Epidermal growth factor receptor-tyrosine kinase inhibitor therapy is effective as first-line treatment of advanced non-small-cell lung cancer with mutated EGFR: A meta-analysis from six phase III randomized controlled trials
Department of Oncology, Shanghai Pulmonary Hospital, Tongji University School of Medicine, Tongji University Cancer Institute, Shanghai 200433, People's Republic of China
Gefiinib and erlotinib are two similar small molecules of selective and reversible epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs), which have been approved for second-line or third-line indication in previously treated advanced Non-small-cell lung cancer (NSCLC) patients. The results of comparing the EGFR-TKI with standard platinum-based doublet chemotherapy as the first-line treatment in advanced NSCLC patients with activated EGFR mutation were still controversial. A meta-analysis was performed to derive a more precise estimation of these regimens. Finally, six eligible trials involved 1,021 patients were identified. The patients receiving EGFR-TKI as front-line therapy had a significantly longer progression-free survival (PFS) than patients treated with chemotherapy [median PFS was 9.5 versus 5.9 months; hazard ratio (HR) = 0.37; 95% confidence intervals (CI) = 0.27–0.52; p < 0.001]. The overall response rate (ORR) of EGFR-TKI was 66.60%, whereas the ORR of chemotherapy regimen was 30.62%, which was also a statistically significant favor for EGFR-TKI [relative risk (RR) = 5.68; 95% CI = 3.17–10.18; p < 0.001]. The overall survival (OS) was numerically longer in the patients received EGFR-TKI than patients treated by chemotherapy, although the difference did not reach a statistical significance (median OS was 30.5 vs. 23.6 months; HR = 0.94; 95% CI = 0.77–1.15; p = 0.57). Comparing with first-line chemotherapy, treatment of EGFR-TKI achieved a statistical significantly longer PFS, higher ORR and numerically longer OS in the advanced NSCLC patients harboring activated EGFR mutations, thus, it should be the first choice in the previously untreated NSCLC patients with activated EGFR mutation.
Lung cancer is the leading cause of cancer-related mortality worldwide, with nearly 1.4 million deaths each year. Of the 1.6 million new cases of lung cancer diagnosed each year, approximately 222,520 individuals (116,750 men and 105,770 women) will be diagnosed with lung cancer and 157,300 individuals (86,220 men and 71,080 women) will die from lung cancer in the USA by the end of 2010.1 Lung cancer is diagnosed at an advanced stage in a majority of patients, which is the primary reason behind the high mortality rate associated with this disease. Early detection continues to be an elusive goal, and substantial numbers of patients diagnosed as localized disease are often unsuitable for curative resection procedures due to concomitant medical illness.2 Non-small-cell lung cancer (NSCLC) accounts for nearly 85% of all cases of lung cancer. The prognosis of the patients with NSCLC remains poor. For these patients with metastatic or recurred disease, the main treatment is platinum-based doublet chemotherapy, however, the result is still dismal.3–5
Epidermal growth factor receptor (EGFR)-dependent pathway plays an important role in the development and progression of human epithelial cancers, including NSCLC. Activation of the EGFR pathway could promote tumor growth and progression, stimulating cancer cell proliferation, invasion and metastasis and inhibiting apoptosis.6 EGFR mutation has been confirmed as predictors of efficacy for EGFR-tyrosine kinase inhibitors (EGFR-TKIs). Gefitinib and erlotinib are two similarly small, orally active, selective and reversible EGFR-TKIs molecules, which have been extensively used in NSCLC. Treatment with EGFR-TKIs is more effective in female, never smoker, Asian and patients with pulmonary adenocarcinomas. In these sub-populations, EGFR-TKIs are associated with favorable objective response rate, progression-free survival (PFS) and overall survival (OS).7–10 Recently, great efforts in clinical and laboratory research have led to a better understanding of predictive factors to the efficacy of these drugs. Remarkably, increasing evidence has been accumulated to strongly support predictive role of EGFR gene mutation in NSCLC patients treated with EGFR-TKIs. As a result, EGFR inhibition strategy, which was originally limited to patients failed from previously standard treatment, has been transferred as first-line strategy in NSCLC patients potentially benefiting from the treatment of EGFR-TKIs. Approximately, 10–15% of NSCLC patients in Europe11 and 30–35% of NSCLC patients in Asia harbor EGFR mutation-positive NSCLC.12 Patients of female, never-smoker and adenocarcinoma have a relatively high incidence of somatic EGFR mutations.13, 14
Several randomized trials have been conducted, with the aim of comparing the EGFR-TKI (gefitinib or erlotinib) with the standard doublet chemotherapy (platinums plus third-generation new agents) as first-line treatment in advanced NSCLC patients with activating EGFR mutation.15–22 However, the small sample size of each single trial might not have sufficient power to detect potentially relevant differences in efficacy of the different arms. Two meta-analysis including four and five randomized controlled trials (RCT),23, 24 comparing EGFR-TKI with chemotherapy in the first-line treatment of advanced NSCLC with mutated EGFR, have revealed EGFR-TKI significantly prolonged PFS and increased overall response rate (ORR) over chemotherapy, while not increasing hematological toxicity.
