Gefitinib plus chemotherapy versus placebo plus chemotherapy in EGFR-mutation-positive non-small-cell lung cancer after progression on first-line gefitinib (IMPRESS): a phase 3 randomised trial

Articles

Gefitinib plus chemotherapy versus placebo plus chemotherapy in EGFR-mutation-positive non-small-cell lung cancer after progression on first-line gefitinib (IMPRESS): a phase 3 randomised trial Jean-Charles Soria, Yi-Long Wu, Kazuhiko Nakagawa, Sang-We Kim, Jin-Ji Yang, Myung-Ju Ahn, Jie Wang, James Chih-Hsin Yang, You Lu, Shinji Atagi, Santiago Ponce, Dae Ho Lee, Yunpeng Liu, Kiyotaka Yoh, Jian-Ying Zhou, Xiaojin Shi, Alan Webster, Haiyi Jiang, Tony S K Mok

Summary Background Optimum management strategies for patients with advanced non-small-cell lung cancer (NSCLC) with acquired resistance to EGFR tyrosine-kinase inhibitors are undefined. We aimed to assess the efficacy and safety of continuing gefitinib combined with chemotherapy versus chemotherapy alone in patients with EGFR-mutation-positive advanced NSCLC with acquired resistance to first-line gefitinib. Methods The randomised, phase 3, multicentre IMPRESS study was done in 71 centres in 11 countries in Europe and the Asia-Pacific region. Eligible patients were aged at least 18 years with histologically confirmed, chemotherapynaive, stage IIIB–IV EGFR-mutation-positive advanced NSCLC with previous disease control with first-line gefitinib and recent disease progression (Response Evaluation Criteria in Solid Tumors version 1.1). Participants were randomly assigned (1:1) by central block randomisation to oral gefitinib 250 mg or placebo once daily in tablet form; randomisation did not include stratification factors. All patients also received the platinum-based doublet chemotherapy cisplatin 75 mg/m² plus pemetrexed 500 mg/m² on the first day of each cycle. After completion of a maximum of six chemotherapy cycles, patients continued their randomly assigned treatment until disease progression or another discontinuation criterion was met. All study investigators and participants were masked to treatment allocation. The primary endpoint was progression-free survival in the intention-to-treat population. Safety was assessed in patients who received at least one dose of study treatment. The study has completed enrolment, but patients are still in follow-up for overall survival. This trial is registered with ClinicalTrials.gov, number NCT01544179. Findings Between March 29, 2012, and Dec 20, 2013, 265 patients were randomly assigned: 133 to the gefitinib group and 132 to the placebo group. At the time of data cutoff (May 5, 2014), 98 (74%) patients had disease progression in the gefitinib group compared with 107 (81%) in the placebo group (hazard ratio 0·86, 95% CI 0·65–1·13; p=0·27; median progression-free survival 5·4 months in both groups [95% CI 4·5–5·7 in the gefitinib group and 4·6–5·5 in the placebo group]). The most common adverse events of any grade were nausea (85 [64%] of 132 patients in the gefitinib group and 81 [61%] of 132 patients in the placebo group) and decreased appetite (65 [49%] and 45 [34%]). The most common adverse events of grade 3 or worse were anaemia (11 [8%] of 132 patients in the gefitinib group and five [4%] of 132 patients in the placebo group) and neutropenia (nine [7%] and seven [5%]). 37 (28%) of 132 patients in the gefitinib group and 28 (21%) of 132 patients in the placebo group reported serious adverse events. Interpretation Continuation of gefitinib after radiological disease progression on first-line gefitinib did not prolong progression-free survival in patients who received platinum-based doublet chemotherapy as subsequent line of treatment. Platinum-based doublet chemotherapy remains the standard of care in this setting. Funding AstraZeneca.

Introduction EGFR tyrosine-kinase inhibitors (TKIs) are standard firstline treatments for patients with non-small-cell lung cancer (NSCLC) with tumours harbouring activating EGFR mutations, whereas platinum-based doublet chemotherapy is standard first-line treatment for patients whose tumours do not harbour mutations.1,2 Findings from several randomised studies have confirmed improved proportions of patients achieving objective responses and better progression-free survival in patients

with EGFR-mutation-positive NSCLC tumours treated with first-line EGFR TKIs as compared with those with EGFR-mutation-negative NSCLC tumours.3–10 However, almost all patients eventually develop acquired resistance,11 and optimum management strategies for patients with acquired resistance to first-line EGFR TKIs are undefined.12 Outside clinical trials, treatment options include systemic chemotherapy alone or continuation of EGFR TKIs in combination with chemotherapy at the time of disease progression.12

www.thelancet.com/oncology Published online July 7, 2015 http://dx.doi.org/10.1016/S1470-2045(15)00121-7

Lancet Oncol 2015 Published Online July 7, 2015 http://dx.doi.org/10.1016/ S1470-2045(15)00121-7 Department of Medicine, Gustave Roussy Cancer Campus and University Paris-Sud, Villejuif, France (Prof J-C Soria MD); Guangdong Lung Cancer Institute, Guangdong General Hospital and Guangdong Academy of Medical Sciences, Guangzhou, China (Prof Y-L Wu MD, J-J Yang MD); Department of Medical Oncology, The Faculty of Medicine, Kinki University, Osakasayama City, Osaka, Japan (Prof K Nakagawa MD); Department of Oncology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea (Prof S-W Kim MD, Prof D H Lee MD); Division of Hematology-Oncology, Department of Medicine, The Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea (Prof M-J Ahn MD); Department of Thoracic Medical Oncology, Peking University School of Oncology, Beijing Cancer Hospital and Institute, Beijing, China (J Wang MD); Department of Oncology, The National Taiwan University Hospital and College of Medicine, Taipei, Taiwan (Prof J C-H Yang MD); Department of Thoracic Cancer, Cancer Centre, West China Hospital, West China Medical School, Sichuan University, Sichuan, China (Prof Y Lu MD); Department of Thoracic Oncology, Kinki-chuo Chest Medical Center, Osaka, Japan (S Atagi MD); Medical Oncology Service, Hospital Universitario 12 de Octubre, Madrid, Spain (S Ponce MD); Department of Medical Oncology, The First

