E: CARDIOPULMONARY MEDICATIONS

HISTORY AND EPIDEMIOLOGY

The term dysrhythmia encompasses an array of abnormal cardiac rhythms that range in clinical significance from merely annoying to instantly life threatening. The word arrhythmia can be defined as a lack of rhythm. Some references state they are synonyms for irregular heartbeat, and others define arrhythmia as irregular heartbeat and dysrhythmia as a disturbance to an otherwise normal rhythm. Throughout the text, we have chosen the term dysrhythmia. Antidysrhythmics include all medications that are used to treat any of these various dysrhythmias. The importance of dysrhythmia management in the modern practice of medicine cannot be overstated because dysrhythmias are among the most common causes of preventable sudden cardiac death.⁵²

For many years, antidysrhythmics were considered among the most rational of the available cardiac medications. This well-earned reputation related to their high efficacy at reducing the incidence of malignant dysrhythmias. Similarly, they are effective at controlling extrasystoles which cause patient discomfort or anxiety. However, this approach changed dramatically following publication of the Cardiac Arrhythmia Suppression Trials (CAST and CAST II)²⁸ and more recently with the rise of mechanical interventions, such as ablation therapy and implantable defibrillators. Cardiac Arrhythmia Suppression Trials assessed the ability of three antidysrhythmics to suppress asymptomatic ventricular dysrhythmias considered to be harbingers of sudden death. The original CAST was discontinued in 1989 before completion, when encainide and flecainide, two of the study medications, not only failed to prevent sudden death but actually increased overall mortality. The CAST II noted similar problems with moricizine.¹⁰² It is clear that the enhanced mortality associated with many antidysrhythmics is a result of their prodysrhythmogenic effects and that virtually all medications of this group carry such risk. Because most patients with atrial fibrillation do not benefit from rhythm conversion compared with control of the ventricular response rate, the use of antidysrhythmics for this indication is less common but is still useful in some populations.⁹¹

In addition to the predictable, mechanism-based adverse effect of each medication, unique and often unanticipated effects also occur.⁸⁷ Experience with overdose of many of these medications is limited, and management is generally based on the underlying pharmacologic principles, existing case reports, and the experimental literature. This chapter focuses on the medications that serve primarily as antidysrhythmics and, with the exception of lidocaine (found in Chap. 64), have few other medicinal indications. Chap. 15 provides a more detailed description of the electrophysiology of dysrhythmias and a discussion of their genesis. In addition, the toxicities from β-adrenergic antagonists and Ca²⁺ channel blockers, which have indications in addition to dysrhythmia control, are discussed separately in Chaps. 59 and 60. The toxicology of cardioactive steroids such as digoxin is discussed in Chap. 62.

CLASSIFICATION OF ANTIDYSRHYTHMICS

Despite an incomplete understanding of the underlying mechanisms of dysrhythmia formation, an abundance of antidysrhythmics were developed, each attempting to alter specific electrophysiologic components of the cardiac impulse generating or conducting system.

Antidysrhythmics modify impulse generation and conduction by interacting with various membrane Na⁺, K⁺, and Ca²⁺ channels. Generally, antidysrhythmics manifest electrophysiologic effects either through alteration of the channel pore or, more commonly, by modification of its gating mechanism (Fig. 57–1). Unfortunately, given their exceedingly complex mechanisms of action, the descriptive terms used to explain their molecular actions are not always completely accurate. For example, the description of an antidysrhythmic as a specific “channel blocker,” although representative of the conceptual action of that medication, is inaccurate because in most cases, the molecule does not actually block the channel but rather prevents the channel from opening or closing properly. Furthermore, many of these medications are active nonspecifically at other channels or on other cells, resulting in divergent clinical actions of those similarly classified.

The Vaughan-Williams classification of antidysrhythmics by electrophysiologic properties emphasizes the connection between the basic electrophysiologic actions and the antidysrhythmic effects.¹⁰³ Although initially proposed as a descriptive model for electrophysiologic actions and not for clinical effects, the Vaughan-Williams classification is commonly invoked as a user-friendly guide to clinical therapy. A more rational classification would match the electrophysiologic effects of the antidysrhythmics with their molecular interactions on different regions of the various ion channels, such as channel gating and pore conductance.

This discussion of antidysrhythmics uses the Vaughan Williams classification, recognizing its shortcomings.¹⁰⁹ The pharmacokinetic properties of the various medications are summarized in Table 57–1.

CLASS I ANTIDYSRHYTHMICS

All antidysrhythmics in Vaughan-Williams class I (A, B, and C) alter Na⁺ conductance through cardiac voltage-gated, fast inward Na⁺ channels (Table 57–1). These xenobiotics bind to the Na⁺ channels and slow their recovery from the open or inactivated state to the resting state (Fig. 57–1). This conversion must occur before the channel can reopen and participate in another depolarization. Consequently, as the proportion of medication-bound Na⁺ channels increases, fewer of these channels are capable of reactivation on the arrival of the next depolarizing impulse. As a result, by reducing the excitability of the myocardium, abnormal rhythms are both prevented and terminated.

Blockade of these Na⁺ channels slows the rise of phase 0 of the cellular action potential, which correlates with a reduction in the rate of depolarization of the myocardial cell (or V_max). Similarly, conduction through the myocardium is slowed, producing a measurable prolongation of the QRS complex on the surface electrocardiogram (ECG). Correspondingly, slowed intramyocardial conduction is associated with reduced contractility, manifesting as negative inotropy. Myocardial depression also results from effects of reduced intracellular Na⁺ on the Na⁺-Ca²⁺ exchange mechanism.⁷³ This in turn reduces the intracellular Ca²⁺ concentration, which is required for adequate contractility.

