The clinical efficacy and broad applicability of local anesthetics are inseparable from an equally thorough understanding of their toxicity. Local anesthetic systemic toxicity (LAST) remains one of the most feared and potentially lethal complications of regional anesthesia and procedural medicine, and its recognition, prevention, and management constitute core competencies for any clinician who performs procedures with local anesthetics, including not only anesthesiologists and regional anesthesia specialists, but emergency physicians, surgeons, dentists, dermatologists, and proceduralists across every specialty.¹ Beyond LAST, local anesthetics carry a spectrum of adverse effects ranging from the immunologically mediated (true allergic reactions, predominantly ester-class), to the hematologic (methemoglobinemia from benzocaine and prilocaine), to the neurotoxic (transient neurologic symptoms from intrathecal lidocaine). This module addresses each of these toxicity domains in depth, provides evidence-based guidance on prevention and treatment, and applies this framework to the special populations, including pregnant patients, neonates and pediatric patients, patients with hepatic disease, and patients with significant cardiac disease, in whom baseline pharmacokinetic and pharmacodynamic differences alter risk profiles and mandate dose modification or agent substitution.

Recognition of patient- and procedure-specific risk factors for LAST is the foundation of prevention, enabling the clinician to modify technique, reduce dose, adjust the choice of agent, and ensure appropriate monitoring and rescue resources are available before the block is performed.¹ Patient-related risk factors include extremes of age (neonates and elderly patients have reduced protein binding, altered volume of distribution, and impaired hepatic clearance of amide local anesthetics), hepatic disease (reducing amide local anesthetic clearance and potentially reducing α₁-acid glycoprotein synthesis, elevating the free fraction), cardiac disease with reduced cardiac output (reducing hepatic blood flow and local anesthetic clearance, and reducing the dilutional buffering that normally limits peak arterial concentrations after bolus injection), low muscle mass and lean body mass (reducing the volume of distribution for lipid-soluble agents), and pregnancy (reducing protein binding due to dilutional hypoproteinemia and altering α₁-acid glycoprotein levels).^1,3

Procedure-related risk factors include injection at highly vascular sites (intercostal, paracervical, caudal epidural), the use of large volumes of concentrated local anesthetic, continuous catheter techniques delivering drug over extended periods, and multiple sequential nerve blocks performed in a single session without accounting for cumulative dose. The risk of intravascular injection is technique-dependent: real-time ultrasound visualization of needle tip position and continuous aspiration before and during injection have significantly reduced (though not eliminated) the risk of intravascular placement, while hydrodissection with small initial aliquots (2–3 mL test doses) before the full volume allows detection of intravascular placement before the full toxic dose is delivered.¹

The initial management of LAST follows the same general priority structure as any acute cardiovascular or neurologic emergency: airway, breathing, and circulation, with simultaneous administration of specific rescue therapy.¹ If CNS excitation or seizure occurs, securing the airway with rapid sequence intubation is the priority, both to prevent hypoxic injury and to facilitate hyperventilation, which raises pH and reduces the fraction of ionized (more toxic) local anesthetic at the Nav channel binding site. Benzodiazepines (midazolam 1–2 mg IV, or diazepam 5–10 mg IV) are the preferred agents for seizure termination because they avoid the cardiac depression and negative inotropy associated with propofol and thiopental; however, if benzodiazepines are unavailable, small doses of propofol (0.5–1 mg/kg, avoiding doses that produce further cardiovascular depression) are acceptable for seizure termination while lipid emulsion is being prepared.¹ Succinylcholine or rocuronium for neuromuscular blockade will terminate the muscular manifestations of seizures but does not terminate the underlying cortical seizure activity and provides no protection from cerebral injury from the seizure itself.

Intravenous lipid emulsion (ILE) therapy is the cornerstone of specific pharmacologic treatment for LAST and has transformed the management of severe local anesthetic toxicity since its introduction into clinical practice following landmark animal studies by Weinberg and colleagues in the late 1990s and early 2000s.⁴ The mechanism by which ILE reverses local anesthetic toxicity is not fully understood and likely involves multiple pathways. The "lipid sink" theory, the most widely cited mechanism, proposes that the intravenously administered lipid emulsion creates a rapidly expanding intravascular lipid phase that sequesters lipid-soluble local anesthetic molecules (particularly bupivacaine, which has very high lipid solubility) from the aqueous plasma compartment where they exert toxicity on cardiac and CNS tissue, reducing the free concentration available to act on ion channels.⁴ Additional proposed mechanisms include direct positive inotropic effects of the lipid emulsion on cardiac mitochondrial function (correcting the bupivacaine-induced impairment of fatty acid oxidation in cardiac myocytes), and restoration of the sodium gradient through mechanisms not dependent on lipid partitioning.

