GLOB-1: a prospective randomised clinical phase III trial comparing vinorelbine–cisplatin with vinorelbine–ifosfamide–cisplatin in metastatic non-small-cell lung cancer patients

P. J. Souquet1,+, E. H. Tan2, J. Rodrigues Pereira3, R. Van Klaveren4, A. Price5, U. Gatzemeier6, M. Jaworski7, J. P. Burillon7 and D. Aubert7

1 Centre Hospitalier Lyon-Sud, France; 2 National Cancer Centre, Singapore; 3 Instituto Arnaldo Vieira de Carvalho, Sao Paulo, Brazil; 4 Kliniek Longarts, Rotterdam, The Netherlands; 5 Western General Hospital, Edinburgh, UK; 6 Krankenhaus, Großhansdorf, Germany; 7 Institut de Recherche Pierre Fabre, France

Received 6 July 2001; revised 20 November 2001; accepted 12 June 2002


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Acknowedgements
 References
 
Background:

The standard doublet, vinorelbine–cisplatin, was compared with a triplet of vinorelbine–ifosfamide–cisplatin, in terms of survival, in patients with advanced non-small-cell lung cancer (NSCLC).

Patients and methods:

From February 1998 to June 1999, 259 chemonaïve patients entered the study and were randomised to receive either vinorelbine–cisplatin (NP; vinorelbine 30 mg/m2 on days 1, 8 and 15 with cisplatin 80 mg/m2 on day 1) or vinorelbine–ifosfamide–cisplatin (NIP; vinorelbine 25 mg/m2 on days 1 and 8, ifosfamide 3 g/m2 on day 1 and cisplatin 75 mg/m2 on day 1), with both regimens being repeated every 3 weeks. All patients had stage IV or relapsed disease and a performance score of 0 or 1.

Results:

The overall response rate was 34.6% for NP and 35.7% for NIP. Median and 1-year survival rates were 10.0 months and 38.4% for NP, and 8.2 months and 33.7% for NIP, respectively. A median of four cycles was administered in each arm. The major World Health Organization grade 3–4 toxicities for NP and NIP, respectively, were: neutropenia (20.3% compared with 9% of cycles), anaemia (4.1% compared with 5% of cycles), nausea and vomiting (22.2% compared with 19.4% of patients) and alopecia (5.6% compared with 29.8% of patients). Four toxic deaths occurred in the NP arm and eight in the NIP arm.

Conclusions:

The different schedules of vinorelbine in the two arms led to a greater survival in the NP arm without impairing the tolerance profile, although this is not statistically significant. This confirms that the two-drug combination NP is a reference treatment for metastatic NSCLC. The role of three-drug combinations remains questionable in this subset of patients.

Key words: chemotherapy, doublet, non-small-cell lung cancer, quality of life, stage IV, triplet


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Acknowedgements
 References
 
Lung cancer is the leading cause of cancer mortality in men throughout the world, with >1 000 000 new cases each year. More than 75% of patients with lung cancer have non-small-cell histology, and 50% present with incurable stage IIIB or IV disease.

Treatment for patients with advanced or metastatic non-small-cell lung cancer (NSCLC) is usually cisplatin-based chemotherapy, which produces a significant improvement in overall survival as compared with best supportive care. In meta-analyses of randomised trials, cisplatin-based chemotherapy increased the median survival by 1.5 months, and the 1-year survival rate improved from 15% to 25% compared with best supportive care [1].

On the basis of reported response rates in phase II trials several agents appear to have interesting levels of activity; however, response rates are notoriously variable in this disease, and poorly correlated with survival. Vinorelbine (Navelbine®; Pierre Fabre Médicament, Boulogne, France) has undergone some of the most thorough testing in randomised trials performed in NSCLC. Depierre et al. [2] included 231 patients to be treated either with vinorelbine 30 mg/m2 weekly, or vinorelbine 30 mg/m2 on days 1, 8 and 15 with cisplatin 80 mg/m2 on day 1 repeated every 3 weeks. A randomised study conducted by Le Chevalier et al. [3] compared standard cisplatin and vindesine with either cisplatin and vinorelbine or with single agent vinorelbine. Patients treated with cisplatin–vinorelbine had a significantly longer median survival of 9.3 months compared with 7.4 months for patients treated with cisplatin and vindesine. In another clinical randomised trial conducted in the USA by the South West Oncology Group (SWOG) [4] vinorelbine–cisplatin resulted in superior survival compared with single-agent cisplatin.

A recently completed phase III SWOG trial [5] comparing two novel doublets, paclitaxel–carboplatin and cisplatin–vinorelbine, demonstrated similar efficacy with a median survival of 32 weeks for both arms and a 1-year survival rate of 38% and 36%, respectively. Survival remains suboptimal for these patients.

