Journal of Thoracic Oncology:
A Retrospective Analysis of VeriStrat Status on Outcome of a Randomized Phase II Trial of First-Line Therapy with Gemcitabine, Erlotinib, or the Combination in Elderly Patients (Age 70 Years or Older) with Stage IIIB/IV Non–Small-Cell Lung Cancer
Stinchcombe, Thomas E. MD*; Roder, Joanna PhD†; Peterman, Amy H. PhD‡; Grigorieva, Julia PhD†; Lee, Carrie B. MD, MPH*; Moore, Dominic T. MPH§; Socinski, Mark A. MD‖
*Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, NC; †Biodesix, Inc. Boulder, CO; ‡Department of Psychology, University of North Carolina at Charlotte, Charlotte, NC; §Division of Biostatistics and Data Management, Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, NC; and ‖Division of Hematology and Oncology, University of Pittsburgh Cancer Institute, Pittsburgh, PA.
This paper was an oral presentation at the Chicago Multidisciplinary Symposium in Thoracic Oncology in Chicago, Illinois, September 6–8, 2012.
Disclosure: Drs. Roder and Grigorieva are employed by Biodesix, Inc., the company that developed and provides VeriStrat testing(, have Biodesix stock options, and hold patents related to the VeriStrat test. The other authors declare no conflicts of interest.
Author for correspondence: Thomas E. Stinchcombe, Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, 170 Manning Drive, Physician’s Office Building, 3rd Floor, Chapel Hill, NC 27599-7305. E-mail: firstname.lastname@example.org
Purpose: In a multicenter randomized phase II trial of gemcitabine (arm A), erlotinib (arm B), and gemcitabine and erlotinib (arm C), similar progression-free survival (PFS) and overall survival (OS) were observed in all arms. We performed an exploratory, blinded, retrospective analysis of plasma or serum samples collected as part of the trial to investigate the ability of VeriStrat (VS) to predict treatment outcomes.
Methods: Ninety-eight patients were assessable, and the majority had stage IV disease (81%), adenocarcinoma histology (63%), reported current or previous tobacco use (84%), and 26% had a performance status (PS) of 2.
Results: In arm A, patients with VS Good (n = 20) compared with VS Poor status (n = 8) had similar PFS (hazard ratio [HR]: 1.21; p = 0.67) and OS (HR: 0.82; p = 0.64). In arm B, patients with VS Good (n = 26) compared with VS Poor (n = 12) had a statistically significantly superior PFS (HR: 0.33; p = 0.002) and OS (HR: 0.40; p = 0.014). In arm C, patients with VS Good (n = 17) compared with Poor (n = 1 5) had a superior PFS (HR: 0.42; p = 0.027) and a trend toward superior OS (HR: 0.48; p = 0.051). In the multivariate analysis for PFS, VS status was statistically significant (p = 0.011); for OS, VS status (p = 0.017) and PS (p = 0.005) were statistically significant. A statistically significant VS and treatment interaction (gemcitabine versus erlotinib) was observed for PFS and OS.
Conclusions: Gemcitabine is the superior treatment for elderly patients with VS Poor status. First-line erlotinib for elderly patients with VS Good status may warrant further investigation.
The management of elderly patients with stage IIIB or IV non–small-cell lung cancer (NSCLC) remains controversial. Cytotoxic chemotherapy remains the standard therapy for the majority of patients, but the optimal treatment for this heterogeneous patient population is unclear. Many “fit” elderly patients tolerate platinum-based therapy and are frequently enrolled in trials of platinum-based therapy. Other elderly patients have cardiopulmonary comorbidities related to tobacco use and comorbidities associated advanced age. These patients represent the “frail” elderly and have difficulty tolerating cytotoxic chemotherapy.1 Trials investigating novel agents, optimal chemotherapy schedules and combinations, and developing biomarkers of efficacy and/or tolerance of therapy for the frail elderly patient population are a priority for the thoracic oncology community.
