During induction of anaesthesia, the intravenous (i.v.) injection of the aminosteroidal neuromuscular blocking drug rocuronium is often associated with withdrawal reactions of the arm into which the drug is injected [1-5]. It is generally accepted that severe burning pain of short duration is the cause of these spontaneous movements and that the drug is not suitable for use in awake patients, e.g. as a subparalysing dose before succinylcholine or in priming [2,4,6,7]. However, even in anaesthetized patients, the drug has been shown to produce an increase in heart rate and sometimes an increase in arterial pressure . The pathophysiological mechanisms leading to this adverse effect are still unclear. Possible explanations comprise activation of nociceptors by unphysiological osmolality or pH of the solution as well as activation by the release of endogenous algogenic mediators such as histamine or bradykinin [1,5,6]. Based on the disagreement on pathophysiological mechanisms, there is controversy about the prevention of rocuronium-induced adverse effects.
Therefore, the aim of this study was to determine the mechanism of the pain induced by rocuronium. We compared different concentrations of rocuronium with vecuronium, an aminosteriodal neuromuscular-blocking drug that causes pain on injection less frequently [9,10], in two different settings. In a first setting, the neuromuscular blocking drugs were compared using dermal microdialysis. In combination with scanning laser Doppler imaging, we determined the time-course of pain sensations, mediator release and protein extravasation, as well as local and axon-reflex vasodilatation, in vivo. In a second setting, direct activation of cutaneous nociceptors by the neuromuscular blocking agents were evaluated using an isolated mouse skin-nerve in vitro preparation and the single-fibre recording technique.
In vivo studies
This part of the study included 10 healthy volunteers with a mean (range) age of 26 (21-37) yr. After obtaining approval from the local Ethics Committee and informed consent, intracutaneous microdialysis was performed using previously described techniques [11,12](Fig. 1). In each subject, 5 single plasmapheresis hollow fibres (diameter 0.4 mm, cut-off 3000 kD; Asahi, Japan) were inserted intracutaneously at a length of 1.5 cm via a 25-G cannula in the volar forearm. All fibres were oriented transversally to the axis of the forearm with at least 3 cm distances. The hollow fibres were perfused with Ringer's solution (Ringerlösung, Bad Homburg, Germany) using a microdialysis pump (Pump 22®; Harvard Apparatus, South Natick, MA, USA) at a flow rate of 4 μL min−1. Tygon® tubes were used to connect the hollow fibres with the syringes (inner diameter 0.38 mm; Cole-Palmer Instrument Co., Vernon Hills, IL, USA). After skin passage, the membranes were inserted in glass capillaries for collecting the dialysate. To minimize outflow resistance, the capillaries were tilted at 5°. The length of the microdialysis fibres being exposed to the air was <3 mm on the inflow and outflow sites.
After a baseline of 60 min, the fibres were simultaneously perfused with rocuronium 1 or 10 mg mL−1, vecuronium 1 or 10 mg mL−1, or a hydrogen chloride-Ringer's solution (NaCl/HCl, pH adjusted to 3.8 as a control) for 30 min in a double-blind and randomly allocated fashion. The stimulation period was followed by a 30 min wash-out period. Pulse oximetry (SPO2), electrocardiograph and non-invasive arterial pressure were monitored throughout and for at least 1 h after the stimulation period.
During the first minute after onset of stimulation, the subjects were asked to rate the itch and pain sensation on numerical rating scales (NRS) ranging from 0 to 10. Separate ratings were obtained for itch (0 = no itch; 10 = maximum itch imaginable) and for pain (0 = no pain; 10 = maximum pain imaginable).
The superficial blood flow of the forearm was measured repeatedly by laser-Doppler imaging (LDI®; Moor Instruments Ltd, Devon, UK). For this purpose, an area of 20 × 10 cm around the injection sites was scanned with a resolution of 16 380 pixels, with each pixel representing a separate Doppler flux measurement. They were recorded for further processing with dedicated software (Moor LDI® Version 3.0). The mean flux was determined separately in an area of 1.5 × 0.5 cm around the respective fibre to determine the local vasodilatation and 0.5 cm beside the fibre to determine the axon-reflex vasodilatation. In contrast to the local vasodilatation, which is primarily evoked by direct vasodilatatory effects of released mediators, the axon-reflex vasodilatation (neurogenic flare) mechanism is initiated by locally induced excitation of nociceptors in the skin, which travel along their cutaneous arborization . When the terminal nerve endings are reached, the incoming action potentials causes release of substance P and calcitonin gene-related peptide (CGRP) with subsequent vasodilatation of small arterioles in the receptive fields of activated nociceptors surrounding the injury site (a few centimetres).