1

Articles

Hospital of China Medical University, Shengyang City, China (Prof Y Liu MD); Division of Thoracic Oncology, National Cancer Centre Hospital East, Chiba, Japan (K Yoh MD); Department of Respiratory Diseases, The First Affiliated Hospital of College of Medicine, Zhejiang University, Zhejiang, China (Prof J-Y Zhou MD); AstraZeneca, Shanghai, China (X Shi MD, H Jiang MD); AstraZeneca, Macclesfield, UK (A Webster MSc); and State Key Laboratory of South China, Hong Kong Cancer Institute, Department of Clinical Oncology, The Chinese University of Hong Kong, Prince of Wales Hospital, Sha Tin, Hong Kong, China (Prof T S K Mok MD) Correspondence to: Prof Tony S K Mok, Department of Clinical Oncology, The Chinese University of Hong Kong, Prince of Wales Hospital, Sha Tin, Hong Kong, China [email protected]

Acquired resistance to EGFR TKIs is defined by the Jackman criteria,11 which adopt the Response Evaluation Criteria in Solid Tumors (RECIST) for radiological progression.13 However, the mechanism of resistance is complex and heterogeneous, and might not be identified by radiological imaging alone.14,15 The most common cause of acquired resistance to EGFR TKIs is the presence of a mutation in the EGFR gene at exon 20—the T790M mutation.16 However, many other causes exist, including MET amplification or a PI3K mutation17 and transformation into small-cell lung cancer.18 Therefore, radiological progression does not always imply that all metastatic sites share the same cause of resistance, and some tumour sites could potentially be sensitive to EGFR TKIs at the time of RECIST progression.19 This potential tumour heterogeneity suggests that continuation of EGFR TKIs in combination with chemotherapy might be beneficial—a hypothesis supported by findings from a retrospective study.20 Goldberg and colleagues20 reported that 48% of patients with tumours resistant to EGFR TKI treatment who were subsequently treated with a combination of chemotherapy and erlotinib achieved a tumour response versus 18% of patients treated with chemotherapy alone. In this study, we aimed to establish whether patients who developed acquired resistance to first-line gefitinib as per RECIST would benefit from continuation of EGFR TKI treatment in addition to platinum-based doublet chemotherapy.

Methods Study design and participants

See Online for appendix

2

The aim of this prospective, randomised phase 3 study (IMPRESS) was to compare continuation of gefitinib (AstraZeneca, Macclesfield, UK) in combination with cisplatin plus pemetrexed versus placebo plus cisplatin plus pemetrexed in patients with EGFR-mutation-positive advanced NSCLC with progression after first-line gefitinib. Patients were enrolled at 71 centres in 11 countries in Europe (France, Germany, Hungary, Italy, Russia, and Spain) and the Asia-Pacific region (China, Hong Kong, Japan, South Korea, and Taiwan; appendix). Eligible patients were aged at least 18 years (≥20 years in Japan), with cytologically or histologically confirmed chemotherapy-naive advanced NSCLC and an activating EGFR mutation as confirmed by local testing, who had achieved a complete or partial response for longer than 4 months or durable stable disease for at least 6 months on first-line gefitinib treatment and had subsequently developed radiological disease progression (ie, acquired resistance), as per RECIST version 1.1,13 within 4 weeks before random allocation. Patients were also required to have a life expectancy of at least 12 weeks and a WHO performance status of 0 or 1. Patient exclusion criteria included NSCLC of predominantly squamous cell histology, a history of interstitial lung disease, any other coexisting malignancies diagnosed within the past 5 years (excluding basal cell carcinoma, cervical cancer

in situ, or completely resected intramucosal gastric cancer), or treatment with another investigational drug 4 weeks or less before random allocation (appendix). All patients provided written, informed consent before enrolment. Study approval was obtained from independent ethics committees at each institution. The study was undertaken in accordance with Good Clinical Practice for Trials on Drugs (in Japan only, the Ministry of Health, Labour and Welfare [MHLW] Ordinance number 28, March 27, 1997, partially revised by MHLW Ordinance and their related notifications, was used), the Declaration of Helsinki, the International Conference on Harmonisation and Good Clinical Practice, applicable regulatory requirements, and AstraZeneca’s policy on bioethics and human biological samples.

Randomisation and masking We used central block randomisation to allocate patients (1:1) to gefitinib or placebo orally once daily. Randomisation did not include stratification factors; stratification by centre was removed in the first amendment of the study protocol on Nov 29, 2011, with a view to avoid potential imbalance of the distribution of patients at the centre level, because the study plan was to include a maximum of about 100 investigator study sites. Before assessment of patient eligibility, the principal investigator at each study site assigned individual patients a unique enrolment number using the interactive web response system or interactive voice response system at screening (visit 1). Once patient eligibility was confirmed, the principal investigator, or designated study site staff, sequentially assigned each patient a unique randomisation and treatment code using the interactive web response system or interactive voice response system at the randomisation visit (visit 2). All study investigators and participants were masked to treatment allocation. To ensure masking of study investigators and participants, all gefitinib and placebo packaging was identical. Apart from for safety reasons, nobody was allowed access to the randomisation scheme or study results until completion of the randomised treatment period to minimise any potential bias in data handling and to safeguard the integrity of the masking of study investigators.

Procedures Patients received either oral gefitinib 250 mg (once a day) or placebo in tablet form. Patients also received intravenous doublet chemotherapy with cisplatin 75 mg/m² plus pemetrexed 500 mg/m² on the first day of each cycle. Chemotherapy was prescribed and given as per local guidance; thus, there was no specific cycle length. Patients received a maximum of six cycles of cisplatin plus pemetrexed in addition to the randomly assigned treatment (gefitinib or placebo), after which the randomly assigned treatment was received until disease progression or another discontinuation criterion was met. Tumour assessments were done using MRI or CT scans 4 weeks or

www.thelancet.com/oncology Published online July 7, 2015 http://dx.doi.org/10.1016/S1470-2045(15)00121-7

Articles

less before the start of study treatment (ie, baseline) and repeated every 6 weeks after randomisation, with response categorised according to RECIST version 1.1.13 Dose reductions of gefitinib or placebo were not allowed; however, dose interruptions for up to 3 weeks were allowed to manage toxic effects. Chemotherapy-related toxic effects and dose modifications were managed as per standard clinical practice and cisplatin plus pemetrexed prescribing information. No pemetrexed maintenance treatment was allowed, because this was not regarded as standard practice at the time of protocol design. Precautionary pretreatment regimens were also undertaken with cisplatin (anti-emetic and hydration treatment as per local prescribing practice) and pemetrexed (vitamin supplements) treatment to reduce toxic effects.

