Gefitinib or Carboplatin–Paclitaxel in Pulmonary Adenocarcinoma

Gefitinib or Carboplatin–Paclitaxel in Pulmonary

Adenocarcinoma

Tony S. Mok, M.D., Yi-Long Wu, M.D., F.A.C.S., Sumitra Thongprasert, M.D., Chih-Hsin Yang, M.D., Ph.D., Da-Tong Chu, M.D., Nagahiro Saijo, M.D., Ph.D., Patrapim Sunpaweravong, M.D., Baohui Han, M.D., Benjamin Margono, M.D., Ph.D., F.C.C.P., Yukito Ichinose, M.D., Yutaka Nishiwaki, M.D., Ph.D., Yuichiro Ohe, M.D., Ph.D., Jin-Ji Yang, M.D., Busyamas Chewaskulyong, M.D., Haiyi Jiang, M.D.,

Emma L. Duffield, M.Sc., Claire L. Watkins, M.Sc., Alison A. Armour, F.R.C.R., and Masahiro Fukuoka, M.D., Ph.D.

ABSTR ACT

From the State Key Laboratory in Oncol-ogy in South China, Sir YK Pao Centre for Cancer, Department of Clinical Oncolo-gy, Chinese University of Hong Kong, Hong Kong (T.S.M.), Guangdong Gener-al Hospital, Guangzhou (Y-L.W., J.-J.Y.), Cancer Hospital, Chinese Academy of Medical Sciences, Beijing (D.-T.C.), and Shanghai Chest Hospital, Shanghai (B.H.) — all in China; Maharaj Nakorn Chiang Mai Hospital, Chiang Mai University, Chiang Mai (S.T., B.C.), and Prince of Songkla University, Songkla (P.S.) — both in Thailand; National Taiwan Uni-versity Hospital, Taipei, Taiwan (C.-H.Y.); National Cancer Center Hospital East, Chiba (N.S., Y.N.), National Kyushu Can-cer Center, Fukuoka (Y.I.), National Can-cer Center Hospital, Tokyo (Y.O.), Astra-Zeneca, Osaka (H.J.), and Kinki University School of Medicine, Osaka (M.F.) — all in Japan; Dr. Soetomo Hospital, Suraba-ya, Indonesia (B.M.); and AstraZeneca, Macclesfield, United Kingdom (E.L.D., C.L.W., A.A.A.). Address reprint requests to Dr. Mok at the Department of Clinical Oncology, Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, HKSAR, China, or at tony@https://www.wendangku.net/doc/417355207.html,.hk.This article (10.1056/NEJMoa0810699) was published on August 19, 2009, and was last updated on September 4, 2009, at https://www.wendangku.net/doc/417355207.html,.

N Engl J Med 2009;361:947-57.

Copyright ? 2009 Massachusetts Medical Society.

BACKGROUND

Previous, uncontrolled studies have suggested that first-line treatment with gefitinib would be efficacious in selected patients with non–small-cell lung cancer.

METHODS

In this phase 3, open-label study, we randomly assigned previously untreated patients in East Asia who had advanced pulmonary adenocarcinoma and who were nonsmok-ers or former light smokers to receive gefitinib (250 mg per day) (609 patients) or car-boplatin (at a dose calculated to produce an area under the curve of 5 or 6 mg per milliliter per minute) plus paclitaxel (200 mg per square meter of body-surface area) (608 patients). The primary end point was progression-free survival.

RESULTS

The 12-month rates of progression-free survival were 24.9% with gefitinib and 6.7% with carboplatin–paclitaxel. The study met its primary objective of showing the nonin-feriority of gefitinib and also showed its superiority, as compared with carboplatin–paclitaxel, with respect to progression-free survival in the intention-to-treat population (hazard ratio for progression or death, 0.74; 95% confidence interval [CI], 0.65 to 0.85; P<0.001). In the subgroup of 261 patients who were positive for the epidermal growth factor receptor gene (EGFR ) mutation, progression-free survival was significantly longer among those who received gefitinib than among those who received carboplatin–pacli-taxel (hazard ratio for progression or death, 0.48; 95% CI, 0.36 to 0.64; P<0.001), whereas in the subgroup of 176 patients who were negative for the mutation, progres-sion-free survival was significantly longer among those who received carboplatin–pacli-taxel (hazard ratio for progression or death with gefitinib, 2.85; 95% CI, 2.05 to 3.98; P<0.001). The most common adverse events were rash or acne (in 66.2% of patients) and diarrhea (46.6%) in the gefitinib group and neurotoxic effects (69.9%), neutropenia (67.1%), and alopecia (58.4%) in the carboplatin–paclitaxel group.

CONCLUSIONS

Gefitinib is superior to carboplatin–paclitaxel as an initial treatment for pulmonary adenocarcinoma among nonsmokers or former light smokers in East Asia. The pres-ence in the tumor of a mutation of the EGFR gene is a strong predictor of a better outcome with gefitinib. (https://www.wendangku.net/doc/417355207.html, number, NCT00322452.)

T h e ne w engl a nd jour na l o f medicine

I nhibitors of the epidermal gro w th

factor receptor (EGFR) tyrosine kinase have clinical efficacy, as compared with the best sup-portive care1 or standard chemotherapy,2 when given as second-line or third-line therapy for ad-vanced non–small-cell lung cancer. Treatment with EGFR tyrosine kinase inhibitors is most effective in women, patients who have never smoked, pa-tients with pulmonary adenocarcinomas, and pa-tients of Asian origin. In these populations, such treatment is associated with favorable rates of ob-jective responses, progression-free survival, and overall survival.1,3,4 These populations also have a relatively high incidence of somatic mutations in the region of the EGFR gene that encodes the tyrosine kinase domain.5,6 Studies have shown that in patients with pulmonary adenocarcinoma who had a base-pair deletion at exon 19 (del746_A750) or a point mutation at exon 21 (L858R), the tumors were highly responsive to EGFR tyrosine kinase inhibitors,7-9 and subsequent studies of first-line therapy with these agents showed objective re-sponse rates of 54.8 to 81.6% and progression-free survival of 9.7 to 13.3 months among patients with these mutations.10-12

On the basis of these and other studies,1,4,13-16 we hypothesized that in a selected population, first-line therapy with an oral EGFR tyrosine ki-nase inhibitor would be at least as effective as chemotherapy with carboplatin–paclitaxel. In this study, we compared the efficacy, safety, and ad-verse-event profile of gefitinib with those of car-boplatin–paclitaxel when these drugs were used as first-line treatment in nonsmokers or former light smokers in East Asia who had adenocarci-noma of the lung. We also examined the role of an EGFR mutation as a predicator of the efficacy of gefitinib or carboplatin–paclitaxel.