The method of laser-Doppler imaging provides a rapid, non-invasive, detailed analysis of intensity and spatial pattern of vasodilatation in the skin. In contrast to the analysis of the visible flare, laser-Doppler imaging is a more sensitive method for investigating changes of superficial blood flow. The flare area was calculated from all pixels around the fibre in which flux values after stimulation with the neuromuscular blocking drug >99% percentile of the baseline distribution.
For the determination of mediator release, dialysate samples for total protein, histamine and tryptase content were taken every 15 min for 120 min. Dialysate samples for bradykinin content were taken every 5 min only during the stimulation period. Samples were frozen at −20°C in polyethylene cups for later analysis. Total protein content was measured photometrically (MRX reader®; Dynatech, Denkendorf, Germany) using Coomassie blue dye for the analysis and bovine serum albumin as a standard. Histamine was analysed using a fibre-based spectrofluorometric assay. The principle and analytical aspects of this assay have been described in detail elsewhere . The limit of detection of this assay is 5 ng mL−1; the fluorescence output is linear up to about 1000 ng mL−1. Bradykinin was analysed using ELISA (Bradykinin EIAH Kit®; Peninsula Laboratories, San Carlos, CA, USA). Tryptase was analysed by enzyme-immunoassay according to the manufacturer's instructions (UniCAP 100®; Pharmacia & Upjohn, Freiburg, Germany). The assay had a sensitivity of 1 ng mL−1 tryptase and a linear range up to 200 ng mL−1.
In vitro studies
The studies were performed using an isolated mouse skin-nerve preparation (Fig. 2). Details of this preparation have been published previously . The saphenous nerve and the hind paw skin area innervated by that nerve were subcutaneously dissected and excised in 18 preparations from 12 animals. The skin was pinned out corium side up in one chamber of an organ bath. The nerve was pulled into a second chamber where filaments were teased and subdivided until a single-unit activity could be recorded via a gold wire electrode in a layer of paraffin oil. The preparation was superfused (16 mL h−1) at 32°C with synthetic interstitial fluid (SIF ) continuously bubbled with carbogen (95% O2/5% CO2).
To measure the conduction velocity (CV), the nerve endings were electrically stimulated via Teflon®-insulated steel microelectrodes (3-7 MΩ). The mechanical thresholds were tested with calibrated von Frey bristles (1-256 mN). For determination of heat sensitivity, a halogen lamp was placed below the translucent bottom of the organ bath and focused on the epidermal side of the receptive field. After placing a metal ring (6.6-9.6 mm i.d.) on the corium side and evacuating its fluid content, a standard heat stimulus was applied. The temperature was raised linearly in 20 s from 32 to 45°C measured via a feedback thermocouple placed on the corium.
After sensory characterization of an identified single nerve fibre, the receptive field on the corium side of the skin was superfused with a pH-neutralized solution (pH 7.3) of one of either neuromuscular blocking drug (10 mg mL−1 dissolved in acetate-buffered saline). The perfusion rate was chosen to produce a turbulent flow (about 3 mL min−1). Before the arrival of a new solution, the chamber was emptied to provide an instantaneous change of fluid; the drug superfusions lasted for 10 min each. Both neuromuscular blocking drugs were tested twice with a 10 min wash-out period between.
After passing through a 100 Hz-2 kHz filter, the spike signals were amplified 10 000-fold. Online computer records and offline analysis of the spike data were made using the up-dated SPIKE/SPIDI® software package . The spikes were assigned to different classes based on their spike form and according to statistical criteria; only one class resulted in a 'clean' single-fibre recording. From each class, a time-frequency plot of the respective spikes or a peristimulus-time histogram can be calculated to analyse the firing pattern of the unit of interest.
