New insight provided by ultrasonography

Since 1953, we have learned that paresthesias are not required to ensure a successful nerve block. For instance, transarterial injection and loss of resistance can be used to anesthetize the axillary brachial plexus and femoral nerve (fascia iliaca compartment block), respectively [1,18]. Neurostimulation and US constitute additional modalities that do not depend on paresthesias to locate and block a nerve. Conversely, the occurrence of paresthesia, in itself, does not guarantee success. For instance, Boezaart and colleagues [19] observed a 10% failure rate with EP-guided interscalene blocks. For axillary blocks, the success rate with EP ranges from 40% to 90% [20,21], with most trials reporting rates between 76% and 82% [22–25]. These findings, which were previously puzzling, can now be explained with US. Neural blockade is mediated by LA; positioning the needle tip close to the nerve can result in failure if circumferential LA spread does not occur around the nerve [26]. In summary, paresthesias no longer ensure anesthesia, nor does anesthesia require paresthesias anymore.

The controversy related to the safety of EP-guided nerve blocks cannot be solved with US at this time. Though first reported in 1978 [27], US has only penetrated routine clinical practice in the last 10 years. Furthermore, US is seldom combined with intentional EP for peripheral nerve blocks. These 2 factors explain the limited contribution of ultrasound guidance to the understanding of EP's safety. However, US has been instrumental in shedding new light on an unexplained phenomenon: the neurophysiologic mechanism of paresthesia. In the past, the latter was conceptualized (simply) as the inevitable consequence of needle-nerve contact. Conversely, a lack of paresthesia during LA injection was routinely used to denote safety, a guarantor that the needle tip was at a safe distance from the nerve. Because US allows the operator to visualize neural structures, regional anesthesiologists have become increasingly aggressive with their choice of targets. For supraclavicular blocks, the needle tip can now be placed inside the neural cluster formed by the confluence of trunks and divisions of the brachial plexus [28,29]. For axillary blocks, Bigeleisen [30] has reported his initial experience with LA injection inside the terminal nerves. Despite these intracluster/intraneural positions of the needle tips, paresthesias did not occur in all subjects. For instance, with ultrasound-guided supraclavicular blocks, Tran and colleagues [28] and Roy and colleagues [29] reported 19.6% and 45.1% incidences of paresthesia, respectively. On the other hand, Bigeleisen [30] recorded a paresthesia with almost all cases of radial and median nerve puncture. By contrast, only 50% and 79% of patients reported paresthesias with penetration of the musculocutaneous and ulnar nerves, respectively. These findings echo a fact well known to most practitioners and US enthusiasts: paresthesias can be conspicuously absent despite sonographic evidence of needle-nerve contact. Russon and Blanco [31] even reported a case whereby LA injection inside the musculocutaneous nerve failed to generate a paresthesia. These recent US-driven observations beg the question: why do paresthesias occur?

The genesis of a paresthesia remains a deceptively complex neurophysiologic phenomenon. An axon generates signals only at its terminal end [32]; thus, mechanical encroachment on its central portion (during peripheral nerve blocks, for example) should theoretically be imperceptible to the patient. Yet paresthesias do arise with needle-nerve contact. Karaca and colleagues [33] hypothesized that paresthesias do not result from contact with the nerve but from mechanical stimulation of the nerves to the nerve (ie, nervi nervorum). This elegant premise would certainly explain why paresthesias arise often, but inconsistently, despite sonographic confirmation of neural contact by the needle tip. In other words, if the latter were to contact the axon in an area devoid of nervi nervorum, the event would not translate into a shock-like...