Methods

STUDY DESIGN AND PATIENTS

The First Line Iressa versus Carboplatin/Paclitaxel in Asia (Iressa Pan-Asia Study [IPASS]) study was a phase 3, multicenter, randomized, open-label, parallel-group study comparing gefitinib (Iressa, AstraZeneca) with carboplatin (Paraplatin, Bristol-Myers Squibb) plus paclitaxel (Taxol, Bristol-Myers Squibb) as first-line treatment in clinically selected patients in East Asia who had advanced non–small-cell lung cancer. The primary end point was pro-gression-free survival. Secondary end points in-cluded overall survival (an early analysis, since follow-up is ongoing), the objective response rate, quality of life, reduction in symptoms, safety, and the adverse-event profile. Evaluations of efficacy according to the baseline biomarker status of EGFR were planned exploratory objectives.

Patients were eligible for inclusion in the study if they were 18 years of age or older, had histo-logically or cytologically confirmed stage IIIB or IV non–small-cell lung cancer with histologic fea-tures of adenocarcinoma (including bronchoalve-olar carcinoma), were nonsmokers (defined as pa-tients who had smoked <100 cigarettes in their lifetime) or former light smokers (those who had stopped smoking at least 15 years previously and had a total of ≤10 pack-years of smoking), and had had no previous chemotherapy or biologic or im-munologic therapy. Other eligibility criteria are described in the Supplementary Appendix, avail-able with the full text of this article at https://www.wendangku.net/doc/417355207.html,. The principal investigators and the members of the steering committee (see the Appendix at the end of this article) designed the study in collabo-ration with the sponsor (AstraZeneca) and super-vised the conduct of the trial. The sponsor col-lected and analyzed the data. The lead academic author had unrestricted access to the data and vouches for the validity and completeness of the results of the trial (see the Supplementary Appen-dix for further details). All patients provided writ-ten informed consent; separate consent was pro-vided for the assessment of EGFR biomarkers. An independent ethics committee at each participat-ing institution approved the study protocol. The study was conducted in accordance with the Dec-laration of Helsinki, the International Conference on Harmonization Guidelines for Good Clinical Practice, applicable regulatory requirements, and AstraZeneca’s policy on bioethics. One planned interim analysis was performed by an independent statistician and reviewed by an independent data and safety monitoring committee (see the Supple-mentary Appendix).

STUDY TREATMENT

Patients were randomly assigned, in a 1:1 ratio, to receive gefitinib (250 mg per day, administered orally) or paclitaxel (200 mg per square meter of body-surface area, administered intravenously over a 3-hour period on the first day of the cycle) fol-

Gefitinib or Carboplatin–Paclitaxel in Pulmonary Adenocarcinoma

lowed immediately by carboplatin (at a dose calcu-lated to produce an area under the concentration–time curve of 5.0 or 6.0 mg per milliliter per minute, administered intravenously over a period of 15 to 60 minutes) in cycles of once every 3 weeks for up to 6 cycles. Randomization was performed with the use of dynamic balancing17 with respect to per-formance status, as assessed by the World Health Organization (WHO) performance scale measur-ing activity (0 or 1, or 2 on a scale of 0 to 4, with lower numbers indicating a higher degree of ac-tivity); smoking status (nonsmoker or former light smoker); sex; and center. Treatment continued un-til progression of the disease, development of un-acceptable toxic effects, a request by the patient or physician to discontinue treatment, serious non-compliance with the protocol, or completion of six chemotherapy cycles. Among patients assigned to gefitinib therapy, those whose tumor progressed were offered the opportunity to switch to treat-ment with carboplatin–paclitaxel; however, if the patient declined or was not a good candidate for that treatment, he or she could receive another approved therapy of the physician’s choice. Among patients who were receiving carboplatin–paclitaxel, further therapy after progression of the disease was at the physician’s discretion.

ASSESSMENTS

Progression-free survival was assessed from the date of randomization to the earliest sign of dis-ease progression, as determined by means of the Response Evaluation Criteria in Solid Tumors (RECIST),18 or death from any cause. Overall sur-vival was assessed from the date of randomiza-tion until death from any cause. Tumor response was assessed every 6 weeks until disease progres-sion. Quality of life was assessed with the use of the Functional Assessment of Cancer Therapy–Lung (FACT–L) questionnaire (in which scores range from 0 to 136, with higher scores indicat-ing better quality of life) and the Trial Outcome Index (TOI, which is the sum of the physical well-being, functional well-being, and lung-cancer sub-scale [LCS] scores of FACT-L; scores range from 0 to 84, with higher scores indicating better qual-ity of life), and symptoms were assessed with the use of the LCS score (scores range from 0 to 28, with higher scores indicating fewer symptoms). The FACT-L questionnaire19 was administered at randomization and at week 1, once every 3 weeks until day 127, once every 6 weeks from day 128 until disease progression, and when the study drug was discontinued. Clinically relevant improvement was predefined as an improvement of six points or more in FACT-L and TOI scores or an improve-ment of two points or more in LCS scores, with the higher scores maintained for at least 21 days.20 Safety and tolerability were assessed according to National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.0. Tumor sam-ples from patients who consented to have biomark-ers assessed were analyzed at two central labora-tories to determine biomarker status, with EGFR mutation status the first priority. Patients were considered to be positive for the EGFR mutation if 1 of 29 EGFR mutations was detected with the use of the amplification refractory mutation sys-tem (ARMS) and the DxS EGFR29 mutation-detec-tion kit.21,22

