|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 22, No 18 (September 15), 2004: pp. 3766-3775 © 2004 American Society of Clinical Oncology. DOI: 10.1200/JCO.2004.03.104 Modulation of Fluorouracil by Leucovorin in Patients With Advanced Colorectal Cancer: An Updated Meta-AnalysisThe Meta-Analysis Group in CancerFrom the Meta-Analysis Group in Cancer Address reprint requests to Pascal Piedbois, MD, PhD, Department of Medical Oncology, Assistance Publique, Hopitaux de Paris, Henri Mondor Hospital, 94000 Creteil, France; e-mail: pascal.piedbois{at}hmn.ap-hop-paris.fr
PURPOSE: The modulation of fluorouracil (FU) by folinic acid (leucovorin [LV]) has been shown to be effective in terms of tumor response rate in patients with advanced colorectal cancer, but a meta-analysis of nine trials previously published by our group failed to demonstrate a statistically significant survival difference between FU and FU-LV. We present an update of the meta-analysis, with a longer follow-up and the inclusion of 10 newer trials. PATIENTS AND METHODS: Analyses are based on individual data from 3,300 patients randomized in 19 trials on an intent-to-treat basis. Two trials had multiple comparisons, leading to a total of 21 pair-wise comparisons. FU doses were similar in both arms in 10 pair-wise comparisons, 15% to 33% higher in the FU-alone arm in six comparisons, and more than 66% higher in five comparisons. RESULTS: Overall analysis showed a two-fold increase in tumor response rates (11% for FU-LV v 21% for FU alone; odds ratio, 0.53; 95% CI, 0.44 to 0.63; P < .0001) and a small but statistically significant overall survival benefit for FU-LV over FU alone (median survival, 11.7 v 10.5 months, respectively; hazards ratio, 0.90; 95% CI, 0.87 to 0.94; P = .004), which were primarily seen in the first year. We observed a significant interaction between treatment benefit and dose of FU, with tumor response and overall survival advantages of FU-LV over FU-alone being restricted to trials in which a similar dose of FU was prescribed in both arms. CONCLUSION: This updated analysis demonstrates, on a large data set, that FU-LV improves both response rate and overall survival compared with FU alone and that this benefit is consistent across various prognostic factors.
Approximately 50% of patients diagnosed with colorectal carcinoma have metastatic or nonresectable disease at time of diagnosis or will develop metastases and/or a local recurrence after their initial diagnosis. Despite the recent development of new chemotherapy compounds, the outcome of patients with advanced colorectal carcinoma (ACC) remains grim. Introduced in clinical practice in the late 1950s, fluorouracil (FU) remains an essential component of chemotherapy regimens in ACC. Substantial progress has been made over the past 20 years in the use of FU. In a series of meta-analyses published in the 1990s, our group, the Meta-Analysis Group in Cancer (MAGIC), showed that FU alone delivered as a bolus infusion has a limited efficacy in patients with ACC, providing a tumor response rate of 10% to 15%, whereas the modulation of FU either by leucovorin (LV)1 or methotrexate2 led to a doubling of tumor response rates. A similar increase in tumor response rate was also observed with the use of FU as an intravenous continuous infusion,3 and an even greater efficacy was reported in patients with metastases confined to the liver treated by locoregional administration of fluoropyrimidines, using hepatic artery infusion.4 Even though the impact on overall survival of these FU administration modalities remains modest,5 a small but statistically significant survival benefit was reported in each of these meta-analyses, with the exception of the FU versus FU-LV meta-analysis.1 Meanwhile, FU-LV has proven to be an efficient regimen in the adjuvant setting for patients with resected stage II or III colon cancer.6-8 Therefore, we hypothesized that the lack of survival benefit reported in the initial FU-LV meta-analysis published in 1992 may have been because of a lack of statistical power and because heterogeneity across the trials was not sufficiently taken into account. We decided to reassess the comparison of FU with FU-LV through an update of the initial FU-LV meta-analysis, including new trials and longer follow-up.
The search for eligible trials and the data collection were performed by MAGIC (Creteil, France). Data were checked in the Department of Biostatistics of Gustave Roussy Institute (Villejuif, France), where analyses were also performed in collaboration with MAGIC.
Eligible Trials On the basis of the findings of our previous analyses,2,3 trials were stratified according to the difference in prescribed FU dose between treatment arms. This stratification by FU dose was not performed in the initial FU-LV meta-analysis.
