Fax: (011) 886 2-28763466
Phase II randomized study of daily gefitinib treatment alone or with vinorelbine every 2 weeks in patients with adenocarcinoma of the lung who failed at least 2 regimens of chemotherapy
Article first published online: 9 MAR 2007
Copyright © 2007 American Cancer Society
Volume 109, Issue 9, pages 1821–1828, 1 May 2007
How to Cite
Chen, Y.-M., Liu, J. M., Chou, T.-Y., Perng, R.-P., Tsai, C.-M. and Whang-Peng, J. (2007), Phase II randomized study of daily gefitinib treatment alone or with vinorelbine every 2 weeks in patients with adenocarcinoma of the lung who failed at least 2 regimens of chemotherapy. Cancer, 109: 1821–1828. doi: 10.1002/cncr.22616
- Issue published online: 18 APR 2007
- Article first published online: 9 MAR 2007
- Manuscript Accepted: 10 JAN 2007
- Manuscript Revised: 9 JAN 2007
- Manuscript Received: 4 DEC 2006
- epidermal growth factor receptor;
- nonsmall cell lung cancer;
- salvage therapy;
The objective of this study was to assess the efficacy of adding chronic, intermittent, low-dose vinorelbine to gefitinib treatment for patients who had adenocarcinoma of the lung who failed ≥2 regimens of chemotherapy.
Patients were randomized into 2 arms: Oral gefitinib 250 mg daily (the G arm) or vinorelbine 15 mg/m2 as an intravenous infusion on Day 1 and oral gefitinib 250 mg daily on Days 2 through 14 every 2 weeks (the GV arm). From August 2004 to October 2005, 48 patients were enrolled. Epidermal growth factor receptor (EGFR) exon 18 through 21 nucleotide sequence analysis and fluorescence in situ hybridization were performed in patients who had tumor tissue specimens available for analysis.
After randomization, each arm had 24 patients. However, 3 patients refused vinorelbine treatment and were given gefitinib treatment only. Thus, 27 patients received G treatment, and 21 patients received GV treatment. Objective response rates were 55.6% in the G arm and 52.4% in the GV arm. All toxicities in both arms were mild. The 1-year progression-free survival rate was 57.1% in the GV arm and 21.2% in the G arm (P = .008). The median survival was 13.3 months in the G arm and 23.4 months in the GV arm (P = .1231). Three of 6 patients (50%) had an exon 19 in-frame deletion, and 2 of 10 patients had EGFR gene high polysomy or amplification (20%).
Gefitinib was highly effective in ethnic Chinese patients with adenocarcinoma of the lung who failed previous platinum and taxane treatment. The addition of low-dose vinorelbine every 2 weeks produced a significantly better 1-year progression-free survival rate. Cancer 2007. © 2007 American Cancer Society.
Gefitinib (Iressa; AstraZeneca Pharmaceuticals, Wilmington, Del) is a selective epidermal growth factor receptor (EFGR) tyrosine kinase inhibitor (EGFR-TKI). It is an orally active agent for advanced nonsmall cell lung cancer (NSCLC) in patients who have failed on a previous platinum-based regimen and taxane treatment.1–3 In 2 large Phase II trials (the Iressa Dose Evaluation in Advanced Lung Cancer [IDEAL] 1 study and the IDEAL 2 study), the results indicated that gefitinib had a substantial effect as a salvage treatment for patients who had failed at least 1 or 2 previous regimens of chemotherapy.1, 2
Results from the IDEAL 1 study indicated that Japanese patients had a better response to gefitinib treatment than Caucasian patients; and results from another study, the Iressa Survival Evaluation in Lung Cancer, indicated that East Asian patients had better response and survival than Caucasian patients and patients who received placebo treatment.4 Once it was determined that the EGFR mutation occurred more frequently in East Asian patients than in Caucasians and that patients who had an EGFR mutation had a better response to gefitinib treatment, researchers understood why gefitinib had better efficacy in East Asian patients than in Caucasians.5–8 In a previous study at our hospital, the investigators reported an EGFR tyrosine kinase mutation/deletion rate of approximately 61% in patients with available specimens.9
Whether or not the combined use of chemotherapeutic agent(s) and gefitinib has an additive or synergistic effect is an interesting issue that has been explored continuously in recent years. Large-scale, randomized Phase III studies of first-line combination use of doublet chemotherapy with gefitinib have taken place but have produced negative results.10, 11 Scagliotti et al.12 performed a prospective study combining gefitinib with single-agent gemcitabine (1200 mg/m2 on Days 1 and 8 every 3 weeks) or vinorelbine (30 mg/m2 on Days 1 and 8 every 3 weeks) in elderly patients with NSCLC. Those authors reported that vinorelbine plus gefitinib was too toxic to be tolerated by their patients. However, the response rate and median survival were higher for those patients who received vinorelbine plus gefitinib treatment.12 In addition, a subset analysis of nonsmokers who received carboplatin/paclitaxel with or without erlotinib (another TKI) in the Tarceva Responses in Conjunction with Paclitaxel and Carboplatin (TRIBUTE) study revealed a markedly greater survival benefit for nonsmokers who received chemotherapy plus erlotinib compared with nonsmokers who received chemotherapy alone (median 22.5 months vs 10 months; P = .01).13, 14