Beginning in 2005, we performed a noncomparative randomized phase II trial of gemcitabine, erlotinib, and erlotinib and gemcitabine in elderly patients (age 70 years or older) with stage IIIB or IV NSCLC.2 At the time the study was designed, single-agent chemotherapy was a standard of care based on prospective phase III trials in elderly patients.3–5 Erlotinib, an epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI), was a standard therapy in the second- and third-line settings, and a single arm phase II trial had revealed promising activity in elderly patients.6,7 A phase III trial of gemcitabine and erlotinib compared with gemcitabine in advanced pancreatic cancer had revealed superior overall survival (OS) with the combination therapy.8 The results of the intent-to-treat patient population have previously been published;2 neither of the investigational arms (erlotinib alone or with gemcitabine) demonstrated sufficient activity to warrant further investigation. The role of EGFR mutation status in the selection of patients for first-line EGFR TKI therapy was not established at the time the trial was designed.9–12
VeriStrat is a commercially available serum or plasma test using matrix-assisted laser desorption ionization mass spectrometry methods. It was developed on a training set of pretreatment serum samples from patients with advanced NSCLC who experienced either long-term stable disease or early progression on gefitinib therapy.13 Mass spectra (MS) from these patients’ serum samples were used to define eight MS features (i.e., peaks), differentiating these two outcome groups. An algorithm utilizing these features and based on k-nearest neighbors classification scheme was created, and its parameters were optimized using additional spectra from the training cohort. The current commercial test uses the same fixed set of parameters established during the development phase. VeriStrat assigns each sample a classification of VeriStrat Good or VeriStrat Poor; when an unequivocal classification cannot be determined (<3% of samples), an indeterminate result is reported.
Validation studies of VeriStrat were performed in a blinded fashion using multiple single-arm cohorts of patients with NSCLC undergoing EGFR TKI therapy.13 Retrospective analysis of available plasma samples from the phase III registrational trial of erlotinib, National Cancer Institute of Canada Clinical Trials Group BR.21, confirmed VeriStrat’s ability to separate patients with advanced NSCLC into groups with better and worse outcomes with erlotinib therapy. VeriStrat status demonstrated prognostic properties and was predictive of response to erlotinib. This study confirmed previous results that VeriStrat classification is not significantly correlated with EGFR or KRAS mutation status, and the absence of a correlation with EGFR gene copy number.14–16 Although VeriStrat classification significantly correlated with certain prognostic characteristics, such as performance status (PS), it maintained a significant correlation with outcomes independent of these potential confounding factors in multivariate analysis.13,14
In this study we performed a retrospective analysis of the clinical outcomes of patients classified as VeriStrat Good and Veristrat Poor in each treatment arm. Although the analysis in the erlotinib arm is similar to previous studies, those in the other treatment arms represent the first studies of VeriStrat testing in gemcitabine and gemcitabine plus erlotinib treated patients and therefore are exploratory.
PATIENTS AND METHODS
Patients were required to have histologic or cytologic diagnosis of NSCLC, AJCC 6th edition stage IIIB or IV disease, age 70 years or older, Eastern Cooperative Oncology Group PS of 0 to 2, and adequate bone marrow, renal, and hepatic function. Patients were required to have measureable disease according to Response Evaluation Criteria in Solid Tumors (RECIST).17 Patients who were unable to provide informed consent or participate in the Health Related Quality of Life (HRQOL) questionnaires were not eligible. There were no eligibility requirements related to histology, history of tobacco use, or EGFR mutation status. This trial was reviewed by the institutional review board of all the participating centers, and patients were required to provide informed consent before any study related tests were performed. The study was registered with Clinicaltrials.gov (NCT00283244). The protocol was amended in December 2007, and a correlative science study with collection of peripheral blood samples was incorporated into the study. Participation was voluntary and patients who agreed to participate in the correlative science study signed a separate institutional review board approved informed consent document. Collection of tumor samples was not required; therefore, tumor samples are not available for analysis for EGFR and KRAS mutational status.
