Published ahead of print on June 26, 2008, doi:10.1164/rccm.200803-389OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200803-389OC
First- or Second-line Therapy with Gefitinib Produces Equal Survival in Non–Small Cell Lung Cancer1 Department of Internal Medicine and 2 Department of Oncology, National Taiwan University Hospital, Taipei, Taiwan; 3 Division of Biostatistics, Graduate Institute of Epidemiology and Center of Biostatistics Consultation, College of Public Health, National Taiwan University, Taipei, Taiwan; 4 Division of Critical Care Medicine, Department of Emergency and Critical Care Medicine, Lo-Tung Poh-Ai Hospital, Yi-Lan, Taiwan; and 5 Department of Radiology and 6 Department of Laboratory Medicine, National Taiwan University Hospital, Taipei, Taiwan Correspondence and requests for reprints should be addressed to Jin-Yuan Shih, M.D., Ph.D., Department of Internal Medicine, National Taiwan University Hospital, No. 7 Chung-Shan South Road, Taipei 10002, Taiwan. E-mail: jyshih{at}ntu.edu.tw
Rationale: Gefitinib is effective in treating patients with non–small cell lung cancer (NSCLC) with epidermal growth factor receptor (EGFR) mutations. Deletions in exon 19 and L858R in exon 21 are the best-documented EGFR mutations that are associated with effective gefitinib responsiveness. Objectives: To clarify the influence of gefitinib timing, we conducted a study to compare the outcomes of different lines of gefitinib treatment in patients with exon 19 deletions or L858R. Methods: We surveyed the clinical data and mutational studies of patients with NSCLC with EGFR mutations in the National Taiwan University Hospital (Taipei, Taiwan). Measurements and Main Results: Three hundred and twenty-eight patients, who received gefitinib for stage IIIb or IV NSCLC, were adequately sequenced for EGFR mutations; 192 patients had mutant EGFR, including 77 patients with exon 19 deletions and 75 patients with L858R. The 152 patients with exon 19 deletions or L858R and who were receiving gefitinib were classified into a chemonaive group (91 patients) or a chemotherapy-treated group (61 patients). Chemonaive status before gefitinib and female sex were associated with clinical response to gefitinib (P = 0.006 and 0.053, respectively). Neither overall survival after the start of antitumor therapy nor progression-free survival after gefitinib therapy was significantly different between these two groups (P = 0.207 and 0.804, respectively). Clinical response to gefitinib was the only factor associated with better overall survival (P = 0.001). Conclusions: This study suggests that gefitinib is effective in patients with EGFR mutations. The gefitinib response rate in chemonaive patients is higher than in chemotherapy-treated patients; however, there is no difference in overall survival.
Key Words: epidermal growth factor receptor • gefitinib • lung cancer • mutations
Non–small cell lung cancer (NSCLC) is the most common cause of cancer-related death in the world, and more than 1 million patients die of lung cancer annually (1). Prognosis is poor for late-stage NSCLC (stages IIIb and IV). Several regimens of chemotherapy have been documented to improve survival, such as first-line cisplatin-based regimens (2) and second-line docetaxel or pemetrexed (3, 4). Chemotherapy after two or more prior regimens has not yet shown obvious benefit. Epidermal growth factor receptor (EGFR), a transmembrane glycoprotein, is involved in the proliferation, angiogenesis, and resistance to apoptosis of cells. Loss of control of EGFR because of deregulation, amplification, or mutations may result in malignant changes of cells (5, 6). Gefitinib (Iressa, ZD1839; AstraZeneca, Wilmington, DE), an inhibitor of EGFR tyrosine kinase, has a novel antitumor effect on NSCLC. It has been documented to be useful in patients previously treated with chemotherapy regimens (7), or in selected groups of patients as first-line therapy, with effects of tumor regression and symptom amelioration (8). Effective response to gefitinib is associated with EGFR mutations, as reported in various studies (9–19). A series of EGFR kinase domain mutations have been discovered in NSCLC, mainly adenocarcinoma (20–22). These mutations exist in exons 18 to 21, which encompass most of the tyrosine kinase–binding domain of EGFR (23). They include substitutions for G719 in the nucleotide-binding loop of exon 18, in-frame deletions within exon 19, in-frame duplication and/or insertion in exon 20, and substitutions for L858 or L861 in the activation loop in exon 21. As reported in the literature, the two major EGFR mutations included deletions in exon 19 and L858R in exon 21 (24). These two mutations were the best-documented mutations associated with an effective patient response to gefitinib (25). The clinical