Survival improvement is still needed, especially when new combinations remain to be investigated. One strategy is to add an active drug with a different mechanism of action to the active doublet of vinorelbine and cisplatin.

The combination of vinorelbine plus ifosfamide and cisplatin has demonstrated a high response rate and has improved 1-year survival in phase II trials. Several studies [611] have reported response rates ranging from 41% to 66%, median survival from 9.8 to 14 months and 1-year survival from 47% to 60%.

We conducted a trial with vinorelbine–ifosfamide–cisplatin (NIP) according to three different schedules: (i) arm A: cisplatin (75 mg/m2 on day 1), ifosfamide (3 g/m2 on day 1), vinorelbine (25 mg/m2 on day 1); (ii) arm B: cisplatin (75 mg/m2 on day 1), ifosfamide (3 g/m2 on day 1), vinorelbine (25 mg/m2 on days 1 and 8); (iii) arm C: cisplatin (75 mg/m2 on day 1), ifosfamide (3 g/m2 on day 1), vinorelbine (25 mg/m2 on days 1 and 8, and 12.5 mg/m2 on day 15). Response rates were 32%, 44% and 67% in groups A, B and C, respectively. Median survival was 6.5, 8.8 and 12.8 months for groups A, B and C, respectively. The 1-year survival observed in group C was 54% [7].

The regimen used for group C was apparently more effective than the other two, but this regimen was not easy to manage, requiring protracted hospitalisations.

Tan et al. [8] have included 78 patients with advanced/metastatic NSCLC in a NIP phase II trial: vinorelbine 25 mg/m2 on days 1 and 8, ifosfamide/Mesna 3 g/m2 on day 1, and cisplatin 50 mg/m2 on day 1 were delivered on a 21-day schedule. The overall response rate was 58%, a median survival of 14 months, with 60% of patients alive at 1 year.

From these phase II trials, NIP regimens seem to be safe and active, and consequently we have designed a randomised phase III trial in inoperable NSCLC patients with the aim of determining whether a more intensive treatment (the three-drug regimen NIP) is better than a standard treatment (the two-drug regimen NP) in terms of survival.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Acknowedgements
 References
 
Patients
The inclusion criteria for patient entry to the study were: (i) histological or cytological evidence of metastatic NSCLC (stage IV disease or NSCLC in relapse after a local treatment) with no prior chemotherapy; (ii) age 18–75 years. Clinical characteristics included: (i) Karnofsky performance score (KPS) >=80% and life expectancy >=3 months; (ii) at least one evaluable lesion; (iii) adequate bone marrow function (granulocyte count >=1.5 x 109/l, platelet count >=100 x 109/l); (iv) adequate liver function (total bilirubin <=1.5 x the upper limit of normal (ULN), transaminases <=2.5 x ULN, unless due to documented liver metastases); and (v) adequate kidney function (creatinine <=130 µmol/l). Patients were excluded from the trial if they had a local relapse suitable for treatment by radiation therapy. Additional clinical features that precluded entry into the trial were: (i) clinical signs of brain metastasis or leptomeningeal involvement; (ii) second malignancy (except adequately treated basal cell carcinoma of the skin and carcinoma in situ of the uterine cervix); (iii) active infectious disease; (iv) uncontrolled hypercalcemia; (v) pregnancy; (vi) neurological disorders that could interfere with the evaluation of neurological toxicity; and (vii) mentally incapacitated patients, family, social or environmental conditions impairing adequate follow-up and protocol compliance. Patients had to be willing and able to complete Visual Analogic Scale (VAS) questionnaires and give written informed consent. Local ethics committee approval had to be obtained.

Treatment
The treatment schedule was as follows: for arm A, vinorelbine 30 mg/m2 on days 1, 8 and 15, and cisplatin 80 mg/m2 on day 1, repeated every 21 days; and for arm B, vinorelbine 25 mg/m2 on days 1 and 8, and cisplatin 75 mg/m2 and ifosfamide 3 g/m2 on day 1, repeated every 21 days.

At least two courses of treatment were given unless rapid disease progression was seen after the first course. A full assessment was performed after two courses: patients with complete or partial response received a maximum of six cycles; in case of stable disease, patients received two further cycles and thereafter the treatment was stopped. No consolidation was planned.