We performed this meta-analysis, including six RCT, to compare the EGFR-TKI with the platinum-based doublet chemotherapy as first-line treatment in advanced NSCLC with activating EGFR mutation. Three of these studies have been published, while the other three studies have been reported in abstract forms. Here, we performed a meta-analysis of the most updated results of these studies to better quantify the clinical benefits of EGFR-TKI over chemotherapy.
Material and Methods
To ensure retrieval of all relevant trials, we used a broad search strategy in which key words and text words related to lung cancer and EGFR-TKI were combined with a validated methodological filter, as described by Dickersin et al.25 The following keywords were used: “gefitinib,” “erlotinib,” “EGFR mutation” and “non-small-cell lung cancer.” We used this strategy to search a variety of electronic databases, including the PubMed database (1966 to June 10, 2011), the Cochrane Controlled Trials Register (issue 3, 2011), EMBASE (1974 to June 10, 2011) and Chinese Biomedical database (1978 to June 10, 2011). We also manually searched the conference proceedings of the American Society of Clinical Oncology (ASCO), the European Society of Medical Oncology and the International Association for the Study of Lung Cancer from the year of 2000 to 2011 for relevant clinical trials. Reference lists from studies selected for this review, and from other published systematic reviews and practice guidelines were also hand-searched.
Only RCT were included in this analysis. Other inclusion criteria included enrolled patients >18 years, pathologically proven NSCLC with EGFR mutation-positive, clinical IIIB–IV stage, previously untreated and that compared the EGFR-TKI with the platinum-based doublet chemotherapy. Trials were included regardless of publication status, date of publication and language. We excluded ongoing studies, nonrandomized studies and studies with not more than 10 patients per study arm.
Study selection and data extraction
Two reviewers (Guanghui Gao and Shengxiang Ren) independently screened the titles and abstracts of all studies identified in the literature search to verify compliance with the inclusion. When this information was unsatisfactory, full-text was retrieved and inclusion criteria were applied. Disagreements between the two reviewers were resolved by consensus involving a third reviewer (Caicun Zhou). The same reviewers who screened the studies independently performed data extraction. The following data were collected from each study: first author's surname, source of trial, treatment, primary endpoint, total numbers of cases and controls, percentage of female, percentage of adenocarcinoma, percentage of never smokers, EGFR mutation type, ORR, PFS and OS, respectively. All included studies, regardless of whether they were published or not, were assessed for internal validity parameters, with particular emphasis on randomization, masking of patients and clinicians, concealment of allocation, documentation of dropouts and withdrawals and intent-to-treat (ITT) analysis. Follow-up was arbitrarily defined as complete if trial enrollment was closed, all patients were accounted and <10% were lost in follow-up.
The primary outcome was PFS (as defined by Therasse et al.26). Secondary outcomes included OS and ORR.
Data analysis and statistical methods
A random-effect model was assumed in all meta-analysis. The hazard ratio (HR) was used for PFS and OS. For binary data, the relative risk (RR) was used for ORR. A statistical test with a value of p < 0.05 was considered significant. An HR > 1 reflects more deaths or progression in the EGFR-TKI regimen. An RR > 1 reflects a favorable outcome in the EGFR-TKI arm for response. A funnel plot was generated or a linear regression test27 was performed to examine whether there was publication bias. Subgroup and sensitivity analysis were done by different type of EGFR-TKI (gefitinib or erlotinib).
Analysis was performed using the STATA SE 11.0 package (StataCorp, College Station, TX). Statistical tests for heterogeneity were one-sided; statistical tests for effect estimates and publication bias were two sided.
