Osimertinib

Osimertinib compared docetaxel-bevacizumab as third-line treatment in EGFR T790M mutated non-small-cell lung cancer

Keke Niea, Zhongfa Zhanga, Chunling Zhangb, Chuanxin Genga, Ling Zhanga, Xiajuan Xub,
Shichao Liua, Songping Wanga, Xingjun Zhuangc, Ketao Lanb, YouXin Jib,⁎
a Department of Radiotherapy, Qingdao Cancer Hospital, Qingdao, 266042, China
b Department of Oncology, The Affiliated Qingdao Central Hospital of Qingdao University, Qingdao, 266042, China
c Department of Oncology, PLA 401 Hospital, Qingdao, 266001, China

Abstract

Objective: To compare the efficacy and toXicity of osimertinib versus docetaxel-bevacizumab as third-line treatment in EGFR T790M mutated NSCLC.

Methods: In this phase 3, open-label, three-center study, we randomly assigned (1:1) previously treated with TKI-chemotherapy or chemotherapy-TKI recurrent or metastatic advanced non-squamous lung cancer patients into two groups. These patients had acquired EGFR T790M resistance mutation confirmed by tumor tissues or serum. One group received oral osimertinib (80 mg/day) and the other group received intravenous infusion docetaxel (75 mg/m2) and bevacizumab (7.5 mg/kg) every 21 days until disease progression, unacceptable toXic effects or patient death. The primary endpoint of this study was progression-free survival (PFS) and the sec- ondary endpoints were response rates, toXicities and overall survival (OS). This trial was registered with ClinicalTrials.gov, number NCT02959749.

Results: A total of 147 patients were treated. Among them, 74 were enrolled in the osimertinib group and 73 were in the docetaxel-bevacizumab group. The median progression-free survival was 10.20 months in the osi- mertinib group versus 2.95 months in the docetaxel-bevacizumab group (hazard ratio 0.23; 95% confidence interval [CI], 0.12–0.38; P < 0.001). The overall response rate in the osimertinib group was significantly better than in the docetaxel-bevacizumab group (61.6%; 95% CI, 55.5–67.7 versus 8.3%; 95% CI, 1.3–15.3; p < 0.001). Because all the progressed patients in the docetaxel-bevacizumab group crossed over to the osi- mertinib group, there was no significant difference in the median OS between two groups at the time of last follow-up (hazard ratio 0.79; 95% CI, 0.38–1.61; P = .551). The main grade 3 or 4 toXic effects were diarrhea (2.7%) and interstitial lung disease (1.4%) in the osimertinib group and alopecia (15.3%), anorexia (12.5%), neutropenia (9.7%) and nausea (8.3%) in the docetaxel-bevacizumab group. Conclusions: Osimertinib had higher response rate, longer PFS and milder side effects than docetaxel-bev- acizumab in third-line therapy in patients with EGFR T790 M positive advanced NSCLC. 1. Introduction Acquired epidermal growth factor receptor (EGFR) T790M mutation is the most common genetic change after resistant to first-generation EGFR tyrosine kinase inhibitor (EGFR-TKI) in non-small-cell lung cancer. After 10–14 months with the treatment of first-generation EGFR-TKI, half of the patients will get disease progression [1,2]. The substitution of threonine with methionine at amino acid position 790 (T790M), as the second mutation in EGFR, is the most common re- sistance mechanism and is detected in tumor cells from more than 50–60% of patients after disease progression. This mutation enhances ATP affinity, reduces the ability of ATP-competitive reversible EGFR- TKI binding to EGFR tyrosine kinase domain, resulting in cancer cells resistant to gefitinib and erlotinib [3]. Osimertinib (Tagrisso, AZD9291, AstraZeneca) is an oral, potent, irreversible EGFR-TKI, which inhibits kinase activity of EGFR sensitive mutation and T790M resistance mutation [4,5]. In vitro and in vivo study, osimertinib showed strong inhibition of both the EGFR sensitive mutation and EGFR T790M resistance mutation. However, it showed less activity on wild type EGFR compared to first-generation EGFR-TKI (e.g., gefitinib or erlotinib) [6]. It had a high response rate in NSCLC patients who had harbored EGFR T790M mutation [7], and was approved for clinical use by FDA in November 2015. Fig. 1. Trial profile. Data cutoff date was May 30, 2016. Overall survival data was obtained in Feb. 15, 2017. For non-squamous NSCLC patients