Whatever the precise mechanism, the clinical evidence for ILE efficacy in reversing bupivacaine-induced cardiovascular collapse is compelling: multiple animal models demonstrate consistent reversal of bupivacaine toxicity with ILE that is superior to epinephrine alone, and numerous human case reports and case series document resuscitation from LAST that had been refractory to standard Advanced Cardiac Life Support (ACLS) measures, following ILE administration.⁴

The American Society of Regional Anesthesia and Pain Medicine (ASRA)-recommended protocol for ILE therapy in LAST is as follows. Intralipid 20% (lipid emulsion 20%) is given as an IV bolus of 1.5 mL/kg (approximately 100 mL for a 70 kg adult) over 2–3 minutes, followed immediately by an infusion at 0.25 mL/kg/min. If cardiovascular stability is not achieved or deterioration recurs, additional boluses of 1.5 mL/kg may be repeated up to twice at 5-minute intervals, and the infusion rate may be doubled to 0.5 mL/kg/min. The infusion is continued for at least 10 minutes after hemodynamic stability is achieved. The recommended upper limit of total lipid dose is approximately 10–12 mL/kg in the first 30 minutes, beyond which the risk of lipid-related complications (lipemia, interference with laboratory assays, potential cardiac depression from lipid overload) is not offset by additional benefit.¹ Critically, ILE should be available wherever local anesthetics are used in significant quantities, including operating rooms, interventional radiology suites, procedure rooms, emergency departments, and any outpatient facility performing regional anesthesia, as a preemptive resource rather than a response that must be located and transported after the event.

Standard ACLS resuscitation is initiated simultaneously with ILE therapy for LAST with cardiovascular compromise. However, several important modifications apply to LAST-specific resuscitation that distinguish it from standard cardiac arrest management.¹ Epinephrine should be used in reduced doses (10–100 μg IV boluses, rather than the standard 1 mg ACLS dose), because large doses of epinephrine in the context of bupivacaine toxicity may worsen ventricular dysrhythmias by increasing myocardial excitability in tissue with impaired conduction recovery, a phenomenon documented in animal models. Vasopressin is not recommended in LAST-related cardiac arrest. Amiodarone should be avoided for ventricular dysrhythmias in the context of bupivacaine toxicity because it adds further sodium and potassium channel blockade to already-compromised cardiac conduction.¹

The goal of resuscitation in LAST should include the recognition that bupivacaine-induced cardiovascular collapse may require prolonged cardiopulmonary resuscitation (CPR), sometimes for 30–60 minutes or longer, while drug redistribution and lipid sequestration reduce myocardial bupivacaine concentrations to sub-toxic levels. Cardiopulmonary bypass or extracorporeal membrane oxygenation (ECMO) has been used successfully to support circulation during refractory LAST while awaiting drug redistribution, and the availability of these resources should be considered when planning high-risk regional anesthetic procedures in facilities capable of providing extracorporeal support.⁴

As introduced in CNS-02, methemoglobinemia is the conversion of functional oxyhemoglobin (Fe²⁺) to non-functional methemoglobin (Fe³⁺) by oxidizing agents, of which benzocaine and prilocaine are the clinically most important local anesthetic examples. Methemoglobin cannot bind or transport oxygen, and its presence in the circulation produces both functional anemia (reduced oxygen-carrying capacity) and leftward shift of the oxygen-hemoglobin dissociation curve of remaining oxyhemoglobin (reduced oxygen release to tissues), a combination that renders methemoglobinemia disproportionately impairing relative to an equivalent reduction in hemoglobin from true anemia.⁶ The severity of clinical manifestations depends on the methemoglobin fraction: at 10–20%, cyanosis is visible but symptoms may be mild; at 20–30%, dyspnea, headache, and fatigue develop; at 30–50%, significant neurologic compromise including confusion, obtundation, and cardiovascular instability emerges; above 50–70%, the syndrome is life-threatening.