STATISTICAL ANALYSIS

The primary end point (progression-free survival) was analyzed with the use of a Cox proportional-hazards model in the intention-to-treat population (all randomly assigned patients) to assess the non-inferiority of gefitinib as compared with carbo-platin–paclitaxel, with the WHO performance sta-tus (0 or 1, or 2), smoking status (nonsmoker or former light smoker), and sex as covariates. For noninferiority to be demonstrated, the 95% con-fidence interval for the hazard ratio had to lie en-tirely below the predefined noninferiority limit of 1.2. We estimated that with a total of 944 pro-gression events, the study would have 80% power to demonstrate noninferiority if the treatments were truly equal, with a two-sided 5% probability of an erroneous demonstration of noninferiority. If the 95% confidence interval for the hazard ra-tio was also below 1, the P value would be less than 0.05 and superiority could be concluded from the same analysis without statistical penalty (closed test procedure).23 Supportive secondary analyses are described in the Supplementary Appendix. Planned subgroup analyses were performed to compare progression-free survival between treat-ments in groups defined according to WHO per-formance status (0 or 1, or 2), smoking status (nonsmoker or former light smoker), sex, age at randomization (<65 years or ≥65 years), disease stage at screening (stage IIIB or IV), and presence or absence of biomarkers. Tests to determine in-

T h e ne w engl a nd jour na l o f medicine

teractions of treatment with covariates were used to identify predictive factors by assessing wheth-er there was a significant difference in the treat-ment effect for progression-free survival (hazard ratio for progression or death) between subgroups. Overall survival was analyzed with the use of methods that were similar to those used for the analysis of progression-free survival. The results of an early analysis are presented; follow-up with respect to overall survival is ongoing. The objec-tive response rate (in the intention-to-treat pop-ulation) and quality of life and rates of symptom reduction (among all patients with a baseline and at least one post-baseline quality-of-life assessment that could be evaluated) were assessed with the use of a logistic-regression model with the same covariates as those considered for progression-free survival to calculate odds ratios and 95% confi-dence intervals. Planned subgroup analyses of the objective response rate were performed with the use of methods that were similar to those used for the analysis of progression-free survival. Adverse events were summarized for all pa-tients who received at least one dose of the as-signed study treatment. The incidence rates of 10 specified safety events (5 that were possibly associ-ated with each study treatment) were compared with the use of Fisher’s exact test; adjustment for multiple comparisons was performed with the use of the method of Westfall and Young.24

R esults

PATIENTS AND TREATMENT

From March 2006 through October 2007, a total of 1217 patients from 87 centers in Hong Kong, elsewhere in China, Indonesia, Japan, Malaysia, the Philippines, Singapore, Taiwan, and Thailand were randomly assigned to a study group (Fig. 1). The two groups were well balanced with respect to de-mographic and baseline characteristics (Table 1). The mean duration of treatment was 6.4 months (median, 5.6; range, 0.1 to 22.8) for gefitinib and 3.4 months (median, 4.1; range, 0.7 to 5.8) for carboplatin–paclitaxel. The median number of treatment cycles in the carboplatin–paclitaxel group was six. At the cutoff date for collection of data (April 14, 2008), a total of 24.5% of the patients in the gefitinib group were continuing to receive the study treatment; all patients in the carboplatin–paclitaxel group had discontinued the drugs. Af-ter discontinuation of the assigned treatment at any time during the study, 38.9% of the patients in the gefitinib group received carboplatin–pacli-taxel, and 39.5% of the patients in the carbo-platin–paclitaxel group received an EGFR tyrosine kinase inhibitor; 10.5% of the patients in the ge-fitinib group and 14.0% of those in the carbo-platin–paclitaxel group received other antican-cer treatments.

EFFICACY

The median follow-up period for the analysis of progression-free survival was 5.6 months. The me-dian progression-free survival was 5.7 months in the gefitinib group and 5.8 months in the carbo-platin–paclitaxel group, approximately coinciding with crossing of the Kaplan–Meier curves. The 12-month rates of progression-free survival were 24.9% with gefitinib and 6.7% with carboplatin–paclitaxel; a total of 950 patients had progression of disease. The study met its primary objective of demonstrating noninferiority and showed the su-periority of gefitinib as compared with carbopla-tin–paclitaxel for progression-free survival (hazard ratio for progression or death, 0.74; 95% confi-dence interval [CI], 0.65 to 0.85; P<0.001). The probability that a patient would be free of disease progression was greater with carboplatin–pacli-taxel in the first 6 months and greater with gefi-tinib in the following 16 months (Fig. 2A). Pro-gression-free survival was longer in the gefitinib group than in the carboplatin–paclitaxel group in all clinical subgroups; the only clinical factor that affected progression-free survival was age (<65 years: hazard ratio, 0.81; 95% CI, 0.70 to 0.95; P = 0.007; ≥65 years: hazard ratio, 0.58; 95% CI, 0.45 to 0.76; P<0.001; P = 0.03 for the interaction of treatment with age) (Fig. 1 in the Supplemen-tary Appendix).

A total of 1038 patients (85.3%) gave their con-sent for biomarker analyses, and 683 patients (56.1%) provided samples. EGFR mutation data for 437 patients (35.9%) could be evaluated. Patients with a tissue sample that could be evaluated had demographic characteristics that were similar to those of the overall population (Table 1 in the Supplementary Appendix). Of the 437 samples, 261 (59.7%) were positive for a mutation. Of these 261 samples, 140 (53.6%) had exon 19 deletions, 111 (42.5%) had a mutation at exon 21 (L858R), 11 (4.2%) had a mutation at exon 20 (T790M), and 10 (3.8%) had other mutations; 11 patients had mul-tiple mutations. The proportions of mutations

Gefitinib or Carboplatin–Paclitaxel in Pulmonary Adenocarcinoma

were well balanced between the two groups (Ta-ble 2 in the Supplementary Appendix).

There was a significant interaction between treatment and EGFR mutation with respect to pro-gression-free survival (P<0.001). Progression-free survival was significantly longer among patients receiving gefitinib than among those receiving carboplatin–paclitaxel in the mutation-positive sub-group (hazard ratio for progression, 0.48; 95% CI, 0.36 to 0.64; P<0.001) (Fig. 2B) and significantly shorter among patients receiving gefitinib than among those receiving carboplatin–paclitaxel in the mutation-negative subgroup (hazard ratio, 2.85; 95% CI, 2.05 to 3.98; P<0.001) (Fig. 2C). Re-sults in the subgroup with unknown EGFR-muta-tion status (hazard ratio with gefitinib, 0.68; 95%

T h e ne w engl a nd jour na l o f medicine

CI, 0.58 to 0.81; P<0.001) (Fig. 2D ) were similar to those for the overall population.

