Clonidine Prolongation of Lidocaine Analgesia after Sciatic Nerve Block in Rats Is Mediated via the Hyperpolarization-activated Cation Current, Not by α-Adrenoreceptors

After Institutional Animal Care and Use Committee approval, experiments were performed on 168 male Sprague-Dawley rats (350–400 g; Sasco, Wilmington, MA). Sciatic nerve block was performed using methods similar to Thalhammer et al.  15Animals were briefly anesthetized with 1.2% isoflurane in oxygen and placed in lateral recumbency. A 16 mm long 25-gauge hypodermic needle, with a prefilled 1 ml tuberculin syringe attached, was inserted percutaneously into the sciatic notch between the greater trochanter and the ischial tuberosity, pointing toward the ischium. Stimulating pulses (0.3 v, 2 ms, 2 Hz, negative polarity) were delivered via  a clip lead attached to the top of the needle, and the syringe/needle was advanced until a vigorous ipsilateral hindleg kick was observed. Then, 0.5 ml test drug solution was injected over 5 s and the needle was removed after an additional 5 s.

Before drug injection, animals were placed on a metal grid and sensory response was evaluated by observation of foot withdrawal when a pin was applied to the plantar midline surface of the ipsilateral hind paw (tibial nerve distribution). Any foot withdrawal, whether sciatic or femoral nerve mediated, was considered a response. Motor response was evaluated by observation of the toe-spreading reflex (sciatic nerve motor fibers), a vestibular reflex induced by lifting the rat and generating a response with the toes extended and spread. For both tests, responses were rated as present or absent.

For the single injection studies, baseline sensory (pin) and motor (toe spread) responses were recorded and 0.5 ml drug was injected into the sciatic notch. Animals were then reevaluated at 10-min intervals for the absence or presence of sensory and motor responses. Because the animals are anesthetized for injection, the initial 10-min evaluation cannot be a precise indicator of the actual onset time, and so the duration of nerve block is calculated as the latency from the end of the injection to the return of response. Drug solutions consisted of 1% lidocaine (commercial preservative-free solution intended for nerve block; Xylocaine-MPF, AstraZeneca, Wilmington, DE) with or without added clonidine hydrochloride or epinephrine hydrochloride at concentrations of 0–25 μg/ml. In control experiments, clonidine or epinephrine alone at 25 μg/ml was injected into the sciatic notch. Animals were also monitored for overt sedation at 30 min postinjection by testing ability to ambulate on a rotating rod (10 rpm, 180 s). Clonidine or epinephrine were also injected intraperitoneally in 0.5 ml volume, 5 min before the sciatic notch injection of lidocaine in some experiments, to determine if clonidine or epinephrine were acting at a distal site (in addition to the sciatic nerve site) reached by systemic absorption. The selective α¹-adrenergic antagonist prazosin (1 or 2 mg/kg) or the selective α²-adrenergic antagonist yohimbine (1, 2, or 4 mg/kg) were injected intraperitoneally 5 min before the sciatic notch injection of lidocaine plus 2.5 μg/ml clonidine to ascertain if receptors outside of the sciatic nerve bed contribute to clonidine modulation of lidocaine nerve block. The drug solutions were prepared by one of the investigators (J.S.K.), and the personnel (D.R.B., J.E.T., D.E.W.) performing the drug injections, behavioral assessments, data recording, and group statistics were blinded to the drug identities.

For adrenoreceptor antagonist studies, animals were first injected into the sciatic notch (0.5 ml) with prazosin hydrochloride (up to 40 μg/ml) or yohimbine hydrochloride (up to 100 μg/ml). The second sciatic notch injection of lidocaine with clonidine or epinephrine was administered 10 min later. In one additional experiment, the time between the first and second injections was increased to 60 min. All lidocaine alone animals for these experiments had a saline preinjection. In control experiments, prazosin 40 μg/ml or yohimbine 100 μg/ml alone was injected into the sciatic notch.