with disease progression after platinum-doublet chemotherapy, docetaxel with or without bev- acizumab are suggested agents for the third-line therapy if they are not exposed before [8–10]. Docetaxel combined with bevacizumab had higher response rate than docetaxel alone and the toXicities were tolerable [11,12]. Bevacizumab (Avastin, Roche) is a vascular endothelial growth factor A (VEGFA) monoclonal antibody, inhibits tumor angio- genesis, and is widely used in colorectal cancer and lung adenocarci- noma. Bevacizumab combined with VEGFA, attenuates VEGFA depen- dent tumor blood vessels formation, normalizes tumor blood vessels, prompts tumor cell apoptosis and finally shrinks tumor [13,14]. Com- bined with platinum-doublet chemotherapy, bevacizumab could in- crease ORR, prolong PFS and OS in Caucasian and Asian patients with advanced stage non-squamous NSCLC [8,15,16]. For patients with disease progression after treated with first line EGFR-TKI and second-line platinum-doublet chemotherapy, or che- motherapy and then EGFR-TKI, the optimal third-line therapy is criti- cally important for continuing patients’ survival. This study compared the efficacy and toXicity of osimertinib and docetaxel-bevacizumab as the third-line therapy in patients with local advanced or metastatic non-squamous NSCLC. 2. Methods 2.1. Study design and patient selection This was a three-center open-label randomized study. Study patients were pathologically, or, cytologically confirmed as having local ad- vanced, or metastatic non-squamous NSCLC with acquired EGFR T790M mutation. EGFR T790M mutation was diagnosed at progression of 2nd line therapy, e.g., after chemotherapy in TKI-chemotherapy treated patients or after TKI in chemotherapy-TKI treated patients. Patients treated with only one-line of therapy (1st generation EGFR-TKI or chemotherapy) were not eligible for this study. EGFR T790M mutation was confirmed by tumor tissues (109/147) or serum (38/147), which was tested by Amplification Refractory Mutation System (ARMS) (AmoyDX kit, China) before randomization. Liquid based EGFR test was done if patient’s tissue was not available. All participant patients were estimated to have a life expectancy of 3 months or longer at the time of registration. All patients had failed with TKI (gefitinib or erlotinib)-chemotherapy (paclitaxel + carboplatin/ cisplatin or carboplatin/cisplatin + pemetrexed) or chemotherapy (paclitaxel + carboplatin/cisplatin or carboplatin/cisplatin + peme- trexed)-TKI (gefitinib or erlotinib). Previously irradiated lesions were not selected as target lesions. None of these patients had received third- generation EGFR-TKI therapy previously. Local therapy must be com- pleted at least 4 weeks prior to the baseline scan. All the patients were 18–80 years of age with Eastern Cooperative Oncology Group (ECOG) performance status ≤2 and with adequate haematological, hepatic and renal function. This study was aimed to compare osimertinib with docetaxel-bevacizumab in third-line therapy in advanced EGFR T790M positive NSCLC. All eligible patients were randomly assigned into two groups, in a 1:1 ratio. Patients were randomized by center. Computerized rando- mization was done by center of the Affiliated Qingdao Central Hospital of Qingdao University using Microsoft EXcel 2007 formula, and was dispensed to researchers case by case. When patient was qualified to the trial and informed consent was signed, the trial center of Affiliated Qingdao Central Hospital of Qingdao University would be informed and randomization would be done. One group was received continuous oral osimertinib 80 mg once daily, the other group was treated with in- travenous infusion docetaxel (75 mg/m2) on day 1 and bevacizumab (7.5 mg/kg) on day 1, every 21 days a cycle. Randomization was per- formed with dynamic balancing [17] with respect to performance status, assessed using the World Health Organization (WHO) perfor- mance scale measure activity, sex, and EGFR mutation type. Osi- mertinib or docetaxel-bevacizumab treatment continued until: (1) the disease progression defined by Response Evaluation Criteria in Solid Tumors (RECIST) criteria 1.1; (2) Eastern Cooperative Oncology Group (ECOG) performance status changed