The differences among class I antidysrhythmics are directly related to their pharmacologic relationships with the Na⁺ channel. However, it is noteworthy that the original subdivision of class I antidysrhythmics was based on clinical observations, not current pharmacologic awareness, accounting for the somewhat illogical ordering of the class I subdivisions.¹⁰⁹ Type IB antidysrhythmics have their highest affinity for inactivated Na⁺ channels. This occurs at the end of depolarization, during early repolarization, and during periods of myocardial ischemia, all situations in which the myocardium is partially depolarized. These xenobiotics also have rapid “on–off” binding kinetics (rapid τ_recovery) and are thus bound only briefly, during late electrical systole, the period during which the Na⁺ channels are predominantly in the inactivated form. They are almost exclusively unbound during electrical diastole, which is the major portion of the cardiac cycle at normal heart rates. However, the degree of binding increases as the heart rate accelerates because the duration of diastole decreases, and the relative proportion of time spent in systole increases; this is termed use dependence. Because all IB antidysrhythmics do not bind to activated Na⁺ channels, in therapeutic doses they do not affect the rate of rise of phase 0 of the action potential, or V_max, and have no effect on the ECG. Alternatively, the class IC antidysrhythmics either act preferentially on activated Na⁺ channels or they release from the Na⁺ channels very slowly (slow τ_recovery) and thus are still bound during the next cardiac cycle. This prolonged channel blockade and reduced channel reactivation results in both greater pharmacologic effects and toxicity, even at slow heart rates. These xenobiotics reduce V_max and prolong the QRS complex. Class IA antidysrhythmics fall between the other two subclasses.

Although in the Vaughan-Williams classification all class I antidysrhythmics are considered primarily Na⁺ channel blockers, many, particularly those in class IA, have important effects on cardiac K⁺ channels. These channels are critical to maintenance of the cardiac action potential and repolarization of the myocardial cell. Slowing of K⁺ efflux prolongs the duration of the action potential and accounts for the persistence of refractoriness, or the time during which the cell is incapable of redepolarization. This effect produces QT interval prolongation and predisposes to the triggering of polymorphic ventricular tachycardia (torsade de pointes).⁵³ Because class IB antidysrhythmics have no effect on myocardial K⁺ channels, they do not alter refractoriness or the QT interval. In fact, class IB antidysrhythmics often reduce the action potential duration, shortening refractoriness. Further discussion of K⁺ channel blockade is found in Chap. 15 and later in the discussion of class III antidysrhythmics.

Class IA Antidysrhythmics: Procainamide, Quinidine, and Disopyramide

Procainamide

Procainamide is used to suppress either atrial or ventricular tachydysrhythmias (Fig. 57–2). Importantly, procainamide undergoes hepatic biotransformation by acetylation to N-acetylprocainamide (NAPA), the rate of which is genetically determined, and rapid acetylators are prone to developing toxicity.⁷⁶ Although NAPA lacks the Na⁺ channel–blocking activity of procainamide, it prolongs the action potential duration through blockade of the K⁺ rectifier currents and is independently responsible for the production of cardiac dysrhythmias.⁴⁵

Rapid intravenous (IV) dosing of procainamide is potentially dangerous because its initial volume of distribution is smaller than its final volume of distribution. Because this initial compartment includes the heart, adverse myocardial effects occur unexpectedly. In addition, procainamide exhibits ganglionic-blocking properties, which produce profound hypotension with rapid IV administration. Thus, to prevent toxicity during medication infusion, the IV loading dose is administered by slow infusion with ECG monitoring. Both procainamide and NAPA are renally eliminated and accumulate in patients with chronic kidney disease (CKD).⁶⁷

Although the chronic use of procainamide is commonly accompanied by the development of antinuclear antibodies and infrequently progresses to drug-induced systemic lupus erythematosis,⁴⁶ this syndrome is not associated with acute poisoning. Furthermore, NAPA has less propensity than procainamide to produce this syndrome.⁴⁵ Other reported adverse effects include seizures and antimuscarinic effects with acute overdose and myopathic pain, hepatitis, thrombocytopenia, and agranulocytosis after long-term use.

Serum procainamide concentrations should be determined as part of therapeutic drug monitoring (therapeutic: 4–12 mcg/mL); NAPA concentrations should also be monitored in patients with CKD (therapeutic: 10–20 mcg/mL). In patients with procainamide overdose, both procainamide and NAPA concentrations should be obtained, although a direct prognostic role is not defined. Because the elimination half-life of procainamide is 3 to 4 hours, which is substantially shorter than that of NAPA (6–10 hours), chronic overdosing typically results in NAPA toxicity.⁵ In this situation, the QT interval, a reflection of K⁺ channel blockade, correlates directly and blood pressure correlates inversely with the degree of poisoning. Severe effects usually do not occur until total (procainamide plus NAPA) serum concentrations are greater than 60 mcg/mL. Because of its structural similarity with amphetamine, some patients with procainamide overdose may have a false-positive urine enzyme-multiplied immunoassay test (EMIT) for amphetamines.¹¹⁰

Quinidine

Quinidine, the d-isomer of quinine, is derived from the bark of the cinchona tree. Because it is a weak base, it is typically formulated as the sulfate or gluconate salt. Quinidine undergoes hydroxylation by the liver, and both active and inactive metabolites are renally eliminated.

Quinidine was once widely used for the management of atrial or ventricular dysrhythmias but has largely fallen out of favor because of its adverse effects. Quinidine has substantial cardiotoxicity that includes intraventricular conduction abnormalities and an increased QT interval. “Quinidine syncope,” in which patients on therapeutic doses of quinidine experience paroxysmal, transient loss of consciousness, is almost exclusively a result of torsade de pointes.⁵¹

Because quinidine shares many pharmacologic properties with quinine (Chap. 55), patients occasionally have cinchonism after either chronic or acute quinidine overdose. This syndrome includes abdominal symptoms, tinnitus, and altered mental status. Quinidine also produces both peripheral and cardiac antimuscarinic effects, which enhance conduction via the atrioventricular (AV) node. Furthermore, as with quinine, quinidine-induced blockade of K⁺ channels in pancreatic islet cells causes uncontrolled insulin release, leading to hypoglycemia. Hypoglycemia occurs more commonly in patients who are critically ill because counterregulatory mechanisms are impaired in this population.⁸¹ The antimuscarinic effects can produce classical anticholinergic findings such as dry mouth and flushed skin.⁵⁷

Serum quinidine concentrations greater than 14 mcg/mL are associated with cardiotoxicity,⁵⁷ as evidenced by a 50% increase in either the QRS ­complex or QT interval. These concentrations have limited utility in clinical management.