When developing a new anticancer drug, an important Phase II objective is to identify the most appropriate target population for a subsequent Phase III trial. In the past, a target population generally was characterized by disease stage and primary site. With molecular targeting drugs, often, this no longer is adequate. Based on the relative good toxicity profiles and high activity observed in our previous Phase II study of treatment with gefitinib alone,15 we decided to conduct a Phase II randomized trial using gefitinib with or without low-dose vinorelbine treatment in ethnic Chinese patients with adenocarcinoma of the lung who failed on previous chemotherapy. Our objective was to investigate whether or not adding vinorelbine to gefitinib treatment would improve the time to disease progression and survival status in a specified group of patients.
MATERIALS AND METHODS
This study was conducted according to existing rules for good clinical practice, and the study protocol was approved by the local ethics committee. Patients with adenocarcinoma of the lung who had failed ≥2 regimens, including taxanes and platinum-based chemotherapy, were entered into the study after they provided informed consent. Eligibility criteria included a histologic or cytologic diagnosis of stage IV adenocarcinoma of the lung in patients who had failed previous chemotherapy with ≥2 regimens (including taxanes and platinum-based chemotherapy); clinically measurable disease, which was defined as bidimensionally measurable lesions; no previous radiotherapy directed at the on measurable lesion(s); adequate bone marrow reserve with a white blood cell count ≥4000/mm3, platelets ≥100,000/mm3, and hemoglobin ≥10 g/dL; and a life expectancy of >2 months. Patients with inadequate liver function (total bilirubin >1.5 times the upper limit of normal and alanine and aspartate aminotransferase levels >3 times the upper limit of normal) or inadequate renal function with creatinine levels >2 mg/dL were excluded from the study.
Baseline evaluations included the patient's history, a physical examination, and a performance score. A complete blood cell count, urinalysis, serum biochemistry profile, electrocardiogram, chest roentgenography, whole-body bone scan, brain computed tomography (CT) scan, and chest (including the liver and adrenal glands) CT scans also were obtained.
Eligible patients were randomized to receive either gefitinib treatment (G arm) or gefitinib plus vinorelbine treatment (GV arm). All patients in the G arm received gefitinib at a fixed daily oral dose of 250 mg. All patients in the GV arm received vinorelbine (15 mg/m2) on Day 1 and gefitinib at a fixed daily oral dose of 250 mg on Days 2 through 14 every 2 weeks. Concomitant use of other chemotherapeutic agents was not allowed. Palliative radiotherapy to the measurable lesion was not used. Treatment was given until patients developed disease progression or of intolerable toxicity.
With regard to dose modifications, gefitinib was stopped temporarily for 1 or 2 weeks if the patient suffered from grade ≥3 toxicity; gefitinib could be restarted from a 50% dose once the toxicity was reduced to grade ≤2, except in patients with drug-induced pneumonitis, who stopped the treatment permanently. A subsequent dose escalation to the original level was allowed provided that the patient tolerated the doses given at the 50% level. Vinorelbine was omitted and replaced with 250 mg gefitinib if the absolute neutrophil count was <1 × 109/L, the platelet count was <75 × 109/L on the day of vinorelbine injection, or if the patient refused the vinorelbine injection.
Subsequent complete blood cell counts and serum biochemistry studies were obtained 2 weeks after the beginning of gefitinib treatment and every 4 weeks thereafter for every patient. In addition, a complete blood cell count was checked every 2 weeks in the GV arm. Drug-related adverse events and toxicities were recorded according to the Common Toxicity Criteria of the National Cancer Institute (version 2.0). The Lung Cancer Symptom Scale was recorded before treatment, every month thereafter, and when the patient completed or went off the study.
