The primary hypothesis is that disease-free survival is improved in patients undergoing resection for tumor thought to be stage I-III primary non-small cell lung cancer in patients with combined general-epidural anesthesia & analgesia as compared to patients receiving general anesthesia and postoperative patient-controlled opioid analgesia. Patients having surgery for resection of potentially curable lung cancer will be randomized to combined general and epidural anesthesia or general anesthesia with opioid analgesia. The primary outcome will be disease-free survival.

Surgery is the primary treatment of lung cancer, but surgery releases tumor cells into the systemic circulation. Whether this minimal residual disease results in clinical metastases is a function of host defense. At least three perioperative factors shift the balance toward initiation and progression of minimal residual disease.

1. Surgery per se depresses cell-mediated immunity, reduces concentrations of tumor-related anti-angiogenic factors (e.g., angiostatin and endostatin), and increases concentrations of pro-angiogenic factors such as VEGF.
2. Anesthesia impairs numerous immune functions, including neutrophil, macrophages, dendritic cells, T-cell, and NK-cell functions.
3. Opioid analgesics inhibit both cellular and humoral immune function in humans, and promote tumor growth in rodents.

The primary hypothesis is that disease-free survival is improved in patients undergoing resection for tumor thought to be stage I-III primary non-small cell lung cancer in patients with combined general-epidural anesthesia & analgesia as compared to patients receiving general anesthesia and postoperative patient-controlled opioid analgesia. Patients having surgery for resection of potentially curable lung cancer will be randomized to combined general and epidural anesthesia or general anesthesia with opioid analgesia. The primary outcome will be disease-free survival.

The effect of combined epidural/general anesthesia versus general anesthesia with opioid analgesia on the primary outcome of disease-free survival (time to the earlier or recurrence or death from any cause) will be assessed uni-variably with Kaplan-Meier analyses and multivariably (primary analysis) with a Cox proportional hazards regression model adjusting for known risk factors for recurrence, including tumor stage, size of tumor, age, sex, receipt of chemotherapy before or after surgery, and clinical site. As usual for this type of analysis, stopping criteria will be based on number of outcome events rather than enrollment.

Inclusion Criteria:

1. Tumor thought to be primary non-small cell lung cancer (stage 1-3) as determined according to the IASLC Lung Cancer Staging Project[ref];
2. Scheduled for potentially curative tumor resection;
3. Written informed consent, including willingness to be randomized to epidural anesthesia/analgesia plus general anesthesia or to general anesthesia and postoperative opioid analgesia;
4. ASA physical status 1-3.

Exclusion Criteria:

1. Any contraindication to epidural anesthesia, (including coagulopathy, abnormal anatomy).
2. Any contraindication to midazolam, propofol, sevoflurane, fentanyl, morphine, or hydromorphone.
3. Age <18 or >85 years old.
4. Other cancer not believed by the attending surgeon to be in long-term remission.