|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2004.02.109 on August 23 2004 © 2004 American Society of Clinical Oncology. Meta-Analysis of Randomized Clinical Trials Comparing Cisplatin to Carboplatin in Patients With Advanced Non–Small-Cell Lung CancerFrom the Department of Medicine II, Okayama University Medical School, Okayama; Department of Epidemiology and Prevention, Aichi Cancer Center Research Institute, Nagoya, Japan. Address reprint requests to Katsuyuki Hotta, MD, Department of Medicine II, Okayama University Medical School, 2-5-1 Shikata-cho, Okayama 700-8558, Japan; e-mail: khotta{at}md.okayama-u.ac.jp
PURPOSE: It remains undetermined whether cisplatin and carboplatin are equally effective for advanced non–small-cell lung cancer (NSCLC). We therefore did a meta-analysis of trials that compared cisplatin-based chemotherapy with carboplatin-based chemotherapy. METHODS: We performed a literature search to identify trials that had investigated the substitution of carboplatin for cisplatin in the treatment of advanced NSCLC. We evaluated these trials for inclusion, rated methodologic quality, and abstracted relevant data. RESULTS: Of 1,191 reports, eight trials (2,948 patients) were identified, five of which investigated drug regimens containing platinum plus a new agent. Cisplatin-based chemotherapy produced a higher response rate, but the survival advantage was not significant (hazard ratio = 1.050; 95% CI, 0.907 to 1.216; P = .515). Subgroup analysis revealed that combination chemotherapy consisting of cisplatin plus a new agent yields 11% longer survival than carboplatin plus the same new agent (hazard ratio = 1.106; 95% CI, 1.005 to 1.218; P = .039). Patients on cisplatin-based chemotherapy frequently developed nausea and vomiting; thrombocytopenia was more frequent during carboplatin-based chemotherapy. No significant difference in treatment-related mortality was observed. CONCLUSION: We found that combination chemotherapy consisting of cisplatin plus a new agent yields a substantial survival advantage compared with carboplatin plus a new agent in patients with advanced NSCLC, although we failed to find any survival difference in an analysis that included both new and old agents. The strength of our conclusion is limited because we used abstracted data, and careful interpretation is thus required. Nevertheless, our results raise a critical point that needs to be evaluated in future studies.
Lung cancer is the leading cause of cancer-related deaths in many countries. Approximately one third of patients with non–small-cell lung cancer (NSCLC) have metastatic disease at the time of diagnosis.1 Cisplatin-based chemotherapy is currently considered to be the standard treatment in advanced NSCLC, with a 10% absolute improvement in the 1-year survival rate compared to supportive care alone.2 However, many medical oncologists remain skeptical about these data and have not routinely used cisplatin-based chemotherapy to treat patients with advanced NSCLC.3 This reluctance may be partly explained by the severe toxicity that is associated with cisplatin-based chemotherapy. In an attempt to circumvent cisplatin-induced toxicities, carboplatin, an analog of cisplatin, was introduced into clinical trials in 1981.4 Indeed, cisplatin has already been replaced by carboplatin for the chemotherapy of a few other malignancies, such as ovarian cancer.5 In patients with advanced NSCLC, carboplatin-based chemotherapy has also been extensively investigated.6-14 A two-drug combination consisting of carboplatin plus paclitaxel has been frequently used in clinical practice as well as in clinical trials, especially in the United States.14 However, it is still unclear whether carboplatin has efficacy equivalent to that of cisplatin or not. Go et al15 reviewed reports directly comparing the effectiveness of cisplatin with that of carboplatin. In their report, carboplatin was shown to possess inferior activity to that of cisplatin in germ cell, head and neck, and esophageal cancers. Furthermore, comparisons between cisplatin and carboplatin in NSCLC have been based on limited data. Accordingly, we performed a meta-analysis to compare the effect of carboplatin-based chemotherapy with that of cisplatin-based chemotherapy on overall survival, response rate, and toxicity in patients with advanced NSCLC.