Chemotherapy doses were reduced for haematological, neurological, hepatic and renal toxicities. Toxicities were graded according to the World Health Organization (WHO) criteria. Changes in dosage were based on blood count results obtained on day 1 of treatment; if neutrophils were <1.5 x 109/l and platelets were <100 x 109/l, treatment was delayed by 1 week. Treament on days 8 or 15 had to be cancelled if neutrophil counts were <1.0 x 109/l and platelets were <100 x 109/l. If treatment could not be given after a 3-week interval because of haematological toxicity, it had to be discontinued and the patient withdrawn from the study. Prophylactic use of colony-stimulating factors was not permitted. Neurological toxicity above grade 2 (including paresthesia, muscle weakness or paralytic ileus) resulted in suspension of treatment; ototoxicity grade 2 or 3 resulted in a 50% dose reduction of cisplatin.

The following dose modifications of vinorelbine were implemented based on AST/ALT (aspartate aminotransferase/alanine aminotransferase) and bilirubin values on day 1 of treatment: if AST/ALT were between 5.1 and 20.0 x ULN or bilirubin was between 1.5 and 3.0 x ULN, dosing was cancelled and the patient was reassessed 1 week later. If AST/ALT were >20.0 x ULN or bilirubin was >3.0 x ULN, vinorelbine was discontinued. Renal impairment also induced vinorelbine dose modifications based upon serum creatinine and creatinine clearance as follows: if serum creatinine was grade >1, the dose was delayed by 1 week and the test repeated; if serum creatinine was greater than the normal value of the centre, the dose was delayed by 1 week and the test repeated. If after a 2-week delay, the creatinine clearance remained at the normal value of the centre, the patient was taken off the study.

If performance status at the time of the next cycle was KPS <70%, treatment was delayed by 1 week. If there was no improvement in KPS, treatment had to be stopped and the patient withdrawn.

The choice of further treatments was at the investigator’s discretion.

Objectives
The main objective was to compare the effect of a three-drug regimen (NIP) with a two-drug regimen (NP) on 1-year survival in patients with inoperable NSCLC previously untreated by chemotherapy.

The secondary aims were: (i) to compare the effect on other measures of efficacy (response rate, duration of response, time to tumour progression, median survival); (ii) to assess the impact of a three-drug regimen (NIP) compared with a two-drug regimen (NP) on patient benefit, defined as changes in KPS, patient weight and lung cancer-related symptoms (pain, dyspnea, cough, haemoptysis, asthenia and anorexia); and (iii) to assess the safety of both regimens.

Assessments of safety and efficacy
Assessment of toxicity was made before each cycle of chemotherapy according to WHO criteria [12]. The data for each patient across all cycles of chemotherapy were used, recording the most severe result.

Appropriate scanning, or physical assessment to determine tumour involvement was completed within 2 weeks before study entry. Perpendicular diameters of representative malignant lesions were measured and recorded, and the extent of evaluable disease was assessed. Physical assessments were made by chest X-ray, bronchoscopy, ultrasound and/or computed tomography (CT) scan of chest and liver; CT scan of the brain and bone imaging was undertaken if symptoms were present. Response assessment was performed after patients had completed two cycles. Patients responding to therapy had a repeated evaluation at least 4 weeks later using the same methods. One-year survival was the primary criterion to assess the efficacy of the two regimens, and the best response was recorded for each patient. All patients were followed up for survival every 3 months until death.

Patient benefit assessment
Each patient was classified as a clinical-benefit responder or non-responder on the basis of the following measures, as described below: (i) main measures of clinical benefit; and (ii) secondary measure of clinical benefit.

Main measures of clinical benefit: change in the lung cancer disease-related symptoms
Symptom improvement was assessed using a VAS, which was given to the patient, taking as baseline the first day of treatment. The intensity of each of the following symptoms was evaluated every 3 weeks: pain, dyspnoea, cough, haemoptysis, asthenia and anorexia. The patient was asked to place a mark along the line to indicate his or her subjective judgement. The score of each VAS was expressed as a percentage.

Main measures of clinical benefit: change in performance status
Performance status was recorded once at baseline, before randomisation, and every 3 weeks as long as the patient remained in the study.

Secondary measure of clinical benefit: change in weight
Patient’s weight was recorded at study entry and every 3 weeks, before each course.

To determine whether a patient was a clinical-benefit responder, the main measures were considered first. A patient was considered as a non-responder if either disease symptoms or performance status worsened. A patient was defined as a responder if either disease symptoms or performance status improved. If both were stable, the patient was considered as stable for main measures.

If a patient was classified as stable for the main clinical-benefit measures, the secondary measure (weight change) was examined. Such a patient was defined to be a clinical-benefit responder if weight change was classified as positive (increase of at least 3% from the baseline). If weight change was not positive, the patient was defined as a clinical-benefit non-responder.

