Lung Cancer (2007) 58, 73—79

avai lab le at www.sc iencedi rec t .com

journa l homepage: www.e lsev ier .com/ locate / lungcan

Phase II trial of carboplatin and paclitaxel innon-small cell lung cancer patients previouslytreated with chemotherapy

Kiyotaka Yoh ∗, Kaoru Kubota, Ryutaro Kakinuma, Hironobu Ohmatsu,Koichi Goto, Seiji Niho, Nagahiro Saijo, Yutaka Nishiwaki

Division of Thoracic Oncology, National Cancer Center Hospital East, 6-5-1 Kashiwanoha, Kashiwa, Chiba 277-8577, Japan

Received 14 November 2006; received in revised form 9 February 2007; accepted 20 April 2007

KEYWORDSNon-small cell lungcancer;Carboplatin;Paclitaxel;Chemotherapy;Second-linetreatment;

Summary The purpose of this phase II trial was to evaluate the efficacy and toxicity ofcarboplatin plus paclitaxel in the treatment of advanced non-small cell lung cancer (NSCLC)previously treated with chemotherapy. Patients with a performance status (PS) of 0 or 1 whohad received one or two previous chemotherapy regimens for advanced NSCLC were eligible.Paclitaxel 200 mg/m2 was infused over 3 h and followed by carboplatin (area under the curve 6)infusion over 1 h, once every 3 weeks. Thirty patients were enrolled. A complete response wasobserved in 1 patient and a partial response in 10 patients, for an overall response rate of 36.7%.The median time to progression was 5.3 months. The median survival time was 9.9 months, and

Toxicity the 1-year survival rate was 47%. Hematological toxicity in the form of grade 3/4 neutropeniaoccurred in 54%, but grade 3 febrile neutropenia developed in only 3%. Non-hematological grade3 toxicities were less frequent. There were no treatment-related deaths. The combination ofcarboplatin plus paclitaxel is an active and well-tolerated regimen for the treatment of NSCLCpatients who have previously been treated with chemotherapy and have a good PS.

. All

IN

© 2007 Elsevier Ireland Ltd

1. Introduction

Lung cancer remains a major cause of death from cancerin many countries. More than half of all patients diagnosedwith non-small cell lung cancer (NSCLC) have advanced stage

∗ Corresponding author. Tel.: +81 4 7133 1111;fax: +81 4 7131 4724.

E-mail address: kyoh@east.ncc.go.jp (K. Yoh).

bts[pc

tb

0169-5002/$ — see front matter © 2007 Elsevier Ireland Ltd. All rights redoi:10.1016/j.lungcan.2007.04.015

rights reserved.

IIB or IV disease at presentation, and patients with advancedSCLC are candidates for systemic chemotherapy. Platinum-ased chemotherapy is considered the standard first-linereatment for patients with advanced NSCLC, and prolongsurvival, palliates symptoms, and improves quality of life1,2]. Many patients with good performance status (PS) when

rogression occurs after first-line chemotherapy are suitableandidates for second-line chemotherapy [3].

The taxanes are an important class of new agents forhe treatment of advanced NSCLC. Paclitaxel, in com-ination with carboplatin, is the most common regimen

served.

7

utibeopasi[

taoasbvtdspapatttpcNIc[

tpt

2

2

TartooiwttwEhipab

Eetsddcbtwt

2

Ts5lChp(biaapmtoarncAaa

pPbasAroc

2

OrwCa

4

sed as first-line chemotherapy for advanced NSCLC, andhis combination has a more favorable toxicity profile ands more convenient to administer than other platinum-ased regimens [4,5]. Docetaxel has been investigated morextensively than any other agent for second-line treatmentf advanced NSCLC, and the results of two randomizedhase III trials of second-line chemotherapy in patients withdvanced NSCLC demonstrated that docetaxel monotherapyignificantly improved survival compared with best support-ve care or other single agents (vinorelbine or ifosfamide)6,7].