Search for Trials We searched for trials that had completed recruitment by December 31, 2001. To avoid publication bias, both published and unpublished trials were identified through a computer-based search of the PubMed database and abstracts from the past 13 conferences of the American Society of Clinical Oncology (ASCO). We searched using the following terms: "lung cancer," "chemotherapy," and "randomized controlled trial." Only references published in English were included. We also examined reference lists of original articles, review articles, and relevant books, and the Physician Data Query registry of clinical trials.
Selection of Trials
Validity Assessment
Data Abstraction All data were checked for internal consistency, and disagreements were resolved by discussion among the investigators. All principal investigators of the trials were contacted to confirm or update the published data. For response assessment, we used only trials that included patients with measurable or assessable disease, and that were analyzed with well-accepted criteria. Toxicity profiles were reported according to the WHO's criteria or the cooperative groups' criteria.
Quantitative Data Synthesis A hazard ratio (HR) was calculated to assess the survival advantage of the carboplatin-based chemotherapy as compared with the cisplatin-based chemotherapy. The crude log HR value and its variance in each trial were calculated using the abstracted survival probabilities in the Kaplan-Meier curve at specific time points according to the methods proposed by Parmar et al.18 Minimum and maximum follow-up times were used to estimate censored subjects under the assumption that censoring happens constantly throughout follow-up. If the minimum follow-up time was not available, time zero was substituted for it. As we assumed constant hazard for the two types of therapy within an individual trial, all the survival probabilities available in each trial were used to obtain a representative HR for each trial instead of limiting time points to specified times. HRs were calculated to show how many times higher the probability of death from any cause was in patients receiving a carboplatin-based chemotherapy as compared with those receiving a cisplatin-based chemotherapy. Therefore, an HR greater than unity indicates that the cisplatin-based chemotherapy is better than the carboplatin-based chemotherapy.
A general variance-based method was used to estimate the summary HR, ORs, and their 95% CIs. We looked for heterogeneity among the trials based on standard methods.19 We also calculated the between-study variation ( All statistical analyses were conducted with STATA version 8 software (College Station, TX). We defined a statistical test with a P value less than .05 as significant.
Trial Flow The flow chart of our study is shown in Figure 1. One of the nine trials retrieved for more detailed evaluation compared cisplatin plus tirapazamine with carboplatin plus tirapazamine.23 Since tirapazamine is not a cytotoxic agent and the effectiveness of tirapazamine for advanced NSCLC has not been determined,24 we excluded this trial from our analysis. Thus, eight trials involving 2,948 patients with advanced NSCLC were ultimately analyzed.6-13
Characteristics of the Eight Trials Baseline characteristics of the eight trials are listed in Table 1. In total, 2,948 patients were randomly assigned to cisplatin-based chemotherapy (1,478 patients) or carboplatin-based chemotherapy (1,470 patients). Patients were stratified by four variables in four trials, by three in one, and by two in three. Clinical stage was used for stratification in all trials.
Of the eight trials, seven were randomized phase III trials,6-11,13 and the remaining one was a randomized phase II trial.12 There was no placebo-controlled double-blind trial. Each of the eight trials was reported in a full paper. Old chemotherapy regimens including etoposide, vindesine, mitomycin C, and vinblastine were investigated in three trials,6,7,14 whereas new chemotherapy regimens including paclitaxel, gemcitabine, and docetaxel were investigated in five trials.8-12 We assessed the quality of the eight trials using the three question instrument reported by Jadad et al.16 There was a statement regarding both randomization and withdrawals in reports on all eight trials, whereas none of the trials were described as double-blind. Therefore, we assigned two points for all trials and judged that study quality was not a source of heterogeneity. Other potential sources of heterogeneity, including use of a new agent as a combination drug, were examined by meta-regression analysis. However, we detected no significant factor.