Statistical methods
This study was designed to recruit 250 patients with 125 patients in each arm. The sample size was calculated as follows: for the patients randomised to vinorelbine–cisplatin (arm A), 1-year survival was estimated at 35%. For a suitable additional benefit from the vinorelbine–ifosfamide/Mesna–cisplatin, the objective was to detect an absolute improvement of 20% in 1-year survival. In order to have 90% confidence, with a one-sided 5% level test, 230 patients were needed to detect the above difference in survival. Considering the likelihood that patients who were lost to follow-up might account for up to 5% of the total number of subjects, a total sample size of 250 eligible patients (125 patients per group) was required.

All relevant baseline assessment data were both tabulated and summarised by frequency and percentage or qualitative item and by calculating the mean and standard deviation, and median and range for quantitative items.

One-year survival was the main end point of this trial. Overall response rate was the secondary end point and was analysed for the intention-to-treat population. Comparison was performed using the chi-square test.

The analysis of safety was focused on the frequency and severity of unwanted side effects or adverse events. Toxicity was graded according to WHO criteria. The analysis was performed for the worst grade by patients and by cycle for each patient.

Duration of response was assessed using the Kaplan–Meier method for patients who achieved an objective response (complete or partial response) according to WHO criteria.

Survival was calculated from randomisation to patient death, and overall survival curves were derived using the Kaplan–Meier method and were compared using the log-rank test .


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Acknowedgements
 References
 
Population description
Two hundred and fifty-nine patients from 37 centres in 16 countries were enrolled over 17 months between February 1998 and June 1999, with 133 patients in the NP arm and 126 in the NIP arm. Patients were well matched for pre-treatment characteristics (Table 1): age, gender, KPS, stage, histological subtype and number of metastatic sites, with the exception of liver metastases which were present in 13 patients in the NP arm and 30 patients in the NIP arm (9.8% compared with 23.8%; P = 0.003).


View this table:
[in this window]
[in a new window]
 
Table 1. Characteristics of randomised patients
 
The mean time between diagnosis and entry to the study was 21 days (range 1 day to 43.6 months in the NP arm and 1 day to 52.7 months in the NIP arm). No patient had received previous chemotherapy for advanced disease. Twenty-four patients (NP, 16; NIP, 8) had undergone only surgery before relapsing, 11 patients (NP, 6; NIP, 5) had been treated with radiation therapy only, and 15 patients (NP, 8; NIP, 7) had received both treatments.

Extent of exposure
Among the 259 patients entered into the study, seven did not receive any treatment; four of them died between randomisation and treatment, two refused treatment after randomisation, and one had a decrease in performance score between randomisation and treatment.

Patients allocated to the NP arm received a median of 4 cycles (range 1–14). The median dose intensity was 17.9 mg/m2 per week (range 9.3–30.7) [odds ratio (OR) 59.8%; 95% confidence interval (CI) 31% to 102%] for vinorelbine, and 24 mg/m2 per week (range 16.9–28.8) (OR 89.9%; 95% CI 63% to 108%) for cisplatin. In the NIP arm the median number of cycles administered was 4 (range 1–11). The median dose intensity was 14.6 mg/m2 per week (range 6.7–17.5) (OR 87.8%; 95% CI 40% to 105%) for vinorelbine, 23.3 mg/m2 per week (range 15.5–33.6) (OR 93.3%; 95% CI 62% to 134%) for cisplatin and 932 mg/m2 per week (range 552–1205) (OR 93.2%; 95% CI 55% to 120%) for ifosfamide.

Four hundred and fifty-two (60.2%) cycles were completed to schedule ±3 days (197 and 255 in the NP and NIP arms, respectively) and 299 (39.8%) were delayed (193 and 106 in the NP and NIP arms, respectively).Treatment delays were mainly due to haematological toxicity (71.9% and 52.9% in the NP and NIP arms, respectively).

Tolerance
Assessment of tolerance was performed in 240 evaluable patients (126 and 114 in the NP and NIP arms, respectively) and over 962 courses (497 and 465 in the NP and NIP arms, respectively).

Nineteen patients were not evaluable for tolerance, seven having not been treated and 12 having received only one cycle without evaluation. Among these 12 patients, eight were in the NIP arm: five patients died (acute renal failure, pulmonary dysfunction, septic shock, pneumonia, sudden death at home as reported in the serious adverse event forms); one was lost to follow-up; and two were unevaluable for unclear reasons. Four were in the NP arm: three patients died (neutropenia without fever, pneumonia, severe abdominal pain as reported in the serious adverse event forms); and one was lost to follow-up (refusal to continue after day 1 of the first cycle).

Haematological toxicity is summarised in Table 2. Grade 3–4 anaemia was observed in 4.1% and 5% of cycles in the NP and NIP arms, respectively; thrombocytopenia was very uncommon (two cases of grade 3 in the NIP arm). The total leukocyte count and the number of neutrophils were assessed at day 21.


