Gemcitabine–docetaxel versus cisplatin–vinorelbine in advanced or metastatic non-small-cell lung cancer: a phase III study addressing the case for cisplatin

J.-L. Pujol1,*, J.-L. Breton2, R. Gervais3, P. Rebattu4, A. Depierre5, J.-F. Morère6, B. Milleron7, D. Debieuvre8, D. Castéra9, P.-J. Souquet10, D. Moro-Sibilot11, E. Lemarié12, R. Kessler13, H. Janicot14, D. Braun15, D. Spaeth16, X. Quantin1 and C. Clary17

1 Montpellier University Hospital, Montpellier; 2 Belfort Hospital, Belfort; 3 Caen Cancer Institute, Caen; 4 Lyon Cancer Institute, Lyon; 5 Besançon University Hospital, Besançon; 6 Bobigny Avicenne University Hospital, Bobigny; 7 Paris Tenon University Hospital, Paris; 8 Vesoul Hospital, Vesoul; 9 Perpignan Saint Pierre Institute, Perpignan; 10 Hospices Civils de Lyon, Lyon; 11 Grenoble University Hospital, Grenoble; 12 Tours University Hospital, Tours; 13 Strasbourg Hautepierre University Hospital, Strasbourg; 14 Clermont Ferrand University Hospital, Clermont Ferrand; 15 Briey Hospital, Briey; 16 Nancy Cancer Institute, Nancy; 17 Nice University Hospital, Nice, France

* Correspondence to: Dr J.-L. Pujol, Montpellier Academic Hospital, Hôpital Arnaud de Villeneuve, 34295 Montpellier Cedex 5, France. Tel/Fax: +33-4-67-33-61-36; Email: pujol{at}cyber-sante.org


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background: This multicenter, randomized, phase III study compared the efficacy, including progression-free survival (PFS), and safety of gemcitabine–docetaxel (GD) combination versus cisplatin–vinorelbine (CV) in the treatment of advanced non-small-cell lung cancer (NSCLC).

Patients and methods: Chemonaïve patients with stage IIIB or IV NSCLC were treated with GD (gemcitabine 1000 mg/m2 days 1 and 8 plus docetaxel 85 mg/m2 day 8, every 3 weeks for eight cycles) or CV (cisplatin 100 mg/m2 day 1 plus vinorelbine 30 mg/m2, days 1, 8, 15 and 22, every 4 weeks for six cycles).

Results: A total of 311 patients were enrolled (155 GD and 156 CV). Neither PFS nor overall survival differed significantly between the two arms (median PFS 4.2 and 4 months; median survival 11.1 and 9.6 months; 1-year survival 46% and 42%, for GD and CV, respectively). For the GD arm compared with the CV arm, the hazard ratio for PFS was 1.04 [95% confidence interval (CI) 0.83–1.32], and for overall survival, it was 0.90 (95% CI 0.70–1.16). Objective response rates did not differ significantly (31% for GD, 35.9% for CV). Myelosupression, emesis and frequency of febrile neutropenia were less pronounced on the GD arm, whereas fluid retention and pulmonary events were more pronounced. The CV arm experienced a higher number of serious adverse events and a lower compliance with the protocol. There was no quality of life (QoL) difference between arms. Median time to definite impairment of health-related QoL was 153 and 168 days in GD and CV arms, respectively.

Conclusions: There was no advantage in PFS with GD compared with CV; however, the CV regimen had higher rate of toxic events, mainly myelosuppression. The herein, non-platinum-containing regimen could be considered as a rational alternative to the cisplatin-based doublet.

Key words: chemotherapy, non-platinum-containing regimen, non-small-cell lung cancer


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Non-small-cell lung cancer (NSCLC) remains an important challenge for medical oncology. The majority of patients with advanced disease (stage IV or IIIB with pleural or pericardial malignant effusion) are ineligible for surgery or radiation therapy. Therefore, chemotherapy is the standard treatment option in advanced or metastatic NSCLC [1Go].

Cisplatin is considered a pivotal drug in the therapy of advanced NSCLC [2Go]. A meta-analysis performed by the Non-Small Cell Lung Cancer Collaborative Group [3Go] suggested that cisplatin-based regimens offer the best survival advantage. Because of the better outcome of patients receiving a doublet regimen compared with patients receiving a single-drug regimen, and owing to the demonstrated synergic effect of third-generation drugs such as gemcitabine [4Go] or vinorelbine [5Go] when combined with cisplatin, the doublet regimen has progressively been adopted as the standard treatment worldwide. Since 1997, the recommendation published by the American Society of Clinical Oncology (ASCO) that ‘chemotherapy given to NSCLC patients should be a platinum-based combination regimen’ has been accepted as the standard worldwide [1Go]. Cisplatin–vinorelbine (CV) is one of the most extensively studied doublets in this setting [5Go–8Go]. The literature suggests that the CV combination, the most frequently used in France at the time the study was planned [9Go], produces similar survival to other standard cisplatin-based doublets. Several phase III studies performed during the past decade used this regimen as a standard arm (e.g. [6Go, 7Go]).

