Our hypothesis is that patients receiving therapy based on their baseline tumor ERCC1 levels and EGFR mutations would attain better disease free survival rates than patients in the control arm receiving noncustomized therapy. Using this selective approach, patients with stage II and IIIA non N2 NSCLC in the genotypic arm with low ERCC1 levels will receive cisplatin plus pemetrexed, and those with high ERCC1 levels will not receive cisplatin-based chemotherapy. If they harbor EGRF mutations they will be treated with erlotinib.

The study will be restricted to non-squamous NSCLC for two mains reasons. First, this will enrich the EGFR mutation rate that is awaited to be higher in these tumors than in squamous cell carcinoma. Second, permetrexed cisplatin combination has a promising efficacy and favorable toxicity profile and is of potential interest in the adjuvant setting of resected non-squamous NSCLC.

Inclusion Criteria:

• NSCLC that is defined as other than predominantly squamous cell histology (squamous cell and/or mixed small cell, non-small cell histology is not permitted)
• Surgically resected NSCLC with pathological stage II or stage IIIA non-N2 (TNM classification 2008)
• Performance status (PS) = 0 or 1
• 18 years <= age < 70 years
• Signed inform consent

Exclusion Criteria:

• Squamous cell carcinoma
• Previous cancer excepted those treated for more than 5 years and considered cured, basocellular skin carcinoma, carcinoma in situ of uterine cervix
• Inadequate renal or cardiac functions
• Pregnant women
• Any of the following within 6 months prior to study enrollment: myocardial infarction, severe/unstable angina pectoris, coronary/peripheral artery bypass graft, NYHA class III or IV congestive heart failure, cerebrovascular accident or transient ischemic attack, grade ≥ 2 peripheral neuropathy, peptic ulcer disease, erosive esophagitis or gastritis