Perineural adjuncts for peripheral nerve block

Key points.

• •An ideal adjunct for perineural injection would increase the duration of analgesia without prolonging motor block.
• •Perineural clonidine, dexmedetomidine, and dexamethasone can increase the duration of analgesia by a mean period of approximately 2, 4.5 and 4–8 h, respectively.
• •Perineural clonidine and dexmedetomidine have been associated with adverse effects such as bradycardia, hypotension and sedation.
• •Superiority of the perineural over the systemic route of adjunct administration is not yet confirmed.
• •None of the drugs studied as potential adjuncts for peripheral nerve block have been approved for perineural administration.

Learning objectives.

By reading this article you should be able to:

• •Describe the limitations of conventional local anaesthetic agents in peripheral nerve blocks.
• •Discuss the characteristics of an ideal adjunct to local anaesthetic agents for perineural use.
• •Recall the effectiveness, adverse effects and safety of individual drugs used as perineural adjuncts to local anaesthetic agents for peripheral nerve blocks.

Moderate to severe postoperative pain remains a widespread but still unresolved problem. Poor control of pain is associated with decreased satisfaction for patients and an increased incidence of postoperative morbidity and mortality. The use of local anaesthetic (LA) agents for peripheral nerve block (PNB) has long been linked to better pain relief compared with opioid-based analgesia alone. Nevertheless, the effectiveness of LAs is limited by their duration of action, which can be insufficient for adequate postoperative pain control. Patients who have had a single-shot PNB can sometimes complain of slightly greater postoperative discomfort between 16 and 24 h compared with those who have had only systemic analgesics. Rebound pain may occur after single-shot PNBs, resulting in unanticipated overnight sleep disturbances, difficulties in compliance with enhanced recovery and physiotherapy protocol, and increased consumption of opioids. The use of opioids is associated with adverse effects such as constipation, nausea and vomiting, pruritus and secondary hyperalgesia. Hence, strategies have been sought to extend the benefits of single-shot PNBs beyond the maximum of 8–16 h.

A continuous PNB involves the percutaneous insertion of a catheter adjacent to a peripheral nerve, plexus, or fascial plane, followed by the administration of LA through the catheter. Benefits include prolonging the duration of a single-shot PNB;, and reducing pain, supplemental analgesic requirements, opioid-related adverse effects and sleep disturbances.¹ However, problems include inaccurate catheter tip placement and secondary block failure;, catheter-related mechanical nerve irritation, catheter knotting, migration, obstruction or shearing;, fluid leakage or inflammation at the insertion site of the catheter;, bacterial catheter colonisation;, infusion pump malfunction;, myonecrosis with repeated large boluses of bupivacaine;, and LA systemic toxicity. Given that the introduction of continuous PNBs in clinical practice necessitates appropriate resource and service provision to support increased costs, training, and follow up, its use has been limited. The use of ‘perineural adjuncts’ is an attractive and technically simple alternative strategy to continuous PNBs in order to extend the benefits of single-shot PNBs. The term perineural adjuncts refers to the coadministration of pharmacological agent(s) with LA(s) around a peripheral nerve, plexus or fascial plane with the aim of affecting the characteristics of the resulting block. Over time, the number of potential perineural adjuncts has increased to a wide variety of drugs.

For all adjuncts, the benefits associated with their perineural administration should be weighed against the capacity for harm. Potential sources of harm include medication errors associated with the need to combine drugs, the possibility of extended duration of motor block, the risk of adverse effects, and the potential for neurotoxicity. Prolonged motor block could delay ambulation, impede physiotherapy, and increase the risk of falls, interfering with postoperative recovery. Of note, falls in the postoperative period are multifactorial in nature and the relative contributions of decreased motor strength, joint proprioception, and sensory ability to the risk of falls has still not been determined.

In this article, we consider how the data for individual drugs used as perineural adjuncts to LAs in PNBs compares with the characteristics of an ideal adjunct for perineural use (Table 1).

Table 1.

Characteristics of an ideal perineural adjunct.