to grade 3; (3) requested by pa- tients or physician; (4) uncontrollable serious adverse effects or death. Crossover of patients from the docetaxel-bevacizumab group to the osimertinib group was permitted in the case of disease progression by RECIST 1.1 or requested by physician. On request was defined as physician’s request based on patients’ condition. Study physician would decide to cross to osimertinib treatment when patient with life-threa- tening situation if crossover was not be done, or all other reasons docetaxel-bevacizumab could not be used continuously. Two dose adjustments of osimertinib were permitted, first to 80 mg every other day, then to 40 mg every other day. Patients treated with docetaxel-bevacizumab were also permitted to two times dose reduc- tion, each of them could be reduced 10% of total dose. Patients were withdrawn from the study if further dose reductions were needed (Fig. 1).Enrolled patients were required to have one or more evaluable target lesions which could be measured in one dimension according to RECIST 1.1. Patients were excluded if they were pre-treated with any third-generation EGFR-TKI therapies.This study was approved by the Ethics Committee of Affiliated Qingdao Central Hospital of Qingdao University, and was performed in compliance with the provisions of Good Clinical Practice guidelines, the Declaration of Helsinki and local laws. Informed consents were ob- tained from all enrolled patients before randomization. 2.2. Outcomes and assessment The primary endpoint of the study was PFS, which was measured from randomization date to disease progression or death. The sec- ondary endpoints were OS, response rates, toXicities, and quality of life (QOL). Tumor responses were assessed using RECIST 1.1 [18] observed during trial period and classified into: complete response (dis- appearance of tumor lesions), partial response (a decrease of at least 30% in the sum of tumor lesion sizes), stable disease (steady state of disease), or progressive disease (an increase ≥20% in the sum of tumor lesions sizes). All adverse events were recorded and classified by grade according to the National Cancer Institute Common Terminology Cri- teria for Adverse Events version 3.0. Tumor measurements with CT scan for lung and abdomen or ul- trasound for subcutaneous lesions and lymph nodes were performed at screening and every 4 weeks thereafter. We performed ultrasound every 4 weeks and CT scans every 2 months, but it was not limited for patients suspected in quick progression. Patients’ compliance, treatment safety,and side effects were accessed at each check point on every 4 weeks. 2.3. Statistical analysis Based on the CoX proportional hazards model, and taking into ac- count the influence of gender (male or female), ECOG performance status score (1 vs. 2), and EGFR exon mutation type (19 del vs. L858R), HR and 95% CI were calculated in the full analysis population. PFS and OS curves were analyzed by using Sigmaplot 11 (Systat software Inc., USA) Kaplan-Meiers log-rank test and the hazard ratio (HR) using CoX proportional hazards model in the intention-to-treat (ITT) people to compare osimertinib with docetaxel-bevacizumab, with mutated EGFR exon types as covariates.The response rate, symptom reduction, and treatment-related ad- verse events were assessed with Fisher’s exact test (all randomly as- signed patients received at least one dose of study drug). 3. Results From April 2015 to May 2016, we screened 212 patients, of which 147 were enrolled. The remaining 65 patients were excluded from the study during screening for they did not meet the inclusion standard. Randomization with dynamic balancing was performed. If an obvious imbalance happened, the next allocation would be allocated to the patient according to the study protocol AppendiX description, but it didn’t happen actually.The two groups were well balanced, with 74 patients were randomly assigned to the osimertinib group, among them, 1 patient did not start, 73 patients received osimertinib; 73 patients were assigned to the docetaxel-bevacizumab group with 1 patient withdrew (Fig. 1). No patient crossed over to the osimertinib group by physician’s request. All recruited patients were Chinese. We