Although clearly established, the survival advantage offered by a cisplatin-based doublet is probably of low magnitude. Moreover, cisplatin causes serious toxicities (e.g. nephrotoxicity, emesis, ototoxicity, peripheral neuropathy [10Go, 11Go]) that are detrimental to patients' quality of life (QoL). Therefore, current QoL-adjusted survival results obtained from cisplatin-based doublet therapy are unsatisfactory from the perspective of both patients and physicians, and the search for new regimens with better tolerability and equivalent or superior survival improvement is warranted. With the development of new cytotoxic compounds yielding unique anti-proliferative activity, there is a chance to develop a non-platinum regimen offering a better efficacy/toxicity ratio in the treatment of advanced NSCLC.

Gemcitabine, a new deoxycytidine analog, and docetaxel, a new synthetic taxane compound, are promising agents in NSCLC [12Go, 13Go]. In randomized studies comparing single-drug regimens plus best supportive care (BSC) versus BSC alone, both of these drugs have demonstrated their ability to improve patient outcome by alleviating symptoms and/or slowing tumor progression [14Go, 15Go].

In a recently published phase I–II study of the gemcitabine–docetaxel (GD) doublet, we determined an optimal combination of gemcitabine 1000 mg/m2 given on days 1 and 8 and docetaxel 85 mg/m2 given on day 8 of a 3-week cycle [16Go]. According to the protocol, neither granulocyte colony-stimulating factor (G-CSF) nor granulocyte–macrophage-CSF was allowed. This platinum-free doublet was active and well-tolerated, with only mild myelosuppression reported.

Therefore, we conducted a phase III study to compare the GD combination with the standard CV regimen in patients with advanced or metastatic NSCLC. The primary objective of the study was to compare progression-free survival (PFS); secondary objectives included evaluation of response rates, overall survival, QoL and safety on both study arms.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patient eligibility
Chemonaïve patients with histologically or cytologically confirmed stage IIIB or IV NSCLC (except bronchio–alveolar carcinoma), not amenable to surgery or curative radiotherapy, were eligible for the study. Prior surgery for recurrent disease or palliative radiotherapy for bone metastases was allowed if discontinued at least 4 weeks prior to randomization and involving <50% of bone marrow volume. Other eligibility criteria included: age 18–75 years; Karnofsky index ≥70; weight loss <10% during the previous 3 months; measurable or evaluable disease outside a previously irradiated area according to WHO recommendations [17Go]; life expectancy of at least 3 months; adequate bone marrow reserve (white blood cell count ≥3 x 109/l, neutrophils ≥2 x 109/l, platelets ≥100 x 109/l, hemoglobin ≥100 g/l); adequate renal function (creatinine <1.25x normal range and creatinine clearance ≥60 ml/min); adequate hepatic function [bilirubin ≤1x upper normal limit; alanine transaminase (ALT) or aspartate transaminase (AST) ≤1.5x normal]; and normal serum calcium.

Exclusion criteria were previous malignant disease (except well-controlled basal cell skin cancer or in situ cervix carcinoma); a history of ileus; active infection; pregnancy or breast feeding; grade 3–4 peripheral neuropathy or deafness as per National Cancer Institute Common Toxicity Criteria (NCI CTC); polysorbate 80 allergy; contra-indication to corticosteroids; and symptomatic central nervous system metastases.

Signed informed consent had to be obtained prior to randomization. The study was approved by the Montpellier University medical ethics committee and was conducted in accordance with ethical principles stated in the most recent version of the Declaration of Helsinki or the applicable guidelines on good clinical practice.

Study design and treatment
The main objective of this study was to compare the PFS between the GD and CV arms. Patients were randomly allocated to receive either GD (investigational chemotherapy) or CV (standard chemotherapy). The study was powered at 80% to detect a PFS improvement of 3 months between the two regimens in favor of the GD arm. Given a hazard ratio (HR) of 0.67 (equivalent to a median PFS of 9 months on GD and 6 months on CV) at the 5% significance level, ~195 events (progressions or deaths) were required. It was anticipated that this number of events would be reached by randomizing 300 patients. Randomization was centralized by an independent academic research institute (Institut Universitaire de la Recherche Clinique, Montpellier, France) according to computer-generated lists. A stratification by center was done.

For patients on the GD arm, gemcitabine 1000 mg/m2 (over 30 min) on days 1 and 8 plus docetaxel 85 mg/m2 (over 60 min) on day 8 (prior to gemcitabine) was delivered every 3 weeks for eight cycles. For the CV arm, a modified Le Chevalier's schedule [9Go] was used to administer cisplatin 100 mg/m2 (over 60–120 min) on day 1 plus vinorelbine 30 mg/m2 (over 10–20 min) weekly (days 1, 8, 15 and 22) every 4 weeks for six cycles, resulting in a 24-week treatment duration similar to GD arm. The injections of both combinations were administered for a total of at least 12 weeks (up to 24 weeks) or until intolerable toxicity or disease progression occurred.

Patients on the GD arm received a premedication of oral corticosteroids; patients on the CV arm received pre- and post-hydration for cisplatin and antiemetic prophylaxis, including 5-HT3 antagonists and corticosteroids. All patients received full supportive care, including G-CSF growth factors for prolonged aplasia (no prophylactic G-CSF was allowed).

Dose adjustments during a cycle were made based on weekly absolute granulocyte counts (AGC) (x109/l) and/or platelet counts (x109/l). If a dose was missed, it was given after a delay of 1 week. GD was delayed for at least 1 week and no more than 3 weeks for an AGC <1.5 or platelets <100. Vinorelbine was reduced by 25% for an AGC <1.2 or platelets <100. Vinorelbine was withheld for an AGC <1 and/or platelets <75.