Available as a preservative-free solution

Chemically compatible with LAs and other perineural adjuncts

Proved mechanism of action

Superiority of perineural administration over all other routes of administration

Effective for all nerve blocks and in combination with all short, intermediate, and long acting LAs Decrease in effective dose of LA Evidence of dose-response relationship with characteristics of resulting block Decrease in times to onset of motor and sensory block Improvement in quality of sensory block

Increase in duration of sensory block to at least 24 h

Absence of prolongation of motor block Increase in duration of analgesia

Absence of systemic adverse effects including respiratory depression, bradycardia, hypotension, hallucinations, sedation, nausea and vomiting, and pruritus

Absence of chondrotoxic, myotoxic, and neurotoxic effects

Attenuation of chondrotoxic, myotoxic, and neurotoxic effects associated with LAs Decrease risk of LA systemic toxicity

Steroids

Dexamethasone

Dexamethasone is a potent long-acting glucocorticoid with minimal mineralocorticoid activity. Its mechanism of action after perineural administration could be secondary to stimulation of glucocorticoid receptors located on the cell membranes of neurones, increasing the expression of inhibitory K⁺ channels and thereby decreasing the excitability of and neuronal transmission in nociceptive unmyelinated C fibres. It may be that its actions are mediated via localised vasoconstriction or systemic anti-inflammatory effects after absorption through the vasculature. Dexamethasone must be administered as a preparation without preservatives such as benzyl alcohol and propylene, both of which can cause neurolytic effects.² Unlike bupivacaine and lidocaine, exposure of ropivacaine to the alkalinity of dexamethasone can result in crystallisation, demonstrating their in vitro incompatibility.

Several meta-analyses have assessed the efficacy of perineural dexamethasone, mainly in the context of brachial plexus block but also dental, peribulbar, sciatic nerve, and transversus abdominis plane blocks. In a systematic review of 29 RCTs, perineural dexamethasone was demonstrated to be associated with an increase in the mean duration of analgesia by approximately 4 h when injected with short- or medium-acting LAs (lidocaine, mepivacaine, or prilocaine) and by 8 h when injected with long-acting LAs (bupivacaine, levobupivacaine, or ropivacaine).³ Furthermore, perineural dexamethasone was related to an increase in the mean duration of motor blockade by approximately 2.5 h when injected with short- or medium-acting LAs and by 4 h when injected with long-acting LAs. Shortening in the times to onset of sensory and motor block were not clinically significant. Perineural dexamethasone was shown to decrease pain scores at rest and on movement, and reduce cumulative morphine consumption at 24 h. In the setting of brachial plexus block, increases in the dose of perineural dexamethasone prolongs the duration of analgesia in a dose-dependent manner until the ceiling dose of 4 mg is reached.⁴

Dexamethasone has been associated with an increase in postoperative blood glucose concentration in diabetic and non-diabetic patients.⁵ Concerns about the potential to cause delayed wound healing and systemic or wound infection have not been confirmed in a recent meta-analysis. In one RCT, the increase in mean serum glucose concentrations with perineural dexamethasone administration was clinically insignificant. Dexamethasone has been linked to a neurotoxic effect in some but not all in vitro studies.⁶ However these findings are subject to the challenges and limitations inherent in the translation of such results from bench to bedside. Histopathological or neurobehavioural changes in in vivo studies and neurological complications in clinical series or trials have not been reported.7, 8 Nevertheless, lack of evidence of harm is not evidence of absence of harm. In order to conclusively determine the safety of a perineural adjunct, given the low incidence of nerve injury after PNB techniques, a total sample size of approximately 16,000 patients would be required to show a doubling of the baseline complication rate of 0.4%.

α₂-adrenoreceptor agonists

α₂-adrenoceptor agonists are characterised by non-selective imidazoline derivatives such as clonidine and dexmedetomidine. Clonidine and dexmedetomidine have an α₂:α₁ activity ratio of 220:1 and 1620:1, respectively, and hence dexmedetomidine is eight times more specific for α₂-adrenoceptors than clonidine.⁹ Given that α₂-adrenoceptors are not located on the axons of peripheral nerves, the mechanism of action of α₂-adrenoceptor agonists after perineural administration is likely to be independent of such receptors. In the refractory phase of an action potential, hyperpolarisation-activated cation currents normally restore the resting potential of the neurone, allowing the return of functional activity. Clonidine and dexmedetomidine block the hyperpolarisation-activated, cyclic nucleotide-gated channels responsible for these currents, maintaining the neurone in a hyperpolarised state. When hyperpolarised, it is unable to generate further action potentials, thereby inhibiting conduction in Aδ and C nerve fibres and producing analgesia. Clonidine could also work, at least in part, through localised vasoconstriction mediated via its less selective activity at α₁-adrenoceptors.

