TOXICITY

Local anesthetics are relatively safe if administered in an appropriate dosage and in the correct anatomic location. However, systemic and localized toxic reactions can occur, usually as a result of accidental intravascular or intrathecal injection or administration of an excessive dose. In addition, specific adverse effects are associated with the use of certain drugs, such as allergic reactions to the aminoester drugs and methemoglobinemia after the use of prilocaine.

Systemic Toxicity (also see Chapter 43 , Chapter 44 , Chapter 45 , and Chapter 58 )

Systemic reactions to local anesthetics primarily involve the CNS and the cardiovascular system. In general, the CNS is more susceptible to the actions of systemic local anesthetics than the cardiovascular system is, and thus the dose and blood level of local anesthetic required to produce CNS toxicity are usually lower than that resulting in circulatory collapse.

593

Central Nervous System Toxicity

The initial symptoms of local anesthetic-induced CNS toxicity are feelings of lightheadedness and dizziness, followed frequently by visual and auditory disturbances such as difficulty focusing and tinnitus. Other subjective CNS symptoms include disorientation and occasional feelings of drowsiness. Objective signs of CNS toxicity are usually excitatory in nature and include shivering, muscular twitching, and tremors initially involving muscles of the face and distal parts of the extremities. Ultimately, generalized convulsions of a tonic-clonic nature occur. If a sufficiently large dose or a rapid intravenous injection of a local anesthetic is administered, the initial signs of CNS excitation are rapidly followed by a state of generalized CNS depression. Seizure activity ceases, and respiratory depression and ultimately respiratory arrest may occur. In some patients, CNS depression without a preceding excitatory phase is seen, particularly if other CNS depressant drugs have been administered.

CNS excitation may be the result of an initial blockade of inhibitory pathways in the cerebral cortex by local anesthetics,^[120] but it can also result from the net stimulation of glutamate release, an excitatory amino acid neurotransmitter. Blockade of the inhibitory pathways allows facilitatory neurons to function in an unopposed fashion, which results in an increase in excitatory activity leading to convulsions. An increase in the dose of local anesthetic leads to inhibition of the activity of both inhibitory and facilitatory circuits and thereby results in a generalized state of CNS depression. Persistent glutamate release can result in receptor desensitization or transmitter depletion, both of which will depress the CNS.

In general, a correlation exists between the anesthetic potency and intravenous CNS toxicity of various drugs. For example, in cats the dose of intravenous procaine required to cause convulsions is approximately seven times greater than the convulsive dose of bupivacaine.^[121] However, bupivacaine is also approximately eight times more potent than procaine as a local anesthetic.^[122] Intravenous infusion studies in human volunteers have also demonstrated an inverse relationship between the intrinsic anesthetic potency of various drugs and the dosage required to induce premonitory signs of CNS toxicity.^{[122] [123]} Convulsions from an inadvertent intravenous bolus of local anesthetic can generally be terminated by small intravenous doses of benzodiazepines, such as midazolam, or by small intravenous doses of thiopental.

Respiratory or metabolic acidosis increases the risk of CNS toxicity from local anesthetics in animals and patients. In cats, the convulsive threshold of various local anesthetics was inversely related to the arterial PCO₂ level.^[121] An increase in PaCO₂ from 25 to 40 mm Hg to 65 to 81 mm Hg decreases the convulsive threshold of procaine, mepivacaine, prilocaine, lidocaine, and bupivacaine by approximately 50%.

An elevation in PaCO₂ enhances cerebral blood flow, and as a result, anesthetic is delivered more rapidly to the brain. In addition, diffusion of carbon dioxide into neuronal cells decreases intracellular pH, which facilitates conversion of the base form of the drugs to the cationic form. The cationic form does not diffuse well across the nerve membrane, so ion trapping will occur, which will increase the apparent CNS toxicity of local anesthetics.

Hypercapnia or acidosis (or both) also decreases the plasma protein binding of local anesthetics.^{[124] [125]} Accordingly, an elevation in PaCO₂ or a decrease in pH will increase the proportion of free drug available for diffusion into the brain. On the other hand, acidosis increases the cationic form of the local anesthetic, which should decrease the rate of diffusion through lipoid barriers.