Search for Relevant Trials The previous meta-analysis included the following nine trials: City of Hope,10 Toronto,11 Genova,12 Bologna,13 Gruppo Italiano Studio Carcinomi Apparato Digerente,14 Roswell Park Cancer Institute,15 Gruppo Oncologico Italiano di Ricerca Clinica,16 Gastrointestinal Tumor Study Group (GITSG),17 and Northern California Oncology Group (NCOG).18 Updated patient data were obtained for the Toronto trial11; update of patient follow-up was not necessary in the remaining trials. The following 13 additional trials were identified: North Central Cancer Treatment Group (NCCTG)-Mayo Clinic,19 Siena,20 Crema,21 Schweizerische Arbeitsgruppe fur Klinische Krebsforschung (SAKK),22 European Organization for Research and Treatment of Cancer (EORTC),23 Southwest Oncology Group (SWOG),24 Arbeitsgemeinschaft Internische Onkologie (AIO),25 Hellenic Cooperative Oncology Group,26 Spain,27 Hungary,28 Bern,29 Arbeitsgemeinschaft Gastroenterologische Onkologie Nordheim-Westfalen,30 and Tubingen.31 Three of the additional trials,29-31 representing a total of 331 patients, were not made available for the meta-analysis. Thus, the present meta-analysis was based on 19 trials, including 3,338 patients (Table 1). The SWOG trial24 and the EORTC trial23 had two different FU-LV arms. Thus, control arms of these two trials were considered twice in analysis, bringing the total number of pair-wise comparisons to 21 and the total number of patients for the calculations to 3,597.
The FU predictive cumulative dose was calculated for each study arm at the end of the 17th week of treatment. The pair-wise comparisons were stratified in three groups according to the difference in predictive cumulative FU dose between treatment arms (Table 1). Group 1 consisted of 10 comparisons (1,791 patients) in which no difference in FU dose was found between the two arms (City of Hope,10 Toronto,11 NCCTG-Mayo Clinic,19 Genova,12 Crema,21 Bologna,13 Gruppo Italiano Studio Carcinomi Apparato Digerente,14 SAKK,22 and EORTC 123) and one comparison in which FU was administered at a slightly lower dose in the FU alone arm (Siena20). Group 2 consisted of six comparisons (834 patients) in which FU was given at a 15% to 33% higher dose in the FU alone arm (Roswell Park Cancer Institute,15 Gruppo Oncologico Italiano di Ricerca Clinica,16 SWOG 1 and 224, AIO,25 and Hellenic Cooperative Oncology Group26). Group 3 consisted of five comparisons (972 patients) in which FU dose was at least 66% higher in the FU alone arm (GITSG,17 NCOG,18 Spain,27 Hungary,28 and EORTC 223).
Protocol for the Meta-Analysis
Data Collection
Patient Characteristics
Tumor Response and Survival Complete response and partial response criteria adopted in all individual trials followed the WHO recommendations.32 For the purpose of the meta-analyses, patients experiencing minimal response, stable disease, or progressive disease were considered to have no response. In all trials, treatment was maintained until disease progression or severe toxicity. Duration of survival was calculated from the date of randomization to the date of death, whatever its cause.
Statistical Analysis Subgroup analyses were performed to check the effects of the treatment on predefined patient characteristics. Median follow-up was computed by the potential follow-up method.39 Prognostic factors for survival were identified through a proportional hazards regression model.40 CIs for the proportion of responders are based on normal approximations. All P values were two-sided.
Tumor Response Analysis Overall tumor response rate for the entire population was 17% (95% CI, 16% to 18%). Six individual trials (City of Hope,10 Toronto,11 NCCTG,19 Crema,21 SAKK,22 and GITSG17) showed a benefit of FU-LV over FU alone in terms of tumor response. In the combined analysis, the tumor response rate was 11% (95% CI, 9% to 12%) for patients allocated to FU alone and 21% (95% CI, 19% to 23%) for patients allocated to FU-LV. This result was statistically significant (P < .0001), with an OR for response of 0.53 (95% CI, 0.44 to 0.63). The ORs for tumor response for individual trials and overall are presented in Figure 1. There was a significant heterogeneity between trials, with a significant interaction between treatment benefit in terms of tumor response and the predictive cumulative FU dose (Fig 1). No statistically significant difference in tumor response rate was demonstrated for trials in which the FU dose was increased (> 15%) in the FU alone arm, although there was a trend toward improved response rates with FU-LV in all three strata. There was no significant residual heterogeneity (P = .38) beyond this one.