Data obtained from the LDI measurements and the microdialysis samples as well as pain ratings were statistically evaluated using ANOVA in a two-way within-subjects (repeated measures) model. Scheffé's post hoc tests were performed when suitable. Correlations were described using the Pearson correlation coefficient (r). In vitro data were evaluated using Friedman ANOVA (repeated measures design) for intraindividual differences during stimulation with the respective neuromuscular blocking drugs. Pairwise comparisons were evaluated using the U-test. Two-sided P ≤ 0.05 were considered as significant. The STATISTICA® software package (Statsoft, Tulsa, OK, USA) was used for statistical analysis.
In vivo studies
Insertion of the microdialysis membranes was well tolerated by all subjects - they felt comfortable and remained haemodynamically stable for the whole observation period. Even during stimulation with the neuromuscular-blocking drugs, no subject showed any signs of muscle relaxation or systemic anaphylactoid reaction.
Immediately after insertion, local vasodilatation with concomitant axon-reflex vasodilatation, as well as a slight increase of mast cell mediators with protein extravasation, was measured (Figs 3 and 4). Intracutaneous stimulation of the skin with the hydrogen chloride solution adapted to pH 3.8 did not induce vasodilatation or mediator release. Furthermore, no pain was reported at the stimulation sites perfused with this acidic solution (Fig. 5).
In contrast, stimulation with the neuromuscular blocking drugs led to a dose-dependant increase in local and axon-reflex vasodilatation as well as mediator release and pain sensation. Both concentrations of rocuronium and vecuronium showed a significant local vasodilatatory effect directly above the microdialysis membranes lasting for the whole stimulation period (P < 0.05) (Fig. 3). Only high concentrations of both neuromuscular blocking drugs could induce axon-reflex vasodilatation (P < 0.01). No differences in vasodilatatory effects were found between the drugs at same concentrations.
Analysis of the mediator release revealed distinct patterns of mast cell activation of both neuromuscular blocking drugs (Fig. 4). Intradermal stimulation with high concentrations led to a significant increase in histamine and tryptase release compared with neuromuscular blocking both drugs at low concentrations (P < 0.01). Furthermore, rocuronium 10 mg mL−1 evoked the most pronounced increase of histamine and a significantly longer-lasting increase in tryptase (P < 0.05). Interestingly, peak tryptase concentrations were found 30 min after finishing the stimulation period.
In parallel to the histamine release, protein extravasation was most pronounced during stimulation with rocuronium 10 mg mL−1 (P < 0.001), whereas rocuronium 1 mg mL−1, as well as both concentrations of vecuronium, showed only a slight - but significant - increase of protein extravasation (P < 0.05).
Peak pain sensation was reported immediately after onset of the stimulation, followed by a rapid decline in its intensity (Fig. 5). Both neuromuscular blocking drugs elicited significantly more pain at higher concentrations compared with low concentrations and control (P < 0.01). No correlations were found between individual pain ratings and the release of mast cell mediators. Even bradykinin release, which was induced by stimulation with high concentrations of rocuronium as well as vecuronium, did not correlate with the pain sensation reported by the subjects (Fig. 5). In contrast, a significant correlation was found between pain ratings and the axonreflex vasodilatation (r = 0.78, P < 0.001).
In vitro studies
In the isolated skin preparations, 12 unmyelinated high-threshold mechanosensitive C-fibres with conduction velocities between 0.3 and 0.9 m s−1 were identified, of which 11 also showed responsiveness to noxious heat ('polymodal nociceptors'). The sensory characterization procedures occasionally induced some transient ongoing activity that was disregarded until each unit was silent for a control period of at least 3 min. The neuromuscular blocking drug was then superfused (n = 6 units each) for 10 min in neutralized solutions at the clinically used concentration of rocuronium (10 mg mL−1). All of the fibres showed weak but consistent excitatory responses with rapid onset, reaching peak discharges (three spikes per 10 s on average) after around 1 min of superfusion and followed by a slow decline over another 5 min in spite of the continued presence of the drugs (Fig. 6). Eight fibres showed singular periods of bursting discharge for 20-40 s during which the firing rate exceeded 10 spikes in 10 s (up to 72 spikes in 10 s). No significant differences were observed in C-fibre discharge rates after stimulation with vecuronium or rocuronium 10 mg mL−1 (P > 0.05, U-test corrected with the Bonferroni procedure). Repeated stimulation with the respective neuromuscular blocking drug in no case induced a second excitatory response. Moreover, the prolonged drug superfusions regularly left the fibres with a marked desensitization to punctuate mechanical (von Frey) stimulation, suggesting a delayed effect of the high drug concentrations.