Belani et al. recently reported that results of a phase IIIrial comparing a carboplatin plus paclitaxel regimen withcisplatin plus etoposide regimen for first-line treatment

f advanced NSCLC [8]. Carboplatin plus paclitaxel yieldedhigher response rate (23% versus 15%), time to progres-

ion (121 days versus 111 days), and overall quality of lifeenefit than cisplatin plus etoposide, but the median sur-ival time was better in the cisplatin plus etoposide armhan in the carboplatin plus paclitaxel arm (274 days and 233ays, respectively [P = 0.086]). The authors reported that aubstantially greater proportion of patients in the cisplatinlus etoposide arm received second-line chemotherapy withtaxane-containing regimen than in the carboplatin plus

aclitaxel arm, and suggested that treatment with tax-nes in a second-line setting may have had an impact onhe survival in their study. Remarkably, more than half ofhe regimens that were used in the second-line setting ofheir study consisted of paclitaxel alone or carboplatin plusaclitaxel, not docetaxel. While the efficacy of paclitaxel-ontaining regimens as first-line chemotherapy for advancedSCLC has been established in many randomized phase

II trials [9], the data on the efficacy of paclitaxel-ontaining regimens in second-line settings are limited10,11].

Based these considerations we conducted a phase II trialo evaluate the efficacy and toxicity of carboplatin plusaclitaxel in the treatment of advanced NSCLC previouslyreated with chemotherapy.

. Patients and methods

.1. Eligibility criteria

he inclusion criteria were: pathologically confirmeddvanced NSCLC patients with measurable disease who hadeceived one or two previous chemotherapy regimens forheir disease. Patients were required to submit evidencef failure of prior chemotherapy. Patients who were previ-usly treated with carboplatin or paclitaxel were excludedf the best response was progressive disease (PD). Patientsho had received prior radiotherapy were eligible provided

hat at least 30 days had elapsed between the comple-ion of radiotherapy and entry into the study. Patientsere also required to be 20—75 years of age, have anastern Cooperative Oncology Group PS of 0 or 1, and

ave adequate organ function as indicated by the follow-ng parameters: absolute neutrophil count ≥1500 mm−3,latelet count ≥100,000 mm−3, hemoglobin ≥9.0 g/dl, ASTnd ALT ≤2.0 × the institutional upper normal limits, totalilirubin ≤1.5 mg/dl, creatinine ≤1.5 mg/dl, PaO2 ≥65 Torr.

struv

K. Yoh et al.

xclusion criteria were: uncontrolled pleural or pericardialffusion, active concomitant malignancy, prior irradiationo areas encompassing more than a third of the pelvis pluspine, active infection, myocardial insufficiency or myocar-ial infarction within the preceding 6 months, uncontrollediabetes mellitus or hypertension, any other condition thatould compromise protocol compliance, pregnancy and/orreast-feeding. All patients were required to provide writ-en informed consent before entry into the study. The studyas approved by the institutional review board of our insti-

uion.

.2. Treatment plan

reatment was started within a week of entry into thetudy. Patients received paclitaxel 200 mg/m2 diluted in00 ml of 0.9% saline as a 3-h intravenous infusion fol-owed by carboplatin (area under the curve [AUC] 6;alvert formula) diluted in 250 ml of 5% glucose as a 1-intravenous infusion, every 3 weeks. All patients were

remedicated with dexamethasone (24 mg i.v.), famotidine20 mg i.v.), and diphenhydramine (50 mg orally) 30 minefore the paclitaxel infusion to prevent a hypersensitiv-ty reaction. A 5-HT3-receptor antagonist was intravenouslydministered as an antiemetic before carboplatin. Ther-py was continued for at least two cycles unless theatient experienced unacceptable toxicity or had PD. Theaximum number of cycles of chemotherapy was six. In

he event of grade 4 leukopenia or thrombocytopenia orf grade 3 neutropenic fever, the dose of carboplatinnd paclitaxel was reduced to AUC 5 and 175 mg/m2,espectively, in the following cycle of chemotherapy. Theext cycle of chemotherapy was started if the neutrophilount was ≥1500 mm−3, the platelet count ≥100,000 mm−3,ST and ALT ≤100 IU/l, total bilirubin ≤2.0 mg/dl, cre-tinine ≤1.5 mg/dl, PS 0 or 1, and the patient wasfebrile.