In three comparisons (EORTC 1,23 AIO,25 and Hungary28), FU was administered as a continuous infusion in both arms. A sensitivity analysis showed that after exclusion of these three trials, the advantage of FU-LV over FU in terms of tumor response remained highly significant (OR, 0.49; 95% CI, 0.40 to 0.59). In a logistic regression model, good performance status (P < .0001), metastases confined to the liver (P = .002), and FU-LV arm allocation (P < .0001) were found to be independent favorable prognostic factors for tumor response. The tumor response rate was 19%, 18%, and 11% for patients with a performance status of 0, 1, and 2 or more, respectively. The tumor response rate was 20% for patients with metastases confined to the liver compared with 16% for patients with other metastatic sites. Subgroup analyses for the predefined patient characteristics showed no significant interactions between the treatment effect and patient characteristics. However, the interaction for performance status showed a strong trend at P > .05, with patients who had a performance status of 2 receiving the maximal benefit (Fig 2).
Survival Analysis For the entire patient population, the median follow-up was 45 months. The median overall survival duration was 11.1 months. Only one individual trial (Crema21) showed a statistically significant survival benefit in favor of the FU-LV regimen. The combined analysis found a statistically significant survival benefit in favor of FU-LV (HR, 0.90; 95% CI, 0.87 to 0.94; P = .004). The survival HRs for individual trials and overall are presented in Figure 3. The median survival time was 10.5 months for patients allocated to FU alone and 11.7 months for patients allocated to the FU-LV. A test for proportionality of the HRs calculated over 6-month periods of time is presented in Table 3 and indicates that the survival differences between FU alone and FU-LV do not evolve constantly over time (test for interaction, P = .03).
Significant heterogeneity between trials was found (test for heterogeneity, P = .03), with a significant interaction between treatment survival benefit and differences in predictive FU cumulative dose between treatment arms (test for interaction, P = .02; and test for trend, P = .005). The survival benefit of biochemical modulation was restricted to trials using the same FU dose in the two treatment arms. No statistically significant difference was found in trials in which the dose of FU was increased (> 15%) in the FU alone arm. The difference of the size of treatment effect between the three trial groups explains most of the heterogeneity. There is no significant residual heterogeneity (P = .11) beyond this one. A sensitivity analysis showed that the heterogeneity was no longer significant after removing the only positive trial13 (test for heterogeneity, P = .46), whereas the benefit in survival in favor of FU-LV persisted (HR, 0.92; 95% CI, 0.86 to 0.99; P = .03). One-year survival rates were 47% for patients allocated to FU-LV versus 37% for patients allocated to FU alone in the same dose of FU group, 55% versus 55% in the 15% to 33% increase in FU group, and 46% versus 46% in the more than 66% increase in FU group. A sensitivity analysis showed that the benefit of FU-LV over FU alone persists after exclusion of the three comparisons in which FU was administered as a continuous infusion in both arms (HR, 0.89; 95% CI, 0.82 to 0.96). Using a Cox regression model, good performance status (P < .0001, when used as categorical variable), metastases confined to the liver (P = .0002) and to the lung (P < .0001), and FU-LV chemotherapy (P = .0001) were independent good prognosis factors for overall survival. One-year survival rates were 63% for patients with a performance status of 0, 45% for patients with a performance status of 1, and 20% for patients with a performance status of 2 or more. The 1-year survival rates were 50% for patients with liver metastases only, 57% for patients with lung metastases only, and 44% for patients with other metastatic sites. Subgroup analyses for treatment effect showed no significant interactions between patient characteristics and treatment benefit (Fig 4). Again, there was the suggestion of an interaction between treatment benefit and performance status, with patients who had a performance status of 2 showing the greatest benefit. Overall survival curves by performance status and treatment are illustrated in Figure 5.
