Platinum-Based Versus Non-Platinum-Based Chemotherapy in Advanced Non-Small-Cell Lung Cancer: A Meta-Analysis of the Published Literature
http://www.100md.com
《临床肿瘤学》
the Kantonsspital St Gallen, Departement Innere Medizin, Fachbereich Onkologie-Haematologie, St Gallen, Switzerland
Princess Margaret Hospital, Toronto, Ontario, Canada
ABSTRACT
PURPOSE: This meta-analysis was performed to compare the activity, efficacy and toxicity of platinum-based versus non-platinum-based chemotherapy in patients with advanced non-small-cell lung cancer.
METHODS: Randomized phase II and III clinical trials comparing first-line palliative platinum-based chemotherapy with the same regimen without platinum or with platinum replaced by a nonplatinum agent were identified by electronic searches of Medline, Embase, and Cancerlit, and hand searches of relevant abstract books and reference lists. Response rates, 1-year survival, and toxicity were analyzed. Subgroups of trials using third-generation agents were compared.
RESULTS: Thirty-seven assessable trials were identified including 7,633 patients. A 62% increase in the odds ratio (OR) for response was attributable to platinum-based therapy (OR, 1.62; 95% CI, 1.46 to 1.8; P < .0001). The 1-year survival rate was increased by 5% with platinum-based regimens (34% v 29%; OR, 1.21; 95% CI, 1.09 to 1.35; P = .0003). No statistically significant increase in 1-year survival was found when platinum therapies were compared to third-generation-based combination regimens (OR, 1.11; 95% CI, 0.96 to 1.28; P = .17). The toxicity of platinum-based regimens was significantly higher for hematologic toxicity, nephrotoxicity, and nausea and vomiting, but not for neurotoxicity, febrile neutropenia rate, or toxic death rate.
CONCLUSION: Response is significantly higher with platinum-containing regimens. One-year survival was not significantly prolonged when platinum-based therapies were compared with third-generation-based combination regimens. Toxicity is generally higher for platinum-based regimens.
INTRODUCTION
Non-small-cell lung cancer (NSCLC) accounts for nearly 80% of all lung cancers, with 1.2 million new cases diagnosed worldwide each year.1 The 5-year survival rate for these patients is approximately 15%, and lung cancer remains the leading cause of cancer-related mortality in the western world. Most patients present with inoperable advanced stage III and IV disease, and many of these patients are candidates for palliative chemotherapy.
A meta-analysis of 52 randomized trials using individual patient data published in 1995 confirmed the value of palliative chemotherapy.2 In 11 randomized trials that compared palliative systemic therapy with best supportive care, a 10% improvement of 1-year survival was found. This benefit was limited to platinum-containing regimens. Other analyses confirmed benefits related to symptom control and quality of life,3 as well as the cost effectiveness of chemotherapy in advanced NSCLC.4 Current international guidelines recommend the use of platinum-based chemotherapy for patients with good performance status,5 and the use of doublets including a platinum plus a third-generation agent (ie, gemcitabine, vinorelbine, a taxane, or camptothecin) has been widely adopted. The meta-analysis of the NSCLC collaborative group was based on regimens containing second-generation agents, and the period of recruitment of the trials analyzed was 1970 to 1988. Therefore, its comparability to regimens in use today must be questioned. Furthermore, platinum compounds, cisplatin in particular, are associated with considerable toxicity that may lead to reluctance on the part of both physicians and patients to accept chemotherapy for incurable NSCLC. Cisplatin may result in neurotoxicity, nephrotoxicity, and severe emesis, and the need for additional hydration complicates the administration of this agent.
Third-generation agents are, in general, better tolerated than their first- and second-generation predecessors, and they have shown single-agent activity that is equal to or greater than cisplatin. The activity and tolerability of these agents led many investigators to evaluate doublet combinations of third-generation agents in the hope that platinum analogs could be eliminated from the treatment of advanced NSCLC. Phase II trials of these novel regimens appeared promising, and so randomized trials comparing them with traditional platinum-based regimens were undertaken by several investigators and cooperative groups. We undertook this meta-analysis to compare response rates, survival, and toxicity of platinum-based chemotherapy with nonplatinum regimens in advanced NSCLC.
METHODS
Eligibility Criteria
To be included in this analysis, trials had to compare a platinum-containing regimen with the same regimen either without platinum or with platinum replaced by a nonplatinum compound. Randomized phase II and III studies were eligible if fully published or available as abstracts. Eligible trials had to address first-line palliative chemotherapy in inoperable advanced stage III and/or stage IV NSCLC, and had to use marketed chemotherapeutic agents of proven efficacy. Platinum was defined as cisplatin or carboplatin.
Identification of Trials
In October 2002, an electronic search of Medline, 1966 to Oct 2002; Cancerlit, mid-1960s to Oct 2002; Embase, 1966 to Oct 2002; and the Cochrane Controlled Trial Register (CCTR) of the third quarter of 2002 was performed. The search terms "NSCLC," "non-small-cell lung cancer," and "Carcinoma, Non-Small-Cell Lung" (medical subject heading MESH) were used, and the searches were limited initially to English publications of randomized controlled trials in humans. The results were then hand searched for eligible trials. We also performed a hand search of the abstract books of the American Society of Clinical Oncology (ASCO), the European Society for Medical Oncology (ESMO), and the European Conference of Clinical Oncology (ECCO), from 1998 until 2002, and the International Association for the Study of Lung Cancer (IASLC) 1997 and 2000. In addition, reference lists of review articles on the treatment of NSCLC, identified by electronic search of Medline and limited to the last 5 years, and of relevant books and guidelines were hand searched. Results were double-checked and arbitrated by a second investigator (F.S.). The authors of all identified trials (or principal investigators or group statisticians where applicable) were asked for data confirmation and/or updates by e-mail and mail.
Data Collection
The following information was collected from the identified trials: year of publication, form of publication (full/abstract); number of patients assessable for response; response rate (complete response plus partial response); number of patients assessable for survival; median survival time with lower and upper 95% CI; 1-year survival with SE, hazard ratio for platinum versus nonplatinum arm (survival), and SE for hazard ratio; number of patients assessable for toxicity; percentage of patients with WHO grade 3 and 4 toxicities combined for hemoglobin, neutrophils, platelets, nephrotoxicity, neurotoxicity (combined sensory and motor), nausea, and vomiting; febrile neutropenia rate; and toxic death rate. Whenever possible, particularly for studies published only in abstract form, updated or confirmed results obtained from the authors were used; if no update was obtained, published data were used.
Statistical Analysis
The analysis was done by pair-wise comparisons of the platinum-containing arms of the identified trials with the respective nonplatinum arms. Among the studies used there were four three-arm trials, which were entered twice in the database with the platinum arm compared individually with each of the other two treatments. In this way, two platinum arms and two nonplatinum arms appear twice in the analysis. To avoid an increased influence of these trials on the overall result, the number of patients in the arm, which was used twice, was divided by two. The only four-arm trial in the meta-analysis compared two regimens at two different doses; therefore, each platinum arm was paired to its natural counterpart defined by the matching treatment doses.
The odds ratios (ORs) for response to treatment, for the different types of toxicity, and for survival at 1 year were Peto ORs calculated based on Statistical Methods programmed in MetaView, version 4, prepared by Jon Deeks on behalf of the Statistical Methods Working Group of the Cochrane Collaboration in December 1999. An OR > 1 reflects a favorable outcome in the platinum arm. Response definition was adopted from the analyzed studies. Stable disease was not considered as response. To be able to use the same methodology for survival at 1 year, we assumed that the follow-up was complete for the first year. We also tested the heterogeneity of the studies, ie, whether the effect size was different between studies. Where the test for heterogeneity was significant, we compared the two treatment groups using the DerSimonian and Laird random effects model. In general, the analyses based on Peto ORs and the analyses based on DerSimonian-Laird were concordant, except for neutropenia and nausea and vomiting for the subgroup analysis pertaining third-generation combination regimens.
Toxicity reports were inconsistent. All toxicities were therefore analyzed in the trials stating the respective toxicities in a per-patient fashion; trials not reporting on the respective toxic effect or reporting in a different way (eg, reporting as toxicities per treatment cycle) were excluded from the analysis of the toxic effect.
The assumption of 1-year complete follow-up may seem unreasonable; therefore, we have also analyzed the survival at 6 months. The estimates for the survival at 6 months were based on survival at 1 year and on the assumption that time to death follows an exponential distribution.
RESULTS
Thirty-eight randomized trials meeting the inclusion criteria were identified,6–43 all in the English literature. One trial was excluded from the analysis because of missing data that could not be obtained through requests to the author.43 The remaining 37 trials reported at least on response rates and were included in the analysis. Four were three-arm trials and one was a four-arm trial; therefore, 42 pair-wise comparisons were possible. The evaluated trials included 7,633 patients, 3,770 treated with platinum-based regimens and 3,863 treated with nonplatinum regimens (Table 1) .