Clonidine

In a meta-analysis of 20 RCTs, the efficacy of perineural clonidine at a dose range of 30–300 μg and most frequently a dose of 150 μg, was evaluated in the context of mainly brachial plexus block but also ankle, cervical plexus, femoral, iliohypogastric, ilioinguinal, and sciatic nerve blocks.¹⁰ Perineural clonidine was associated with an increase in the mean duration of analgesia by approximately 2 h, sensory block by 1.25 h and motor block by 2.5 h, irrespective of whether intermediate-acting (lidocaine, mepivacaine, or prilocaine) or long-acting (bupivacaine or ropivacaine) LAs were injected. Clonidine also increased the likelihood of adverse effects such as bradycardia, arterial hypotension, fainting or orthostatic hypotension, and sedation. Undesirable adverse effects could preclude the use of perineural clonidine in higher-risk patients and for procedures associated with changes in HR and BP; they may also increase the need for perioperative monitoring and might interfere with care pathways put in place to facilitate safe discharge. No evidence of a dose-response relationship was found for either beneficial or harmful effects. In an in vitro study, clonidine at estimated clinical concentrations and higher concentrations had no effect and an increased effect in relation to the neurotoxicity of ropivacaine in rats, respectively.⁶ In a follow up in vivo study, exposure of rats to sciatic nerve blocks containing bupivacaine and clonidine did not result in histopathological or neurobehavioural changes.⁷ The optimal dose of perineural clonidine, at which beneficial effects are maximised and harmful effects are minimised, has not yet been determined.

Dexmedetomidine

In a meta-analysis of 34 RCTs, the efficacy of perineural dexmedetomidine was reviewed in the setting of brachial plexus block.¹¹ Perineural dexmedetomidine was demonstrated to be associated with an increase in the mean duration of analgesia by approximately 4.5 h, sensory block by 4 h, and motor block by 3 h. Furthermore, perineural dexmedetomidine was related to a decrease in the mean time to onset of sensory block by 9 min and motor block by 8 min. Perineural dexmedetomidine increased the odds of adverse effects such as bradycardia, hypotension, and sedation. No evidence of a dose-response relationship was found for either beneficial or harmful effects. Fewer studies have investigated the administration of perineural dexmedetomidine in relation to lower limb blocks, but these have reported similar findings.¹² It has been suggested that dexmedetomidine, unlike clonidine, can extend sensory block without prolonging motor block, producing a differential sensory motor effect, possibly as a result of a greater inhibitory effect on Aδ and C nerve fibres relative to motor neurones.¹³ In an in vivo study examining sciatic nerve blocks in rats, high dose dexmedetomidine added to bupivacaine did not result in histopathological changes and actually appeared to be neuroprotective with reduced perineural inflammation.¹⁴ This neuroprotective effect could be secondary to, at least in part, the inhibition of activated nuclear factor NF-κB, which is responsible for the transcription of pro-inflammatory cytokines, and the modulation of mast cell degranulation. Consistent with this, the use of perineural dexmedetomidine in clinical trials has not been associated with neurological complications to date. The optimal dose of perineural dexmedetomidine, at which the duration of sensory block is maximised and the haemodynamic adverse effects are minimised, has been shown to be 50–60 μg.¹¹

Meta-analyses have suggested perineural dexmedetomidine to be superior to clonidine in terms of indices of block characteristics, but inferior to dexamethasone (Table 2).15, 16 Compared with clonidine, perineural administration of dexmedetomidine was associated with an increase in the mean duration of analgesia by approximately 3.5 h, sensory block by 3 h, and motor block by 2.75 h.¹⁵ Shortening in the times to onset of sensory and motor block were not clinically significant. Such results may be explained by the more pronounced inhibitory effect dexmedetomidine has on neuronal action potentials compared with clonidine. However, perineural dexmedetomidine increased the likelihood of adverse effects such as bradycardia and sedation. Conversely, a systematic review of perineural dexamethasone compared with dexmedetomidine found an increase in the mean duration of analgesia by approximately 2.5 h without prolongation of either sensory or motor block and with a decreased risk of hypotension and sedation.¹⁶

Table 2.