For I^hchannel studies, the I^hblocker ZD 7288 (14, 56, 280, or 650 μg/ml) was first injected into the sciatic notch (0.5 ml), and 60 min later the second sciatic notch injection of lidocaine with or without clonidine was performed to determine the role of I^hchannel blockade in extending the duration of nerve block. A 60-min delay was chosen between the first and second injections because even with desheathed nerve preparations at least 20 min of perfusion for ZD 7288 is required to effect I^hand the duration of action can be 2 or more hours.10,14As the adenylyl cyclase activator forskolin and the cell-permeable cAMP analog 8-Br-cAMP have been shown to enhance I^hin nerve cells in vitro ,16forskolin (21 or 63 μg/ml) or 8-Br-cAMP (1.28 or 4.3 mg/ml) were injected into the sciatic notch and 10 min later the second sciatic notch injection of lidocaine with or without clonidine was performed to determine if I^henhancement could decrease the duration of nerve block.

For all multiple injection experiments, duration of nerve block was timed to the end of the second injection. ZD 7288 was obtained from Tocris (Ballwin, MO) and all other drugs were from Sigma (St. Louis, MO). All drugs were directly dissolved in 0.9% sodium chloride injection, except forskolin, which was initially solubilized in dimethylsulfoxide, and then diluted in saline (final dimethylsulfoxide concentration 2%).

Animals received drug injections on either side at 5–7 day intervals, alternating between the left and right leg on different days. Multiple injections administered days apart were allowed because a previous study had demonstrated that a second lidocaine injection onto the sciatic nerve 24 h after the first injection produces an identical functional block.17In addition, in a previous study in which 1% lidocaine was infused into a silicone rubber cuff surrounding the rat sciatic nerve three times a day for 3 days, we found no nerve fiber degeneration, no damage to underlying muscle, and no loss of motor function.18 

In the dose–response studies, duration of nerve block at different added clonidine or epinephrine doses was compared by analysis of variance with post hoc  comparison to the lidocaine alone group with Dunnett’s test. Student t  test was used for comparison of nerve block duration after sciatic notch injection of added clonidine or epinephrine with duration after systemic injection of the same dose. In the receptor antagonist and I^hmodulator experiments, the sensory block durations attributable to different drug combinations were compared by analysis of variance, followed by post hoc  comparison with the Tukey-B test. All data in figures are shown as mean ± SEM. Statistical analysis was performed by one investigator (J.S.K.).

The dose response curve for duration of nerve block with clonidine added to 1% lidocaine is shown in figure 1. In these experiments, the duration of sensory block for 1% lidocaine alone was 69 ± 2 min, and the duration of motor block was 72 ± 2 min. For comparative purposes, the dose–response curve for duration of nerve block with epinephrine added to 1% lidocaine is shown in figure 2. In these experiments, the duration of sensory (pinprick withdrawal) block for 1% lidocaine alone was 68 ± 3 min, and the duration of motor (toe-spreading reflex) block was 71 ± 2 min. With either drug the dose–response curve is monotonic, with the higher dose always producing a longer duration of blockade. None of the drug combinations affected the toe-spreading reflex in the contralateral limb. Because there was no difference between the duration of block using the pinprick test versus  the toe-spreading reflex, for either lidocaine/clonidine (fig. 1) or lidocaine/epinephrine (fig. 2), only pinprick testing results are subsequently displayed for the α-adrenoreceptor and the I^hmodulator studies.

Sciatic notch injection of clonidine or epinephrine alone at 25 μg/ml did not produce measurable sensory or motor blockade, nor did either produce overt sedation (rotating rod test). Systemic injection (intraperitoneal) of 25 μg/ml epinephrine before lidocaine injection did not extend the duration of sensory or motor block. However, systemic clonidine injection at the 25 μg/ml dose extended the duration of motor block by 14 min. At lower doses (15 μg/ml or less) there was no extended motor block with systemic clonidine preinjection. At the 25 μg/ml dose, systemic clonidine injection demonstrated a trend toward increased duration of sensory block (10 min increase, P = 0.067), and at lower doses (15 μg/ml or less) there was no increased duration of sensory block compared with lidocaine alone (intraperitoneal saline preinjection).