Dose adjustments for subsequent cycles were made based on the toxicity seen in the previous cycle. For the GD arm, treatment was delayed for a maximum of 3 weeks for AGC <1 and platelets <100. For the CV arm, treatment was delayed for AGC <1.5 and platelets <100. Otherwise, for both arms, full doses were given unless any of the following were noted: grade 4 neutropenia lasting >1 week; grade 3–4 neutropenia associated with fever ≥38.2 °C at two readings; or grade 3 or 4 thrombocytopenia. After recovery from these events, both the day 1 and day 8 doses were given at 75% of the previous dose; if grade 4 neutropenia persisted after dose reductions, treatment was discontinued.

Both regimens reduced doses to 50% to 75% for grade 3 nonhematological toxicity (except nausea/vomiting and alopecia) and omitted doses for any grade 4 non-hematological toxicity. For bilirubin >1x upper normal values, all drugs except cisplatin were delayed. Gemcitabine, docetaxel and vinorelbine doses were held for transaminases >5x upper normal values, or alkaline phosphatase (ALP) >5x upper normal values. For ALP 1–2.5 and transaminase 2.5–5, or ALP 2.5–5 and transaminase <5, doses of docetaxel or gemcitabine were reduced by 25% and vinorelbine by 50%. For grade 3–4 peripheral neurological toxicity, doses of vinorelbine and docetaxel were omitted; if the neuropathy was confirmed after a 3-week delay, patients in the GD arm received gemcitabine only and patients in the CV arm were withdrawn. Cisplatin was discontinued for grade ≥2 perception deafness. For creatinine clearance of 40–60 ml/min, cisplatin was reduced by 50%, whereas it was omitted for creatinine clearance <40.

Baseline treatment and QoL assessments
Before and during the study, the disease status of each patient was assessed with the following procedures: medical history and physical examination (including measurements of weight), evaluation of Karnofsky performance status and QoL, tumor measurement of palpable or visual lesions, chest X-ray, radiological tests [computed tomography (CT) scan of the thorax and of the adrenals, brain CT scan if necessary, bone nuclear medicine scan if necessary (due to pain) and CT scan of the liver], full blood count and blood chemistries, vital signs, electrocardiogram and audiogram. The same assessment method used to determine disease status at baseline was used consistently throughout the study.

Conventional assessment of tumor response every other cycle would have resulted in a bias insofar as tumor response measurements would have been performed every 6 weeks in the GD group and every 8 weeks in the CV group. Therefore, patients were assessed every 6 weeks using WHO criteria [17Go] whichever the allocated group, according to an individual plan that was transmitted by fax from the methodological center at time of randomization. All analyses were performed on an intention-to-treat basis. There was no deviation from random allocation and, as stated in the results section, all patients received at least one infusion of therapy consequently ascertaining efficacy analyses. Response had to be confirmed 3–6 weeks after its first observance. After completion of the treatment program, patients were followed up every 3 months.

PFS was defined as the time from the date of randomization to the date of progression or the date of death from any cause, and was censored at the date of last follow-up for patients who were still alive without progression. Survival was defined as the time from the date of randomization to the date of death, and was censored at the date of the last follow-up for patients who were still alive.

The QoL of patients was assessed every cycle by use of European Organization for Research and Treatment of Cancer Quality of Life Questionnaire C30 (EORTC QLQ-C30) and the Lung Cancer (LC) module 13 QoL scale [18Go]. All randomized patients who had completed at least one baseline and one additional post-baseline questionnaire were qualified for analysis. Time to definite deterioration in health-related QoL, assessed by the score of global health status, was analyzed, with definite deterioration defined as: (i) impairment of global health ≥20% of the baseline score with subsequent scores ≤80% of the baseline score; or (ii) death occurring within the period lasting from the date of randomization to the end of randomly assigned chemotherapy plus 10 days. Time to definite deterioration in health-related QoL was defined as the time from the date of randomization to the first date of definite deterioration in health-related QoL, and was censored at the date of last health-related QoL assessment for patients who were alive and who did not have definite deterioration in health-related QoL [19Go].

Relative dose intensity was defined as the ratio of the actual mean dose per unit of time to the planned mean dose per unit of time [20Go]. For each patient, dose intensity was calculated taking into account the actual time of treatment plus 3 weeks on the GD arm and plus 4 weeks on the CV arm. Percentage of cumulative chemotherapy dose actually received on each treatment arm was calculated according to the following formula: (mean received cumulative dose/planned cumulative dose) x100.

All patients who received at least one dose were evaluated for safety. Toxicity was assessed at the end of each cycle according to the NCI CTC scale.

Statistical considerations
All analyses were performed on an intention-to-treat basis. Patients who received at least one infusion of therapy were included in efficacy analysis (full analysis set). PFS was estimated by using the Kaplan–Meier method [21Go], and Cox's proportional hazard model [22Go] was used to study the adjusted treatment effect. Covariates used in the Cox model included age, gender, specific organ metastatic involvement, Karnofsky index, stage of disease, center and histopathological diagnosis. Adjusted HR of treatment and 95% confidence interval (CI) were provided. Additional supporting analysis included the log-rank test.