This updated meta-analysis has demonstrated that, in patients with ACC, the advantage of FU-LV over FU alone is not limited to tumor response. The present study found that, in addition to a two-fold increase in tumor response rate (OR, 0.53; P < .0001), there is a small but statistically significant improvement in overall survival (HR, 0.90; P = .004) in favor of FU-LV. Compared with the initial meta-analysis, the inclusion of more recent trials did not significantly increase the median survival duration, despite the recent availability of second-line chemotherapy regimens. In fact, the treatment benefit was primarily observed in the first year (HR, 0.84; 95% CI, 0.76 to 0.94, for time period of 6 to 12 months) but not for the time period of 12 to 18 months (HR, 0.94; 95% CI, 0.83 to 1.06). The 1- and 2-year survival rates were 43% and 17%, respectively, for the patients allocated to FU alone and 49% and 17%, respectively, for the patients allocated to FU-LV. These findings provide indirect evidence that the chemotherapy regimen chosen as first-line therapy has a definitive impact, albeit small, on survival. A possible FU dose effect was hypothesized after the results of two previous meta-analyses.1,2 The initial FU versus FU-LV meta-analysis1 and the FU versus FU plus methotrexate meta-analysis2 generated the hypothesis that the benefit of FU modulation could be compensated by an increase of the FU dose in the FU alone arm. Therefore, a stratification of trials according to FU doses was prospectively planned in the statistical design of the present study to further address this issue. The updated meta-analysis confirmed a clear interaction between treatment benefit and FU dose, both for tumor response (P < .0001) and survival (P = .02). For both end points, the advantage of the modulation was restricted to trials using similar FU dosage in both arms. In this group of trials, 1-year survival rates were 37% for patients allocated to FU alone versus 47% for patients allocated to FU-LV, and 2-year survival rates were 12% versus 14%, respectively. Thus, the data presented here could suggest that high-dose FU can be substituted with FU modulation by LV. This statement would not take into consideration the increase of toxicity that is observed with high-dose FU. Although no data were recorded on toxicity in the meta-analysis, individual trials have clearly shown that the toxicity induced by high-dose FU is not acceptable. In the NCOG trial18 for instance, treatment efficacy was similar in the high-dose FU alone and the FU-LV arms, but toxicity was much higher in the FU alone arm. This greater toxicity of high-dose FU alone compared with FU-LV was not only observed for grades 3 and 4 hematologic toxicity (21% v 12%, respectively) but also for nonhematologic toxicity (42% v 24%, respectively). The authors concluded that "experimental regimens had superior therapeutic ratios of benefits versus toxicity."18 Analyses based on the actual administered dose would have needed the collection of chemotherapy doses delivered at every course of treatment. Because a retrospective collection of this information on 3,300 patients was not feasible, we used projected doses instead of actually administered doses. The potential limitations of this approach, especially when individual patient toxicity data were not available, must be underlined. However, the interaction between treatment benefit and FU dose is so clear that our findings concerning the importance of FU dose seem reliable. The assessment of prognostic factors is one of the advantages of meta-analyses based on individual patient data; however, such analyses are limited to those parameters prospectively recorded. Because of the time period during which the trials included in the meta-analyses were conducted, only main clinical characteristics were available for prognostic factors analyses. Performance status was already reported in our meta-analyses as the most relevant prognostic factors for tumor response2,3 and overall survival.1-4,34 Metastases confined to the liver or to the lung were also reported as favorable independent prognostic factors for overall survival in a meta-analysis.34 Rectum as the primary tumor site was associated with higher tumor response rate34 and better survival.3,34 The present meta-analysis confirms the relevance of simple clinical factors, such as performance status and site of metastases, for the outcome of patients with ACC treated with fluoropyrimidines. Should new biologic markers be prospectively collected in large-scale individual clinical trials, future meta-analyses will be able to better establish the prognosis of patients with ACC. The meta-analysis did not identify clear subgroups of patients who benefit more from FU-LV regimen. The interaction test between treatment effect and performance status had a borderline significant effect both for tumor response (P = .05) and survival (P = .08). Because most of the patients with a performance status of 2 belong to the group of trials with the same FU dose (68%) and 23% belong to the Siena trial, this result should be interpreted with caution. The updated meta-analysis confirms the value of FU-LV in the management of patients with ACC.
Writing Committee; P. Thirion (Saint Luc Hospital, Ireland), S. Michiels, J.P. Pignon (Institut Gustave Roussy, Villejuif, France), M. Buyse (International Drug Development Institute, Brussels, Belgium), A.C. Braud  (Henri Mondor Hospital, Créteil, France), R.W. Carlson (Stanford University, Stanford, CA), M. O’Connell, P. Sargent (Mayo Clinic, Rochester, MN), and P. Piedbois (Henri Mondor Hospital, Créteil, France).