We have demonstrated that the aminosteroidal neuromuscular blocking drugs rocuronium and vecuronium induce similar responses by direct activation of cutaneous nociceptors. No correlations were found between pain sensations and mediator release in the skin. However, the clinically used concentration of rocuronium (10 mg mL−1) evoked the most pronounced mast cell degranulation with concomitant protein extravasation.
Rocuronium administered to patients during the induction of general anaesthesia can induce pronounced pain reactions with withdrawal reactions of the arm into which the drug was injected. Incidences of pain-related movements up to 50-80% were observed [4,6,18]. Even vecuronium was noted to cause pain reactions during injection, although to a lesser extend [9,10]. It is generally accepted that pain sensations following i.v. injection of different drugs in peripheral veins are mediated via chemonociceptors in the venous wall . These nociceptors were assumed to respond to the unphysiological osmolality or pH of solutions as well as to endogenous mediators such as bradykinin [20,21]. However, mechanisms of the pain induced by aminosteroidal neuromuscular blocking drugs remained unclear.
We therefore used intradermal microdialysis of human skin, which has proven to be an excellent tool for simultaneous investigation of sensory and vascular effects as well as of mediator release induced by different drugs in vivo[11,12,22]. To date no differences between skin nociceptors and nociceptors in deeper tissues have been described. Thus, it can be assumed that the skin is a valid model also for nociception in deeper tissues. As observed upon i.v. administration, we found intense burning pain shortly after onset of intradermal perfusion with rocuronium 10 mg mL−1. However, similar pain ratings were observed during stimulation with vecuronium 10 mg mL−1. We suggest that the lesser extent of withdrawal reactions seen under clinical conditions is most likely due to its lower neuromuscular ED95, leading to smaller doses (and concentrations) of vecuronium being needed for the neuromuscular block during the induction of anaesthesia. Dilution of the respective drugs caused significant less pain sensation, which again supports a concentration-dependent effect of aminosteroidal neuromuscular blocking agents in respect to pain.
Unphysiological osmolality, or the pH of the neuromuscular blocking drug, can elicit pain sensations. However, the osmolality of the formulations used for rocuronium and vecuronium ranged between 260 and 320 mOsmol kg−1. Klement and Arndt showed that drugs injected into a segment of a dorsal hand vein did not cause pain at >1000 mOsmol kg−1, even when the perfusion time lasted for 10 min . Therefore, it seems unlikely that osmolality is of relevance in the induction of pain after i.v. administration of aminosteroidal neuromuscular blocking drugs.
In contrast, the low pH of 4 in both drug solutions might well be a factor for the induction of pain, since it has been shown that constant perfusion of isolated hand vein segments with acidic solutions of pH 4, or less, caused pain on injection, which increased linearly with lower pH . Polymodal nociceptors are the main target of protons, producing pain as well as pinprick hyperalgesia and a flare reaction [19,23]. Interestingly, while all solutions tested in our study had a pH < 4.2, pain sensations were only reported after stimulation with rocuronium and vecuronium at concentrations of 10 mg mL−1. This can be attributed most probably to the high buffering capacity of the skin, as found by measuring the pH in the eluates. In line with these results, application of the nonsteroidal neuromuscular blocking agent atracurium at 1 and 10 mg mL−1 (pH 3.5) in an identical microdialytical set-up did not provoke any pain, but instead produced mast cell degranulation and consecutive itching . However, it has to be taken into account that even after a large rapid i.v. bolus injection of an acidic solution close to the injection site, the pH of the plasma in clinical practice is well above the pH of an injected solution. Our findings are in agreement with the observations of Borgeat and Kwiatkowski that patients receiving i.v. saline adjusted to pH 4 reported no pain sensation, while patients receiving rocuronium 10 mg mL−1 did so . We agree with the authors that the pH of the solution is not the major cause of pain after i.v. administration of aminosteroidal neuromuscular blocking drugs, since pain ratings in our study were not correlated with the acidity of the tested solutions.
