Triple- vs Double-Agent Chemotherapy for Advanced Non–Small-Cell Lung Cancer

July 1, 2000

In our previous phase I/II studies, both the cisplatin (Platinol), gemcitabine (Gemzar), and vinorelbine (Navelbine) (PGV), and cisplatin, gemcitabine, and paclitaxel (Taxol) (PGT) regimens produced a median survival of

ABSTRACT: In our previous phase I/II studies, both the cisplatin (Platinol), gemcitabine (Gemzar), and vinorelbine (Navelbine) (PGV), and cisplatin, gemcitabine, and paclitaxel (Taxol) (PGT) regimens produced a median survival of approximately 1 year in patients with advanced non–small-cell lung cancer (NSCLC). The present phase III study compared the median survival of patients treated with these triple-drug regimens to that of patients receiving cisplatin plus vinorelbine (PV) or cisplatin plus gemcitabine (PG). Accrual for the trial began in 1997 and by December 1998, a total of 240 patients (stage IIIB, 98; stage IV, 142) had been enrolled. An interim survival analysis was performed in April 1999. Overall, 151 patients had died. The median survival rates of patients in the PGV, PG, and PV arms were 51, 42, and 35 weeks, respectively. At the time of this analysis, the median survival of patients in the PGT arm could not be assessed; however, the 1-year projected survival rate was 58%. At multivariate Cox analysis, the estimated risk of death for patients receiving PGV compared with those receiving PV was 0.35 (95% CI, 0.16-0.77, P = .0058). Overall response rates were 47% in the PGV arm, 30% in the PG arm, 25% in the PV arm, and 58% in the PGT arm. Severe neutropenia and vomiting were significantly more frequent in patients who received PV than in those who received PGV. The PV regimen produced a significantly shorter survival compared with the PGV combination. Since this difference in survival complied with one of the early stopping rules, accrual in the PV arm was discontinued. Enrollment in the PGV, PG, and PGT arms is ongoing. [ONCOLOGY 14(Suppl 4):35-40, 2000]


A number of chemotherapy agents have been shown to be active in non–small-cell lung cancer (NSCLC).[1,2] However, the long-term prognosis for patients is still poor. After the results of a recent meta-analysis,[3] there is wide agreement about the benefit of including cisplatin (Platinol) in first-line combination treatment of this disease.

The cisplatin and vinorelbine (Navelbine) (PV),[4,5] cisplatin (or carbo-platin) and paclitaxel (Taxol) (PT),[6-9] or cisplatin and gemcitabine (Gemzar) (PG)[10-11] combinations have been widely tested in the last few years. Although the addition of a new agent to cisplatin has resulted in improvement in survival when compared with cisplatin alone, there has been no statistically significant difference in survival with the use of these “new” regimens vs the “older” regimens.[9,12-16] However, there is increasing interest in testing triple-drug combinations including combinations with at least two new agents.

Our group devised the cisplatin, gemcitabine, and vinorelbine (PGV) regimen for the management of patients with advanced NSCLC. In this regimen, we scheduled all three drugs on days 1 and 8 every 3 weeks so that we were able to deliver a full dose of chemotherapy with moderate toxicity.[17] This regimen produced more than a 50% overall response rate, and a 50-week median survival in patients with advanced NSCLC.[18] Therefore, we embarked on a three-arm phase III trial to compare this new combination with PV and PG. An interim analysis was planned when the first 60 patients in each arm were evaluable for response.

In a parallel phase I/II study, our group also investigated the cisplatin, gemcitabine, and paclitaxel (PGT) combination[19] administered every 3 weeks. In this regimen, we used the same doses of cisplatin and gemcitabine as in the previous trial, and split the paclitaxel dose on days 1 and 8. In view of the promising antitumor activity demonstrated by the PGT regimen in this phase II study, we decided to incorporate this combination in the phase III trial currently underway. In this article, we present the results of the interim analysis.

