The cell membrane’s resting potential is negative and close to the potential determined by potassium alone (-70 mV). During the transmission of an action potential, Na⁺ moves into the cell through open Na⁺ channels depolarizing the membrane and bringing its potential to -20 mV or more.

Local anesthetics are compounds that have the ability to interrupt the transmission of the action potential in excitable membranes. They bind to specific receptors in the Na⁺ channels and their action, at clinically recommended doses, is reversible. Conduction can still continue, although at a slower pace, with up to 90% of receptors blocked.

All local anesthetics are neurotoxic, if injected intraneurally, and the damage caused is directly related to the degree of hydrostatic pressure reached inside the axoplasma. Local anesthetics injected around nerves could also be toxic as result of the concentration of the agent and the duration of the exposure (e.g., cauda equina after intrathecal local anesthetics).

The local anesthetics available in clinical practice are usually racemic mixtures, that is a mixture of both R and S enantiomers. Exceptions are lidocaine, levo-bupivacaine and ropivacaine. The S isomer appears to have similar efficacy than the R isomer, but lesser cardiac toxicity.

Historical perspective

Anesthesia by compression was common in the antiquity. Cold as an anesthetic was widely used until the 1800s, and then it came cocaine. The native Indians of Peru chewed coca leaves and knew about their cerebral-stimulating effects. The leaves of erythroxylon coca were taken to Europe where Niemann in Germany isolated cocaine in 1860. Carl Koller, a contemporary and friend of Sigmund Freud, is credited with the introduction of cocaine as a topical ophthalmic local anesthetic in Austria in 1884. In 1888 Koller came to the US and established a successful ophthalmology practice at Mount Sinai Hospital in New York until the year of his death in 1944.

Recognition of cocaine’s cardiovascular side effects, as well as its potential for dependency and abuse, led to a search for better local anesthetic drugs. Cocaine is a good topical local anesthetic that also produces vasoconstriction and for this reason it is still used by some, for topical anesthesia of the nose and other mucous membranes. Cocaine blocks the reuptake of cathecolamines from nerve endings. Total dose should not exceed 100 mg (2.5 mL of a 4% solution), to avoid systemic effects (hypertension, tachycardia, cardiac arrhythmias).

Ropivacaine is the only other local anesthetic able to produce some vasoconstriction, and that effect is weak.

Some highlights on local anesthesia

1850s Invention of the syringe and hypodermic hollow needle.

1884 Halsted, an American surgeon, blocks the brachial plexus with a solution of cocaine under direct surgical exposure.

1885 Wood, in the United Kingdom, is credited with the introduction of conduction anesthesia through hypodermic injection.

1897 Epinephrine is isolated by John Abel at Johns Hopkins Medical School.

1897 Braun in Germany relates cocaine toxicity with systemic absorption and advocates the use of epinephrine.

1898 Bier is set to receive the first planned spinal anesthesia from his assistant Hildebrandt. After CSF is obtained, the syringe is found not fit the needle and therefore no injection could be performed. Bier then performs the first spinal anesthesia on Hildebrandt using cocaine. They both experience the first spinal headaches.

1908 Bier introduces the intravenous peripheral nerve block (Bier block) with procaine.

1911 Hirschel performs the first percutaneous axillary block.

1911 Kulenkampff performs the first percutaneous supraclavicular block.

1922 Gaston Labat from France, a disciple of Pauchet, introduces in the US his book “Regional Anesthesia Its Technic and Clinical Application”, the first manual of regional anesthesia published in America.

1923 Labat founds the first American Society of Regional Anesthesia.

1953 Daniel Moore, from Virginia Mason Clinic in Seattle, publishes his book “Regional Block”.

1975 Alon Winnie, L. Donald Bridenbaugh, Harold Carron, Jordan Katz, and P. Prithvi Raj establish the current American Society of Regional Anesthesia (ASRA) in Chicago.

1976 The first ASRA meeting is held in Phoenix, Arizona.

1976 Regional Anesthesia Journal, volume 1, number 1 is published.

1983 Winnie introduces his book, Plexus Anesthesia, Perivascular Techniques of Brachial Plexus Block.

Date of introduction in clinical practice of some local anesthetics:

1905 procaine; 1932 tetracaine; 1947 lidocaine; 1955 chloroprocaine (last ester type that is still in clinical use); 1957 mepivacaine; 1963 bupivacaine; 1997 ropivacaine; 1999 levobupivacaine.

Chemical structure of local anesthetics

Local anesthetics are weak bases with a pka above 7.4 and poorly soluble in water. They are commercially available as acidic solutions (pH 4-7) of hydrochloride salts, which are hydrosoluble. A typical local anesthetic molecule is composed of two parts, a benzene ring (lipid soluble, hydrophobic) and an ionizable amine group (water soluble, hydrophilic). These two parts are linked by a chemical chain, which can be either an ester (-CO-) or an amide (-HNC-). This is the basis for the classification of local anesthetics as either esters or amides.

Injecting local anesthetics in the proximity of a nerve(s) triggers a sequential set of events, which eventually culminates with the interaction of some of their molecules with receptors located in the Na⁺ channels of nerve membranes. The injected local anesthetic volume spreads initially by mass movement, moving across “points of least resistance”, which do not necessarily lead into the desired nerve(s). This fact emphasizes the importance of injecting in close proximity to the target nerve(s). The local anesthetic solution then diffuses through tissues; each layer acting as a physical barrier. In the process part of the solution gets absorbed into the circulation. Finally a small percentage of the anesthetic reaches the target nerve membrane, at which point the different physicochemical properties of the individual anesthetic will dictate the speed, duration and nature of the interaction with the receptors.

Physicochemical properties-activity relationship

1. Lipid solubility: determines both the potency and the duration of action of local anesthetics, by facilitating their transfer through membranes and by keeping the drug close to the site of action and away from metabolism. In addition, the local anesthetic receptor site in Na⁺ channels is thought to be hydrophobic, so its affinity for hydrophobic drugs is greater. Hydrophobicity also increases toxicity, so the therapeutic index of more lipid soluble drugs is decreased.

Protein binding: local anesthetics are bound in large part to plasma and tissue proteins. The bound portion is not pharmacologically active. The plasmatic unbound fraction is responsible for systemic toxicity. The most important binding proteins in plasma are albumins and alpha-1-acid glycoprotein (AAG). Although albumin has a greater binding capacity than AAG, the latter has a greater affinity for drugs with pka higher than 8, the case for most local anesthetics. Newborn infants have very low concentration of AAG, only reaching adult values by 10 months of age. The elderly and debilitated also frequently have decreased levels of albumin and other plasma proteins. These patient populations could be at increased risk for toxicity.

On the other hand, AAG levels increase during stress and for several days after the postoperative period. Higher levels of AAG lead to decreased levels of unbound fraction of local anesthetics and a decreased potential for local anesthetic toxicity. However, changes in protein binding are only clinically important for drugs highly protein-bound, such as bupivacaine, which is 96% bound, and sufentanil and alfentanil, which are both 92% bound (Booker et al, Br J Anaesth 1996; 76:365-8).