At the time of analysis, 20 of the trials were fully published, 6-8,10-17,19-22,28,33,35,37,41 16 were published in abstract form, 18,23-27,29-32,34,36,38-40,42 and one was published as a brief communication.9 For 14 trials, an update or data confirmation was obtained from the authors.16,19,23,27,28,29,32,33,35–40 New information mainly concerned originally missing data or confirmation of published data. In two cases, response information was changed,23,27 and in three cases survival information was changed.23,33,25 For the 23 trials for which no update was received, the data were used as originally published. Because of incomplete reporting of survival and toxicity, these analyses are based on lower numbers of trials.
In 17 of the analyzed pairs, platinum-containing combinations were compared with single agents, nine of them third-generation agents. None used platinum as the single agent. Twenty-five pairs incorporated a third-generation agent in the nonplatinum arm (9 third-generation single agents and 16 third-generation-containing combinations), and 14 combinations consisted of third-generation agents only.
Response
Response rate was stated in all 37 trials analyzed, which included 7,633 patients. A significant 62% increase in the OR for response was attributable to platinum-based chemotherapy compared with nonplatinum regimens (OR, 1.62; 95% CI, 1.46 to 1.8; P < .0001; Fig 1). The subgroup analysis of platinum-based regimens compared with single-agent nonplatinum compounds (16 trials including 3,221 patients) also favored platinum compounds (OR, 2.62; 95% CI, 2.22 to 3.09; P < .0001). When the analysis was restricted to trials using any third-generation agent in the nonplatinum arm, the increase in the odds of a response was 55% in favor of platinum-based regimens (23 trials including 5,451 patients; OR, 1.55; 95% CI, 1.38 to 1.75; P < .0001) and in the trials that had only third-generation combination regimens as nonplatinum therapy, the benefit was 17% in favor of platinums (14 trials including 3,204 patients; OR, 1.17; 95% CI, 1.01 to 1.36; P = .042; Fig 2).
Survival
Information on 1-year survival was available for 30 trials including 6,504 patients. Overall, the 1-year survival for the platinum-containing regimens was 34% (95% CI, 33% to 36%), and 29% (95% CI, 27% to 30%) for the nonplatinum therapies. This corresponded to a 21% increase in OR in favor of platinum-based regimens (OR, 1.21; 95% CI, 1.09 to 1.35; P = .0003; Fig 3). This effect was also statistically significant for the comparison of platinum-containing regimens versus single-agent nonplatinum therapies (11 trials including 2,616 patients; survival at 1 year, 35% [95% CI, 33 to 37] v 25% [95% CI, 23 to 27]; OR, 1.38; 95% CI, 1.17 to 1.63; P = .0001) and for platinum regimens versus third-generation-based nonplatinum regimens (20 trials including 5,191 patients; survival at 1 year, 37% [95% CI, 36 to 39] v 31% [95% CI, 29 to 32]; OR, 1.18; 95% CI, 1.05 to 1.32; P = .0057). However, when single-agent trials were excluded and platinum-based therapies were compared with third-generation-based combination regimens only, no statistically significant difference could be found (14 trials including 3,307 patients; 1-year survival, 36% [95% CI, 34 to 38] for platinum regimens v 35% [95% CI, 33 to 38] for nonplatinum regimens; OR, 1.11; 95% CI, 0.96 to 1.28; P = .17; Fig 4).
The results of the survival analysis were changed neither by the updates obtained by the study authors nor by the 6-month survival analysis (data not shown).
Toxicity
The adverse effects of chemotherapy were analyzed as WHO grade 3 and 4 toxicity combined, and as the percentage of treated patients who experienced such toxicity. Considerable variability in the completeness of toxicity reporting was found among the studies. Overall, platinum-based therapy was associated with a significant (approximately two-fold) increase in the OR for anemia (14 trials including 3,322 patients; OR, 1.99; 95% CI, 1.53 to 2.6; P < .0001) and neutropenia (22 trials including 5,280 patients; OR, 1.99; 95% CI, 1.76 to 2.26; P < .0001) and an approximately three-fold increase in the OR for thrombocytopenia (19 trials including 4,903 patients; OR, 3.56; 95% CI, 2.89 to 4.38; P < .0001), nephrotoxicity (18 trials including 4,384 patients; OR, 3.09; 95% CI, 1.88 to 5.06; P < .0001), and nausea/vomiting (21 trials including 5,358 patients; OR, 3.42; 95% CI, 2.86 to 4.08; P < .0001). No statistically significant difference was found for neurotoxicity, febrile neutropenia, and toxic death rate. However, it should be noted that the toxic death rate, although lacking statistical significance, was two-fold higher with platinum-based therapy (13 trials including 3,111 patients; OR, 2.03; 95% CI, 0.91 to 4.53; P = .082).
When the toxicity of platinum-based regimens was compared with third-generation-based combination regimens, a statistically significant increase with platinum-based therapy was seen in anemia, neutropenia, thrombocytopenia, nausea and vomiting, and toxic death rate, but not in nephrotoxicity, neurotoxicity, or febrile neutropenia rate (Table 2). In this analysis, the heterogeneity P value was significant for neutropenia (< .0001), thrombocytopenia (< .0001), and nausea and vomiting (.0085). DerSimonian-Laird analysis gave nonsignificant P values for neutropenia and nausea and vomiting, whereas the difference in thrombocytopenia remained significant.
DISCUSSION
Historically, platinum compounds have been the backbone of all treatment regimens in the chemotherapy of NSCLC. In the advanced disease setting, platinum-based regimens were the first to improve survival as well as quality of life when compared with best supportive care.44 Platinum-based combination therapy represents a standard of care.5 Currently in most developed countries, platinum compounds usually are combined with third-generation agents, and second-generation agents are no longer commonly used. However, the important meta-analysis from the NSCLC collaborative group that showed conclusively that chemotherapy for NSCLC improved survival over best supportive care was published in 1995, and did not include studies of most of the new and more active third-generation agents.2 Despite the clear survival benefit demonstrated by the meta-analysis, there continued to be reluctance on the part of both treating physicians and patients themselves to accept the toxicity of platinum-based therapy in view of the small survival gains that could be expected and the high potential for a negative impact on quality of life.
However, since the publication of the 1995 meta-analysis, several important therapeutic developments have taken place. First, the third-generation agents (vinorelbine, gemcitabine, the taxanes, and camptothecins) were introduced and are now common doublet partners of the platinums in the treatment of NSCLC. Third-generation agents are, in general, better tolerated than platinums, and offer single-agent activity that is comparable, if not superior, to platinum agents. Secondly, the third-generation agents have even demonstrated a survival advantage in the second-line setting, and such treatment is now an established option for patients with advanced NSCLC.45 Thirdly, improved supportive-care medications are better able to control or prevent severe emesis, anemia, neutropenia, and febrile neutropenia. Finally, classical outcomes such as response rates and survival times are being complemented more and more by the appreciation of symptomatic benefits and quality of life measurements. In many palliative situations, stabilization of disease with good quality of life may be of greater importance than tumor shrinkage that may be achieved only with significant treatment-related toxicity.
In 1998, Lilenbaum published a meta-analysis comparing single-agent treatment to combination chemotherapy in advanced NSCLC patients.46 Although an improvement in objective response rate was documented with combination regimens, survival was prolonged only modestly with combination therapy, and not significantly so when third-generation agents were used. The first step in the investigation of third-generation agents was to combine them with classical compounds, typically platinum agents, or to use them as single agents with the intention of reducing treatment toxicity. As a second step, several third-generation combinations were tested, and some were then compared to platinum-based regimens. Results were conflicting with some, although not all, studies, suggesting that nonplatinum regimens might be inferior.31,47 However, despite trends in favor of platinum-based therapy, no study showed a statistically significant difference.
In view of the variable results reported in the trials of platinum-based versus non-platinum-based treatment, we felt that a meta-analysis was indicated to clarify the role of third-generation nonplatinum chemotherapy for NSCLC. Our results clearly show the superiority of platinum-containing regimens in terms of objective response rate, and this superiority was found throughout the subgroup-analyses performed. In particular, platinum-regimens were more effective even when compared with third-generation regimens, and also when single-agent regimens were excluded from the analysis.
Our results also confirm that platinum-based therapy is generally associated with higher toxicity, particularly nausea and vomiting, hematologic toxicity, and nephrotoxicity. Importantly, no statistically significant increase in febrile neutropenia rate or in toxic death rate was shown. Therefore, platinum-based regimens can be administered as safely as nonplatinum therapies when patients are selected correctly. No statistically significant difference in neurotoxicity could be documented for platinum regimens. This may be due to the fact that only grade 3 and 4 toxicities were collected for this analysis. Furthermore, neurotoxicity can be caused by several nonplatinum compounds, the taxanes in particular.