analyzed enrolled patients’ age, gender, smoking status, performance status, EGFR mutation types according to centers with Fisher’s exact text, there was no significant difference. All patients were in stage IV by AJCC 2007 staging system. A total of 70.7% of patients (104 of 147) were women, and 93.2% (137 of 147) had adenocarcinoma on histology (Table 1). The mean age of this study was slightly younger than in other studies, this might be influ- enced by study period, the study time span or lung cancer incidence in the area. Although patients’ ECOG performance status score 0–2 were qualified for the study, but no score 0 patient was enrolled. This might result from enrolled patients were all in advanced stages and had re- ceived at least one prior EGFR-TKI and received prior platinum-doublet cytotoXic chemotherapy. EGFR T790M mutation was detected in tumor tissue or serum in all of the enrolled patients. The last follow-up date was Feb 15, 2017. In this study, 1 patient did not start in the osimertinib group and 1 patient withdrew in the docetaxel + bevacizumab group. A total of 255 cycles of docetaxel + bevacizumab were delivered to 72 patients, average of 3.49 (median 3) cycles for each person. There was no dif- ference between ITT and full analysis population. The median PFS was 10.20 months and 2.95 months for the osimertinib group and the docetaxel-bevacizumab group respectively (HR 0.23; 95% CI, 0.12–0.38; P < 0.001) (Fig. 2A). The median OS was not reached in either group (HR 0.79; 95% CI, 0.38–1.61; P = 0.551) (Fig. 2B). The PFS and OS in the osimertinib group were also analyzed according to EGFR mutation type. Patients had a consistent benefit from osimertinib therapy, whatever T790M mutation + exon 19 del or T790M muta- tion + L858R mutation their tumor harbored (Fig. 3 A–B). The response rate was 61.6% (95% CI, 55.5–67.7) in the osimertinib group and 8.3% (95% CI, 1.3–15.3) in the docetaxel-bevacizumab group respectively (P < 0.001). Two patients reached complete response in the osimertinib group but none in the docetaxel-bevacizumab group. There was no significant difference in response rate for EGFR T790M status determined by blood or tissue. The disease control rate (complete response + partial response + stable disease) was 87.7% (95% CI, 79.8–95.4) in the osimertinib group, significantly higher than that in the docetaxel-bevacizumab group 43.1% (95% CI, 35.1–50.9) (P < 0.001) (Table 2). Qualities of life (QOL) of patients were also accessed between the two groups at baseline and at last follow-up, there was no significant difference. The main adverse effect of osimertinib was rash or acne of the skin, which happened in 35.6% patients, higher than the incidence in the docetaxel-bevacizumab group (1.4%, P < 0.001) (Table 3). The in- cidences of diarrhea, and dry skin were higher in the osimertinib group than in the docetaxel-bevacizumab group (26.0% vs. 6.9%, 28.8% vs.1.4%), but the anorexia, nausea and vomiting were much higher in the docetaxel-bevacizumab group (Table 3). Nine patients in docetaxel- bevacizumab group had dose reductions because of neutropenia or GI toXicities; two of them had a second dose reduction because of grade 4 neutropenia. One patient suffered mild epistaxis after 7-cycle of doc- etaxel-bevacizumab treatment; but there was no deep vein thrombosis, pulmonary embolism or neuropathy happened in this group. Three patients suffered from interstitial lung disease (ILD), two of them discontinued osimertinib study, and one was cured by dose reduction and steroids therapy. Five patients suffered from asymptomic mild prolonged Q-T interval on ECG after 2–3 months osimertinib and this phenomenon lasted until the end of the treatment. One patient dis- continued study because of persistent grade 3 diarrhea after two times dose reductions of osimertinib. There was no treatment-related death in either group (Table 3). All patients had brain MRI or brain CAT-scan at disease progression. SiX patients in the docetaxel-bevacizumab group developed brain metastasis and all crossed over to osimertinib treat- ment. Two patients in the osimertinib group had brain metastasis were treated by brain irradiations. Fig. 2. Kaplan-Meier analysis of progression-free survival (A) and overall sur- vival (B) in the full analysis set. HR = hazard ratio. 