Survival was analyzed in a similar way to PFS. Time to definite deterioration in health-related QoL was estimated using the Kaplan–Meier method. The result of the log-rank test comparing the arms was provided. Sensitivity analyses were performed with deterioration levels set at 5% and 10% (not presented).

For qualitative variables, arms were compared using the {chi}2-test or Fisher's exact test. For quantitative variables, a Wilcoxon rank sum test was performed. The statistical analysis was carried out using SAS software version 8.2 (SAS institute, Inc., Cary, NC, USA). No interim analysis was planned. Two-sided tests were provided with a type I error of 5%.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patient characteristics
From September 1999 to April 2001, 25 French institutions randomized and treated 311 patients, 155 in the GD arm and 156 in the CV arm. Patient characteristics were well balanced between the two arms with respect to several prognostic factors (Table 1). However, there was a trend towards a higher proportion of stage IV disease in the CV arm, and more patients in the CV arm had received prior palliative radiotherapy (P=0.01). Complete disease characteristics regarding metastatic profile are shown in Table 2.


View this table:
[in this window]
[in a new window]
 
Table 1. Baseline patient and disease characteristics (n=311)

 

View this table:
[in this window]
[in a new window]
 
Table 2. Description of metastatic profiles (n=311)

 
Twenty-two (14%) and 16 (10%) patients on the GD and CV arms, respectively, had at least one protocol deviation at inclusion. Main reasons for ineligibility consisted of hypercalcemia in 11 cases, insufficient hematological functions in five cases (none of them affected by neutropenia, a single patient with grade 1 thrombocytopenia, four patients with anemia), six cases of abnormal hepatic functions (AST or ALT >2-fold normal values, hyperbilirubinemia in one case) and low creatinine clearance in five cases. One patient was misdiagnosed as having stage IV disease with lung metastases and was retrospectively found to have stage IA disease with satellite infectious nodules. All patients were included in the intention-to-treat analyses.

Treatment administration
All 311 patients were treated by at least one infusion of the randomly allocated treatment. A total of 1333 cycles were delivered: 735 and 598 in GD and CV arms, respectively. The median number of chemotherapy courses received in the GD arm was significantly higher than that received in the CV arm (five and four, respectively; Wilcoxon rank sum test, P <0.001). Median durations in the treatment program, were similar inasmuch as the GD course was shorter than CV by 1 week.

The percentage of dose reduction was lower in the GD arm compared with the CV arm (8.5% versus 17%). Adverse events possibly or probably related to the study drugs led to dose reductions for 5.2% of patients in the GD arm (fluid retention in 37.5% of cases and febrile neutropenia in 37.5%) and for 10.3% of patients in the CV arm (febrile neutropenia in 75% of the cases). Adverse events possibly or probably related to the study drugs led to a delay in the treatment program for 3.9% and 5.1% of the patients in the GD and CV arms, respectively.

In the GD arm, mean relative dose intensities were 96% and 91% for gemcitabine and docetaxel, corresponding to 641.5 and 25.6 mg/m2/week of gemcitabine and docetaxel, respectively. Compliance with the protocol was higher in the GD arm compared with the CV arm in that the relative dose intensity for cisplatin was 95% but that achieved for vinorelbine was only 65%, corresponding to 23.7 and 19.5 mg/m2/week of cisplatin and vinorelbine, respectively.

We observed a statistical trend towards a lower proportion of patients in the GD arm completing the 24-week program (18.1% versus 26.3% in the GD and CV arms, respectively; {chi}2-test, P=0.08) (Table 3). However, up to 40% of the patients in the GD arm received at least six cycles (18 weeks) of chemotherapy. Disease progression was the main cause of discontinuation. Interestingly, the reasons for protocol discontinuation differed significantly between the two arms: a higher proportion of patients in the GD regimen (24.5%) discontinued chemotherapy for safety reasons compared with the CV regimen (14.1%). Reasons for discontinuation are indicated in the section on safety below.


View this table:
[in this window]
[in a new window]
 
Table 3. Main reasons for treatment discontinuation

 
Survival
Minimum follow-up time was 15 months. A total of 302 patients experienced progressive disease or death. PFS did not differ significantly between the two arms (median PFS 4.2 and 4 months for GD and CV arms, respectively; log-rank test, P=0.65) (Figure 1). In the Cox model, a poor performance status was the only significant prognostic determinant of poor PFS: 70% versus 100%; HR 2.29, 95% CI 1.40–3.74. The HR (95% CI) of progression for patients in the GD arm was 1.04 (0.83–1.32).



View larger version (18K):
[in this window]
[in a new window]
 
Figure 1. Progression-free survival by treatment arm. Gem, gemcitabine; Doc, docetaxel; Cis, cisplatin; Vin, vinorelbine; CI, confidence interval.

 
Overall survival analysis took into account 261 events. Patients in the GD arm did not prove to have a significantly longer survival than patients in the CV arm, with 1-year survival rates 46% and 42%, and median survival 11.1 and 9.6 months, respectively (log-rank test, P=0.47) (Figure 2). In the Cox model, age ≤60 years (HR 1.33; 95% CI 1.03–1.72) and poor Karnofsky index (70% versus 100%; HR 3.69; 95% CI 2.18–6.25) were significant determinants of poor survival. HR (95% CI) for patients in the GD arm was 0.90 (0.70–1.16).