Patients and Methods

Eligibility Criteria

Chemotherapy-naive patients with histologically or cytologically proven locally advanced (stage IIIB) or metastatic (stage IV) NSCLC were eligible for this trial. Patients had to be £ 70 years of age and have adequate bone marrow function (absolute neutrophil count ³ 2 × 109/L, platelet count ³ 100 × 109/L, and hemoglobin level ³ 10.0 g/L), adequate liver function (bilirubin level < 2 times the upper limit of normal, oxaloacetic transaminase enzyme (AST) and/or alanine aminotransferase (ALT) < 3 times the upper limit of normal, prothrombin time < 1.5 times control), and creatinine clearance ³ 60 mL/min. The presence of severe cardiac arrhythmia or heart failure, second- or third-degree heart block, or acute myocardial infarction within 4 months prior to study entry were considered exclusionary. Central nervous system metastases, if asymptomatic, were not considered exclusionary.

 Patients were also required to have a performance status £ 1 on the Eastern Cooperative Oncology Group (ECOG) scale, and a life expectancy of at least 12 weeks. All patients gave their written informed consent, and the trial was approved by the Ethical Committee for Biologic Research of the National Tumor Institute of Naples.

Pretreatment Evaluation

The pretreatment evaluation included a complete history and physical examination, electrocardiogram, chest x-ray, respiratory tests, fiberoptic bronchoscopy, and computed tomography (CT) of the chest, brain, and upper abdomen. A radionuclide scan of bone was also performed as necessary to document disease extent. Laboratory studies included a complete blood cell count with white blood cell differential and platelet count, full chemistry profile, prothrombin time, partial thromboplastin and thrombin time, and urinalysis.

Treatment Regimens

Patients were randomized to receive: (1) cisplatin 50 mg/m², gemcitabine 1,000 mg/m², and vinorelbine 25 mg/m² on days 1 and 8, with recycling every 3 weeks (PGV arm); (2) cisplatin 100 mg/m² on day 1 and gemcitabine 1,000 mg/m² on days 1, 8, and 15 every 4 weeks (PG arm); (3) cisplatin 120 mg/m² on days 1 and 29 and vinorelbine 30 mg/m²/week for 10 weeks (PV arm), or (4) cisplatin 50 mg/m², gemcitabine 1,000 mg/m², and paclitaxel 125 mg/m² (over 1 hr) on days 1 and 8 every 3 weeks (PGT arm).

All patients received antiemetic prophylaxis consisting of HT3-receptor antagonists plus 20 mg of dexamethasone. In the PGT arm, prophylaxis for hypersensitivity to paclitaxel consisted of dexamethasone 20 mg intravenously (IV), ranitidine (Zantac) 50 mg IV, and promethazine 50 mg intramuscularly (IM), administered 30 minutes before the start of paclitaxel infusion.

Response and Toxicity Evaluation

Study patients underwent assessment for tumor response after a similar period of induction chemotherapy (after three cycles in the PGV and PGT arms and two cycles in the PG and PV arms). An additional two or three cycles, respectively, were administered only in patients achieving a complete or partial response according to World Health Organization (WHO) criteria. Chest irradiation was performed (if indicated) in stage IIIB patients showing stable disease and in progressing patients when considered useful by the physician.

The WHO grade scale was also used to record toxicity. Toxicity was assessed prior to each cycle of chemotherapy, and hematologic studies were performed weekly to determine toxicity at nadir. Data showing severest toxicity for each patient in all cycles of chemotherapy were used for toxicity analysis.

Statistical Methods and Study Design

Survival was the primary end point of the study. The study has a power of 80% to recognize a 50% prolongation in the median survival of patients receiving the triple-drug regimens when compared with PV and PG (12 months vs 8 months). Each arm must enroll about 120 people. An interim analysis was planned after the first 60 patients in each arm had undergone a minimum follow up at 26 weeks. The study would be concluded early if the triple-drug regimen did not demonstrate a statistically significant reduction in the risk of death (P < .44), when compared with either PV or PG (null hypothesis accepted). We also planned to close accrual early in any two-drug arm that demonstrated a significant increase in the risk of death (P < .01) as compared with one or both of the triple-drug regimens (null hypothesis rejected).[20]

Overall survival was measured from the date of study entry to the date of death or last follow up. Survival curves were estimated using the Kaplan-Meier product-limit method,[21] and compared by Cox analysis,[22] with age (< 65 years vs older), performance status (0 vs 1), stage (IIIB vs IV), histology (squamous vs others), and weight loss (> 5% vs < 5% of body weight) as covariates.

