OBJECTIVES: I. Determine whether whole body fludeoxyglucose F 18 (F-18 2-fluoro-2-deoxy-D-glucose; FDG) positron emission tomography (PET) can quantitate the response to induction (preoperative) chemotherapy in patients with esophageal or non-small cell lung cancer (NSCLC) by correlating changes in PET scan images with surgical staging, frequency of complete resection, disease-free survival, and overall survival. II. Correlate PET scan results with the multiple conventional complementary imaging modalities of thoracic and/or abdominal CT, bone scans, and cranial MRI. III. Evaluate the use of PET scanning to uncover disease sites undetected by current imaging modalities.

OUTLINE: This is a nonrandomized study. The choice of chemotherapeutic regimen is at the discretion of the medical oncologist. If the tumor is resectable after treatment with chemotherapy, the patients undergoes the appropriate resection as defined by the thoracic surgeon. All patients have positron emission tomography (PET) scans done within 2 weeks before the first dose of chemotherapy and again about 2-3 weeks after the third or fourth dose of chemotherapy. Patients are asked to not eat or drink 6 hours before coming into the hospital for the PET scan. They are administered an injection of a solution of fluorodeoxyglucose F 18 (FDG) and then undergo a PET scan.

PROJECTED ACCRUAL: An estimated 75 patients (50 lung cancer and 25 esophageal cancer patients) will be accrued into this protocol over 24-30 months.