View larger version (19K):
[in this window]
[in a new window]
 
Figure 2. Overall survival by treatment arm. Gem, gemcitabine; Doc, docetaxel; Cis, cisplatin; Vin, vinorelbine; CI, confidence interval.

 
Tumor response
All 311 patients were evaluated for tumor response as per the intention-to-treat analysis. Objective response rate did not significantly differ between the GD arm (31%) and the CV arm (35.9%). In the CV arm, 2.6% of the patients achieved a complete response, versus none in the GD arm (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 4. Best WHO response

 
Quality of life
A total of 125 and 114 patients in GD and CV arm, respectively, completed at least one baseline and one post-baseline questionnaire. Compliance with completion of the questionnaires reached >80%. Time to definite impairment of health-related QoL did not differ significantly between the two arms. Median time to definite impairment of health-related QoL was 153 and 168 days in GD and CV arms, respectively (Figure 3).



View larger version (15K):
[in this window]
[in a new window]
 
Figure 3. Time to definite deterioration of health-related quality of life scale by treatment arm. Gem, gemcitabine; Doc, docetaxel; Cis, cisplatin; Vin, vinorelbine.

 
Safety
Grade 3–4 laboratory toxicities are detailed in Table 5. Myelosupression was more pronounced in the CV regimen compared with the GD arm, particularly anemia and neutropenia. A total of 81 patients (52.3%) assigned to GD experienced at least one event of grade 3–4 neutropenia, compared with 130 patients (83.3%) in the CV arm (P <0.001). Severe modification of hepatic or renal functions were observed, infrequently and consisted of laboratory toxicity only.


View this table:
[in this window]
[in a new window]
 
Table 5. National Cancer Institute Common Toxicity Criteria grade 3–4 toxicities (http://ctep.cancer.gov)

 
Toxicities of the GD regimen leading to early discontinuation mainly consisted of mild events such as fluid retention (10 patients; 6.5%), skin allergic rashes or asthma crisis (seven patients; 4.5%), hepatitis (one patient with only hepatic enzyme increase and documented recovery) and miscellaneous (12 patients; 7.5%). However, suspected GD-induced pneumonitis was reported in eight patients (5.2%); these pulmonary events occurred mostly late in the chemotherapy program (after cycle 4) and were reported as probably drug-related according to the investigator. Among these observations, follow-up proved five documented recoveries and one toxicity-related death. In the CV arm, early discontinuations were induced by hematological toxicity (six patients), renal insufficiency (four), auditory impairment (three) or miscellaneous (nine).

A lower number of serious adverse events possibly or probably related to the study drugs were reported in the GD arm compared with the CV arm (21.9% and 35.9%, respectively; P=0.007). Those serious adverse events reported in the latter arm were predominantly febrile neutropenia, infections and vomiting episodes (Table 5). More pulmonary events were observed in the GD arm compared with the CV arm. Deaths following an adverse event were twice as low in the GD arm compared with the CV arm (2.6% versus 5.1%, respectively); among these deaths, one (pulmonary edema) and three (two septic shocks concomitant with neutropenia, one interstitial pneumopathy in context of disease progression) were drug-related, respectively. Likewise, fewer early deaths occurred in the GD regimen compared with the CV (4.5% and 8.3%, respectively; P=0.17).

More than 75% of the patients assigned to both groups received post-study treatment consisting of palliative radiotherapy (bone metastatis or mediastinal syndrome) and/or second- and third-line treatment. Second-line chemotherapy choice differed according to the allocated group: up to 60% of the patients randomly assigned to the GD group subsequently received a vinorelbine–platinum combination as the predominant second line, whereas patients assigned to the CV group mainly received docetaxel or gemcitabine, or the combination of both.


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
By comparing GD and CV, this phase III study addressed the case of cisplatin in the treatment of advanced and metastatic NSCLC. The study has been designed to detect an improvement by 3 months of PFS for patients receiving the GD regimen. This phase III study did not demonstrate a difference in PFS between the GD and CV combinations. However, toxicity profiles differed significantly between the two groups: myelosupression was more severe in the CV group, whereas pulmonary events were more pronounced on the GD group. A significantly higher number of serious adverse events were observed in the CV arm.

The impact in terms of QoL was difficult to evaluate owing to the low reliability of the EORTC QLQ-C30 questionnaire regarding early effects of chemotherapy, which usually occurs the week following the chemotherapy whereas the EORTC QLQ-C30 questionnaire investigates the week before chemotherapy.

Because the study was not powered to test a non-inferiority hypothesis, it is not possible to conclude that GD and CV are equivalent in terms of survival. Whether or not the higher toxicity of the CV regimen was related to the platinum compound is difficult to establish. Although combination-related emesis was probably a consequence of cisplatin, myelosuppression depended on the combination of vinorelbine and cisplatin. This combination is thought to cause a more grade 3–4 neutropenia compared with other cisplatin-based doublets such as gemcitabine–cisplatin [6Go], docetaxel–cisplatin [7Go] or paclitaxel–carboplatin [8Go]. Nevertheless, our study allowed us to conclude that, compared with a well-established cisplatin-based doublet, the GD regimen had a good safety profile. Patients randomly assigned to the CV regimen were at higher risk of serious adverse events when compared with GD, whereas patients randomly assigned in the GD group were at higher risk of toxicity-induced early treatment discontinuation. This is contradictory only in appearance insofar as these latter toxicities were graded as mild events, albeit justifying per-protocol chemotherapy interruption. Dose reduction was more frequently applied with vinorelbine in comparison with either gemcitabine or docetaxel. Consequently, specific toxicities related to the latter cytotoxic agents were more apparent.