Patients were randomly assigned to treatment arms A, B, or C; patients were stratified based on gender, smoking history (never or light smoking history versus current or former tobacco use), and PS (2 versus 0 or 1). Patients assigned to arm A received gemcitabine 1200 mg/m2 intravenously on days 1 and 8 every 21 days until disease progression, unacceptable toxicity, or a maximum of four cycles. At the time of disease progression patients were offered erlotinib 150 mg daily until disease progression or unacceptable toxicity as part of the trial. Patients assigned to arm B received erlotinib 150 mg daily until disease progression or unacceptable toxicity. Patients assigned to arm C received gemcitabine 1000 mg/m2 intravenously on days 1 and 8 every 21 days and erlotinib 100 mg daily; patients received gemcitabine until disease progression, unacceptable toxicity, or a maximum of four cycles. After four cycles (in the absence of disease progression or unacceptable toxicity), patients continued erlotinib until disease progression or unacceptable toxicity. Details about the dose adjustments for gemcitabine and erlotinib are available in the previous publication.2
Patients were required to have a staging computed tomography scan of the chest and abdomen (including the liver and adrenals) within 4 weeks of trial enrollment. Bone scan, positron emission tomography, and computed tomography scan or magnetic imaging of the brain were not required and were performed if clinically indicated. Disease status according to RECIST was assessed after cycles 2 and 4 and at 6 months from the beginning of therapy, and then every 2 months or if clinically indicated. Disease status was assessed by the investigator. Patients underwent laboratory and physical examinations and toxicity assessment with each cycle.2 Patients were evaluated using the Functional Assessment of Cancer Therapy for Lung Cancer (FACT-L) which consists of the FACT-General and the lung cancer–specific subscale (LCS).18,19 The FACT-L was administered at the screening visit (within 2 weeks of the day 1 of the first treatment cycle) or day 1 of the first treatment cycle (baseline), after each cycle (21 days), after the completion of treatment or at the time the patient was withdrawn from the study. The Trial Outcome Index-Lung (TOI-L) was the primary HRQOL analyzed.
Mass spectrometry was performed in a fully blinded manner on 110 available pretreatment plasma and serum samples sent to Biodesix (Boulder, CO). Sample aliquots were diluted 1:10 in HPLC-grade water and mixed (1:1 v/v) with matrix solution (25 mg/ml sinapinic acid dissolved in 50/50/0.1% acetonitrile:water:trifluoroacetic acid). The dilute sample-matrix mixture was spotted in triplicate on a matrix-assisted laser desorption ionization target in randomly assigned plate positions. Spectra were acquired on a Bruker Flextreme mass spectrometer. Each replicate spectrum consisted of an average of 2000 individual spectra collected from various locations within the spot. The MS were then processed by the VeriStrat classification algorithm, which is identical to the one previously described by Taguchi et al.13 The VeriStrat test performed was the same as that provided commercially, and testing is conducted by Biodesix, Inc. in their Clinical Laboratory Improvement Amendments (CLIA)-certified laboratory. No adjustments were made to the test, which has remained fully locked since development in 2005. Testing was carried out blinded to all clinical information.
Statistical Methods and Study Design
This retrospective analysis was designed as an exploratory study of the use of the VeriStrat proteomic test in first-line elderly patients with advanced NSCLC treated with gemcitabine, erlotinib, or the combination of gemcitabine and erlotinib. Statistical analyses were performed using SAS 9.2 (Cary, NC) and PRISM (GraphPad, La Jolla, CA). Progression-free survival (PFS) and OS were summarized using the Kaplan–Meier method and compared between treatment arms and VeriStrat groups using the log-rank test. Hazard ratios (HRs) between groups for time-to-event variables were calculated using Cox proportional hazard methods. The impact of baseline prognostic factors on outcome was explored with Cox regression models using forward selection, backward elimination, or stepwise selection of covariates with a fixed selection parameter of α = 0.1. Correlations of categorical variables with VeriStrat classification were assessed using Fisher’s exact test or a χ2 test. Statistical significance was set at a level of 0.05 for all analyses.
