Neurolytic agents, such as phenol and alcohol, have been used to ablate peripheral nerves to treat pain and spasticity [12345678910111213]. These agents were, however, not specific for neural tissue and complications have been seen involveing damage to surrounding soft tissue [41112]. These complications have included paralysis, soft tissue (muscle, vascular, skin) damage, and in the case of alcohol can cause pain on injection [41112]. Radiofrequency ablation (RFA) has become a safer alternative method for neurolysis especially for non-cancer pain. For smaller nerves such as ilioinguinal and lateral femoral cutaneous, RFA may be technically difficult given the small area of ablation and difficulty localizing these nerves.

Lidocaine has been demonstrated to be neurotoxic in multiple animal and cell culture studies [1428]. Ready et al. [18] in 1985 performed intrathecal injections of 2-32% lidocaine on rabbits and followed neurologic function for 7 days prior to harvesting the spinal cords for histology. They reported persistent neurologic deficit and major histologic changes starting at 8% concentration. Most animal studies, however, have used concentrations of 5% or less to study neurotoxicity since this the most common concentration used clinically. Previous studies used direct intra-neural injection, intrathecal injection, desheathed nerves, and cell cultures. Only one study, by Kalichman et al. [16], injected lidocaine peri-neurally by piercing the connective tissue separating the neural tissue from overlying muscle in rat sciatic nerves. They reported endoneurial edema, collapsed myelin sheaths, and axonal degeneration at 48 hours. This study, however, used a lower concentration than seen in previous studies, at 3% lidocaine.

Although animal models have shown evidence of neurotoxicity at clinically used concentrations of 5% or less, evidence of neurotoxicity in humans at these concentrations appeared to be much less than expected. This has been well documented in the use intrathecal lidocaine for spinal anesthesia and the phenomenon of transient neurologic symptoms (TNS). Schneider et al. [28] reported 4 cases of TNS using 5% lidocaine. Keld et al. [29] compared 5% lidocaine versus 0.5% bupivacaine, and found that lidocaine caused TNS in 26% of patients versus 3% with bupivacaine. Several prospective studies noted TNS incidence of 4-33% with lidocaine [3031].

Our study studied the utility of higher concentration 10% lidocaine as an alternative, neuro-specific, rapid-onset, and safer alternative to chemical neurolysis with phenol/alcohol in the canine sciatic nerve model using standard ultrasound guided peri-neural nerve block technique. The 10% concentration of lidocaine of 10% was chosen to study, since the use of neurolytic agents in clinical practice usually require injection of a contrast media to confirm correct placement and to rule out vascular uptake. The addition of these agents will reduce the local concentration of 10% lidocaine but keep the concentration above or at the 8% concentration needed for neurolysis as per the study by Ready et al. [18]. Given the previous study by Ready et al. [18], our hypothesis was that higher concentrations above the 5% that caused inconsistent TNS may cause rapid (defined in our study as within 20 minutes) histologic changes making 10% lidocaine a possible alternative agent compared to alcohol or phenol. This was a pilot study only and is limited to evaluation at one time point without direct comparison to other agents or different concentrations.

After Henry Ford Hospital (Detroit, MI) institutional review board and institutional animal care and use committee animal approval, a freshly euthanized dog (purpose bred mongrel 24 kg) was procured for the injection. The dog was placed in right lateral decubitus position and the left gluteal area was shaved and prepped with chloroprep and draped in sterile fashion (Fig. 1). Using a Venue 50 ultrasound from General Electric Healthcare (Buckinghamshire, UK), the left sciatic nerve was identified. A 22 gauge 3.5 spinal needle was guided by ultrasound adjacent to the sciatic nerve using the "in plane" technique (Fig. 2). Ten ml of 10% lidocaine preservative free compounded by Health Dimensions Compounding Pharmacy (Farmington MI) was injected under live ultrasound and demonstrated peri-neural spread (Fig. 2). After waiting 20 minutes, a scalpel was used to dissect along the spinal needle down to the sciatic nerve with gross inspection of the sciatic nerve and surrounding muscle noted (Fig. 3A). A 3 cm segment was excised and preserved in 10% buffered formalin fixative solution. The dog was placed in the right lateral decubitus position to identify the right gluteal region. The process was repeated on the right side except the sciatic nerve was injected with 10 cc 0.9% preservative free saline (Figs. 2, 3B). Both samples underwent progressive dehydration and infusion of paraffin after which they were placed on paraffin blocks. The six sections were cut at 4 µm and stained with hemoxylin and eosin. Microscopic review was performed by a pathologist from Henry Ford Hospital who was blinded to which experimental group each sample was in. The examination was performed using an Olympus BX40 microscope at 400× magnification.

The lidocaine injected nerve, on gross visualization, showed changes with loss of gross architecture on visual inspection (Fig. 3A) while the saline injected nerve did not (Fig. 3B). No gross visual changes were seen in the surrounding muscle or soft tissue seen in either saline or lidocaine injection sites (Fig. 3). On the histologic slides, the saline injected sample showed normal neural tissue (Fig. 4A). The lidocaine injected sample showed basophilic degeneration with marked cytoplasmic vacuolation in the nerve fibers with separation of individual fibers and endoneurial edema (Fig. 4B).