Evaluation of response was performed after 4 weeks of treatment and every 8 weeks thereafter. Types of responses were assessed with the use of the Response Evaluation Criteria in Solid Tumors (RECIST).16 Responding patients and patients with stable disease continued treatment until they developed disease progression.
This study was designed to enroll at least 20 qualified patients in each arm. This calculation assumed that the true 1-year progression-free survival rate for the best treatment was 10% better than the rate for the other treatments. We assumed that the smaller 1-year progression-free survival rate was 5% and that the higher rate was 15%, with a power of 0.85 and a P value of 0.05, and that each treatment group required 20 qualified patients.17 Survival was measured from administration of the first dose until the date of death or last follow-up. For statistical analysis, the Kaplan-Meier method with a log-rank test was used for univariate survival analysis. For multivariate survival analysis of all 48 patients, a Cox regression analysis was performed that included the variables sex, performance status, current treatment as third-line or later, smoking, response or no response to current treatment, response or no response to first-line chemotherapy, response or no response to last-line chemotherapy, skin rash, diarrhea, paronychia, and treatment with or without vinorelbine. For the statistical analysis of the Lung Cancer Symptom Scale, the Mann-Whitney test was used for a comparison of the 2 treatment arms, and the Wilcoxon signed-rank test was used for comparison before and after treatment. The SPSS statistical software program was used for the analyses.
Fluorescence in situ hybridization studies for EGFR gene copy numbers were performed using the EGFR/chromsome 7 centromere (CEP7) dual-color probes (Vysis, Downer's Grove, Ill) according to the manufacturer's instructions. Hybridization occurred with the orange EGFR probe and the green CEP7 probe. Specimens that measured 3 or 4 μm thick were placed on silane-coated slides; then, they were baked at 56°C for 2 hours, deparaffinized, pretreated (1 M sodium thiocyanate in 1 M Tris-HCl, pH 8.0), and digested with pepsin (Sigma-P7012) (4 mg/mL in 0.9% NaCl solution, pH 2.0) for approximately 12 to 14 minutes, and fixed (in 10% buffered formalin) and denatured in solution (formamide, 20 × standard saline citrate, and H2O at 7:1:2 dilution, pH 7.0–7.5) for 5 minutes. After dehydration with 70%, 85%, and 100% ethanol, specimens were hybridized overnight at 37°C. The definition of EGFR gene amplification or high polysomy was based on that reported by Cappuzzo et al.18
EGFR mutation analysis was performed with nucleotide sequence analysis. The VarientSEQr Resequencing Primer Set (Applied Biosystems, Foster City, Calif) was selected for mutational analysis of the tyrosine kinase domain: exons 18 through 21 of the EGFR gene. DNA was extracted, exons 18 through 21 were amplified, and uncloned polymerase chain reaction (PCR) fragments were sequenced and analyzed in both sense and antisense directions for the presence of heterozygous mutations. All sequence variants were confirmed by multiple, independent PCR amplifications, and forward and reverse sequencing reactions were done. Each 10-μL PCR reaction contained 10 ng of DNA, 1 μL HotStar buffer (with 1.5 mM MgCl2), 1 μL nucleotide triphosphates, 0.1 U HotStarTaq polymerase (Qiagen, Valencia, Calif), and 2 μL primer mix. The PCR program was set at 95°C for 15 minutes, followed by 40 cycles at 94°C for 30 seconds, 60°C for 45 seconds, 72°C for 45 seconds, and finally 72°C for 10 minutes. Sequencing reaction consisted of diluted PCR products mixed with M13 primer and Big Dye Terminator Mix (Applied Biosystems), followed by electrophoresis in an ABI Prism 3700 DNA Analyzer. Chromatograms were analyzed using SeqScape software (version 2.1; Applied Biosystems). Genomic DNA was derived from formalin-fixed and paraffin blocks for mutational analysis of the EGFR coding sequence. Normal control DNA provided by ABI was used for wild-type controls. All the sequence variations were confirmed by multiple, independent PCR amplifications and repeated sequencing reactions.
From August 2004 to October 2005, 48 patients were enrolled. Twenty-four patients were randomized into the GV arm. However, 3 patients refused vinorelbine treatment and received gefitinib treatment only; thus, 27 patients received G treatment and 21 patients received GV treatment. Among 21 patients in the GV arm, vinorelbine was stopped before disease progression in 4 patients because of port-A occlusion after 1 injection, 5 injections, 20 injections, and 23 injections, respectively; after 4 injections in 1 patient because of mucositis; and after 7 injections in another patient because of a persistent sensation of fatigue.