Outcomes The primary endpoint was progression-free survival, defined as time from randomisation to objective tumour progression according to RECIST version 1.1 investigator assessment, or death from any cause. Secondary endpoints were the proportion of patients who achieved an objective response (ie, complete plus partial response), the proportion of patients who achieved disease control (ie, complete plus partial response plus stable disease) according to RECIST version 1.1 investigator assessment at least 6 weeks after random allocation, overall survival (defined as the time from randomisation to death from any cause), safety and tolerability, and health-related quality of life (HRQoL). Patients who had not progressed or died at the time of analysis were censored at the time of their last measurable RECIST version 1.1 assessment. Preplanned exploratory objectives were investigation of biomarkers in samples from patients with acquired resistance to gefitinib for potential differentiators of relative treatment effects; these data will be reported separately. We also undertook a secondary, supportive central review of progression-free survival. Safety and tolerability were assessed by adverse events and clinical laboratory results, classified according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE; version 4.0).21 HRQoL was assessed using the Functional Assessment of Cancer Therapy—Lung (FACT-L) questionnaire, the FACT-L Lung Cancer Subscale (LCS), and Trial Outcome Index (TOI; the sum of the physical wellbeing, functional wellbeing, and LCS domain scores of the FACT-L). Questionnaires were given to patients at visits 2–6, then every 6 weeks until progression, at progression, and at 8 weeks after progression. HRQoL improvement was defined as the percentage of all analysed patients with an overall response of improved (two visit responses of improved [change in score of at least +6 on the FACT-L or TOI or of at least +2 on the LCS] a minimum of 21 days apart without an intervening visit response of worsened [change in score of –6 or less on the FACT-L or TOI or of –2 or less on the LCS]). Time to worsening was defined for FACT-L, LCS,

287 patients assessed for eligibility

22 excluded 21 eligibility criteria not fulfilled 1 patient decision

265 randomly assigned

133 assigned to gefitinib group (ITT population)

132 assigned to placebo group (ITT population)

1 eligibility criteria not fulfilled

132 received at least one dose of study drug (safety population)

132 received at least one dose of study drug (safety population)

132 underwent HRQoL assessment at baseline and at least once after baseline (HRQoL population)

132 underwent HRQoL assessment at baseline and at least once after baseline (HRQoL population)

109 discontinued 18 patient decision 8 adverse events 82 objective disease progression 1 other

23 ongoing study treatment at data cutoff (May 5, 2014)

114 discontinued 4 patient discontinuation 9 adverse events 98 objective disease progression 3 other

18 ongoing study treatment at data cutoff (May 5, 2014)

Figure 1: Patient flow diagram HRQoL=health-related quality of life. ITT=intention to treat.

and TOI as the interval from the date of randomisation to the first visit response of worsened without a subsequent visit response of improved or no change (change in score other than that qualifying for improved or worsened for FACT-L, LCS, or TOI) within 21 days or to the date of death if no more than 12 weeks after the last assessable HRQoL assessment. If a visit score of worsened was not reported at the time of analysis, time to worsening was censored at the last non-missing assessment visit.

Statistical analysis The primary endpoint of progression-free survival was calculated in all enrolled patients (ie, the intention-totreat [ITT] population) and analysed using a Cox proportional hazards model adjusted for age (<65 years vs ≥65 years) and previous response to gefitinib (stable disease vs partial or complete response), because these were prognostic variables deemed potentially important. To show superiority of gefitinib in combination with cisplatin plus pemetrexed versus cisplatin plus pemetrexed alone, we estimated—independently of the Cox proportional hazards model—that 250 patients

www.thelancet.com/oncology Published online July 7, 2015 http://dx.doi.org/10.1016/S1470-2045(15)00121-7

3

Articles

Gefitinib group Placebo group (n=133) (n=132) Country of enrolment China

60 (45%)

58 (44%)

France

4 (3%)

5 (4%)

Germany

3 (2%)

3 (2%)

Hong Kong

2 (2%)

3 (2%)

Hungary

2 (2%)

2 (2%)

Italy

7 (5%)

9 (7%)

Japan

12 (9%)

11 (8%)

Russia

2 (2%)

0

South Korea

21 (16%)

21 (16%)

Spain

11 (8%)

11 (8%)

Taiwan

9 (7%)

9 (7%)

Age (years)

60 (33–79)

58 (35–79)

Age ≥65 years

43 (32%)

34 (26%)

Sex Female

87 (65%)

84 (64%)

Male

46 (35%)

48 (36%)

WHO performance status 0

55 (41%)

53 (40%)

1

78 (59%)

79 (60%)

Never smoker

88 (66%)

91 (69%)

Adenocarcinoma histological abnormalities

126 (95%)

131 (99%)

Metastatic at baseline

124 (93%)

119 (90%)

44 (33%)

31 (23%)

Brain metastases at baseline Previous response to first-line gefitinib Complete response

4 (3%)

2 (2%)

Partial response

87 (65%)

98 (74%)

Stable disease

42 (32%)

32 (24%)

Time to progression after initial gefitinib treatment ≤10 months

52 (39%)

58 (44%)

>10 months

81 (61%)

74 (56%)

Exon 19 deletion

85 (64%)

86 (65%)

L858R mutation

40 (30%)

42 (32%)

L861Q mutation

2 (2%)

2 (2%)

G719X (G719S/A/C) mutation

0

4 (3%)

T790M mutation

2 (2%)

0

S768I mutation

0

1 (1%)

Exon 20 insertions

2 (2%)

0

EGFR mutation subtype*

Other

10 (8%)

6 (5%)

Data are number (%) or median (range). Data are for the intention-to-treat population. *Patients’ tumours could have harboured more than one EGFR mutation subtype.