Many of the nonplatinum regimens analyzed were single-agent regimens. Therefore, it could be argued that these treatments would not be expected to be equitoxic to combination regimens. Still, when subgroup analyses were restricted to modern, third-generation combinations, again the same pattern of toxicity differences persisted, favoring nonplatinum treatment.
An improvement in toxicity profile might be expected to be associated with an improvement in overall quality of life during chemotherapy. Unfortunately most of the trials in this meta-analysis did not include formal quality-of-life assessment. However in the trial reported by Gridelli et al in 2003, quality of life was the primary end point of the study, and interesting results were obtained.47 Patients treated in the nonplatinum arm had less toxicity. Conversely, they also had lower rates of symptom improvement. The net result was that there were no significant differences in overall quality of life between the platinum and nonplatinum arms.
As with response, survival analysis showed the superiority of platinum regimens. The subgroup analysis of 1-year survival rates favored platinum-based therapy even when compared with modern third-generation regimens. However, this was not true when single-agent therapies were excluded from the analysis. The trials that evaluated third-generation-containing combination regimens had 1-year survival rates that were similar to platinum-based regimens.
This meta-analysis was not based on individual patient data and was not subjected to an open external evaluation procedure. Meta-analyses based on published data tend to overestimate treatment effects compared with individual patient data analyses. However, analyses using individual patient data may include fewer studies if all authors do not agree to submit their full databases to the analyzing group. Another drawback of analyses based on individual patient data is the time-consuming review process, as shown with the meta-analysis of the NSCLC Collaborative Group.
In our analysis, the quality of data was improved by our efforts to obtain updated information from the authors of all included trials. Although the risk of publication bias exists in any meta-analysis, whether based on individual patient data or not, we feel that this was not an important aspect of our study, as many positive and negative trials were included in the analysis.
Because of inconsistent reporting of study results among the trials, some comparisons could not be performed using all 7,633 patients in the study, and were restricted to patients for whom results were available. With respect to both response and survival, we could not limit our analyses to intention-to-treat populations as the total number of patients randomized per arm was not always reported. Therefore, for consistency among studies, we elected to use the assessable patients for our analyses. Even greater variability was seen in the reporting of toxicities despite the fact that nonplatinum regimens were designed to reduce toxicity and improve quality of life. We did not identify any significant trends toward greater or lesser reporting of toxicity in trials that had small or large treatment effects (data not shown), and in almost all studies platinum-containing regimens were associated with higher toxicity rates than nonplatinum combinations. Quality-of-life results were reported so infrequently and inconsistently that it was not possible to undertake comparisons in this study.
Despite the limitation of this type of meta-analysis, these results suggest that the standard of platinum-based chemotherapy for advanced NSCLC may be challenged by modern combination regimens. Given the better tolerability of nonplatinum regimens and their similar effects on survival, modern third-generation combination regimens are valid options for patients who have concerns about toxicity and for patients for whom response in classical terms of tumor-shrinkage is not the primary goal of treatment. However, whether the added cost of purely third-generation combination regimens can be justified by their improved tolerability cannot be answered by this meta-analysis. Finally, in clinical situations in which cure is the goal of treatment, including the administration of chemotherapy as induction or adjuvant therapy, response may be a critical factor contributing to overall survival, and so it may still be prudent to include a platinum agent in the regimen.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Consultant/Advisory Role: Frances A. Shepherd, Eli Lilly, Aventis, GlaxoSmithKline. Stock Ownership: Frances A. Shepherd, Eli Lilly. Honoraria: Giannicola D'Addario, Aventis, Eli Lilly; Thomas Cerny, Aventis; Frances A. Shepherd, Eli Lilly, Aventis, GlaxoSmithKline. Research Funding: Frances A. Shepherd, Eli Lilly. Expert Testimony: Frances A. Shepherd, Eli Lilly, Aventis. For a detailed description of these categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration form and the Disclosures of Potential Conflicts of Interest section of Information for Contributors found in the front of every issue.
Acknowledgment
We thank the following investigators for supporting this meta-analysis by providing updated information on their studies (in alphabetical order): Dr F.J. Barata, Centro Hospitalar, Coimbra, Portugal; Dr R. Berardi, University of Ancona, Italy; Dr Y.M. Chen, Veterans General Hospital, Taipei, Taiwan; Dr V. Georgoulias, University General Hospital of Heraklion, Greece; Dr F.A. Greco, Sarah Cannon Cancer Center, Nashville, KY; Dr C. Gridelli, National Cancer Institute, Naples, Italy; Dr P. Kosmidis, Hygeia Hospital, Athens, Greece; Dr E. Laack, University Hospital, Eppendorf, Germany; Dr T. Le Chevalier, Institut Gustave Roussy, Villejuif, France; Dr R.C. Lilenbaum, Mount Sinai Cancer Center, Miami Beach, FL; Dr T.S. Mok, Chinese University of Hong Kong, Hong Kong; Dr M. Satouchi, West Japan Thoracic Oncology Group, Osaka, Japan; Dr M. Wiesenfeld, North Central Cancer Treatment Group, Rochester, NY.
NOTES
Published in part in abstract form and orally presented at the 10th World Conference on Lung Cancer, Vancouver, Canada, August 10-14, 2003.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Parkin DM: Global cancer statistics in the year 2000. Lancet Oncol 2:533-543, 2001
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
Bunn PA, Kelly K: New chemotherapeutic agents prolong survival and improve quality of life in non-small cell lung cancer: A review of the literature and future directions. Clin Cancer Res 4:1087-1100, 1998
Jaakkimainen L, Goodwin PJ, Pater J, et al: Counting the costs of chemotherapy in a National Cancer Institute of Canada randomized trial in non-small-cell lung cancer. J Clin Oncol 8:1301-1309, 1990
American Society of Clinical Oncology: Clinical practice guidelines for the treatment of unresectable non-small-cell lung cancer. J Clin Oncol 15:2996-3018, 1997
Fuks JZ, Aisner J, Van Echo DA, et al: Randomized study of cyclophosphamide, doxorubicin, and etoposide (VP16-213) with or without cisplatinum in non-small cell lung cancer. J Clin Oncol 1:295-301, 1983
Elliott JA, Ahmedzai S, Hole D, et al: Vindesine and cisplatin combination chemotherapy compared with vindesine as a single agent in the management of non-small cell lung cancer: A randomized study. Eur J Cancer Clin Oncol 20:1025-1032, 1984
Robert F, Omura GA, Birch R, et al: Randomized phase III comparison of three doxorubicin-based chemotherapy regimens in advanced non-small cell lung cancer: A Southeastern Cancer Study Group trial. J Clin Oncol 2:391-395, 1984
Popkin JD, Hong WK, Cersosimo RJ, et al: A prospective randomized trial of combination vindesine and cisplatin versus single-agent vindesine in advanced non-small cell lung cancer. Pharmacotherapy 5:20-22, 1985
Shinkai T, Saijo N, Tominaga K, et al: Comparison of vindesine plus cisplatin or vindesine plus mitomycin in the treatment of advanced non-small cell lung cancer. Cancer Treat Rep 69:945-951, 1985
Einhorn LH, Loehrer PJ, Williams SD, et al: Random prospective study of vindesine versus vindesine plus high-dose cisplatin versus vindesine plus cisplatin plus mitomycin C in advanced non-small-cell lung cancer. J Clin Oncol 4:1037-1043, 1986
Harvey VJ, Slevin ML, Cheek SP, et al: A randomized trial comparing vindesine and cisplatinum to vindesine and methotrexate in advanced non small cell lung carcinoma. Eur J Cancer Clin Oncol 23:1615-1619, 1987
Luedke DW, Einhorn L, Omura GA, et al: Randomized comparison of two combination regimens versus minimal chemotherapy in nonsmall-cell lung cancer: A Southeastern Cancer Study Group Trial. J Clin Oncol 8:886-891, 1990