Disopyramide

Disopyramide is more likely than other class IA antidysrhythmics to produce negative inotropy and congestive heart failure. This effect is noted both in patients receiving therapeutic dosing⁹⁸ and in those who overdose and is likely related to the blockade of myocardial Ca²⁺ channels caused by disopyramide.⁴⁸ The current use of disopyramide capitalizes on its negative inotropic effects for the treatment of patients with hypertrophic cardiomyopathy who remain symptomatic despite first-line treatment such as β-adrenergic antagonists.³⁰ The mono-N-dealkylated metabolite of disopyramide produces the most pronounced antimuscarinic effects of the class,^57,106 accounting for the occasional, although unproven, use of disopyramide to treat neurocardiogenic syncope.⁸⁹ Lethargy, confusion, or hallucinations are prominent in overdose.

Electrophysiologic abnormalities similar to those associated with poisoning from other class IA antidysrhythmics occur, including intraventricular conduction abnormalities, torsade de pointes, and other ventricular dysrhythmias. Disopyramide causes hyperinsulinemic hypoglycemia through its antagonism of K⁺ channels in the pancreatic islet cells.¹

Management of Class IA Antidysrhythmic Toxicity

Management concentrates on assessment and correction of cardiovascular dysfunction. After airway evaluation and IV line placement, 12-lead ECG and continuous ECG monitoring are of paramount importance. Appropriate gastrointestinal (GI) decontamination is recommended when the patient is sufficiently stabilized. Activated charcoal is recommended for early-presenting patients with a protected airway and an ingestion of a potentially toxic amount of medication. Gastric lavage is reasonable for patients with life-threatening ingestions who present within several hours depending on the severity and probability of retained ingestant. Decontamination whole-bowel irrigation is reasonable if a sustained-release preparation is involved (Chap. 5 and Antidotes in Depth: A2).

For patients who have widening of the QRS complex duration, bolus administration of IV hypertonic sodium bicarbonate is recommended (Antidotes in Depth: A5). Depolarization is accelerated, and the QRS complex duration is reduced by enhancing rapid Na⁺ ion influx through the myocardial Na⁺ channels.¹⁰ However, hypokalemia from the use of sodium bicarbonate may further prolong the QT interval, requiring careful monitoring of the serum K⁺ and ECG. Class IA antidysrhythmic-induced hypotension is treated primarily with rapid infusion of 0.9% NaCl to expand intravascular volume and to simultaneously increase myocardial contractility by enhancing the Starling force. Hypotension in the setting of QRS complex prolongation responds favorably to hypertonic sodium bicarbonate, which enhances inotropy by both accelerating depolarization and raising intravascular volume. In refractory cases, dobutamine or a catecholamine such as epinephrine or norepinephrine is recommended based on the assessment of the hemodynamic requirements for inotropy versus pressor effect. Epinephrine, by analogy with chloroquine and tricyclic antidepressants, may be superior to norepinephrine.^21,58 Extracorporeal life support can improve hemodynamics, but their use has not been systematically evaluated, but use in patients with standards indications, if readily available, seems reasonable. Because disopyramide also blocks Ca²⁺ channels, Ca²⁺ administration is reportedly beneficial,² and although evidence to support this antidotal effect is lacking, it is reasonable to administer in appropriate doses. Glucagon effectively reversed myocardial depression in canine models, but it has not been evaluated in humans because such its use remains experimental at this time.⁷⁵

We recommend treating patients with stable ventricular dysrhythmias occurring in the setting of class IA antidysrhythmic poisoning with hypertonic sodium bicarbonate or lidocaine. Although it seem counterintuitive to administer another class I antidysrhythmic to a patient already poisoned by a class I antidysrhythmic, there is sound theoretical and experimental literature to support the use of lidocaine in this setting.¹¹¹ Because lidocaine is a class IB antidysrhythmic with rapid on–off receptor kinetics, it displaces the “slower” class IA antidysrhythmic from the binding site on the Na⁺ channel, effectively reducing channel blockade. Sodium bicarbonate enhances conduction through the myocardium, promoting spontaneous termination of the ventricular dysrhythmia.³⁴ Magnesium sulfate and overdrive pacing can prevent recurrent torsade de pointes after initial spontaneous resolution or defibrillation. Medications that must be avoided in treating patients with dysrhythmias associated with class IA poisoning include other class IA and IC antidysrhythmics, as well as the β-adrenergic antagonists and Ca²⁺ channel blockers, all of which exacerbate conduction abnormalities or produce hypotension.

The roles of activated charcoal hemoperfusion, hemofiltration, and continuous arteriovenous hemodiafiltration are inadequately defined, but these are reasonable and appropriate means to remove NAPA.⁶⁷ There is not sufficient clinical evidence to support the use of hemodialysis or hemoperfusion for quinidine or disopyramide poisoning.^1,43

Class IB Antidysrhythmics: Lidocaine, Tocainide, Mexiletine, and Moricizine

Lidocaine

Lidocaine is an aminoacyl amide that is a synthetic derivative of cocaine (Fig. 57–3). Its predominant clinical uses are as a local anesthetic and, for mechanistically similar reasons, to control ventricular dysrhythmias. The high frequency of lidocaine-related medication errors relates in part to its wide use in the past as well as the availability of “amps” of varying quantities, designed for specific uses such as preparation of IV infusions or for local anesthesia.⁵⁴ Many of the current cases of lidocaine toxicity result from its use in liposuction, discussed later. Lidocaine prevents myocardial reentry and subsequent dysrhythmias (Chap. 15) by preferentially suppressing conduction in compromised tissue.⁹⁹ After an IV bolus, lidocaine rapidly enters the central nervous system (CNS) but quickly redistributes into the peripheral tissue with a distribution half-life of approximately 8 minutes.¹¹ Lidocaine is 95% dealkylated by hepatic CYP3A4 to an active metabolite, monoethylglycylxylidide (MEGX) and, subsequently, to the inactive glycine xylidide (GX). Glycine xylidide is further metabolized to monoethylglycine and xylidide. Monoethylglycylxylidide, although less potent as a Na⁺ channel blocker than lidocaine, bioaccumulates because of its substantially longer half-life.⁸

Patients with massive lidocaine exposures develop both CNS and cardiovascular effects, generally in that order. Because of its rapid entry into the CNS, acute lidocaine poisoning typically produces CNS dysfunction, with paresthesias or convulsion, as the initial manifestation.^33,86,95 Concomitant respiratory arrest generally occurs. Shortly after the CNS effects, depression in the intrinsic cardiac pacemakers leads to sinus arrest, AV block, intraventricular conduction delay, hypotension, or cardiac arrest. If the patient is supported through this period, the xenobiotics rapidly distributes away from the heart, and spontaneous cardiac function returns.