Systemic preinjection of prazosin, 1 mg/kg intraperitoneal, did not alter the duration of sensory block of lidocaine plus 2.5 μg/ml clonidine. Preinjection of prazosin, 2 mg/kg intraperitoneal, increased the duration of sensory block (114 ± 6 min with prazosin versus  92 ± 2 min with saline preinjection). Prazosin, 4 mg/kg intraperitoneal, was not tested because that dose decreased ambulation time on the rotating rod. Systemic preinjection of yohimbine, 1, 2, or 4 mg/kg intraperitoneal, did not alter the duration of sensory block of lidocaine plus 2.5 μg/ml clonidine. At 4 mg/kg yohimbine intraperitoneal, the block duration was 99 ± 6 min versus  101 ± 5 min with saline preinjection.

Epinephrine at 1.6 μg/ml increased the sensory block duration by approximately 50% of the maximum effect (seen at the 25 μg/ml dose). Using this intermediate 1.6 μg/ml dose, the α-adrenergic antagonists prazosin or yohimbine were given before the lidocaine/epinephrine combination. Prazosin at 10 μg/ml reduced the duration of sensory block of the lidocaine/epinephrine combination to that of lidocaine alone (fig. 3). Yohimbine at 40 μg/ml had no effect on the duration of the lidocaine/epinephrine induced sensory block.

Clonidine at 2.5 μg/ml increased the sensory block duration by approximately 50% of the maximum effect (seen at the 25 μg/ml dose). Using this intermediate 2.5 μg/ml dose, the α-adrenergic antagonists prazosin or yohimbine were given before the lidocaine/epinephrine combination. Neither prazosin at 40 μg/ml nor yohimbine up to 100 μg/ml altered the duration of the sensory block of the lidocaine/clonidine combination (fig. 4). The same result was obtained when the time between the first and second injections was increased to 60 min or with a 15 μg/ml dose of clonidine. Sciatic notch injection of prazosin or yohimbine alone at these same doses did not produce any sensory or motor blockade, nor did either produce overt sedation.

When the I^hblocker ZD 7288 was injected into the sciatic notch at 650 μg/ml, the duration of lidocaine sensory block increased by 39 min (fig. 5). A similar increase in block (35 min) could be obtained with the addition of 15 μg/ml clonidine to lidocaine. Preinjection of ZD 7288 before that same lidocaine/clonidine combination did not produce any additional prolongation of sensory block duration compared with the lidocaine/clonidine combination alone. A lower dose of ZD 7288 (280 μg/ml) followed by lidocaine produced a smaller increase in lidocaine sensory block (20 min). The lowest doses of ZD 7288 (14 or 56 μg/ml) did not prolong lidocaine nerve block duration. Preinjection of forskolin at 63 μg/ml (150 μm) decreased the duration of lidocaine sensory nerve block by 19 min (fig. 6A). When forskolin at 63 μg/ml was given before the combination of lidocaine and 15 μg/ml clonidine, the duration of sensory nerve block was decreased by 54 min. A lower dose of forskolin (21 μg/ml) followed by lidocaine also reduced the duration of sensory block (13 min). Preinjection of 8-Br-cAMP at 4.3 mg/ml (10 mm) decreased the duration of lidocaine sensory block by 22 min (fig. 6B). When 8-Br-cAMP at 4.3 mg/ml was given before the combination of lidocaine and 15 μg/ml clonidine, the duration of sensory nerve block was decreased by 34 min. A lower dose of 8-Br-cAMP (1.28 mg/ml) followed by lidocaine did not reduce the duration of lidocaine sensory block, although that dose still decreased the duration of lidocaine/clonidine block by 34 min.