Rosso R, Salvati F, Ardizzoni A, et al: Etoposide versus etoposide plus high-dose cisplatin in the management of advanced non-small cell lung cancer: Results of a prospective randomized FONICAP trial: Italian Lung Cancer Task Force. Cancer 66:130-134, 1990
Depierre A, Chastang C, Quoix E, et al: Vinorelbine versus vinorelbine plus cisplatin in advanced non-small cell lung cancer: A randomized trial. Ann Oncol 5:37-42, 1994
Le Chevalier T, Brisgand D, Douillard JY, et al: Randomized study of vinorelbine and cisplatin versus vindesine and cisplatin versus vinorelbine alone in advanced non-small-cell lung cancer: Results of a European multicenter trial including 612 patients. J Clin Oncol 12:360-367, 1994
Brocato N, Bruno MF, Araujo CE, et al: Treatment of non-small cell lung cancer with ifosfamide (IFO)+ 4'-epiadriamycin (EPI)+platinum vs. IFO+EPI: A GETLAC Study: Grupo de Estudio y Tratamiento Latinoamericano del Cancer Study. Oncology 52:24-31, 1995
Gil Deza E, Balbiani L, Coppola F, et al: A Phase III study of navelbine (NVB) vs NVB plus cisplatin in non small cell lung cancer (NSCLC) Stage IIIB or IV. Proc Am Soc Clin Oncol 15:394, 1996 (abstr 1193)
Gridelli C, Perrone F, Palmeri S, et al: Mitomycin C plus vindesine plus etoposide (MEV) versus mitomycin C plus vindesine plus cisplatin (MVP) in stage IV non-small-cell lung cancer: A phase III multicentre randomised trial: The "Gruppo Oncologico Centro-Sud-Isole" (G.O.C.S.I.). Ann Oncol 7:821-826, 1996
Splinter TA, Sahmoud T, Festen J, et al: Two schedules of teniposide with or without cisplatin in advanced non-small-cell lung cancer: A randomized study of the European Organization for Research and Treatment of Cancer Lung Cancer Cooperative Group. J Clin Oncol 14:127-134, 1996
Jeremic B, Shibamoto Y, Acimovic L, et al: Prolonged administration of oral etoposide alone or with intravenous carboplatin in stage IV non-small cell lung cancer: A randomized trial. Lung Cancer 18:179-188, 1997
Mencoboni M, Lerza R, Castello G, et al: Feasibility and toxicity of combination chemotherapy with ifosfamide, vinorelbine, cisplatin versus ifosfamide, vinorelbine in patients with advanced non small cell lung cancer. Anticancer Res 17:2795-2798, 1997
Wiesenfeld M, Marks R, Grill J, et al: A randomized phase II study of topotecan (TOPA) and cisplatin (CDDP) with filgrastim (G-CSF) versus topotecan and paclitaxel (Taxol) with filgrastim in patients with advanced non-small cell lung cancer. Proc Am Soc Clin Oncol 16:488a, 1997 (abstr 1755)
Galvez C, Galmarini F: Gemcitabine/cisplatin (G/P) vs gemcitabine/vinorelbine (G/V) in the treatment of non-small cell lung cancer (NSCLC): Stage IIIB and IV. Ann Oncol 11:115, 2000 (suppl 4; abstr 526P)
Gorospe A, Cristal-Luna G, Li R, et al: Initial comparative study on the efficacy of single agent gemcitabine versus combination gemcitabine and cisplatin in Filipino patients with advanced non-small cell lung cancer. Lung Cancer 29:70, 2000 (suppl 1; abstr 225)
Komiya T, Fukuoka M, Negoro S, et al: Randomized phase III trial compared cisplatin and irinotecan (CPT-P) versus cisplatin and vindesine (VDS-P) versus irinotecan alone (CPT) in patients with advanced non-small cell lung cancer (NSCLC)—The final report. Lung Cancer 29:28, 2000 (suppl 1; abstr 88)
Berardi R, Porfiri E, Massidda B, et al: Gemcitabine (GEM) and cisplatin (PL) versus gemcitabine alone in stage IV non small cell lung cancer (NSCLC): Preliminary results of a randomized multicenter phase III study. Proc Am Soc Clin Oncol 20:347a, 2001 (abstr 1385)
Georgoulias V, Papadakis E, Alexopoulos A, et al: Greek Oncology Cooperative Group (GOCG) for Lung Cancer. Platinum-based and non-platinum-based chemotherapy in advanced non-small-cell lung cancer: A randomised multicentre trial. Lancet 357:1478-1484, 2001
Satouchi M, Takada Y, Takeda K, et al: Randomized phase II study of docetaxel (DOC) plus cisplatin (CDDP) versus DOC plus irinotecan in advanced non-small cell lung cancer (NSCLC): A West Japan Thoracic Oncology Group (WJTOG) Study. Proc Am Soc Clin Oncol 20:329a, 2001 (abstr 1312)
Sculier JP, Lafitte JJ, Lecomte J, et al: A three-arm phase III randomised trial comparing combinations of platinum derivatives, ifosfamide and/or gemcitabine in stage IV non-small-cell lung cancer. Ann Oncol 13:874-882, 2002
Van Meerbeeck JP, Smit EF, Lianes P, et al: A EORTC randomized phase III trial of three chemotherapy regimens in advanced non-small cell lung cancer (NSCLC). Proc Am Soc Clin Oncol 20:308a, 2001 (abstr 1228)
Barata FJ, Parente B, Cunha J, et al: A phase II study, multicenter, randomized of carboplatin plus gemcitabine (CG) versus vinorelbine plus gemcitabine (VG) in advanced non small cell lung cancer (NSCLC). Ann Oncol 13:143, 2002 (suppl 5; abstr 523P)
Chen YM, Perng RP, Lee YC, et al: Paclitaxel plus carboplatin, compared with paclitaxel plus gemcitabine, shows similar efficacy while more cost-effective: A randomized phase II study of combination chemotherapy against inoperable non-small-cell lung cancer previously untreated. Ann Oncol 13:108-115, 2002
Georgoulias V, Ardavanis A, Agelidou M, et al: Preliminary analysis of a multicenter phase III trial comparing docetaxel (D) versus docetaxel/cisplatin (DC) in patients with inoperable advanced and metastatic non-small cell lung cancer (NSCLC). Proc Am Soc Clin Oncol 21:291a, 2002 (abstr 1163)
Greco FA, Gray JR, Thompson DS, et al: Prospective randomized study of four novel chemotherapy regimens in patients with advanced nonsmall cell lung carcinoma: A Minnie Pearl Cancer Research Network trial. Cancer 95:1279-1285, 2002
Gridelli C, Shepherd F, Perrone F, et al: Gemvin III: A phase III study of gemcitabine plus vinorelbine (GV) compared to cisplatin plus vinorelbine or gemcitabine chemotherapy (PCT) for stage IIIb or IV non-small cell lung cancer (NSCLC): An Italo-Canadian study. Proc Am Soc Clin Oncol 21:292a, 2002 (abstr 1165)
Kosmidis P, Mylonakis N, Nicolaides C, et al: Paclitaxel plus carboplatin versus gemcitabine plus paclitaxel in advanced non-small-cell lung cancer: A phase III randomized trial. J Clin Oncol 20:3578-3585, 2002
Laack E, Dierlamm T, Dickgreber N, et al: Randomized phase III study of gemcitabine + vinorelbine (GV) versus gemcitabine + vinorelbine + cisplatin (GVP) in metastatic non-small cell lung cancer (NSCLC). Ann Oncol 13:128, 2002 (suppl 5; abstr 469O)
Lilenbaum RC, Herndon J, List M, et al: Single-agent (SA) versus combination chemotherapy (CC) in advanced non-small cell lung cancer (NSCLC): A CALGB randomized trial of efficacy, quality of life (QOL), and cost-effectiveness. Proc Am Soc Clin Oncol 21:1a, 2002 (abstr 2)
Mok TS, Zee B, Nguyen B, et al: A prospective randomized study comparing toxicity and quality of life of two gemcitabine-based regimens (with or without cisplatin) for treatment of advanced non-small-cell lung cancer. Proc Am Soc Clin Oncol 21:218b, 2002 (abstr 2692)
Sederholm C: Gemcitabine versus gemcitabine/carboplatin in advanced non-small cell lung cancer: Preliminary findings in a phase III trial of the Swedish Lung Cancer Study Group. Semin Oncol 29:50-54, 2002 (suppl 9)
Stathopoulos GP, Veslemes M, Marosis K, et al: A randomized phase III trial in inoperable and metastatic non small cell lung cancer (NSCLC). Ann Oncol 13:144, 2002 (suppl 5; abstr 526P)
Wachters FM, Van Putten JW, Kramer H, et al: Randomized multicenter phase III trial comparing gemcitabine in combination with cisplatin or epirubicine as first-line treatment in advanced non-small-cell lung cancer. Proc Am Soc Clin Oncol 21:296a, 2002 (abstr 1181)
Ginsberg RJ, Vokes EE, Rosenzweig K: Non-small cell lung cancer, in De Vita VT, Hellman S, Rosenberg SA (ed): Cancer: Principles and Practice of Oncology. 6th edition. Philadelphia, PA, Lippincott Williams and Wilkins, 2001, pp 925-982
Shepherd FA, Fossella FV, Lynch T, et al: Docetaxel (Taxotere) shows survival and quality-of-life benefits in the second-line treatment of non-small cell lung cancer: A review of two phase III trials. Semin Oncol 28:4-9, 2001 (suppl 2)
Lilenbaum RC, Langenberg P, Dickersin K: Single agent versus combination chemotherapy in patients with advanced nonsmall cell lung carcinoma: A meta-analysis of response, toxicity, and survival. Cancer 82:116-126, 1998
Gridelli C, Gallo C, Shepherd FA, et al: Gemcitabine plus vinorelbine compared with cisplatin plus vinorelbine or cisplatin plus gemcitabine for advanced non-small-cell lung cancer: A phase III trial of the Italian GEMVIN Investigators and the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 21:3025-3034, 2003(Giannicola D'Addario, Mel)
Princess Margaret Hospital, Toronto, Ontario, Canada
ABSTRACT
PURPOSE: This meta-analysis was performed to compare the activity, efficacy and toxicity of platinum-based versus non-platinum-based chemotherapy in patients with advanced non-small-cell lung cancer.
