Ultrasound‐guided block and the incidence of intraneural injection

Sermeus et al. suggested that a tangential approach to needle/nerve contact “should be included in algorithms for performing peripheral nerve block in order to prevent puncture and intraneural injection” 1. We share their concerns with regard to the risk of inadvertent needle trauma and intraneural injection, and would like to invite the authors to reply to some observations we have about their study. The authors state that after “i an extensive search of published articles, our group did not find any studies that evaluated the risk of puncture and intraneural injection based on the approach taken to the nerve”. In fact, in 2015, our group published two ultrasound-guided brachial plexus injection techniques at the interscalene groove 2, performing 52 peri-plexus and intra-plexus interscalene injections in 26 cadavers, all with sonographic and histological confirmation of needle tip and injectate. Our intention was to have our needle tip in either the periplexus position 3, or between the nerve roots, without violating the epineurium around each of the nerve roots. As such, all of our intra-plexus needle trajectories were ‘tangential’ to the nerve roots. Despite this, we found a 11.5% risk of sub-epineurial (below the epineurium) ink in the intra-plexus group, while we found no sub-epineurial ink in the peri-plexus group. Though we applaud the number of injections performed by the authors (158 in five cadavers), the sensitivity of ultrasound to detect sub-epineurial needle or injectate placement with low volume injectate is poor. Even expert observers have been shown to miss one in six histologically-verified intra-neural 0.5 ml injections 4, and to our knowledge, the ability to sonographically distinguish between intra- and extraneural injection below 0.5 ml 5, such as the 0.1–0.2 ml used in this study, has never been quantified. As none of the 158 injections were verified by histology, it is possible, if not likely, that the authors may have underestimated the risk of intra-neural injection with the tangential approach. The authors state that only 57% agreement (four out of seven) between blinded observers was used to define sonographically-defined intra-neural needle or injectate, effectively requiring only one person to break the tie. This suggests the authors may agree regarding the lack of sensitivity of ultrasound image interpretation to reliably determine needle tip position relative to essential elements of the neural architecture. To overcome the issues with sonographic insensitivity in this paper, the authors chose to perform histology after 20 injections, six direct and 14 tangential, in a single cadaver. We have concerns with the methodology used. The authors state, “Two approaches were performed at each nerve/level, separated by 2–3 cm.” Our own research has demonstrated that very small volumes of ink (0.1 ml) may pass liberally through long lengths of a nerve (greater than 2 cm) and involve all fascicles within a nerve in the sub-perineurial space 2. With injections spaced so closely together, there may have been contamination from one site to another, such that ink may be found well beyond the site of injection. For the same reasons, we are also concerned that the appearance of nerves on ultrasound at sites close to the previous injection may have been altered by the injectate at the adjacent level. The volume injected, albeit small, may have altered the nerve borders, echogenicity and the ability to discern nerve swelling. We are concerned that on the basis of only 20 histologically verified injections at eight sites, that the authors conclude that the tangential approach reduces nerve penetration, especially as the authors themselves found a 14% subepineurial rate for the ‘safer’ tangential approach in the limited number of histology evaluations they performed. The nature of nerves at different levels within the brachial plexus and the subsequent risk of needle intrusion, intra-neural injection and fascicular damage may be quite different at the C5 root, versus the axillary region 6. Thus, each location must be meticulously studied with sufficient histological validation to make such a strong conclusion regarding the superiority of one technique over another. The study suggests that targeting the nerve tangentially, rolling the nerve as the needle passes above or below it's ill defined border, is preferable to slowly skewering it. This is a long way from concluding that the tangential technique be adopted as part of an algorithm for safe nerve blockade. We would suggest that ultrasound be used to avoid the borders of nerves altogether and to optimise spread of local anesthetic around nerve structures as much as possible. It is not surprising to us that this study seemed to confirm the alarming frequency with which the neural tissues are violated when needle-nerve contact is encouraged regardless of the trajectory or the bevel direction. We strongly support research to improve the safety of ultrasound-guided nerve blocks, and are grateful to the authors for focus and commitment to this issue. Future cadaveric research examining techniques should always be verified with histology whenever possible, while further clinical studies are needed to look at how far away we can be from nerves structures while achieving reliable neural blockade.