View this table:
[in this window]
[in a new window]
 
Table 2. Percentage haematological and clinical toxicities (WHO grade 3–4)
 
While 20.8% of patients in the NP arm and 8.9% of patients in the NIP arm developed grade 4 neutropenia, this represented 7.6% of administered courses in the NP arm and 2.7% in the NIP arm; however, seven patients developed fatal infections during periods of neutropenia. Four toxic deaths occurred in the NP arm, and eight occurred in the NIP arm (P = 0.2). Eight deaths (three and five in the NP and NIP arms, respectively) also occurred in the 19 patients unevaluable for tolerance.

Clinical tolerance is reported in Table 2. Gastrointestinal tract toxicities included grade 3–4 nausea and vomiting observed in 22.2% and 19.4% of treated patients in the NP and NIP arms, respectively (7.7% and 7.7% of cycles, respectively), and no grade 4 toxicity was observed in the NP arm; four patients in the NP arm and two in the NIP arm experienced grade 3 diarrhoea. WHO grade 3–4 mucositis was not experienced by any patient in the NP arm, but by five patients in the NIP arm.

In assessable patients, alopecia was observed in 80 patients in the NP arm (seven cases of grade 3) and in 87 patients in the NIP arm (34 cases of grade 3).

No patient showed significant elevation of bilirubin. Rare grade 3–4 elevation of transaminases or serum alkaline phosphatase was observed (0.4% of cycles only in the NP arm for transaminases, and 0.2% of cycles in the NP arm compared with 0.5% in the NIP arm for serum alkaline phosphatase).

Clinical benefit
Globally, no difference was apparent between the NP and NIP arms as 70% and 57% of patients for NP and 68% and 62% of patients for NIP can be considered as responders after two and four courses, respectively. The completion of the VAS decreased along the course of treatment; 69% of patients had evaluable VAS scores after two courses, and only 46% had them after four courses.

Overall, in both arms, improvement was noticed in all patients, even after only one course, for pain, cough, dyspnoea and haemoptysis. With respect to asthenia and anorexia, a slight deterioration appeared after the first course in both arms, and more predominantly after the second course.

Survival
At the cut-off date of January 2001, 217 patients had died, 111 in the NP arm and 106 in the NIP arm, and the median follow-up for the 42 living patients was 23.2 months (range 18.6–31.8 months). Survival-related parameters are reported in Table 3.


View this table:
[in this window]
[in a new window]
 
Table 3. Survival-related parameters
 
There was no difference in survival between the two arms (Figures 1 and 2). Median progression free survival was 5.0 and 4.4 months in the NP and NIP arms, respectively (log-rank test, P = 0.87). Median survival was 10.0 months for NP and 8.2 months for NIP (log-rank test, P = 0.24). The estimated 1-year survival rate was 38.4% for NP and 33.7% for NIP (log-rank test, P = 0.24).



View larger version (15K):
[in this window]
[in a new window]
 
Figure 1. Progression-free survival for all patients (cut-off date 1 January 2001).

 


View larger version (15K):
[in this window]
[in a new window]
 
Figure 2. Overall survival for all patients (cut-off date 1 January 2001).

 
Tumour response
The response data of the 259 patients included in the analysis are summarised in Table 4. Forty-six partial responses were observed in the NP arm, while in the NIP arm, one patient achieved a complete response and 44 patients a partial response, giving objective response rates of 35.7% (95% CI 27% to 44%) and 34.6% (95% CI 27% to 43%), respectively. The median duration of response was 8.1 and 9.4 months for NP and NIP, respectively. Disease control (responders with stable disease) was 64.7% (95% CI 57% to 73%) for the NP arm and 59.5% (95% CI 51% to 68%) for the NIP arm.


View this table:
[in this window]
[in a new window]
 
Table 4. Response by treatment arm
 
No difference was found in terms of response between the two arms ({chi}2 test, P = 0.85).


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Acknowedgements
 References
 
Over the past 10 years, there have been considerable changes in the treatment of NSCLC, especially concerning the management of stage IV disease which comprises more than half of all NSCLC patients.

In 1995, a large meta-analysis confirmed the role of chemotherapy (especially platinum-based chemotherapy) in this setting [1], with an increased median survival of 1.5 months over best supportive care.

The combination of vinorelbine and cisplatin has been studied in several large phase III studies over the last 10 years, demonstrating high levels of efficacy, achieving an overall response of 30%, median survivals of around 8–9 months, and 1-year survival rates around 35%. These results have established vinorelbine–cisplatin as a standard treatment for NSCLC stage IV disease.