The flow chart of our study is shown in Figure 1. The literature search identified 2,431 trials, of which 53 were considered potentially relevant. The remaining 53 articles were selected for analysis and were evaluated in more detail. Of these, 47 were excluded for the following reasons: 21 trials were non-randomized comparison; 14 trials were RCTs, but were used in second-line or third-line treatment; seven trials were RCTs, but the treatment arms were EGFR-TKIs plus chemotherapies, not single EGFR-TKIs in treatment arm versus chemotherapies in control arm; five trials were RCTs, but which were used in sequential or maintained therapy. The remaining six RCT (Table 1), which involved 1,021 adult patients, met all the inclusion and exclusion criteria and were included in the meta-analysis. Among which, three studies have been published, while the other three studies have been reported in abstract forms. They were all Phase III RCTs and there was no placebo-controlled double-blinded trial.
Table 1. Baseline characteristics of the 6 trials comparing Epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKI) with Chemotherapy for patients with previously untreated NSCLC with mutated EGFR
The chemotherapy regimens used included: paclitaxel and carboplatin; gemcitabine and cisplatin; gemcitabine and carboplatin; docetaxel and cisplatin; docetaxel and carboplatin. All trials compared the EGFR-TKI schedule with the standard platinum-based doublet chemotherapy schedule (Table 1).
The PFS of EGFR-TKI arm ranged from 8.4 months to 13.1 months and the PFS of chemotherapy arm ranged from 4.6 months to 6.7 months. The median PFS was 9.5 months versus 5.9 months between the two groups. The pooled HR for PFS showed that there was a statistically significant improvement when EGFR-TKI regimen patients were compared with chemotherapy regimen patients [HR = 0.37; 95% confidence interval (CI) = 0.27–0.52; p < 0.001; Fig. 2].
Subgroup and sensitivity analysis of gefitinib or erlotinib versus chemotherapy also revealed that there was a significant difference in the two regimens (HR = 0.43; 95% CI = 0.32–0.58; p < 0.001 and HR = 0.26; 95% CI = 0.10–0.67; p = 0.005, respectively; Fig. 2).
Overall survival (five trials, 852 patients)
Five trials including 852 patients reported the data of OS. The OS of EGFR-TKI arm ranged from 18.9 months to 30.9 months and the OS of chemotherapy arm ranged from 14.4 months to >30.9 months. The median OS was 30.5 months in the patients receiving EGFR-TKI versus 23.6 months in the patients treated with chemotherapy. The pooled HR for OS showed that there was no significant difference in the two groups (HR = 0.94; 95% CI = 0.77–1.15; p = 0.57; Fig. 3).
Subgroup and sensitivity analysis of gefitinib versus chemotherapy also revealed that there was no significant difference in the two groups (HR = 0.97; 95% CI = 0.78–1.20; p = 0.78; Fig. 3).
The ITT analysis demonstrated that the ORR of the EGFR-TKI group was 66.60% (345/518) significantly higher than the 30.62% (154/503) in chemotherapy group (RR = 5.68; 95% CI = 3.17–10.18; p < 0.001; Fig. 4).
Furthermore, subgroup and sensitivity analysis of gefitinib or erlotinib versus chemotherapy also revealed that there was a significant difference in the two regimens (RR = 3.82; 95% CI = 2.28–6.39; p < 0.001 and RR = 11.99; 95% CI = 6.80–21.15; p < 0.001, respectively; Fig. 4).
An extensive search strategy was made to minimize the potential for publication bias. Publication bias was not found according to both Begg's test (0.452 for EGFR-TKI vs. platinum-based doublet chemotherapy) and Egger's test (0.864 for EGFR-TKI vs. platinum-based doublet chemotherapy).
We could concluded the major results from this meta-analysis comparing EGFR-TKI with platinum-based doublet chemotherapy in a total of 1,021 patients with advanced NSCLC patients harboring EGFR mutation: First, the treatment of EGFR-TKI significantly improved PFS than chemotherapy. Second, patients receiving EGFR-TKI therapy had statistically significantly higher ORR than patients treated with chemotherapy. Third, subgroup analysis revealed statistically significant and robust data for prolonging PFS and improving ORR in patients treated with either gefitinib or erlotinib versus chemotherapy.
Our results showed that there was no statistically significant difference between the two groups in OS. There might be several possible explanations: only five randomized controlled studies reported the OS and three of them had been reported in abstract forms. The non-sufficient data about OS and the relatively shorter time of follow-up might influence statistical result. More importantly, differences in OS are potentially affected by the subsequent drugs cross-over, and a relevant number of patients assigned to chemotherapy arm finally received an EGFR-TKIs (gefitinib or erlotinib) as second or third-line treatment after disease progression. The high proportion of cross-over may decrease the benefit associated with the administration of gefitinib.28 For example, 39% of patients allocated into chemotherapy arm in IPASS study and 95% of patients with chemotherapy group in Maemondo et al.19 study finally received gefitinib. So, PFS but not OS will be more properly to be a primary endpoint in NSCLC. Although the survival data are not statistically significant difference between the two arms, there is a trend that these patients with EGFR-TKI had a numerically longer OS time. Patients with advanced NSCLC treated with chemotherapy had a median OS of around 8–10 months. Now, patients with EGFR mutations treated with gefitinib or erlotinib have a median OS of around 18–30 months.