The fraction of drug bound to protein in plasma correlates with the duration of action of local anesthetics: bupivacaine (95%) = ropivacaine (94%)> tetracaine (85%) > mepivacaine (75%) > lidocaine 65%) > procaine (5%) and 2-chloroprocaine (negligible). This suggests that the binding site for the local anesthetic molecule in the sodium channel receptor protein, may share a similar sequence of amino acids with the plasma protein binding site.

Drugs as lidocaine, tetracaine, bupivacaine and morphine (e.g., DepoDur) have been incorporated into liposomes to prolong their duration of action. Liposomes are vesicles with two layers of phospholipids, which slow down the release of the drug.

Pka: determines the ratio between the ionized (cationic) and the uncharged (base) forms of the drug. The pka of local anesthetics ranges from 7.6 to 9.2. By definition the pka is the pH at which 50% of the drug is ionized and 50% is present as a base. The pka generally correlates with the speed of onset of most local anesthetics. The closer the pka is to physiologic pH, the faster the onset. For example, lidocaine with a pka of 7.7 is 25% non-ionized at pH 7.4. Its onset is therefore faster than bupivacaine, whose pka of 8.1 makes it only 15% non-ionized at that pH. One important exception is 2-chloroprocaine that, despite its pka of 9.1, has a very rapid onset. This is usually attributed to the relatively high concentrations used in clinical practice (3%) that are possible thanks to its low toxicity. It is also claimed that 2-chloroprocaine has better “tissue penetrability”.

Mechanism of action and sodium channels

The non-charged hydrophobic fraction (B), which exists in equilibrium with the hydrophilic charged portion (BH⁺), crosses the lipidic nerve membrane and initiates the events that lead to Na⁺ channel blockade. Once inside the cell, the pka of the drug and the intracellular pH dictate a new equilibrium between the two fractions. Because of the relative more acidic intracellular environment, the relative proportion of charged fraction (BH⁺) increases. This hydrophilic, charged fraction is the active form on the Na⁺channel.

The Na⁺channel is a protein structure that communicates the extracellular of the nerve with its axoplasm. It consists of four repeating alpha subunits and two beta subunits, beta-1 and beta-2. The alpha subunits are involved in ion movement and local anesthetic activity. It is generally accepted that the main action of local anesthetics involves interaction with specific binding sites within the Na⁺channel. Local anesthetics may also block to some degree calcium and potassium channels as well as N-methyl-D-aspartate (NMDA) receptors. Local anesthetics do not ordinarily affect the membrane resting potential.

The Na⁺ channels seem to exist in three different states, closed (resting), open and inactivated. Under adequate stimulation, the protein molecules of the channel undergo conformational changes, from the resting state to the ion-permeable state or open state, allowing the inflow of extracellular Na⁺, which depolarizes the membrane. After a few milliseconds the channel goes then through a transitional inactivated state, where the proteins leave the channel closed and ion-impermeable. With repolarization the proteins revert to their resting configuration.

Other drugs, like tricyclic antidepressants (amitriptyline), meperidine, volatile anesthetics and ketamine, also exhibit Na⁺ channel-blocking properties. Tetrodotoxin and other biotoxins also interact with the Na⁺channels,althoughtheir actions are exerted on the extracellular side of the channel.

Frequency-dependent blockade

Local anesthetics show more affinity for open Na⁺ channels. When a nerve is experiencing a high frequency of depolarization, like during spontaneous pain or voluntary muscle contractions, it becomes more sensitive to blockade, because the chances of interaction, between local anesthetics molecules and Na⁺channels, increase.

The concept of frequency-dependent blockade also explains the greater susceptibility to blockade exhibited by small sensory fibers, as they generate long action potential (5 ms) at high frequency. Motor fibers on the other hand generate short action potentials (0.5 ms) at lower frequency making them more difficult to block.

Pregnancy and local anesthetics

Increased sensitivity to local anesthetics, demonstrated as faster onset and more profound block, may be present during pregnancy.

Alterations in protein binding of bupivacaine, may result in increased concentrations of active unbound drug in the pregnant patient, increasing its potential for toxicity.

Placental transfer is also more active for lipid soluble local anesthetics. In any case, agents with a pka closer to physiologic pH have a higher placental transfer. For example the umbilical vein/maternal vein ratio for mepivacaine is 0.8 (pka 7.6) while for bupivacaine is 0.3 (pka 8.1).

In the presence of fetal acidosis, local anesthetics cross the placenta and become ionized in higher proportion than at normal pH. The ionized fraction cannot cross back to the maternal circulation, originating what is called “ion trapping”. Therefore, 2-chloroprocaine, with its very short maternal and fetal half-lives, is an ideal local anesthetic in the presence of fetal acidosis.

Fiber size and pattern of blockade

As a general rule small nerve fibers are more susceptible to local anesthetics than large fibers. However, other factors like myelinization and relative position of the fibers within a nerve (mantle versus core) may also play a role. The depolarization in myelinated fibers is saltatory. About three nodes of Ranvier need to be blocked in order to block the transmission of the action potential.

The smallest nerve fibers are nonmyelinated and are blocked more readily than larger myelinated fibers. However at similar size, myelinated fibers are blocked before nonmyelinated fibers. In general autonomic fibers, small nonmyelinated C fibers (mediating pain, temperature and touch), and small myelinated A delta fibers (mediating pain and cold temperature) are blocked before A alpha, A beta and A gamma fibers (motor, propioception, touch, and pressure).

It has been speculated that in large nerve trunks, motor fibers would be usually located in the outer portion (mantle) of the nerve bundle, therefore more “accessible” to local anesthetics. This would help explain why motor fibers tend to be blocked before sensory fibers in large mixed nerves. In contrast, the frequency-dependence of local anesthetic action would favor block of small sensory fibers, as they generate long action potentials at high frequency, whereas motor fibers generate short action potentials at lower frequency.

(Figure from Morgan’s Clinical Anesthesiology, 3^rd edition, 2006, reproduced with permission)

Modulating local anesthetic action

pH adjustment

The ionized fraction of local anesthetics is the active form in the Na⁺channel, although the rate-limiting step in this cascade is membrane penetration of local anesthetics in its non-ionized form. Unfortunately, only a small proportion of local anesthetic in solution exists in the non-ionized state. Changes in pH can theoretically reduce the onset time by increasing its proportion. At a pH of 5.0 to 5.5 the cation/base ratio is 1000:1, at a pH of 7.4 the same ratio becomes 60:40. The limiting factor for pH adjustment is the solubility of the base form before reaching precipitation. The most lipid soluble agents, like bupivacaine and ropivacaine, cannot be alkalinized above a pH of 6.5 because they precipitate.