Pretreatment evaluation included a medical history, ahysical examination, vital signs, height and body weight,S, complete blood count, biochemical studies, arteriallood gas analysis, electrocardiogram, chest radiographnd computed tomography scan (CT), abdominal ultra-ound or CT, and brain magnetic resonance imaging or CT.

complete blood count, biochemical studies, and chestadiograph were performed weekly during the first cyclef chemotherapy, and 2 weekly starting with the secondycle.

.3. Response and toxicity assessment

bjective tumor response was assessed as completeesponse (CR), partial response (PR), stable disease ≥8eeks (SD), or PD according to the Response Evaluationriteria in Solid Tumors. Measurable lesions were defineds lesions whose longest diameter was ≥2 cm. Imaging

tudies were repeated every 4 weeks until the objec-ive tumor response was confirmed. All responses wereeviewed by an independent radiologist. Toxicity was gradedsing National Cancer Institute-Common Toxicity Criteriaersion 2.0.

75

Table 1 Patient characteristics

Characteristic No. of patients (%)

Patients enrolled 30

SexMale 26Female 4

Age, yearsMedian 60Range 39—75

ECOG performance status0 71 23

StageIIIB 11IV 19

HistologyAdenocarcinoma 21Squamous cell carcinoma 7Large cell carcnioma 2

Prior treatmentPlatinum-based chemotherapy 28 (93)Docetaxel 5 (16)Chest radiotherapy 4 (13)

No. of prior chemotherapy regimens

Taanpmpatients who had received two regimens (P > 0.99). Accord-ing to the treatment-free interval since the final dose of theprior chemotherapy regimen, the response rate of patientswhose interval was 3 months or more was 33% (7/21), com-

Table 2 Treatment efficacy (n = 30)

No. of patients %

ResponseOverall response rate 11 36.7

Complete response 1 3.3Partial response 10 33.3Stable disease 12 40Progressive disease 5 16.7Not evaluable 2 6.7

Survival

Carboplatin plus paclitaxel as second-line chemotherapy

2.4. Statistical analysis

The primary endpoint of this study was the response rate,defined as the proportion of patients whose best responsewas CR or PR among all enrolled patients in the intent-to-treat analysis. The secondary end points were toxicity andoverall and progression-free survival (PFS) from the date ofenrollment in this study.

According to Simon’s minimax two-stage phase II studydesign, the treatment program was designed for a minimalresponse rate of 5% and to provide a significance level of 0.05with a statistical power of 80% in assessing the activity of theregimen according to a 20% response rate. The upper limitfor first-stage drug rejection was no response in 13 evaluablepatients. The upper limit for second-stage drug rejectionwas three responses in 27 evaluable patients. Overall sur-vival time was defined as the interval between enrollment inthis study and death or the most recent follow-up visit. PFSwas defined as the interval between enrollment in this studyand the first documented PD, death, or the most recentfollow-up visit. Survival was estimated by the Kaplan—Meieranalysis method. All comparisons between proportions wereperformed by Fisher’s exact test.

3. Results

3.1. Patient characteristics

Between October 2002 and November 2003, 30 patientswere enrolled in this study, and their characteristics areshown in Table 1. Twenty-six (87%) patients were men, and21 (70%) patients had adenocarcinoma. Median age was 60years. The majority of the patients (93%) had received priorplatinum-based chemotherapy, and seven (23%) patients hadreceived two prior chemotherapy regimens. The platinum-based chemotherapy regimens that had been used were:cisplatin plus vinorelbine (n = 26), cisplatin plus gemcitabine(n = 1), and carboplatin plus gemcitabine (n = 1). There were15 (50%) responders to any of the prior chemotherapy reg-imens and 12 of them had experienced a response (CR/PR)to cisplatin-based chemotherapy. Twenty-one (70%) patientshad a treatment-free interval of 3 or more months since thefinal dose of the prior chemotherapy regimen.