Triplet combinations have been widely developed in the field of NSCLC treatment in order to improve both response and survival. The two most commonly used triplets have been MVP (mitomycin C with vindesine and cisplatin) and MIC (mitomycin C with ifosfamide and cisplatin). Crino [13] has compared these two triplets to cisplatin and etoposide, which was considered to be a standard doublet in the early 1990s. The two triplets offered significant advantages in terms of response rate and survival estimates analysed by the log-rank test, with a slight superiority for MIC. As a result, Crino et al. [14] decided to compare MIC with a newly emerging doublet, gemcitabine–cisplatin. This doublet offered a significant superiority in terms of response but failed to demonstrate any advantage in terms of survival, either assessed as median survival or 1-year survival.

Vinorelbine has been studied extensively in triplet combinations with several active agents against NSCLC. Among all vinorelbine-based triplets, the largest experience has been reported with cisplatin and ifosfamide.

The present protocol was designed to detect an absolute improvement of 20% in the 1-year survival for the triplet. Concerning the schedules used, NIP was given every 3 weeks with vinorelbine administered on days 1 and 8, based on the previous phase II experience. In the NP arm, vinorelbine was administered weekly (according to the phase III regimens), but cisplatin was repeated every 3 weeks in order to have an equivalent schedule between both arms, and to avoid any discrepancy in terms of tolerance. The results of the study show that the toxicity profile of the two arms is acceptable. It should be noticed that the toxicities have not been evaluated at the nadir but at day 21; in the NP arm, vinorelbine was able to be administered on day 15 to 54.3% of the patients (34.9% of the cycles).

In terms of efficacy, response rates are similar in both arms, with confirmation for NP of activity around 35%, but results are lower than previously reported phase II trials for NIP at 35.7%. More surprisingly, NP generated superior results compared with NIP for survival, with a 2-month benefit for mean survival, and 38.2% versus 32.7%, respectively, for 1-year survival (although this was not statistically significant). This trend in favour of NP might be explained by the different schedules of vinorelbine (25 mg/m2 on days 1 and 8 for NIP, and 30 mg/m2 once weekly for NP), which allowed an increased dose-intensity in the NP arm; this would support the suggestion by Banerjee [15] that dosage of vinorelbine correlates with survival.

Due to the arrival of several new cytotoxics, many new triplets have been tested in recent years. Vinorelbine has been combined in several phase II studies with gemcitabine and cisplatin/ifosfamide, leading to response rates of 44–65% and median and 1-year survival of 8.6–13 months and 38–65%, respectively [1619]. Combinations of gemcitabine, ifosfamide and cisplatin [20, 21], or gemcitabine, paclitaxel (or docetaxel) and a platinum salt are numerous and all provide phase II results in the same range in terms of response and survival, generally at the cost of high haematological toxicity (high rate of transfusion and/or requirement for growth factors). With a 2-week schedule of gemcitabine, paclitaxel and cisplatin, Sorensen et al. [22] reported an overall response rate of 54% in 43 patients, with a mean survival of 10.7 months and 1-year survival of 47%; 77% of the patients experienced grade 3–4 neutropenia and 12 patients had grade 3 neurotoxicity. In a trial involving 71 patients treated with gemcitabine, paclitaxel and carboplatin, Burris et al. [23] found 44% of patients with grade 3–4 thrombocytopenia (nine patients required platelet transfusion), 32% of patients with grade 3–4 anemia, 16% with sepsis, 41% with grade 3–4 fatigue and 26% with grade 3–4 myalgia. With a combination of gemcitabine, docetaxel and carboplatin (supported with growth factors), Pectasides et al. [24] found that of 45 patients in total, 46.6% had neutropenia, 28.8% had thrombocytopenia and 22.2% had neurotoxicity, all grade 3–4.

Comella et al. [25] compared two new doublets, vinorelbine–cisplatin and gemcitabine–cisplatin, with a new triplet vinorelbine–gemcitabine–cisplatin. The triplet showed a significant advantage in terms of survival compared with the doublets, after only 60 patients had been accrued in each arm. The schedule of vinorelbine–cisplatin in this trial was quite toxic, using a higher dose of cisplatin at 120 mg/m2 on days 1 and 29. Alberola et al. [26] have reported the results of a large phase III study comparing a platinum-based doublet (gemcitabine–cisplatin) with a platinum-based triplet (gemcitabine–cisplatin–vinorelbine) and sequential non-platinum-based doublets (vinorelbine–gemcitabine followed by vinorelbine–ifosfamide). As in our study, the triplet failed to demonstrate any survival advantage (Table 5), even against the non-platinum arm.


View this table:
[in this window]
[in a new window]
 
Table 5. Comparison of doublet combinations with triplet combinations
 
In total, it seems that the use of triplet combinations in the setting of metastatic NSCLC remains questionable, mainly as a result of the tolerance profile. Other randomised trials are needed to decide whether the addition of a third drug could be beneficial for stage IV patients; optimisation of the dose intensity, use of growth factors, and quality of life studies could be paths to follow.

