Background: EGFR-driven NSCLCs are dependent on the EGFR pathway for proliferation which EGFR TKIs potently block. Progression occurs when resistance to erlotinib develops—through the emergence of EGFR—T790M or MET gene amplification. This event is likely specific to any given individual tumor clone and therefore on relapse only some tumor clones carry resistance mutations, others likely remain dormant while the EGFR pathway is inhibited providing rationale for maintenance therapy. This concept is supported by clinical observations demonstrating of “tumor flares” in patients taken off of EGFR TKI therapy on progression. Similar analogies exist in prostate cancer/maintenance of castrate state, breast cancer/maintenance trastuzumab. This study seeks preliminary evidence that such a strategy can lead to meaningful benefits in advanced non-small cell lung cancer. Methods: Randomized, phase II study of pemetrexed chemotherapy versus pemetrexed chemotherapy plus erlotinib in patients who derived a clinical benefit from erlotinib but now have evidence of progressive disease (NCT00660816). Pemetrexed is given at standard dose of 500 mg/m2 every 3 weeks to a maximum of 8 cycles while erlotinib is given at 150 mg oral daily dose on days 2-19 of each cycle. Our main hypothesis is that maintenance erlotinib in this patient population could extend PFS by 50%, from 3 to 4.5 months and we seek to enroll 78 patients by randomized phase II screening design to allow 80% power to detect such a difference. Standard eligibility criteria include progression following at least twelve weeks of treatment with single-agent erlotinib during which time the patients experienced a clinical benefit. 23 of planned 78 patients have been enrolled and current enrollment averages 1-2 new patients per month. An amendment to allow either pemetrexed or docetaxel as the chemotherapy base is under review to further enhance accrual. Translational studies, including EGFR and other gene mutation analyses will also be performed.