Intraneural injections were once considered harbingers for neural injury, and the practice of regional anesthesia focused on their avoidance. Stimulation techniques centered on maximizing proximity of needles and catheters to nerves without piercing them,^1,2 and studies examined the risks to nerves involved with elicitation of paresthesias.³ The use of ultrasound for regional anesthesia has offered new insights into the relationship between intraneural (into the epineurium) and intrafascicular (within the perineurium) injections of local anesthesia and their relationship with postoperative neurologic complications. Initial studies looked at the effect of intraneural injections in animal models, but more recently, their effect in humans has been described.

Bigeleisen⁴ promoted the idea that intraneural injections of local anesthesia did not necessarily yield postoperative neural dysfunction. He performed ultrasound-guided axillary nerve blocks on 26 patients undergoing surgery on the base of the thumb. Using a 22G short-beveled needle under direct visualization, he attempted to inject each of the four nerves (radial, median, ulnar, musculocuTaneous) with 2 to 3 mL of local anesthetic. Nerve swelling was considered evidence of an intraneural injection and was observed in 72 of the 104 injections, whereas the remaining injections occurred immediately outside the epineurium (Fig. 13.1). Complete surgical anesthesia was achieved in 100% of patients. Postoperatively and at 6-month follow-up examinations, none of the patients reported paresthesias or dysesthesias in the distribution of the four injected nerves. Bigeleisen offers the explanation that fascicles in the axillary nerves are separated by large amounts of stroma between the fascicles. Intraneural injections performed with blunt needles then do not result in neurologic damage because the nerves are able to swell and because the needles do not readily penetrate the perineurium, which is much stronger than the epineurium (Fig. 13.2).⁵ There are other reports of intraneural injections without neurologic sequelae. Upon review of video ultrasound images of an axillary block, Russon and Blanco noted that an intraneural injection of 7 mL of levobupivacaine into the musculocuTaneous nerve had occurred.⁶ That patient reported no subjective or objective neurologic deficit 6 months after the event. An inadvertent intraneural injection of 35 mL of local anesthesia into the femoral nerve, with no adverse sequelae, has also been published.⁷ In a separate study, 34 patients received ultrasound-guided intraneural supraclavicular blocks. Included in this cohort were seven patients with long-sTanding diabetes, three of whom had polyneuropathy. All of these patients also had successful blocks without any measurable neurologic injuries at 6 months.⁸

Although a “donut” sign may represent adequate spread of local anesthesia for many nerve blocks, this is not true for an interscalene block. In an attempt to define what constitutes adequate local anesthesia spread for an interscalene block, Spence et al.⁹