The Eastern Cooperative Oncology Group (ECOG) E 1594 trial [27] has recently demonstrated that four platinum-based doublets provide equivalent efficacy results that are in line with previous published phase III studies: it is obvious that a sort of limit has been reached [28] for the level of effectiveness of treatment.

Cisplatin, despite being considered the ‘key drug’ in NSCLC, is associated with several distressing toxicities in this subset of patients. There is a need for new approaches to both improve the results and to decrease the toxicities. One approach could be to move to consolidation after three or four cycles of cisplatin-based combination, but to administer only the second drug. Another promising approach could be to explore non-platinum combinations. Kosmidis [29] has shown favourable results for paclitaxel–gemcitabine compared with paclitaxel–carboplatin. Based on the above, we have decided to implement a new phase III trial comparing vinorelbine–gemcitabine with vinorelbine–carboplatin. Recruitment is on-going into this trial.

In conclusion, the place for the use of a classical triplet combination in metastatic NSCLC is still not resolved. The classical triplets do not show any survival superiority in large randomised trials. So it may be necessary to consider new triplets in which the third compound has a different mechanism of action and toxicity profile. In this setting, agents such as an epidermal growth factor tyrosine kinase receptor inhibitor, or monoclonal anti-Her-2 neu antibody might be considered. Results of ongoing trials are awaited.


    Acknowedgements
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Acknowedgements
 References
 
This study was conducted according to Good Clinical Practice standards and supported by the Institut de Recherche Pierre Fabre with their provision of vinorelbine (Navelbine®; Pierre Fabre Médicament, Boulogne, France), documentation of treatment, data management and statistical analysis.


    Footnotes
 
+ Correspondence to: Dr P. J. Souquet, Centre Hospitalier Lyon Sud, Department of Pneumology, F-69495 Pierre Benite Cedex, France. Tel: +334-7886-1328; Fax: +334-7886-6571; E-mail: pierre-jean.souquet{at}chu-lyon.fr Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Acknowedgements
 References
 
1. Non-Small Cell Lung Cancer Collaborative Group. Chemotherapy in non-small cell lung cancer using updated data on individual patients from 52 randomised clinical trials. Br Med J 1995; 311: 899–909.[Abstract/Free Full Text]

2. Depierre A, Chastang CL, Quoix E et al. Vinorelbine versus vinorelbine plus cisplatin in advanced non-small-cell lung cancer: a randomized trial. Ann Oncol 1994; 5: 37–42.[Abstract]

3. Le Chevalier T, Brisgand D, Douillard JY et al. Randomized study of Navelbine® and cisplatin versus vindesine and cisplatin versus Navelbine® 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]

4. 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]

5. Kelly K, Crowley JJ, Bunn P 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]

6. Baldini E, Tibaldi C, Chella A et al. Phase II study of vinorelbine/ifosfamide/cisplatin for the treatment of advanced non-small cell lung cancer. Ann Oncol 1996; 7: 747–749.[Abstract]

7. Souquet PJ, Fournel P, Bohas CH et al. Cisplatin and ifosfamide with various doses of vinorelbine (Navelbine®) in advanced non-small cell lung cancer. Semin Oncol 1996; 23 (2 Suppl): 8–10.

8. Tan EH, Ang PT, Wee J et al. Vinorelbine, ifosfamide and cisplatin in advanced non-small cell lung cancer. Acta Oncol 1999; 38: 619–622.[ISI][Medline]

9. Rey F, Astoul P, Marqueste L et al. Cisplatin, ifosfamide and vinorelbine combination chemotherapy in stage III–IV non-small cell lung cancer—a phase II study. Am J Oncol 1998; 21: 518–522.[ISI]

10. Barone C, Corsi DC, Pozzo C et al. Vinorelbine and alternating cisplatin and ifosfamide in the treatment of non-small cell lung cancer. Oncology 2000; 58: 25–30.[ISI][Medline]

11. Montalar J, Morales S, Vadell C et al. Vinorelbine, ifosfamide and cisplatin as essential treatment in patients with inoperable non-small cell lung cancer. Lung Cancer 2000; 29 (Suppl 1): Abstr 180.