Response
Overall Survival Data on overall survival were available for all eight trials (2,903 patients; Table 3). Survival analyses were carried out based on intention-to-treat analysis in four trials, whereas 11, five, 24, and five patients, respectively, in the trials reported by Klastersky et al,6 Jelic et al,7 Schiller et al,8 and Mazzanti et al,12 had been excluded from the survival analysis. The most common reason for the exclusion of patients from survival analysis was incorrect clinical stage. Cisplatin-based chemotherapy was associated with only a 5% improvement in overall survival as compared with carboplatin-based chemotherapy, and this difference was not statistically significant (HR, 1.050; 95% CI, 0.907 to 1.216; P = .515; Fig 2). A funnel plot and rank correlation test regarding survival confirmed the absence of publication bias (Z = 0.37; P = .71). On the other hand, subset analysis of the five trials8-12 revealed that the combination chemotherapy consisting of cisplatin plus a new agent yielded an 11% superior survival as compared with that of carboplatin plus a new agent. This difference was statistically significant (HR, 1.106; 95% CI, 1.005 to 1.218; P = .039; Fig 3).
Toxicity Eight trials including 2,899 patients provided toxicity profile results. Complete data for neutropenia were not obtained in two trials6,13 and those for nephrotoxicity were not available in one trial.9 Cisplatin-based chemotherapy frequently led to grade 3 or more of nausea and vomiting (OR, 2.51; 95% CI, 1.76 to 3.56), while grade 3 or greater thrombocytopenia was significantly more frequent with carboplatin-based chemotherapy (OR, 0.58; 95% CI, 0.39 to 0.87). The risk of grade 3 or greater neutropenia and grade 3 or greater nephrotoxicity was almost comparable between the two modalities (OR, 0.94; 95% CI, 0.66 to 1.35 and OR, 2.82; 95% CI, 0.88 to 9.05, respectively). No significant difference in the number of treatment-related deaths was observed between the two modalities; there were 54 treatment-related deaths (3.9%) among 1,380 patients treated with cisplatin-based chemotherapy and 40 (2.9%) among 1,366 patients treated with carboplatin-based chemotherapy. This represents a 1.4-fold increase in the risk of treatment-related death in patients receiving cisplatin-based chemotherapy, but this difference was not statistically significant (OR, 1.36; 95% CI, 0.89 to 2.07). Similar results were obtained for subgroup analysis of the five trials that investigated the two-drug combinations of platinum plus a new agent.8-12
In the present meta-analysis, we failed to demonstrate that cisplatin-based chemotherapy produces a significant survival advantage as compared with carboplatin-based chemotherapy in patients with advanced NSCLC. Then, we further analyzed the regimens containing platinum plus a new agent, because the combination chemotherapy regimens consisting of platinum plus etoposide,6 mitomycin C and vindesine,7 or mitomycin C and vinblastine13 are outdated, as defined in the ASCO guidelines.17 In this second analysis, we demonstrated that combination chemotherapy consisting of cisplatin plus a new agent yields a significant survival benefit compared with that of carboplatin plus a new agent. These results suggest that cisplatin has a possible advantage in the treatment of advanced NSCLC compared with carboplatin, if platinum is combined with a new agent. Physicians should carefully interpret these results when they apply them in clinical practice because toxicity profiles were quite different between the two modalities. Because carboplatin-based chemotherapy frequently led to thrombocytopenia, only patients with adequate hematologic function should be treated with carboplatin-based chemotherapy. On the other hand, only patients with sufficient renal function should be allowed to receive cisplatin-based chemotherapy since severe nephrotoxicity was observed in patients receiving cisplatin-based chemotherapy, though only patients with adequate renal function were accrued in this meta-analysis. We also note that patients receiving cisplatin-based chemotherapy developed nausea and vomiting more