In the animal model, the dose response curve for clonidine or epinephrine added to 1% lidocaine did not demonstrate any U-shaped characteristics in which higher doses would be less effective than lower doses in producing nerve block. Both dose–response curves represent monotonic functions. Our animal study does not provide an explanation for the reported finding that lower doses of clonidine added to lidocaine could be more effective in patients than higher doses of clonidine under certain conditions.4,5 

Although epinephrine is slightly more potent than clonidine, no direct comparison is made between the two curves, as our study demonstrated that the two compounds act by different mechanisms. Notably, however, the epinephrine dose–response curve is similar to that measured in humans when 1% lidocaine was infiltrated subcutaneously with epinephrine in four doses from 0.3–20 μg/ml, and the half-maximum response (pinprick) was observed at a concentration of about 1.25 μg/ml.19 

Sciatic nerve injections of 25 μg/ml clonidine alone did not produce sensory or motor block. This is consistent with in vitro  experiments that demonstrate a local anesthetic effect of clonidine only at high doses (500 μm = 134 μg/ml)8,9and an in vivo  sciatic notch injection experiment showing that 1000 μm clonidine alone did not cause detectable nerve block using the hot plate test.20One clinical study also showed that local application of clonidine alone (150 μg in 15 ml) onto the brachial plexus did not produce a local anesthetic effect.21Although 25 μg/ml clonidine alone did not produce overt sedation, it may still have effects in the central nervous system related to systemic absorption. This is evidenced by the increased duration of motor block and the slight increase in sensory block with intraperitoneal clonidine combined with lidocaine, at the highest dose. The toe-spreading reflex is initiated by vestibular input, and higher doses of clonidine may have sedative effects in the brain.6,22In patients, doses of 150 μg or greater in nerve block mixtures produce significant sedation.2,4Therefore, the effects of large doses of clonidine on nerve block can be attributable to both local and central actions. Because systemic doses of clonidine 15 μg/ml or lower did not extend the duration of lidocaine nerve block, central nervous system inhibition does not appear to be an important factor over most of the useful clonidine dose range. This is supported by antagonist experiments showing that systemically administered α-adrenoreceptor antagonists did not reduce clonidine prolongation of lidocaine nerve block. The increase in lidocaine/clonidine sensory nerve block duration with 2 mg/kg prazosin preinjection may be attributable to mild sedation or motor impairment at that prazosin dose because other rodent studies have shown that prazosin, 2 mg/kg, decreases locomotion,23and our 4 mg/kg dose caused impairment on the rotating rod. In the dose–response curves, there was no ambiguity with epinephrine at higher doses, as intraperitoneal epinephrine at 25 μg/ml did not extend the duration of lidocaine sensory or motor block.

It is widely believed that the effect of epinephrine on extending the duration of lidocaine nerve block is the result of vasoconstriction,24,25and our study is in accord with that mechanism. The α¹-adrenergic antagonist prazosin completely blocked the increased duration of analgesia and motor block obtained when epinephrine was added to lidocaine in the local injection mixture. Interestingly, the α²-adrenergic antagonist yohimbine was without measurable effect, although α²-adrenergic receptors are mediators of vasoconstriction in some blood vessels.26,27 

Although clonidine could have local vasoconstrictive effects that might slow absorption of lidocaine, one clinical study did not show any prolongation of plasma lidocaine concentration when a clonidine/lidocaine admixture was used for brachial plexus block, whereas the epinephrine/lidocaine combination did produce this effect.2However, in another study with blockade of the human dorsal cutaneous nerves, the tissue lidocaine concentration was increased when 10 μg/ml clonidine was added to 1% lidocaine compared to lidocaine alone, suggesting that the prolongation of nerve block was at least partially pharmacokinetically mediated.7Nevertheless, we did not find any clonidine-induced prolongation of nerve block that could be attributed to α-adrenergic receptors. Although we did not examine the complete displacement of the dose–response curves with the α-adrenergic antagonists, the absence of any effect at an intermediate dose (2.5 μg/ml) or a high dose (15 μg/ml) makes it unlikely that our methodology was insufficiently sensitive to detect α-adrenergic interaction. It is possible that the deep sciatic nerve site in the animal study differs from the superficial tissue bed of human cutaneous nerves with respect to blood vessels subject to vasoconstriction24or that there may be a species difference in α-adrenoreceptor distribution. Another consideration is that peripheral nerve block involves a balance between the amount of drug outside the peripheral nerve and that within the perineurial space,28and this balance may also vary among different injection sites.