12. World Health Organization (WHO). Handbook for reporting results of cancer treatment. Geneva: WHO 1979.

13. Crino L, Clerici M, Figoli F et al. Chemotherapy of advanced non-small-cell lung cancer: a comparison of three active regimens: a randomized trial of the Italian Oncology Group for Clinical Research (GOIRC). Ann Oncol 1995; 6: 347–353.[Abstract]

14. Crino L, Scagliotti GV, Ricci S et al. Gemcitabine and cisplatin versus mitomycin, ifosfamide and cisplatin in advanced non-small cell lung cancer: a randomized phase III study of the Italian Lung Cancer Project. J Clin Oncol 1999; 17: 3522–3530.[Abstract/Free Full Text]

15. Banerjee S, Dahrouges S, Toniak E et al. Vinorelbine in the treatment of non-small cell lung cancer: association of response with survival. Proc Am Soc Clin Oncol 2001; 20: 2860.

16. Comella P, Frasci G, Panza N et al. Cisplatin, gemcitabine and vinorelbine combination therapy in advanced non-small cell lung cancer: a phase II randomized study of the Southern Italy Cooperative Oncology Group. J Clin Oncol 1999; 17: 1526–1534.[Abstract/Free Full Text]

17. Ginopoulos P, Mastronikolis NS, Giannios J et al. A phase II study with vinorelbine, gemcitabine and cisplatin in the treatment of patients with stage IIIb–IV non-small cell lung cancer. Lung Cancer 1999; 23: 31–37.[ISI][Medline]

18. Comella G, Frasci G, Panza N et al. Cisplatin–gemcitabine vs cisplatin–gemcitabine–vinorelbine vs cisplatin–gemcitabine–paclitaxel in advanced non-small-cell lung cancer: final analysis of a Southern Italy Cooperative Oncology Group (SICOG) phase III trial. Ann Oncol 2000; 11 (Suppl 4): Abstr 4860.

19. Huidobro G, Grande C, Casal J. Gemcitabine, cisplatin, vinorelbine in advanced non-small-cell lung cancer: a single institution phase II trial. Ann Oncol 2000; 11 (Suppl 4): Abstr 525P.

20. Boni C, Bisagni G, Savoldi L et al. Treatment of stage IIIb–IV non-small-cell lung cancer with gemcitabine, ifosfamide, cisplatin: a phase II study. Ann Oncol 1998; 9 (Suppl 4): Abstr 439P.

21. Mohedano N, Medina NB, Sanchez P et al. Phase II study of gemcitabine in combination with cisplatin and ifosfamide in advanced non-small-cell lung cancer. Ann Oncol 1998; 9 (Suppl 4): Abstr 432P.

22. Sorensen JB, Stenbygaard LE, Dombernowsky P et al. Paclitaxel, gemcitabine and cisplatin in non-resectable non-small cell lung cancer. Ann Oncol 1999; 10: 1043–1049.[Abstract]

23. Burris HA III, Hainsworth JD, Erland JB et al. Phase II trial evaluating triplet chemotherapy using gemcitabine, paclitaxel and carboplatin in the treatment of patients with advanced non-small cell lung cancer. Semin Oncol 2000; 27 (Suppl 2): 9–13.

24. Pectasides D, Aspropotamitis A, Halikia A et al. Combination chemotherapy with carboplatin, docetaxel and gemcitabine in advanced non-small cell lung cancer: a phase II study. J Clin Oncol 1999; 17: 3816–3821.[Abstract/Free Full Text]

25. Comella P, Frasci G, Panza N et al. Randomized trial comparing cisplatin, gemcitabine and vinorelbine with either cisplatin and gemcitabine, or cisplatin and vinorelbine with either cisplatin and gemcitabine, or cisplatin and vinorelbine in advanced non-small cell lung cancer: interim analysis of a phase III trial of the Southern Italy Cooperative Oncology Group. J Clin Oncol 2000; 18: 1451–1457.[Abstract/Free Full Text]

26. Alberola V, Camps C, Provencia M et al. Cisplatin/gemcitabine vs cisplatin/gemcitabine/vinorelbine vs sequential doublets of gemcitabine/vinorelbine followed by ifosfamide/vinorelbine in advanced non-small cell lung cancer: results of a Spanish Lung Cancer Group phase III trial (GEPC.98-02). Proc Am Soc Clin Oncol 2001; 20: 1229 (Abstr).

27. Shiller JH, Harrington D, Sandler A et al. A randomized phase III trial of four chemotherapy regimens in advanced non-small cell lung cancer. Proc Am Soc Clin Oncol 2000; 19: (Abstr).

28. Shepherd FA. Chemotherapy for non-small cell lung cancer: have we reached a new plateau? Semin Oncol 1999; 26 (Suppl D): 3–11.[ISI][Medline]

29. Kosmidis PA et al. A randomized phase III trial of paclitaxel plus carboplatin versus paclitaxel plus gemcitabine in advanced non-small cell lung cancer: a preliminary analysis. Lung Cancer 2000; 29 (Suppl 2): 147.[ISI][Medline]