SummaryThis study used a radionuclide imaging technique to map the spread and density contours of phenol in glycerin injected into the epidural space of cancer patients. Correlations were made between phenol injectate volume, sequence of injection, position of patient, and resultant epidural spread and analgesic outcome. Fifteen patients with cancer pain (average age of 61 years) were treated with serial epidural phenol in glycerin injections. Phenol in glycerin is miscible with [99mTc]sulfur colloid. An assumption was made that the admixture injected epidurally was inseparable prior to absorption and the spread was recorded by continuous gamma camera observations with computer collection for 30 min postinjection. The spread of injectate volumes of 2, 3, and 4 ml were compared and further correlations made between observed spread and sequence of injections (first to fifth in series within each patient) and position of patient. Small volumes of phenol may spread extensively in the epidural space (3 ml spreads a mean 13.6 segments) with wide variation among patients. Initial phenol injections spread further than subsequent injections. Maximum spread is achieved by 15 min postinjection and epidural distribution is mostly uniform, independent of patient position. Good analgesia was obtained in 14 patients (93%). Epidural neurolysis using serial injections of small volumes of phenol in glycerin is an effective, safe technique for cancer pain relief. Injectate volumes larger than 3 ml may be unnecessary and potentially dangerous.