courses and survival of patients with lung cancer, who have EGFR mutations as deletions in exon 19 or L858R, can be changed with gefitinib treatment. Mutations in exon 18 and exon 20 are much rarer, and more studies are required to document their responsiveness to gefitinib treatment. Whether gefitinib offers more benefit when it is given to chemonaive or chemotherapy-treated patients remains controversial. In the study conducted by Chang and coworkers, the lack of previous chemotherapy was an independent predictive factor for gefitinib response (26). Nevertheless, pooled analysis performed by Costa and coworkers revealed that mixed gefitinib treatment of chemonaive and chemotherapy-treated patients was not inferior to gefitinib treatment of only chemonaive patients in both response rate and progression-free survival (27). Little information about overall survival was presented in these studies. Moreover, confounding factors, such as sex, smoking status, or types of histology, might exist between chemonaive and chemotherapy-treated groups (28). Among these confounding factors for analysis, the status of EGFR mutations might be the most substantial. To clarify the difference between gefitinib treatment in chemonaive and in chemotherapy-treated patients, we performed a retrospective analysis of the prognosis of patients with NSCLC who underwent gefitinib therapy. We further focused on patients with mutant EGFR consisting of deletions in exon 19 and L858R in exon 21, the two EGFR mutations with promising gefitinib responsiveness, to eliminate the influence of other mutations and wild-type EGFR in our analysis.
Patients The study group included patients with lung cancer diagnosed at the National Taiwan University Hospital (Taipei, Taiwan) between January 2004 and June 2007. All patients, with stage IIIb or IV NSCLC, who received gefitinib before chemotherapy (chemonaive) or after chemotherapy (chemotherapy treated) were identified from the records of the Department of Pharmacy in the hospital. The timing and order of the various chemotherapy regimens, which depended on the physicians' discretion, were determined after thorough discussion with the patients. Clinical data of patients, including demographic information, cancer cell type, smoking status, and image studies, were reviewed. Nonsmokers were defined as those who had smoked fewer than 100 cigarettes in their lifetime. Date of diagnosis, all chemotherapy received, and responsiveness to chemotherapy were recorded. Lung cancer histology was defined according to the World Health Organization (Geneva, Switzerland) pathology classification (29). Tumor specimens obtained by either surgical or needle biopsy/aspiration procedures, including primary lung tumors, malignant effusion cell blocks, and other distant metastases, were sequenced for mutational analysis. Written informed consent for the use of tissues in molecular analysis was acquired from patients at the time of procurement of tumor specimens. This study was approved by the Institutional Review Board of the National Taiwan University Hospital.
Antitumor Chemotherapy and Evaluation of Effectiveness
Mutational Analysis of EGFR Lung cancer tissues were obtained at surgery, immediately snap frozen in liquid nitrogen, and stored until use. Total mRNA was extracted from resected cancer tissue with an RNA extraction kit (RNeasy Mini Kit; Qiagen). The four exons (exons 18–21) that encode the tyrosine kinase domain of the EGFR gene were amplified with primers and PCR conditions as reported (35). Reverse transcription (RT)-PCR amplicons were purified and sequenced. All of the sequencing reactions were performed in both forward and reverse directions, using tracings from at least two PCRs.
Statistical Analyses
Patients with NSCLC with EGFR Mutations and Gefitinib Treatment A total of 698 patients with NSCLC had received gefitinib treatment between January 2004 and June 2007. Specimens from patients, who had measurable disease and received gefitinib therapy for stage IIIb or IV NSCLC, were examined for mutations of the EGFR tyrosine kinase domain. Patients with specimens that were inadequate for sequencing were excluded, and finally 328 patients were adequately sequenced and included in our study. The tissues processed for mutation examination included 68 surgical specimens, 117 needle biopsies (either echo guided, computed tomography guided, or bronchoscopic), and 143 cell block preparations of pleural effusion. Clinical characteristics of the 328 patients are listed in Table 1. Of these 328 patients, 192 had EGFR mutations. The mutations were more frequent in nonsmokers than in smokers (64.6 vs. 37.8%; P < 0.001), in adenocarcinomas than in nonadenocarcinomas (62.5 vs. 14.8%; P < 0.001), and in females than in males (64.2 vs. 47.8%; P = 0.004).