Between March 2006 and May 2010, 146 eligible patients were enrolled and initiated trial therapy. Plasma or serum samples were available from 110 patients, and 98 were assessable for analysis (Fig. 1). The clinical characteristics of the patients are presented in Table 1.2 Of the 98 patients in the VeriStrat analysis cohort, 63 were classified as VeriStrat Good and 35 as Poor. Tumor samples were not available for EGFR and KRAS mutational status making an analysis of Veristrat Status and tumor mutational status impossible. Of the 28 patients in the gemcitabine arm, 12 patients received second-line erlotinib as part of the protocol therapy. In the erlotinib and gemcitabine and erlotinib combination arms, 14 and 13 patients received second-line therapy off of the protocol. An analysis of PFS and OS within the treatment arms was performed (Fig. 2 and Table 2). In the gemcitabine arm, patients classified as Good (n = 20) compared with Poor (n = 8) experienced similar PFS and OS. In contrast, in the erlotinib arm, patients classified as Good (n = 26) compared with Poor (n = 12) experienced significantly longer PFS (HR: 0.33; 95% CI: 0.16–0.70; log-rank p = 0.002; median PFS of 89 and 22 days, respectively) and OS (HR: 0.40; 95% CI: 0.19–0.85; log-rank p = 0.014; median OS of 255 and 51 days, respectively). In the gemcitabine and erlotinib arm, patients classified as Good (n = 17) compared with Poor (n = 15) experienced a statistically significantly longer PFS (HR: 0.42; 95% CI: 0.19–0.93; log-rank p = 0.027) and a trend toward an improvement in OS (HR: 0.48; 95% CI: 0.23–1.02; log-rank p = 0.051).
When outcomes between treatment the erlotinib alone and gemcitabine alone arms were compared, the HR for PFS between for the VeriStrat Good and Poor groups were 0.60 (95% CI: 0.31–1.15; log-rank p = 0.12) and 2.13 (95% CI: 0.83–5.52; log-rank p = 0.11), respectively (HR < 1 favors erlotinib). The corresponding results for OS for the VeriStrat Good and Poor groups between erlotinib and gemcitabine arms were HR = 0.66 (95% CI: 0.35–1.24; log-rank p = 0.19) and HR = 1.62 (95% CI: 0.64–4.07; log-rank p = 0.30), respectively. It should, however, be noted that in the gemcitabine monotherapy arm, nine of the patients classified as Good and three classified as Poor went on to receive erlotinib therapy as a second-line therapy. Outcomes on the erlotinib and gemcitabine combination arm lay numerically in between those of the two monotherapy arms for both the VeriStrat groups.
In the gemcitabine alone among patients with VeriStrat Good and Poor, the overall response rate (ORR) was 6% and 0%, respectively, and the disease control rate (DCR), defined as ORR and stable disease rate, was 60% and 62%, respectively. In the erlotinib arm, the ORR among patients with Veristrat Good and Poor was 5% and 0%, respectively, and the DCR was 62% and 25%, respectively. In the combination arm, the ORR among patients with VeriStrat Good and Poor was 15% and 10%, respectively, and the DCR was 65% and 47%, respectively. No evidence of association between VeriStrat classification and ORR or DCR was found in any of the treatment arms.
Analysis of the association of patient characteristics with VeriStrat classification revealed a statistically significant correlation between VeriStrat Good status and a PS of 0 or 1 as well as adenocarcinoma histology compared with other histologies (Table 3). To adjust for these and other possible confounding factors, multivariate analysis was performed for PFS and OS (Table 4). In the multivariate analysis for PFS, VeriStrat status (Good versus Poor) was associated with longer PFS (HR: 0.51; 95% CI: 0.30–0.86; p = 0.011). There was a statistical trend for shorter PFS associated with PS of 2 compared with PS 0 or 1 (HR: 1.69; 95% CI: 0.98–2.92; p = 0.058). In the multivariate analysis for OS, VeriStrat status (Good versus Poor) was associated with longer OS (HR: 0.53; 95% CI: 0.32–0.90; p = 0.017), and PS of 2 compared with 0 or 1 was associated with shorter OS (HR: 2.20; 95% CI: 1.27–3.83; p = 0.005). There was a borderline significant difference between stage IIIB and IV, with worse survival for patients with stage IV compared with stage IIIB disease (HR: 1.76; 95% CI: 0.93–3.32; p = 0.080).
