Epidural anesthesia reduced by 34% the sevoflurane required for adequate depth of anesthesia. This effect was not a result of systemic lidocaine absorbtion, but may have been caused by deafferentation by epidural anesthesia or direct rostral spread of local anesthetic within the cerebrospinal fluid. Lower-than-expected concentrations of volatile agents may be sufficient during combined epidural-general anesthesia.

Overall, most spinal drugs in clinical use have been poorly studied for spinal cord and nerve root toxicity. Laboratory studies indicate that all local anesthetics are neurotoxic in high concentrations and that lidocaine and tetracaine have neurotoxic potential in clinically used concentrations. However, spinal anesthesia (including lidocaine and tetracaine) has a long and enviable history of safety. Spinal analgesics such as morphine, fentanyl, sufentanil, clonidine, and neostigmine seem to have a low potential for neurotoxicity based on laboratory and extensive clinical use. Most antioxidants, preservatives, and excipients used in commercial formulations seem to have a low potential for neurotoxicity. In addition to summarizing current information, we hope that this review stimulates future research on spinal drugs to follow a systematic approach to determining potential neurotoxicity. Such an approach would examine histologic, physiologic, and behavioral testing in several species, followed by cautious histologic, physiologic, and clinical testing in human volunteers and patients with terminal cancer refractory to conventional therapy.

For outpatient knee arthroscopy, anesthesia can be provided adequately with regional or general anesthesia. Epidural and general anesthesia provide equal recovery times and patient satisfaction, whereas spinal anesthesia may prolong recovery and have increased side effects. The choice of anesthesia may depend primarily on the patient's interest in being alert or asleep during the procedure.