Comparison of the effect of perineural adjuncts on indices of block characteristics and incidence of adverse effects. All data have been extracted from meta-analyses. ND, no difference; OR, odds ratio; RR, risk ratio.

Outcome	Dexamethasone	Clonidine	Dexmedetomidine	Adrenaline	Buprenorphine	Magnesium
Duration of analgesia (min)	+402	+123	+264	+66	+518	+125
Onset of sensory block (min)	−1	−2	−9		ND	ND
Duration of sensory block (min)	+419	+74	+228			+107
Onset of motor block (min)	−1	ND	−8		−0.3	−1
Duration of motor block (min)	+241	+141	+192		+13	+90
Block failure	ND	ND		ND		ND
Pain scores at less than or equal to 24 h	Lower		Lower		Lower
Cumulative postoperative opioid consumption at 24 h (mg)	−19		−10	Not studied	Not studied
Adverse effects	Increase in mean blood glucose concentration by 3.8 mg dl−1	Bradycardia (OR 3.1) Arterial hypotension (OR 3.6) Orthostatic hypotension (OR 2.3) Sedation (OR 5.1)	Bradycardia (OR 7.4) Sedation (OR 11.8)	Hypertension Tachycardia	PONV (RR 5) Pruritus (RR 6)

Adrenaline (epinephrine)

Adrenaline is one of the oldest perineural adjuncts to LAs. It is widely believed to hasten the onset, increase the duration of block characteristics, and delay the systemic uptake of LA, thereby reducing the risk of LA systemic toxicity. It can further serve as a marker of intravascular injection. Its mechanism of action after perineural administration is thought to be, at least in part, related to vasoconstriction secondary to α₁-adrenoceptor stimulation. In a meta-analysis of five RCTs, perineural adrenaline was associated with an increase in the mean duration of analgesia by approximately 1 h.¹⁷ Adrenaline can produce a substantial decrease in blood flow to the peripheral nerve, particularly when administered in combination with LA, increasing the susceptibility to neurotoxicity. Guidelines from the American Society of Regional Anesthesia (ASRA) recommend modifications of the anaesthetic technique to include eliminating or reducing the concentration of adrenaline in patients with acquired peripheral neuropathy. Adverse effects of perineural administration are tachycardia and hypertension, but these are rare in our experience.

Buprenorphine

Buprenorphine is a partial MOP (μ) opioid receptor agonist and KOP (κ) opioid receptor antagonist, which has analgesic and antihyperalgesic properties. Its mechanism of action after perineural administration is likely to be secondary to concentration-dependent block of voltage-gated sodium channels, inhibiting the generation of action potentials in a similar manner to LAs, and interaction with MOP opioid receptors on the axons of unmyelinated C fibres. In a meta-analysis of 13 RCTs, the efficacy of perineural buprenorphine was assessed in the context of mainly brachial plexus block, but also femoral and sciatic nerve blocks.¹⁸ Perineural buprenorphine was associated with an increase in the mean duration of analgesia by approximately 8.5 h and a slightly longer duration of motor block of 13 min. Its main adverse effects were postoperative nausea and vomiting (PONV) and pruritus. However, none of the patients in the studies included in the meta-analysis received prophylaxis against these. Other opioid-related adverse effects were poorly reported. In an in vitro study, buprenorphine at estimated clinical concentrations had no effect in relation to the neurotoxicity of ropivacaine in rats.⁶ In a follow-up in vivo study, exposure of rats to sciatic nerve blocks containing bupivacaine and buprenorphine did not result in histopathological or neurobehavioural changes.⁷

Magnesium

Magnesium is an N-methyl-d-aspartate (NMDA) receptor antagonist that has been found to increase the excitation threshold in peripheral nerves, more so in myelinated Aβ than unmyelinated C fibres. Its mechanism of action after perineural administration could be secondary to the effects of its positive divalent charge on the neuronal membrane or its role as a physiological calcium antagonist. In a meta-analysis of seven RCTs, the efficacy of perineural magnesium was evaluated in the setting of mainly brachial plexus block, but also femoral and thoracic paravertebral blocks.¹⁹ Perineural magnesium was associated with an increase in the mean duration of analgesia by approximately 2 h, duration of sensory block by 1.75 h and duration of motor block by 1.5 h. Perineural magnesium did not increase the risk of PONV. Studies examining the safety of magnesium have shown inconsistent results, with some revealing the potential for neurotoxicity.