METHODS: Randomized phase II and III clinical trials comparing first-line palliative platinum-based chemotherapy with the same regimen without platinum or with platinum replaced by a nonplatinum agent were identified by electronic searches of Medline, Embase, and Cancerlit, and hand searches of relevant abstract books and reference lists. Response rates, 1-year survival, and toxicity were analyzed. Subgroups of trials using third-generation agents were compared.
RESULTS: Thirty-seven assessable trials were identified including 7,633 patients. A 62% increase in the odds ratio (OR) for response was attributable to platinum-based therapy (OR, 1.62; 95% CI, 1.46 to 1.8; P < .0001). The 1-year survival rate was increased by 5% with platinum-based regimens (34% v 29%; OR, 1.21; 95% CI, 1.09 to 1.35; P = .0003). No statistically significant increase in 1-year survival was found when platinum therapies were compared to third-generation-based combination regimens (OR, 1.11; 95% CI, 0.96 to 1.28; P = .17). The toxicity of platinum-based regimens was significantly higher for hematologic toxicity, nephrotoxicity, and nausea and vomiting, but not for neurotoxicity, febrile neutropenia rate, or toxic death rate.
CONCLUSION: Response is significantly higher with platinum-containing regimens. One-year survival was not significantly prolonged when platinum-based therapies were compared with third-generation-based combination regimens. Toxicity is generally higher for platinum-based regimens.
INTRODUCTION
Non-small-cell lung cancer (NSCLC) accounts for nearly 80% of all lung cancers, with 1.2 million new cases diagnosed worldwide each year.1 The 5-year survival rate for these patients is approximately 15%, and lung cancer remains the leading cause of cancer-related mortality in the western world. Most patients present with inoperable advanced stage III and IV disease, and many of these patients are candidates for palliative chemotherapy.
A meta-analysis of 52 randomized trials using individual patient data published in 1995 confirmed the value of palliative chemotherapy.2 In 11 randomized trials that compared palliative systemic therapy with best supportive care, a 10% improvement of 1-year survival was found. This benefit was limited to platinum-containing regimens. Other analyses confirmed benefits related to symptom control and quality of life,3 as well as the cost effectiveness of chemotherapy in advanced NSCLC.4 Current international guidelines recommend the use of platinum-based chemotherapy for patients with good performance status,5 and the use of doublets including a platinum plus a third-generation agent (ie, gemcitabine, vinorelbine, a taxane, or camptothecin) has been widely adopted. The meta-analysis of the NSCLC collaborative group was based on regimens containing second-generation agents, and the period of recruitment of the trials analyzed was 1970 to 1988. Therefore, its comparability to regimens in use today must be questioned. Furthermore, platinum compounds, cisplatin in particular, are associated with considerable toxicity that may lead to reluctance on the part of both physicians and patients to accept chemotherapy for incurable NSCLC. Cisplatin may result in neurotoxicity, nephrotoxicity, and severe emesis, and the need for additional hydration complicates the administration of this agent.
Third-generation agents are, in general, better tolerated than their first- and second-generation predecessors, and they have shown single-agent activity that is equal to or greater than cisplatin. The activity and tolerability of these agents led many investigators to evaluate doublet combinations of third-generation agents in the hope that platinum analogs could be eliminated from the treatment of advanced NSCLC. Phase II trials of these novel regimens appeared promising, and so randomized trials comparing them with traditional platinum-based regimens were undertaken by several investigators and cooperative groups. We undertook this meta-analysis to compare response rates, survival, and toxicity of platinum-based chemotherapy with nonplatinum regimens in advanced NSCLC.
METHODS
Eligibility Criteria
To be included in this analysis, trials had to compare a platinum-containing regimen with the same regimen either without platinum or with platinum replaced by a nonplatinum compound. Randomized phase II and III studies were eligible if fully published or available as abstracts. Eligible trials had to address first-line palliative chemotherapy in inoperable advanced stage III and/or stage IV NSCLC, and had to use marketed chemotherapeutic agents of proven efficacy. Platinum was defined as cisplatin or carboplatin.
Identification of Trials
In October 2002, an electronic search of Medline, 1966 to Oct 2002; Cancerlit, mid-1960s to Oct 2002; Embase, 1966 to Oct 2002; and the Cochrane Controlled Trial Register (CCTR) of the third quarter of 2002 was performed. The search terms "NSCLC," "non-small-cell lung cancer," and "Carcinoma, Non-Small-Cell Lung" (medical subject heading MESH) were used, and the searches were limited initially to English publications of randomized controlled trials in humans. The results were then hand searched for eligible trials. We also performed a hand search of the abstract books of the American Society of Clinical Oncology (ASCO), the European Society for Medical Oncology (ESMO), and the European Conference of Clinical Oncology (ECCO), from 1998 until 2002, and the International Association for the Study of Lung Cancer (IASLC) 1997 and 2000. In addition, reference lists of review articles on the treatment of NSCLC, identified by electronic search of Medline and limited to the last 5 years, and of relevant books and guidelines were hand searched. Results were double-checked and arbitrated by a second investigator (F.S.). The authors of all identified trials (or principal investigators or group statisticians where applicable) were asked for data confirmation and/or updates by e-mail and mail.
Data Collection
The following information was collected from the identified trials: year of publication, form of publication (full/abstract); number of patients assessable for response; response rate (complete response plus partial response); number of patients assessable for survival; median survival time with lower and upper 95% CI; 1-year survival with SE, hazard ratio for platinum versus nonplatinum arm (survival), and SE for hazard ratio; number of patients assessable for toxicity; percentage of patients with WHO grade 3 and 4 toxicities combined for hemoglobin, neutrophils, platelets, nephrotoxicity, neurotoxicity (combined sensory and motor), nausea, and vomiting; febrile neutropenia rate; and toxic death rate. Whenever possible, particularly for studies published only in abstract form, updated or confirmed results obtained from the authors were used; if no update was obtained, published data were used.
Statistical Analysis
The analysis was done by pair-wise comparisons of the platinum-containing arms of the identified trials with the respective nonplatinum arms. Among the studies used there were four three-arm trials, which were entered twice in the database with the platinum arm compared individually with each of the other two treatments. In this way, two platinum arms and two nonplatinum arms appear twice in the analysis. To avoid an increased influence of these trials on the overall result, the number of patients in the arm, which was used twice, was divided by two. The only four-arm trial in the meta-analysis compared two regimens at two different doses; therefore, each platinum arm was paired to its natural counterpart defined by the matching treatment doses.
The odds ratios (ORs) for response to treatment, for the different types of toxicity, and for survival at 1 year were Peto ORs calculated based on Statistical Methods programmed in MetaView, version 4, prepared by Jon Deeks on behalf of the Statistical Methods Working Group of the Cochrane Collaboration in December 1999. An OR > 1 reflects a favorable outcome in the platinum arm. Response definition was adopted from the analyzed studies. Stable disease was not considered as response. To be able to use the same methodology for survival at 1 year, we assumed that the follow-up was complete for the first year. We also tested the heterogeneity of the studies, ie, whether the effect size was different between studies. Where the test for heterogeneity was significant, we compared the two treatment groups using the DerSimonian and Laird random effects model. In general, the analyses based on Peto ORs and the analyses based on DerSimonian-Laird were concordant, except for neutropenia and nausea and vomiting for the subgroup analysis pertaining third-generation combination regimens.
Toxicity reports were inconsistent. All toxicities were therefore analyzed in the trials stating the respective toxicities in a per-patient fashion; trials not reporting on the respective toxic effect or reporting in a different way (eg, reporting as toxicities per treatment cycle) were excluded from the analysis of the toxic effect.
The assumption of 1-year complete follow-up may seem unreasonable; therefore, we have also analyzed the survival at 6 months. The estimates for the survival at 6 months were based on survival at 1 year and on the assumption that time to death follows an exponential distribution.
RESULTS
Thirty-eight randomized trials meeting the inclusion criteria were identified,6–43 all in the English literature. One trial was excluded from the analysis because of missing data that could not be obtained through requests to the author.43 The remaining 37 trials reported at least on response rates and were included in the analysis. Four were three-arm trials and one was a four-arm trial; therefore, 42 pair-wise comparisons were possible. The evaluated trials included 7,633 patients, 3,770 treated with platinum-based regimens and 3,863 treated with nonplatinum regimens (Table 1) .
