W. Queisser, M.E. Heim, et al. Phase III study of 5-FU and carmustine versus 5-FU, carmustine, and doxorubicin in advanced gastric cancer. Cancer Treat Rep. 1986;70:477-479
  • 58 M. Hayakawa, K. Morise, Y. Oka, et al. FEP and FAP combination chemotherapy in advanced gastric cancer. Research Group for Gastric Cancer. Gan To Kagaku Ryoho. 1988;15:3073-3079
  • 59 T. Sasai, M. Kurihara, Y. Kubo, et al. UFT + MMC and UFT + ACNU therapy in advanced gastric cancer. Gan To Kagaku Ryoho. 1988;15:141-146
  • 60 Gastrointestinal Tumor Study Group (GITSG). Triazinate and platinum efficacy in combination with 5-fluorouracil and doxorubicin: results of a three-arm randomized trial in metastatic gastric cancer. J Natl Cancer Inst. 1988;80:1011-1015
  • 61 Y. Goto, T. Toyota, S. Asaki, et al. Randomized-controlled study of treatment with UFT-MMC or UFT-ACR in advanced gastric cancer. Tohoku Study Group of Cancer Treatment for the Digestive Organs. Gan To Kagaku Ryoho. 1989;16:2227-2233
  • 62 K. Kikuchi, A. Wakui, H. Shimizu, Y. Kunii. Randomized controlled study on chemotherapy with 5-FD, ADM plus CDDP in advanced gastric carcinoma. Gan To Kagaku Ryoho. 1990;17:655-662
  • 63 E. Diaz-Rubio, J. Jimeno, E. Aranda, et al. Etoposide (E) + epirubicin (E) + cisplatin (P) combination chemotherapy (EEP) in advanced gastric cancer: negative impact on clinical outcome. Spanish Cooperative Group for GI Tumor Therapy (T.T.D.). Ann Oncol. 1992;3:861-863
  • 64 W. Scheithauer, G. Kornek, M. Heina, et al. Palliative chemotherapy versus best supportive care in patients with metastatic gastric cancer: a randomized trial. Ann Hematol. 1997;72 A181.107
  • 65 G. Colucci, F. Giotta, E. Maiello, et al. Efficacy of the association of folinic acid and 5-fluorouracil alone versus folinic acid and 5-fluorouracil plus 4-epidoxorubicin in the treatment of advanced gastric cancer. Am J Clin Oncol. 1995;18:519-524 Crossref
  • 66 Y.J. Bang, E. Van Cutsem, A. Feyereislova, et al. Trastuzumab in combination with chemotherapy versus chemotherapy alone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label, randomised controlled trial. Lancet. 2010;376:687-697 Crossref
  • 67 D.H. Koo, B.Y. Ryoo, H.J. Kim, et al. A prognostic model in patients who receive chemotherapy for metastatic or recurrent gastric cancer: validation and comparison with previous models. Cancer Chemother Pharmacol. 2011;68:913-921 Crossref
  • 68 Nisen P, Freeman A, Rockhold F. Opening the gate to the walled garden: providing access to patient level data from clinical trials, submitted for publication.
  • 69 NIH Data Sharing Policy and Implementation Guidance. < http://grants.nih.gov/grants/policy/data_sharing/data_sharing_guidance.htm >; 2012 [accessed 15.12.12].
  • 70 F. Godlee, T. Groves. The new BMJ policy on sharing data from drug and device trials. BMJ. 2012;345:e7888 Crossref

Footnotes

a Membership and addresses of the GASTRIC Group are listed in the Appendix.

b Correspondence to: Dr K Oba, Translational Research and Clinical Trial Center, Hokkaido University Hospital, Kita 14, Nishi 5, Kita-ku, Sapporo, Hokkaido 060-8648, Japan; tel.: +81 11 706 7413; e-mail: k.oba@huhp.hokudai.ac.jp