4. Discussion As the trial was designed in 2014, the guidelines recommended therapy for EGFR or ALK positive NSCLC with advanced stage was TKI or chemotherapy alternation, or platinum-doublet chemotherapy for EGFR or ALK wild type or mutation unknown. Docetaxel was the standard third-line therapy for recurrent or metastatic patients if it had not been used previously. Bevacizumab was commonly combined with chemotherapy in first-line or maintenance therapy in NSCLC, it pro- longed patients’ PFS and OS when it administered with paclitaxel and carboplatin [15,16]. We didn’t choose paclitaxel plus bevacizumab because 65.3% patients (96/147) were exposed to paclitaxel in previous first-line or second-line chemotherapy. Fig. 3. Kaplan-Meier analysis of progression-free survival (A) and overall survival (B) according to EGFR mutation type. HR = hazard ratio. EGFR T790M mutation is the most common cause of acquired drug resistance of the first-line EGFR-TKI. Monotherapy with osimertinib was associated with a response rate of 61%, PFS 9.6 months, with limited skin and gastrointestinal adverse effects among patients with positive EGFR T790M [4]. Although the studies of osimertinib in treatment of NSCLC after resistant to gefitinib or erlotinib was suc- cessfully carried out, to our knowledge, this is the first study of osi- mertinib in third-line treatment of patients with non-small-cell lung cancer harboring EGFR T790M. Median progression-free survival for NSCLC patients receiving third-line therapy is about 2–4 months [19,20]. In our study, the response rate was 61.6% and PFS was 10.20 months in the osimertinib treated patients, and response rate was 8.3% and PFS was 2.95 months in the docetaxel-bevacizumab treated patients (hazard ratio for pro- gression in the osimertinib group, 0.23; 95% CI, 0.12–0.38; P < 0.001). Patients in the osimertinib group had a much higher ORR and a much longer PFS than in the docetaxel-bevacizumab group. More than half of the patients were still alive at the time of last follow-up; the median overall survival was 15.65 months in the osimertinib group (21.6% mutation), and was 15.29 months in the docetaxel-bevacizumab group (23.3% mutation). The OS were almost the same in the two groups; this might result from patients crossed over to the osimertinib treatment after disease progression with docetaxel-bevacizumab. These results confirmed EGFR T790M was a positive predictive factor for osimertinib therapy even in third-line therapy and a negative factor for chemotherapy or VEGF monoclonal antibody. The docetaxel-bevacizumab group had a lower response rate (8.3%), this scenario indicated patients with EGFR. T790M mutation were not only resistant to the first-generation TKI (e.g. gefitinib or erlotinib) therapy but also resistant to chemotherapy and VEGF monoclonal antibody. With the previous studies, NSCLC patients harbored EGFR exon 19 del had better prognosis than EGFR L858R mutation after treated with erlotinib, gefitinib or afatinib [21–24]. Gene polymorphism, pathway activation and smoking status are associated with poor response to EGFR-TKI in patients with EGFR L858R mutation lung adenocarcinoma [21,25,26]. Through analyzing EGFR mutation type, we did not find survival difference between EGFR exon 19 del or exon L858R mutation in patients treated by osimertinib (Fig. 3). The median PFS of 41 patients with exon 19 del and T790M mutation and 33 patients with L858R and T790M mutation were 9.54 months and 10.97 months respectively (hazard ratio for progression, 1.272; 95% CI, 0.76–2.33; P = .361). The median OS was not reached (hazard ratio for death, 0.78; 95% CI, 0.26–2.27; P = .638). After studying each subgroup’s 95% CIs (s), we found that there was no interaction between the two exons mutations; less likely detected a sig- nificant difference between them. Because this study was not designed for subgroups power analysis, EGFR exon 19 del or exon L858R sub- groups were compared descriptively rather than statistically. Since the sample of the two subgroups sizes were small, probably too small to analyze adequate power. Further study to compare the treatment re- sults