The composition of mutations in the 192 patients was as follows: 77 patients had a single mutation consisting of exon 19 deletions; 75 patients had a single mutation, L858R; 8 patients had a single mutation in exon 20; and the remaining 32 patients had other mutations (either single or complex mutations) (see Table E1 in the online supplement). Demographic data and response to gefitinib are listed in Table 2. The total gefitinib response rate of these mutant-EGFR patients was 62.5%. The response rate (12.5%) was notably lower in those with single exon 20 mutations. The duration of progression-free survival after the start of gefitinib was obviously shorter for those with single exon 20 mutations (Figure 1A; P = 0.003 by log-rank test), whereas overall survival rates after the start of systemic antitumor therapy were not statistically different (Figure 1B; P = 0.608 by log-rank test). Furthermore, it was noteworthy that no significant difference in overall survival existed between patients with L858R and those with exon 19 deletions (Figure 1B).
Patients with Single Mutation of L858R or Deletions in Exon 19 To avoid the influence of other mutations and wild-type EGFR on our analysis, patients with L858R or deletions in exon 19 were identified for further analysis. A total of 152 patients with the single mutation of L858R or deletions in exon 19 were included for further analysis. For outcome assessment of gefitinib therapy performed with different timing, the patients were divided into two groups: chemonaive before gefitinib treatment (n = 91) and chemotherapy-treated before gefitinib treatment (n = 61). Patient characteristics are summarized in Table 3. In this population with single EGFR mutations of L858R or deletions in exon 19, the response rate was 67.1% and the disease control rate was 88.2%. The median overall survival was 16.1 months.
The clinicopathological data, including age, sex, smoking status, cell type, and clinical staging, were not significantly different between the two groups (Table 3). The response rate and disease control rate with gefitinib were higher for the chemonaive group than for the chemotherapy-treated group (response rate: 75.8 vs. 54.1%, P = 0.005; disease control rate: 93.4 vs. 80.3%, P = 0.014). There was no difference in progression-free survival after gefitinib treatment between the chemonaive group and the chemotherapy-treated group (median progression-free survival, 8.2 vs. 7.0 mo; P = 0.804; Figure 2A). Median overall survival times after the start of first-line antitumor therapy, either gefitinib or other regimens, were 16.9 months for the chemonaive group and 15.0 months for the chemotherapy-treated group (P = 0.207; Figure 2B).
Among the patients with L858R or deletions in exon 19, male nonsmokers (n = 23) and female nonsmokers (n = 107) had no significant difference in gefitinib response rates (60.9 vs. 70.0%; P = 0.333) and outcomes (median progression-free survival: 6.0 vs. 7.9 mo, P = 0.840; overall survival: 14.9 vs. 16.8 mo, P = 0.460). Moreover, male patients with adenocarcinoma (n = 41) and female patients with adenocarcinoma (n = 108) also had no significant difference in gefitinib response rates (56.1 vs. 71.3%; P = 0.079) and outcomes (median progression-free survival: 6.5 vs. 7.7 mo, P = 0.858; overall survival: 14.9 vs. 17.0 mo, P = 0.230).
Predictive Factors for Response to Gefitinib among Patients with L858R or Deletions in Exon 19
Predictive Factors for Overall Survival Clinical factors were assessed for predictive value of overall survival after first-line antitumor therapy in the NSCLC group with EGFR mutations consisting of L858R or deletions in exon 19 and gefitinib therapy (Table 5). We found that response to gefitinib was the only factor associated with better overall survival outcome (P = 0.001 by log-rank test). Other factors did not show significant correlation with survival.