When the clinical data from the gemcitabine and erlotinib arms were analyzed using adjusted multivariate analysis with covariate selection, the interaction between treatment and VeriStrat classification was shown to be significant (p < 0.001), together with the treatment arm (p = 0.028) and sex (p = 0.029) for PFS (Table 5). Similar analysis for OS showed that the interaction was again significant (p = 0.017), this time together with PS (p = 0.010) and disease stage (p = 0.047) (Table 5). The significance of the interaction between erlotinib or gemcitabine treatment and VeriStrat classification indicates that there is a differential benefit from these therapies between patients with Good and Poor VeriStrat classification, that is, the VeriStrat Good group benefits more from erlotinib, whereas Veristrat Poor group benefits more from gemcitabine.
We performed an exploratory analysis investigating if VeriStrat Good status and the longer PFS were associated with an improvement in HRQOL or lung cancer symptoms. In the gemcitabine alone arm, the PFS was similar in patients with VeriStrat Good and Poor status, and we performed an exploratory analysis to see if there was a difference in HRQOL or lung cancer symptoms related to VeriStrat status without the potential confounding factor of difference in PFS. These analyses are retrospective and exploratory and were not designed to test a specific hypothesis. There did not appear to be any significant differences in best response to treatment analysis as assessed by the FACT-L, TOI-L, or the LCS (Table 6). In the gemcitabine arm (n = 28) and erlotinib arm (n = 36), statistically significant differences were not observed between Veristrat Good and Poor groups and also differences were not observed on the TOI-L, FACT-L, and LCS at any of the four individual time points or in the longitudinal HRQOL trend analysis (Fig. 3).
The development of predictive biomarkers for currently available therapies is an area of intense investigation in thoracic oncology. EGFR mutations and anaplastic lymphoma kinase rearrangements are predictive of clinical benefit of EGFR TKI therapy and crizotinib, respectively. It is, however, estimated that only approximately 20% of patients with adenocarcinoma will have these molecular characteristics.20 Patients with EGFR mutant tumors derive tremendous benefit from EGFR TKI therapy, but there is a significant clinical need to further define who benefits from EGFR TKI therapy in patients with EGFR wild-type tumors. VeriStrat is a commercially available test in the United States; it has previously been demonstrated to be associated with longer PFS and OS in patients treated with EGFR TKIs and to be predictive of response to erlotinib. Our retrospective analysis confirmed that patients with a VeriStrat Good compared with Poor status have statistically significant longer PFS and OS when treated with erlotinib. The median PFS and OS observed among elderly patients with VeriStrat Good status suggests benefit for first-line EGFR TKI therapy in this patient population. This, however, is a retrospective analysis, and the number of patients in this cohort is small (n = 26). VeriStrat Good status was associated with good PS and adenocarcinoma histology, which are also good prognostic factors. Unfortunately, we do not have additional clinical data to determine if there is an association with VeriStrat status with other clinical factors such as weight loss, site or number of metastases, and comorbidities. The association with VeriStrat Good status and good PS has been observed in other studies and warrants further investigation.13,21
Of concern, the patients in the erlotinib arm with Veristrat Poor status had poor median PFS and OS, 22 and 51 days, respectively. Importantly, only 14 of patients enrolled in the erlotinib arm received second-line therapy, and VeriStrat Poor status was associated with worse PS and nonadenocarcinoma histology.2 Patients in this cohort experienced rapid disease progression on first-line erlotinib and may not have received second-line therapy related to poor PS, comorbidities, symptomatic decline, and patient and/or physician decision. The reasons for the poor efficacy are most likely multifactorial, but the data indicate erlotinib is not an acceptable first-line treatment option for this patient population. Patients with Veristrat Poor and Good status had similar PFS and OS with single agent gemcitabine, and a significant interaction of treatment and VeriStrat status was detected. These data suggest that the VeriStrat test is predictive with respect to treatment, not merely prognostic, and that the patients with a VeriStrat Poor status received greater clinical benefit from single agent gemcitabine compared with erlotinib than those with VeriStrat Good status. Patients in the gemcitabine and erlotinib arm with VeriStrat Good compared with VeriStrat Poor status experienced an improvement in PFS and a trend toward improvement in OS. The lack of the association between VeriStrat status and improved PFS and OS in the gemcitabine alone arm in this analysis and the absence of an association between VeriStrat status and chemotherapy observed in previous trials suggests that the erlotinib component of the therapy is responsible for this observation.