An additional appendix is included in the full-text version of this article, available online at www.jco.org. It is not included in the PDF (via Adobe® Acrobat Reader®) version.
The following collaborators participated in this study: J.H. Doroshow (City of Hope, Duarte), C. Erlichman, S. Fine, T. Gadalla (Princess Margaret Hospital, Toronto), M. O'Connell, D. Sargent (North Central Cancer Treatment Group-Mayo Clinic), M.T. Nobile , R. Rosso, M.R. Sertoli, A. Sobrero (Istituto Nazionale per la Ricerca sul Cancro, Genova), R. Petrioli (Siena), E. Bobbio-Pallavicini, C. Porta (Crema and Pavia), A. Martoni, A. Cricca, F. Pannuti, M. Casadio (Ospedale Policlinico G.Orsola-M. Malpighi, Bologna), R. Labianca, G. Pancera, B. Cesana, G. Luporini, G. Beretta, G. Martignoni (Gruppo Italiano Studio Carcinomi Apparato Digerente), M.M. Borner, M. Castiglione, M. Bacchi, R. Maibach (Schweizerische Arbeitsgruppe fur Klinische Krebsforschung), C.H. Kohne, R. Sylvester (European Organization for Research and Treatment of Cancer), P. Creaven, N. Petrelli, Y. Rustum, P. Burke (Roswell Park Cancer Institute, Buffalo, and Gastrointestinal Tumor Study Group), F. Di Costanzo, M. Bacchi, S. Gasperoni (Gruppo Oncologico Italiano di Ricerca Clinica), C.G. Leichman, J. Benedetti (Southwest Oncology Group), C.H. Kohne (Arbeitsgemeinschaft Internische Onkologie), H.P. Kalofonos (Hellenic Cooperative Oncology Group), F. Valone, B. Brown, R.W. Carlson (Northern California Oncology Group), A. Abad (Spain), G. Pajkos (Hungary), M. Buyse (International Drug Development Institute, Brussels, Belgium), J.P. Pignon, S. Michiels, Elodie Carmona (Institut Gustave Roussy, Villejuif, France), P. Thirion (Saint Luc Hospital, Ireland), and P. Piedbois, A.C. Braud, Y. Piedbois, S. Laurent (Henri Mondor Hospital, Créteil, and AERO, France).
The authors indicated no potential conflicts of interest.
We thank Youri Piedbois for his help identifying relevant studies to be included in the meta-analysis and Elodie Carmona for her help in checking the data. This meta-analysis is dedicated to the memory of our colleagues, Anne-Chantal Braud, MD, and Maria-Teresa Nobile, MD, who participated actively in this work and died accidentally before this article was accepted for publication. Their smiles and professional expertise will continue to influence all who knew them.
Deceased. Supported by a grant from the French "Programme Hospitalier de Recherche Clinique" and by the European Association for Research in Oncology. Presented in part at the 39th Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, May 31-June 3, 2003. Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Advanced Colorectal Cancer Meta-Analysis Project: Modulation of 5-fluorouracil by leucovorin in patients with advanced colorectal cancer: Evidence in terms of response rate. J Clin Oncol 10:896-903, 1992[Abstract]
2. Advanced Colorectal Cancer Meta-Analysis Project: Meta-analysis of randomized trials testing the biochemical modulation of 5-fluorouracil by methotrexate in metastatic colorectal cancer. J Clin Oncol 12:960-969, 1994
3. Meta-Analysis Group in Cancer: Efficacy of intravenous continuous infusion of fluorouracil compared with bolus administration in advanced colorectal cancer. J Clin Oncol 16:301-308, 1998
4. Meta-Analysis Group in Cancer: Reappraisal of hepatic arterial infusion in the treatment of non resectable liver metastases from colorectal cancer. J Natl Cancer Inst 88:252-258, 1996 5. Buyse M, Thirion P, Carlson RW, et al: Relation between tumour response to first-line chemotherapy and survival in advanced colorectal cancer: A meta-analysis. Lancet 356:373-378, 2000[CrossRef][Medline]
6. Moertel CG, Fleming TR, Macdonald JS, et al: Fluorouracil plus levamisole as effective adjuvant therapy after resection of stage III colon carcinoma: A final report. Ann Intern Med 122:321-326, 1995