The clinical implication of this effect of hypercapnia and acidosis on toxicity deserves emphasis. Seizures produce hypoventilation and a combined respiratory and metabolic acidosis, which further exacerbates the CNS toxicity. In the setting of local anesthetic toxic reactions, it is essential to provide prompt assisted ventilation and circulatory support as needed to prevent or correct hypercapnia and acidosis, as well as prevent or correct hypoxemia, which also exacerbates CNS toxicity. Based on the discussion just presented, it should be apparent that clinicians performing major conduction blockade should make a routine practice of having the following ready at hand: monitoring equipment; an oxygen tank or wall oxygen outlet; airway equipment, including at minimum a bag-mask circuit for delivery of positive-pressure ventilation; and drugs to terminate convulsions, such as midazolam, lorazepam, diazepam, or thiopental.

Cardiovascular System Toxicity

Local anesthetics can exert direct actions on both the heart and peripheral blood vessels, as well as indirect actions on the circulation by blockade of sympathetic or parasympathetic efferent activity.

Direct Cardiac Effects

The primary cardiac electrophysiologic effect of local anesthetics is a decrease in the rate of depolarization in the fast-conducting tissues of Purkinje fibers and ventricular muscle.^{[120] [121]} This reduction in rate is believed to be due to a decrease in the availability of fast sodium channels in cardiac membranes. Action potential duration and the effective refractory period are also decreased by local anesthetics.^[126] However, the ratio of the effective refractory period to the duration of the action potential is increased in both Purkinje fibers and ventricular muscle.

The electrophysiologic effects of various agents differ qualitatively. Bupivacaine depresses the rapid phase of depolarization (V̇max) in Purkinje fibers and ventricular muscle to a greater extent than lidocaine does.^[126] In addition, the rate of recovery from a use-dependent block is slower in bupivacaine-treated papillary muscles than in lidocaine-treated muscles.^[127] This slow rate of recovery results in incomplete restoration of Na⁺ channel availability between action potentials, particularly at high heart rates. In contrast, recovery from lidocaine is complete, even at rapid heart rates. These differential effects of lidocaine and bupivacaine have been advanced as explanations of the antiarrhythmic properties of lidocaine and the arrhythmogenic potential of bupivacaine.

Electrophysiologic studies in intact dogs and in humans have shown that high blood levels of local anesthetics will prolong the conduction time through various

594

All local anesthetics exert a dose-dependent negative inotropic action on cardiac muscle.^{[128] [129]} This depression of cardiac contractility is proportional to the conduction-blocking potency of the various drugs in isolated nerves^[129] and in dogs^{[130] [131] [132]} ( Table 14-11 ). Thus, bupivacaine and tetracaine are more potent cardiodepressants than lidocaine, which in turn is a more potent cardiode-pressant than chloroprocaine.

Local anesthetics may depress myocardial contractility by affecting calcium influx and triggered release. For example, procaine blocks the intracellular release of calcium in isolated sarcoplasmic reticulum preparations.^[133] However, in the isolated guinea pig heart, an increase in the extracellular concentration of calcium failed to reverse the negative inotropic action of bupivacaine or lidocaine. ^[134]

Voltage-clamp studies show that lidocaine inhibits cardiac sarcolemmal Ca²⁺ currents as well as Na⁺ currents.^[135] This action alone should be antagonized by an increase in extracellular Ca²⁺ , so it is likely that the negative inotropy of local anesthetics involves several mechanisms, not just the blockade of inward currents.

Direct Peripheral Vascular Effects

Local anesthetics exert a biphasic effect on peripheral vascular smooth muscle. Low concentrations of lidocaine and bupivacaine produce vasoconstriction in the cremaster muscle of rats, whereas high concentrations increase arteriolar diameter, indicative of vasodilation.

In vivo studies have also demonstrated that small doses of local anesthetics decrease peripheral arterial flow without any change in arterial blood pressure whereas larger doses increase blood flow.^[136] Cocaine is the only local anesthetic that consistently causes vasoconstriction at all concentrations because of its ability to inhibit the uptake of norepinephrine by premotor neurons and thus potentiate neurogenic vasoconstriction.^{[137] [138]}

Some studies in anesthetized animals and in isolated heart-lung preparations have found increases in pulmonary vascular resistance after infusion of local anesthetic.^{[131] [132]} The addition of lidocaine to cultured smooth muscle cells acutely elevates intracellular Ca²⁺ and could be a mechanism for local vasoconstriction. It has not been determined to what degree these pulmonary vascular effects in vivo are direct actions on pulmonary vascular smooth muscle or to what degree they reflect responses to circulatory or respiratory depression from drug acting on the CNS.