In vitro  studies with desheathed peripheral nerve suggest that I^hcurrents, which are among the determinants of nerve after-potentials, may explain how clonidine potentiates local anesthetic action. The I^hblocker ZD 7288 produces slowing of conduction in rat C-fibers.14When ZD 7288 at 10 μm (3 μg/ml) is applied to the isolated vagus nerve, activity-dependent hyperpolarization is enhanced, presumably by blocking the depolarizing effect of I^h.10Clonidine 50 μm (13 μg/ml) produces a similar enhancement of activity-dependent hyperpolarization. However, combining the two drugs did not produce any additional hyperpolarization, implying that clonidine enhances this hyperpolarization by inhibiting I^h.10We obtained the same combination of effects with the in vivo  animal model. The only quantitative difference was that the concentration of ZD 7288 used in our study was 650 μg/ml. The likely reason for this difference is that ZD 7288 is a relatively hydrophilic compound,13and therefore a much higher concentration is required to allow penetration through the perineurium and access to nerve fibers, unlike the in vitro  experiments with desheathed nerve. The 15 μg/ml clonidine dose was used in this experiment, rather than a 25 μg/ml dose, as the systemic injection experiments showed that at the higher dose, part of the effect of clonidine may be attributable to central effects not directly related to the peripheral nerve activity. Compounds that stimulate adenylyl cyclase (forskolin) or are cAMP analogs (8-Br-cAMP) have been shown in vitro  to enhance I^hin neurons16,29or in dorsal root ganglion cells.30It is not known if these compounds have the same effect in vitro  on peripheral nerve. The decreased duration of nerve block seen in our experiments when either compound was injected before lidocaine/clonidine is consistent with the hypothesis that in vivo  nerve blockade is modulated via  I^hchannel activity. Further confirmation of these results with electrophysiological studies would reinforce this hypothesis.

The exact doses of clonidine used in the dose–response part of our rat study cannot be precisely compared with those used in clinical studies because the injection volume and site and amount of perfused tissue are different. However, the amount of added clonidine (2.5 μg/ml) required to produce a moderate enhancement of sensory nerve block duration is consistent with the sensitivity seen in clinical studies.2,4,5,31One limitation of our study is that animals received more than one sciatic nerve injection (5–7 days apart), and so it is possible that there was some carryover effect of lidocaine or clonidine not anticipated by the authors. Also, as a result of the use of electrical stimulation for nerve location, the animals had to be briefly (<2 min) anesthetized during performance of the nerve block procedure. However, it has been shown that there is no difference in duration of sciatic nerve blockade performed under brief general anesthesia versus  when awake, in animals injected with a local anesthetic (bupivacaine).32Clonidine also binds to imidazoline receptors in the brainstem and has antagonist actions at the I¹receptor subtype.33However, no I¹selective antagonists are available.34When such compounds become available, it would be worthwhile to perform additional antagonist studies to examine if the I¹receptor can also mediate lidocaine/clonidine analgesia in a manner similar to the imidazoline receptor mediated antihypertensive effect.33 

Our study is also in accord with clinical studies showing no advantage to using clonidine instead of epinephrine in lidocaine admixtures with respect to duration of nerve block.2,35However, in patients with hypertension, cardiovascular disease, or hyperthyroidism,2or where there is concern about local ischemia as a result of vasoconstriction, clonidine can be an effective alternative to epinephrine.

In conclusion, our results suggest that I^hchannel blockade, and not α-adrenoreceptor activity, contributes to clonidine enhancement of in vivo  lidocaine block of the sciatic nerve.