frequently, which might lead to a deterioration in quality of life (QOL). Because the primary role of chemotherapy in patients with advanced NSCLC is palliative, the influence on patients' QOL is an important issue in determining the true value of new therapy. However, formal QOL assessments were performed in only three of the eight trials.10,11,13 Additionally, the compliance for QOL assessment was generally poor. In the report by Fossella et al,11 only 926 (76%) of the 1,218 accrued patients were assessed for QOL. Accordingly, further studies will be necessary to assess any difference in QOL between the two modalities. Several technical issues have to be mentioned regarding this meta-analysis. One major limitation is the data source we used. Analyses were based on abstracted data and not on individual patient data (IPD). In general, an IPD-based meta-analysis would give a more robust estimation for the association,25 therefore, one needs to interpret our results with care, especially for a positive association in a subgroup analysis. Clearly, further investigation using IPD should be conducted to examine main effects as well as other end points, such as interaction between subgroups and main effect. Publication bias is a significant threat to the validity of meta-analysis. Although we detected no evidence of publication bias using graphical and statistical methods, it is difficult to completely rule out this possibility. Heterogeneity among trials can be another limitation of our meta-analysis, although we applied a random-effect model that takes possible heterogeneity into consideration. The absence of a statistically significant difference in the meta-regression analysis we used to examine heterogeneity may justify the analysis. However, as the number of trials was limited, careful interpretation of heterogeneity is necessary. Regarding HR estimation, we applied the Kaplan-Meier curve-based method which has substantially good correlations with alternative methods.18 We did not find any statistical inconsistencies between results in the original report and in the HR analysis that we did. Therefore, we can say that the overall HR results we obtained in this study are valid. In conclusion, we demonstrated that combination chemotherapy consisting of cisplatin plus a new agent produces a significant survival advantage compared with that of carboplatin plus the same new agent in patients with advanced NSCLC, although we failed to demonstrate a survival advantage in an overall analysis that included both new agents and old agents. Although our conclusions should be interpreted cautiously, our results nevertheless raise a critical point regarding the long-standing debate on whether cisplatin-based chemotherapy or carboplatin-based chemotherapy is superior for advanced NSCLC. Further evaluation regarding this issue is now strongly needed.
The authors indicated no potential conflicts of interest.
We wish to thank Dr Jean Klastersky and Dr Marianne Paesmans, Dr Svetislav Jelic, Dr Joan Schiller, Dr Ronghui Xu, Dr Rafael Rosell, Dr Frank Fossella, Dr Paola Mazzanti, and Dr Adriano Paccagnella for their support, data provision, and comments on our analyses.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Stanley K, Stjernsward J: Lung cancer—A worldwide health problem. Chest 96:1S-5S, 1989
2. Chemotherapy in non-small cell lung cancer: A meta-analysis using updated data on individual patients from 52 randomized clinical trials. Non-Small Cell Lung Cancer Collaborative Group. BMJ 311:899-909, 1995
3. Raby B, Pater J, Mackillop WJ: Does knowledge guide practice? Another look at the management of non-small-cell lung cancer. J Clin Oncol 13:1904-1911, 1995 4. Muggia FM: Overview of carboplatin: Replacing, complementing, and extending the therapeutic horizons of cisplatin. Semin Oncol 16:7-13, 1989 (2 Suppl 5)[Medline] 5. Bunn PA Jr: Chemotherapy for advanced non-small-cell lung cancer: Who, what, when, why? J Clin Oncol 20:23S-33S, 2002 (18 suppl) 6. Klastersky J, Sculier JP, Lacroix H, et al: A randomized study comparing cisplatin or carboplatin with etoposide in patients with advanced non-small-cell lung cancer: European Organization for Research and Treatment of Cancer Protocol 07861. J Clin Oncol 8:1556-1562, 1990[Abstract] 7. Jelic S, Mitrovic L, Radosavljevic D, et al: Survival advantage for carboplatin substituting cisplatin in combination with vindesine and mitomycin C for stage IIIB and IV squamous-cell bronchogenic carcinoma: A randomized phase III study. Lung Cancer 34:1-13, 2001[Medline]