We did not make meta-analysis about adverse events in our article. Although only three trials in our meta-analysis reported the toxicities, the toxicities were mild and acceptable in EGFR-TKI arm than in chemotherapy arm. Diarrhea and rash were more frequent in the EGFR-TKI group, whereas hematologic toxic effects (neutropenia, anemia and thrombocytopenia) were more frequent in the chemotherapy group. Although diarrhea and rash had a high frequency, most of them, were Grade 1 or 2 without any medical intervention. Interstitial lung disease is an EGFR-TKI related lethal disease, however, no >1% patients treated with EGFR-TKI would develop the interstitial lung disease.
Lacking “head to head” RCT comparing gefitinib to erlotinib, now it is hard for us to select gefitinib or erlotinib for patients with advanced NSCLC. However, some retrospective study had found that there was no significant difference between gefitinib and erlotinib. Kim et al.29 reported a study of comparing gefitinib and erlotinib in 342 patients with NSCLC who have failed prior chemotherapy. The result revealed that the ORR (38% vs. 32.2%), OS (12.6 vs. 12.1 months) and PFS (4.6 vs. 2.7 months) were similar in the two groups. Similarly, Wu et al.30 also found type of TKIs did not influence treatment outcomes in patients with EGFR mutation or wild-type EGFR, although there was an obvious disparity in drug selection between gefitinib and erlotinib in his study.
Cetuximab is an IgG1 chimeric human-mouse monoclonal antibody directed against EGFR, which has been approved to be used in NSCLC patients. Recently, two meta-analysis,31, 32 which compared cetuximab plus platinum-based doublet chemotherapy with chemotherapy alone as first-line treatment of NSCLC, have revealed cetuximab plus platinum-based doublet chemotherapy significantly prolonged OS and increased ORR than chemotherapy alone. Chen et al.31 reported cetuximab plus platinum-based doublet chemotherapy prolonged OS (median OS was 10.9 vs. 9.8 months; HR = 0.87, 95% CI = 0.78–0.96; p = 0.004) and increased ORR (RR= 1.32; 95% CI = 1.15–1.52; p < 0.001) over chemotherapy alone. Adverse events in the combination arms were similar in incidence to those of the chemotherapy alone arms, with the exception of an increased incidence of rash and diarrhea. Ibrahim et al.32 also reported similar result. However, translational research did not define a biomarker to predict the efficacy of cetuximab. In an era of personalized therapy, further research is warranted to identify predictive markers of cetuximab benefit in NSCLC.
Although publication bias was not found according to funnel plot, Begg's test and Egger's test, these results should be interpreted very cautiously because there were only six RCTs. Our analysis was limited to the data presented by authors of the source studies. In some cases we had incomplete information. Our meta-analysis was based on aggregate study and substudy data, not on individual patient data. Despite the limitations of our study, we believe that it makes an important contribution to the NSCLC field. Before our meta-analysis, no system review and meta-analysis had reported the difference between the two arms, our analysis confirms that the EGFR-TKI schedule was better than the chemotherapy schedule in PFS and ORR in NSCLC patients harboring activating EGFR mutation.
The findings of this meta-analysis will benefit a substantial part of NSCLC patients, especially the Chinese NSCLC patients. As we know, the morbidity and mortality of NSCLC is quite high in China, together with the huge national population, which make China the biggest NSCLC patients population country in the world. The frequency of EGFR mutations in Chinese NSCLC patients is also significantly higher than that in Caucasian. That is the reason why it only took 2 years for the Phase III OPTIMAL clinical trial to release it results. As a result, these results will in turn help clinicians to choose a proper drug for NSCLC patients with activated EGFR mutation.
In conclusion, The EGFR-TKI regimen significantly prolonged PFS and increased ORR when compared with platinum-based doublet chemotherapy in the previously untreated advanced NSCLC patients with EGFR mutation. As EGFR mutations are predictor of benefit from TKIs, prospective EGFR mutation should be routinely tested before the initiation of treatment in advanced NSCLC patients.