the basis of results of the phase 3, open-label, IPASS study7 that compared gefitinib with carboplatin plus paclitaxel in Asian patients. Planned subgroup analyses were undertaken to compare progression-free survival between treatments, including region (Asia-Pacific or Europe), time from first progression to randomisation (>2 weeks or ≤2 weeks), smoking history (never or present plus former), previous response to gefitinib (stable disease or partial plus complete response), EGFR mutation subtype (exon 19 deletion, L858R mutation), age (≥65 years or <65 years), sex (male or female), disease stage at diagnosis (locally advanced or metastatic), WHO performance status (0 or 1), time to progression on first-line gefitinib (>10 months or ≤10 months), and brain or CNS metastatic disease at baseline (yes or no). No interim analysis was done for progression-free survival. Objective responses and disease control were summarised in the ITT population and analysed by a logistic regression model adjusted for the same covariates as for progression-free survival. Overall survival was calculated in the ITT population and analysed as for progression-free survival. Safety and tolerability were assessed in all patients who received at least one dose of gefitinib (ie, the safety population), with adverse events summarised according to system organ class and the Medical Dictionary for Regulatory Activities. HRQoL analyses were done in all patients with a baseline HRQoL assessment and at least one post-baseline HRQoL assessment (ie, the quality-oflife population). Time to worsening was analysed using a Cox proportional hazards model that included terms for treatment received and covariates as defined for progression-free survival. HRQoL improvements were summarised descriptively by treatment and analysed as for objective responses, with no specific differences in HRQoL between treatment groups hypothesised. We undertook all analyses using SAS software (version 9.1.3 or higher), with a two-sided 5% significance level. An independent data monitoring committee did several analyses on Jan 21, 2013, July 18, 2013, and Jan 27, 2014, when 58, 155, and 165 patients had been recruited, and followed up with these patients to confirm the safety of gefitinib in combination with cisplatin plus pemetrexed. The independent data monitoring committee concluded that they had no safety concerns of the combination and did not request any changes to the study. This trial is registered with ClinicalTrials.gov, number NCT01544179.

Table 1: Demographics and baseline clinicopathological characteristics

Role of the funding source would need to be randomly assigned to achieve 190 progression-free survival events (76% maturity). This estimate was based on 90% power, with a two-sided 5% significance level, assuming a hazard ratio (HR) of 0·63 and a median progression-free survival of 9·5 months for gefitinib versus 6 months for placebo, on 4

This study was sponsored by AstraZeneca, who coordinated the trial, managed the database, and undertook the analyses. In collaboration with AstraZeneca, the study was designed by J-CS and TSKM and data were interpreted by the study steering committee (J-CS, Y-LW, KN, and TSKM). All data were collected and entered directly by study investigators into the web-based data capture system

www.thelancet.com/oncology Published online July 7, 2015 http://dx.doi.org/10.1016/S1470-2045(15)00121-7

Articles

Results Between March 29, 2012, and Dec 20, 2013, 265 patients were randomly assigned: 133 to the gefitinib group and 132 to the placebo group (figure 1). 264 patients received treatment (132 in each group). The appendix lists study recruitment sites, principal investigators, and numbers of patients enrolled. Patient sociodemographic and clinicopathological characteristics were generally well balanced between treatment groups (table 1). 206 (78%) of 265 patients were from the Asia-Pacific region. Use of doublet or triplet chemotherapy and of EGFR TKI treatment after discontinuation of the study drug was greater in the placebo group than in the gefitinib group (table 2). Median duration of follow-up for the primary analysis of progression-free survival was 11·2 months (IQR 8·0–15·0). Median duration of chemotherapy treatment was 117·0 days (IQR 81·0–127·0) in the gefitinib group and 122·0 days (46·5–132·0) in the placebo group. Median duration of gefitinib treatment was 152·5 days (IQR 92·5–216·5) and median duration of placebo treatment was 161·5 days (73·0–212·0). The median number of chemotherapy cycles was five in both the gefitinib (IQR three to six) and placebo (two to six) groups. At the time of data cutoff (May 5, 2014), 41 (15%) of 265 patients were still receiving study treatment: 23 patients in the gefitinib group and 18 patients in the placebo group. At the time of data cutoff, 98 (74%) of 133 patients had disease progression in the gefitinib group compared with 107 (81%) of 132 in the placebo group. There was no statistically significant difference in progression-free survival between the gefitinib and placebo groups (HR 0·86, 95% CI 0·65–1·13; p=0·27). Median progressionfree survival was 5·4 months in both groups (95% CI gefitinib group 4·5–5·7, placebo group 4·6–5·5); figure 2). Findings from subgroup analyses of progression-free survival were generally consistent with the overall result (figure 3). Independent masked central review confirmed this result (data not shown). 42 (32%) of 133 patients in the gefitinib group and 45 (34%) of 132 in the placebo group achieved an objective response (odds ratio [OR] 0·92, 95% CI 0·55–1·55; p=0·76). 112 (84%) of 133 patients in the gefitinib group and 104 (79%) of 132 in the placebo group achieved disease control (OR 1·39, 95% CI 0·74–2·62; p=0·31). Overall survival was immature at the time of data cutoff for progression-free survival, and not conclusive. 87 (33%) of 265 patients (50 [38%] of 133 in the gefitinib

Gefitinib group (n=133)

Placebo group (n=132)

Patients remaining on study treatment at data cutoff

23 (17%)

28 (21%)

Patients who received treatment after discontinuation*

61 (46%)

72 (55%)

Platinum alone†

0

Platinum-based regimen (doublet or triplet)

5 (4%)

2 (2%) 17 (13%)

Single-drug cytotoxic‡

27 (20%)

27 (20%)

EGFR tyrosine-kinase inhibitor§

27 (20%)

39 (30%)

Others

19 (14%)

17 (13%)

Data are number (%). Data are for the intention-to-treat population. *Patients could have received more than one treatment after discontinuation of the study drug and so can appear in more than one row. †Carboplatin, cisplatin, or nedaplatin. ‡Docetaxel, paclitaxel, pemetrexed, gemcitabine, or vinorelbine. §Gefitinib (nine in the gefitinib group and 14 in the placebo group), erlotinib (16 in the gefitinib group and 22 in the placebo group), or AZD9291 (two in the gefitinib group and three in the placebo group).