At the time of analysis, 20 of the trials were fully published, 6-8,10-17,19-22,28,33,35,37,41 16 were published in abstract form, 18,23-27,29-32,34,36,38-40,42 and one was published as a brief communication.9 For 14 trials, an update or data confirmation was obtained from the authors.16,19,23,27,28,29,32,33,35–40 New information mainly concerned originally missing data or confirmation of published data. In two cases, response information was changed,23,27 and in three cases survival information was changed.23,33,25 For the 23 trials for which no update was received, the data were used as originally published. Because of incomplete reporting of survival and toxicity, these analyses are based on lower numbers of trials.
In 17 of the analyzed pairs, platinum-containing combinations were compared with single agents, nine of them third-generation agents. None used platinum as the single agent. Twenty-five pairs incorporated a third-generation agent in the nonplatinum arm (9 third-generation single agents and 16 third-generation-containing combinations), and 14 combinations consisted of third-generation agents only.
Response
Response rate was stated in all 37 trials analyzed, which included 7,633 patients. A significant 62% increase in the OR for response was attributable to platinum-based chemotherapy compared with nonplatinum regimens (OR, 1.62; 95% CI, 1.46 to 1.8; P < .0001; Fig 1). The subgroup analysis of platinum-based regimens compared with single-agent nonplatinum compounds (16 trials including 3,221 patients) also favored platinum compounds (OR, 2.62; 95% CI, 2.22 to 3.09; P < .0001). When the analysis was restricted to trials using any third-generation agent in the nonplatinum arm, the increase in the odds of a response was 55% in favor of platinum-based regimens (23 trials including 5,451 patients; OR, 1.55; 95% CI, 1.38 to 1.75; P < .0001) and in the trials that had only third-generation combination regimens as nonplatinum therapy, the benefit was 17% in favor of platinums (14 trials including 3,204 patients; OR, 1.17; 95% CI, 1.01 to 1.36; P = .042; Fig 2).
Survival
Information on 1-year survival was available for 30 trials including 6,504 patients. Overall, the 1-year survival for the platinum-containing regimens was 34% (95% CI, 33% to 36%), and 29% (95% CI, 27% to 30%) for the nonplatinum therapies. This corresponded to a 21% increase in OR in favor of platinum-based regimens (OR, 1.21; 95% CI, 1.09 to 1.35; P = .0003; Fig 3). This effect was also statistically significant for the comparison of platinum-containing regimens versus single-agent nonplatinum therapies (11 trials including 2,616 patients; survival at 1 year, 35% [95% CI, 33 to 37] v 25% [95% CI, 23 to 27]; OR, 1.38; 95% CI, 1.17 to 1.63; P = .0001) and for platinum regimens versus third-generation-based nonplatinum regimens (20 trials including 5,191 patients; survival at 1 year, 37% [95% CI, 36 to 39] v 31% [95% CI, 29 to 32]; OR, 1.18; 95% CI, 1.05 to 1.32; P = .0057). However, when single-agent trials were excluded and platinum-based therapies were compared with third-generation-based combination regimens only, no statistically significant difference could be found (14 trials including 3,307 patients; 1-year survival, 36% [95% CI, 34 to 38] for platinum regimens v 35% [95% CI, 33 to 38] for nonplatinum regimens; OR, 1.11; 95% CI, 0.96 to 1.28; P = .17; Fig 4).
The results of the survival analysis were changed neither by the updates obtained by the study authors nor by the 6-month survival analysis (data not shown).
Toxicity
The adverse effects of chemotherapy were analyzed as WHO grade 3 and 4 toxicity combined, and as the percentage of treated patients who experienced such toxicity. Considerable variability in the completeness of toxicity reporting was found among the studies. Overall, platinum-based therapy was associated with a significant (approximately two-fold) increase in the OR for anemia (14 trials including 3,322 patients; OR, 1.99; 95% CI, 1.53 to 2.6; P < .0001) and neutropenia (22 trials including 5,280 patients; OR, 1.99; 95% CI, 1.76 to 2.26; P < .0001) and an approximately three-fold increase in the OR for thrombocytopenia (19 trials including 4,903 patients; OR, 3.56; 95% CI, 2.89 to 4.38; P < .0001), nephrotoxicity (18 trials including 4,384 patients; OR, 3.09; 95% CI, 1.88 to 5.06; P < .0001), and nausea/vomiting (21 trials including 5,358 patients; OR, 3.42; 95% CI, 2.86 to 4.08; P < .0001). No statistically significant difference was found for neurotoxicity, febrile neutropenia, and toxic death rate. However, it should be noted that the toxic death rate, although lacking statistical significance, was two-fold higher with platinum-based therapy (13 trials including 3,111 patients; OR, 2.03; 95% CI, 0.91 to 4.53; P = .082).
When the toxicity of platinum-based regimens was compared with third-generation-based combination regimens, a statistically significant increase with platinum-based therapy was seen in anemia, neutropenia, thrombocytopenia, nausea and vomiting, and toxic death rate, but not in nephrotoxicity, neurotoxicity, or febrile neutropenia rate (Table 2). In this analysis, the heterogeneity P value was significant for neutropenia (< .0001), thrombocytopenia (< .0001), and nausea and vomiting (.0085). DerSimonian-Laird analysis gave nonsignificant P values for neutropenia and nausea and vomiting, whereas the difference in thrombocytopenia remained significant.
DISCUSSION
Historically, platinum compounds have been the backbone of all treatment regimens in the chemotherapy of NSCLC. In the advanced disease setting, platinum-based regimens were the first to improve survival as well as quality of life when compared with best supportive care.44 Platinum-based combination therapy represents a standard of care.5 Currently in most developed countries, platinum compounds usually are combined with third-generation agents, and second-generation agents are no longer commonly used. However, the important meta-analysis from the NSCLC collaborative group that showed conclusively that chemotherapy for NSCLC improved survival over best supportive care was published in 1995, and did not include studies of most of the new and more active third-generation agents.2 Despite the clear survival benefit demonstrated by the meta-analysis, there continued to be reluctance on the part of both treating physicians and patients themselves to accept the toxicity of platinum-based therapy in view of the small survival gains that could be expected and the high potential for a negative impact on quality of life.
However, since the publication of the 1995 meta-analysis, several important therapeutic developments have taken place. First, the third-generation agents (vinorelbine, gemcitabine, the taxanes, and camptothecins) were introduced and are now common doublet partners of the platinums in the treatment of NSCLC. Third-generation agents are, in general, better tolerated than platinums, and offer single-agent activity that is comparable, if not superior, to platinum agents. Secondly, the third-generation agents have even demonstrated a survival advantage in the second-line setting, and such treatment is now an established option for patients with advanced NSCLC.45 Thirdly, improved supportive-care medications are better able to control or prevent severe emesis, anemia, neutropenia, and febrile neutropenia. Finally, classical outcomes such as response rates and survival times are being complemented more and more by the appreciation of symptomatic benefits and quality of life measurements. In many palliative situations, stabilization of disease with good quality of life may be of greater importance than tumor shrinkage that may be achieved only with significant treatment-related toxicity.
In 1998, Lilenbaum published a meta-analysis comparing single-agent treatment to combination chemotherapy in advanced NSCLC patients.46 Although an improvement in objective response rate was documented with combination regimens, survival was prolonged only modestly with combination therapy, and not significantly so when third-generation agents were used. The first step in the investigation of third-generation agents was to combine them with classical compounds, typically platinum agents, or to use them as single agents with the intention of reducing treatment toxicity. As a second step, several third-generation combinations were tested, and some were then compared to platinum-based regimens. Results were conflicting with some, although not all, studies, suggesting that nonplatinum regimens might be inferior.31,47 However, despite trends in favor of platinum-based therapy, no study showed a statistically significant difference.
In view of the variable results reported in the trials of platinum-based versus non-platinum-based treatment, we felt that a meta-analysis was indicated to clarify the role of third-generation nonplatinum chemotherapy for NSCLC. Our results clearly show the superiority of platinum-containing regimens in terms of objective response rate, and this superiority was found throughout the subgroup-analyses performed. In particular, platinum-regimens were more effective even when compared with third-generation regimens, and also when single-agent regimens were excluded from the analysis.
Our results also confirm that platinum-based therapy is generally associated with higher toxicity, particularly nausea and vomiting, hematologic toxicity, and nephrotoxicity. Importantly, no statistically significant increase in febrile neutropenia rate or in toxic death rate was shown. Therefore, platinum-based regimens can be administered as safely as nonplatinum therapies when patients are selected correctly. No statistically significant difference in neurotoxicity could be documented for platinum regimens. This may be due to the fact that only grade 3 and 4 toxicities were collected for this analysis. Furthermore, neurotoxicity can be caused by several nonplatinum compounds, the taxanes in particular.
Many of the nonplatinum regimens analyzed were single-agent regimens. Therefore, it could be argued that these treatments would not be expected to be equitoxic to combination regimens. Still, when subgroup analyses were restricted to modern, third-generation combinations, again the same pattern of toxicity differences persisted, favoring nonplatinum treatment.