The clinical recognition of methemoglobinemia is aided by its characteristic presentation: central cyanosis that does not improve or worsens with supplemental oxygen (because the problem is not a lack of inspired oxygen but the inability of the hemoglobin to carry it), and pulse oximetry readings that plateau at approximately 85% regardless of the true methemoglobin fraction (because the standard two-wavelength pulse oximeter cannot distinguish methemoglobin from oxyhemoglobin at 940 nm and misidentifies it as approximately 85% saturation). Co-oximetry, available on arterial blood gas analyzers, directly measures the methemoglobin fraction and is the diagnostic test of choice whenever methemoglobinemia is suspected.⁶

Treatment for clinically significant methemoglobinemia (methemoglobin fraction >20–25%, or lower if the patient is symptomatic or has underlying cardiopulmonary disease) is methylene blue 1–2 mg/kg IV administered over 5–10 minutes. Methylene blue is reduced by NADPH (generated by the hexose monophosphate shunt) to leukomethylene blue, which in turn donates electrons to the NADPH methemoglobin reductase system, accelerating the enzymatic reduction of Fe³⁺ back to Fe²⁺ at rates far exceeding the normal spontaneous reduction rate.⁶ The response to methylene blue is typically rapid, with cyanosis resolving within 15–30 minutes, and a repeat dose may be given if the response is inadequate. Methylene blue is ineffective and contraindicated in glucose-6-phosphate dehydrogenase (G6PD)-deficient patients because NADPH generation requires an intact hexose monophosphate shunt; in these patients, alternative therapies including ascorbic acid (which can reduce methemoglobin by an NADPH-independent mechanism, albeit more slowly) and exchange transfusion are used. Methylene blue itself is an oxidizing agent at doses above 7 mg/kg and can paradoxically worsen methemoglobinemia at excessive doses, a practical reminder that careful dosing is essential.⁶

Pregnant patients at term present a constellation of pharmacokinetic and pharmacodynamic changes that significantly alter local anesthetic behavior and risk profile.³ Plasma protein binding is reduced by dilutional hypoproteinemia, with albumin falling by approximately 15–20% at term, elevating the free fraction of highly protein-bound agents such as bupivacaine and increasing the concentration available to cross the placenta and exert systemic effects. α₁-Acid glycoprotein (AAG), the primary binding protein for basic local anesthetics, is actually elevated in the neonate and reduced relative to adult levels in the fetus, which may contribute to higher unbound fetal drug concentrations despite the lower total fetal plasma levels measured in studies. The engorged epidural venous plexus, secondary to aortocaval compression and elevated inferior vena cava pressure, reduces subarachnoid and epidural space volume, increasing the cephalad spread of both spinal and epidural local anesthetics for a given dose and mandating dose reduction compared to non-pregnant patients.³

The increased cardiac output of pregnancy (30–50% above baseline at term) means that systemic absorption of local anesthetic produces higher peak concentrations more rapidly, and the reduced functional residual capacity of term pregnancy reduces the oxygen reserve available during any period of respiratory compromise from local anesthetic-induced central depression or high spinal. These collective changes increase the vulnerability of pregnant patients to LAST and high spinal block, making careful dose titration, particularly for epidural dosing, and meticulous technique imperative in obstetric regional anesthesia.

Neonates present a pharmacologically distinct population for local anesthetic management, with differences in every pharmacokinetic domain that are clinically meaningful in the context of regional anesthetic dosing.⁸ The apparent volume of distribution for local anesthetics is larger in neonates relative to body weight, reflecting the higher total body water content and lower fat and protein content of the neonate. Hepatic metabolic capacity for amide local anesthetics is substantially reduced in the first weeks of life; CYP3A4 (cytochrome P450 3A4) and CYP1A2 (cytochrome P450 1A2) activity reach adult levels only by 3–12 months of age, resulting in prolonged plasma half-lives and reduced clearance; lidocaine half-life in neonates is approximately 3 hours compared to 1.5 hours in adults, and accumulation during continuous infusion or repeated dosing is more likely. α₁-Acid glycoprotein (AAG) levels are approximately 50% of adult levels in neonates, reducing protein binding and elevating free drug fractions. Neonatal hemoglobin F is more susceptible to oxidation than adult hemoglobin A, making neonates particularly vulnerable to methemoglobinemia from benzocaine and prilocaine, as discussed in CNS-02.

The practical implication of these differences is that maximum doses of local anesthetics in neonates and infants are lower than in older children and adults on a per-kilogram basis, continuous infusion rates must be reduced, and the duration of monitoring after regional techniques should be extended to account for prolonged drug clearance.⁸ Caudal epidural anesthesia, the most commonly performed regional technique in pediatric anesthesia, uses bupivacaine 0.25% at 1 mL/kg (maximum 20 mL) as the standard single-shot formulation for perineal and lower extremity procedures; ropivacaine 0.2% at 1 mL/kg is increasingly preferred for its reduced motor block and marginally better cardiac safety profile.