Acute lidocaine toxicity is generally related to excessive or inappropriate parenteral therapeutic dosing. Common settings include inadvertent IV or intraarterial administration instead of the intended route and excessive subcutaneous administration during laceration repair or other procedures.¹⁰⁵ Acute lidocaine toxicity is reported with topical tracheal application of lidocaine used for bronchoscopy,¹¹³ during circumcision,⁸⁶ and during ureteral application during ureteroscopic stone extraction.⁷⁸ The typical CNS manifestations of submassive lidocaine poisoning include drowsiness, weakness, a sensation of “drifting away,” euphoria, diplopia, decreased hearing, paresthesias, muscle fasciculations, and seizures. The more severe of these effects develop when serum lidocaine concentrations exceed 5 mcg/mL and are often preceded by paresthesias or somnolence. Any of these symptoms should, therefore, prompt the clinician to examine the patient’s medication administration history or medication-infusion rate. Apnea and seizures, as well as hypotonia in neonates, are reported to result from submassive acute lidocaine toxicity.⁸⁵

A related form of toxicity and death results from subcutaneous and adipose administration of lidocaine during tumescent liposuction.⁸³ In this technique, a large volume of dilute lidocaine is used to distend subcutaneous fat before liposuction.⁹ Although in some reports, the cause of death was controversial,⁸² postmortem lidocaine concentrations were commonly elevated, and it is likely that lidocaine metabolites were also involved in the adverse events.⁵⁵ Interestingly, proponents of this procedure suggest that lidocaine doses up to a maximum of 55 mg/kg are safe,⁹ but the conventional recommended limit for subcutaneous lidocaine with epinephrine is only 5 to 7 mg/kg. Of significant concern is that the recommended doses used for liposuction procedures do not consider the ability of lidocaine to saturate the CYP3A4 enzymes. When saturation occurs, elimination lags behind absorption, and lidocaine toxicity results.

Numerous publications unequivocally demonstrate the toxicity associated with orally administered lidocaine despite its poor bioavailability.^20,115 Some of the toxicity is attributed to MEGX. Because of the relatively high concentration of viscous lidocaine (typically 4%), this preparation is overrepresented in reports of oral lidocaine poisoning.¹¹⁵ As little as 15 mL of 2% viscous lidocaine in a 3-year-old child (estimate, 300 mg or 21.4 mg/kg/dose) is reported to cause seizures.

Chronic lidocaine toxicity most commonly occurs as a result of therapeutic misadventure in patients on lidocaine infusions, generally in a critical care unit. Toxicity after appropriate dosing is most likely to occur in patients with reduced hepatic blood flow as occurs with congestive heart failure, liver disease, or concomitant therapy with CYP3A4 and CYP1A2 inhibitors (Chap. 11).¹⁰⁴ Adverse reactions to lidocaine also increase with advancing age, decreasing body weight, and increasing infusion rate. Chronic lidocaine toxicity occurs in 6% to 15% of patients receiving infusions at 3 mg/min for several days.⁹³ Partly for this reason, lidocaine is no longer routinely used to prevent dysrhythmias in the immediate postmyocardial infarction period. The clearance of lidocaine falls after approximately 24 hours of the start of an infusion, and this effect may be caused by competition for hepatic metabolism between lidocaine and its metabolites.

Mexiletine

Mexiletine, originally developed as an anorectic, was found to have antidysrhythmic, local anesthetic, and anticonvulsant activity.¹⁷ It is currently available in oral form for the management of ventricular dysrhythmias and is used for the management of chronic neuropathic pain. Its chemical structure and electrophysiologic properties are similar to those of ­lidocaine. Mexiletine, a base, is absorbed in the small intestine; therefore, its absorption is increased when the gastric contents are alkalinized. Congestive heart failure and cirrhosis, as well as therapy with cimetidine or disulfiram, decrease the clearance of mexiletine.⁶² Its metabolism, predominantly through CYP2D6, is accelerated by concomitant use of phenobarbital, rifampin, and phenytoin.

Adverse effects with therapeutic dosing are primarily neurologic and are similar to those that occur with lidocaine. The few reported cases of mexiletine overdose describe prominent cardiovascular effects such as complete heart block, torsade de pointes, and asystole.^23,37 Neurotoxicity resulting from overdose includes self-limited seizures, generally in the setting of cardiotoxicity. Moreover, a single case report described a patient with mexiletine poisoning who experienced status epilepticus without any hemodynamic or ECG abnormalities.⁷² Mexiletine is reported to produce a false-positive result on the amphetamine immunoassay of the urine.^23,59

Moricizine

Moricizine possesses the general qualities of class I antidysrhythmics but is difficult to specifically subclassify because it has properties that place it in both classes IB or IC.²² Historically, it is discussed as a class IB antidysrhythmic, as it is here. The parent medication undergoes extensive and rapid metabolism. Dose-related lengthening of PR interval and QRS complex are expected, as are hemiblocks, bundle blocks, and sustained ventricular tachydysrhythmias. Experience in the setting of myocardial infarction during CAST II suggests that it is a prodysrhythmic.¹⁰² Clinical experience with overdose is limited but is expected to be similar to that of other class I antidysrhythmics.

Management of Class IB Antidysrhythmic Toxicity

The focus of the initial management for IV lidocaine-induced cardiac arrest is continuous cardiopulmonary resuscitation to allow lidocaine to redistribute away from the heart. Apart from this setting, management of hemodynamic compromise includes fluid replacement and other conventional strategies. Catecholamine administration is recommended for the management of resistant hypotension, and if refractory toxicity persists, insertion of an intraaortic balloon assist pump or cardiac bypass.³⁹ Cardiopulmonary bypass, which does not directly enhance elimination, maintains hepatic perfusion, thereby allowing the lidocaine to be metabolized.³⁹ When bradydysrhythmias do not respond to atropine, chronotrope such as norepinephrine or isoproterenol is recommended. External pacing or insertion of a transvenous pacemaker is reportedly successful in some instances, but the myocardium is often refractory to electrical capture. Lidocaine-induced seizures, and those related to lidocaine analogs, are generally brief in nature and do not require specific therapy. For patients requiring treatment, an IV benzodiazepine is recommended; rarely, a barbiturate is required. Similarly, although IV lipid emulsion is often described as useful for the resuscitation of patients with life-threatening local anesthetic overdose (local anesthetic systemic toxicity) its use for lidocaine poisoned patients is limited to case reports and likely unnecessary given the rapid time course of recovery. As such, lipid emulsion is not recommended for lidocaine poisoning except in cases of severe and prolonged toxicity that are refractory to other therapies.⁴⁹ Lipid emulsion was used in the successful resuscitation from cardiac arrest after local infiltration of lidocaine for an incision and drainage procedure.¹⁰⁵ It was also used to reverse the CNS toxicity of lidocaine used for local anesthesia during procedures (Antidotes in Depth: A23).⁶³ Enhanced elimination techniques are limited after IV poisoning because of the rapid time course of poisoning.