Patient accrual in the PGV, PG, and PV arms began in April 1997. From October 1997 on, patients were also randomly assigned to the PGT arm. By January 16, 1999, 286 patients were enrolled in the study; 10 patients did not meet the eligibility criteria, and 9 were not evaluable because of the absence of demographics and follow-up data. On April 15, 1999, we performed an interim analysis of the survival of the first 60 evaluable patients in each arm.



The characteristics of the first 240 eligible and evaluable patients assigned to the four arms are summarized in Table 1. The majority of patients were male (92%), and the median age was 62 years (range 32 to 70). About 60% of patients had stage IV disease, and half had a squamous histology. At diagnosis, a weight loss of more than 5% was registered in about 25% of patients. Brain metastases were present in 10 patients. Main baseline characteristics were well balanced across the four arms.


As of April 15, 1999, 128 (PGV = 33, PG = 42, PV = 53) patients had died in the original three arms of the study. Median follow-up was 69 weeks (range 26 to 102 weeks). Actuarial survival curves are shown in Figure 1. Median survival was 51 weeks in the PGV arm, 42 weeks in the PG arm, and 35 weeks in the PV arm. The 1-year probability of survival was 45%, 40%, and 34% for each arm, respectively. At multivariate Cox regression analysis, the estimated risk of death for patients receiving PGV compared with those receiving PV was 0.35 (95% confidence interval [CI], 0.16-0.77; P = .0058).

Of the 60 patients receiving the PGT regimen, 23 had died as of April 15, 1999. Median follow up was 48 weeks (range 13 to 80 weeks); 12 patients were at risk of death at 1 year. The survival curve for this arm is illustrated in Figure 2. The median survival time has not yet been reached, with a survival probability of 81% at 6 months and 58% at 1 year.


The response data are summarized in Table 2. There were two complete and 26 partial responses in the PGV arm, for an overall response rate of 47% (95% CI, 34%-60%). No complete responses occurred in the PG and PV arms, where the overall response rates were 30% (95% CI, 19%-43%) and 25% (95% CI, 15%-38%), respectively. Overall response rates according to stage IIIB vs IV were 54% and 42% in the PGV arm, 36% and 26% in the PG arm, and 29% and 22% in the PV arm, respectively.

Failures occurred within 3 months in 25%, 33%, and 50% of patients in the PGV, PG, and PV arms, respectively. Two complete and 33 partial responses were recorded in the PGT arm, for an overall response rate of 58% (95% CI, 45%-71%).

The overall response rates for patients with stage IIIB and stage IV disease were 63% and 50%, respectively. An additional 14 patients achieved stable disease after three cycles, while 11 patients failed treatment.

Treatment Compliance and Toxicity

Overall, 2 patients in the PV arm died as a result of treatment-related toxicity. Chemotherapy was discontinued early due to toxicity in 3 patients in the PGV arm, 3 patients in the PG arm, and 10 patients in the PV arm. For patients who received the PGT regimen, only 3 failed to complete the first three treatment cycles. Severe hematologic and nonhematologic toxicity events associated with these regimens are reported in Table 3.


In light of the very promising results obtained in phase II studies, the cis-platin/vinorelbine,[4,5,16] cisplatin or carboplatin (Paraplatin)/paclitaxel,[6-9] or cisplatin/gemcitabine regimens have been widely used.[10,11] Unfortunately, subsequent phase III trials did not demonstrate that these two-drug regimens substantially modified the prognosis of advanced NSCLC.[12-14,16]

The purpose of the present phase III trial was to evaluate whether one or both of the new triple-drug combinations (PGV and PGT) we devised could produce a meaningful survival gain when compared with the two-drug combinations, PG and PV. This is the first report of the comparison of a three-drug regimen (including two new drugs) with standard two-drug combinations. On the basis of the survival data of this interim analysis, we concluded that the PGV regimen produced a significantly longer median survival compared with the PV regimen.

However, a definitive conclusion about the comparison with the PG combination cannot be made yet. We wish to point out that the wide difference between survival curves of the PGV and PV arms was not due to an underestimation of the activity of the PV regimen in the present study. Indeed, both median and 1-year survival rates observed in the present study were superimposable on those reported in the two consecutive Southwest Oncology Group (SWOG) trials.[6,18]

On the other hand, the survival curve of patients who received PGV in our study was absolutely identical to the survival curve we previously reported in a phase II study in a similar population.[18] Notably, these results were obtained in case series with good prognosis. Elderly or unfit patients were not included in these studies, because we strongly believe that a tailored therapeutic approach is advisable for frail patients.