Amide local anesthetics depend entirely on hepatic metabolism for elimination, and hepatic disease at all levels of severity reduces clearance to a clinically relevant degree.⁹ Hepatic disease affects local anesthetic pharmacokinetics through two mechanisms: reduced hepatic enzyme capacity (from hepatocellular damage reducing CYP3A4 and CYP1A2 activity) and reduced hepatic blood flow (from portal hypertension, intrahepatic shunting, and reduced cardiac output in advanced cirrhosis). Both mechanisms reduce clearance, prolong plasma half-life, and increase steady-state plasma concentrations during repeated or continuous dosing. Additionally, advanced hepatic disease reduces albumin synthesis and AAG levels, reducing protein binding and elevating the free fraction of local anesthetic, further increasing toxicity risk at any given total plasma concentration. In patients with Child-Pugh class B or C cirrhosis, amide local anesthetic doses should be reduced by 25–50% from standard recommendations for any technique involving repeated dosing or continuous infusion. Single-shot peripheral nerve blocks are generally safer in hepatic disease than continuous catheter techniques because the systemic exposure is time-limited; the lowest effective dose should be used and monitoring should be extended.⁹

Ester local anesthetics, which are hydrolyzed by plasma pseudocholinesterase and tissue esterases rather than hepatic CYP enzymes, are not significantly affected by hepatic parenchymal disease at clinical doses, though pseudocholinesterase synthesis may be mildly reduced in advanced cirrhosis. For patients requiring topical or infiltration anesthesia in the context of severe hepatic disease, chloroprocaine or procaine avoids the hepatic metabolism concern entirely.

Patients with significant cardiovascular disease, including reduced ventricular function, significant coronary artery disease, cardiac conduction abnormalities, or cardiac device implantation, require particular care with local anesthetic selection and dosing.¹⁰ The reduced cardiac output of heart failure reduces hepatic blood flow and amide local anesthetic clearance, predisposing to accumulation with repeated dosing. The sodium channel blockade that constitutes the therapeutic mechanism of local anesthetics can interact additively with the sodium channel effects of antiarrhythmic drugs (particularly Class I agents such as flecainide, mexiletine, and the sodium channel-blocking properties of amiodarone), reducing the threshold for cardiac conduction toxicity. Patients with pre-existing prolonged QRS complex (QRS) or QTc intervals, or with bundle branch block, are at higher baseline risk of ventricular dysrhythmia from any sodium channel-blocking drug, including local anesthetics at toxic concentrations. In these patients, the selection of ropivacaine over racemic bupivacaine for large-volume or long-duration regional anesthetic procedures represents a rational pharmacologic precaution, and the availability of lipid emulsion and resuscitation resources should be assured before the block is performed.¹⁰

The clinical management of regional anesthesia in patients with implanted pacemakers or ICDs requires consideration of whether the electrical activity of the device may be affected by the electromagnetic noise of nerve stimulators used during block placement (a consideration largely obviated by ultrasound guidance) and whether the device's bradycardia-pacing function will maintain cardiac output if significant bradycardia results from a high spinal or thoracic epidural sympathectomy.

Renal disease does not directly impair the hepatic metabolism of amide local anesthetics, and for single-shot peripheral nerve blocks or single-dose neuraxial anesthesia in patients with chronic kidney disease, dose adjustment is generally not required. However, renal failure has two indirect pharmacokinetic consequences that become clinically relevant during continuous infusions or repeated dosing. First, amide local anesthetic metabolites, including monoethylglycinexylidide (MEGX) and glycinexylidide (GX) from lidocaine, and 3-hydroxy ropivacaine and 2′,6′-pipecoloxylidide (PPX) from ropivacaine, are cleared by renal excretion as glucuronide conjugates. In severe renal failure (estimated glomerular filtration rate below 15 mL/min) or in patients on dialysis, these conjugates accumulate; MEGX has approximately 25–80% of the pharmacologic activity of lidocaine, meaning that metabolite accumulation during prolonged lidocaine infusion in dialysis patients can contribute to CNS effects even when the parent drug concentration appears acceptable.¹

Second, renal failure commonly produces hypoalbuminemia and altered AAG levels, reducing protein binding and elevating the free fraction of local anesthetic. In a patient with nephrotic syndrome and serum albumin of 1.8 g/dL, the free fraction of bupivacaine may be substantially higher than predicted from the nominal total plasma concentration, shifting the effective toxic threshold downward. The practical recommendation for patients with significant renal impairment (chronic kidney disease (CKD) stage 4–5 or on dialysis) undergoing continuous perineural catheter infusions is to use reduced infusion rates, monitor for early CNS symptoms, and consider ropivacaine over bupivacaine given ropivacaine’s somewhat lower intrinsic cardiotoxicity. Ester agents remain unaffected by renal function for practical purposes, since their hydrolysis occurs in plasma and tissue rather than hepatically, and the resulting metabolites are water-soluble and excreted without pharmacologic consequence in any degree of renal insufficiency.¹