8. Schiller JH, Harrington D, Belani CP, et al: Comparison of four chemotherapy regimens for advanced non-small-cell lung cancer. N Engl J Med 346:92-98, 2002 9. Zatloukal P, Petruzelka L, Zemanova M, et al: Gemcitabine plus cisplatin vs gemcitabine plus carboplatin in stage IIIb and IV non-small cell lung cancer: A phase III randomized trial. Lung Cancer 41:321-331, 2003[CrossRef][Medline]
10. Rosell R, Gatzemeier U, Betticher DC, et al: Phase III randomised trial comparing paclitaxel/carboplatin with paclitaxel/cisplatin in patients with advanced non-small-cell lung cancer: A cooperative multinational trial. Ann Oncol 13:1539-1549, 2002
11. Fossella F, Pereira JR, von Pawel J, et al: Randomized, multinational, phase III study of docetaxel plus platinum combination versus vinorelbine plus cisplatin for advanced non-small-cell lung cancer: The TAX326 Study Group. J Clin Oncol 21:3016-3024, 2003 12. Mazzanti P, Massacesi C, Rocchi MB, et al: Randomized, multicenter, phase II study of gemcitabine plus cisplatin versus gemcitabine plus carboplatin in patients with advanced non-small cell lung cancer. Lung Cancer 41:81-89, 2003[Medline] 13. Paccagnella A, Favaretto A, Oniga F, et al: Cisplatin versus carboplatin in combination with mitomycin and vinblastine in advanced non small cell lung cancer. A multicenter, randomized phase III trial. Lung Cancer 43:83-91, 2004[CrossRef][Medline]
14. Kelly K, Crowley J, Bunn PA Jr, et al: Randomized phase III trial of paclitaxel plus carboplatin versus vinorelbine plus cisplatin in the treatment of patients with advanced non-small-cell lung cancer: A Southwest Oncology Group Trial. J Clin Oncol 19:3210-3218, 2001
15. Go RS, Adjei AA: Review of the comparative pharmacology and clinical activity of cisplatin and carboplatin. J Clin Oncol 17:409-422, 1999 16. Jadad AR, Moore RA, Carroll D, et al: Assessing the quality of reports of randomized clinical trials: Is blinding necessary? Control Clin Trials 17:1-12, 1996[CrossRef][Medline]
17. Pfiste DG, Johnson DH, Azzoli CG, et al: American Society of Clinical Oncology Treatment of Unresectable Non–Small-Cell Lung Cancer Guideline: Update 2003. J Clin Oncol 22:330-353, 2004 18. Parmar MB, Torri V, Stewart L: Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints. Stat Med 17:2815-2834, 1998[CrossRef][Medline] 19. Higgins JP, Thompson SG: Quantifying heterogeneity in a meta-analysis. Stat Med 21:1539-1558, 2002[CrossRef][Medline] 20. DerSimonian R, Laird N: Meta-analysis in clinical trials. Control Clin Trials 7:177-188, 1986[CrossRef][Medline] 21. Begg CB, Mazumdar M: Operating characteristics of a rank correlation test for publication bias. Biometrics 50:1088-1101, 1994[CrossRef][Medline]
22. Egger M, Smith GD, Schneider M, et al: Bias in meta-analysis detected by a simple, graphical test. BMJ 315:629-634, 1997 23. Bissett D, Nicolson M, Falk S, et al: Randomised phase 2 study of tirapazamine with either cisplatin or carboplatin in advanced NSCLC. Proc Am Soc Clin Oncol 20:346a, 2001 (abstr 1381) 24. Williamson SK, Crowley JJ, Lara PN, et al: S0003: Paclitaxel/carboplatin (PC) v PC + tirapazamine (PCT) in advanced non-small cell lung cancer (NSCLC). A phase III Southwest Oncology Group (SWOG) Trial. Proc Am Soc Clin Oncol 22:622, 2003 (abstr 2502) 25. Clarke MJ, Stewart LA: Obtaining individual patient data from randomized controlled trials, in Egger M, Smith GD, and Altman DG (eds): Systematic Reviews in Health Care. London, UK, BMJ Publishing Group, 2001, pp 109-121 Submitted February 17, 2004; accepted April 29, 2004.
Related Article
Related Editorial
Related Correspondence
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
2) from the Q statistic according to the method described by DerSimonian and Laird.