The mean patient age was 62.7 years in the G arm and 63 years in the GV arm. All patients had stage IV lung adenocarcinoma. The clinical characteristics of these patients are listed in Table 1. There was no statistical difference in the clinical characteristics between patients in the 2 treatment arms in terms of sex, age, performance status, current treatment as third-line or later treatment, and a history or no history of smoking. All patients were assessable for toxicity profile and treatment response.
|Variable||No. of patients (%)||P*|
|Gefitinib||Gefitinib plus vinorelbine|
|No. of patients||27||21|
|Mean age [range], y||62.7 [35–85]||63 [39–80]||.848|
|WHO performance status|
|1||16 (59.3)||16 (76.2)||.18|
|2||10 (37)||5 (23.8)|
|Third line||18 (66.7)||16 (76.2)||.835|
|Fourth line||8 (29.6)||3 (14.3)|
|Fifth line||1 (3.7)||2 (9.5)|
|Yes||13 (48.1)||11 (52.4)||.777|
|No||14 (51.9)||10 (47.6)|
After 8 weeks of treatment, 15 patients achieved a partial response (PR) in the G arm (overall response rate, 55.6%), and 11 patients achieved a PR in the GV arm (overall response rate, 52.4%; P = .837). Stable disease (SD) was reported in 6 patients (22.2%) and 7 patients (33.3%) in the G and GV arms, respectively; and progressive disease was reported in 6 patients (22.2%) and 3 patients (14.3%), respectively. The response rate did not correlate with the patients' performance status in either treatment arm.
After a median follow-up of 18 months, the median time to disease progression was 7.1 months in the G arm and 12.8 months in the GV arm (P = .1331). The 1-year progression-free survival rate was 21.2% in the G arm and 57.1% in the GV arm (P = .008) (Fig. 1). The median survival was 13.3 months in the G arm and 23.4 months in the GV arm (P = .1231) (Fig. 2), and the 1-year survival rates were 51.3% and 75.3%, respectively (P = .133).
When all 48 patients were considered together, survival was significantly better in those who had a better performance status (P = .0003) and those who responded to current treatment (P = .0038). Survival was longer in women, in those patients whose current treatment was third-line, in nonsmokers, in patients who responded to first-line or last-line chemotherapy, in patients who had gefitinib side effects (including skin rash, diarrhea, or paronychia), and in patients who received vinorelbine treatment; however, those differences were without statistical significance (Table 2). The multivariate Cox regression analysis that included sex, treatment arm, performance status, current treatment as third-line or later, smoking, response or no response to current treatment, response or no response to first-line treatment, response or no response to last-line treatment, skin rash, diarrhea, and paronychia showed that only performance status (P = .0058) and response to current treatment (P = .0468) had statistical significance.
|Variable||No. of patients||Median survival, months||P (Log-rank)|
|Third line||34||Not reached||.7361|
|Response to present treatment*|
|Response to first-line chemotherapy|
|Response to last-line chemotherapy|
Fourteen patients had pathology samples available for analysis. However, only 6 patients had qualified samples for mutation analysis, and 10 patients had samples available for EGFR fluorescence in situ hybridization (FISH) examinations. EGFR FISH and mutation studies showed that 3 of 6 patients (50%) had exon 19 in-frame deletions (Glu746-Thr750), and 2 of 10 patients (20%) had EGFR gene amplification. The 3 patients who had exon 19 deletions did not have EGFR gene high polysomy or amplification. The 2 patients who had EGFR gene amplification did not have a mutation analysis performed. Two of 3 patients who had EGFR mutation and both patients with EGFR gene amplification had a PR to treatment, with a maximal overall response rate of 80% in patients who had EGFR mutation or amplification.