The intradermal stimulation with the neuromuscular blocking drugs led to a dose-dependant increase of histamine and tryptase release from the dermis, causing vasodilatation as well as protein extravasation. Rocuronium 10 mg mL−1 evoked the most pronounced mast cell degranulation with concomitant protein extravasation. Taking into account the similar structural formula of vecuronium, it cannot be excluded that this mismatch might be due to different solvents rather than to the aminosteroid itself. However, activation of mast cells is known to produce pure itch sensations in healthy subjects [12,25], making histamine unlikely to induce pain after administration of aminosteroidal neuromuscular blocking agents.
In contrast, burning pain of short lasting duration seen after administration of rocuronium is often seen after application of bradykinin. Bradykinin is a very effective endogenous excitatory agent of polymodal nociceptors in human beings , producing burning pain after intradermal injection of the skin as well as after i.v. infusion into an isolated hand vein segment . Although bradykinin concentrations in the skin increased significantly during the microdialytic stimulation period, no correlations were found between pain sensation and bradykinin release. Moreover, the delayed release of bradykinin did not match the instantaneous pain sensation observed in the subjects. This mismatch cannot be attributed to a delay caused by the microdialysis technique itself. It has been shown before that even for microdialysis membranes with a lower cut-off (2 kD), the mediator increase in the dialysate reflected a step-like increase of the external mediator concentration with a delay of <2 min .
The pain sensations as well as the axon-reflex vasodilatation are assumed to result from direct activation of unmyelinated cutaneous C-fibre terminals with concomitant release of the vasodilatatory neuropeptide CGRP in their wide-branching arborization . Because of the absence of blood and vascular reactions, the isolated skin-nerve preparation used to record from the primary afferents does not allow for many indirect secondary actions of drugs on nerve endings. Just the mast cell degranulation with histamine release would be a conceivable mechanism. However, histamine only weakly excites a very small subpopulation of nociceptors causing itch , whereas the neuromuscular blocking agents caused weak discharge in every single and higher frequent bursts in the vast majority of C-fibres recorded. The mean peak discharge (three spikes per 10 s) under rocuronium or vecuronium compares best with the effect magnitude of nicotine (10−1 mol), that causes pain in human skin, and is in contrast to stronger algogenic agents such as bradykinin (10−1 mol) and protons (pH 6.1) that induce mean peak discharge rates around one spike s−1 in the isolated rat skin-nerve preparation [23,29,30]. However, the latter agents affect less than half of the polymodal nociceptor population. Thus, the discharge induced by the neuromuscular blocking drugs may well have exceeded the pain thresholds in our subjects - by convergence on spinal dorsal horn neurons. It may also cause the withdrawal reactions in patients. Indeed, nociceptor discharge and pain ratings showed a similar time-course, and tachyphylaxis was observed upon repeated administration of the drugs, which reflects clinical reports that a second injection of rocuronium causes significantly less or no pain . This suggests a desensitizing, thus perhaps receptor-mediated, principle of action on nociceptive terminals. As a first speculation one could think of cholinergic receptors that are the primary target of the neuromuscular blocking drugs, since nociceptors are immunocytochemically as well as functionally equipped with subtypes of both nicotinic and muscarinic acetylcholine receptors . However, further studies are necessary to determine the mechanisms of receptor signalling involved in pain induced by aminosteroidal neuromuscular blocking drugs.
The work was supported in part by a grant for scientific research to W. K. from the Medical Faculty of the University Erlangen-Nürnberg (ELAN) and by the Deutsche Forschungsgemeinschaftt (SFB 353, Projects B12 and C4).