Our study included a higher proportion of patients with locally advanced disease compared to the SWOG trials.[5,16] However, it must be noted that the superiority of the triple-drug combinations becomes even more evident when only patients with stage IV disease are considered. Indeed, PGV yielded a median survival more than 4 months longer than that of patients in the PV arm.

In this article, we have also provided data concerning patients who received the PGT combination, although the randomization in this arm started 6 months after the phase III trial began. Because that delay required a different follow-up duration, survival results in this arm are not mature enough to reach firm conclusions.

The triple- and double-drug regimens also produced widely different overall response rates. To provide a similar treatment duration (8 to 10 weeks) before restaging, we decided to administer two cycles of PV or PG and three cycles of PGV or PGT, and to stop treatment in nonresponding patients. The overall response rates of patients in the PV (25%) and PG (30%) arms in our study were slightly lower than those reported in previous randomized studies.[4,5,12,13,16] However, the cumulative dose of gemcitabine planned in the PG arm was identical to that scheduled in the PGV and PGT arms, and the vinorelbine cumulative dose in the PV arm was much higher than that projected in the PGV regimen. Moreover, even in responding patients, the addition of three more cycles of chemotherapy did not result in a clear improvement in survival.[23]

In general, the toxicity of both triple-drug regimens was manageable. Only three patients in the PGV and PGT arms had to suspend treatment early because of toxicity. That compares favorably with the compliance of patients in the PV arm.

Indeed, the high dose of cisplatin given in a single day and the weekly administration of full doses of vinorelbine frequently caused patients in this arm to refuse treatment. Severe emesis, constipation, fatigue, and sometimes renal toxicity were the main reasons for noncompliance. In addition, the incidence of severe neutropenia was significantly more frequent in the PV than in the PGV arm.

A relevant nonhematologic toxicity was sometimes associated with the PGT regimen. In particular, severe fatigue and a troublesome peripheral neuropathy were observed in 30% and 44% of patients in this arm, respectively.

Therefore, only a final comparison of the quality of life of patients in our study will establish whether the rapid tumor shrinkage and symptom control may be counterbalanced by a worse toxicity profile.


According to the results of the present interim analysis, the triple-drug PGV combination produced a significant gain in survival in patients with advanced NSCLC compared with cisplatin and vinorelbine treatment. The triple-drug PGT regimen seems to be associated with an even better survival gain, although caution is required considering the shorter follow up of these patients.

 In view of these data, accrual in the PV arm has been closed, while the other arms will remain open until the planned sample is reached. This will allow us to evaluate whether the two triple-drug regimens can produce a significant survival gain compared with the cisplatin and gemcitabine regimen.

Acknowledgments: We wish to thank Dr. John Perchard for the statistical analysis, Dr. Sabrina Caiazzo for the data management, and Ms. Liliana Gallifuoco for her secretarial assistance.


1. Gralla R: New directions in non–small-cell lung cancer. Semin Oncol 17(suppl 7):20-29, 1990.

2. Lilenbaum RC, Green MR: Novel chemotherapeutic agents in the treatment of non–small-cell lung cancer. J Clin Oncol 11:1391-1402, 1993.

3. Non–Small-Cell Lung Cancer Cooperative Group: Chemotherapy in non–small-cell lung cancer: A meta-analysis using updated data on individual patient data from 52 randomised clinical trials. Br Med J 311:899-909, 1995.

4. Le Chevalier T, Brisgand D, Douillard JY, et al: Randomized study of vinorelbine and cis-platin vs vindesine and cisplatin vs vinorelbine alone in advanced non–small-cell lung cancer: Results of a European Multicenter Trial including 612 patients. J Clin Oncol 12:360-366, 1994.

5. Wozniak AJ, Crowley JJ, Balcezark SP, et al: Randomized trial comparing cisplatin with cisplatin plus vinorelbine in the treatment of advanced non–small-cell lung cancer (NSCLC): A Southwest Oncology Group study. J Clin Oncol 16:2459-2465, 1998.