Comparison of the perineural and systemic routes of adjunct administration

It is still uncertain whether the placement of drugs in a perineural and hence more targeted manner improves their efficacy or safety as adjuncts compared with systemic administration. If the perineural and systemic administration of drugs were to be equivalent in extending analgesia, then systemic use would obviate any potential risk of nerve injury associated with perineural injection.

For dexamethasone, i.v. administration at a dose of 0.1–0.2 mg kg⁻¹ has been related to decreased postoperative pain and opioid consumption.²⁰ Perineural compared with i.v. administration of dexamethasone in one meta-analysis was associated with an overall increase in the mean duration of analgesia by 3 h, but on subgroup analysis this was statistically significant only with bupivacaine and not with ropivacaine.²¹ No difference in the incidence of PONV was found between perineural and i.v. dexamethasone. However, perineural dexamethasone in another meta-analysis did not provide an incremental benefit over the i.v. route.²² Surprisingly, in a recent study of healthy volunteers, addition of perineural or i.v. dexamethasone to ulnar nerve block with ropivacaine did not show any significant benefits over saline.²³ These results may not be representative of clinical practice in the absence of an operative insult. If the mechanism of action of perineural dexamethasone is solely dependent on subsequent systemic distribution, then the lack of an operative stimulus might have prevented the anti-inflammatory effects of perineural and i.v. dexamethasone. In the case of α₂-adrenoceptor agonists, the results of studies comparing the perineural with the i.m. or i.v. route of administration are inconsistent with regard to the relative duration of analgesia.¹³ In a systematic review, perineural administration of buprenorphine resulted in an increase in the mean duration of analgesia by 7 h compared with i.m. administration, but there were no differences in postoperative pain or PONV.¹⁸

Other perineural adjuncts

Hyaluronidase is an enzyme that can be administered in conjunction with LA to decrease the time to onset of ophthalmic blocks and provide improved akinesia and analgesia. It degrades hyaluronic acid, a glycosaminoglycan that attaches to proteoglycans in the orbital connective tissue and otherwise hinders the spread of LA. Sodium bicarbonate can reduce the time to onset of neuronal block by alkalinisation of the solution, increasing its pH closer to the pKa of the LA and thus favouring the non-ionised form of the LA that is able to penetrate the peripheral nerve to reach its site of action.

In view of either conflicting or limited evidence and worries about their potential for adverse effects or neurotoxicity, consideration of the administration of other drugs as perineural adjuncts in PNBs is not encouraged (Table 3).

Table 3.

Summary of evidence for other adjuncts used for perineural administration.

Adjunct	Evidence
Fentanyl	Conflicting findings for effectiveness overall but possible efficacy when administered with bupivacaine Adverse effects reported include hypercapnia, bradycardia, and sedation
Morphine	Conflicting findings for effectiveness No clear demonstration of superiority over systemic administration
Tramadol	Conflicting findings for effectiveness No clear demonstration of superiority over systemic administration
Ketamine	Lack of demonstrated effectiveness Significant adverse effect profile including drowsiness, hallucinations, and nausea Evidence of in vitro and in vivo neurotoxicity
Midazolam	Limited demonstration of effectiveness Evidence of in vitro and in vivo neurotoxicity No increase in neurological symptoms after intrathecal injection in humans
Neostigmine	Lack of demonstrated effectiveness Significant adverse effect profile including nausea and vomiting Evidence of in vitro and in vivo neurotoxicity

Our experience of perineural adjuncts

In all patients who are about to undergo soft tissue or superficial operative procedures that are not associated with a functional limitation, we do not administer i.v. or perineural adjuncts in combination with PNB. If the operative procedure is likely to be associated with increased postoperative pain, for example after surgery involving the bone, or would be expected to result in a functional limitation, for instance subsequent to hallux valgus correction, we consider perineural administration of preservative-free dexamethasone at a dose of 4 mg, or i.v. dexamethasone at a dose of 0.15 mg kg⁻¹ or 8 mg. In our experience, after perineural or i.v. dexamethasone in conjunction with PNB, 90% and 75% of patients remain free of pain for more than 24 h after ankle and shoulder surgery, respectively.

