Does Dexmedetomidine Have a Perineural Mechanism of Action When Used as an Adjuvant to Ropivacaine?: A Paired, Blinded, Randomized Trial in Healthy Volunteers

PERIPHERAL nerve blocks are frequently performed as postoperative pain treatment or for surgery. Ropivacaine is one of the most commonly used local anesthetics providing analgesia of 9 to 14 h.^1,2  The relatively short duration of single-shot peripheral nerve blocks frequently leaves the patient in pain when the block wears off. In order to prolong the duration of nerve blocks, the α₂-receptor agonist dexmedetomidine has been investigated as an adjuvant to different local anesthetics in a number of trials. Most trials show a prolongation of the nerve block,^3–7  as well as time to the first request of analgesics^3,5–7  when coadministering perineural dexmedetomidine with local anesthetics compared to local anesthetics alone.

Although two previous studies on rats indicate that the effect of dexmedetomidine as an adjuvant to local anesthetics is primarily peripheral,^8,9  the evidence from human studies is less straightforward. Marhofer et al.¹⁰  demonstrated a profound prolongation of the nerve block when administering low-dose dexmedetomidine perineurally but also a lesser prolongation with systemic administration in healthy volunteers. In contrast, a trial in patients undergoing arthroscopic shoulder surgery showed a similar prolongation of an interscalene block whether dexmedetomidine was administered perineurally or systemically.¹¹  Dexmedetomidine has centrally mediated analgesic effects at the cerebral and spinal levels through an α₂-receptor mechanism.¹²  When dexmedetomidine is administered perineurally, it is absorbed and redistributed and may exert systemically mediated effects. Thus, none of the previous human studies offers complete control of the systemic contribution of the adjuvant, and further, both human trials investigating systemic effects^10,11  used lower than optimal doses of dexmedetomidine.¹³ 

We have developed a model offering control of the systemic contribution of dexmedetomidine. This control is achieved by performing simultaneous bilateral saphenous nerve blocks in healthy volunteers with ropivacaine plus dexmedetomidine in one thigh and ropivacaine alone in the other thigh. The dexmedetomidine administered on one side will be absorbed and redistributed systemically and subsequently equally affect each nerve block. Consequently, a longer duration of the nerve block receiving dexmedetomidine would solely be attributable to a perineural effect.

This randomized paired trial in healthy volunteers was designed to test the hypothesis that dexmedetomidine prolongs the duration of a peripheral nerve block by a peripheral mechanism when controlling for possible systemic effects.

We conducted a paired, blinded randomized trial in healthy volunteers. The Danish National Board of Health, København S, Denmark (EudraCT 2014-005651-89; Food and Drug Administration equivalent), the Regional Ethics Committee, Sorø, Denmark (VEK SJ-438), and the data protection agency approved the study. The study was registered at www.clinicaltrials.gov (NCT02488473) by the principal investigator (J.H.A.) on June 30, 2015. The Copenhagen Good Clinical Practice Unit monitored the trial, which was conducted according to the Declaration of Helsinki. The trial report complied with the Consolidated Standards of Reporting Clinical Trials guidelines (fig. 1, Consolidated Standards of Reporting Clinical Trials diagram).¹⁴  We conducted the trial at the postanesthesia care unit (Department of Anesthesiology, Zealand University Hospital, Køge, Denmark) in the period August to September 2015.

Eligible volunteers were males, more than 18 yrs old, and had American Society of Anesthesiologists physical status I. Exclusion criteria were inability to read and speak Danish, allergies to the involved drugs, weekly alcohol consumption of more than 21 units, medical abuse, use of any analgesics within 48 h or consumption of opioids within 4 weeks, neuromuscular defects, former surgery, or other trauma to the lower extremities. Signed informed consent was obtained from all subjects before enrollment. Upon screening, the principal investigator included the participants consecutively.

Before block performance, we assessed the ability to discriminate temperature with an alcohol swab and pinprick in the sensory area supplied by the saphenous nerve on the medial aspect of the crus halfway between the knee and ankle. In addition, we performed three other sensory tests:

A computer-controlled thermode placed on the medial aspect of the crus was heated to 45°C for 30 s, and the volunteers rated the worst pain experienced during the test using a visual analog scale (VAS) score (0 to 100 mm). Pain scores were considered normalized when VAS exceeded baseline values ± 10 mm.