A total of 94 cycles of chemotherapy were administered,and the median number of cycles per patient was three(range, 1—6). Four patients had received only one cycle oftreatment either because of toxicity (two patients, grade3 rash), the patient’s refusal (one patient), or PD (onepatient).

3.2. Response and survival

Two patients were not evaluable for response because theprotocol treatment had been terminated because of toxi-city (grade 3 rash) during the first cycle of chemotherapy,

and they subsequently received further chemotherapy with-out PD. There was 1 CR and 10 PRs among the 30 patients,and the objective response rate in the intent-to-treat anal-ysis was 36.7% (95% confidence interval [CI], 19.9—56.1%)(Table 2). Treatment outcomes of all patients are listed in

1 232 7

able 3. The response rate of patients who experiencedresponse (CR/PR) to prior cisplatin-based chemother-

py was 43% (6/14), as opposed to 23% (3/13) among theon-response patients (P = 0.41). The response rate of theatients who had received one prior chemotherapy regi-en was 39% (9/23), as opposed to 28% (2/7) among the

Median (months) 9.91 year (%) 47

Progression-free survivalMedian (months) 5.3

76K.

Yohet

al.

Table 3 Treatment outcomes of all patients

Patient No. Prior first-line therapy Prior second-line therapy Time from lasttherapy (months)

CBDCA + PTX,best response

PFS (months) Survival(months)

Regimen Best response Regimen Best response

1 CDDP + VNR SD DOC PD 1.8 SD 1.4 25.22 CBDCA + GEM NE Gefitinib PD 0.8 PR 3.8 8.83 CDDP + VNR SD — 6.8 SD 7.6 18.14 CDDP + GEM PR — 9.5 PR 7.5 33.8+5 CDDP + VNR SD — 4.8 SD 2.8 7.06 CDDP + VNR + DOC + RT PR — 6.0 PR 8.0 21.67 GEM + VNR SD — 23.0 PD 1.2 7.88 CDDP + VNR + RT PR — 13.6 SD 6.7 25.0+9 CDDP + VNR SD — 5.0 SD 2.1 3.7

10 CDDP + VNR SD — 5.0 PD 1.2 6.711 CDDP + VNR PR — 8.9 NE 1.1 3.312 CDDP + VNR SD Gefitinib CR 1.9 SD 6.3 6.313 CDDP + VNR PR — 5.4 NE 1.0 13.414 CDDP + VNR PR — 1.7 SD 4.8 5.715 CDDP + VNR + RT PR — 9.3 SD 5.0 15.716 CDDP + VNR SD — 2.8 PR 3.7 15.817 CDDP + VNR SD DOC + GEM SD 3.8 SD 5.3 21.6+18 CDDP + VNR + DOC + RT PR — 3.9 SD 4.5 9.019 CDDP + VNR PR — 12.9 PR 9.4 16.020 CDDP + VNR PR — 11.5 CR 24.8+ 24.821 CDDP + VNR PD — 1.1 PR 9.2 23.6+22 CDDP + VNR SD DOC SD 4.5 PD 2.3 5.523 Gefitinib SD — 0.9 PR 8.8 12.724 CDDP + VNR PR — 11.1 PR 5.3 10.225 CDDP + VNR PR Gefitinib PR 4.4 PR 5.5 9.926 CDDP + VNR NE — 11.7 PR 7.0 12.227 CDDP + VNR PR — 5.4 SD 6.2 9.428 CDDP + VNR SD — 0.8 PD 1.4 2.529 CDDP + VNR PR — 4.4 PD 0.2 8.430 Gefitinib PD CDDP + VNR PD 0.9 SD 3.1 3.3

CBDCA, carboplatin; PTX, paclitaxel; PFS, progression-free survival; CDDP, cisplatin; VNR, vinorelbine; GEM, gemcitabine; DOC, docetaxel; RT, chest radiotherapy; SD, stable disease; NE,not evaluable; PR, partial response; PD, progressive disease; CR, complete response.