DiFazio et al (Anesth Analg 1986:65; 760-64) demonstrated a more than 50% decrease in onset of epidural anesthesia, when the pH of commercially available lidocaine with epinephrine was raised from 4.5 to 7.2, by the addition of bicarbonate. Capogna et al (Reg Anesth 1995; 20: 369-377) randomized 180 patients to study the effects of alkalinizing lidocaine, mepivacaine and bupivacaine for nerve blocks. They concluded that alkalinization of lidocaine and bupivacaine shortens the onset of epidural; alkalinization of lidocaine shortens the onset of axillary block and alkalinization of mepivacaine shortens the onset of sciatic/femoral blocks. However, when only small changes in pH can be achieved, because of the limited solubility of the base, only small decreases in onset time will occur, as when plain bupivacaine is alkalinized.

It is generally accepted that adding bicarbonate to local anesthetics, may speed the onset of local anesthetics solutions that have epinephrine added by the manufacturer (vials have a lower pH), while the effect would be negligible when fresh epinephrine is added to a plain solution.

Chloroprocaine plus 1 mL of sodium bicarbonate for 30 mL of solution raises the pH to 6.8. Adding 1 mL of sodium bicarbonate per 10 mL of lidocaine or mepivacaine raises the pH of the solution to 7.2 and adding 0.1 mL of bicarbonate per 10 mL of bupivacaine raises the pH of the solution to 6.4 (from Mulroy’s Regional Anesthesia, 3^rd edition, 2002).

Carbonation

Another approach to shortening onset time has been the use of carbonated local anesthetic solutions. These solutions contain large amounts of carbon dioxide, which readily diffuses into the axoplasm of the nerve, lowering the pH and favoring the formation of the cationic active form of the local anesthetic inside the cell. Carbonated solutions are not available in the United States.

LOCAL ANESTHETICS ADDITIVES

Vasoconstrictors

Epinephrine is the most common vasoconstrictor added to local anesthetics to prolong the anesthetic effect and to decrease absorption. Epinephrine is also used to detect intravascular injection. Without beta-blockers on board, 15 mcg of epinephrine should produce a 30% increase in heart rate within 30 seconds.

Vasoconstrictors may also improve the quality and density of the block, especially with spinal and epidural anesthesia. This has been demonstrated with tetracaine, lidocaine and bupivacaine. The mechanism is unclear. Epinephrine may simply increase the amount of local anesthetic available by reducing absorption. It could also have some local anesthetic effect by means of its α2-agonist actions. Subarachnoid epinephrine also delays voiding and discharge readiness.

The prolongation of effect in peripheral nerve blocks can be 30-60%, depending on site of injection and type of local anesthetics (more vascular sites like intercostal see more effect, and intermediate agents like lidocaine benefit more). Peripherally epinephrine does not have any significant alpha-2 effect.

In general, epinephrine added to spinal anesthesia prolongs the effect of the less lipid soluble agents like lidocaine and mepivacaine (20-30%). The exception to this rule is tetracaine, a highly lipid soluble agent, that gets the largest prolongation of all spinal local anesthetics (up to 60% in lumbar dermatomes).

The usual dose of intrathecal epinephrine is 200 mcg, but doses as small as 50 mcg can be sufficient. In the epidural space the usual dose is 5 mcg/mL. Epinephrine, other than intrathecal, is absorbed systemically and may produce adverse cardiovascular effects. In small doses the beta-adrenergic effects predominate, with increased cardiac output and heart rate. Dose larger than 0.25 mg (250 mcg) may be associated with arrhythmias or other undesirable cardiac effects.

The potential risk for peripheral nerve ischemia, as a result of epinephrine acting on epineural vessels and vaso nervorum has to be balanced against the lower risk of systemic toxicity, the ability to detect intravascular injection and the prolongation of action. According to Neal (Reg Anesth Pain Med 2003;28:124-134) adding 5 mcg/mL (1:200,000 dilution) prolongs the duration of lidocaine for peripheral nerve blocks from 186 min to 264 min. Adding only 2.5 mcg/mL (1:400,000 dilution) prolongs the block to 240 min (almost the same prolongation), without apparent effect on nerve blood flow. Patients with micro angiopathy (e.g., diabetics), who could be at increase risk for neural ischemia secondary to vasoconstriction, potentially could benefit from the use of more diluted epinephrine. Adding only 1:400,000 epinephrine to local anesthetic solutions for nerve blocks has become the standard in our practice, in both diabetics and non-diabetics patients.

Intrathecal epinephrine does not lead to cord ischemia, because it does not decrease spinal cord blood flow, although it decreases epidural blood flow (Kosody R, et al. Can Anaesth Soc J; 31: 503-8, 1984).

In fact spinal cord ischemia due to epinephrine is “improbable because the cord vessels are autoregulated and show very minimal response to endogenous or exogenous vasoactive agents” (Neal JM In: Regional Anesthesia, The Requisites. Elsevier Mosby, Philadelphia 2004, pp 25-31)

Although epinephrine-containing local anesthetics are usually contraindicated in areas of terminal circulation (e.g., digits) this recommendation is not based on hard evidence. Anecdotal use of epinephrine-containing solutions in digits is cited in the literature. Lalonde et al published a multicenter study including 3,110 consecutive cases of use of epinephrine in the fingers and hand from 2002-2004. The authors (surgeons) defined “low dose” epinephrine as 1:100,000 and they reported no instance “of digital tissue loss” (J Hand Surg 2005; 30:1061-67). At this time we do not recommend this practice.

Dilution/concentration issues

By definition a 1:1,000 dilution means 1 g solute in 1,000 mL of solution. That is the same than to say 1 mg/mL or 1,000 mcg/mL.

Therefore:

1:10,000 equals 100 mcg/mL
1:100,000 equals 10 mcg/mL
1:200,000 equals 5 mcg/mL
1:400,000 equals 2.5 mcg/mL

Opioids

Neuraxial use: The addition of opioids to local anesthetics has a synergistic effect, both in anesthesia and postoperative analgesia (especially visceral pain). They block pain pathways without significantly affecting motor or sympathetic fibers.

The hydrophilic opioid morphine can be used in doses of 0.1-0.3 mg spinal and 1-3 mg epidural. It has a slow onset of 45 min, providing an analgesic action that lasts 12-24 h. Morphine reaches the brainstem and 4^th ventricle slowly. Delayed respiratory depression (8-10 h) is a risk with all neuraxial opioids, but it is more frequently seen with hydrophilic drugs like morphine, and in susceptible populations like the elderly and debilitated. Neuraxial morphine is also associated with higher incidence (40-50%) of nausea and vomiting than systemic opioids, more pruritus (60-80%, 20% of it severe), and delayed voiding. It is not suitable for outpatients.

Short-acting opioids, such as fentanyl and sufentanil, when added to spinal anesthetics can also intensify the block, and prolong the duration of anesthesia, beyond the duration of local anesthetics. Respiratory depression with these agents is rare and usually early (within 4 h). Sufentanil spinal can be used in doses of 2.5-10 mcg. Fentanyl spinal is used in doses of 10-25 mcg and 25-150 mcg epidural. Onset occurs at 5-15 min, peak effect at 10-20 min and duration of 1-3 h. Hypotension, pruritus, nausea and vomiting are some common side effects.