7. O'Connell MJ, Mailliard JA, Kahn MJ, et al: Controlled trial of fluorouracil and low-dose leucovorin given for 6 months as postoperative adjuvant therapy for colon cancer. J Clin Oncol 15:246-250, 1997 8. International Multicentre Pooled Analysis of Colon Cancer Trials (IMPACT) Investigators: Efficacy of adjuvant fluorouracil and folinic acid in colon cancer. Lancet 345:939-944, 1995[CrossRef][Medline]
9. Sobrero AF, Aschele C, Bertino JR: Fluorouracil in colorectal cancer: A tale of two drugs—Implications for biochemical modulation. J Clin Oncol 15:368-381, 1997 10. Doroshow JH, Multhauf P, Leong L, et al: Prospective randomized comparison of fluorouracil versus fluorouracil and high-dose continuous infusion leucovorin calcium for the treatment of advanced measurable colorectal cancer in patients previously unexposed to chemotherapy. J Clin Oncol 8:491-501, 1990[Abstract] 11. Erlichman C, Fine S, Wong A, et al: A randomized trial of fluorouracil and folinic acid in patients with metastatic colorectal carcinoma. J Clin Oncol 6:469-475, 1988[Abstract] 12. Nobile MT, Rosso R, Sertoli MR, et al: Randomised comparison of weekly bolus 5-fluorouracil with or without leucovorin in metastatic colorectal carcinoma. Eur J Cancer 28A:1823-1827, 1992 13. Martoni A, Cricca A, Guaraldi M, et al: Weekly regimen of 5-FU vs 5-FU + intermediate dose folinic acid in the treatment of advanced colorectal cancer. Anticancer Res 12:607-612, 1992[Medline]
14. Labianca R, Pancera G, Aitini E, et al: Folinic acid + 5-fluorouracil (5-FU) versus equidose 5-FU in advanced colorectal cancer. Phase III study of 'GISCAD' (Italian Group for the Study of Digestive Tract Cancer). Ann Oncol 2:673-679, 1991
15. Petrelli N, Herrera L, Rustum Y, et al: A prospective randomized trial of 5-fluorouracil versus 5-fluorouracil and high-dose leucovorin versus 5-fluorouracil and methotrexate in previously untreated patients with advanced colorectal carcinoma. J Clin Oncol 5:1559-1565, 1987
16. Di Costanzo F, Bartolucci R, Calabresi F, et al: Fluorouracil-alone versus high-dose folinic acid and fluorouracil in advanced colorectal cancer: A randomized trial of the Italian Oncology Group for Clinical Research (GOIRC). Ann Oncol 3:371-376, 1992 17. Petrelli N, Douglass HO Jr, Herrera L, et al: The modulation of fluorouracil with leucovorin in metastatic colorectal carcinoma: A prospective randomized phase III trial. J Clin Oncol 7:1419-1426, 1989[Abstract] 18. Valone FH, Friedman MA, Wittlinger PS, et al: Treatment of patients with advanced colorectal carcinomas with fluorouracil alone, high-dose leucovorin plus fluorouracil, or sequential methotrexate, fluorouracil, and leucovorin: A randomized trial of the Northern California Oncology Group. J Clin Oncol 7:1427-1436, 1989[Abstract] 19. Poon MA, O'Connell MJ, Moertel CG, et al: Biochemical modulation of fluorouracil: Evidence of significant improvement of survival and quality of life in patients with advanced colorectal carcinoma. J Clin Oncol 7:1407-1417, 1989[Abstract] 20. Petrioli R, Lorenzi M, Aquino A, et al: Treatment of advanced colorectal cancer with high-dose intensity folinic acid and 5-fluorouracil plus supportive care. Eur J Cancer 31A:2105-2108, 1995 21. Bobbio-Pallavicini E, Porta C, Moroni M, et al: Folinic acid does improve 5-fluorouracil activity in vivo. Results of a phase III study comparing 5-fluorouracil to 5-fluorouracil and folinic acid in advanced colon cancer patients. J Chemother 5:52-55, 1993[Medline]
22. Borner MM, Castiglione M, Bacchi M, et al: The impact of adding low-dose leucovorin to monthly 5-fluorouracil in advanced colorectal carcinoma: Results of a phase III trial. Ann Oncol 9:535-541, 1998