Is CV an adequate standard combination when designing a trial attempting to compare cisplatin-based and cisplatin-free regimens? The patients allocated to the CV arm present with outcomes in the range usually observed with CV regimens in other trials. The pivotal study on vinorelbine in NSCLC compared this drug from the third generation combined with cisplatin, with vindesine–cisplatin, a classical second-generation-containing regimen [9Go]. Patients receiving the CV regimen proved to have a longer survival. Since then, CV has been compared in different studies with other cisplatin-third-generation drug combinations: paclitaxel–carboplatin, gemcitabine–cisplatin, docetaxel–cisplatin and docetaxel–carboplatin. Most of the studies did not demonstrate survival difference, so are those cisplatin-based doublets equivalent? In a recent meta-analysis of gemcitabine–platinum chemotherapy versus other platinum-containing combinations, a significant reduction in overall mortality in favor of a gemcitabine-based regimens was observed given an HR (95% CI) of 0.90 (0.84–0.96) with an absolute benefit at 1 year of 3.9% [23Go]. However, a subset analysis also showed that when compared with other third-generation doublets, the benefit offered by gemcitabine–platinum therapy is lower, with an HR (95% CI) of 0.93 (0.86–1.01). Therefore, there are several clues in the literature suggesting that the CV combination, the most frequently used in France at the time this study was planed, produces similar survival in comparison with other standard cisplatin-based doublets. The weekly dosage of 30 mg/m2 for vinorelbine, as used in this study, was congruent with the planned schedule described in the pivotal vinorelbine study [9Go]. Current schedules of CV combinations favor either a weekly administration of 25 mg/m2 associated with cisplatin 100 mg/m2 every 4 weeks, or day 1 and 8 administrations of 25 mg/m2 associated with cisplatin 75–80 mg/m2 every 3 weeks. Indirect comparisons between the CV regimen of this study and those schedules with lowering vinorelbine dosages suggest that anticancer effects are similar in terms of response rate and survival. Therefore, our CV could be regarded as a valid standard cisplatin-based doublet.

The newest cytotoxic compounds such as taxanes, vinorelbine and gemcitabine yield a better efficacy/toxicity ratio and could be safely combined in doublet regimens.

Gemcitabine and taxane combinations have been evaluated extensively owing to their different mechanisms of cellular activity and their distinct toxicity. The only overlapping toxicity between gemcitabine and taxanes consisted of possible severe interstitial pneumonitis in 3% to 5% of patients [24Go]. Up to now, there are two reported studies comparing the gemcitabine–paclitaxel doublet with a cisplatin-based regimen [25Go, 26Go]. In the Hellenic study, 329 patients were randomly assigned to receive carboplatin–paclitaxel or gemcitabine–paclitaxel [25Go]. There were no differences in survival, response rate or toxicity according to conventional statistical tests. In addition, retrospective analysis of cost did not demonstrate a difference between the two arms. The EORTC group reported a randomized study comparing the gemcitabine–paclitaxel doublet with two different cisplatin-based doublets [26Go]. In that trial, 480 patients were randomly allocated to receive either: (arm A) paclitaxel–cisplatin; (arm B) gemcitabine–cisplatin; or (arm C) gemcitabine–paclitaxel. The planned doses and schedules for paclitaxel were identical in arms A and C. Similarly, gemcitabine was delivered according to an identical procedure in arms B and C, and cisplatin in arms A and B. There was no statistical difference regarding response rates when the three groups were compared (31%, 36% and 27% in arms A, B and C, respectively). However, patients randomly assigned in arm C proved to have a shorter overall survival and PFS (6.9 and 3.9 months) when compared with arm A (8.1 and 4.4 months) or arm B (8.8 and 5.6 months). There was a trend towards a significant difference in overall survival favoring arm A in comparison with arm C (P <0.1). Whether or not this trend resulted from the specific taxane used in this combination could not be determined, as no direct comparison of paclitaxel and docetaxel has been carried out in this setting.

The GD combination has been particularly developed by the Hellenic Oncology Group. The feasibility of this combination has been reported extensively [27Go]. Aside from this publication, numerous other phase I–II studies have evaluated different dosages and schedules of the GD doublet [16Go, 24Go]. All schedules share a similar gemcitabine delivery, on days 1 and 8 of a 21-day cycle; however, they differed in the modalities of delivery of docetaxel. Most of the studies favored delivery of docetaxel on day 8 instead of day 1. The day 8 docetaxel delivery avoided subsequent gemcitabine omission (no overlapping myelosuppression) and appeared to be well tolerated [16Go].

Another study aimed at comparing the GD combination with cisplatin–docetaxel. A total of 441 patients were randomly allocated to receive either docetaxel 100 mg/m2 and cisplatin 80 mg/m2 every 21 days, or docetaxel 100 mg/m2 on day 1 and gemcitabine 1100 mg/m2 on days 1 and 8. G-CSF was given in both arms as a primary prophylaxis for neutropenia. The study hypothesis was designed to detect a 15% response rate difference in favor of the GD arm [28Go]. Taking into account fully eligible patients (instead of all randomly assigned patients), no statistical difference was observed when response rate (33% and 35% for GD and docetaxel–cisplatin, respectively), time to progression (9 versus 8 months; P=0.778) and overall survival (9.5 versus 10 months; P=0.980) were considered. It is noteworthy that the main difference between the two regimens lay in the toxicity profile, in that more patients in the docetaxel–cisplatin arm were affected by grade 3–4 neutropenia, whereas more patients in the GD arm suffered from edema and fluid retention. Considering the fact that no survival difference could be detected between the two arms, the authors concluded that the two drug combinations had comparable activity.