Carboplatin plus paclitaxel as second-line chemotherapy

Fig. 1 Kaplan—Meier curve for overall survival.

agtat

4

Dfor5lrrrmbg[pdow4otvpaasf

pb[paat3cpaO

Fig. 2 Kaplan—Meier curve for progression-free survival.

pared with 44% (4/9) in patients in whom it was less than 3months (P = 0.68).

The median follow-up time was 24 months. The mediansurvival time (MST) was 9.9 months (range, 2.5—33.8months), and the 1-year survival rate was 47% (95%CI, 29—65%). The median PFS was 5.3 months. TheKaplan—Meier curve for overall survival and for PFS is shownin Figs. 1 and 2, respectively. Nineteen patients (63%)received at least one subsequent chemotherapy regimen,and their regimens are shown in Table 4. Fourteen of themwere treated with gefitinib, and a PR was achieved in threeof them.

3.3. Toxicity

The common toxicities associated with carboplatin plus

paclitaxel are listed in Table 5. Grade 3/4 neutropeniaoccurred in 54% of the patients in our study, but grade 3febrile neutropenia developed in only 3%. Grade 3/4 anemiaand thrombocytopenia were observed in five patients (16%)

Table 4 Post-study chemotherapy

Regimen No. ofpatients

Responder(%)

Gefitinib 14 3 (21)Docetaxel 9 0Gemcitabine plus viborelbine 1 0

tt35HiNmpsaboe

77

nd two patients (13%), respectively. Non-hematologicalrade 3 toxicities were less frequent. Grade 3 hypona-remia was observed in five (16%) patients, but they werell asymptomatic. Grade 2 neuropathy occurred in 33% ofhe patients. There were no treatment-related deaths.

. Discussion

ocetaxel, pemetrexed, and erlotinib have been approvedor second-line treatment of advanced NSCLC on the basisf the results of phase III trials [6,7,12,13]. Hanna et al.eported a phase III study comparing 3-weekly pemetrexed00 mg/m2 with 3-weekly docetaxel 75 mg/m2 as second-ine treatment for advanced NSCLC. The overall responseate with pemetrexed and docetaxel was 9.1% and 8.8%,espectively, and MST was 8.3 months and 7.9 months,espectively. Although efficacy in terms of the outcome aseasured by survival time and response rate was similar foroth treatments, the pemetrexed group experienced lessrades 3—4 hematological toxicity and alopecia of all grades12]. In the trial reported by Shepherd et al. 731 NSCLCatients previously treated with chemotherapy were ran-omized to receive either erlotinib at a dose of 150 mg dailyr placebo, and the response rate in the erlotinib groupas 8.9%. MST was 6.7 months in the erlotinib group and.7 months in the placebo group (P < 0.001). The resultsf their trial showed that erlotinib significantly prolongedhe survival of patients with advanced NSCLC who had pre-iously been treated with chemotherapy [13]. Despite theositive results of these phase III trials, the response rate ofdvanced NSCLC to second-line chemotherapy remains low,nd the life expectancy of advanced NSCLC patients remainshort. Alternative effective chemotherapy option is neededor second-line treatment of advanced NSCLC.

The combination of carboplatin plus paclitaxel hasroved effective as one of the standard platinum-based dou-let regimens for first-line treatment of advanced NSCLC4,5,14]. However, since the efficacy of carboplatin plusaclitaxel used in a second-line setting had hardly beenssessed, in the present study we evaluated the efficacynd toxicity of carboplatin plus paclitaxel in the second- orhird-line treatment of advanced NSCLC. The results in the0 patients with advanced NSCLC previously treated withhemotherapy indicated that the combination of carboplatinlus paclitaxel yielded an objective response rate of 36.7%nd an MST of 9.9 months, with a 1-year survival rate of 47%.ur study had not included patients who were treated withhe platinum/taxane combination chemotherapy. Most ofhe toxicity observed in our study was hematological. Grade/4 neutropenia, anemia, or thrombocytopenia occurred in4, 16, or 13% of the patients in our study, respectively.ematological toxicity of carboplatin plus paclitaxel used

n first-line treatment for Japanese patients with advancedSCLC has been reported that grade 3/4 neutropenia, ane-ia, or thrombocytopenia occurred in 88, 15, or 11% of theatients [15]. The toxicity observed in our study appeared

imilar to that of carboplatin plus paclitaxel, which wasdministered as the first-line treatment, although the num-er of patients in our study was not large. The combinationf carboplatin plus paclitaxel seems to be effective and tol-rable, not only as first-line therapy for advanced NSCLC but