Table 2: Treatment after discontinuation

100

Gefitinib group Placebo group

90 80 Progression-free survival (%)

at the study site. In collaboration with AstraZeneca, these study data were interpreted by the study steering committee, after which the clinical study report was produced by AstraZeneca under the guidance and approval of the steering committee. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

70 60 50 40 30 20 10

HR 0·86 (95% CI 0·65–1·13); p=0·27*

0 0 Number at risk Gefitinib 133 Placebo 132

2

4

6

8

10

12

14

12 5

6 4

0 0

Time since randomisation (months) 110 100

88 85

40 39

25 17

Figure 2: Kaplan-Meier plot of progression-free survival Data are for the intention-to-treat population. HR=hazard ratio. *Primary Cox analysis with covariates. A HR <1 suggests a lower risk of progression with gefitinib than with placebo.

group and 37 [28%] of 132 in the placebo group) had died by data cutoff; follow-up is ongoing. The placebo group seemed to have longer overall survival than the gefitinib group (HR 1·62, 95% CI 1·05–2·52; p=0·03). Median overall survival was 17·2 months (95% CI 15·6 to not reached) in the placebo group versus 14·8 months (10·4 to 19·0) in the gefitinib group. We also did ad-hoc progression-free and overall survival analyses adjusting for the presence or absence of brain metastases at baseline (75 of 265 patients had brain metastases at baseline); the adjusted progression-free survival HR was 0·80 (95% CI 0·61–1·06; p=0·13) and the adjusted overall survival HR was 1·55 (1·00–2·41; p=0·05). Table 3 summarises the CTCAE adverse events that occurred in more than 10% of patients and adverse events of special interest. The appendix lists all grade 1–2 adverse events of more than 10% frequency and all adverse events of grade 3, 4, or 5. 126 (95%) of 132 patients in the gefitinib group and 130 (98%) of 132 patients in the

www.thelancet.com/oncology Published online July 7, 2015 http://dx.doi.org/10.1016/S1470-2045(15)00121-7

5

Articles

Number of events/sample size (%) All patients

HR (95% CI)

Gefitinib

Placebo

98/133 (74%)

107/132 (81%)

0·86 (0·65–1·13)

Age ≥65 years

32/43 (74%)

25/34 (74%)

1·01 (0·60–1·73)

<65 years

66/90 (73%)

82/98 (84%)

0·76 (0·55–1·06)

Sex Female

61/87 (70%)

67/84 (80%)

0·83 (0·58–1·17)

Male

37/46 (80%)

40/48 (83%)

0·83 (0·53–1·31)

Region Europe

20/29 (69%)

25/30 (83%)

0·95 (0·52–1·70)

Asia

78/104 (75%)

82/102 (80%)

0·80 (0·58–1·09)

Previous response to gefitinib Partial or complete response

70/91 (77%)

81/100 (81%)

0·97 (0·70–1·34)

Stable disease

28/42 (67%)

26/32 (81%)

0·59 (0·35–1·02)

Present or former

36/45 (80%)

34/41 (83%)

1·16 (0·72–1·86)

Never

62/88 (70%)

73/91 (80%)

0·70 (0·50–0·98)

95/129 (74%)

106/131 (81%)

0·83 (0·62–1·09)

>2 weeks

66/87 (76%)

66/79 (84%)

0·79 (0·56–1·11)

≤2 weeks

32/46 (70%)

41/53 (77%)

0·89 (0·56–1·41)

Smoking history

Disease stage at diagnosis* Metastatic Time from progression to randomisation

EFGR mutation subtype Exon 19 deletion

65/86 (76%)

71/86 (83%)

0·76 (0·54–1·07)

L858R mutation

28/40 (70%)

32/42 (76%)

1·08 (0·65–1·80)

Time to progression for initial gefitinib ≤10 months

39/52 (75%)

52/58 (90%)

0·86 (0·56–1·31)

>10 months

59/81 (73%)

55/74 (74%)

0·84 (0·58–1·21)

Not brain or CNS

64/89 (72%)

80/101 (79%)

0·84 (0·60–1·16)

Brain or CNS

34/44 (77%)

27/31 (87%)

0·66 (0·40–1·10)

Site of disease at baseline

WHO performance status score 0

38/55 (69%)

42/53 (79%)

0·68 (0·44–1·06)

1

60/78 (77%)

65/79 (82%)

0·95 (0·67–1·35) 0·10

1·00 Favours gefitinib

10·00 Favours placebo

Figure 3: Forest plot of progression-free survival subgroup analyses Data are for the intention-to-treat population. Number (%) of events are provided for illustrative purposes only. The subgroup analyses shown were undertaken using the same statistical methods (Cox model) as for the overall population, which take account of censoring. HR=hazard ratio. *Only five patients had locally advanced disease and so calculation of HR (95% CI) was not possible.

placebo group experienced adverse events. The most common adverse events reported were generally similar between the two treatment groups, with most being grade 1 or 2. Of the adverse events of special interest, diarrhoea, stomatitis, vomiting, and anaemia were reported more frequently by patients in the gefitinib group than those in the placebo group, with most being grade 1 or 2. Rash and hepatic dysfunction (ie, neutropenia, leucopenia, or nausea) were reported at a similar frequency between the two treatment groups. No adverse events of interstitial lung disease were reported. 37 (28%) of 132 patients in the gefitinib group and 28 (21%) of 132 patients in the placebo group reported serious adverse events. Adverse events leading to discontinuation were reported by ten (8%) of 132 patients in the gefitinib group and 13 (10%) of 132 in the placebo 6

group. Adverse events of grade 3 or higher were reported in 59 (45%) of 132 patients in the gefitinib group and 55 (42%) of 132 in the placebo group. 13 adverse events resulted in death (five in the gefitinib group and eight in the placebo group), of which two deaths in the gefitinib group were deemed causally related to gefitinib and cisplatin plus pemetrexed treatment (pneumonia and dyspnoea), and one in the placebo group was considered causally related to cisplatin plus pemetrexed treatment (haemoptysis). Ten (8%) patients in the gefitinib group and 11 (8%) in the placebo group died during the 30-day follow-up period after the last dose of study treatment. Grade 5 adverse events were experienced by five patients in the gefitinib group (one pneumonia, one dyspnoea, and three respiratory failure) and eight patients in the placebo group (one pneumonia, one cerebral infarction,

www.thelancet.com/oncology Published online July 7, 2015 http://dx.doi.org/10.1016/S1470-2045(15)00121-7

Articles

Gefitinib group (n=132)

Total*

Placebo group (n=132)

Grade 1–2

Grade 3

Grade 4

Grade 5

Grade 1–2

Grade 3

Grade 4

Grade 5

67 (51%)

48 (36%)

6 (5%)

5 (4%)

75 (57%)

42 (32%)

5 (4%)

8 (6%)

Adverse events with frequency >10%† Nausea‡

80 (61%)

5 (4%)

0

0

75 (57%)

6 (5%)

0

0

Decreased appetite

64 (48%)

1 (1%)

0

0

42 (32%)

3 (2%)

0

0

Vomiting‡

49 (37%)

6 (5%)

0

0

41 (31%)

3 (2%)

0

0

Anaemia‡

31 (23%)

11 (8%)