This analysis is exploratory and has several deficiencies. The retrospective nature of this analysis and the association of VeriStrat Good status with certain prognostic factors could have created imbalances of prognostic factors in the different subsets; however, multivariate analyses may have adequately adjusted for possible confounding factors. Imbalances in the rate and type of poststudy therapy may have impacted the OS results, and type of second-line therapy was only mandated in the single agent gemcitabine arm. The lack of mandatory tumor collection as part of the eligibility criteria makes assessment of any correlation between VeriStrat status and molecular characteristics such as EGFR or KRAS mutation status impossible. The majority of patients enrolled had a history of tobacco use (approximately 85% of patients), 40% of patients had a tumor with nonadenocarcinoma histology, and no responses were observed in the erlotinib alone arm. These data are suggestive that the rate of EGFR mutant NSCLC was low in this trial. There was no significant difference in the HRQOL outcomes in the initial trial, and we wanted to test the hypothesis that the significant efficacy difference between the Veristrat Good and Poor subgroups in the erlotinib arm may have resulted in differences in HRQOL. The sample size for this analysis was small, and the short PFS observed in the VeriStrat Poor group treated with erlotinib limited the number of HRQOL assessments available for each patient. The utility of these analyses is limited and the purpose of including the results was to provide preliminary data for future studies.
This is the first report of the comparison of VeriStrat status in the elderly patients treated with erlotinib or with single-agent chemotherapy that demonstrates how a serum or plasma test, with no need of biopsy, may be particularly useful. Patients with VeriStrat Poor status should not receive erlotinib therapy but may be good candidates for gemcitabine treatment. Since this trial was developed in 2005, a phase III trial compared carboplatin and weekly paclitaxel to single-agent chemotherapy (gemcitabine or vinorelbine); a statistically significant improvement in PFS and OS was observed with the combination treatment.22 This trial established double-agent platinum-based therapy as a standard therapy for appropriate patients. Many elderly patients are, however, “frail” and may not be candidates for platinum-based therapy, and the optimal management for this patient population remains unclear. A prospective phase II trial comparing erlotinib to single-agent chemotherapy in elderly patients as first-line therapy who are not candidates for double-agent platinum-based therapy with VeriStrat Good status may be worth pursuing to further investigate the use of this minimally invasive test to select patients for therapy. Since this trial would be exploratory, a potential primary end point would be PFS with the goal of demonstrating an improvement in PFS of 1.5 times or greater than the control arm before pursing larger validation trials. An ongoing randomized trial in the second-line setting is prospectively stratifying patients based on VeriStrat status and is comparing erlotinib to docetaxel or pemetrexed.23 Until the results of the prospective clinical trial are available, clinicians should not use VeriStrat status for selection erlotinib or chemotherapy in routine clinical practice.
1. Hurria A, Togawa K, Mohile SG, et al. Predicting chemotherapy toxicity in older adults with cancer: a prospective multicenter study. J Clin Oncol. 2011;29:3457–3465
2. Stinchcombe TE, Peterman AH, Lee CB, et al. A randomized phase II trial of first-line treatment with gemcitabine, erlotinib, or gemcitabine and erlotinib in elderly patients (age ≥70 years) with stage IIIB/IV non-small cell lung cancer. J Thorac Oncol. 2011;6:1569–1577
3. Gridelli C, Gallo C, Shepherd FA, et al. Gemcitabine plus vinorelbine compared with cisplatin plus vinorelbine or cisplatin plus gemcitabine for advanced non-small-cell lung cancer: a phase III trial of the Italian GEMVIN Investigators and the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol. 2003;21:3025–3034