In this study, we recruited only patients with NSCLC with EGFR mutations consisting of L858R or deletions in exon 19 for analysis, to minimize the influence of EFGR mutation status, probably the most important confounding factor for gefitinib response. Among these patients, the rate of response to gefitinib treatment was 67.1% and the median overall survival time was 16.1 months. We also found that chemonaive status before gefitinib treatment and female sex were potential predictive factors for gefitinib response among such patients. Gefitinib was effective both in chemonaive or chemotherapy-treated patients, leading to similar overall survival. The response rate to gefitinib treatment among Western patients with NSCLC is about 10% (7, 36), much lower than the response rate of 20 to 30% among East Asian patients (26). This discrepancy may be due to EGFR mutations. With mutant EGFR, the gefitinib response rate among East Asian patients is about 60 to 80%, but falls to 0 to 30% among East Asian patients without mutant EGFR (13, 32, 37). Similarly, in the large-scale study (1,692 patients) conducted by Thatcher and coworkers, a beneficial effect of gefitinib was displayed in never-smokers and Asian patients (38). This might also be due to the existence of EGFR mutations. EGFR mutations would be the most important confounding factor in analyzing the effects of gefitinib treatment. This is the reason why this current study focused only on patients with EGFR mutations consisting of L858R or deletions in exon 19. In prior studies, factors predictive of gefitinib response were female sex, nonsmoking status, and adenocarcinoma histology (28, 39). Before the emerging understanding of EGFR mutations, these factors were important references for physicians in choosing patients responsive to gefitinib treatment. However, these factors might be effective via the medium of EGFR mutations. In the large-scale study conducted by Shigematsu and coworkers, 130 of a total of 617 patients with lung cancer had EGFR mutations, distributed between exon 18 and exon 21 (40). These mutations were also associated with these clinical features: nonsmoking status, female sex, and adenocarcinoma histology. In the present study, we included only patients with the L858R mutation of deletions in exon 19. Nonetheless, we found that female sex was still associated with gefitinib response among these patients. Several studies have documented the influence of gefitinib timing (26, 28, 39). In the phase II study conducted by Fukuoka and coworkers, 216 patients, who had undergone at least one or two chemotherapy regimens before receiving gefitinib, were evaluated (39). The response rate did not change with the use of one or two previous regimens, but the effect of overall survival was not mentioned in the report. Chang and coworkers analyzed a group of 428 patients with lung cancer who received gefitinib, without regard of EGFR mutations (26). They found that being free of previous chemotherapy before gefitinib treatment was an independent predictive factor for gefitinib response. Long-term survival was associated with responsiveness to gefitinib. However, the overall survival rates did not differ significantly between chemonaive and chemotherapy-treated patients. Overall survival was even shorter for the chemonaive group in that study (6.3 vs. 7.9 mo). In the study by Yang and coworkers, composed of 196 chemonaive patients who received gefitinib for advanced NSCLC, the response rate of 42% was considered higher than the response rates of chemotherapy-treated patients reported in the literature (28). In all these previous studies, the populations were composed of patients with and without mutant EGFR. The outcome analyses of these prior studies might be influenced to some extent by the status of EGFR mutations. To our knowledge, the present study is the first to assess the overall survival of different gefitinib timing while minimizing the effects of EGFR mutation status. In this study, although chemonaive patients had a higher gefitinib response rate than did chemotherapy-treated patients, overall survival rates were similar between the two groups. Although this is not a prospective study, our observation strongly suggests that there is no difference in overall survival among patients with EGFR mutations when gefitinib is given to either chemonaive or chemotherapy-treated patients. When a crossover effect cannot be avoided, progression-free survival, quality of life, and pharmacoeconomics may be other end points in clinical trial designs testing the therapeutic effect of EGFR tyrosine kinase inhibitors in NSCLC. Overall survival alone may not be an appropriate primary end point. Some limitations might be of concern because of the retrospective design of the study with a selective group of patients. First, the timing of gefitinib treatment depended on the clinical decisions of physicians caring for the patients. The doctors might have chosen patients with potential response to gefitinib, especially the subgroup of female patients, patients with adenocarcinomas, and nonsmokers. Bias might exist in the decisions made. However, the demographic data of chemonaive and chemotherapy-treated patients were not significantly different in our analysis. Second, the cell types of our study group were mostly adenocarcinoma. The timing of gefitinib therapy in patients with NSCLC with other cell types and EGFR mutations should be discussed in further studies. Another limitation is the relatively small study population of our study. More data are needed for clinical application. In conclusion, EGFR mutations may influence the response to gefitinib therapy. In the present study, consisting only of patients with L858R or deletions in exon 19 of EGFR, gefitinib timing and female sex were predictive factors for clinical response of gefitinib. However, whether gefitinib was used in chemonaive patients or chemotherapy-treated patients did not influence overall survival. Besides, given no significant difference in progression-free survival after gefitinib, whether gefitinib was given as the first-line treatment or after chemotherapy does not alter the likelihood of gefitinib usage. Further large-scale, prospective studies are needed to clarify these findings.
The authors thank Dr. Gene Alzona Nisperos for editing this manuscript. The authors also thank the Department of Medical Research of the National Taiwan University Hospital.
Supported by grants from the National Science Council (Taiwan), NSC-95-2314-B-002-110-MY3 and NSC-95-2314-B-002-113-MY3, to J.-Y.S. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200803-389OC on June 26, 2008 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form March 8, 2008; accepted in final form June 25, 2008
Related articles in AJRCCM:
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