0

0

28 (21%)

5 (4%)

0

0

Constipation

34 (26%)

0

0

0

35 (27%)

0

0

0

Diarrhoea‡

41 (31%)

3 (2%)

0

0

18 (14%)

1 (1%)

0

0

Neutropenia‡

20 (15%)

8 (6%)

1 (1%)

0

21 (16%)

5 (4%)

2 (2%)

0

Fatigue

28 (21%)

0

0

0

23 (17%)

0

0

0

Leucopenia‡

24 (18%)

3 (2%)

0

0

19 (14%)

3 (2%)

0

0

Asthenia

13 (10%)

2 (2%)

0

0

26 (20%)

4 (3%)

0

0

8 (6%)

5 (4%)

3 (2%)

0

12 (9%)

10 (8%)

0

0 0

Decreased neutrophil count‡ Pyrexia

22 (17%)

0

0

0

14 (11%)

0

0

Cough

18 (14%)

0

0

0

15 (11%)

0

0

0

Decreased white blood cell count‡

10 (8%)

7 (5%)

0

0

11 (8%)

2 (2%)

0

0

Headache

10 (8%)

0

0

0

18 (14%)

1 (1%)

0

0

Dyspnoea

14 (11%)

1 (1%)

0

1 (1%)

7 (5%)

1 (1%)

1 (1%)

1 (1%)

Back pain

9 (7%)

2 (2%)

0

0

14 (11%)

0

0

0

0

0

0

1 (1%)

0

0

Rash‡

14 (11%)

0

0

0

11 (8%)

Stomatitis‡

13 (10%)

1 (1%)

0

0

4 (3%)

Additional adverse events of special interest§ Increased ALT concentration¶

16 (12%)

1 (1%)

0

0

21 (16%)

2 (2%)

0

0

Increased AST concentration¶

30 (23%)

0

0

0

29 (22%)

0

0

0

Interstitial lung disease

0

0

0

0

0

0

0

0

Haemoglobin decreased

0

0

0

0

1 (1%)

0

0

0

Data are number of patients (%). Data are for the safety population. Patients are listed according to their worst grade. Grades are according to the Common Terminology Criteria for Adverse Events. Events are listed according to the Medical Dictionary for Regulatory Activities preferred term. Adverse events with a frequency >10% are presented, regardless of treatment attribution. ALT=alanine aminotransferase. AST=aspartate aminotransferase. *Patients with any adverse event (not just those with >10% frequency). †Patients with several events in the same category are counted only once in that category. Patients with events in more than one category are counted once in each of those categories. ‡Events of special interest. §Causally related adverse events, sorted by international system organ class, except for increased ALT or AST concentrations. ¶Hepatic chemistry changes: patients worsening from baseline to any grade or a specified grade.

Table 3: Adverse events

one cerebrovascular accident, one cardiac failure, one dyspnoea, one haemoptysis, one gastric perforation, and one multiorgan failure). The appendix summarises the completion rates for the total FACT-L and TOI; reasons for missing HRQoL data were not recorded. Improvement in HRQoL was not significantly different between the gefitinib and placebo groups, as assessed by TOI (29·0% in the gefitinib group vs 30·2% in the placebo group; OR 0·92, 95% CI 0·53−1·59; p=0·77), FACT-L (35·5% vs 38·0%; 0·91, 0·54−1·53; p=0·73), or LCS (43·5% vs 42·6%; 1·01, 0·61−1·68; p=0·96), although such differences are descriptive only. Time to worsening of HRQoL as assessed by TOI was 12·1 weeks in the gefitinib group compared with 9·4 weeks in the placebo group (HR 0·91, 95% CI 0·68–1·21; p=0·51), as assessed by FACT-L was 12·0 weeks compared with 8·9 weeks (0·92, 0·69–1·23; p=0·58), and as assessed by LCS was 14·6 weeks compared with 9·1 weeks (0·89, 0·66–1·20; p=0·44).

Discussion IMPRESS—to our knowledge, the first randomised study to investigate optimum treatment strategies for patients with acquired resistance to first-line EGFR TKIs—did not show a significant improvement in progression-free survival with continuation of gefitinib in addition to platinum-based doublet chemotherapy versus chemotherapy alone. Independent masked central review (data not shown) confirmed this result. Furthermore, no statistical differences were noted between treatment groups for objective response or disease control, and improvement in HRQoL was also similar between groups. Therefore, platinum-based doublet chemotherapy remains the standard of care in this setting. The number of deaths at the time of progression-free survival analysis was fewer than expected at 87 (33%) of 265 compared with 125 estimated in the study protocol. The protocol estimate was based on results of the IPASS study;7 however, more recent studies have reported

www.thelancet.com/oncology Published online July 7, 2015 http://dx.doi.org/10.1016/S1470-2045(15)00121-7

7

Articles

longer overall survival.4,8 This difference might be due to several reasons, including more aggressive local treatment for brain metastases, or patient enrolment onto clinical trials with second-generation or thirdgeneration TKIs. Thus, the better than expected survival might not be unique to the IMPRESS study. Overall survival in IMPRESS was immature at the time of analysis, and thus the data presented here are not conclusive; however, overall survival seemed to be better in the placebo group than in the gefitinib group. The difference in overall survival was not related to toxic effects—most deaths were disease related and reported more than 30 days after the final dose of study treatment. A similar number of deaths were reported during treatment or within 30 days of the last dose in both treatment groups. After treatment, more patients in the placebo group than the gefitinib group were exposed to EGFR TKIs (20% in the gefitinib group vs 30% in the placebo group) and to doublet chemotherapy (4% in the gefitinib group vs 13% in the placebo group). Furthermore, there was an imbalance in a baseline characteristic in IMPRESS that was not adjusted for in the overall survival analysis; namely, more patients had brain metastases in the gefitinib group than in the placebo group (33% vs 23%). A median overall survival of 16 months has been reported in patients with EGFRmutation-positive NSCLC treated with EGFR TKIs who had brain metastases at baseline,22 which is shorter than the range of overall survival (19–35 months) reported in other studies for patients with EGFR-mutation-positive advanced NSCLC treated with first-line gefitinib.23,24 A post-hoc analysis in which brain metastasis was added as a covariate showed a slightly smaller difference in overall survival, and we plan to undertake further exploratory analyses when mature overall survival data are available; in particular, we will assess the T790M status of patients at the time of progression. Gefitinib in combination with cisplatin plus pemetrexed was well tolerated, with a safety and tolerability profile consistent with that known for each treatment. Rash was reported with a similar incidence in both treatment groups (gefitinib 11% vs placebo 8%) and at a lower incidence than in gefitinib-naive patients (33−79%),25 possibly because patients had received previous treatment with gefitinib at the time of study entry. Some adverse events generally expected with chemotherapy (ie, vomiting, stomatitis, anaemia, and fatigue) were reported with a higher frequency in the gefitinib group than in the placebo group. Investigation of the reasons why the primary endpoint of progression-free survival was similar between treatment groups is important, including whether IMPRESS was representative of the overall population of patients with EGFR-mutation-positive NSCLC and acquired resistance to EGFR TKIs, and how applicable the resistance criteria used in this trial were to the heterogeneous range of biological mechanisms for resistance. 8