Extended-release epidural morphine (DepoDur): is a new liposomal formulation designed for epidural use, providing 48 h of pain relief. DepoDur was approved in 2004. It is supplied in a 2 mL vial containing 10 mg/mL dose in sterile saline. It is only approved as a single lumbar epidural dose prior to surgery or after clamping of the umbilical cord during C-section. The recommended dose is 10 mg for C-section, 10-15 mg for lower abdominal surgery and 15 mg for major orthopedic surgery of the lower extremities. Respiratory depression is dose-related. The most common adverse events reported during clinical trials were decreased oxygen saturation, hypotension, urinary retention, nausea and vomiting, constipation and pruritus.

Peripheral nerve blocks: The usefulness of opioids in peripheral nerve blocks is mostly unsupported by the evidence. Opioids have been shown useful when injected in the intra-articular space.

Clonidine

Alpha-2 agonists have central (sedation, analgesia, bradycardia) and peripheral effects (vasoconstriction/vasodilation with net hypotension, anti shivering, diuresis). The site for sedative action is the locus ceruleus of the brain stem, while the principal site for analgesia seems to be the spinal cord.

The main alpha-2 effect on the heart is decreased tachycardia by blocking cardioacelerator fibers, and bradycardia through a vagomimetic effect. In the periphery clonidine produces both vasodilation via sympatholysis and vasoconstriction through receptors on smooth muscle. The cause for its anti shivering and diuretic effects are yet to be established.

Side effects, including sedation, hypotension and bradycardia, limit alpha-2 agonists use. Small doses of clonidine (50-75 mcg) have shown to significantly prolong analgesia in spinal, epidural, IV regional, and peripheral nerve blocks, both when injected along local anesthetics and when given orally. Injected intrathecally, they also can delay voiding and can produce orthostasis. Side effects do not occur often at clonidine doses below 1.5 mcg/kg or a total dose less than 150 mcg.

Iskandar et al in France in 2001, showed that adding 50 mcg of clonidine to selected nerves (median and musculocutaneous) prolonged mepivacaine sensory anesthesia by 50%, compared to placebo, after a mid-humeral block, without prolonging motor effect. Because the prolongation was observed only in the nerves that received clonidine they postulated that the effect must be peripheral and not central through absorption.

Dexmedetomidine

It is a more selective alpha-2 agonist agent with an alpha-2:alpha-1 receptor ratio of 1,600:1, seven times greater than that of clonidine. Its elimination half-life is only 2 h compared to more than 8 h for clonidine. Dexmedetomidine may offer extended analgesia with lesser side effects. Currently it is approved only for sedation in the ICU.

Neostigmine

It is an acetylcholinesterase inhibitor that prevents the breakdown of acetylcholine promoting its accumulation. Acetylcholine is an endogenous spinal neurotransmitter that induces analgesia. Neostigmine does not cause neural blockade nor have any action on opioid receptors.

Spencer Liu et al in 1999 (Anesthesiology 1999; 90:710-717) studied the effects of different doses of neostigmine added to bupivacaine spinal. They reported that 50 mcg of neostigmine increased sensory and motor anesthesia, but also delayed discharge time and was accompanied by 67% nausea and up to 50% vomiting. Lower doses did not show analgesic effect, but still had significant rates of side effects (nausea and vomiting).

N-methyl-D-aspartate (NMDA) receptor antagonists

Activation of NMDA receptors makes the neurons of the spinal cord more responsive to all types of input including pain stimuli (central sensitization). NMDA receptor antagonists, like ketamine, have shown analgesic activity. In fact in IV regional 0.1 mg/kg of ketamine is superior to clonidine (1 mcg/kg) in preventing tourniquet pain.

Errando in Spain showed that commercially available ketamine containing benzethonium chloride is toxic in swine (Reg Anesth Pain Med 1999; 24:146-52). Preservative-free solutions of ketamine have proven safe.

Hyaluronidase

It breaks down collagen bonds potentially facilitating the spread of local anesthetic through tissue planes. However, the evidence shows that at least in the epidural space it can decrease the quality of anesthesia. Its use seems limited to retrobulbar blocks.

Dextran

Dextran and other high-molecular-weight compounds have been advocated to increase the duration of local anesthetics. The evidence is lacking.

METABOLISM OF LOCAL ANESTHETICS

Ester local anesthetics

They are rapidly hydrolyzed at the ester linkage by plasma pseudocholinesterase, the same enzyme that hydrolyses acetylcholine and succinylcholine. The hydrolysis of 2-chloroprocaine is about four times faster than procaine, which in turn is hydrolyzed about four times faster than tetracaine. However, even tetracaine has a metabolic half-life of only 2.5-3.0 min (Tetzlatt JE. In: Ambulatory Anesthesia Perioperative Analgesia. New York, McGraw-Hill, 2005, p 193).

In individuals with atypical plasma pseudocholinesterase the half-life of these drugs is prolonged and potentially could lead to plasma accumulation. Cerebrospinal fluid does not contain esterase enzymes, so if an ester is used for spinal anesthesia (e.g., tetracaine) its termination of action depends on blood absorption.

The hydrolysis of all ester local anesthetics leads to the formation of para-aminobenzoic acid (PABA), which is associated with a low potential for allergic reactions. Allergic reactions may also develop from the use of multiple dose vials of amide local anesthetics that contain methylparaben (PABA derivative) as a preservative.

As opposed to other ester type anesthetics, cocaine is partially metabolized in the liver and partially excreted unchanged in the urine.

Amide local anesthetics

They are transported into the liver before their biotransformation. The two major factors controlling the clearance of amide local anesthetics by the liver are hepatic blood flow and hepatic function.

The metabolism of local anesthetics as well as that of many other drugs occurs in the liver by the cytochrome P-450 enzyme system. Because of the liver large metabolic capacity it is unlikely that drug interaction would affect the metabolism of local anesthetics. The rate of metabolism is agent specific (prilocaine > lidocaine > mepivacaine > ropivacaine > bupivacaine).

The metabolism of amide local anesthetics is relatively fast, although slower than esters. Elimination half-life for lidocaine is 1.5-2 h. Drugs such as general anesthetics, norepinephrine, cimetidine, propranolol and calcium channel blockers can decrease hepatic blood flow and potentially increase the elimination half-life of amides. Similarly, decreases in hepatic function caused by a lowering of body temperature, immaturity of the hepatic enzyme system in the fetus, or liver damage (e.g., cirrhosis) can lead to decreased rate of hepatic metabolism of the amides. Renal clearance of unchanged local anesthetic is a minor route of elimination (e.g., lidocaine is only 3% to 5% recovered unchanged in the urine of adults, while bupivacaine is 10% to 16%).