Safe practice requires that the clinician calculate and track the total local anesthetic dose against established maximum dose guidelines before any significant regional anesthetic procedure. The following reference values represent current widely cited recommendations and should be understood as population-based estimates requiring individualization based on site of injection, patient characteristics, and clinical context:³ Lidocaine: 4.5 mg/kg without epinephrine (maximum 300 mg), 7 mg/kg with epinephrine (maximum 500 mg). Bupivacaine: 2.5 mg/kg without epinephrine (maximum 175 mg), 3 mg/kg with epinephrine (maximum 225 mg). Ropivacaine: 3 mg/kg without epinephrine (maximum 200 mg), 4 mg/kg with epinephrine (maximum 250 mg). Mepivacaine: 5 mg/kg without epinephrine (maximum 400 mg), 7 mg/kg with epinephrine (maximum 550 mg). Chloroprocaine: 11 mg/kg without epinephrine (maximum 800 mg), 14 mg/kg with epinephrine (maximum 1000 mg). Prilocaine: 6 mg/kg without epinephrine (maximum 400 mg), 8.5 mg/kg with epinephrine (maximum 600 mg), the methemoglobinemia threshold at approximately 600 mg total dose in healthy adults. These limits should be reduced by 20–30% in elderly patients, neonates, patients with hepatic disease, and patients with low lean body mass, and dose calculations should always be based on lean body weight rather than total body weight in obese patients.

Tumescent anesthesia refers to the subcutaneous infiltration of very large volumes of highly dilute lidocaine solution containing epinephrine, used primarily in liposuction, dermatologic procedures, and body-contouring surgery. The technique was developed by Klein in the late 1980s and involves infiltrating volumes of 1–3 liters or more of solution containing lidocaine at concentrations of 0.05–0.1% (500–1000 mg/L) with epinephrine at 1:1,000,000 concentration (1 mg/L).¹ The key pharmacologic principle underlying the safety of tumescent anesthesia at doses that would be catastrophic by other routes is the combination of extreme dilution, subcutaneous location (the least vascular injection site in the absorption hierarchy), and high-concentration epinephrine-mediated vasoconstriction. These factors together produce an extraordinarily slow systemic absorption profile: peak plasma lidocaine concentrations after tumescent infiltration may not be reached for 12–14 hours after injection, compared to minutes for peripheral nerve block injection, and the peak concentrations achieved with even large tumescent doses remain below the CNS toxic threshold when the technique is performed correctly.

The Klein formula for maximum tumescent lidocaine dose is 35–55 mg/kg, a limit that represents a fundamentally different paradigm from the 4.5–7 mg/kg limits applicable to conventional injection routes, and reflects the unique pharmacokinetics of the tumescent compartment rather than a relaxation of safety standards.¹ These limits apply specifically to the tumescent technique with epinephrine 1:1,000,000; they do not apply to subcutaneous lidocaine without epinephrine, to any injection technique with higher vascularity, or to combinations of tumescent and other forms of local anesthetic administration in the same session. A clinician who adds a peripheral nerve block to a patient who has already received 35 mg/kg of tumescent lidocaine is now operating beyond the pharmacokinetic safety model that justifies the tumescent limit, because the nerve block lidocaine is absorbed by a more vascular route and reaches systemic circulation on a much faster timeline than the tumescent component.

The risk of LAST in tumescent anesthesia is real despite the favorable pharmacokinetics, and occurs primarily under two circumstances: when the total milligram dose substantially exceeds the Klein limits, and when epinephrine at 1:1,000,000 is omitted or substituted at a lower concentration, eliminating the vasoconstriction that is the principal determinant of slow absorption. Systemic epinephrine effects from tumescent infiltration (tachycardia, hypertension, anxiety, tremor) are common and must be distinguished from early LAST. Because the absorption delay means that peak lidocaine concentrations arrive hours after the procedure has ended, LAST from tumescent anesthesia may present in the recovery area or even after the patient has been discharged from an outpatient facility, a pattern fundamentally different from the immediately post-injection LAST seen with peripheral nerve blocks. Monitoring must extend beyond the procedural period, and facilities performing tumescent liposuction should have protocols for extended post-procedure monitoring and LAST rescue resources including lipid emulsion immediately available.¹