The objective response rate in the overall popu-lation was significantly higher with gefitinib than with carboplatin–paclitaxel (43.0% vs. 32.2%; odds ratio, 1.59; 95% CI, 1.25 to 2.01; P<0.001) (Table 3 in the Supplementary Appendix) and numerically or statistically greater with gefitinib in all clini-cal subgroups. The objective response rate was 71.2% with gefitinib versus 47.3% with carbo-platin–paclitaxel in the mutation-positive subgroup (P<0.001) and 1.1% (one patient) versus 23.5%, respectively, in the mutation-negative subgroup (P = 0.001) (Table 3 in the Supplementary Ap-pendix).

Overall survival in this early analysis (450 pa-tients [37.0%] died, with follow-up ongoing) was similar between the two groups in the overall population (hazard ratio for death in the gefitinib group, 0.91; 95% CI, 0.76 to 1.10) (Fig. 2A in the Supplementary Appendix). Median survival was 18.6 months among patients receiving gefitinib and 17.3 months among patients receiving carbo-platin–paclitaxel. After observing the results with respect to progression-free survival, we performed an analysis of overall survival according to muta-tion status, although this analysis included only 81 deaths in the mutation-positive subgroup and 94 in the mutation-negative subgroup. The hazard ratios with gefitinib were 0.78 (95% CI, 0.50 to 1.20) in the mutation-positive subgroup and 1.38 (95% CI, 0.92 to 2.09) in the mutation-negative subgroup (Fig. 2B and 2C in the Supplementary Appendix).

Significantly more patients in the gefitinib group than in the carboplatin–paclitaxel group had a clinically relevant improvement in quality of life, as assessed by scores on the FACT-L ques-tionnaire (odds ratio, 1.34; 95% CI, 1.06 to 1.69; P = 0.01) and by scores on the TOI (odds ratio, 1.78; 95% CI, 1.40 to 2.26; P<0.001) (Fig. 3). Rates of reduction in symptoms, as assessed on the basis of the LCS scores, were similar between patients who received gefitinib and those who received carboplatin–paclitaxel (odds ratio with gefitinib, 1.13; 95% CI, 0.90 to 1.42; P = 0.30) (Fig. 3). Re-sults according to mutation status are provided in Figure 3 in the Supplementary Appendix.

SAFETY AND ADVERSE-EVENT PROFILE

Table 2 lists the most common adverse events. Gefitinib, as compared with carboplatin–paclitaxel,

was associated with a lower rate of grade 3 or 4

* Percentages may not sum to 100 because of rounding.? Ethnic group was self-reported.

? Other East Asian refers to patients who belong to East Asian ethnic groups other than Chinese and Japanese.

§ The World Health Organization (WHO) performance status measures level of activity and is assessed on a scale of 0 to 4, with lower numbers indicating a higher degree of activity.

? All patients had Stage IIIB or IV disease at entry.

Gefitinib or Carboplatin–Paclitaxel in Pulmonary Adenocarcinoma

mon Terminology Criteria for Adverse Events (28.7% vs. 61.0%), a lower rate of adverse events leading to discontinuation of the drug (6.9% vs. 13.6%), and a lower rate of dose modification due to tox-ic effects (16.1% vs. 35.2% for carboplatin and 37.5% for paclitaxel). Adverse events leading to death occurred in 3.8% of the patients treated with gefitinib and in 2.7% of the patients treated with paclitaxel–carboplatin; serious adverse events, in-cluding death, occurred in 16.3% and 15.6% of patients in the two groups, respectively; and seri-curred in 13.8% and 13.1% of patients in the two groups, respectively. The incidences of rash or acne, diarrhea, and elevated liver aminotransferase lev-els were significantly higher with gefitinib than with carboplatin–paclitaxel, whereas the incidenc-es of neurotoxic effects, nausea and vomiting, and hematologic toxic effects were significantly higher with carboplatin–paclitaxel (Table 4 in the Supplementary Appendix). Interstitial-lung-disease events (i.e., the acute respiratory distress syndrome, interstitial lung disease, pneumonitis, or radiation

T h e ne w engl a nd jour na l o f medicine

pneumonitis) occurred in 16 patients treated with gefitinib (2.6%), 3 of whom died, and in 8 pa-tients treated with carboplatin-paclitaxel (1.4%), 1 of whom died.

Discussion

Platinum-based combination chemotherapy, such as carboplatin–paclitaxel, is the standard first-line therapy for advanced non–small-cell lung can-cer.25,26 The results of this trial showed that gefi-tinib by itself is superior to carboplatin–paclitaxel in a selected population of East Asian patients. As initial treatment of non–small-cell lung can-cer in East Asian nonsmokers or former light smokers with pulmonary adenocarcinoma, gefi-tinib, as compared with carboplatin–paclitaxel, prolonged progression-free survival, increased the objective response rate, reduced toxic effects, and improved quality of life. The overall benefit was driven primarily by the subgroup of patients with EGFR mutations; in this subgroup, patients treated with gefitinib, as compared with those treated with carboplatin–paclitaxel, had a remarkably high objective response rate (71.2%) and prolonged progression-free survival (hazard ratio for progres-sion or death, 0.48; 95% CI, 0.36 to 0.64; P<0.001). In the subgroup of patients without EGFR muta-tions, the objective response rate with gefitinib was 1.1%, and progression-free survival favored chemotherapy (hazard ratio with gefitinib, 2.85; 95% CI, 2.05 to 3.98; P<0.001). These contrasting outcomes probably explain the change over time in treatment effect for progression-free survival in the overall population. The initial superiority of carboplatin–paclitaxel was attributed to the ben-efit that the EGFR-mutation–negative subgroup re-ceived from chemotherapy but not from gefitinib, whereas prolonged progression-free survival in the EGFR-mutation–positive subgroup explained the subsequent improvement favoring gefitinib. Cross-ing of the Kaplan–Meier curves did not occur in the mutation-positive subgroup or the mutation-negative subgroup.