All patients who were enrolled on the study were eligible for toxicity evaluation. The toxicities were few and were mild in severity in both arms. Only 1 patient in the GV arm suffered from grade 3 leukopenia or neutropenia (Table 3). In terms of nonhematologic toxicities, more patients suffered from a sensation of fatigue (P = .027) in the GV arm than in the G arm (Table 4). However, toxicities grade ≥2 were rare. Only 1 patient in the G arm suffered from grade 3 pneumonitis, and that patient responded well to steroid treatment.
|Toxicity/Grade*||No. of patients (%)|
|Gefitinib (n = 27)||Gefitinib plus vinorelbine (n = 21)|
|Leukopenia||24 (88.9)||1 (3.7)||2 (7.4)||0||0||12 (57.1)||7 (33.3)||1 (4.8)||1 (4.8)||0|
|Neutropenia||25 (92.6)||1 (3.7)||1 (3.7)||0||0||10 (47.6)||8 (38.1)||2 (9.5)||1 (4.8)||0|
|Anemia||14 (51.9)||8 (29.6)||5 (18.5)||0||0||6 (28.6)||13 (61.9)||2 (9.5)||0||0|
|Thrombocytopenia||24 (88.9)||2 (7.4)||1 (3.7)||0||0||20 (95.2)||0||1 (4.8)||0||0|
|Toxicity/Grade*||No. of patients (%)|
|Gefitinib (n = 27)||Gefitinib plus vinorelbine (n = 21)|
|Fatigue||25 (92.6)||2 (7.4)||0||0||0||15 (71.4)||3 (14.3)||3 (14.3)||0||0|
|Skin rash||12 (44.4)||6 (22.2)||8 (29.6)||1 (3.7)||0||14 (66.7)||1 (4.8)||4 (19)||2 (9.5)||0|
|Dry skin||21 (77.8)||2 (7.4)||3 (11.1)||1 (3.7)||0||17 (81)||2 (9.5)||2 (9.5)||0||0|
|Paronychia||22 (81.5)||4 (14.8)||1 (3.7)||0||0||17 (81)||3 (14.3)||1 (4.8)||0||0|
|Diarrhea||22 (81.5)||3 (11.1)||2 (7.4)||0||0||18 (85.7)||2 (9.5)||1 (4.8)||0||0|
|Constipation||25 (92.6)||0||1 (3.7)||1 (3.7)||0||20 (95.2)||0||1 (4.8)||0||0|
In total, 45 patients (25 patients in the G arm and 20 patients in the GV arm) completed the Lung Cancer Symptom Scale questionnaire at baseline, after 2 months of treatment, and/or after going off study. The results of the completed Lung Cancer Symptoms Scale showed that there was no statistically significant difference in the scales between the G arm and the GV arm, either before treatment, or 2 months after treatment, or when the patient went off study, and whether the measure was scored by the patients (9 items) or by the observers (6 items), including the categories of loss of appetite, fatigue, cough, dyspnea, hemoptysis, pain, disease severity, daily activity, and quality of life. When all of the treated patients were considered together, there was no difference in the scores for all items between baseline and 2 months after treatment, except loss of appetite and fatigue, which had a slight and significant decrease. There was also a slight and significant decease in the scores for all items except hemoptysis between baseline and after the patient had gone off study.
Two large, randomized Phase III studies of chemotherapy with or without gefitinib in patients with advanced chemotherapy-naive NSCLC, known as the Iressa NSCLC Trial Assessing Combination Treatment (INTACT) 1 and INTACT 2 trials, were performed to determine whether or not gefitinib can enhance the antitumor effect of chemotherapy and, thus, prolong patient survival more than chemotherapy alone.10, 11 In the INTACT 1 study, the chemotherapy regimen was cisplatin plus gemcitabine at the usual dose and schedule.10 Patients were assigned randomly to receive chemotherapy plus a placebo, chemotherapy plus gefitinib 250 mg daily, or chemotherapy plus gefitinib 500 mg daily. Surprisingly, the results demonstrated that there were no differences in response rate, overall survival, or progression-free survival across the 3 treatment arms. In the INTACT 2 trial, a different chemotherapy regimen was used.11 Again, there were no differences in response rate, overall survival, or progression-free survival across the 2 treatment arms. The results of those 2 trials demonstrated that, in patients with chemotherapy-naive NSCLC, the addition of gefitinib to conventional chemotherapy does not enhance the clinical benefit of combination chemotherapy alone. Recently, it has come to light that EGFR TKIs induce apoptosis in mutant cell lines when they are given with chemotherapy but result in G1 arrest in EGFR wild-type tumors19; and, because approximately 50% to 60% of our patient population carried EGFR mutant tumors,9 the high response rates and prolonged survival were not surprising. In addition, our patients received gefitinib on Days 2 through 14 and vinorelbine on Day 1, which effectively achieved pharmacodynamic separation and high response rates even in our patients with EGFR wild-type tumors; and this supports the high 40% response rate (1 CR, 3 PRs, and 3 SD in 10 patients) achieved on a Phase I clinical trial of docetaxel on Day 1 and erlotinib on Days 2 through 16 repeated every 3 weeks.19 Our small study with an appropriate control arm should make a very important contribution toward justifying the initiation of large Phase III studies,20, 21 because several randomized Phase III studies have produced negative results when a TKI was added to doublet chemotherapy.10, 11, 13
