Tumor necrosis factor (TNF)-alpha has been strongly implicated as a major contributing factor for the development of radiculopathy. In animal studies, the application of TNF-alpha to nerve roots results in pain behavior indicative of radiculopathy. The use of TNF-alpha inhibitors (etanercept and infliximab) have been shown to prevent this pain behavior. Open-label studies in humans have shown both etanercept and infliximab provide excellent, long-term relief in patients with acute radiculopathy from herniated disc. However, a recent placebo-controlled study failed to demonstrate any significant difference from placebo. The investigators have already established the safety of neuraxial etanercept in a trial that has just been completed (not yet published). The objective of this study is to determine whether small doses of epidural etanercept, an anti-TNF-a medication, is an effective treatment for LBP caused by nerve root irritation (i.e., radiculopathy).

As per the wishes of the Dept. of the Army and Walter Reed Army Medical Center Dept. of Clinical Investigation, patients will be randomized in a 3:1 ratio to receive 2 transforaminal epidural etanercept or saline injections at 2-week intervals. Both patients and physicians will be blinded as to the injectate and treatment group. There will be 3 study groups. Group I will receive either 2 mg of etanercept or saline per injection. Group II will receive either 4 mg of etanercept or saline per injection. Group III will receive either 6 mg of etanercept or saline per injection. In each group there will be 8 patients: 6 who receive etanercept and 2 who receive saline. As per a previous study we just completed, etanercept doses will not be escalated until all 6 patients have completed their 1-month follow-up visits without any evidence of toxicity or complications.

Inclusion Criteria:

1. Chronic low back pain of radicular origin of > 2 months but < 1 year duration.
2. Failure of conservative therapy to include physical and pharmacotherapy.
3. MRI evidence of a herniated disc corresponding to the patient's radicular symptoms.
4. Normal white blood cell count (drawn in 1 blood vial).

Exclusion Criteria:

1. Uncontrolled coagulopathy.
2. Pregnancy, which will be ruled out by a urine pregnancy test if any question as to the patient's status exists.
3. Allergy to contrast dye.
4. Unstable medical condition (e.g., unstable angina or congestive heart failure).
5. Rheumatoid arthritis, Crohn's disease or spondylarthropathy.
6. Unstable neurological condition (e.g., multiple sclerosis)
7. Systemic infection
8. Age < 18 or > 70 years.