Background and objectives Accumulating evidence indicates that amitriptyline decreases pain sensation when administered orally, intraperitoneally, or for sciatic nerve block. Previous reports of intrathecal administration of amitriptyline have yielded inconsistent results. The failure of amitriptyline to provide antinociception may partly be related to its high logP (octanol-water partition coefficient) and consequent poor spread within the cerebrospinal fluid. We evaluated spinal block after various concentrations of amitriptyline administered intrathecally in a fixed high volume.

Methods We administered 100 μL of 5, 10, 15.9 (0.5%), 25, 50, or 100 mmol/L amitriptyline hydrochloride solution or 100 μL of 15.4 mmol/L (0.5%) bupivacaine hydrochloride solution intrathecally to rats. The neurologic deficit was evaluated by antinociceptive, motor, and proprioceptive responses, and the spinal cord was examined for histopathologic changes.

Results Doses of 100 μL amitriptyline at 15.9 mmol/L (0.5%) and 25 mmol/L produced longer complete nerve block than did bupivacaine at 15.4 mmol/L (0.5%); 5 and 10 mmol/L amitriptyline produced only partial nerve block. However, with 100 μL intrathecal amitriptyline at 50 and 100 mmol/L, many rats did not fully recover from spinal block. Severe axonal degeneration, myelin breakdown, and replacement of neuronal structures by vacuoles were seen in the spinal root section of animals injected with concentrations higher than 25 mmol/L amitriptyline.

Conclusions At lower doses, intrathecal injection of high volumes of amitriptyline results in long-acting spinal block. At higher doses, intrathecal amitriptyline results in irreversible neurologic deficit. Therefore, we do not recommend the use of intrathecal amitriptyline because of a very low therapeutic index.