After oral poisoning by a class IB antidysrhythmic, activated charcoal is recommended. Lidocaine and its metabolites are not well cleared by hemodialysis,²⁷ and there are no adequate data to support the use of extracorporeal removal for mexiletine or moricizine. There is one case report of a patient with a large oral mexiletine overdose who was successfully treated with hemodialysis, but how to apply this more broadly remains unclear.⁴

Class IC Antidysrhythmics: Flecainide and Propafenone

Flecainide

Flecainide, a derivative of procainamide, is orally administered to maintain sinus rhythm in patients with structurally normal hearts who have atrial fibrillation or supraventricular tachycardia.¹⁰⁰ Kidney disease, medication interactions, and congestive heart failure all decrease the clearance of flecainide and its active metabolite. Additionally, alkaluria reduces its clearance, presumably by enhanced tubular reuptake of nonionized flecainide. Therapeutic doses produce left ventricular dysfunction with worsening congestive heart failure. This is presumably a result of the negative inotropic effect of flecainide, which itself relates to its antagonistic effects on Ca²⁺ channels. Furthermore, sudden dysrhythmic death occurs, particularly in patients with underlying ischemic heart disease.¹⁰⁰

Flecainide toxicity is associated with an increase in QRS complex duration, prolongation of the PR interval, and prolongation of the QT interval. The duration of the QRS complex has significant prognostic value: a QRS duration of 200 ms or greater is predictive of the need for mechanical circulatory support.¹⁰⁸ The expected consequences of these electrophysiologic disturbances include bradycardia, premature ventricular contractions, and ventricular fibrillation. Occasionally, supraventricular tachycardia can present with a bizarre axis and is often mistaken for ventricular tachycardia, which results in inappropriate treatment.¹⁰⁸ The combination of marked QRS complex and PR interval changes, associated with minimal QT interval prolongation, is characteristic of flecainide toxicity and contrasts with those described with other antidysrhythmics.

Propafenone

Propafenone bears a structural resemblance to propranolol,³⁶ as well as similar qualitative, but not quantitative, electrophysiologic properties.³² Propafenone blocks fast inward Na⁺ channels, is a weak β-adrenergic antagonist, and is an L-type Ca²⁺ channel blocker.³² Its long half-life allows the accumulation of propafenone, particularly in patients with the slow metabolizer pharmacogenetic variant of CYP2D6 or those exposed to xenobiotics that inhibit this enzyme.⁶⁴ Propafenone overdose produces sinus bradycardia, ventricular dysrhythmias, and negative inotropy. The ECG often shows a right bundle branch block pattern, first-degree AV block, and prolongation of the QT interval. Generalized seizures have occurred.⁷⁷

Management of Class IC Antidysrhythmic Toxicity

Management concentrates on assessment and correction of cardiovascular dysfunction. After airway evaluation and IV line placement, 12-lead ECG and continuous ECG monitoring are of paramount importance. Appropriate GI decontamination is recommended when the patient is sufficiently stabilized. Activated charcoal is recommended for early-presenting patients with a protected airway and an ingestion of a potentially toxic amount of medication. Gastric lavage is reasonable for patients with life-threatening ingestions who present within several hours depending on the severity and probability of retained ingestant. Decontamination should also include whole-bowel irrigation are recommended if a sustained-release preparation is involved (Chap. 5 and Antidotes in Depth: A2).

Standard management strategies for hypotension, such as fluid resuscitation, and for seizures, with benzodiazepines as well as careful attention to maintaining the airway as indicated. Additionally, therapy for hypotension and the ECG manifestations of class IC poisoning include IV hypertonic sodium bicarbonate to overcome the Na⁺ channel blockade.⁶⁰ The utility of calcium salts for the treatment of propafenone toxicity is undefined. Several reports of overdose in humans verify QRS complex narrowing in response to hypertonic sodium bicarbonate administration on flecainide^12,50,66 and propafenone.⁷⁷ The renal elimination of flecainide is reduced by urinary alkalinization, suggesting that sodium chloride, in equimolar doses, is superior to sodium bicarbonate; however, the routine attempt to enhance urinary clearance with saline is not supported by evidence and is not recommended.⁷⁰ The administration of other class IC or IA antidysrhythmics is contraindicated because of their additive blockade of the Na⁺ channel. Similarly, the administration of phenytoin to a child with propafenone poisoning was associated with a prolongation of the QRS complex, which initially responded to sodium bicarbonate, but the patient subsequently developed bradyasystolic arrest.⁶⁹ However, amiodarone was successful in the setting of flecainide-induced ventricular fibrillation refractory to other therapy.⁹⁶ As with β-adrenergic antagonists, an animal model and human case reports suggest that hyperinsulinemic euglycemic therapy is beneficial after propafenone poisoning.^7,116 In patients with refractory bradycardia, pacing should be attempted with the caveat that the efficacy of an external or internal pacemaker may be limited because of the xenobiotic-induced increased electrical pacing threshold of the ventricle. Successful therapy with cardiopulmonary bypass or extracorporeal membrane oxygenation (ECMO) is reported and is recommended in patients with severe toxicity when available.^6,13,25 Intravenous fat emulsion therapy was reportedly successful in two patients with severe flecainide poisoning²⁹ and in propafenone poisoning.^7,107 Although the safety and efficacy of lipid emulsion therapy in this setting remains undefined, it is reasonable for those who are seriously ill, deteriorating, and refractory to conventional therapy.

Extracorporeal removal is not expected to be beneficial for patients with flecainide poisoning. Although hemodialysis was successful in removing propafenone after overdose, additional studies are needed to determine its clinical benefit.¹⁴ It is not recommended at this time.