4. . Effects of vinorelbine on quality of life and survival of elderly patients with advanced non-small-cell lung cancer. The Elderly Lung Cancer Vinorelbine Italian Study Group. J Natl Cancer Inst. 1999;91:66–72
5. Frasci G, Lorusso V, Panza N, et al. Gemcitabine plus vinorelbine versus vinorelbine alone in elderly patients with advanced non-small-cell lung cancer. J Clin Oncol. 2000;18:2529–2536
6. Shepherd FA, Rodrigues Pereira J, Ciuleanu T, et al.National Cancer Institute of Canada Clinical Trials Group. Erlotinib in previously treated non-small-cell lung cancer. N Engl J Med. 2005;353:123–132
7. Jackman DM, Yeap BY, Lindeman NI, et al. Phase II clinical trial of chemotherapy-naive patients > or = 70 years of age treated with erlotinib for advanced non-small-cell lung cancer. J Clin Oncol. 2007;25:760–766
8. Moore MJ, Goldstein D, Hamm J, et al.National Cancer Institute of Canada Clinical Trials Group. Erlotinib plus gemcitabine compared with gemcitabine alone in patients with advanced pancreatic cancer: a phase III trial of the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol. 2007;25:1960–1966
9. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med. 2004;350:2129–2139
10. Paez JG, Jänne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science. 2004;304:1497–1500
11. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med. 2009;361:947–957
12. Azzoli CG, Baker S Jr, Temin S, et al.American Society of Clinical Oncology. American Society of Clinical Oncology Clinical Practice Guideline update on chemotherapy for stage IV non-small-cell lung cancer. J Clin Oncol. 2009;27:6251–6266
13. Taguchi F, Solomon B, Gregorc V, et al. Mass spectrometry to classify non-small-cell lung cancer patients for clinical outcome after treatment with epidermal growth factor receptor tyrosine kinase inhibitors: a multicohort cross-institutional study. J Natl Cancer Inst. 2007;99:838–846
14. Carbone D, Seymour L, Ding K, et al. Serum proteomic prediction of outcomes in advanced NSCLC patients treated with erlotinib/placebo in the NCIC clinical trials group BR.21 trial. J Thorac Oncol. 2010;5:S80
15. Amann JM, Lee JW, Roder H, et al. Genetic and proteomic features associated with survival after treatment with erlotinib in first-line therapy of non-small cell lung cancer in Eastern Cooperative Oncology Group 3503. J Thorac Oncol. 2010;5:169–178
16. Chung CH, Seeley EH, Roder H, et al. Detection of tumor epidermal growth factor receptor pathway dependence by serum mass spectrometry in cancer patients. Cancer Epidemiol Biomarkers Prev. 2010;19:358–365
17. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst. 2000;92:205–216
18. Cella DF, Tulsky DS, Gray G, et al. The Functional Assessment of Cancer Therapy scale: development and validation of the general measure. J Clin Oncol. 1993;11:570–579
19. Cella DF, Bonomi AE, Lloyd SR, Tulsky DS, Kaplan E, Bonomi P. Reliability and validity of the functional assessment of cancer therapy-lung (FACT-L) quality of life instrument. Lung Cancer. 1995;12:199–220
20. Kris MG, Johnson BE, Kwiatkowski DJ, et al. Identification of driver mutations in tumor specimens from 1,000 patients with lung adenocarcinoma: The NCI’s Lung Cancer Mutation Consortium (LCMC). J Clin Oncol. 2011;29:abstr CRA7506
21. Carbone DP, Ding K, Roder H, et al. Prognostic and predictive role of the VeriStrat plasma test in patients with advanced non-small-cell lung cancer treated with erlotinib or placebo in the NCIC Clinical Trials Group BR.21 trial. J Thorac Oncol. 2012;7:1653–1660
22. Quoix E, Zalcman G, Oster JP, et al. Carboplatin and weekly paclitaxel doublet chemotherapy compared with monotherapy in elderly patients with advanced non-small-cell lung cancer: IFCT-0501 randomised, phase 3 trial. Lancet. 2011;378:1079–1088
23. Sorlini C, Barni S, Petrelli F, et al. PROSE: randomized proteomic stratified phase III study of second line erlotinib versus chemotherapy in patients with inoperable non–small cell lung cancer (NSCLC). J Clin Oncol. 2011;29:TPS214
Proteomics; Biomarkers; Epidermal growth factor receptor tyrosine kinase inhibitors; Elderly
© 2013International Association for the Study of Lung Cancer
Highlight selected keywords in the article text.