of osimertinib between patients with EGFR 19 del + T790M mu- tation and EGFR L858R + T790M mutation is needed in the future. Most patients’ EGFR T790M mutation was diagnosed by tumor tissues; there were no significant difference in response rate or survival for EGFR T790M status determined by blood or tissue.There were siX patients in the docetaxel-bevacizumab group and two patients in the osimertinib group had brain metastasis at progres- sion. The incidence of brain metastasis was lower than previous re- ported [27]. This might be most patients were still in the study, and patients of docetaxel-bevacizumab group crossed over to osimertinib treatment after disease progression. Currently, sequentially using first-generation EGFR-TKI then changes to osimertinib is recommended by guideline if non-small-cell lung cancer harboring sensitive EGFR mutation (EGFR 19 del or EGFR L858R) changes to resistant mutation (EGFR T790M). First-line treat- ment with osimertinib showed significant superiority of osimertinib over erlotinib or gefitinb in untreated EGFR mutated advanced non- small-cell lung cancer in patients with or without CNS metastases [28]. Based on this study, FDA approval osimertinib for the first-line therapy in EGFR mutant NSCLC is expected later this year. The mechanism of resistance to third generation EGFR-TKI is much more complicated; and the survival result is gloomy if resistance occurs. Our data showed that, osimertinib could prolong PFS in NSCLC patients harboring EGFR T790M mutation whether their previously treatment sequence of TKI- chemotherapy or chemotherapy-TKI used. In summary, this study established the superiority of osimertinib when compared with docetaxel-bevacizumab, showed that osimertinib was a valid therapeutic option for patients with non-small-cell lung cancer harboring EGFR T790M mutation after failed to chemotherapy and first-generation EGFR-TKI. Authors’ contributions KN and YJ designed the study. YJ, ZZ, CZ, CG, LZ, XJ, SL, SW en- rolled and follow-up patients. XZ and KL did statistical analysis. YJ gave research consultation and helped to draft the manuscript. All authors read and approved the final manuscript. Ethics approval and consent to participate Yes. Consent for publication Yes. Conflict of interest All authors declared there was no conflict of interest. Acknowledgements We sincerely appreciate the superb help by Rong Wang and Lei Peng on data collection at Department of Medical Library, the Affiliated Central Hospital of Qingdao University. We also thank Professor Wendy Wang of Feinstein Institute for Medical Research and Professor John Califra in New York helped in editing the manuscript. This study was supported by Grants from National Health and Family Planning Commission of Peoples Republic of China, Medical Technology Research Center for Health Development Grants W2012FZ007(YJ). References [1] Z. Yu, T.J. Boggon, S. Kobayashi, et al., Resistance to an irreversible epidermal growth factor receptor (EGFR) inhibitor in EGFR-mutant lung cancer reveals novel treatment strategies, Cancer Res. 67 (21) (2007) 10417–10427. [2] S. Kobayashi, H. Ji, Y. Yuza, et al., An alternative inhibitor overcomes resistance caused by a mutation of the epidermal growth factor receptor, Cancer Res. 65 (16) (2005) 7096–7101. [3] K. Suda, R. Onozato, Y. Yatabe, T. Mitsudomi, EGFR T790M mutation: a double role in lung cancer cell survival? J. Thorac. Oncol. 4 (1) (2009) 1–4. [4] P.A. Janne, J.C. Yang, D.W. Kim, et al., AZD9291 in EGFR inhibitor-resistant non- small-cell lung cancer, N. Engl. J. Med. 372 (18) (2015) 1689–1699. [5] T. Jiang, C. Zhou, Clinical activity of the mutant-selective EGFR inhibitor AZD9291 in patients with EGFR inhibitor-resistant non-small cell lung cancer, Transl. Lung Cancer Res. 3 (6) (2014) 370–372. [6] D.A. Cross, S.E. Ashton, S. Ghiorghiu, et al., AZD9291, an irreversible EGFR-TKI, overcomes T790M-mediated resistance to EGFR inhibitors in lung cancer, Cancer Discov. 4 (9) (2014) 1046–1061. [7] T.M. Kim, A. Song, D.W. Kim, et al., Mechanisms of acquired resistance to AZD9291: a mutation-selective, irreversible EGFR inhibitor, J. Thorac. Oncol. 