TORCH Study: How Much Longer Should We Continue to Use Erlotinib in Unselected Patients With Non–Small-Cell Lung Cancer?

Abstract

TO THEEDITOR: Gridelli et al 1 recently published the final results of the TORCH (Tarceva or Chemotherapy) study, a multicenter randomized phase III trial designed to demonstrate noninferiority, in terms of overall survival (OS), between first-line erlotinib followed by cisplatin/gemcitabine at progression, and the standard inverse sequence. Patients with stage IIIB (with pleural effusion or supraclavicular nodes) or IV non–small-cell lung cancer (NSCLC) with Eastern Cooperative Oncology Group performance status of 0 to 1 were eligible. To test the hypothesis, 760 patients were randomly assigned during a 3-year period (between December 2006 and November 2009) in Italy and Canada. However, at the first planned interim analysis, the independent data monitoring committee recommended early study termination because of ethical reasons, as a result of substantial inferiority of the experimental arm. In fact, the data showed a significantly reduced total progression-free survival (PFS) for patients treated with erlotinib compared with those treated with chemotherapy (6.4v 8.9 months, respectively; hazard ratio, 1.21). In addition, first PFS (time from start of treatment to the signs of first progression) was 2.2 months for the experimental arm versus 5.4 months for the standard arm (hazard ratio, 1.53). Indeed, the authors claim that there was a statistically significant interaction of treatment effect with sex (P .014), smoking status (P .001), and epidermal growth factor receptor (EGFR) mutation status (P .006). In this regard (data shown in Fig 3 of the article), the authors should explain how it was possible that among the 20 patients who were found to be positive for EGFR mutations and were treated with chemotherapy first, a median PFS of 6.9 months (95% CI, 6.6 to 9.6 months) could translate into 32.5 months of median OS (95% CI, 17.3 to not achieved). Conversely, among the 19 patients who were positive for EGFR mutations and received erlotinib first, a median PFS of 9.7 months (95% CI, 5.7 to 18.2) translated into only 18.1 months of median OS (12.4 months to not achieved). If these data were confirmed in a larger number of patients, instead of the handful of patients enrolled onto the TORCH study for whom EGFR mutation status was evaluated, it could be suggested that this subset of patients be treated with chemotherapy first and then with erlotinib to obtain the observed prolonged median OS (32.5 months). However, who could accept this biologic nonsense? It would be interesting to hear from the authors on this point, even if they concluded that first-line erlotinib followed by second-line chemotherapy is not recommended compared with the reverse sequence, in the treatment of unselected patients with advanced