The use of induction chemotherapy is feasible and effective. It is also logistically beneficial for decreasing micrometastases and radiation-related toxicity by decreasing tumor burden before definite locoregional concurrent therapy. Previously the investigators conducted several phase II study of IP chemotherapy in advanced NSCLC and demonstrated that IP chemotherapy has a promising activity and readily manageable toxicity profile. Given the encouraging activity of IP chemotherapy in the advanced stage setting, the investigators postulated that their further investigation in the stage III setting might lead to further prolongation of survival times. In addition to cisplatin, Irinotecan has been demonstrated to act as radiation sensitizers in preclinical and clinical setting. Therefore, their use with concurrent radiotherapy might lead to radiation sensitization and improved locoregional control.

Concurrent Chemoradiation therapy is widely accepted as a standard treatment of locally advanced unresectable stage III NSCLC. When compared with the result of radiation therapy alone of CALGB 8433 trial (i.e., 9.7 months), the median survival times have almost doubled over the last 2 decades, but rarely exceeded 18 months after chemoradiation therapy in most randomized trials. On the other hand, a significant portion of patients had to endure the side effects of grade 3/4 esophagitis and also pneumonitis, which resulted in treatment-related deaths in some cases. There is a great need to develop more effective but less toxic treatment strategies. Recently, molecular-targeted therapy using EGFR-TKIs brought new enthusiasm to the NSCLC therapy. The investigators observed a median survival time of 20.1 months in chemo-naïve never-smoker Korean patients with adenocarcinoma of the lung. The benefit of EGFR-TKI was also demonstrated in never-smokers who participated in the phase III trial of carboplatin/paclitaxel with or without Erlotinib (TRIBUTE). Despite a lack of benefit in the overall patient population, Erlotinib conferred a survival benefit to those who had never smoked cigarettes, In this analysis, 105 patients who were identified as never smokers had a median survival of 10 months, similar to the entire study population, when treated with carboplatin/paclitaxel plus placebo. However, for the patients in this subpopulation who were treated with Erlotinib and the same chemotherapy regimen, the median survival increased to 22.5 months (P = 0.01). Furthermore, EGFR mutation was associated with significantly higher response rate and longer survival as compared with those without EGFR mutation. More importantly, the median survival time of those patients with EGFR mutation-positive tumors exceeded 20 months in the majority of the studies. These results are very provocative given the fact that only the patients with stage IIIb not amenable to chemoradiation therapy and stage IV NSCLC patients were included in the study and in many studies, the majority of the patients were heavily pre-treated with multiple chemotherapy regimens. The investigators postulate that if the case were properly selected, EGFR-TKI would significantly improve the overall survival of the patients with locally advanced unresectable stage III NSCLC. The investigators therefore propose a randomized phase II trial to evaluate the efficacy and toxicity of EGFR-TKI Erlotinib in selected group of NSCLC patients with EGFR mutation-positive stage III tumors. The use of induction chemotherapy is feasible and effective. It is also logistically beneficial for decreasing micrometastases and radiation-related toxicity by decreasing tumor burden before definite locoregional concurrent therapy. Previously the investigators conducted several phase II study of IP chemotherapy in advanced NSCLC and demonstrated that IP chemotherapy has a promising activity and readily manageable toxicity profile. Given the encouraging activity of IP chemotherapy in the advanced stage setting, the investigators postulated that their further investigation in the stage III setting might lead to further prolongation of survival times. In addition to cisplatin, Irinotecan has been demonstrated to act as radiation sensitizers in preclinical and clinical setting. Therefore, their use with concurrent radiotherapy might lead to radiation sensitization and improved locoregional control.

Inclusion Criteria:

1. Histologically or cytologically confirmed NSCLC: it is recommended to obtain adequate tissue samples for EGFR mutation analysis.
2. Unresectable stage IIIA (N2) or stage IIIB NSCLC defined as:unresectability confirmed by Surgeon /Stage IIIa T1-3 N2/Stage IIIb T1-4 N3/Stage IIIb T4 N2
3. Age 18 years over.
4. ECOG performance status of 0 or 1.
5. Tumor work-up: within 4 weeks prior 1st day of treatment: chest X-ray; CT of chest, liver, and adrenal glands; bone scan; brain MRI
6. Measurable or un-measurable disease (according to RECIST criteria), documented by CT, MRI, X-ray, or physical exam, as appropriate.
7. Hematology (within 1 week before 1st day of treatment)Absolute Neutrophil Count ³2.0 x 109/L; Platelet ³100 x 109/L; Hemoglobin ³10 g/dl
8. Liver function test (within 1 week before 1st day of treatment)Serum bilirubin £1 x UNL; AST & ALT £2.5 x UNL
9. Renal function (within 1 week before 1st day of treatment)Serum creatinine £1 x UNL. In case of borderline value, 24h creatinine clearance should be > 60 mL/min.
10. Pulmonary function (within 4 weeks before 1st day of treatment)FEV1 ³ 1 Liter
11. ECG without significant abnormalities within 4 weeks before 1st day of treatment.
12. Written informed consent.

Exclusion Criteria:

1. T4 with malignant pleural effusion.
2. Any prior therapy (chemotherapy, immunotherapy, biologic therapy such as EGFR-targeted therapy, radiotherapy) for lung cancer.
3. History of prior malignancies, except for cured non-melanoma skin cancer, curatively treated in-situ carcinoma of the cervix or other cancer curatively treated and with no evidence of disease for at least 5 years.
4. Unintended weight loss > 10% within the last 3 months.
5. Other serious concomitant illness or medical conditions:
6. Congestive heart failure or angina pectoris except if it is medically controlled. Previous history of myocardial infarction within 1 year from study entry, uncontrolled hypertension or arrhythmia.
7. History of significant neurological or psychiatric disorders including dementia or seizures.
8. Active infection requiring IV antibiotics.
9. Active ulcer, unstable diabetes mellitus or other contra-indication of corticosteroid therapy.
10. Significant gastrointestinal abnormalities, including requirement for intravenous nutrition, active peptic ulcer disease, prior surgical procedures affecting absorption.
11. Pregnant or lactating women-Patients (male or female) with reproductive potential not implementing adequate contraceptive measures.
12. Concurrent treatment with any other experimental anti-cancer drugs.
13. Concurrent use of phenytoin, carbamazepin, barbiturates, or rifampin.
14. Mental condition rendering the patient unable to understand the nature, scope, and possible consequence of the study.
15. Patient unlikely to comply with protocol, i.e., uncooperative attitude, inability to return for follow-up visits, and not likely to complete the study.