Our study end point was PFS. It was chosen taking into account current standard therapy of this disease, which frequently involves resorting to second-line therapy at time of progression. The second-line therapy was left to each investigator's policy, and therefore PFS was expected to more accurately reflect the treatment allocation. These second-line regimens have been chosen taking into account clinical criteria and disease status at time of progression. Therefore, they gave an opportunity to determine what combinations were considered as feasible for the investigator panel.

Of interest, the most frequently chosen second-line regimen for patients who received GD was a vinorelbine–platinum combination, whereas docetaxel and gemcitabine given sequentially or concomitantly were the drugs most widely proposed to patients for whom disease progressed after CV. Therefore, second-line chemotherapy, as determined on a patient-by-patient basis according to each center's policy, resulted in a substitute of crossover therapy for a majority of patients receiving post-study therapy. Which sequence is best, GD followed by CV or the reverse, cannot be established from this study.

In conclusion, this study did not demonstrate an advantage in PFS for GD compared with CV. However, the CV regimen had higher rate of toxic events, mainly myelosuppression. The non-platinum-containing regimen reported here could be considered as a rational alternative to a cisplatin-based doublet.

Recently, ASCO published the updated 2003 guidelines for treatment of NSCLC [29Go]. For stage IV NSCLC, it is recommended that ‘first-line chemotherapy given to patients with advanced NSCLC should be a two-drug combination regimen. Non-platinum-containing chemotherapy regimens may be used as alternatives to platinum-based regimens in the first line’. Are we to leave the cisplatin era in the treatment of advanced NSCLC? In spite of the ASCO recommendations, the question could be considered as still debatable. Our study is a contribution to answering this question. As observed in the various aforementioned studies, the sizes of differences between cisplatin- and non-cisplatin-containing regimens are so small that a meta-analysis could be the best way to progress in this issue.


    Acknowledgements
 
The authors wish to thank Dr Farid Khial and Mrs Marie Cécile Bozonnat for technical assistance, Dr Maria Aguirre, Dr Caroline Germa and Nathalie Heurtebize for statistical assistance, and the staff of Languedoc Mutualité Nouvelle Technologie for technical support. Research was sponsored by Eli Lilly France and supported by grants from Aventis France and Languedoc Mutualité Nouvelle Technologie. Drugs were provided by Eli Lilly (gemcitabine) and Aventis (docetaxel and cisplatin). This work was presented previously at the 10th World Conference on Lung Cancer, 14 August 2003, Vancouver, Canada.

Participating centers and specialists were as follows: Montpellier University Hospital (J. L. Pujol, X. Quantin); Belfort Hospital (J. L. Breton); Caen Cancer Institute (R. Gervais, A. Rivière); Lyon Cancer Institute (P. Rebattu); Besançon University Hospital (A. Depierre); Bobigny Avicenne University Hospital (J. F. Morère); Paris Tenon University Hospital (B. Milleron); Vesoul Hospital (D. Debieuvre); Perpignan Saint Pierre Institute (D. Castéra); Lyon Sud University Hospital (P. J. Souquet); Grenoble University Hospital (D. Moro-Sibilot); Tours University Hospital (E. Lemarié); Strasbourg Hautepierre University Hospital (R. Kessler); Clermont Ferrand University Hospital (H. Janicot); Briey Hospital (D. Braun); Nancy Cancer Institute (D. Spaeth); Nice University Hospital (C. Clary); Lille Catholic University Hospital (P. Maes); Carcassonne Hospital (J. L. Labouret); Lille University Hospital (B. Wallaert); Nice Cancer Institute (M. Poudenx); Marseille Saint Joseph Hospital (J. F. Pollet); Catherine de Sienne Nantes Hospital (C. El Kouri): Hôtel Dieu Paris University Hospital (J. M. Bréchot): Nançy University Hospital (J. F. Bic).

Received for publication September 26, 2004. Revision received November 25, 2004. Accepted for publication December 3, 2004.


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. American Society of Clinical Oncology. Clinical practice guidelines for the treatment of unresectable non-small-cell lung cancer. Adopted on May 16, 1997, by the American Society of Clinical Oncology. J Clin Oncol 1997; 8: 2996–3018.

2. Bunn PA. The expanding role of cisplatin in the treatment of non-small-cell lung cancer. Semin Oncol 1989; 16 (4 Suppl 6): 10–21.