The study population of IMPRESS was representative of the general population of patients with EGFRmutation-positive NSCLC and we included measures to ensure continuation of TKI treatment with minimum interruption. However, this study is limited by the fact that we did not classify the different types of progression. So far, we have learnt that patients with radiological progression on EGFR TKI therapy may be heterogeneous, with single-site, oligo-site, or multisite progression; each can be symptomatic or asymptomatic,26 and may need differing clinical management approaches.15 As shown in the ASPIRATION study,27 patients with slow progression or oligo-site progression should remain on first-line erlotinib, and the median progression-free survival may extend from 11·0 to 13·1 months. Our study results might have been different if we either excluded this group of patients or allowed treatment with TKIs beyond RECIST progression before randomisation. Furthermore, rebiopsies of samples to define the mechanisms of acquired resistance, such as T790M mutation, were not mandatory in our study. With the upcoming study results of AURA 2 and 3 (NCT02094261 and NCT02151981) and TIGER 2 and 3 (NCT02147990 and NCT02322281), subsequent treatment after first-line EGFR TKI failure could include AZD9291 and CO1686, respectively. If either of these drugs become the standard treatment for T790M-positive progression, the IMPRESS data might be applicable only to patients with non-T790M-related progression. Findings from our study suggest that progression-free survival of platinum-based doublet chemotherapy (cisplatin plus pemetrexed) in patients with acquired resistance to EGFR TKI treatment (median 5·4 months) is somewhat comparable to progression-free survival of first-line doublet chemotherapy (median 6·9 months).28 However, further data for doublet chemotherapy in the second-line setting after progression on first-line EGFR TKI treatments are needed to confirm this finding, such as the ongoing phase 3 open-label AURA 3 study of AZD9291 versus platinum-based doublet chemotherapy in patients with EGFR-mutation-positive NSCLC (with tumours harbouring a T790M mutation) and experience of disease progression on previous EGFR TKI therapy (NCT02151981). Contributors J-CS and TSKM were the principal investigators of the study and wrote the manuscript. Y-LW and KN contributed to protocol design, recruitment of patients, and data interpretation. S-WK, J-JY, M-JA, JW, JC-HY, YLu, SA, SP, DHL, YLi, KY, and J-YZ contributed to the recruitment of patients and data interpretation. XS contributed to management of the study and data interpretation. Statistical analysis was done under the direction of AW. HJ contributed to the protocol design and data interpretation. All authors reviewed and approved the final version of this manuscript. Declaration of interests J-CS is an adviser for and has received honoraria from AstraZeneca, Eli Lilly, Genentech-Roche, MSD, Merck Serono, Bristol-Myers Squibb, Pfizer, Taiho, Boehringer Ingelheim, Novartis, GlaxoSmithKline, Clovis Oncology, Amgen, Janssen, Servier, and Sanofi. Y-LW has received speaker fees from AstraZeneca, Roche, Eli Lilly, Pfizer, and Sanofi.

www.thelancet.com/oncology Published online July 7, 2015 http://dx.doi.org/10.1016/S1470-2045(15)00121-7

Articles

KN has received consultant fees from AstraZeneca and Eli Lilly. M-JA has been an adviser or consultant for AstraZeneca, Boehringer Ingelheim, Clovis Oncology, and Eli Lilly; and has received research funding from AstraZeneca and Eli Lilly. JW has received consultant fees from AstraZeneca, Eli Lilly, Pfizer, Roche, and Sanofi. JC-HY has received honoraria from AstraZeneca, Boehringer Ingelheim, Eli Lilly, Pfizer, and Roche; has been an adviser or consultant for Astellas, AstraZeneca, Bayer, Boehringer Ingelheim, Celgene, Clovis Oncology, Daiichi-Sankyo, Eli Lilly, Merck, MSD, Novartis, Pfizer, and Roche/Genentech; and has received research funding from Boehringer Ingelheim. YLu has received consultant fees from AstraZeneca, Eli Lilly, Roche, and Pfizer; and research funding from AstraZeneca and Pfizer. SA has received honoraria or consultant fees from AstraZeneca, Eli Lilly Japan, Chugai Pharmaceutical, Taiho Pharmaceutical, Boehringer Ingelheim, and Pfizer Japan. KY has received speaker fees from Chugai Pharmaceutical, Eli Lilly Japan, AstraZeneca, Taiho Pharmaceutical, Nippon Boehringer Ingelheim, and Pfizer Japan. XS, AW, and HJ are employees of and hold shares in AstraZeneca. TSKM has been an adviser for or received honoraria from AstraZeneca, Roche, Eli Lilly, Merck Serono, Eisai, Bristol-Myers Squibb, AVEO, Pfizer, Taiho, Boehringer Ingelheim, Novartis, GSK Biologicals, Clovis Oncology, Amgen, Janssen, and BioMarin Pharmaceutical; has been on speaker bureaus for AstraZeneca, Roche, Eli Lilly, Boehringer Ingelheim, Merck Serono, Pfizer, and Amgen International; and has received research funding from AstraZeneca. S-WK, J-JY, SP, DHL, YLi, and J-YZ declare no competing interests. Acknowledgments This study was funded by AstraZeneca. We thank the patients and investigators for their participation in this study. We thank the independent data and safety committee members for their critical review: M Reck, S Lu, and M Buyse. We thank L Brown, from Complete Medical Communications, who provided medical writing support funded by AstraZeneca. References 1 Mok T, Yang JJ, Lam KC. Treating patients with EGFR-sensitizing mutations: first line or second line—is there a difference? J Clin Oncol 2013; 31: 1081–88. 2 Reck M, Popat S, Reinmuth N, De Ruysscher D, Kerr KM, Peters S. Metastatic non-small-cell lung cancer (NSCLC): ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2014; 25 (suppl 3): iii27–39. 3 Han JY, Park K, Kim SW, et al. First-SIGNAL: first-line single-agent Iressa versus gemcitabine and cisplatin trial in never-smokers with adenocarcinoma of the lung. J Clin Oncol 2012; 30: 1122–28. 4 Yang JCH, Schuler M, Yamamoto N, et al. LUX-Lung 3: a randomized, open-label, phase III study of afatinib versus pemetrexed and cisplatin as first-line treatment for patients with advanced adenocarcinoma of the lung harboring EGFR-activating mutations. Proc Am Soc Clin Oncol 2012; 30 (18 suppl): LBA7500 (abstr). 5 Maemondo M, Inoue A, Kobayashi K, et al. Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR. N Engl J Med 2010; 362: 2380–88. 6 Mitsudomi T, Morita S, Yatabe Y, et al. Gefitinib versus cisplatin plus docetaxel in patients with non-small-cell lung cancer harbouring mutations of the epidermal growth factor receptor (WJTOG3405): an open label, randomised phase 3 trial. Lancet Oncol 2010; 11: 121–28. 7 Mok TS, Wu Y-L, Thongprasert S, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009; 361: 947–57. 8 Wu YL, Zhou C, Hu CP, et al. Afatinib versus cisplatin plus gemcitabine for first-line treatment of Asian patients with advanced non-small-cell lung cancer harbouring EGFR mutations (LUX-Lung 6): an open-label, randomised phase 3 trial. Lancet Oncol 2014; 15: 213–22. 9 Zhou C, Wu Y-L, Chen G, et al. Erlotinib versus chemotherapy as first-line treatment for patients with advanced EGFR mutationpositive non-small-cell lung cancer (OPTIMAL, CTONG-0802): a multicentre, open-label, randomised, phase 3 study. Lancet Oncol 2011; 12: 735–42.