The dibucaine number

People with atypical plasma pseudocholinesterase exhibit prolonged recovery after a dose of succinylcholine or mivacurium. Dibucaine is an amide local anesthetic that helps to identify those patients. Dibucaine binds strongly to normal plasma pseudocholinesterase inhibiting its action. This inhibition is reported as a number from 1-100 representing the percentage of normal enzyme inhibition, the larger the number the larger the proportion of normal enzyme. A number of 80 or higher means that dibucaine is able to inhibit at least 80% of the enzyme and that the patient is a normal homozygous. A dose of succinylcholine will last 4-6 min. A dibucaine number of 50 means that the patient is heterozygous and that the effect of succinylcholine will be prolonged to up to 30 min. A number of 20 is related to the homozygous atypical enzyme and the effect of succinylcholine could be expected to last up to 6 h (incidence 1:3,300).

LOCAL ANESTHETIC TOXICITY

The capacity of a local anesthetic to produce systemic toxicity is directly related to plasma level of unbound drug. This plasma level depends on:

Total dose
Net absorption, which depends on: vasoactivity of the drug, site vascularity and use of a vasoconstrictor
Metabolism and elimination of the drug from the circulation

Brown et al reported a 1.2 in 10,000 incidence of systemic toxicity after epidural anesthesia and 19 in 10,000 after peripheral nerve blocks.

According to Mather et al, central nervous system (CNS) and cardiovascular (CV) effects are “poorly correlated with arterial drug concentrations” and better correlated with the “respective regional venous drainage”. According to them, lung uptake reduces the drug concentration by 40% and slower injection (3 min compared to 1 min) achieves similar decreases (Reg Anesth Pain Med 2005; 30: 553-66). CNS signs of toxicity usually precedes CV manifestations.

Peak local anesthetic blood levels are directly related to the dose administered at any given site. However the vascularity of the site at similar doses is very important in determining different plasma levels. The absorption of local anesthetics from different sites is from highest to lowest: endotracheal > intercostal > caudal > epidural > plexus blocks > sciatic/femoral > subcutaneous infiltration.

Generally the administration of a 100-mg dose of lidocaine in the epidural or caudal space results in approximately a 1 mcg/mL peak blood level in an average adult. The same dose injected into less vascular areas (e.g., brachial plexus axillary approach or subcutaneous infiltration) produces a peak blood level of app 0.5 mcg/mL. The same dose injected in the intercostal space produces a 1.5 mcg/mL plasma level.

Peak blood levels may also be affected by the rate of biotransformation and elimination. In general this is the case only for very actively metabolized drugs such as 2-chloroprocaine, which has a plasma half-life of about 45- 60 seconds.

For amide local anesthetics like lidocaine peak plasma levels after regional anesthesia primarily result from absorption and in general occur within 1 h (please see difference with tumescent anesthesia).

Rodriguez et al studied 10 end-stage renal disease patients coming for A-V fistula (Eur J Anaesthesiol 2001; 18: 171-6). The patients received an axillary block with a total of 650 mg of plain mepivacaine. Plasma levels were studied during 150 min. Peak levels of 8.28 mcg/mL (range 3.83-11.21) were obtained (normal 5 mcg/mL) within 60 min and decreased steadily thereafter. Patients did not exhibit signs of toxicity despite these high plasma levels. This is in contrast with a case report by Tanoubi et al (Ann Fr Anesthe Reanim 2006; 25: 33-5), where an end-stage renal patient for A-V fistula received an axillary block with 375 mg (25 mL) of 1.5% mepivacaine and the patient presented with dysarthria, mental confusion and loss of consciousness without convulsions or arrhythmia. Mepivacaine plasma level at the time of symptoms was 5.1 mcg/mL.

Tumescent (diluted) anesthesia for liposuction

The use of highly diluted concentrations of lidocaine (0.1% or less) plus epinephrine (usually 1 mg per liter or 1:1,000,000) allows for painless and bloodless liposuction procedures. Lidocaine bounds to tissue proteins in this subdermal drug reservoir from where it is subsequently slowly released into the systemic circulation.

Diluted lidocaine, along with epinephrine-induced vasoconstriction, makes systemic uptake so slow as to match the liver maximum lidocaine clearance capacity of 250 mg/h. Therefore, according to de Jong, “the blood level remains below 5 mcg/mL toxic threshold, despite the administration of many times (e.g., 35 mg/kg) the conventional upper dose limit of undiluted full strength lidocaine” (Int J Cosmetic Surg 2002; 4: 3-7).

Peak plasma levels of lidocaine using tumescent technique occur between 5-17 hours compared to less than 1 h for common infiltration.

Central nervous system toxicity

Toxic plasma levels are usually produced by inadvertent intravascular injection. This is the basis for fractionating of the dose and the use of vasoconstrictors. Toxic plasma levels could rarely result from the slow absorption from the injection site. A sequence of symptoms may include numbness of the tongue, lightheadedness, tinnitus, restlessness, tachycardia, convulsions and respiratory arrest. Interestingly enough, there is no evidence that patients suffering from seizure disorders are at any increased risk for CNS local anesthetic toxicity, including seizures.

The site of action for local anesthetic-induced seizures seems to be the amigdala, part of the limbic system, in the base of the brain.

Cardiovascular system toxicity

· The cardiovascular manifestations usually follow the CNS effects (therapeutic index). The exception is bupivacaine, which can produce cardiac toxicity at sub convulsant concentrations.

· Rhythm and conduction are rarely affected by lidocaine, mepivacaine and tetracaine, but bupivacaine and etidocaine can produce ventricular arrhythmias.

· EKG shows a prolongation of PR and widening of the QRS

· The incidence of CV toxicity with local anesthetics is higher in pregnancy due to higher proportion of unbound fraction.

· CV toxicity is increased under conditions of hypoxia and acidosis.

Toxic plasma concentration thresholds

The following are accepted plasma levels of selected local anesthetics, above which systemic effects are expected in humans:

Lidocaine 5 mcg/mL; mepivacaine 5 mcg/mL; bupivacaine 1.5 mcg/mL; ropivacaine 4 mcg/mL

Management of systemic toxicity

The best treatment for toxic reactions is prevention. When local anesthetic-induced seizures occur, hypoxia, hypercarbia and acidosis develop rapidly. ABC (Airway, Breathing and Circulation) is the mainstay of treatment. Administration of O₂ by mask, or ventilation support by bag and mask, is often all that is necessary to treat seizures. If seizures interfere with ventilation, benzodiazepines, propofol or thiopental can be used. The use of succinylcholine effectively facilitates ventilation and, by abolishing muscular activity, decreases the severity of acidosis. However neuronal seizure activity is not inhibited and therefore, cerebral metabolism and oxygen requirements remain increased.

In an interesting study by Mayr et al, out of Innsbruck, Austria (Anesth Analg 2004; 98: 1426-3), the authors induced cardiac arrest in 28 pigs by administering 5 mg/kg of 0.5% bupivacaine and stopping ventilation until asystole occurred. CPR was initiated after 1 min of cardiac arrest. After 2 min the animals received every 5 min either epinephrine alone; vasopressin alone; epinephrine plus vasopressin or placebo IV. In the vasopressin/epinephrine group all pigs survived and in the placebo group all pigs died. In the vasopressin alone 5 of 7 survived and in the epinephrine group 4 of 7 survived. This is in line with current ACLS recommendations of using one single dose of 40U of vasopressin IV before using epinephrine.