1. Lockey D, Coleman P. Pain during injection of rocuronium bromide. Anaesthesia
2. Moorthy SS, Dierdorf SF. Pain on injection of rocuronium bromide. Anesth Analg
3. Robertson EN. Pain on administration of vecuronium. Anaesthesia
4. Steeger MAH, Robertson EN. Pain on injection of rocuronium bromide. Anesth Analg
5. Joshi GP, Whitten CW. Pain on injection of rocuronium bromide. Anesth Analg
6. Borgeat A, Kwiatkowski D. Spontaneous movements associated with rocuronium: is pain on injection the cause? Br J Anaesth
7. Shevchenko Y, Jocson JC, McRae VA, et al.
The use of lidocaine for preventing the withdrawal associated with the injection of rocuronium in children and adolescents. Anesth Analg
8. Robertson EN, Hull JM, Verbeek AM, Booij LHDJ. A comparison of rocuronium and vecuronium: the pharmacodynamic, cardiovascular and intra-ocular effects. Eur J Anaesthesiol
9. Kent AP, Bricker SRW, Coleman P. Pain during injection of vecuronium. Anaesthesia
10. Chow LH, Ho CM, Yang YC, Lee TY, Lui PW. Vecuronium dissolved in normal saline exaggerates pain on intravenous injection. Chung-Hua-I-Hsueh-Tsa-Chih-Taipei
11. Schmelz M, Luz O, Averbeck B, Bickel A. Plasma extravasation and neuropeptide in human skin as measured by intradermal microdialysis. Neurosci Lett
12. Lischetzki G, Rukwied R, Handwerker HO, Schmelz M. Nociceptor activation and protein extravasation induced by inflammatory mediators in human skin. Eur J Pain
13. Koltzenburg M, Handwerker HO. Differential ability of human cutaneous nociceptors to signal mechanical pain and to produce vasodilatation. J Neurosci
14. Petersen LJ, Poulsen LK, Sondergaard J, Skov PS. The use of cutaneous microdialysis to measure substance P-induced histamine release in intact human skin in vivo. J Allergy Clin Immunol
15. Reeh PW. Sensory receptors in mammalian skin in an in vitro
preparation. Neurosci Lett
16. Bretag A. Synthetic interstitial fluid for isolated mammalian tissue. Life Sci
17. Forster C, Handwerker HO. Automatic classification and analysis of microneurographic spike data using a PC/AT. J Neurosci Methods
18. Cheong KF, Wong WH. Pain on injection of rocuronium: influence of two doses of lidocaine pre-treatment. Br J Anaesth
19. Arndt JO, Klement W. Pain evoked by polymodal stimulation of hand veins in humans. J Physiol
20. Klement W, Arndt JO. Pain on i.v. injection of some anaesthetic agents is evoked by the unphysiological osmolality or pH of their formulations. Br J Anaesth
21. Kindgen-Milles D, Klement W, Arndt JO. The nociceptive systems of skin, paravascular tissue and hand veins of humans and their sensitivity to bradykinin. Neurosci Lett
22. Schmelz M, Zeck S, Raithel M, Rukwied R. Mast cell tryptase in dermal neurogenic inflammation. Clin Exp Allergy
23. Steen KH, Reeh PW, Anton F, Handwerker HO. Protons selectively induce lasting excitation and sensitization to mechanical stimulation of nociceptors in rat skin, in vitro. J Neurosci
24. Koppert W, Blunk JA, Petersen LJ, Skov P, Rentsch K, Schmelz M. Different patterns of mast cell activation by muscle relaxants in human skin. Anesthesiology
25. Rukwied R, Lischetzki G, McGlone F, Heyer G, Schmelz M. Mast cell mediators other than histamine induce pruritus in atopic dermatitis patients: a dermal microdialysis study. Br J Dermatol
26. Kress M, Reeh PW. Chemical excitation and sensitization in nociceptors. In: Belmonte C, Cervero F, eds. Neurobiology of Nociceptors.
Oxford, UK: University Press, 1996: 258-297.
27. Petersen LJ, Poulsen LK, Sondergaard J, Skov PS. The use of cutaneous microdialysis to measure substance P-induced histamine release in intact human skin in vivo. J Allergy Clin Immunol
28. Kress M, Guthmann C, Averbeck B, Reeh PW. Calcitonin-gene related peptide and prostaglandin E2 but not Substance P release induced by antidromic nerve stimulation from rat skin, in vitro. Neuroscience
29. Bernardini N, Sauer SK, Haberberger R, Fischer MJ, Reeh PW. Excitatory nicotinic and desensitizing muscarinic (M2) effects on C-nociceptors in isolated rat skin. J Neurosci
30. Liang YF, Haake B, Reeh PW. Sustained sensitisation and recruitment of rat cutaneous nociceptors by bradykinin and a novel theory of its excitatory action. J Physiol