An improvement in toxicity profile might be expected to be associated with an improvement in overall quality of life during chemotherapy. Unfortunately most of the trials in this meta-analysis did not include formal quality-of-life assessment. However in the trial reported by Gridelli et al in 2003, quality of life was the primary end point of the study, and interesting results were obtained.47 Patients treated in the nonplatinum arm had less toxicity. Conversely, they also had lower rates of symptom improvement. The net result was that there were no significant differences in overall quality of life between the platinum and nonplatinum arms.
As with response, survival analysis showed the superiority of platinum regimens. The subgroup analysis of 1-year survival rates favored platinum-based therapy even when compared with modern third-generation regimens. However, this was not true when single-agent therapies were excluded from the analysis. The trials that evaluated third-generation-containing combination regimens had 1-year survival rates that were similar to platinum-based regimens.
This meta-analysis was not based on individual patient data and was not subjected to an open external evaluation procedure. Meta-analyses based on published data tend to overestimate treatment effects compared with individual patient data analyses. However, analyses using individual patient data may include fewer studies if all authors do not agree to submit their full databases to the analyzing group. Another drawback of analyses based on individual patient data is the time-consuming review process, as shown with the meta-analysis of the NSCLC Collaborative Group.
In our analysis, the quality of data was improved by our efforts to obtain updated information from the authors of all included trials. Although the risk of publication bias exists in any meta-analysis, whether based on individual patient data or not, we feel that this was not an important aspect of our study, as many positive and negative trials were included in the analysis.
Because of inconsistent reporting of study results among the trials, some comparisons could not be performed using all 7,633 patients in the study, and were restricted to patients for whom results were available. With respect to both response and survival, we could not limit our analyses to intention-to-treat populations as the total number of patients randomized per arm was not always reported. Therefore, for consistency among studies, we elected to use the assessable patients for our analyses. Even greater variability was seen in the reporting of toxicities despite the fact that nonplatinum regimens were designed to reduce toxicity and improve quality of life. We did not identify any significant trends toward greater or lesser reporting of toxicity in trials that had small or large treatment effects (data not shown), and in almost all studies platinum-containing regimens were associated with higher toxicity rates than nonplatinum combinations. Quality-of-life results were reported so infrequently and inconsistently that it was not possible to undertake comparisons in this study.
Despite the limitation of this type of meta-analysis, these results suggest that the standard of platinum-based chemotherapy for advanced NSCLC may be challenged by modern combination regimens. Given the better tolerability of nonplatinum regimens and their similar effects on survival, modern third-generation combination regimens are valid options for patients who have concerns about toxicity and for patients for whom response in classical terms of tumor-shrinkage is not the primary goal of treatment. However, whether the added cost of purely third-generation combination regimens can be justified by their improved tolerability cannot be answered by this meta-analysis. Finally, in clinical situations in which cure is the goal of treatment, including the administration of chemotherapy as induction or adjuvant therapy, response may be a critical factor contributing to overall survival, and so it may still be prudent to include a platinum agent in the regimen.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Consultant/Advisory Role: Frances A. Shepherd, Eli Lilly, Aventis, GlaxoSmithKline. Stock Ownership: Frances A. Shepherd, Eli Lilly. Honoraria: Giannicola D'Addario, Aventis, Eli Lilly; Thomas Cerny, Aventis; Frances A. Shepherd, Eli Lilly, Aventis, GlaxoSmithKline. Research Funding: Frances A. Shepherd, Eli Lilly. Expert Testimony: Frances A. Shepherd, Eli Lilly, Aventis. For a detailed description of these categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration form and the Disclosures of Potential Conflicts of Interest section of Information for Contributors found in the front of every issue.
Acknowledgment
We thank the following investigators for supporting this meta-analysis by providing updated information on their studies (in alphabetical order): Dr F.J. Barata, Centro Hospitalar, Coimbra, Portugal; Dr R. Berardi, University of Ancona, Italy; Dr Y.M. Chen, Veterans General Hospital, Taipei, Taiwan; Dr V. Georgoulias, University General Hospital of Heraklion, Greece; Dr F.A. Greco, Sarah Cannon Cancer Center, Nashville, KY; Dr C. Gridelli, National Cancer Institute, Naples, Italy; Dr P. Kosmidis, Hygeia Hospital, Athens, Greece; Dr E. Laack, University Hospital, Eppendorf, Germany; Dr T. Le Chevalier, Institut Gustave Roussy, Villejuif, France; Dr R.C. Lilenbaum, Mount Sinai Cancer Center, Miami Beach, FL; Dr T.S. Mok, Chinese University of Hong Kong, Hong Kong; Dr M. Satouchi, West Japan Thoracic Oncology Group, Osaka, Japan; Dr M. Wiesenfeld, North Central Cancer Treatment Group, Rochester, NY.
NOTES
Published in part in abstract form and orally presented at the 10th World Conference on Lung Cancer, Vancouver, Canada, August 10-14, 2003.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Parkin DM: Global cancer statistics in the year 2000. Lancet Oncol 2:533-543, 2001
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
Bunn PA, Kelly K: New chemotherapeutic agents prolong survival and improve quality of life in non-small cell lung cancer: A review of the literature and future directions. Clin Cancer Res 4:1087-1100, 1998
Jaakkimainen L, Goodwin PJ, Pater J, et al: Counting the costs of chemotherapy in a National Cancer Institute of Canada randomized trial in non-small-cell lung cancer. J Clin Oncol 8:1301-1309, 1990
American Society of Clinical Oncology: Clinical practice guidelines for the treatment of unresectable non-small-cell lung cancer. J Clin Oncol 15:2996-3018, 1997
Fuks JZ, Aisner J, Van Echo DA, et al: Randomized study of cyclophosphamide, doxorubicin, and etoposide (VP16-213) with or without cisplatinum in non-small cell lung cancer. J Clin Oncol 1:295-301, 1983
Elliott JA, Ahmedzai S, Hole D, et al: Vindesine and cisplatin combination chemotherapy compared with vindesine as a single agent in the management of non-small cell lung cancer: A randomized study. Eur J Cancer Clin Oncol 20:1025-1032, 1984
Robert F, Omura GA, Birch R, et al: Randomized phase III comparison of three doxorubicin-based chemotherapy regimens in advanced non-small cell lung cancer: A Southeastern Cancer Study Group trial. J Clin Oncol 2:391-395, 1984
Popkin JD, Hong WK, Cersosimo RJ, et al: A prospective randomized trial of combination vindesine and cisplatin versus single-agent vindesine in advanced non-small cell lung cancer. Pharmacotherapy 5:20-22, 1985
Shinkai T, Saijo N, Tominaga K, et al: Comparison of vindesine plus cisplatin or vindesine plus mitomycin in the treatment of advanced non-small cell lung cancer. Cancer Treat Rep 69:945-951, 1985
Einhorn LH, Loehrer PJ, Williams SD, et al: Random prospective study of vindesine versus vindesine plus high-dose cisplatin versus vindesine plus cisplatin plus mitomycin C in advanced non-small-cell lung cancer. J Clin Oncol 4:1037-1043, 1986
Harvey VJ, Slevin ML, Cheek SP, et al: A randomized trial comparing vindesine and cisplatinum to vindesine and methotrexate in advanced non small cell lung carcinoma. Eur J Cancer Clin Oncol 23:1615-1619, 1987
Luedke DW, Einhorn L, Omura GA, et al: Randomized comparison of two combination regimens versus minimal chemotherapy in nonsmall-cell lung cancer: A Southeastern Cancer Study Group Trial. J Clin Oncol 8:886-891, 1990