Little information is available regarding the treatment of local anesthetic cardiovascular toxicity in humans. Animal data suggest:

1. Vasopressin 40 U, IV, single dose, one time only followed by, if needed, high doses of epinephrine (1 mg IV every 3-5 minutes) to support heart rate and blood pressure.

2. Atropine may be useful for bradycardia.

3. DC cardioversion is often successful.

4. Ventricular arrhythmias are probably better treated with amiodarone than with lidocaine. Amiodarone is used as for ACLS, 150 mg over 10 min, followed by 1 mg/min for 6 hrs then 0.5 mg/min. Supplementary infusion of 150 mg as necessary up to 2 g. For pulseless VT or VF, initial administration is 300 mg rapid infusion in 20-30 mL of saline or dextrose in water.

Bupivacaine toxicity and use of lipid emulsion to treat it

Bupivacaine cardiac toxicity was highlighted in an editorial report by Albright in 1979, in which he described several cases of refractory cardiac arrests in association with the use of bupivacaine (Anesthesiology 1979; 51:285-7). In 2003, Weinberg and colleagues from the University of Illinois, published an interesting paper describing the use of a 20% lipid emulsion in combination with cardiac massage to successfully return normal hemodynamics to 9 out of 9 dogs, after asystole brought by a bolus injection of 10 mg/kg of bupivacaine (Reg Anesth Pain Med 2003; 28:198-202).

The results of this study led them to recommend treating bupivacaine-associated cardiac arrest with a 20% lipid emulsion IV. The treatment protocol includes a 1 mL/kg bolus of 20% lipid emulsion (such as intralipid), followed by an infusion of 0.25 mL/kg/min for 10 min, and the continuation of basic life support. The bolus can be repeated every 5 min, up to three times as needed. The maximum dose of 20% lipid emulsion is not known, but the authors suggest that more than 8 mL/kg would not likely be needed, nor successful, if lower doses do not work. This protocol will deliver a significant volume load to the patient. The paper was accompanied by an editorial by Groban and Butterworth from Wake Forest, in Winston-Salem, North Carolina. They believe that the most likely mechanism of action of lipid emulsion is that “in some way the lipid is serving to more rapidly remove LA molecules from whatever binding site serves to produce the cardiovascular depression that has come to be known as bupivacaine toxicity”.

ACLS protocols must be followed with prompt defibrillation and use of pressors like vasopressin followed by epinephrine, to support coronary perfusion if necessary. Amiodarone should be favored over lidocaine to treat arrhythmias and initiate the lipid emulsion at the “earliest sign of severe local anesthetic-induced cardiac toxicity.

In 2006 Rosenblatt et al (Anesthesiology 2006; 105: 217-8) published a case report of successful use of 20% lipid emulsion (Intralipid, Baxter Pharmaceuticals) on a 58-year old male who developed a cardiac arrest, presumably linked to bupivacaine after an interscalene block. They described that after 20 min of cardiac compressions and with the patient in asystole, 100 mL of intralipid IV was given resulting in an apparent “immediate” return of patient’s rhythm. This dose is higher than the recommended 1 mL/kg. A continuous infusion of intralipid was given at 0.5 mL/kg/min for 2 h. The patient was extubated 2.5 after hours after the episode, without any apparent neurological sequelae. In an accompanying editorial, Weinberg suggested having 20% lipid emulsion available in all sites where local anesthetics are used.

Also in 2006, soon after Rosenblatt’s report, Litz et al reported a case of successful use of intralipid after ropivacaine-induced asystole. More recently, in March 2008, McCutchen and Gerancher reported a case of ventricular tachycardia treated with 150 mg of amiodarone, 10 mL of 20% intralipid and a synchronized countershock of 120 J, after which there was a prompt return to normal sinus rhythm. The authors speculate that the use of intralipid might have prevented the patient from going into cardiac arrest.

In summary:

1. Evidence is accumulating on the beneficial effect of a 20% lipid emulsion to treat bupivacaine-related cardiac toxicity.

2. Propofol has the same vehicle than intralipid, but only half the concentration (10%). Giving propofol probably will not provide enough lipids, but instead it will produce a negative inotropic effect due to the presence of the active ingredient di-isopropylphenol (anesthetic action), exacerbating cardiac depression. Therefore, propofol is not indicated to treat local anesthetic-induced cardiac toxicity.

Maximum dose

Regional anesthesiologists perform peripheral nerve blocks with an amount of local anesthetics that usually exceeds what traditionally have been considered “maximum recommended doses”. However the traditional recommendations are based on extrapolation from animal data that do not necessarily apply to clinical practice. According to Rosenberg et al, the common recommendations for maximum doses, as suggested by the literature, “are not evidence based” (Reg Anesth Pain Med 2004; 29: 564-575), and according to Milroy have proven to be “poor approximation of safety” (Reg Anesth Pain Med 2005; 30: 513-515).

It is known that peak plasma levels do not correlate with patient size or body weight. Many practitioners have called to review these guidelines to better reflect the reality of clinical practice. The American Society of Regional Anesthesia convened a “Conference in Local Anesthetic Toxicity” with a panel of experts in 2001, to discuss the subject. Many papers related to that conference have been published. In a review article by Rosenberg et al (just cited) the authors argue that the safe doses instead should be block specific and related to patient’s age (e.g., epidural), organ dysfunction (especially for repeated doses) and whether the patient is pregnant. They suggest also adding epinephrine 2.5 to 5 mcg/mL, when not contraindicated.

The fact is that most of the systemic toxicity occurs with unintentional direct intravascular injection (Mather et al, Reg Anesth Pain Med 2005; 30: 553-566).

Methemoglobinemia

Normal hemoglobin contains an iron molecule in the reduced or ferrous form (Fe²⁺), the only form suitable for oxygen transport by hemoglobin. When hemoglobin is oxidized, the iron molecule is converted into the ferric state (Fe³⁺) or methemoglobin. Methemoglobin lacks the electron that is needed to form a bond with oxygen and therefore it is incapable of oxygen transport. Because red blood cells are continuously exposed to various oxidant stresses, blood normally contains approximately 1% methemoglobin levels. Prilocaine and benzocaine can oxidize the ferrous form of the hemoglobin to the ferric form, creating methemoglobin. It is more frequently seen with nitrates like nitroglycerin. When MetHb exceeds 4 g/dL cyanosis can occur.

Prilocaine doses of more than 600 mg are needed to produce clinically significant methemoglobinemia. Depending on the degree, methemoglobinemia can lead to tissue hypoxia. The oxyHb curve shifts to the left (P50 < 27 mmHg). MetHb has a larger absorbance than Hb and 0₂Hb at 940 nm, but simulates Hb at 660 nm. In the presence of high MetHb concentrations the SaO₂ falsely approaches 85%, independent of the actual arterial oxygenation. Diagnosis needs clinical suspicion and confirmation by blood analysis.