Off-label use of perineural adjuncts in PNB

The concept of licensing of drugs in humans was introduced to safeguard their effectiveness, quality, and safety in response to certain examples of drug toxicity, such as phocomelia in the developing fetus after the ingestion of thalidomide by pregnant women. Once drugs have been licensed by regulatory authorities such as the European Medicines Agency, Food and Drug Administration (FDA), and the Medicines and Healthcare product Regulatory Agency (MHRA), the approved indications, dosing, routes of administration, and patient populations are specified. However, it is common and legal for drugs to be prescribed in a manner different to that recommended in the label or product licence, that is, off-label.²⁴ Despite its routine use, i.v. dexamethasone for the prophylaxis of PONV, for instance, is an unlicensed indication. It has been advised by the General Medical Council (GMC) in the UK that the prescribing of an unlicensed drug may be indicated in the presence of adequate evidence or experience of using the drug, and where no suitably licensed drug would otherwise meet the patient's need. Further guidance on this is available from the MHRA.²⁵ Similarly, the FDA in the USA has recommended that the use of an off-label drug must be based on reasonable scientific rationale and sound clinical judgement.²⁴ Moreover, the GMC has advised that informed consent should be obtained from the patient for the prescription of an off-label drug.

None of the aforementioned drugs studied as potential adjuncts in PNB have been approved for perineural administration by any of the regulatory agencies. Therefore, their perineural use remains off-label. Nevertheless, it is unlikely that a label for perineural administration would ever be sought by pharmaceutical companies because of the associated costs, effort and time required. Moreover, generic drugs may not have the funding foundations needed for a licence to be pursued. In view of this, with the consistent and supporting evidence of large and well conducted RCTs, the characteristics of a prospective adjunct drug must resemble those of an ideal perineural adjunct, showing convincing clinical benefit over systemic administration, before routine off-label perineural use is considered. Until such time, caution is advised.

Current and future directions

Over the last few years, our understanding of the capacity of individual drugs to act as perineural adjuncts to LAs for PNBs has continued to progress. Dexamethasone and dexmedetomidine are particularly promising, but neither has been shown so far to be similar enough to an ideal perineural adjunct to justify their routine perineural use for PNB. Concerns have focused on the adverse effects and the potential for neurotoxicity of prospective adjunct drugs, and whether perineural administration is superior to systemic injection. Studies that have investigated perineural adjuncts to date all have limitations: these include small size, which increases the possibility of incorrectly rejecting the null hypothesis and publication bias; inconsistencies in the definition and the assessment of outcomes; variations in the method of nerve location; and the type and volume of LAs used. Further trials should consider these concerns and limitations and, importantly, endeavour to use standardised definitions for successful outcomes.

Declaration of interest

E.A. has received grants from the Swiss Academy for Anaesthesia Research, B. Braun Medical AG, and the Swiss National Science Foundation to support his clinical research. He has further received an honorarium from B. Braun Medical AG. N.D. and K.E. declare that they have no conflicts of interest.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Biographies

Neel Desai MBChB BSc FHEA FRCA MRCP MRCS PG Cert Med Ed Dip Reg Anaesth is a consultant anaesthetist at Guy's and St Thomas' NHS Foundation Trust. He has interests in obstetric, regional, and vascular anaesthesia.

Eric Albrecht MD PD-MER DESA is the program director for regional anaesthesia at Lausanne University Hospital. He is chairman of the Regional Anaesthesia Scientific Subcommittee of the European Society of Anaesthesiology (ESA) and a council representative of European Society of Regional Anaesthesia (ESRA). His field of research includes acute postoperative pain, regional anaesthesia, and sleep apnoea syndrome.

Kariem El-Boghdadly BSc FRCA EDRA MSc is a consultant anaesthetist at Guy's and St Thomas' NHS Foundation Trust. He is editor of Anaesthesia Reports and on the editorial board of Anaesthesia. His research interests include airway management and regional anaesthesia.

Matrix codes: 1A02, 2G01, 3A09

Appendix A. Supplementary data

The following is the Supplementary data to this article: