![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 285, Issue 2, 413-421, May 1998
Departments of Anesthesiology, Anatomy, Medicine, and Oral Surgery and Division of Neuroscience, University of California, San Francisco, California
 |
Abstract |
|---|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Evidence from both clinical studies and animal models suggests that the local anesthetic, lidocaine, is neurotoxic. However, the mechanism of lidocaine-induced toxicity is unknown. To test the hypothesis that toxicity results from a direct action of lidocaine on sensory neurons we performed in vitro histological, electrophysiological and fluorometrical experiments on isolated dorsal root ganglion (DRG) neurons from the adult rat. We observed lidocaine-induced neuronal death after a 4-min exposure of DRG neurons to lidocaine concentrations as low as 30 mM. Consistent with an excitotoxic mechanism of neurotoxicity, lidocaine depolarized DRG neurons at concentrations that induced cell death (EC50 = 14 mM). This depolarization occurred even though voltage-gated sodium currents and action potentials were blocked effectively at much lower concentrations. (EC50 values for lidocaine-induced block of tetrodotoxin-sensitive and -resistant voltage-gated sodium currents were 41 and 101 µM, respectively.) At concentrations similar to those that induced neurotoxicity and depolarization, lidocaine also induced an increase in the concentration of intracellular Ca++ ions ([Ca++]i; EC50 = 21 mM) via Ca++ influx through the plasma membrane as well as release of Ca++ from intracellular stores. Finally, lidocaine-induced neurotoxicity was attenuated significantly when lidocaine was applied in the presence of nominally Ca++-free bath solution to DRG neurons preloaded with 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA). Our results indicate: 1) that lidocaine is neurotoxic to sensory neurons; 2) that toxicity results from a direct action on sensory neurons; and 3) that a lidocaine-induced increase in intracellular Ca++ is a mechanism of lidocaine-induced neuronal toxicity.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Recent
reports of neurologic injury after continuous spinal anesthesia suggest
that clinically relevant concentrations of local anesthetics such as
lidocaine can cause nerve injury (Auroy et al., 1997
; Rigler
et al., 1991). Specifically, there is evidence that Cauda
Equina syndrome, characterized by incontinence, perineal sensory loss
and motor weakness in the legs, is the result of neurotoxicity
associated with the spinal administration of local anesthetics (Rigler
et al., 1991).
Studies of local anesthetic-induced nerve injury have eliminated
several possible mechanisms of toxicity. For example, glucose, included
in the local anesthetic solution to control the spread of spinally
administered agents, is not neurotoxic (Sakura et al.,
1995b), which suggests that toxicity results from an action of the
local anesthetic itself. In addition, Lambert and colleagues (1994)
demonstrated that irreversible conduction block induced by lidocaine in
an isolated frog nerve is not the result of either a failure to wash
the local anesthetic from the nerve, or a breakdown of the nerve
membrane integrity. Finally, Sakura and colleagues (1995a) concluded
that local anesthetic-induced neurotoxicity in vivo does not
result from a blockade of voltage-gated sodium channels.
The aim of the present study was to determine whether lidocaine has direct neurotoxic effects on sensory neurons and to investigate the possible involvement of Ca++ in this effect. To address this aim, we performed in vitro histological, electrophysiological and ratiometric fluorometrical experiments on isolated sensory neurons from the adult rat. We observed lidocaine-induced neuronal death after exposure of cultured DRG neurons to lidocaine concentrations as low as 30 mM for as little as 4 min. In addition, we provide evidence which indicates that lidocaine-induced increase in the intracellular concentration of free Ca++ ion ([Ca++]i) contributes to lidocaine-induced neuronal death.
| |
Methods |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Cell culture.
Primary cultures of dissociated adult rat DRG
neurons were prepared by methods described previously (Gold et
al., 1996a). Male Sprague-Dawley rats (150-250 g, Bantin and
Kingman, Fremont CA) were anesthetized deeply with an i.p. injection of
Na-pentobarbital (60 mg/kg); lumbar
(L1-L6). DRG were removed,
and rats subsequently were sacrificed by an overdose of
Na-pentobarbital. DRGs were desheathed in ice-cold MEM-BS composed of
90% minimal essential medium (MEM; Gibco BRL, Gaithersburg, MD), 10%
heat-inactivated fetal bovine serum and 1000 U/ml each of penicillin
and streptomycin. DRGs were then incubated 120 min at 37°C in MEM-BS,
to which collagenase P (Boehringer Mannheim, Indianapolis, IN) was
added to a final concentration of 0.125%. DRGs then were incubated 10 min at 37°C in Ca++- and
Mg++-free Hanks' balanced salt solution
containing (in g/l): glucose, 1; KCl, 0.4; NaCl, 8.0;
KH2PO4, 0.06;
Na2HPO4·7H2O,
0.09; Phenol Red, 0.01; NaHCO3, 0.35, and 0.25%
trypsin (Worthington, Bristol, UK) and 0.025%
ethylenediaminetetraacetic acid (Sigma, St Louis, MO). Then trypsin
activity was inhibited by the addition of MEM-BS containing 0.125%
MgSO4 and 80 µg/ml soybean trypsin inhibitor (Sigma), and DRGs were dissociated by trituration with a fire-polished Pasteur pipette. DRG cells were plated onto glass coverslips, previously coated by a solution of 5 µg/ml mouse laminin (Gibco BRL)
and 0.1 mg/ml poly-DL-ornithine (Sigma). The cells were
incubated in MEM-BS at 37°C, 3% CO2 and 90%
humidity, and were studied between 6 and 24 hr after plating.
Cell survival. Trypan blue exclusion was used to determine the effects of lidocaine on DRG neuron survival. After 15 to 20 hr incubation, coverslips were placed in chambers (500 µl volume) continuously perfused with normal bath solution (table 1, BS 1) or nominally Ca++-free bath solution (table 1, BS 3). The perfusion solution was changed to a solution containing lidocaine (1-100 mM) or choline (30-100 mM) (table 1, BS 1 or BS 3 containing lidocaine or choline) 3 to 5 min later. After exposure to test solutions, the perfusion solution was changed again to normal bath solution and the coverslips were perfused continuously for 1 hr. Coverslips were then placed in trypan blue solution (0.1-0.4%) for 10 min, briefly rinsed in normal bath solution and then placed in chilled 4% paraformaldehyde (in PBS) for 30 min at 4°C. Finally, coverslips were rinsed in PBS, dehydrated and mounted on glass slides. All solutions used before fixation were at room temperature.
|
). A 1-hr wash period between lidocaine exposure and trypan blue staining was used to ensure
that trypan blue staining reflected an irreversible decrease in cell
viability rather than a transient decrease in plasma membrane integrity.
Electrophysiology. To study effects of lidocaine on the electrophysiological properties of DRG neurons, conventional whole-cell voltage-clamp techniques were used. Patch pipettes filled with electrode solution (ES, table 1) had resistance of 1 to 4 megohm. The recording chamber (500-µl volume) was perfused continuously (1-2 ml/min) with bath solution. All experiments were performed at room temperature (21-24°C).
For current-clamp experiments an electrode solution with 140 mM KCl was used (table 1, ES 1) in the presence of normal bath solution (table 1, BS 1). To record voltage-gated Na+ currents in isolation, KCl in the electrode solution was replaced by CsCl (table 1, ES 2). To maintain control of the membrane potential while recording voltage-gated Na+ currents, the concentration of Na+ in the bath solution was reduced (table 1, BS 2). All components of the bath and electrode solutions were obtained from Sigma (St. Louis, MO). Cells were voltage- or current-clamped by use of an Axopatch 1B amplifier (Axon Instruments, Foster City, CA). Data were filtered with a 4-pole Bessel filter (see figure legends) and digitized. Series resistance was estimated from the settling rate of the voltage clamp and the membrane capacitance. Estimates for series resistance ranged between 0.3 and 5 megohm. Series resistance was compensated (>80%) by use of amplifier circuitry. Data obtained from neurons in which uncompensated series resistance resulted in voltage-clamp errors greater than 5 mV were discarded from further analysis. A P/4 protocol was used for leak subtraction; to ensure the linearity of leakage currents around the holding potential for the P/4 protocol, the resting conductance for each cell was determined at potentials ranging from
100 to 60 mV. Data were fit by a nonlinear least-square method.
Dose-response data were fit by a Michaelis-Menten equation that related
fractional inhibition of current, F, to drug concentration: F = [1 (Idrug/Icontrol)] = [(Fmax × Dn)/(Dn + EC50n)], characterized by
EC50 (concentration of drug producing a
half-maximal effect), Fmax (maximum
fraction of inhibition) and n, a Hill coefficient used to
account for possible cooperativity. Idrug = the peak current in the presence of lidocaine;
Icontrol = the peak current in the absence
any of drug; and D = the lidocaine concentration. A
similar equation was fit to the dose-response relationship of
lidocaine-induced membrane depolarization. To compare the effects of
different concentrations of lidocaine on the resting membrane potential
of DRG neurons, we calculated the fractional change in membrane
potential induced by each concentration of lidocaine in each neuron.
The fractional change in membrane potential was determined by dividing
the difference between the membrane potential at the start of the
experiment (Vrest) and the membrane
potential in the presence of a given concentration of lidocaine
(Vlido) by the difference between
Vrest and the membrane potential in the
presence of 100 mM lidocaine (V100 lido):
fractional depolarization (Fd) at each
concentration of lidocaine = (Vrest Vlido)/(Vrest V100 lido). The equation used to fit the
dose dependence of depolarization was Fd = Dn/(Dn + EC50n).
Ratiometric fluorimetry. After 15 to 20 hr in culture, cells were loaded 10 min with 2.5 µM Fura-2 (with 0.025% pluronic acid, Molecular Probes), after which cells were placed in normal bath solution (table 1) for 10 min and then placed in a chamber continuously perfused at 2 ml/min with normal bath solution (table 1, BS 1). Fura-2 fluorescence was measured with an ICCD camera (Solamere Technology, Salt Lake City UT). Data were analyzed and presented as a Fura-2 fluorescence ratio, because this ratio is directly proportional to [Ca++]i for concentrations of Ca++ less than 10% of the saturating concentration for Fura-2 (~39 µM). We were able to readily detect increases in [Ca++]i, as demonstrated by the increase in the Fura-2 fluorescence ratio in response to the bath application of lidocaine and 50 mM K+. In addition, we were able to readily detect decreases in [Ca++]i, as demonstrated by a decrease in the fluorescence ratio in response to the application of nominally Ca++-free bath solution (table 1, BS 3). Therefore, the resting [Ca++]i within the DRG neurons we have studied is neither close to the saturating concentration for Fura-2, nor below the threshold for which increase in [Ca++]i may be detected (~5 nM in our system).
Statistics. Data are presented as a mean ± standard error of the mean (S.E.M.). A one-way analysis of variance was used to determine the presence of statistically significant differences between groups. When a significant group effect was detected, post hoc comparisons were performed with a Scheffe test.
| |
Results |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Cell death. A 15-min exposure to lidocaine induced death of cultured DRG neurons as assayed by failure to exclude trypan blue (see fig. 1). This effect was dose-dependent, increasing dramatically at lidocaine concentrations greater than 10 mM. The number of neurons stained with trypan blue 1 hr after a 15-min exposure to 30 mM lidocaine was larger than the number of neurons stained after a 15-min exposure to 100 mM lidocaine (fig. 1A). In contrast to the dose-dependent increase in the number of neurons killed after a 15-min exposure to lidocaine, when data are analyzed for neurons killed as a percentage of the total neurons counted (fig. 1B), the percentage of neurons killed after exposure to 100 mM lidocaine was larger than that observed after exposure to 30 mM lidocaine. The difference between the dose dependence of the number of neurons killed versus percentage of neurons killed reflects the fact that there was a dose-dependent decrease in the number of neurons remaining on the coverslip.
|
Lidocaine depolarizes DRG neurons. Many neurotoxic agents are excitotoxins, i.e., they damage neurons by causing excessive excitation. To determine whether lidocaine can exert direct effects on DRG neurons that might underlie an excitotoxic action, we tested the hypothesis that lidocaine depolarizes DRG neurons. We observed that lidocaine, at concentrations greater than 3 mM, depolarizes DRG neurons (fig. 2, A, B and C). This effect is dose-dependent, with an EC50 of 14 mM (fig. 2C). The DRG neurons exhibited a maximally depolarized potential beyond which lidocaine could not drive the membrane potential, regardless of drug concentration or duration of application. This potential was +3 ± 2.1 mV (n = 16). Neurons depolarized to near +3 mV by high concentrations (>30 mM) of lidocaine failed to repolarize within 10 min after washout of the lidocaine (observed in 11 of 11 neurons tested).
|
54.6 ± 1.3 mV) of these five neurons previously exposed to 100 mM lidocaine was not significantly different from the resting potential (54 ± 2.3 mV) of "naïve"
neurons (i.e., those never exposed to lidocaine, P > .05). In addition, it was possible to evoke a "normal" action
potential from each of these five neurons. Furthermore, each of these
five neurons subsequently was depolarized irreversibly by the
application of 100 mM lidocaine. These observations suggest that the
apparent irreversibility of the lidocaine-induced depolarization might be due to an interaction between lidocaine-induced effects and changes
in the neuron caused by dialysis of the cytoplasm by the whole-cell
patch-clamp electrode.
Lidocaine-induces block of voltage-gated
Na+ current and the somatic action
potential.
Our observation of an excitatory effect of lidocaine on
DRG neurons contrasts with the conventional view of local anesthetics as inhibitory agents that act primarily to reduce voltage-gated sodium
currents. To compare the excitatory and inhibitory actions of this
local anesthetic, we performed voltage-clamp recordings to measure
quantitatively the effect of lidocaine on the two principal voltage-gated sodium currents in DRG neurons (fig.
3A). TTX-S INa is a
rapidly activating and inactivating low-threshold current blocked by
nanomolar concentrations of tetrodotoxin, whereas TTX-R INa is a more slowly activating and inactivating
high-threshold current resistant to micromolar concentrations of TTX.
Currents were evoked from a potential of 100 mV to remove
steady-state inactivation (fig. 3A; Gold et al., 1996b).
Both TTX-S INa and TTX-R
INa were blocked by lidocaine in a dose-dependent
fashion (fig. 3, B and C) with half-maximal concentrations equal to 41 and 101 µM, respectively (fig. 3C). These results are similar to
previously reported values (Roy and Narahashi, 1992). The application of 1 mM lidocaine, a concentration that completely blocked
voltage-gated Na+ current also completely blocked
the somatic action potential (fig. 3C, insert). The effects of 1 mM
lidocaine on voltage-gated sodium currents and somatic action
potentials were fully and rapidly reversible (fig. 3B). The
concentration of lidocaine required to completely block voltage-gated
Na+ currents (1 mM) is more than an order of
magnitude lower than the concentration at which we begin to see
lidocaine-induced cell death.
|
Lidocaine stimulates Ca++ influx and
mobilization of internal stores in DRG neurons.
Because DRG
neurons were depolarized by lidocaine at concentrations similar to
those at which we observed lidocaine-induced cell death, and because
Ca++-dependent mechanisms contribute to
excitotoxic cell death in a variety of neurotoxic models (Orrenius and
Nicotera, 1994), it is possible that an increase in intracellular
Ca++ concentration
([Ca++]i) contributes to
lidocaine-induced cell death. Furthermore, it is possible that the
source of the increase in
[Ca++]i was extracellular
Ca++ possibly entering neurons via
voltage-gated Ca++ channels activated by the
lidocaine-induced depolarization of the plasma membrane. To investigate
these possibilities we determined whether lidocaine-induced cell death
could be diminished by exposing DRG cells to 30 mM lidocaine in the
presence of nominally Ca++-free bath solution. No
significant increase in cell survival was observed when lidocaine was
applied in the absence of extracellular Ca++
(data not shown).
;
Kutchai et al., 1994), and therefore, also may block an
endoplasmic reticulum Ca++-ATPase, possibly
resulting in an increase of
[Ca++]i via
increased efflux of Ca++ from intracellular
stores. Therefore, we hypothesized that lidocaine toxicity might result
from lidocaine-induced release of Ca++ from
intracellular stores. To test this hypothesis, we used the Ca++-sensitive dye, Fura-2, to determine whether
lidocaine induces an increase in
[Ca++]i via
release from intracellular stores.
Lidocaine dose-dependently increased
[Ca++]i in neurons (fig.
4). Examples of the changes in
[Ca++]i induced by brief
(30-sec) applications of increasing concentrations of lidocaine in two
different neurons is shown in figure 4A. As suggested by the examples
presented in figure 4A, there was variability among DRG neurons in the
sensitivity to lidocaine with respect to the change in
[Ca++]i. This variability
is demonstrated clearly by plotting the fraction of neurons responding
to lidocaine with an increase in
[Ca++]i versus
lidocaine concentration (fig. 4B). A neuron was considered responsive
to lidocaine if the change in
[Ca++]i 30 sec after the
application of lidocaine was more than two standard deviations larger
than the mean change in
[Ca++]i observed during
the same period in the absence of lidocaine. More than 50% of DRG
neurons are responsive to the application of 10 mM lidocaine (fig. 4B).
|
|
|
Increased [Ca++]i contributes to lidocaine-induced neurotoxicity. To further investigate the contribution of the lidocaine-induced increase in [Ca++]i to lidocaine-induced neurotoxicity, lidocaine was applied in the presence of nominally Ca++-free bath solution (table 1, BS 3) to DRG neurons preloaded with the Ca++ buffer, BAPTA (see "Methods"). To confirm that the BAPTA loading protocol resulted in an intracellular concentration of BAPTA sufficient to inhibit the lidocaine-induced increase in [Ca++]i, the change in [Ca++]i was measured in response to 30 mM lidocaine. In neurons (n = 19) preloaded with BAPTA, 30 mM lidocaine induced minimal change in [Ca++]i (fig. 7A).
|
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Our results indicate that application of lidocaine to DRG neurons at concentrations greater than 10 mM causes neuronal death. At a concentration of 30 mM, a 4-min application of lidocaine is sufficient to induce neuronal death. Concentrations of lidocaine inducing neuronal death are more than an order of magnitude greater than those required for complete and reversible blockade of voltage-gated sodium channels and action potential generation. Lidocaine, at concentrations that cause neuronal death, also causes rapid depolarization of the neuronal membrane as well as increase in [Ca++]i that apparently results from Ca++ influx through the plasma membrane in addition to Ca++ release from intracellular stores. Buffering the lidocaine-induced increase in [Ca++]i significantly attenuates lidocaine-induced neuronal death.
Our measurement of lidocaine-induced toxicity likely underestimated the extent of lidocaine-induced cell death because of three factors. First, we only counted neurons that remained attached to the coverslip. One hour after a 15-min exposure to 100 mM lidocaine, the number of neurons remaining on the coverslips was reduced markedly. This loss of neurons likely reflects an increase in cell death rather than a nontoxic lidocaine-induced disruption of cell adhesion, as suggested by our observations during the Ca++ imaging experiments. Loss of neurons from the coverslip was associated with the lysis of the neurons. Second, we performed cell counts relatively soon after exposure to lidocaine; when counts were performed 24 hours after lidocaine exposure, there was almost a complete loss of neurons from the coverslips (data not shown). Third, failure to exclude trypan blue represents rather severe damage to neurons; assays for metabolic processes might have revealed a more widespread but less severe neuronal injury as opposed to death. In future studies, such assays may be particularly useful for identifying underlying mechanisms of toxicity.
Many of our observations support the suggestion that lidocaine-induced toxicity is the result of a direct action of the compound itself, rather than the vehicle or the method of administration. First, because crystalline lidocaine was dissolved directly in HEPES-buffered bath solution in which the pH was adjusted to 7.2, toxicity associated with the acidic nature of a nonbuffered solution was avoided. Second, the high osmolality of a 100 mM lidocaine solution was not the cause of neuronal death, because cell death was not caused by exposure to bath solution containing 100 mM choline. Third, because we performed all experiments in continuously perfused chambers, it is unlikely that cell death was an indirect effect that occurred secondarily to a lidocaine-induced release of cytotoxic compounds from neuronal or non-neuronal cells: any such compounds would be diluted rapidly and removed by the flowing bath solution. Furthermore, when studied within 24 hr of plating (as in the present experiments) the vast majority of DRG neurons in this preparation do not contact each other or non-neuronal cells, thereby avoiding the possibility of direct interactions among the cells.
Our results indicate that lidocaine toxicity is not the result of an
immediate and irreversible breakdown in the integrity of the plasma
membrane, as would be reflected by a rapid and permanent loss in
membrane potential. The mechanism of the lidocaine-induced depolarization currently is unknown, but may reflect the simultaneous blockade of ion channels and pumps responsible for the maintenance of
neuronal resting potential; lidocaine-induced depolarization has been
reported previously by other investigators (Lambert et al.,
1994). Although the axonal membrane repolarizes after wash of lidocaine
(Lambert et al., 1994), an irreversible conduction block was
observed in axons after 3 min exposure of the axon to lidocaine at
concentrations as low as 40 mM (Bainton and Strichartz, 1994). That we
observed recovery of somal action potential after exposure of DRG
neurons to even higher concentrations of lidocaine raises the
possibility that different mechanisms underlie neuronal death and loss
of conduction. However, because an increase in [Ca++]i can disrupt
cytoskeleton (Neely and Gesemann, 1994; Uto et al., 1994;
Waxman et al., 1993), and disruption of cytoskeleton can
decrease excitability (Sakai et al., 1985; Waxman et
al., 1993), a lidocaine-induced increase in
[Ca++]i possibly could
contribute to both cell death (see below) and long-lasting conduction
failure.
Lidocaine reportedly blocks high threshold voltage-gated
Ca++ channels in DRG neurons with an
IC50 of 
2.8 mM (Sugiyama and Muteki, 1994).
Therefore, these channels should be blocked by concentrations of
lidocaine that cause sufficient depolarization to activate them.
Whether lidocaine induces Ca++ entry into DRG
neurons via low-threshold, voltage-gated
Ca++ or via other
Ca++-permeable channels has yet to be determined.
The release of Ca++ from internal stores may
reflect the inhibition of a Ca++-ATPase of
endoplasmic reticulum, as lidocaine has been shown to block a
Ca++-ATPase of the sarcoplasmic reticulum (Karon
et al., 1995; Kutchai et al., 1994). Similar to
the effects of Ca++-ATPase inhibitors such as
thapsigargin, the lidocaine-induced increase in
[Ca++]i is transient, in
some neurons decaying to base line within 15 min, (presumably
reflecting a depletion of internal Ca++ stores),
and is followed by a transient increase in
[Ca++]i, after wash of
lidocaine [presumably reflecting "capacitative Ca++ entry" that may be associated with the
refilling of internal Ca++ stores (Clapham,
1995)]. Although it also is possible that lidocaine uncouples the
mitochondrial electrochemical gradient to release stored
Ca++, it previously has been demonstrated that,
under resting conditions, mitochondria in DRG neurons store very little
releasable Ca++ (Werth and Thayer, 1994).
Our observation that preventing the lidocaine-induced increase in
[Ca++]i significantly
attenuates lidocaine-induced neurotoxicity strongly suggests that the
increase in [Ca++]i is an
underlying mechanism of toxicity. This suggestion is consistent with
the observation made by Bainton and Strichartz (1994) that L-type
Ca++ channel blockers made nerves more resistant
to the toxic effects of lidocaine. However, we cannot exclude the
possibility that the apparent decrease in base-line
[Ca++]i in BAPTA-loaded
neurons exposed to Ca++-free bath (see fig. 7A)
also may have contributed to the decrease in lidocaine-induced
neurotoxicity.
Consistent with the delayed nature of
Ca++-dependent neurotoxicity, for most neurons
lidocaine-induced death was not evident during, or immediately after,
application of lidocaine. For example, after washout of lidocaine, the
membrane potential returned to the original resting potential, and 50 mM K+ was able to evoke an increase in
[Ca++]i in six of seven
neurons 15 min after wash of the lidocaine. In addition, the
lidocaine-induced increase in
[Ca++]i returns to
resting level after wash of lidocaine. Our results suggest that a 4-min
increase in [Ca++]i is
sufficient to induce cell death. That an increase in
[Ca++]i as short as 5 min
is sufficient to induce delayed neuronal death has been well documented
(Manev et al., 1991; Randall and Thayer, 1992). Although the
cellular processes engaged by the lidocaine-induced increase in
[Ca++]i that result in
cell death have yet to be identified, there is evidence that
capsaicin-induced delayed death of DRG neurons depends on the
activation of Ca++-dependent proteases (Chard
et al., 1995).
The incidence of nerve injury resulting from the clinical use of
lidocaine is relatively low, possibly because the concentration of
lidocaine at the neuronal membrane rarely reaches levels necessary to
induce injury. With the usual peripheral routes of administration there
are large barriers to diffusion (Ritchie et al., 1965). Consequently, intraneuronal lidocaine concentration is only 1.6% of
the injected concentration at full block (Popitz-Bergez et al., 1995).
Dilution of the local anesthetic in the relatively large volume of the
subarachnoid space is a major factor contributing to the minimization
of local anesthetic concentrations at the neuronal membrane after
spinal administration (Rigler et al., 1991). Importantly, diffusion barriers are considerably lower for spinal administration of
anesthetics compared with those for peripheral administration, as
suggested by the observation that lidocaine has an
EC50 of 230 µM for inhibition of evoked
action potentials in isolated dorsal roots (Jaffe and Rowe, 1996).
Furthermore, spinally administered local anesthetics often are
distributed nonhomogeneously (Drasner et al., 1994; Robinson
et al., 1994). The maldistribution of anesthetic is a
potential problem that may be exacerbated with the use of hyperbaric
solutions specifically designed to control the distribution of spinally
administered local anesthetics. Therefore, given our results indicating
that brief exposure of neurons to relatively low concentrations of
lidocaine may result in neurotoxicity, it is not unreasonable to
suggest that neurotoxic levels of lidocaine may be approached
clinically with routine spinal administration of this local anesthetic.
Indeed, it has been reported that, 5 min after administration of a 5%
solution of lidocaine, the mean cerebrospinal fluid concentration was
close to 16 mM (Van Zundert et al., 1996). Thus, while
diffusion and dilution may limit the concentration of lidocaine at the
neuronal membrane, and therefore minimize the clinical sequela
resulting from the use of lidocaine, our results begin to define an
upper limit on the therapeutic window governing the use of lidocaine;
under conditions in which barriers to diffusion have been removed or
dilution is limited, lidocaine may become neurotoxic.
Summary and conclusions. We have demonstrated for the first time that lidocaine can act directly on mammalian sensory neurons to cause membrane depolarization, an increase in [Ca++]i, and neurotoxicity at doses relevant to the clinical use of lidocaine. Importantly, we have provided evidence indicating that the lidocaine-induced increase in [Ca++]i is an underlying mechanism of lidocaine-induced neurotoxicity. Better understanding of the mechanisms of toxicity might aid in the development of pharmacological strategies to avoid the neurologic injury that complicates the use of spinally administered local anesthetics.
| |
Footnotes |
|---|
Accepted for publication January 5, 1998.
Received for publication July 16, 1997.
1 This work was supported by National Institutes of Health grants NS21647, NS07265, NS07101 and GM51887.
Send reprint requests to: Michael S. Gold, Ph.D., University of Maryland at Baltimore, Department of OCBS, Room 5-A-12, Dental School, 666 W. Baltimore Street, Baltimore, MD 21201.
| |
Abbreviations |
|---|
BAPTA, 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid; DMSO, dimethyl sulfoxide; DRG, dorsal root ganglion; HEPES, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid; PBS, phosphate-buffered saline; TTX-R INa, tetrodotoxin-resistant sodium current; TTX-S INa, tetrodotoxin-sensitive sodium current.
| |
References |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Umbrain, L. Shi, M.-H. Lauwers, I. Smolders, Y. Michotte, and F. Camu Intrathecal lidocaine elevates prostaglandin E2 levels in cerebrospinal fluid: a microdialysis study in freely moving rats Br. J. Anaesth., November 1, 2008; 101(5): 716 - 722. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Hiruma, K. Shimizu, T. Takenami, H. Sugie, and T. Kawakami Effects of clonidine on lidocaine-induced inhibition of axonal transport in cultured mouse dorsal root ganglion neurones Br. J. Anaesth., November 1, 2008; 101(5): 659 - 665. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Onizuka, T. Kasaba, and M. Takasaki The Effect of Lidocaine on Cholinergic Neurotransmission in an Identified Reconstructed Synapse Anesth. Analg., October 1, 2008; 107(4): 1236 - 1242. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Lirk, I. Haller, H. P. Colvin, L. Lang, B. Tomaselli, L. Klimaschewski, and P. Gerner In Vitro, Inhibition of Mitogen-Activated Protein Kinase Pathways Protects Against Bupivacaine- and Ropivacaine-Induced Neurotoxicity Anesth. Analg., May 1, 2008; 106(5): 1456 - 1464. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Lirk, I. Haller, H. P. Colvin, S. Frauscher, L. Kirchmair, P. Gerner, and L. Klimaschewski In Vitro, Lidocaine-Induced Axonal Injury Is Prevented by Peripheral Inhibition of the p38 Mitogen-Activated Protein Kinase, but Not by Inhibiting Caspase Activity Anesth. Analg., December 1, 2007; 105(6): 1657 - 1664. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Karpie and C. R. Chu Lidocaine Exhibits Dose- and Time-Dependent Cytotoxic Effects on Bovine Articular Chondrocytes In Vitro Am. J. Sports Med., October 1, 2007; 35(10): 1621 - 1627. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Hogan Trying to understand spinal anesthesia/Une tentative de comprehension de la rachianesthesie Can J Anesth, August 1, 2007; 54(8): 607 - 612. [Full Text] [PDF]
|
|
|
|
|
|
A. Dogrul, S. E. Gulmez, M. S. Deveci, H. Gul, M. H. Ossipov, F. Porreca, and F. C. Tulunay The Local Antinociceptive Actions of Nonsteroidal Antiinflammatory Drugs in the Mouse Radiant Heat Tail-Flick Test Anesth. Analg., April 1, 2007; 104(4): 927 - 935. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Du, M. Yu, N. D. Volkow, A. P. Koretsky, J. S. Fowler, and H. Benveniste Cocaine Increases the Intracellular Calcium Concentration in Brain Independently of Its Cerebrovascular Effects. J. Neurosci., November 8, 2006; 26(45): 11522 - 11531. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Muguruma, S. Sakura, and Y. Saito Epidural Lidocaine Induces Dose-Dependent Neurologic Injury in Rats Anesth. Analg., October 1, 2006; 103(4): 876 - 881. [Abstract] [Full Text] [PDF]
|