Methemoglobinemia is easily treated by the administration of methylene blue (1-2mg/kg of a 1% solution over 5 min) or less successfully with ascorbic acid (2 mg/kg).

Allergy

True allergy (type I or IgE mediated) to local anesthetics is rare and presents within minutes after the exposure. It is relatively more frequent with esters, which are metabolized to para-amino-benzoic acid (PABA). PABA is frequently used in the pharmaceutical and cosmetic industries. Allergy to amide local anesthetics is exceedingly rare. There is no cross allergy between esters and amides. However use of methylparaben as a preservative in multidose vials can elicit allergy in patients allergic to PABA.

Delayed hypersensitivity reactions (type IV) are T-cell mediated and present 24 to 48 h after exposure. There are few cases in the literature of delayed hypersensitivity to lidocaine, but recent reports suggest it may be more frequent than previously reported. The North American Contact Dermatitis Group found that 0.7 % of patients who were patch tested in 2001-02 demonstrated delayed allergy to lidocaine (ASRA News, February 2006).

Eutectic mixture of local anesthetics (EMLA)

EMLA cream is a 1:1 mixture of 5% lidocaine and 5% prilocaine. One gram of EMLA contains 25 mg of lidocaine, 25 mg of prilocaine, an emulsifier, a thickener and distilled water. EMLA is a liquid at room temperature, containing up to 80% concentration of the uncharged base form of local anesthetic, which confers better dermal penetration. Anesthesia onset takes between 45 to 60 minutes. Its main use is in children. One or 2 grams of EMLA cream are applied per 10 cm² of skin and covered with an occlusive dressing (maximum application area 2000 cm²or 100 cm² in children less than 10 kg).

Drug interactions

Local anesthetics potentiate the effects of non-depolarizing muscle relaxants. Simultaneous administration of succinylcholine and ester local anesthetics, both metabolized by pseudocholinesterases, may potentiate the effect of each other. Cimetidine and propranolol decrease hepatic blood flow and amide local anesthetic clearance increasing the potential for systemic toxicity. Opioids and alpha-2 adrenergic agonists potentiate the effects of local anesthetics and vice versa.

Profile summary of selected agents

1. Procaine:
Type: ester

Pka: 8.9

Protein biding: 5%

Characteristics: intermediate onset, low potency, short duration. Very short half-life (20 sec).

Other: it provides a short-duration spinal (potential benefit on outpatients).

2. 2-Chloroprocaine:

Type: ester

Pka: 9.3

Protein binding: negligible

Characteristics: very fast onset, despite high pka (ability to use higher concentrations could be the reason). Short duration (it has 30 minutes 2-segment regression in epidural). Very short half-life (30 sec).

Other: The original preparation contained sodium metabisulfite as a preservative. It was associated with serious neurological deficits when a large injection, planned for epidural, ended intrathecally. A second preservative, ethylenediamine tetra-acetic acid (EDTA) was associated with severe muscle spasm after epidural in ambulatory patients. EDTA chelates ionized calcium and this side effect may be secondary to action on paraspinal muscles.

The present solution is prepared without preservatives, and no back spasms have been reported.

3. Tetracaine:

Type: ester

Pka: 8.6

Protein binding: 85%

Characteristics: slow onset, high potency, long duration. Short plasma half-life (2.5 to 4 min).

Other: early experience with this product at high doses resulted in CNS toxicity, giving it a bad reputation, mostly undeserved. We still use it occasionally in our practice, as lyophilized crystals dissolved in liquid mepivacaine for a final concentration of 0.2% tetracaine. It prolongs duration of surgical anesthesia in peripheral nerve blocks to 4-6 h. Tetracaine also is the drug that gets the longest prolongation from adding epinephrine to spinal anesthesia (up to 60% in the lumbar dermatomes).

4. Cocaine:

Type: ester

Pka: 8.6

Protein binding: ?

Characteristics: slow onset, short duration. Elimination half life 60-90 min. Urinary excretion of unchanged cocaine is usually less than 1%, but it can be up to 9% especially in acid urine. At the end of 4 hours, most of the drug is eliminated from the plasma. Cocaine metabolites (benzoylecgonine and ecgonine) may be present in the urine for 24-36 hours, but some metabolites may be identified for up to 144 h after administration (Ellenhom and Barceloux, 1988).

Other: It produces vasoconstriction, while most of the LA with the exception of ropivacaine, produce some degree of vasodilation. It interferes with the reuptake of cathecolamines, resulting in hypertension, tachycardia, arrhythmias and myocardial ischemia. It is used mainly for topical anesthesia of the nose. Doses below 100 mg (2.5 mL) are usually safe.

Cocaine can potentiate cathecolamine-induced arrhythmias by halothane, theophylline or antidepressants. Cocaine can induce coronary vasospasm and potential myocardial ischemia, without the need for coronary artery disease. Mixtures of lidocaine and phenylephrine are safer alternatives.

5. Benzocaine:

Type: ester

Pka 3.5

Characteristics: slow onset, short duration and the only LA with a secondary amine structure that limits its ability to pass through membranes (topical use only).

Other: Doses higher than 300 mg can induce methemoglobinemia.

6. Lidocaine:

Type: amide

Pka: 7.8

Protein binding: 65%

Characteristics: intermediate onset and duration, elimination half-life 45-60 min. Other: it is versatile (topical, infiltration, IV regional, neuraxial, antiarrhythmic) and widely used. Spinal use is associated with around 30% of TNS, especially with lithotomy position, knee arthroscopy and obesity. Lowering the concentration does not eliminate the problem with doses larger than 40 mg. Doses of 25-40 mg highly reduce the incidence of TNS.

7. Mepivacaine:

Type: amide

Pka: 7.6

Protein binding: 75%

Characteristics: intermediate onset and duration. Elimination half-life is 2-3 h in adults and 8-9 h in neonates.

Other: It seems to produce less vasodilation than lidocaine. It has been used in spinal anesthesia. It has lower (but not zero) incidence of TNS.

It is the agent we most commonly use for peripheral nerve blocks. A 1.5% of plain solution provides a short onset and dense surgical anesthesia lasting 2-3 h (3-4 h with 1:400,000 epinephrine). Prolonged postoperative analgesia, as with all other LA, is negligible after single-shot blocks.

8. Bupivacaine:

Type: amide

Pka: 8.1

Protein binding: 95%

Characteristics: high potency, slow onset, long duration. Elimination half-life 3-3.5 h in adults and around 8 h in neonates.

Other: lower concentrations (0.25% and less) produce analgesia with increased motor sparing (desirable in outpatients and obstetrics). Commercial bupivacaine is a 50:50 racemic mixture of the R and S enantiomers. Cardiac arrest associated with bupivacaine is difficult to treat possibly due to its high protein binding and high lipid solubility (please see toxicity).

9. Ropivacaine:

Type: amide

Pka: 8.2

Protein binding: 94%

Characteristics: onset and duration similar to bupivacaine, with slight lesser potency. Elimination half-life 1-3 h in adults.

Like bupivacaine, it is chemically related to mepivacaine, but as opposed to most local anesthetics, it is supplied as the pure S enantiomer of the drug. The S enantiomer is associated with less cardiac toxicity, intermediate between that of lidocaine and bupivacaine.

Other: It is a weak vasoconstrictor (only one other than cocaine). At lower concentrations (less than 0.5%) it may show a greater selectivity for sensory than motor blockade than bupivacaine.

10. Levobupivacaine:

Type: amide

Pka: 8.1

Protein binding: 97%

Characteristics: S enantiomer of bupivacaine, very similar to ropivacaine.

Not available at this time in the US.

References

Lou L, Sabar R, Kaye A: Local Anesthetics, In: Raj P (ed), Textbook of Regional Anesthesia. New York, Churchill Livingstone, 2002, pp 177-213
Brown DL, Fink R: The History of Neural Blockade and Pain Management, In: Cousins MJ, Bridenbaugh PO (eds), Neural Blockade, 3^rd edition. Philadelphia, Lippincott-Raven, 1998, pp 3-32
Strichartz GR: Neural Physiology and Local Anesthetic Action, In: Cousins MJ, Bridenbaugh PO (eds), Neural Blockade, 3^rd edition. Philadelphia, Lippincott-Raven, 1998, pp 35-54
Tucker GT, Mather LE: Properties, Absorpton, and Disposition of Local Anesthetic Agents, In: Cousins MJ, Bridenbaugh PO (eds), Neural Blockade, 3^rd edition. Philadelphia, Lippincott-Raven, 1998, pp 55-95
Covino BG, Wildsmith JAW: Clinical Pharmacology of Local Anesthetic Agents, In: Cousins MJ, Bridenbaugh PO (eds), Neural Blockade, 3^rd edition. Philadelphia , Lippincott-Raven, 1998, pp 97-128
Morgan GE, Mikhail MS, Murray MJ: Clinical Anesthesiology, 4^th edition. New York, McGraw-Hill, 2006, pp 263-288
Mulroy MF: Regional Anesthesia, 3^rd edition. Philadelphia, Lippincott Williams & Wilkins, 2002, pp 1-63
Liu SS, Joseph RS: Local Anesthetics, In: Barash PG, Cullen BF, Stoelting RK (eds), Clinical Anesthesia. Philadelphia, Lippincott Williams & Wilkins, 2006, pp 453-471
DiFazio CA, Carron H, Grosslight KR, et al. Comparison of ph-adjusted lidocaine solutions for epidural anesthesia. Anesth Analg 1986; 65: 760-64
Hilgier M. Alkalinization of bupivacaine for brachial plexus block. Reg Anesth 1985;10: 59-61
Booker PD, Taylor C, Saba G. Perioperative changes in α1-acid glycoprotein concentrations in infants undergoing major surgery. Br J Anaesth 1996; 76: 365-368
Stoelting RK: Pharmacology and Physiology in Anesthetic Practice, 3^rd edition. Philadelphia , Lippincott-Raven, 1999, pp158-181
Tetzlaff JE: Local anesthetics and adjuvants for ambulatory anesthesia. In: Steele SM, Nielsen KC, Klein SM (eds), Ambulatory Anesthesia Perioperative Analgesia. New York, McGraw-Hill, 2005, pp 193-205
Weinberg GL et al. Lipid emulsion infusion rescues dogs from bupivacaine-induced cardiac toxicity. Reg Anesth Pain Med 2003; 28:198-202
Mayr VD, Raedler C, Wenzel V, et al. A comparison of epinephrine and vasopressin in a porcine model of cardiac arrest after rapid intravenous injection of bupivacaine. Anesth Analg 2004; 98:1426-3
Neal JM. Effects of epinephrine in local anesthetics on the central and peripheral nervous systems: Neurotoxicity and neural blood flow. Reg Anesth Pain Med 2003; 28:124-134
Rosenberg PH, Veering VT, Urmey WF. Maximum recommended doses of local anesthetics: A multifactorial concept. Reg Anesth Pain Med 2004; 29: 564-575.
Mulroy MF. Local anesthetics: Helpful science, but don’t forget the basic clinical steps (editorial). Reg Anesth Pain Med 2005; 30: 513-515.
Mather LE, Copeland SE, Ladd LA. Acute toxicity of local anesthetics: Underlying pharmacokinetic and pharmacodynamic concepts (A review article). Reg Anesth Pain Med 2005; 30: 553-566
Horlocker TT. One hundred years later, I can still make your heart stop and your legs weak: the relationship between regional anesthesia and local anesthetic toxicity. Reg Anesth Pain Med 2002; 27(6): 543-4
Mulroy MF. Systemic toxicity and cardiotoxicity from local anesthetics: Incidence and preventative measures (editorial). Reg Anesth Pain Med 2002; 27(6): 556-61
Horlocker TT, Wedel DJ. Local, anesthetic toxicity-Does product labeling reflect actual risk? Reg Anesth Pain Med 2002; 27(6): 562-567
Weinberg GL. Current concepts in resuscitation of patients with local anesthetic cardiac toxicity. Reg Anesth Pain Med 2002; 27(6): 568-575
Myer Leonard. Carl Koller: Mankind’s greatest benefactor? The story of local anesthesia. J Dent Res 1998; 77:535-8
De Jong R. Tumescent anesthesia: lidocaine dosing dichotomy. Int J Cosmetic Surg 2002; 4: 3-7
Nordstrom H, Stange K. Plasma lidocaine levels and risks after liposuction with tumescent anaesthesia. Acta Anaesthesiol Scand 2005; 49: 1487-1490
Rosenblatt MA, Abel M, Fischer GW, et al. Successful use of a 20% lipid emulsion to resuscitate a patient after a presumed bupivacaine-related cardiac arrest. Anesthesiology 2006; 105:217-8
Litz RJ, Popp M, Stehr SN, et al. Succesful resuscitation of a patient with ropivacaine-induced asystole after axillary plexus block using lipid infusion. Anaesthesia 2006; 61: 800-801
Rodriguez J, Quintela O, Lopez-Rivadulla M, et al. High doses of mepivacaine for brachial plexus block in patients with end-stage chronic renal failure. A pilot study. Eur J Anaesthesiol 2001; 18: 171-176
Kamibayashi T, Maze M. Clinical uses of a₂-adrenergic agonists. Anesthesiology 2000; 93:1345-9
Tanoubi I, Vialles N, Cuvillon P, et al. Systemic toxicity with mepivacaine following axillary block in a patient with terminal kidney failure. Ann Fr Anesth Reanim 2006; 25:33-5
McCutchen T, Gerancher JC. Early Intralipid Therapy May Have Prevented Bupivacaine-Associated Cardiac Arrest. Reg Anesth Pain Med 2008; 33: 178-180