Rosso R, Salvati F, Ardizzoni A, et al: Etoposide versus etoposide plus high-dose cisplatin in the management of advanced non-small cell lung cancer: Results of a prospective randomized FONICAP trial: Italian Lung Cancer Task Force. Cancer 66:130-134, 1990
Depierre A, Chastang C, Quoix E, et al: Vinorelbine versus vinorelbine plus cisplatin in advanced non-small cell lung cancer: A randomized trial. Ann Oncol 5:37-42, 1994
Le Chevalier T, Brisgand D, Douillard JY, et al: Randomized study of vinorelbine and cisplatin versus vindesine and cisplatin versus vinorelbine alone in advanced non-small-cell lung cancer: Results of a European multicenter trial including 612 patients. J Clin Oncol 12:360-367, 1994
Brocato N, Bruno MF, Araujo CE, et al: Treatment of non-small cell lung cancer with ifosfamide (IFO)+ 4'-epiadriamycin (EPI)+platinum vs. IFO+EPI: A GETLAC Study: Grupo de Estudio y Tratamiento Latinoamericano del Cancer Study. Oncology 52:24-31, 1995
Gil Deza E, Balbiani L, Coppola F, et al: A Phase III study of navelbine (NVB) vs NVB plus cisplatin in non small cell lung cancer (NSCLC) Stage IIIB or IV. Proc Am Soc Clin Oncol 15:394, 1996 (abstr 1193)
Gridelli C, Perrone F, Palmeri S, et al: Mitomycin C plus vindesine plus etoposide (MEV) versus mitomycin C plus vindesine plus cisplatin (MVP) in stage IV non-small-cell lung cancer: A phase III multicentre randomised trial: The "Gruppo Oncologico Centro-Sud-Isole" (G.O.C.S.I.). Ann Oncol 7:821-826, 1996
Splinter TA, Sahmoud T, Festen J, et al: Two schedules of teniposide with or without cisplatin in advanced non-small-cell lung cancer: A randomized study of the European Organization for Research and Treatment of Cancer Lung Cancer Cooperative Group. J Clin Oncol 14:127-134, 1996
Jeremic B, Shibamoto Y, Acimovic L, et al: Prolonged administration of oral etoposide alone or with intravenous carboplatin in stage IV non-small cell lung cancer: A randomized trial. Lung Cancer 18:179-188, 1997
Mencoboni M, Lerza R, Castello G, et al: Feasibility and toxicity of combination chemotherapy with ifosfamide, vinorelbine, cisplatin versus ifosfamide, vinorelbine in patients with advanced non small cell lung cancer. Anticancer Res 17:2795-2798, 1997
Wiesenfeld M, Marks R, Grill J, et al: A randomized phase II study of topotecan (TOPA) and cisplatin (CDDP) with filgrastim (G-CSF) versus topotecan and paclitaxel (Taxol) with filgrastim in patients with advanced non-small cell lung cancer. Proc Am Soc Clin Oncol 16:488a, 1997 (abstr 1755)
Galvez C, Galmarini F: Gemcitabine/cisplatin (G/P) vs gemcitabine/vinorelbine (G/V) in the treatment of non-small cell lung cancer (NSCLC): Stage IIIB and IV. Ann Oncol 11:115, 2000 (suppl 4; abstr 526P)
Gorospe A, Cristal-Luna G, Li R, et al: Initial comparative study on the efficacy of single agent gemcitabine versus combination gemcitabine and cisplatin in Filipino patients with advanced non-small cell lung cancer. Lung Cancer 29:70, 2000 (suppl 1; abstr 225)
Komiya T, Fukuoka M, Negoro S, et al: Randomized phase III trial compared cisplatin and irinotecan (CPT-P) versus cisplatin and vindesine (VDS-P) versus irinotecan alone (CPT) in patients with advanced non-small cell lung cancer (NSCLC)—The final report. Lung Cancer 29:28, 2000 (suppl 1; abstr 88)
Berardi R, Porfiri E, Massidda B, et al: Gemcitabine (GEM) and cisplatin (PL) versus gemcitabine alone in stage IV non small cell lung cancer (NSCLC): Preliminary results of a randomized multicenter phase III study. Proc Am Soc Clin Oncol 20:347a, 2001 (abstr 1385)
Georgoulias V, Papadakis E, Alexopoulos A, et al: Greek Oncology Cooperative Group (GOCG) for Lung Cancer. Platinum-based and non-platinum-based chemotherapy in advanced non-small-cell lung cancer: A randomised multicentre trial. Lancet 357:1478-1484, 2001
Satouchi M, Takada Y, Takeda K, et al: Randomized phase II study of docetaxel (DOC) plus cisplatin (CDDP) versus DOC plus irinotecan in advanced non-small cell lung cancer (NSCLC): A West Japan Thoracic Oncology Group (WJTOG) Study. Proc Am Soc Clin Oncol 20:329a, 2001 (abstr 1312)
Sculier JP, Lafitte JJ, Lecomte J, et al: A three-arm phase III randomised trial comparing combinations of platinum derivatives, ifosfamide and/or gemcitabine in stage IV non-small-cell lung cancer. Ann Oncol 13:874-882, 2002
Van Meerbeeck JP, Smit EF, Lianes P, et al: A EORTC randomized phase III trial of three chemotherapy regimens in advanced non-small cell lung cancer (NSCLC). Proc Am Soc Clin Oncol 20:308a, 2001 (abstr 1228)
Barata FJ, Parente B, Cunha J, et al: A phase II study, multicenter, randomized of carboplatin plus gemcitabine (CG) versus vinorelbine plus gemcitabine (VG) in advanced non small cell lung cancer (NSCLC). Ann Oncol 13:143, 2002 (suppl 5; abstr 523P)
Chen YM, Perng RP, Lee YC, et al: Paclitaxel plus carboplatin, compared with paclitaxel plus gemcitabine, shows similar efficacy while more cost-effective: A randomized phase II study of combination chemotherapy against inoperable non-small-cell lung cancer previously untreated. Ann Oncol 13:108-115, 2002
Georgoulias V, Ardavanis A, Agelidou M, et al: Preliminary analysis of a multicenter phase III trial comparing docetaxel (D) versus docetaxel/cisplatin (DC) in patients with inoperable advanced and metastatic non-small cell lung cancer (NSCLC). Proc Am Soc Clin Oncol 21:291a, 2002 (abstr 1163)
Greco FA, Gray JR, Thompson DS, et al: Prospective randomized study of four novel chemotherapy regimens in patients with advanced nonsmall cell lung carcinoma: A Minnie Pearl Cancer Research Network trial. Cancer 95:1279-1285, 2002
Gridelli C, Shepherd F, Perrone F, et al: Gemvin III: A phase III study of gemcitabine plus vinorelbine (GV) compared to cisplatin plus vinorelbine or gemcitabine chemotherapy (PCT) for stage IIIb or IV non-small cell lung cancer (NSCLC): An Italo-Canadian study. Proc Am Soc Clin Oncol 21:292a, 2002 (abstr 1165)
Kosmidis P, Mylonakis N, Nicolaides C, et al: Paclitaxel plus carboplatin versus gemcitabine plus paclitaxel in advanced non-small-cell lung cancer: A phase III randomized trial. J Clin Oncol 20:3578-3585, 2002
Laack E, Dierlamm T, Dickgreber N, et al: Randomized phase III study of gemcitabine + vinorelbine (GV) versus gemcitabine + vinorelbine + cisplatin (GVP) in metastatic non-small cell lung cancer (NSCLC). Ann Oncol 13:128, 2002 (suppl 5; abstr 469O)
Lilenbaum RC, Herndon J, List M, et al: Single-agent (SA) versus combination chemotherapy (CC) in advanced non-small cell lung cancer (NSCLC): A CALGB randomized trial of efficacy, quality of life (QOL), and cost-effectiveness. Proc Am Soc Clin Oncol 21:1a, 2002 (abstr 2)
Mok TS, Zee B, Nguyen B, et al: A prospective randomized study comparing toxicity and quality of life of two gemcitabine-based regimens (with or without cisplatin) for treatment of advanced non-small-cell lung cancer. Proc Am Soc Clin Oncol 21:218b, 2002 (abstr 2692)
Sederholm C: Gemcitabine versus gemcitabine/carboplatin in advanced non-small cell lung cancer: Preliminary findings in a phase III trial of the Swedish Lung Cancer Study Group. Semin Oncol 29:50-54, 2002 (suppl 9)
Stathopoulos GP, Veslemes M, Marosis K, et al: A randomized phase III trial in inoperable and metastatic non small cell lung cancer (NSCLC). Ann Oncol 13:144, 2002 (suppl 5; abstr 526P)
Wachters FM, Van Putten JW, Kramer H, et al: Randomized multicenter phase III trial comparing gemcitabine in combination with cisplatin or epirubicine as first-line treatment in advanced non-small-cell lung cancer. Proc Am Soc Clin Oncol 21:296a, 2002 (abstr 1181)
Ginsberg RJ, Vokes EE, Rosenzweig K: Non-small cell lung cancer, in De Vita VT, Hellman S, Rosenberg SA (ed): Cancer: Principles and Practice of Oncology. 6th edition. Philadelphia, PA, Lippincott Williams and Wilkins, 2001, pp 925-982
Shepherd FA, Fossella FV, Lynch T, et al: Docetaxel (Taxotere) shows survival and quality-of-life benefits in the second-line treatment of non-small cell lung cancer: A review of two phase III trials. Semin Oncol 28:4-9, 2001 (suppl 2)
Lilenbaum RC, Langenberg P, Dickersin K: Single agent versus combination chemotherapy in patients with advanced nonsmall cell lung carcinoma: A meta-analysis of response, toxicity, and survival. Cancer 82:116-126, 1998
Gridelli C, Gallo C, Shepherd FA, et al: Gemcitabine plus vinorelbine compared with cisplatin plus vinorelbine or cisplatin plus gemcitabine for advanced non-small-cell lung cancer: A phase III trial of the Italian GEMVIN Investigators and the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 21:3025-3034, 2003(Giannicola D'Addario, Mel)