10 (12) (2015) 1736–1744. [8] R.G. Zinner, C.K. Obasaju, D.R. Spigel, et al., PRONOUNCE: randomized, open- label, phase III study of first-line pemetrexed+carboplatin followed by main- tenance pemetrexed versus paclitaxel+carboplatin+bevacizumab followed by maintenance bevacizumab in patients ith advanced nonsquamous non-small-cell lung cancer, J. Thorac. Oncol. 10 (1) (2015) 134–142. [9] C. Manegold, Bevacizumab for the treatment of advanced non-small-cell lung cancer, EXpert Rev. Anticancer Ther. 8 (5) (2008) 689–699. [10] C. Kerr, Bevacizumab and chemotherapy improves survival in NSCLC, Lancet Oncol. 6 (5) (2005) 266. [11] Y. Takagi, Y. Hosomi, F. Oshita, et al., Feasibility study of docetaxel plus bev- acizumab as first line therapy for elderly patients with advanced non-small-cell lung cancer: Thoracic Oncology Research Group (TORG) 1014, BMC Cancer 15 (2015) 740. [12] K. Kurishima, H. Watanabe, H. Ishikawa, H. Satoh, N. Hizawa, A retrospective study of docetaxel and bevacizumab as a second- or later-line chemotherapy for non-small cell lung cancer, Mol. Clin. Oncol. 7 (1) (2017) 131–134. [13] A. Sandler, Bevacizumab in non small cell lung cancer, Clin. Cancer Res. 13 (Pt. (2)) (2007) s4613–s4616. [14] L. Horn, A. Sandler, Chemotherapy and antiangiogenic agents in non-small-cell lung cancer, Clin. Lung Cancer 8 (Suppl. 2) (2007) S68–S73. [15] C. Zhou, Y.L. Wu, G. Chen, et al., BEYOND: a randomized, double-blind, placebo- controlled, multicenter, phase III study of first-line carboplatin/paclitaxel plus bevacizumab or placebo in chinese patients with advanced or recurrent non- squamous non-small-cell lung cancer, J. Clin. Oncol. 33 (19) (2015) 2197–2204. [16] A. Sandler, R. Gray, M.C. Perry, et al., Paclitaxel-carboplatin alone or with bev- acizumab for non-small-cell lung cancer, N. Engl. J. Med. 355 (24) (2006) 2542–2550. [17] S.J. Pocock, R. Simon, Sequential treatment assignment with balancing for prog- nostic factors in the controlled clinical trial, Biometrics 31 (1) (1975) 103–115. [18] E.A. Eisenhauer, P. Therasse, J. Bogaerts, et al., New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1), Eur. J. Cancer 45 (2) (2009) 228–247. [19] T. Shukuya, R. Ko, K. Mori, et al., Prognostic factors in non-small cell lung cancer patients who are recommended to receive single-agent chemotherapy (docetaxel or pemetrexed) as a second- or third-line chemotherapy: in the era of oncogenic dri- vers and molecular-targeted agents, Cancer Chemother. Pharmacol. 76 (4) (2015) 771–776. [20] G.Z. Ying, J.S. Chang, L.S. Cui, et al., Third-line therapy in advanced non-small cell lung cancer, J. BUON 18 (4) (2013) 899–907. [21] T. Fukuhara, M. Maemondo, A. Inoue, et al., Factors associated with a poor re- sponse to gefitinib in the NEJ002 study: smoking and the L858R mutation, Lung Cancer 88 (2) (2015) 181–186. [22] X. Chen, Q. Zhu, L. Zhu, et al., Clinical perspective of afatinib in non-small cell lung cancer, Lung Cancer 81 (2) (2013) 155–161. [23] J.M. Sun, Y.W. Won, S.T. Kim, et al., The different efficacy of gefitinib or erlotinib according to epidermal growth factor receptor exon 19 and exon 21 mutations in Korean non-small cell lung cancer patients, J. Cancer Res. Clin. Oncol. 137 (4) (2011) 687–694. [24] G.J. Riely, W. Pao, D. Pham, et al., Clinical course of patients with non-small cell lung cancer and epidermal growth factor receptor exon 19 and exon 21 mutations treated with gefitinib or erlotinib, Clin. Cancer Res. 12 (Pt. (1)) (2006) 839–844. [25] C.H. Chiu, H.L. Ho, H. Doong, et al., MLH1 V384D polymorphism associates with poor response to EGFR tyrosine kinase inhibitors in patients with EGFR L858R- positive lung adenocarcinoma, Oncotarget 6 (10) (2015) 8407–8417. [26] P. Ozenne, D. Dayde, E. Brambilla, B. Eymin, S. Gazzeri, p14(ARF) Inhibits the growth of lung adenocarcinoma cells harbouring an EGFR L858R mutation by ac- tivating a STAT3-dependent pro-apoptotic signalling pathway, Oncogene 32 (8) (2013) 1050–1058. [27] F. Hsu, C.A. De, D. Anderson, et al., Patterns of spread and prognostic implications of lung cancer metastasis in an era of driver mutations, Curr. Oncol. 24 (4) (2017) 228–233. [28] J.C. Soria, Y. Ohe, J. Vansteenkiste, et al., Osimertinib in untreated EGFR-mutated advanced non-small-cell lung cancer, N. Engl. J. Med. 378 (2) (2018) 113–125.