Lynch et al. found specific EGFR mutations that correlated with tumor response to gefitinib.7 In the Iressa Survival Evaluation in Lung Cancer trial (ISEL; https://www.wendangku.net/doc/417355207.html, number, NCT00242801), the objective response rate for gefitinib-treated patients was 37.5% among the 16 patients with a tumor bearing an EGFR mutation as compared with 2.6% among the 116 patients without a mu-tation.27 Our trial confirms the predictive value of EGFR mutations for the responsiveness of pul-monary adenocarcinoma to gefitinib as compared with carboplatin–paclitaxel. The difference in the rates of objective response between gefitinib-treated patients with an EGFR mutation and those without an EGFR mutation (71.2% vs. 1.1%) was remarkable. The rate of an objective response to first-line gefitinib in our study is similar to rates reported in other studies in which patients were selected according to EGFR-mutation status, in-cluding patients in Western countries.10,12,28 Se-quist et al. screened patients (who were selected on the basis of clinical characteristics) for an EGFR mutation and reported an objective response rate of 54.8% among 31 gefitinib-treated patients who were positive for an EGFR mutation, only 2 of whom were Asian.12 However, in our study, ob-jective response rates among patients without an EGFR mutation were lower than expected, given the results of previous studies.16,29 One possible explanation is our use of ARMS, a more sensitive technique for detecting EGFR mutations.21,22 When Zhu et al. used ARMS to reanalyze 148 samples

Gefitinib or Carboplatin–Paclitaxel in Pulmonary Adenocarcinoma

that had previously been classified as negative for an EGFR mutation, they found 11 new samples with exon 19 mutations.30 Another possible ex-planation is that studies that showed higher re-sponse rates among mutation-negative patients were not always conducted in previously untreated patients. Mutation-negative status that is deter-mined in a diagnostic sample obtained at the time of the initial presentation may change during sub-sequent tumor progression or during the course of chemotherapy.31

Our findings suggest that, whenever possible, EGFR -mutation status should be determined before the initial treatment of pulmonary adenocarcino-ma. Ethnic origin, smoking status, and histologic findings help to identify patients who have a high likelihood of having an EGFR mutation; in this study, 59.7% of the tumors in a clinically selected population had EGFR mutations, as compared with 12.1% and 14.8% in the unselected populations in the ISEL and Iressa in NSCLC Trial Evaluating Response and Survival versus Taxotere (INTEREST; NCT00076388) studies, respectively.2,27

The efficacy of gefitinib seen in this study was coupled with lower incidences of alopecia, nausea, vomiting, neurotoxic symptoms, and myelosup-pression than those seen with carboplatin–pacli-

taxel. Among 607 patients who received gefitinib

* Calculations were based on 1196 patients who received at least one dose of the study treatment. The Common Terminology Criteria (CTC) grade is defined on the basis of the National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.0. Events are included if they occurred in at least 10% of patients in either treatment group, either while the patients were receiving treatment or during the 28-day follow-up, and if there was at least a 5% difference be-tween groups. There were other adverse events that occurred in few patients and that may or may not have been relat-ed to the study drug. NA denotes not available.

? This is a group term (sum of high-level and preferred terms, according to the definitions in the Medical Dictionary for Regulatory Activities ).

? Data are from the laboratory reports of 599 patients who were taking gefitinib and 577 who were taking carboplatin– paclitaxel. Events were included if there was a worsening in the laboratory value (absolute neutrophil count in the case of neutropenia, hemoglobin in the case of anemia, and white-cell count in the case of leukopenia) from baseline to CTC grade 3 or 4.

T h e ne w engl a nd jour na l o f medicine

and who were included in the safety analysis, in-terstitial-lung-disease events developed in only 16 (2.6%), 3 of whom (0.5%) died.

In summary, this study shows that first-line therapy with gefitinib as compared with carbo-platin–paclitaxel prolongs progression-free surviv-al, increases the objective response rate, and im-proves quality of life among clinically selected patients with non–small-cell lung cancer. The presence of an EGFR mutation was a robust predic-tor of improved progression-free survival with ge-fitinib, as compared with carboplatin–paclitaxel, and of the benefit of gefitinib with respect to the objective response rate, indicating that patients in whom an EGFR mutation has been identified will benefit most from first-line therapy with gefitinib. Supported by AstraZeneca.

Dr. Mok reports receiving consulting fees from Roche, Astra-Zeneca, Pfizer, and Eli Lilly, lecture fees from Roche, AstraZen-eca, and Eli Lilly, and a research grant to the Chinese Lung Cancer Research Foundation from AstraZeneca, Hong Kong; Dr. Wu, receiving consulting fees from AstraZeneca, Roche, Eli Lil-ly, and Pfizer and lecture fees from AstraZeneca, Roche, and Eli Lilly; Dr. Thongprasert, receiving consulting fees from Astra-Zeneca, Pfizer, and Sanofi-Aventis, lecture fees from AstraZen-eca, Eli Lilly, and Sanofi-Aventis, and grant support from Astra-Zeneca, Sanofi-Aventis, and Pfizer; Dr. C.-H. Yang, receiving consulting fees and lecture fees from AstraZeneca; Dr. Saijo, owning equity in Takeda and receiving grant support from As-traZeneca, Chugai, Takeda, and Taiho; Dr. Ichinose, receiving lecture fees from Chugai, Kyowa Hakko, AstraZeneca, and Sanofi-Aventis; Dr. Ohe, receiving lecture fees from AstraZene-ca, Eli Lilly, Chugai, Bristol-Myers Squibb, and Sanofi-Aventis; Dr. Chewaskulyong, receiving lecture fees from AstraZeneca and Roche; Dr. Jiang, Miss Duffield, Miss Watkins, and Dr. Ar-mour, being salaried employees of AstraZeneca and owning eq-uity in AstraZeneca; and Dr. Fukuoka, receiving lecture fees from AstraZeneca and Chugai. No other potential conflict of interest relevant to this article was reported.

We thank the patients and investigators for their participation in this study and Annette Smith, Ph.D., from Complete Medical Communications, who provided editing support funded by Astra Z eneca.