23. Kohne CH, Wils J, Lorenz M, et al: Randomized phase III study of high-dose fluorouracil given as a weekly 24-hour infusion with or without leucovorin versus bolus fluorouracil plus leucovorin in advanced colorectal cancer: European Organization for Research and Treatment of Cancer Gastrointestinal Group Study 40952. J Clin Oncol 21:3721-3728, 2003 24. Leichman CG, Fleming TR, Muggia FM, et al: Phase II study of fluorouracil and its modulation in advanced colorectal cancer: A Southwest Oncology Group study. J Clin Oncol 13:1303-1311, 1995[Abstract] 25. Kohne CH, Schoffski P, Wilke H, et al: Effective biomodulation by leucovorin of high-dose infusion fluorouracil given as a weekly 24-hour infusion: Results of a randomized trial in patients with advanced colorectal cancer. J Clin Oncol 16:418-426, 1998[Abstract] 26. Kalofonos HP, Nicolaides C, Samantas E, et al: A phase III study of 5-fluorouracil versus 5-fluorouracil plus interferon alpha 2b versus 5-fluorouracil plus leucovorin in patients with advanced colorectal cancer: A Hellenic Cooperative Oncology Group (HeCOG) study. Am J Clin Oncol 25:23-30, 2002[CrossRef][Medline] 27. Abad A, Garcia P, Gravalos C, et al: Sequential methotrexate, 5-fluorouracil (5-FU), and high dose leucovorin versus 5-FU and high dose leucovorin versus 5-FU alone for advanced colorectal cancer. A multi-institutional randomized trial. Cancer 75:1238-1244, 1995[CrossRef][Medline] 28. Pajkos G, Izso J, Kristo K, et al: Biochemical modulation of 5-fluorouracil (FU) by leucovorin (LV) and/or interferon alpha-2a (IFN) in metastatic colorectal cancer (MCC). 7th International Congress on Anti-Cancer Treatment, Paris, France, February 6-9, 1997 (abstr P 275) 29. Borner M, Brand B, Lang M, et al: Low-dose leucovorin (LD-LV) significantly changes the effect of fluorouracil (FU): Results of a randomized pilot study in advanced colorectal carcinoma (ACC). Proc Am Soc Clin Oncol 11:183, 1992 (abstr 546) 30. Löffler TM, Korsten FW, Reis HE, et al: Fluorouracil al. monotherapie oder in kombination mit folinsäure in der behandlung des metastasierten kolorektalen karzinoms. Dtsch Med Wochenschr 117:1007-1013, 1992[Medline] 31. Steinke B, Gunther E, Hirschmann WD, et al: Fluorouracil versus folinic acid/fluorouracil in advanced colorectal cancer: Preliminary results of a randomized trial. Semin Oncol 19:141-147, 1992 (suppl 3)[Medline] 32. Miller AB, Hoogstraten B, Staquet M, et al: Reporting results of cancer treatment. Cancer 47:207-214, 1981[CrossRef][Medline] 33. Meta-Analysis Group in Cancer: Toxicity of 5-fluorouracil in patients with advanced colorectal cancer: Effect of administration schedule and prognostic factors. J Clin Oncol 16:3537-3541, 1998[Abstract] 34. Thirion P, Piedbois P, Buyse M, et al: Alpha-interferon does not increase the efficacy of 5-fluorouracil in advanced colorectal cancer. Br J Cancer 84:611-620, 2001[CrossRef][Medline] 35. Mantel H, Haenszel W: Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 22:719-748, 1959 36. Cox DR: The analysis of binary data. London, United Kingdom, Methuen, 1970 37. Peto R, Pike MC, Armitage P, et al: Design and analysis of randomized clinical trials requiring prolonged observation of each patient: II. Analysis and examples. Br J Cancer 35:1-39, 1977[Medline]
38. Non-Small Cell Lung Cancer Collaborative Group: Chemotherapy in non-small cell lung cancer: A meta-analysis using updated data on individual patients from 52 randomised clinical trials. BMJ 311:899-909, 1995 39. Schemper M, Smith TL: A note on quantifying follow-up in studies of failure time. Control Clin Trials 17:343-346, 1996[CrossRef][Medline] 40. Cox DR: Regression models and life tables. J R Stat Soc B 34:187-220, 1972 Submitted March 15, 2004; accepted June 18, 2004.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2004 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
) to FU-LV (± IFN-
2 heterogeneity tests were used to test for statistical heterogeneity across trials. An indirect comparison was performed between trials according to the difference in predicted cumulative FU dose between treatment arms through a test for trend and for interaction. Residual heterogeneity after accounting for predicted cumulative FU dose was tested using the difference between the 