A subset analysis of the paclitaxel/carboplatin With or Without Erlotinib in Advanced NSCLC study in the TRIBUTE study showed that nonsmokers had better survival when they received both chemotherapy and erlotinib treatment compared with chemotherapy alone (median survival, 22.5 months vs 10 months; P = .01).13, 14 In the INTACT 2 study, there was a trend toward improved survival in the subgroup of patients with adenocarcinoma who had received chemotherapy for ≥90 days in the gefitinib 250 mg daily arm (P = .05).11 Thus, it is both conceivable and possible that, if only a small proportion of patients are sensitive to gefitinib or erlotinib, then the diluting effect of the lack of selection may make the discerning of small differences in the INTACT, TRIBUTE, and TALENT studies impossible. A sharper definition of such susceptible subgroups of patients certainly will help in the further development of this type of agent. The selection of Asian patients with adenocarcinoma who failed ≥2 chemotherapy regimens, as in our study, and had no overlapping schedule of gefitinib and vinorelbine is another quick way to screen whether or not a new regimen is feasible for further testing.
Although there has been documentation of ineffectiveness of single-agent vinorelbine as salvage chemotherapy,22 it was 1 of the few chemotherapeutic agents that had not been administered previously to our selected patient group; in addition, although severe myelosuppression leading to an unacceptable rate of febrile neutropenia had been reported in patients on concurrent gefitinib/vinorelbine/cisplatin or gefitinib/vinorelbine, very high response rates have been reported,12, 23 so that low-dose vinorelbine was selected for this clinical trial. The safety profiles in the current study of gefitinib plus low-dose vinorelbine were highly tolerable in our patients. The high efficacy of gefitinib plus low-dose vinorelbine deserves further attention and study, because this regimen still was effective after the patients had failed at least 2 chemotherapy regimens.
Many patients in the GV arm in our study stopped vinorelbine treatment before they developed disease progression. Replacement of intravenous vinorelbine with oral vinorelbine should be considered in future studies to prevent early stopping of vinorelbine treatment because of intravenous administration issues.
In our previous chemotherapy study, the Lung Cancer Symptoms Scale showed a significant decrease in symptom scores, which indicates that disease severity or symptoms had worsened after the patients had received 2 cycles of chemotherapy.24 In contrast, the current results indicated that gefitinib-based treatment produced no change in symptom scores after 2 months of treatment, indicating that there was better effective prolongation of symptom deterioration.
In conclusion, the addition of low-dose vinorelbine to gefitinib has shown high efficacy in patients with adenocarcinoma lung cancer who have failed 2 previous regimens of chemotherapy. Because there have been 4 negative Phase III randomized trials of EGFR TKIs with chemotherapy (INTACT 1 and 2, TRIBUTE, and TALENT), only studies in selected EGFR mutation-enriched patient populations can be justified at this time for further clinical trials that combine chemotherapy with EGFR TKIs.
- 4Gefitinib plus best supportive care in previously treated patients with refractory advanced non-small-cell lung cancer: results from a randomised, placebo-controlled, multicentre study (Iressa Survival Evaluation in Lung Cancer). Lancet. 2005; 366: 1527–1537., , , et al.
- 8Clinical application of epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKI) against non-small cell lung cancer (NSCLC). J Cancer Mol. 2005; 1: 83–91..
- 12Gefitinib (ZD1839) combined with gemcitabine or vinorelbine as single-agent in elderly patients with advanced non-small cell lung cancer (NSCLC) [abstract]. J Clin Oncol (Meeting Abstracts). 2004; 22: 7081., , , et al.
- 14Long survival of never smoking non-small cell lung cancer (NSCLC) patients (pts) treated with erlotinib HCl (OSI-774) and chemotherapy: sub-group analysis of TRIBUTE [abstract]. J Clin Oncol (Meeting Abstracts). 2004; 22: 7061., , , et al.
- 21Role of stratification and randomization in early trials of targeted agents. Educ Book Am Soc Clin Oncol. 2006; 24: 119–121..