APPENDIX

Members of the First Line Iressa versus Carboplatin/Paclitaxel in Asia (Iressa Pan-Asia Study [IPASS]) Study Organization were as follows: Steering Committee: T.S. Mok, M. Fukuoka, S. Thongprasert, Y.-L. Wu, C.-H. Yang, D.-T. Chu, N. Saijo, H. Jiang, C.L. Watkins, A.A. Armour (K.F. To, pathologist, advisor to steering committee). Independent Data and Safety Monitoring Committee: A. Chang, K. Egu-chi, M. Buyse, S. Zuckerman. International Coordinating Investigators: T.S. Mok, M. Fukuoka. Study Personnel: S. Rigby, study coor-dinator and study delivery leader; H. Jiang, study physician; P. Magill, study physician; E.L. Duffield, biostatistician. Investigators:China — C. Bojun, X. Cai, X. Cai, Q. Chen, X. Chen, Y. Chen, Z. Chen, W. Cheng, X. Chongrui, D. Chu, T. Chu, J. Dai, Z. Ding, J. Duan, M. Fan, Y. Fan, J. Feng, X. Fu, M. Gao, A. Gu, J. Gu, Z. Guan, B. Han, A. Hao, Z. He, W. Hong, X. Hong, M. Hou, C. Huang, J. Huang, P. Huang, Y. Huang, Y. Huang, Y. Huang, W. Huimin, L. Jia, H. Jian, G. Jiang, L. Jiang, S. Jiao, B. Jin, M. Jin, A. Li, C. Li, H. Li, L. Li, M. Li, R. Li, T. Li, Z. Li, H. Liang, M. Liao, R. Liao, J. Liu, X. Liu, Z. Liu, F. Lou, G. Lou, S. Lu, L. Mei, Q.-Y. Meng, J. Ni, M. Oiu, H. Pan, J. Pei, L. Peng, J. Qi, M. Qi, J. Qian, H. Qiu, J. Shen, Q. Song, Y. Song, S. Sun, X. Tan, B. Wang, B. Wang, H. Wang, H. Wang, H. Wang, K. Wang, L. Wang, L. Wang, M. Wang, W. Wang, X. Wang, Y. Wang, B. Wu, Y. Wu, C. Xie, R. Xie, Y. Xin, L. Xu, Z. Xu, B. Yan, J. Yang, L. Yang, Z. Yi, S. Yu, J. Zhang, J. Zhang, L. Zhang, L. Zhang, W. Zhang, X. Zhang, Y. Zhang, Y. Zhang, Y. Zhang, X. Zhao, Y. Zhao, W. Zhen, Z. Zhen, Y. Zheng, H. Zhong, R. Zhong, C. Zhou, Q. Zhou, T. Zhou, J. Zhu, Y. Zhu, Z. Zhu, W. Zhuang, L. Zou; Hong Kong (China)— S.K. Au, L. Chan, S. Cheung, K.-C. Chow, S.M. Chow, D. Chua, C.K. Kwan, K.C. Lam, T.C. Lam, D. Lee, R. Liu, S.H. Lo, P. Lui, T.S. Mok, P. Poon, C. Tang, K.F. To, Y.C. Tse, Y. Tung, H. Wong, M. Wong, S. Yau; Indonesia — J. Aphridasari, B. Boediwarsono, S. En-darjo, A. Febriani, H. Harijadi, A. Hudoyo, A. Kosasih, J. Kurnianda, B. Kusnan, H. Lunardhi, B. Margono, A. Mudigno, N. Nurhadi, A. Rima, K. Soedarko, C. Soeharti, J. Sugiri, M. Suprapto, E. Surjanto, E. Syahruddin, I. Tedjasukmana, P. Wibowo, K. Widayati, P. Widjanarko, L. Wulandari; Japan — Y. Akashi, K. Aoe, N. Aono, G. Asai, K. Asai, K. Asami, S. Atagi, T. Baba, K. Chikamori, H. Daga, S. Doi, M. Ebisawa, M. Endo, T. Endo, T. Fujieda, M. Fujii, S. Fujita, D. Fujiwara, S. Fukumoto, M. Fukuoka, H. Fukushima, C. Fukuy-ama, S. Fukuyama, S. Fukuyama, S. Fumita, K. Goto, E. Hagiwara, K. Hanioka, F. Hara, D. Harada, M. Harada, T. Harada, Y. Harada, A. Hata, Y. Hattori, M. Hayashi, S. Hibino, Y. Higashi, T. Hirano, N. Hirata, T. Hirata, T. Hishima, H. Honda, T. Horai, A. Horiike, Y. Hosomi, E. Ichihara, S. Ichihara, Y. Ichikawa, Y. Ichimaru, S. Ichinose, Y. Ichinose, S. Igawa, M. Iguchi, S. Ihara, K. Ijichi, T. Ikeda, Y. Ikezawa, Y. Imabashi, H. Imadate, Y. Imahashi, N. Imai, Y. Imai, F. Imamura, M. Inaba, T. Inoue, Y. Inoue, M. Ishida, G. Ishii, Y. Ishikawa, H. Ito, M. Ito, T. Ito, T. Iwasa, K. Iyama, S. Kajikawa, N. Kajiwara, M. Kakihana, T. Kakugawa, T. Kameya, S. Kanda, H. Kaneda, K. Kasahara, H. Kashihara, T. Kashii, K. Kashiwabara, N. Katakami, H. Katayama, N. Katayama, T. Kato, S. Kawabata, Y. Kawada, T. Kawaguchi, M. Kawahara, O. Kawai, Y. Kawai, H. Kenmotsu, Y. Kida, H. Kimura, T. Kimura, T. Kimura, E. Kin, A. Kinoshita, D. Kishino, C. Kitagawa, M. Kitaichi, A. Kitamura, K. Kitamura, M. Kitaoka, K. Kiura, H. Kiyota, S. Kobayashi, T. Kodama, T. Koga, Y. Kogure, Y. Koh, H. Kohrogi, S. Komatsu, T. Kometani, K. Komuta, A. Kubo, T. Kubo, Y. Kubo, K. Kubota, M. Kubota, K. Kudo, S. Kudo, H. Kunitoh, T. Kurata, Y. Kusunoki, S. Kyo, T. Maeda, T. Marutsuka, M. Maruyama, J. Matsubayashi, K. Matsumoto, M. Matsumoto, Y. Matsumoto, Y. Matsuno, H. Minato, S. Mitsuoka, K. Miyajima, E. Miyauchi, M. Miyazaki, T. Miyazaki, K. Mori, R. Morinaga, S. Moritani, H. Murakami, M. Murakami, T. Murakami, K. Murase, T. Nagano, S. Nagase, Y. Nagatsuka, Y. Naito, K. Naka-gawa, R. Nakajima, Y. Nakamura, Y. Nakanishi, S. Nanjo, M. Nakao, M. Nara, R. Naya, S. Negoro, S. Niho, D. Niino, R. Nishihira, H. Nishimori, R. Nishimura, T. Nishimura, K. Nishino, M. Nishio, Y. Nishiwaki, K. Nishiyama, N. Nogami, H. Nokihara, M. Nomura, N. Nomura, K. Nozaki, N. Ochi, Y. Ogata, A. Ogino, T. Ogura, C. Ohbayashi, Y. Ohe, T. Ohira, H. Ohmatsu, S. Ohta, T. Ohta, F. Ohyan-agi, K. Okabe, T. Okabe, I. Okamoto, K. Okamoto, S. Okamoto, T. Okamoto, W. Okamoto, T. Okamura, Y. Okano, M. Oki, K. Ok-ishio, M. Okuno, H. Omiya, M. Omori, A. Ono, M. Osawa, A. Osoegawa, K. Otsuka, A. Oya, I. Oze, S. Saeki, N. Saijo, T. Saijo, T. Saishouji, E. Saito, H. Saito, H. Saji, H. Saka, E. Sasaki, J. Sasaki, T. Sato, T. Sato, M. Satouchi, Y. Segawa, A. Sekine, I. Sekine, R. Seo, T. Seto, M. Shibuya, T. Shimada, T. Shimokata, T. Shimokawa, T. Shinkai, T. Shinohara, H. Shirane, Y. Sogo, A. Sugawara, K. Sugi, M. Sugishita, N. Suko, M. Sumitani, T. Syukuya, M. Tabata, K. Tachibana, R. Tachikawa, H. Tada, A. Tagawa, T. Tagawa, M. Takada, S. Takada, H. Takahashi, K. Takahashi, S. Takahashi, T. Takahashi, K. Takayama, H. Takeda, K. Takeda, M. Takeda, Y. Takeshima, Y. Takeuchi, K. Takezawa, N. Takigawa, A. Tamiya, D. Tamura, T. Tamura, T. Tamura, C. Tanai, K. Tanaka, T. Tashiro, N. Teramoto, M. Terashima, Y. Tochino, S. Tokunaga, Y. Tomita, M. Tsuboi, M. Tsujimoto, K. Tsujino, Y. Tsukamoto, H. Tsukuda, M. Tsuno, J. Tsuru-tani, K. Tsuta, A. Tsuya, J. Uchida, O. Uchida, J. Uchino, S. Ueda, K. Uehira, K. Ueno, H. Ueoka, S. Umemura, K. Urata, S. Ushijima,