78 K. Yoh et al.

Table 5 Hematological and non-hematological toxicity (n = 30)

Toxicity NCI-CTC Version 2.0, grade

0—1 2 3 4

n % n % n % n %

Leukopenia 11 37 10 33 9 30 0 0Neutropenia 10 33 4 13 14 47 2 7Anemia 7 23 18 60 3 10 2 7Thrombocytopenia 27 90 1 3 2 7 0 0Febrile neutropenia 29 97 — 1 3 0 0Nausea 27 90 3 10 0 0 —Fatigue 30 100 0 0 0 0 0 0Neuropathy 20 67 10 33 0 0 0 0Arthralgia 21 70 8 27 1 3 0 0Rash 28 93 0 0 2 6 0 0Infection 29 97 0 0 1 3 0 0Arrhythmia 29 97 0 0 1 3 0 0Alopecia 21 70 9 30 — —AST/ALT 29 97 1 3 0 0 0 0

apc

dasmipsemraNcwcfiotbpc

oimotpThcwT

scd

C

N

A

TM

R

Hyponatremia 25 83 —

s second-line therapy as well if the patients had not beenreviously treated with the platinum/taxane combinationhemotherapy.

Hotta at al. reported a meta-analysis based on abstractedata to compare the effect of carboplatin-based chemother-py with that of cisplatin-based chemotherapy on overallurvival, response rate, and toxicity in the first-line treat-ent of patients with advanced NSCLC [16]. The results

ndicated that combination chemotherapy consisting of cis-latin plus a third generation agent produced a significanturvival benefit compared with carboplatin plus a third gen-ration agent, although the toxicity profiles of the twoodalities were quite different. Recently, Pignon et al.

eported a pooled analysis from five randomized clinical tri-ls of cisplatin-based chemotherapy in completely resectedSCLC patients [17]. Their analysis suggested that adjuvantisplatin-based chemotherapy improved survival in patientsith NSCLC. Based on the results of their meta-analysis,isplatin-based chemotherapy should be recommended asrst-line therapy for patients with advanced NSCLC. More-ver, in view of the results of our own study, we speculatehat the combination of carboplatin plus paclitaxel maye suitable as second-line treatment for advanced NSCLCatients who had experienced progression after first-lineisplatin-based chemotherapy.

Care must be exercised in interpreting the favorableutcome in our study. One concern is that it was a single-nstitution phase II study, and therefore patient selectionay have influenced the outcome. The responders to any

f the prior chemotherapy regimens accounted for 50% ofhe 30 patients enrolled in this study, and about 80% of theatients had received only one prior chemotherapy regimen.

he selection criteria, such as an ECOG PS of 0 or 1, may alsoave contributed to this favorable outcome. Another con-ern is that our study had included only five patients whoere previously treated with chemotherapy using taxanes.herefore, the efficacy of carboplatin plus paclitaxel as the

5 17 0 0

econdary therapy after chemotherapy using taxanes is notlear. A further randomized study is warranted to be able toraw definitive conclusions about our results.

onflict of interest statement

one declared.

cknowledgement

his work was supported in part by a grant from the Bristol-yers Squibb Company.

eferences

[1] Non-small Cell Lung Cancer Collaborative Group. Chemother-apy in non-small cell lung cancer: a meta-analysis usingupdated data on individual patients from 52 randomised clini-cal, trials. BMJ 1995;311:899—909.

