• Phase II multicenter study (N = 150) demonstrating feasibility of customized biomarker-based (EGFR and ERCC1) adjuvant therapy. Study was unsuccessful primarily because the ERCC1 assay was determined to be unreliable
• Patients with ERCC1-negative tumors received cisplatin/pemetrexed and those with EGFR mutation-positive tumors received erlotinib. Patients with ERCC1-positive tumors were closely monitored

This creatively designed and highly anticipated prospective trial essentially failed because of the lack of reproducibility of the ERCC1 assay. These findings provide an important lesson not only about the potential pitfalls of molecular assessments but also the temptation to move early data into clinical practice (which a fair number of clinicians did with ERCC1). Ironically, perhaps what was most clinically useful about this study was learned from the control arm of cisplatin/pemetrexed, on which the tolerability of treatment and intended dose delivery closely mirrored findings from the TREAT study (also discussed in this interactive monograph). In that regard, the study supports the common practice of using cisplatin/pemetrexed in the adjuvant treatment of lung adenocarcinoma.