Gefitinib or Carboplatin–Paclitaxel in Pulmonary Adenocarcinoma

J. Usuda, K. Wakasa, K. Waseda, K. Watanabe, H. Wataya, K. Yamada, I. Yamamoto, N. Yamamoto, N. Yamamoto, R. Yamamoto, Y. Yamane, K. Yamashiro, K. Yamazaki, H. Yanai, M. Yasugi, T. Yazawa, K. Yoh, T. Yoshida, M. Yoshimura, S. Yoshimura, T. Yoshinaga, K. Yumine, Y. Zen; Malaysia — A. Abdul Muttalif, N. Abdullah, H. Ahmad Zaharah, A. Awang Abdullah, J. Azizi Bin Abdul Rahman, I. Beevi, I. Hyder Ali, C.N. Choy, K.T. Chua, J. Dharmaratnam, S. Govindaraju, F.N. Lau, C.H. Leow, C.K. Liam, Y.K. Pang, S. Poospara-jah, B. Rajendran, R. Raman, K. Ratnavelu, H. Sahat, V. Selvaraju, K. Sivaraman Kannan, C.K. Tiong, A. Zaatar; Philippines — A. Abe-lardo, F. Agustin, V. Butalid, P. Caguioa, V. Chan, G. Cornelio, D. Dizon, A. Faundo, K. Gutierez, J. Holaysan, M. Madrid, A. Malilong, A. Ong-Cornel, P. Pua, B. Ramos, E. Tan, D. Tudtud, N. Uy, A. Villalon, K. Villanueva, E. Villegas; Singapore — S.S. Leong, D. Lim, E.H. Tan, Y.O. Tan, H.K. Tan, C.K. Toh; Taiwan— T.-Y. Chao, H.-C. Chen, K.-Y. Chen, P.-J. Chen, Y.-C. Chen, Y.-M. Chen, C.-Y. Chung, C.-C. Ho, C.-L. Ho, M.-L. Ho, R.-K. Hsieh, C. Hsu, W.-Y. Kao, H.-P. Kuo, C.-H. Lai, H.-C. Lin, J.-T. Lin, M.-C. Lin, Y.-L. Lin, Z.-Z. Lin, M.-J. Peng, R.-P. Perng, J.-Y. Shih, C.-C. Wang, C.-H. Wang, J.-L. Wang, Y.-H. Wang, C.-L. Wu, C.-H. Yang, P.-C. Yang, C.-T. Yu, C.-J. Yu; Thailand — C. Akewanlop, V. Ariyaprakai, T. Ativitavas, T. Chalermchai, C. Chantranuwat, C. Charoentum, B. Chewasukulyong, T. Dudsadeeprasert, S. Geater, M. Huntrakoon, S. Juthong, N. Keerativitayanant, N. Kiatikajornthada, C. Kularbkaew, S. Laohavanij, N. Lertprasertsuke, J. Maneechavakajorn, W. Mitarnum, A. Phunmanee, P. Punyarit, M. Rochanawutanon, K. Seetalarom, E. Sirachainan, A. Sookprasert, N. Soparatanapaisarn, V. Srimuninnimit, V. Sriuranpong, P. Sunpaweravong, C. Suthipintawong, H. Suwanrusme, S. Suwanvecho, K. Thammakumpee, S. Thongprasert, V. Viriyachaiyo, N. Voravud, S. Wongbunnate.

References

Shepherd FA, Rodrigues Pereira J, 1.

Ciuleanu T, et al. Erlotinib in previously treated non–small-cell lung cancer. N Engl J Med 2005;353:123-32.

Kim ES, Hirsch V, Mok T, et al. Gefi-2.

tinib versus docetaxel in previously treat-ed non-small-cell lung cancer (INTEREST): a randomised phase III trial. Lancet 2008; 372:1809-18.

Park K, Goto K. A review of the bene-3.

fit-risk profile of gefitinib in Asian patients with advanced non-small-cell lung can-cer. Curr Med Res Opin 2006;22:561-73.

Thatcher N, Chang A, Parikh P, et al.

4.

Gefitinib plus best supportive care in pre-viously treated patients with refractory ad-vanced non-small-cell lung cancer: results from a randomised, placebo-controlled, multicentre study (Iressa Survival Evalua-tion in Lung Cancer). Lancet 2005;366: 1527-37.