CLASS III ANTIDYSRHYTHMICS

Amiodarone, Dofetilide, Dronedarone, and Ibutilide

The class III antidysrhythmics prevent and terminate reentrant dysrhythmias by prolonging the action potential duration and effective refractory period without slowing conduction velocity during phase 0 or 1 of the action potential. This effect on the action potential is generally caused by blockade of the rapidly activating component of the delayed rectifier K⁺ current, which is responsible for repolarization.

The class III antidysrhythmics in use today prolong repolarization of both the atria and ventricles. Thus, common ECG effects at therapeutic doses include prolongation of the PR and QT intervals and abnormal T and U waves. Chap. 15 contains a detailed discussion of the pharmacologic mechanisms of class III antidysrhythmics. Also, Chaps. 16 and 59 discuss sotalol.

Amiodarone and Dronedarone

Amiodarone (Fig. 57–4) is an iodinated benzofuran derivative that is structurally similar to both thyroxine and procainamide. Forty percent of its molecular weight is iodine. Dronedarone is an analog of amiodarone that does not contain iodine.⁶⁰ The 2015 revision of the Advanced Cardiovascular Life Support guidelines places significant emphasis on the early IV administration of amiodarone for ventricular dysrhythmias. Both medications are used primarily to terminate or prevent atrial fibrillation, and although dronedarone is less effective and has safety issues,²⁴ it is not associated with many of the potentially severe adverse effects of amiodarone.⁹²

Although amiodarone and dronedarone have multiple pharmacologic effects, their efficacy is primarily the result of their class III antidysrhythmic effects. They also have weak α- and β-adrenergic antagonist activity and can block both L-type Ca²⁺ channels and inactivated Na⁺ channels. Amiodarone is slowly absorbed by the oral route and concentrates in the liver, lung, and adipose tissue. It has a very long elimination half-life, measured in weeks, and steady-state pharmacokinetics do not occur until after 1 month of use. Dronedarone has a half-life of about 24 hours and a smaller volume of distribution.

The ECG effects of amiodarone differ based on the route of medication administration. Therapeutic oral doses prolong the PR and QT intervals but do not alter the QRS complex. Intravenous dosing prolongs the PR interval but has few other ECG manifestations. Ventricular dysrhythmias and sinus bradycardia are the most serious cardiac complications of therapeutic doses of amiodarone. Ventricular tachycardia reported to be resistant to cardioversion and pharmacologic interventions occur³⁵ but are surprisingly uncommon given the frequency and extent to which the QT interval prolongation occurs. The ability of amiodarone to compete for P-glycoprotein is responsible for several drug interactions, including elevated digoxin and cyclosporin concentrations and enhanced anticoagulation effectiveness of warfarin⁶¹ (Chap. 9).

The diverse complications associated with long-term amiodarone therapy do not occur after short-term IV use. Chronic therapy with oral amiodarone causes substantial pulmonary, thyroid, corneal, hepatic, and cutaneous toxicity caused by bioaccumulation in these organs. Dronedarone rarely causes hepatic or pulmonary toxicity.²⁶ Many of these effects appear to be dose related, but because of the wide range of bioavailabilities and metabolic patterns among different patients, as well as the overlap between therapeutic and toxic serum concentrations, therapeutic drug monitoring is of limited benefit.⁴¹ Pneumonitis, the most consequential extracardiac adverse effect, occurs in up to 5% of patients taking the xenobiotic therapeutically. Amiodarone pneumonitis rarely develops within days of initiating therapy but typically occurs only after years of therapy. Its occurrence appears to be dose related: a daily dose of greater than 400 mg is a risk factor, and pneumonitis is rare in those taking less than 200 mg daily. Focusing on using the minimal effective dose has reduced the incidence of pneumonitis.³¹ Oxygen supplementation speeds the development of pneumonitis, which explains the initial belief that patients with chronic lung disease are at increased risk for amiodarone pneumonitis. Manifestations of pneumonitis include dyspnea, cough, hemoptysis, crackles, hypoxia, and radiographic changes.¹⁸ A chest computed tomography scan is the most helpful initial diagnostic test for pneumonitis but is not useful for monitoring purposes, which is often done with diffusing capacity of carbon monoxide.⁴¹ Bronchoalveolar lavage typically reveals interstitial pneumonitis with many macrophages and a characteristic finely vacuolated foamy cytoplasm, but confirmation of the diagnosis requires open lung biopsy.

Thyroid dysfunction, either amiodarone-induced thyrotoxicosis (AIT) or amiodarone-induced hypothyroidism (AIH), occurs in approximately 4% of patients.¹⁹ Amiodarone-induced hypothyroidism is more common than AIT when iodine intake is sufficient.⁴¹ Amiodarone-induced hypothyroidism is likely caused by an exaggerated Wolff-Chaikoff effect, in which iodine—in this case from amiodarone—inhibits the organification and release of thyroid hormone. Amiodarone-induced thyrotoxicosis appears to exist in two distinct forms: type I AIT, which occurs in patients with abnormal thyroid glands and iodine-induced excessive thyroid hormone synthesis and release, and type II AIT, in which destructive thyroiditis leads to release of thyroid hormone from the damaged follicular cells. The relative prevalence of the two forms of AIT is unknown. Amiodarone also reduces the effect of thyroid hormone on peripheral tissue. The diagnosis is confirmed with standard thyroid function testing (thyroid-stimulating hormone, total triiodothyronine and free thyroxine¹⁰¹ (Chap. 53).

Corneal microdeposits are extremely common during chronic therapy and leads to vision loss.⁶⁸ Abnormal elevation of hepatic enzymes occurs in more than 30% of those on long-term therapy, and the hepatotoxicity progresses to cirrhosis. Periodic monitoring of aminotransferases is recommended.⁴¹ Hepatotoxicity has occurred after initial loading of amiodarone.⁸⁴ Slate-gray or bluish discoloration of the skin is common, particularly in sun-exposed portions of the body.⁸⁸

Dofetilide

Dofetilide is approved for conversion of atrial fibrillation or atrial flutter to a normal sinus rhythm. Dofetilide increases the effective refractory period more substantially in atrial tissue than in ventricular fibers, accounting for this clinical indication.⁹⁰ Dofetilide has no known effect on Ca²⁺ or Na⁺ channels, nor does it result in β-adrenergic antagonism. Dofetilide increases the QT interval but does not change either the PR interval or QRS complex in humans. Heart rate and blood pressure are also not appreciably affected.

