In the spring of 2004, the association between somatic mutations in the epidermal growth factor receptor (EGFR) in non– small-cell lung cancer (NSCLC) and sensitivity to the EGFR tyrosine kinase inhibitor gefitinib was first described. Since that time, numerous retrospective studies have confirmed these original observations. Patients with EGFR mutations treated with gefitinib have a 60% to 90% chance of achieving a radiographic partial response and a median time to progression of approximately 12 months (Table 1). However, these were all retrospective observations that were limited by multiple potential biases including patient selection based on the availability of tumor material, differences in prior treatments, and the methods of EGFR mutation detection. In addition, other biomarkers associated with gefitinib’s clinical benefit, including increased EGFR copy number and EGFR expression as detected by immunohistochemistry, have also emerged over the last 2 years. The contribution and role of EGFR mutations remain to be definitively determined from ongoing prospective studies. In this issue, Inoue et al report the first prospective clinical trial in which chemotherapy-naive advanced NSCLC patients with somatic EGFR mutations were treated with gefitinib as their first systemic therapy. They started the study in June 2004, performed mutational analyses in 99 patients (75 patients were assessable), and found EGFR mutations in 25 of the patients (33%). The mutation frequency in this study is consistent with the previously published somatic EGFR mutation frequency in Japanese lung cancer patients. The 16 patients with EGFR mutations had a response rate of 75% and a median progression-free survival time of 9.7 months, which is consistent with the retrospective observations (Table 1). Inoue et al also found the same types of EGFR mutations as previously identified, with in-frame deletion of exon 19 being the most common type of mutation (Table 1). Despite the small sample size, the study met its primary protocol-defined end point of a 70% response rate, and this observation in a prospective trial confirms the previous retrospective observations associating EGFR mutations and the efficacy of gefitinib. What are the implications of this study? The findings of Inoue et al leave little doubt about the predictive value of EGFR somatic mutations (exon 19 deletions and L858R) for radiographic response after treatment with gefitinib. However, their study does not address whether these patients will survive longer when treated with gefitinib compared with chemotherapy. This will require a phase III trial comparing the outcome of patients with EGFR somatic mutations treated with chemotherapy with the outcome of patients treated with gefitinib. In addition, the role of adding systemic chemotherapy to EGFR tyrosine kinase inhibitors for this patient population remains to be determined and is being evaluated in ongoing studies. Inoue et al also demonstrate the feasibility of EGFR mutation detection in the routine clinical care of chemotherapy-naive patients with advanced NSCLC. The Japanese population is more likely to have a somatic EGFR mutation than the white population (30% in Japanese patients compared with 10% to 15% in white patients). This increased mutation frequency, combined with the rapid turn around time (median, 7 days) for EGFR sequencing, clearly helped this study. In the United States or Western Europe, approximately 300 patients would need to be screened (rather than the 99 patients screened in the study by Inoue et al) to find patients with sufficient tumor material for EGFR sequencing and, ultimately, patients with EGFR mutations to obtain a similar number of assessable patients. Nevertheless, these findings suggest that this may be a worthwhile effort in Western populations as well. Inoue et al found the most common types of EGFR mutations (exon 19 deletions and L858R) in their patient population. To date, more than 20 different EGFR mutations have been described. The predictive value for radiographic response to gefitinib treatment of EGFR mutations, other than exon 19 deletions or L858R, still needs to be determined from ongoing clinical studies. In the study by Inoue et al, there were four patients with NSCLC and EGFR mutations who either had stable (n 2) or progressive disease (n 2) after treatment with gefitinib. What is different about these patients compared with the patients with EGFR mutations who respond to treatment with gefitinib? There are no obvious differences in the types of EGFR mutations among these patients compared with the other 12 patients who achieved a partial response. In addition, the DNA from the tumors of the patients with stable or progressive disease did not contain the previously described EGFR T790M mutation associated with gefitinib resistance. Several recent studies have demonstrated that, both in EGFR mutant NSCLC cell lines and in patients with JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 24 NUMBER 21 JULY 2