3. 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 1995; 311: 899–909.[Abstract/Free Full Text]

4. Sandler AB, Nemunaitis J, Denham C et al. Phase III trial of gemcitabine plus cisplatin versus cisplatin alone in patients with locally advanced or metastatic non-small-cell lung cancer. J Clin Oncol 2000; 18: 122–130.[Abstract/Free Full Text]

5. Wozniak AJ, Crowley JJ, Balcerzak SP et al. Randomized trial comparing cisplatin with cisplatin plus vinorelbine in the treatment of advanced non-small-cell lung cancer: a Southwest Oncology Group study. J Clin Oncol 1998; 16: 2459–2465.[Abstract]

6. Scagliotti GV, De Marinis F, Rinaldi M et al. Phase III randomized trial comparing three platinum-based doublets in advanced non-small-cell lung cancer. J Clin Oncol 2002; 20: 4285–4291.[Abstract/Free Full Text]

7. Fossella F, Pereira JR, von Pawel J et al. Randomized, multinational, phase III study of docetaxel plus platinum combinations versus vinorelbine plus cisplatin for advanced non-small-cell lung cancer: the TAX 326 study group. J Clin Oncol 2003; 21: 3016–3024.[Abstract/Free Full Text]

8. 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 2001; 19: 3210–3218.[Abstract/Free Full Text]

9. 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 1994; 12: 360–367.[Abstract]

10. Cisplatin. In Dorr RT, Fritz WL (eds): Cancer Chemotherapy Handbook. New York, NY: Elsevier 1980; 315–323.

11. McAlpine D, Johnstone BM. The ototoxic mechanism of cisplatin. Hear Res 1990; 47: 191–203.[CrossRef][ISI][Medline]

12. Abratt RP, Bezwoda WR, Falkson G et al. Efficacy and safety profile of gemcitabine in non-small-cell lung cancer: a phase II study. J Clin Oncol 1994; 12: 1535–1540.[Abstract]

13. Cerny T, Kaplan S, Pavlidis N et al. Docetaxel (Taxotere) is active in non-small-cell lung cancer: a phase II trial of the EORTC Early Clinical Trials Group (ECTG). Br J Cancer 1994; 70: 384–387.[ISI][Medline]

14. Roszkowski K, Pluzanska A, Krzakowski M et al. A multicenter, randomized, phase III study of docetaxel plus best supportive care versus best supportive care in chemotherapy-naive patients with metastatic or non-resectable localized non-small cell lung cancer (NSCLC). Lung Cancer 2000; 27: 145–157.[CrossRef][ISI][Medline]

15. Anderson H, Hopwood P, Stephens RJ et al. Gemcitabine plus best supportive care (BSC) vs BSC in inoperable non-small cell lung cancer—a randomized trial with quality of life as the primary outcome. UK NSCLC Gemcitabine Group. Non-Small Cell Lung Cancer. Br J Cancer 2000; 83: 447–453.[CrossRef][ISI][Medline]

16. Rebattu P, Quantin X, Ardiet C et al. Dose-finding, pharmacokinetic and phase II study of docetaxel in combination with gemcitabine in patients with inoperable non-small cell lung cancer. Lung Cancer 2001; 33: 277–287.[CrossRef][ISI][Medline]

17. World Health Organization. WHO Handbook for Reporting the Results of Cancer Treatment, Vol. 48. Geneva, Switzerland: WHO 1979.

18. Aaronson NK, Ahmedzai S, Bergman B et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993; 85: 365–376.[Abstract]

19. Staquet MJ, Hays RD, Fayers PM (eds): Quality of Life Assessment in Clinical Trials. Methods and Practice. Oxford: Oxford University Press 1998.

20. Hryniuk WM, Goodyear M. The calculation of received dose intensity. J Clin Oncol 1990; 8: 1935–1937.[ISI][Medline]

21. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–481.[ISI]

22. Cox JD, Yesner RA. Causes of treatment failure and death in carcinoma of the lung. Yale J Biol Med 1981; 54: 201–207.[ISI][Medline]

23. Le Chevalier T, Brown A, Natale R et al. Gemcitabine in the treatment of Non-Small Cell Lung Cancer (NSCLC): A meta-analysis of survival and progression free survival data. Lung Cancer 2003; 41 (Suppl 2): S70 (Abstr O-239).

24. Dunsford ML, Mead GM, Bateman AC et al. Severe pulmonary toxicity in patients treated with a combination of docetaxel and gemcitabine for metastatic transitional cell carcinoma. Ann Oncol 1999; 10: 943–947.[Abstract]

25. 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 2002; 20: 3578–3585.[Abstract/Free Full Text]

26. Smit EF, van Meerbeeck JP, Lianes P et al. Three-arm randomized study of two cisplatin-based regimens and paclitaxel plus gemcitabine in advanced non-small-cell lung cancer: a phase III trial of the European Organization for Research and Treatment of Cancer Lung Cancer Group—EORTC 08975. J Clin Oncol 2003; 21: 3909–3917.[Abstract/Free Full Text]

27. Georgoulias V, Kouroussis C, Androulakis N et al. Front-line treatment of advanced non-small-cell lung cancer with docetaxel and gemcitabine: a multicenter phase II trial. J Clin Oncol 1999; 17: 914–920.[Abstract/Free Full Text]

28. Georgoulias V, Papadakis E, Alexopoulos A et al. Platinum-based and non-platinum-based chemotherapy in advanced non-small-cell lung cancer: a randomised multicentre trial. Lancet 2001; 357: 1478–1484.[CrossRef][ISI][Medline]

29. American Society of Clinical Oncology. Treatment of unresectable non-small-cell lung cancer guideline: update 2003. Adopted on September 8, 2003, by the American Society of Clinical Oncology. J Clin Oncol 2004; 22: 330–353.[Free Full Text]