Although limited data are available, the expected and reported adverse cardiac events include ventricular tachycardia, particularly torsade de pointes.¹¹² The approximate incidence of torsade de pointes in patients receiving high therapeutic doses of the medication is 3%.³ For this reason, the US Food and Drug Administration has in place strict requirements for the use of dofetilide, such as an individualized dose initiation algorithm and mandatory hospitalization for initial therapy.⁵⁶

Overdose data reported by the manufacturer include two cases. One patient reportedly ingested 28 capsules and experienced no events, but a second patient inadvertently received two supratherapeutic doses 1 hour apart and experienced fatal ventricular fibrillation.⁸⁰ A 33-year-old man ingested 5 mg (20 capsules) and developed QT interval prolongation within 1 hour of ingestion but had no dysrhythmia during his 4-day hospital stay.¹⁶ One patient with a 90-fold intentional overdose experienced nonsustained ventricular tachycardia, frequent multifocal premature ventricular contractions, and ventricular bigeminy, which were successfully treated with potassium and magnesium.⁴⁷

Ibutilide

Ibutilide is an antidysrhythmic with predominant class III activity used for the rapid conversion of atrial fibrillation and flutter to normal sinus rhythm. Because of its extensive first-pass metabolism, ibutilide can only be administered parenterally. Its metabolic pathways are not well understood but do not involve the isoenzymes CYP3A4 or CYP2D6. Pharmacokinetic data thus far do not indicate that age, sex, hepatic, or CKD necessitates adjustment of recommended dosage of ibutilide. In addition to its effects on the delayed rectifier current, ibutilide activates a slow inward Na⁺ current.⁷¹

Ibutilide increases the QT interval and causes torsade de pointes, especially in patients with congenital long-QT syndrome and in women.⁴² Although ibutilide enhances the efficacy of transthoracic cardioversion for atrial fibrillation, its use in patients with ejection fractions below 20% is associated with an increased incidence of sustained polymorphic ventricular tachycardia. Acute kidney injury, including biopsy-identified crystals, is reported in association with ibutilide cardioversion, but a causal relationship is not yet definitive.³⁸ Acute overdose information, only available in limited form (four patients) through the manufacturer, suggests that ventricular dysrhythmias and high-degree AV conduction abnormalities should be expected.⁷⁹

Management of Class III Antidysrhythmic Toxicity

Treatment experience with class III antidysrhythmic overdose is limited. Isoproterenol, magnesium sulfate, and overdrive pacing can prevent recurrent torsade de pointes after initial spontaneous resolution or defibrillation. However, prolonged therapy may be required for amiadorone overdose because of the long half-life of drug.⁹⁴ Administration of class IB ­antidysrhythmics or propranolol for the control of monomorphic ventricular tachycardia cannot be recommended on theoretical grounds. Paradoxically, amiodarone may reduce the risk of torsade de pointes associated with the other class III antidysrhythmics.⁹⁷ This effect is likely mediated by the beneficial effects of amiodarone on the dispersion of myocardial repolarization and its Ca²⁺ channel–blocking activity.

Oral activated charcoal is recommended if patients present shortly after overdose. Hemodialysis is not expected to be beneficial in general, either because of extensive protein binding or because of large volumes of distribution (Table 57–1). A neonate survived cardiovascular collapse with the use of ECMO after an iatrogenic IV amiodarone overdose.⁴⁴ Although the benefit versus risk profile of this intervention is undefined, it is reasonable to implement ECMO if it is available or reasonably attainable through transfer. Lipid emulsion sequesters amiodarone and is demonstrated to be beneficial in swine models.¹¹⁴ It is only reasonable administered to patients with severe amiodarone poisoning refractory to standard resuscitative efforts. However, when amiodarone is used as a treatment for poisoning by another xenobiotic, lipid emulsion therapy will reduce the effectiveness of the amiodarone because of subsequent sequestration.⁷⁴

ANTIDYSRHYTHMICS-UNCLASSIFIED

Adenosine, a nucleoside found in all cells, is released from myocardial cells under physiologic and pathophysiologic conditions. It is administered as a rapid IV bolus to terminate reentrant supraventricular tachycardia. The effects of adenosine are mediated by its interaction with specific G protein–coupled adenosine (A₁) receptors that activate acetylcholine-sensitive outward K⁺ current in the atrium, sinus nodes, and AV nodes. The resultant hyperpolarization reduces the rate of cellular firing. Adenosine also reduces the Ca²⁺ currents, and its antidysrhythmic activity results from its effect in increasing AV nodal refractoriness and from inhibiting delayed afterdepolarizations elicited by sympathetic stimulation.⁶⁵

Adverse effects of adenosine administration are very common and include transient asystole, dyspnea, chest tightness, flushing, hypotension, and atrial fibrillation. Although bronchospasm occurs after pulmonary vascular administration, it is not reported after routine IV use. Dyspnea, and probably chest tightness, is related to adenosine stimulation of the pulmonary vagal C fibers.¹⁵ Fortunately, most of the adverse effects of adenosine are transient because of its rapid metabolism to inosine by both extracellular and intracellular deaminases. The clinical effects are potentiated by dipyridamole, an adenosine uptake inhibitor,⁴⁰ and by denervation hypersensitivity in cardiac transplant recipients. Methylxanthines block adenosine receptors (Chap. 63). In this setting, larger-than-usual doses of adenosine are required to produce an antidysrhythmic effect. Overdose of adenosine is not reported. Treatment is supportive because of the rapid elimination of the medication.

Digoxin, a cardioactive steroid, is often considered a member of this class. A complete discussion of digoxin can be found in Chap. 62.

Magnesium ion is occasionally included in this class as well because of its ability to prevent the development of recurrent torsade de pointes. A complete discussion of magnesium sulfate can be found in Antidotes in Depth: A16.

SUMMARY

• In the overdose setting, the class IA, B, and C antidysrhythmics all produce Na⁺ channel blockade, which can induce profound wide-complex cardiac dysrhythmias and morbidity if not treated judiciously.

