Dexamethasone and peripheral nerve blocks: back to basic (science)

Prolongation of analgesia after surgery under regional anaesthesia is an attractive goal for clinicians. Many investigators have sought the ‘holy grail’ of an analgesic adjuvant that both prolongs pain relief and avoids side effects after a single-shot peripheral nerve block. Although many agents have failed this test (opioids, ketamine, clonidine, etc.), the perineural addition of dexamethasone to local anaesthetic has been shown in several studies to prolong the analgesic effect, and its use has become common in clinical practice around the world. Not surprisingly, much research has been performed with the laudable aim of providing Level 1 evidence¹ via RCT design and systematic review and meta-analysis (SRMA).2, 3, 4, 5, 6 Despite this, we still do not have an adequate answer as to whether perineural dexamethasone is superior to systemic administration alone, and the mechanisms of action of dexamethasone in these clinical contexts have yet to be defined. The fact that multiple RCTs and SRMAs have failed to provide clinicians with an answer to the aforementioned question7, 8 should, perhaps, not be a surprise.9, 10 The susceptibility of SRMAs in general¹¹ (and specifically with regard to dexamethasone and peripheral nerve block⁷) to the effects of publication bias, selection bias, and study heterogeneity means that SRMAs on the same topic conducted by different research groups can reach different conclusions.

Furthermore, the lack of safety data in the literature, the ‘off-label’ use of dexamethasone as an analgesic adjuvant for local anaesthetics and the increased risk of medication error by adding an agent of uncertain benefit to local anaesthetic solutions are of concern. Although perineural steroids have been in use for many years both centrally and peripherally in the chronic pain clinic, there is a paucity of data in the literature to support the safety of this practice. In contrast, i.v. doses of dexamethasone in excess of 0.11 mg kg⁻¹ are known to form a safe and effective part of a multimodal postoperative analgesic¹² and anti-emetic¹³ regimen, provided the prescriber acknowledges the potential impact on glycaemic control in the immediate perioperative period.¹⁴

In this issue of the British Journal of Anaesthesia, Marhofer and colleagues¹⁵ add a valuable piece to the puzzle by presenting their findings on the comparative effects of i.v. and perineural administration of dexamethasone on the duration of sensory block achieved with ropivacaine 0.75%. Using a randomised triple-blinded crossover design, 24 volunteers underwent ultrasound-guided ulnar nerve block with ropivacaine on three separate occasions: once with the addition of perineural dexamethasone and i.v. saline, once accompanied by perineural saline and i.v. dexamethasone, and once with perineural and i.v. saline. Using a visual analogue scale in response to pin-prick testing, the investigators found no significant difference in sensory block duration between the three groups.

Given the wealth of recent literature in this area, the finding that neither dexamethasone groups produced block duration longer than saline placebo is surprising. Some of the reasons for this finding (which seems to contradict earlier clinical data¹⁶) may be related to the volunteer nature of this study with no surgical incision. It is possible that the absence of a surgical insult may have prevented dexamethasone having an opportunity to exert its known anti-inflammatory effects,¹⁷ and this may also explain the lack of effect on the primary outcome in the i.v. group. If we hypothesise that the action of perineural dexamethasone is exclusively by subsequent systemic distribution, then the absence of an effect in the perineural group is also explained.

Whereas Marhofer and colleagues¹⁵ used a volunteer model to define and test sensory block duration precisely, previous clinical trials in this area have used patient-focused surrogates for this outcome (e.g. time to first analgesic request or postoperative verbal rating pain scores). The difference in outcome measures is important. The findings of Marhofer and colleagues¹⁵ raise the possibility that the effectiveness of dexamethasone seen in clinical trial outcome data was attributable to systemic absorption and anti-inflammatory/analgesic effects, rather than as a result of a direct effect on nervous tissue.¹⁸ If, however, the mechanism by which the effect of perineural dexamethasone seen in patients is local to the nerve (e.g. by direct action on nerve impulse transmission), then the absence of an observable effect in the perineural group volunteers of Marhofer and colleagues is difficult to explain.

There is a paucity of evidence relating to the effects of dexamethasone, and other glucocorticoids, on nerve tissue. Dexamethasone applied topically to rat sciatic nerve in vivo results in reduced nerve blood flow.¹⁹ Methylprednisolone applied topically to rat plantar nerves in vivo reduced C-fibre, but not A-beta fibre, transmission.²⁰ In a rat in vivo neuropathic pain model, methylprednisolone applied topically to chronically constricted sciatic nerves reduced some signs of neuropathic pain (heat hyperalgesia and mechano-allodynia, but not mechano-hyperalgesia).²¹

We also question the logic of placing identical doses of steroid perineurally and intravenously. Many studies have placed doses of several milligrams of dexamethasone perineurally. We suggest that, if i.v. dexamethasone has a direct effect on a blocked nerve, then the concentration of steroid reaching that nerve should be very much less than that achieved by placing the local anaesthetic directly onto the nerve. We suggest that this, taken in context with the results of Marhofer and colleagues,¹⁵ also points to a depot (non-local) anti-inflammatory analgesic effect of the steroid rather than a direct effect on the nerve.

The study of Marhofer and colleagues¹⁵ reinforces the need for further preclinical research into the effects of perineural dexamethasone on nervous tissue. Such research should identify the effects, if any, of dexamethasone on axonal transmission and neural blood flow (in the presence and absence of local anaesthetic), and the degree to which perineural dexamethasone is systemically distributed. Until such research is undertaken, we do not believe that the current literature provides support for perineural (in preference to systemic) administration of dexamethasone during peripheral nerve block. Given the analgesic and anti-emetic properties of i.v. dexamethasone, we remain supportive of its use via the i.v. route in anaesthesia, with or without peripheral nerve block.

Authors' contributions

All authors contributed to the drafting, revision, and final approval of the manuscript.

Declarations of interest

DWH and NMB declare that they have no conflicts of interest. JGH is the Associate Editor in Chief of the British Journal of Anaesthesia and is a member of the journal's editorial board. He accepts fees for advice to the Crown Prosecution Service, the police, solicitors, and coroners. CJLM is an associate board member of the British Journal of Anaesthesia.