[2] Bunn Jr PA, Kelly K. New chemotherapeutic agents prolong sur-vival and improve quality of life in non-small cell lung cancer:a review of the literature and future directions. Clin CancerRes 1998;4:1087—100.

[3] Huisman C, Smit EF, Giaccone G, Postmus PE. Second-linechemotherapy in relapsing or refractory non-small-cell lungcancer: a review. J Clin Oncol 2000;18:3722—30.

[4] Kelly K, Crowley J, Bunn Jr PA, et al. Randomized phase III trialof paclitaxel plus carboplatin versus vinorelbine plus cisplatinin the treatment of patients with advanced non-small-celllung cancer: a Southwest Oncology Group trial. J Clin Oncol2001;19:3210—8.

[5] Schiller JH, Harrington D, Belani CP, et al. Comparison of four

chemotherapy regimens for advanced non-small-cell lung can-cer. N Engl J Med 2002;346:92—8.

[6] Shepherd FA, Dancey J, Ramlau R, et al. Prospective ran-domized trial of docetaxel versus best supportive carein patients with non-small-cell lung cancer previously

[

[

[

[

Carboplatin plus paclitaxel as second-line chemotherapy

treated with platinum-based chemotherapy. J Clin Oncol2000;18:2095—103.

[7] Fossella FV, DeVore R, Kerr RN, et al. Randomized phase IIItrial of docetaxel versus vinorelbine or ifosfamide in patientswith advanced non-small-cell lung cancer previously treatedwith platinum-containing chemotherapy regimens. The TAX320 Non-Small Cell Lung Cancer Study Group. J Clin Oncol2000;18:2354—62.

[8] Belani CP, Lee JS, Socinski MA, et al. Randomized phase IIItrial comparing cisplatin-etoposide to carboplatin-paclitaxel inadvanced or metastatic non-small cell lung cancer. Ann Oncol2005;16:1069—75.

[9] Bunn Jr PA. Chemotherapy for advanced non-small-cell lungcancer: who, what, when, why? J Clin Oncol 2002;20:23—33.

[10] Hainsworth JD, Thompson DS, Greco FA. Paclitaxel by 1-hourinfusion: an active drug in metastatic non-small-cell lung can-cer. J Clin Oncol 1995;13:1609—14.

[11] Sculier JP, Berghmans T, Lafitte JJ, et al. A phase II study testing

paclitaxel as second-line single agent treatment for patientswith advanced non-small cell lung cancer failing after a first-line chemotherapy. Lung Cancer 2002;37:73—7.

[12] Hanna N, Shepherd FA, Fossella FV, et al. Randomized phaseIII trial of pemetrexed versus docetaxel in patients with non-

[

79

small-cell lung cancer previously treated with chemotherapy.J Clin Oncol 2004;22:1589—97.

13] Shepherd FA, Pereira JR, Ciuleanu T, et al. Erlotinib in pre-viously treated non-small-cell lung cancer. N Engl J Med2005;353:123—32.

14] Scagliotti GV, De Marinis F, Rinaldi M, et al. Phase III randomizedtrial comparing three platinum-based doublets in advancednon-small-cell lung cancer. J Clin Oncol 2002;20:4285—91.

15] Ohe Y, Ohashi Y, Kubota K, et al. Randomized phase IIIstudy of cisplatin plus irinotecan versus carboplatin pluspaclitaxel, cisplatin plus gemcitabine, and cisplatin plusvinorelbine for advanced non-small-cell lung cancer: Four-Arm Cooperative Study in Japan. Ann Oncol 2007;18:317—23.

16] Hotta K, Matsuo K, Ueoka H, et al. Meta-analysis of ran-domized clinical trials comparing Cisplatin to Carboplatin inpatients with advanced non-small-cell lung cancer. J Clin Oncol

2004;22:3852—9.

17] Pignon JP, Tribodet H, Scagliotti GV, et al. Lung Adjuvant Cis-platin Evaluation (LACE): a pooled analysis of five randomizedclinical trials including 4584 patients. Proc Am Soc Clin Oncol2006;24:7008.