This was the lowest temperature that was perceived as warm. The thermode was heated at 1°C/s from 32°C to a maximum of 52°C. Participants pushed a button when the thermode felt warm, and the temperature was recorded. The test was considered normalized when the measured temperature was less than the baseline value +2°C.

This represents the lowest temperature that was perceived as painful. The thermode was heated at 1°C/s from 32°C to a maximum of 52°C. Participants pushed a button when the thermode felt painful, and the corresponding temperature was recorded. The test was considered normalized when detection thresholds were less than the baseline value +2°C.

The latter two tests were repeated four times with a 4- to 6-s pause between each measurement, and the mean value was used. We used a Modular Sensory Analyzer Thermal Stimulator (2.5 cm², Thermotest; Somedic A/B, Sweden) for these tests.

We performed the sensory tests every hour post block, at the same site as the baseline measurements, until values normalized, besides from 4 to 10 h post block (approximately midnight to 6 am) where the participants were allowed to sleep. Warmth detection threshold and heat pain detection threshold were omitted, as long as the VAS score during tonic heat stimulation was 0 (during the first many hours). These values cannot be measured when the VAS score during tonic heat stimulation is 0. As the block started to resolve and VAS scores during tonic heat stimulation became more than 0, the temperature discrimination test was repeated every 30 min, and warmth detection threshold and heat pain detection threshold were performed hourly. The principal investigator (J.H.A.) or the subinvestigator (H.S.I.) performed all tests.

We secured intravenous (IV) access before block performance and monitored all participants using continuous electrocardiography, noninvasive blood pressure, and peripheral oxygen saturation. We scored the level of sedation using a verbal ranking scale 0 to 3 (0 = no sedation, 1 = light sedation, 2 = moderate sedation, and 3 = severe sedation). Each participant received bilateral ultrasound–guided saphenous nerve blocks using the Philips Sparq ultrasound system (Philips, The Netherlands) for needle visualization. After skin preparation with chlorhexidine, the femoral artery was visualized at the midthigh level in the short axis below the sartorius muscle. We inserted a 22-G × 80-mm needle (Pajunk^® SonoPlex Stim cannula, Germany) in plane with the linear probe transfixing the sartorius muscle. The study medication was injected slowly anterolateral to the femoral artery in the adductor canal in 5-ml aliquots preceded by aspiration. We chose the adductor canal approach since this technique results in a high success rate when blocking the saphenous nerve.¹⁵  The volunteers received a right-sided followed immediately by a left-sided block.

The Pharmacy of Region Hovedstaden (Herlev, Denmark) prepared the computer-generated randomization list and the study medication. Subjects were randomized to receive perineural ropivacaine plus dexmedetomidine in the right or left thigh, in a 1:1 ratio. The opposite thigh received ropivacaine plus saline. They prepared one box for each participant marked with sequential numbers. Each box contained two vials, identical in appearance, one containing 1 ml dexmedetomidine, 100 μg/ml, and the other 1 ml isotonic saline. According to the randomization list, these vials were marked left and right legs. Each volunteer received bilateral saphenous nerve blocks with 20 ml ropivacaine, 0.5%, mixed with 1 ml dexmedetomidine, 100 μg/ml, in one thigh and 20 ml ropivacaine, 0.5%, with 1 ml isotonic saline in the other thigh in a blinded fashion. Investigators, participants, and all other personnel were unaware of the allocation and remained blinded until completion of data analysis. The study medication was drawn in syringes marked left and right legs by the principal investigator (J.H.A.) and checked by the subinvestigator (H.S.I.).

The primary outcome measure was the duration of nerve block assessed by temperature discrimination ability with an alcohol swab.

The secondary outcome measure was the duration of sensory block assessed by pinprick, maximum pain during tonic heat stimulation, warmth detection threshold, and heat pain detection threshold. VAS scores during tonic heat stimulation were measured 1 h after recovery of temperature discrimination in order to detect a possible rebound pain (which would be indicated by an increase in pain scores, compared to baseline measurement).