10

11

12

13

14 15

16

17

18

19

20

21

22

23

24

25

26

27

28

Rosell R, Carcereny E, Gervais R, et al. Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non-small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial. Lancet Oncol 2012; 13: 239–46. Jackman D, Pao W, Riely GJ, et al. Clinical definition of acquired resistance to epidermal growth factor receptor tyrosine kinase inhibitors in non-small-cell lung cancer. J Clin Oncol 2010; 28: 357–60. Gainor JF, Shaw AT. Emerging paradigms in the development of resistance to tyrosine kinase inhibitors in lung cancer. J Clin Oncol 2013; 31: 3987–96. Eisenhauer EA, Therasse P, Bogaerts J, et al. New Response Evaluation Criteria in Solid Tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009; 45: 228–47. Chong CR, Jänne PA. The quest to overcome resistance to EGFR-targeted therapies in cancer. Nat Med 2013; 19: 1389–400. Yang JJ, Chen HJ, Yan HH, et al. Clinical modes of EGFR tyrosine kinase inhibitor failure and subsequent management in advanced non-small cell lung cancer. Lung Cancer 2013; 79: 33–39. Pao W, Miller VA, Politi KA, et al. Acquired resistance of lung adenocarcinomas to gefitinib or erlotinib is associated with a second mutation in the EGFR kinase domain. PLoS Med 2005; 2: e73. Engelman JA, Zejnullahu K, Mitsudomi T, et al. MET amplification leads to gefitinib resistance in lung cancer by activating ERBB3 signaling. Science 2007; 316: 1039–43. Sequist LV, Waltman BA, Dias-Santagata D, et al. Genotypic and histological evolution of lung cancers acquiring resistance to EGFR inhibitors. Sci Transl Med 2011; 3: 75ra26. Yu HA, Arcila ME, Rekhtman N, et al. Analysis of tumor specimens at the time of acquired resistance to EGFR-TKI therapy in 155 patients with EGFR-mutant lung cancers. Clin Cancer Res 2013; 19: 2240–47. Goldberg SB, Oxnard GR, Digumarthy S, et al. Chemotherapy with erlotinib or chemotherapy alone in advanced non-small cell lung cancer with acquired resistance to EGFR tyrosine kinase inhibitors. Oncologist 2013; 18: 1214–20. National Cancer Institute. National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. 2010. http://ctep.cancer.gov/protocolDevelopment/electronic_ applications/ctc.htm (accessed Jan 23, 2015). Park SJ, Kim HT, Lee DH, et al. Efficacy of epidermal growth factor receptor tyrosine kinase inhibitors for brain metastasis in non-small cell lung cancer patients harboring either exon 19 or 21 mutation. Lung Cancer 2012; 77: 556–60. Douillard J-Y, Ostoros G, Cobo M, et al. First-line gefitinib in caucasian EGFR mutation-positive NSCLC patients: a phase IV, open label, single arm study. Br J Cancer 2014; 110: 55–62. Mitsudomi T, Morita S, Yatabe Y, et al. Updated overall survival results of WJTOG 3405, a randomized phase III trial comparing gefitinib (G) with cisplatin plus docetaxel (CD) as the first-line treatment for patients with non-small cell lung cancer harboring mutations of the epidermal growth factor receptor (EGFR). Proc Am Soc Clin Oncol 2012; 30 (15 suppl): 7521 (abstr). Hirsch V. Managing treatment-related adverse events associated with EGFR tyrosine kinase inhibitors in advanced non-small-cell lung cancer. Curr Oncol 2011; 18: 126–38. Gandara DR, Li T, Lara PN, et al. Acquired resistance to targeted therapies against oncogene-driven non-small-cell lung cancer: approach to subtyping progressive disease and clinical implications. Clin Lung Cancer 2014; 15: 1–6. Park K, Ahn M, Yu C, et al. ASPIRATION: first-line erlotinib (E) until and beyond RECIST progression (PD) in Asian patient (pts) with EGFR mutation-positive (MUT+) NSCLC. Ann Oncol 2014; 25 (suppl 4): iv426-iv470 (abstr 1223O). Sequist LV, Yang JC, Yamamoto N, et al. Phase III study of afatinib or cisplatin plus pemetrexed in patients with metastatic lung adenocarcinoma with EGFR mutations. J Clin Oncol 2013; 31: 3327–34.

www.thelancet.com/oncology Published online July 7, 2015 http://dx.doi.org/10.1016/S1470-2045(15)00121-7

9