Kosaka T, Yatabe Y, Endoh H, Kuwano 5.

H, Takahashi T, Mitsudomi T. Mutations of the epidermal growth factor receptor gene in lung cancer: biological and clinical impli-cations. Cancer Res 2004;64:8919-23.

Shigematsu H, Lin L, Takahashi T, et 6.

al. Clinical and biological features associ-ated with epidermal growth factor recep-tor gene mutations in lung cancers. J Natl Cancer Inst 2005;97:339-46.

Lynch TJ, Bell DW, Sordella R, et al.

7.

Activating mutations in the epidermal growth factor receptor underlying respon-siveness of non–small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129-39.

Paez JG, J?nne PA, Lee JC, et al. EGFR 8.

mutations in lung cancer: correlation with clinical response to gefitinib therapy. Sci-ence 2004;304:1497-500.

Pao W, Miller V, Zakowski M, et al.

9.

EGF receptor gene mutations are common in lung cancers from “never smokers” and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A 2004;101:13306-11.

Inoue A, Suzuki T, Fukuhara T, et al.

10.

Prospective phase II study of gefitinib for chemotherapy-naive patients with ad-vanced non-small-cell lung cancer with epidermal growth factor receptor gene mutations. J Clin Oncol 2006;24:3340-6.

Rosell R, Moran T, Queralt C, et al.

11.

Screening for epidermal growth factor

receptor mutations in lung cancer.

N Engl J Med 2009;361. DOI: 10.1056/

NEJMoa0904554.

Sequist LV, Martins RG, Spigel D, et

12.

al. First-line gefitinib in patients with ad-

vanced non-small-cell lung cancer harbor-

ing somatic EGFR mutations. J Clin Oncol

2008;26:2442-9. [Erratum, J Clin Oncol

2008;26:3472.]

Chang G-C, Chen K-C, Yang T-Y, et al.

13.

Activity of gefitinib in advanced non-small-

cell lung cancer with very poor performance

status. Invest New Drugs 2005;23:73-7.

Kimura H, Kasahara K, Shibata K, et

14.

al. EGFR mutation of tumor and serum in

gefitinib-treated patients with chemother-

apy-naive non-small cell lung cancer. J Tho-

rac Oncol 2006;1:260-7.

Lee DH, Han JY, Lee HG, et al. Gefi-

15.

tinib as a first-line therapy of advanced or

metastatic adenocarcinoma of the lung in

never-smokers. Clin Cancer Res 2005;11:

3032-7.

Yang CH, Yu CJ, Shih JY, et al. Specific

16.

EGFR mutations predict treatment outcome

of stage IIIB/IV patients with chemother-

apy-naive non-small-cell lung cancer re-

ceiving first-line gefitinib monotherapy.

J Clin Oncol 2008;26:2745-53.

Pocock SJ, Simon R. Sequential treat-

17.

ment assignment with balancing for

prognostic factors in the controlled clini-

cal trial. Biometrics 1975;31:103-15.

Therasse P, Arbuck SG, Eisenhauer

18.

EA, et al. New guidelines to evaluate the

response to treatment in solid tumors.

J Natl Cancer Inst 2000;92:205-16.

Cella DF, Bonomi AE, Lloyd SR, Tulsky

19.

DS, Kaplan E, Bonomi P. Reliability and va-

lidity of the Functional Assessment of Can-

cer Therapy-Lung (FACT-L) quality of life

instrument. Lung Cancer 1995;12:199-220.

Cella D, Eton DT, Fairclough DL, et al.

20.

What is a clinically meaningful change on

the Functional Assessment of Cancer Ther-

apy-Lung (FACT-L) Questionnaire? Results

from Eastern Cooperative Oncology Group

(ECOG) Study 5592. J Clin Epidemiol 2002;

55:285-95.

Newton CR, Graham A, Heptinstall

21.

LE, et al. Analysis of any point mutation

in DNA: the Amplification Refractory Mu-

tation System (ARMS). Nucleic Acids Res

1989;17:2503-16.

Whitcombe D, Theaker J, Guy SP,

22.

Brown T, Little S. Detection of PCR prod-

ucts using self-probing amplicons and fluo-

rescence. Nat Biotechnol 1999;17:804-7.

Morikawa T, Yoshida M. A useful test-

23.

ing strategy in phase III trials: combined

test of superiority and test of equivalence.

J Biopharm Stat 1995;5:297-306.

Westfall PH, Young SS. Resampling-

24.

based multiple testing. New York: Wiley,

1993.

Pfister DG, Johnson DH, Azzoli CG,

25.

et al. American Society of Clinical Oncol-

ogy treatment of unresectable non-small-

cell lung cancer guideline: update 2003.

J Clin Oncol 2004;22:330-53.

Schiller JH, Harrington D, Belani CP,

26.

et al. Comparison of four chemotherapy

regimens for advanced non–small-cell

lung cancer. N Engl J Med 2002;346:92-8.

Hirsch FR, Varella-Garcia M, Bunn PA

27.

Jr, et al. Molecular predictors of outcome

with gefitinib in a phase III placebo-con-

trolled study in advanced non-small-cell

lung cancer. J Clin Oncol 2006;24:5034-42.

Tamura K, Okamoto I, Kashii T, et al.

28.

Multicentre prospective phase II trial of

gefitinib for advanced non-small-cell lung

cancer with epidermal growth factor re-

ceptor mutations: results of the West

Japan Thoracic Oncology Group trial

(WJTOG0403). Br J Cancer 2008;98:907-14.

Han S-W, Kim T-Y, Hwang PG, et al.

29.

Predictive and prognostic impact of epi-

dermal growth factor receptor mutation

in non-small-cell lung cancer patients treat-

ed with gefitinib. J Clin Oncol 2005;23:

2493-501.

Zhu CQ, da Cunha SG, Ding K, et al.

30.

Role of KRAS and EGFR as biomarkers of

response to erlotinib in National Cancer

Institute of Canada Clinical Trials Group

Study BR.21. J Clin Oncol 2008;26:4268-75.

Maheswaran S, Sequist LV, Nagrath S,

31.

et al. Detection of mutations in EGFR in

circulating lung-cancer cells. N Engl J Med

2008;359:366-77.

Copyright ? 2009 Massachusetts Medical Society.