Using data from a previous study by our group with a similar setup, we found a mean duration of the saphenous nerve block of 22 h with an SD of 4 h. Setting the minimal relevant difference to 4 h, α = 0.05 and β = 0.10, a sample size of 18 volunteers would be required. Twenty-one volunteers were included to compensate for possible dropouts.

Statistical analysis was performed using the IBM© SPSS© Statistics version 23 for Mac (IBM Corp., USA). Data were entered twice and merged to check for discrepancies.

According to visual inspection of histograms, normal Q–Q plots, box plots, and Shapiro–Wilk test, all data except rebound VAS scores were approximately normally distributed.

Data are presented as means with a 95% CI. Comparisons are made using the paired Student’s t test. Since the data set is relatively small, we also analyzed the data using the Wilcoxon signed-rank test.

We defined block failure as a preserved ability to sense cold using an alcohol swab 2 h post block. A partial block was defined as the inability to discriminate temperature when stimulated with an alcohol swab, but VAS score more than 0 during tonic heat stimulation, 2 h after block performance.

Sedation and hemodynamic data are presented descriptively since the volunteers received the same amount of dexmedetomidine.

We screened and included 21 healthy volunteers from August to September 2015. All participants received the assigned treatment, and there were no dropouts. No partial blocks or block failures were observed. All participants could be analyzed for every outcome. Demographic data of the volunteers are presented in table 1.

There was a statistically significant difference in the duration of sensory block assessed with an alcohol swab: mean duration in the leg receiving ropivacaine plus dexmedetomidine was 22 h (95% CI, 21 to 24) compared to 20 h (95% CI, 19 to 21) in the leg receiving ropivacaine plus placebo with a mean difference of 2 h (95% CI, 1 to 3; P = 0.001; fig. 2).

The duration of sensory block was also significantly longer in the leg receiving ropivacaine plus dexmedetomidine compared to ropivacaine plus placebo when assessed by pinprick, pain during tonic heat stimulation, and warmth detection threshold but not heat pain detection threshold (table 2). Analyzing the data using Wilcoxon signed-rank test did not alter these results for any parameter.

Rebound VAS scores during tonic heat stimulation were lower after both treatments compared to baseline VAS scores. The median ΔVAS score was −4 mm in the leg receiving ropivacaine plus placebo, and the ΔVAS score was −2 mm in the leg receiving ropivacaine plus dexmedetomidine, with no difference between values according to Wilcoxon signed-rank test (P = 0.29).

One volunteer experienced minor paresthesia in a 5- × 5-cm area on the medial aspect of the lower leg lasting for a week. At 3-month checkup, there was still discrete numbness in that area, but at 5 months, all symptoms had resolved. One volunteer experienced a heart rate of 32 and blood pressure 80/40 mmHg 1 h post block accompanied by dizziness requiring atropine 0.5 mg IV. We observed a heart rate lower than 40 in three participants (trained athletes), occurring during nighttime when the participants were asleep, between 5 and 10 h after block performance. All incidents resolved quickly upon awakening, and none required treatment. Hemodynamic data are presented in figures 3 and 4. Mild to moderate sedation occurred in all volunteers during the first 4 h post block (fig. 5).

In this trial, we have demonstrated that perineural coadministration of dexmedetomidine and ropivacaine leads to a significant prolongation of a saphenous nerve block of 2 h, attributable to a peripheral mechanism when controlling for systemic effects.

This is less than the a priori defined minimal relevant difference of 4 h. Only 7 of 21 participants experienced a sensory nerve block lasting 4 h longer on the side receiving ropivacaine and dexmedetomidine.