6. Klastersky J, Sculier JP: Dose-finding of paclitaxel (Taxol) plus cisplatin in patients with non–small-cell lung cancer. Lung Cancer 12 (suppl 2):S117-S125, 1995.

7. Langer CJ, Leighton JC, Comis RL, et al: Paclitaxel and carboplatin in combination in the treatment of advanced non–small-cell lung cancer: A phase II toxicity, response, and survival analysis. J Clin Oncol 13:1860-1870, 1995.

8. Bonomi P, Kim K, Kluger J, et al: Comparison of survival for stage IIIB vs IV non–small-cell lung cancer patients treated with etoposide/cisplatin vs Taxol/cisplatin: An Eastern Cooperative Group trial (abstract 1631). Proc Am Soc Clin Oncol 16:454a, 1997.

9. Giaccone G, Splinter TA, Debruyne C, et al: Randomized study of paclitaxel-cisplatin vs cisplatin/teniposide in patients with advanced non–small-cell lung cancer. The European Organization for Research and Treatment of Cancer Lung Cancer Cooperative Group. J Clin Oncol 16(6):2133-2141, 1998.

10. Crinò L, Scagliotti G, Marangolo M, et al: Cisplatin/gemcitabine combination in advanced non–small-cell lung cancer. A phase II study. J Clin Oncol 15:297-303, 1997.

11. Abratt RP, Bezwoda WR, Goedhals L, et al: Weekly gemcitabine with monthly cisplatin: Effective chemotherapy for advanced non–small-cell lung cancer. J Clin Oncol 15:744-750, 1997.

12. Cardenal F, Cabrerizo PL, Anton A, et al: Randomized phase III study of Gemcitabine/cisplatin vs etoposide/cisplatin in the treatment of locally advanced or metastatic non–small-cell lung cancer. J Clin Oncol 17:12-18, 1999.

13. Crinò 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 17:3522-3530, 1999.

14. Belani CP, Natale RB, Lee JS, et al: Randomized phase III trial comparing cisplatin/etoposide vs carboplatin/paclitaxel in advanced and metastatic non–small-cell lung cancer (abstract 1751). Proc Am Soc Clin Oncol 17:455a, 1998.

15. Sandler A, Nemunaitis J, Dehnam C, et al: Phase III study of cisplatin with or without gemcitabine in patients with advanced NSCLC (abstract 1747). Proc Am Soc Clin Oncol 17:454a, 1998.

16. Kelly K, Crowley J, Bunn PA, et al: A randomised phase III trial of paclitaxel plus carboplatin vs vinorelbine plus cisplatin in untreated advanced non–small-cell lung cancer (NSCLC). A Southwest Oncology Group (SWOG) trial (abstract 1777). Proc Am Soc Clin Oncol 18:461a, 1999.

17. Frasci G, Panza N, Comella P, et al: Cis-platin, gemcitabine and vinorelbine in locally advanced or metastatic non–small-cell lung cancer. Ann Oncol 8:1045-1048, 1997.

18. Comella P, Frasci G, Panza N, et al: Cis-platin, gemcitabine and vinorelbine combination therapy in advanced non–small-cell lung cancer (NSCLC). A phase II randomised study of the Southern Italy Cooperative Oncology Group phase II randomised study. J Clin Oncol 17:1525-1534, 1999.

19. Frasci G, Panza N, Comella P, et al: Cis-platin, gemcitabine and paclitaxel in locally advanced or metastatic non–small-cell lung cancer. A phase I-II study. J Clin Oncol 17(8):2316-2325, 1999.

20. Schaid DJ, Wieand S, Therneau TM: Optimal two-stage screening designs for survival comparisons. Biometrika 77:507-513, 1990.

21. Kaplan ES, Meier P: Non parametric estimation for incomplete observations. J Am Stat Assoc 53:557-580, 1958.

22. Cox DR: Regression models and life tables. J Royal Stat Assoc 34:187-220, 1972.

23. Smith IE, O’Brien MER, Norton A, et al: Duration of chemotherapy for advanced non–small-cell lung cancer. A phase III randomised trial of 3 vs 6 courses of mitomycin C, vinblastine, cisplatin (abstract 1759). Proc Am Soc Clin Oncol 17:457a, 1998.