Our study design provides control of the systemic contribution of dexmedetomidine in the prolongation of a nerve block by ensuring that the systemic effects of dexmedetomidine are equal in both nerve blocks. This is achieved by performing simultaneous, bilateral saphenous nerve blocks with ropivacaine, adding dexmedetomidine on one side only. As the perineurally administered dexmedetomidine on one side is absorbed and redistributed systemically, the systemic effects are the same for each block. The nerve block receiving ropivacaine plus placebo is, therefore, in fact, a systemic control group. The only difference between the two legs/blocks is the higher perineural concentration of dexmedetomidine on one side. Consequently, a longer duration of the nerve block including dexmedetomidine would be attributable to a peripheral effect. Several other studies have assessed systemic effects in a unilateral setup.^8,10,11  The bilateral setup for systemic control in the current study has only been employed in one previous trial, and like us, they found a peripheral mechanism of action when controlling for systemic effects.¹⁶  However, that study was performed in rats, using a supraclinical dose of dexmedetomidine as an adjuvant to bupivacaine for a sciatic nerve block.

A two-group design where the intervention group receives perineural ropivacaine with dexmedetomidine and the placebo group receives ropivacaine alone is commonly used in adjuvant studies. Most trials report a sensory prolongation of the nerve block^3–7  and a prolonged time to the first request of analgesics.^3,5–7  That type of trial setup is not designed to evaluate where dexmedetomidine exerts its actions. The observed prolongation in the dexmedetomidine group in these studies may be caused by a peripheral mechanism of action or by central effects as absorption and systemic redistribution of the perineurally administered dexmedetomidine take place. Fritsch et al.¹⁷  measured plasma levels of dexmedetomidine after perineural administration in a trial comparing the addition of 150 µg dexmedetomidine to ropivacaine with ropivacaine alone in an interscalene nerve block. Plasma dexmedetomidine peaked 30 min after perineural administration, but 3 h later, the concentration of dexmedetomidine was low. The clinical difference in the duration of the nerve blocks became apparent from 8 h after block performance when dexmedetomidine levels, according to the half-life of approximately 2 h, would be very low. They concluded that the block-prolonging effects of dexmedetomidine are not systemic in origin. A limitation to this study is that, for targets behind a transporter barrier, such as in the central nervous system, the plasma concentration of a drug may not be very relevant.

Marhofer et al.¹⁰  performed a volunteer study. Participants received an ulnar nerve block using 3 ml ropivacaine, 0.75%, with either 20 μg dexmedetomidine perineurally, IV, or no dexmedetomidine, thereby assessing both perineural and systemic effects, but without a paired, bilateral setup. They found a prolongation of the sensory nerve block of 60% in the perineural group but also a 10% prolongation in the group receiving systemic dexmedetomidine when compared to placebo. They suggested a primarily peripheral mechanism of action, supporting our findings.

A clinical trial employed the triple-group design on ambulatory shoulder surgery patients.¹¹  All patients received an interscalene nerve block using 15 ml ropivacaine, 0.5%, with 0.5 μg/kg epinephrine plus dexmedetomidine perineurally, IV, or no dexmedetomidine. All received general anesthesia. The duration of analgesia defined as self-reported time to first pain at the surgical site was statistically longer in the perineural group (10.9 h) and the systemic group (9.8 h) compared to the placebo group (6.7 h). They concluded the IV route to be noninferior to the perineural route of administration in prolonging the duration of the nerve block. The trial used wide noninferiority limits and was not primarily powered to show noninferiority. They may have used too low a dose of dexmedetomidine (0.5 μg/kg) in order to produce an optimal prolongation of a nerve block peripherally, which would be expected from recent dose-finding studies in healthy volunteers¹³  and patients.¹⁸  Patients registered when they regained normal strength of the arm as a measure of motor block. There was no difference of this parameter between groups. The results presented in that clinical study, with a prolongation of time to first pain and no effect on motor block, may, in theory, represent the systemic analgesic effects of dexmedetomidine,¹²  with no effect on the nerve block.

The mechanism of how systemic administration of dexmedetomidine prolongs the duration of a nerve block is not fully determined. Although dexmedetomidine has central α2-mediated analgesic effects, an animal trial showed that the effect of dexmedetomidine was caused by blockade of the hyperpolarization-activated cation current peripherally and not by its central or peripheral α₁- or α₂-agonistic properties.⁹  While our study supports the conception that the effect of perineural dexmedetomidine is primarily peripheral, the prolongation of the nerve block was only modest when controlling for a systemic effect.

Besides efficacy, adverse events and hemodynamic safety should be considered when deciding whether to administer dexmedetomidine perineurally or systemically.

We report a case of transient nerve damage. We employed an ultrasound approach to the saphenous nerve block, using a slow, low injection pressure technique, and the participant reported no pain on needle placement or during injection. We are unable to determine whether this nerve affection is related to toxicity of dexmedetomidine or needle damage. A follow-up study after adductor canal block did not find any evidence of nerve damage in 97 patients.¹⁹  In an animal study, no nerve damage or perineural inflammation was observed at 24 h and 14 days, respectively, after nerve blocks with dexmedetomidine.¹⁶  Two participants reported paresthesia lasting for 72 h in a volunteer study using a dexmedetomidine dose of 150 µg (38 µg/ml).¹³  No overrepresentation of nerve injury has been reported in clinical studies using dexmedetomidine.^11,17 

The hemodynamic changes observed in our study, with a decrease in blood pressure and bradycardia among healthy volunteers, were easily managed but may pose a problem for patients with cardiac disease. In the study by Abdallah et al.,¹¹  there was no advantage by either mode of administration—IV or perineural—with regard to hemodynamic stability. However, the trial was not powered to investigate these side effects. The sedation observed could present a challenge in some patients but may also prove beneficial as some patients prefer receiving sedation during surgery utilizing peripheral nerve blocks as the sole anesthetic. The question is whether the magnitude of the difference in the duration of the nerve block between the two modes of administration is large enough to warrant off-label perineural use of dexmedetomidine. This question applies to all adjuvants as none are presently approved for perineural use by the Food and Drug Administration or European Medicines Agency.

The strength of our trial design is the control of the systemic contribution of dexmedetomidine, allowing us to investigate whether dexmedetomidine exerts peripheral effects or not. However, the design makes us unable to conclude on any systemic effect of dexmedetomidine, which would have required a group receiving ropivacaine alone (without the systemic contribution of dexmedetomidine from the contralateral side). A pure ropivacaine group was employed in the trials by Abdallah et al.¹¹  and Marhofer et al.,¹⁰  enabling them to comment on the magnitude of the systemic effects. Another limitation of our trial is that we used an almost exclusively sensory block, and therefore, we cannot conclude on the effects on motor block. We may have diluted the dexmedetomidine too much as we added 100 µg dexmedetomidine to 20 ml ropivacaine in a saphenous nerve block, yielding a perineural concentration of 5 µg/ml dexmedetomidine. It is unknown whether it is the perineural dose or the concentration of dexmedetomidine that determines its peripheral effects. Several trials found a prolongation of a nerve block using lower concentrations of dexmedetomidine than we did.^11,18,20,21  A dose-dependent increase in the duration of a sciatic nerve block when adding dexmedetomidine to ropivacaine in rats was demonstrated by Brummet et al.²²  Keplinger et al.¹³  performed a pharmacodynamic dose-ranging study and found that dexmedetomidine prolonged the duration of an ulnar nerve block in a dose-dependent manner, with a ceiling effect around 100 µg at a perineural concentration of 25 µg/ml. We cannot exclude the possibility that we would see a longer prolongation had our concentration of dexmedetomidine been higher. If this is the case and if there is a dose-dependent effect on side effects, then dexmedetomidine may be most applicable in low-volume nerve blocks. Whether the peripheral mechanism of action observed when adding dexmedetomidine to ropivacaine can be generalized to other local anesthetics is unknown.

In conclusion, dexmedetomidine prolongs the duration of a saphenous nerve block by a peripheral mechanism, when controlling for possible systemic effects, but not to our a priori defined minimal relevant difference of 4 h.

The authors express their gratitude to the nurses at the Postanesthesia Care Unit, University Hospital Zealand, Køge, Denmark, for their invaluable help.

Supported in part by the Regional Research Unit, Region Zealand, Roskilde, Denmark, and Department of Anesthesiology, University Hospital Zealand, Køge, Denmark.

The authors declare no competing interests.

Full protocol available from Dr. Andersen: hessel@dadlnet.dk. Raw data available from Dr. Andersen: hessel@dadlnet.dk.