• The class IC antidysrhythmics are considerably more toxic than the other class I antidysrhythmics, although even the class IB antidysrhythmics are lethal in overdose. Dysrhythmias are their primary toxicologic effect.

• The class III antidysrhythmics in overdose cause malignant dysrhythmias, particularly torsade de pointes. Amiodarone has many noncardiac effects, particularly pneumonitis and thyroid effects, which limit its therapeutic usefulness.

• Proper management of both adverse effects and overdose can be accomplished only by understanding the pharmacokinetics and toxicokinetics of these medications.

Acknowledgments

Neal A. Lewin, MD; Mary Ann Howland, PharmD; and Harold Osborn, MD, (deceased) contributed to this chapter in a previous edition.

REFERENCES

INTRODUCTION

Although magnesium is a divalent cation like other metals, we will refer to it as magnesium in this text. It is an essential cofactor in more than 350 enzyme reactions in cardiac, neurologic, neuromuscular, and endocrine processes, as well as in basic energy, structural, nucleic acid, and signal transduction pathways.^107,136 Hypomagnesemia results in nausea, vomiting, weakness, muscle spasms, neurologic and muscular excitation (tremor, hyperreflexia, and tetany), and cardiac dysrhythmias. Hypermagnesemia produces cardiovascular effects, including conduction disturbances, hypotension, and cardiac arrest; respiratory depression; gastrointestinal (GI) complaints of nausea, vomiting, and thirst; and neuromuscular sequelae, including weakness, paralysis, and central nervous system (CNS) depression.^4,77,131 Magnesium as chloride, citrate, hydroxide, oxide or sulfate is used to repair xenobiotic-associated hypomagnesemia and as an adjunctive treatment for cardiovascular toxins, fluoride toxicity, pesticide poisoning, and alcohol use disorders.

HISTORY

In 1679, Nehemiah Grew described the process of obtaining the eponymous salts in the springs of Epsom through evaporation, which he later published in 1695 to be magnesium sulfate (MgSO₄).¹⁰⁵ Later as an antidote, MgSO₄ was used as a therapy for tetanus in 1906.⁶³ In 1923, MgSO₄ was introduced to treat the convulsions of nephritis in children, which was later extended to the advanced chronic nephritis in adults.¹³¹ On the basis of reports in 1925 and 1926, MgSO₄ came into common practice for the prevention and control of seizures in severe toxemia of pregnancy; its use for this indication was “grandfathered” by the US Food and Drug Administration (FDA).²⁰ The cardiac utility of MgSO₄ was demonstrated in 1935 and 1943 as therapy to ­suppress “paroxysmal tachycardia.”^13,139 Magnesium sulfate also continues in use as a uterine tocolytic, a bronchodilator a migraine therapy and as a nutritional adjunct to prevent hypomagnesemia in hyperalimentation.

PHARMACOLOGY

Chemistry and Preparation

Magnesium has the atomic number 12, with a molecular weight of 24.3 Da. Despite having a smaller atomic radius than calcium, magnesium binds water tighter, with two hydration shells, leading to a radius 400 times larger than its dehydrated form.⁶⁴ This property underlies its calcium antagonism, despite similar chemical reactivity and charge.⁶⁴ For clinical use, magnesium is prepared as a salt, typically in combination with chloride, citrate, hydroxide oxide, or sulfate. Magnesium as MgSO₄ is typically formulated in its heptahydrate form MgSO₄•7H₂O, with a molecular weight of 120.38 Da alone and 246.47 when accounting for water. It occurs as colorless crystals or white powder, which is freely soluble in water.

Related Agents

The use of magnesium as an oral saline cathartic (eg, magnesium citrate, magnesium hydroxide, and MgSO₄) is discussed in Antidotes in Depth: A2.

Pharmacokinetics

Magnesium is the fourth most common plasma cation after sodium, potassium, and calcium but follows potassium as the second most prevalent intracellular cation.³⁷ Adults contain approximately 1 mole (21–28 g) of magnesium; homeostasis balances intestinal absorption and renal excretion.¹⁰⁷ A normal serum magnesium concentration is 1.5 to 2.0 mEq/L (1.7–2.4 mg/dL; 0.7–1 mmol/L).¹⁶ Magnesium is approximately 30% to 40% protein bound.^16,82 Extracellular magnesium represents only 1% of the body’s total stores, with the remaining magnesium found in bone (∼60%), muscle (∼20%), and soft tissues (∼19%). Serum values do not always accurately reflect tissue concentrations and can be normal despite intracellular deficiency.³⁷ This discrepancy was assessed by determining retained magnesium in response to an intravenous (IV) load: those retaining more than 30% of an 800-mg load of IV magnesium were considered magnesium depleted; those retaining less than 20% were replete.¹²⁰ Magnesium is not metabolized, but only ionized magnesium is physiologically active. After being absorbed, magnesium is eliminated by the kidney and occurs at a rate proportional to the plasma concentration and glomerular filtration. As kidney function declines, compensatory decreases in tubular resorption occur; however, in end-stage kidney disease, limited magnesium excretion can lead to hypermagnesemia.¹³⁶

Pharmacokinetic studies of MgSO₄ typically have involved pregnant women with preeclampsia. After a 4-g loading dose followed by a 1-g/h continuous maintenance infusion, serum magnesium concentrations approximately doubled from 0.74 to 0.85 mmol/L to 1.48 to 1.70 mmol/L at 30 minutes and remained constant for at least 24 hours.⁹¹ When a 4-g loading dose is followed by a 2-g/h continuous maintenance infusion, magnesium concentrations reach 1.73 to 2.25 mmol/L.⁹¹ After IV MgSO₄ administration, the apparent volume of distribution (V_d) increases rapidly and becomes constant within 2 hours in healthy nonpregnant individuals but may not become constant for up to 4 hours in pregnant women.²¹ Given intramuscularly, [MgSO₄] does not peak until 90 to 120 minutes followed by a slow decline back to baseline concentrations over 4 to 8 hours, unless repeated intramuscular (IM) magnesium is provided at 4 hours.^21,82,91

Pharmacodynamics and Mechanisms of Action

The critical roles and functions of magnesium are more completely reviewed in Chap. 12. Selected key magnesium actions provide the physiologic basis for antidotal strategies. Adenosine triphosphate (ATP) avidly binds and requires magnesium to participate in enzymatic reactions.¹⁰³ Separately, magnesium acts as a calcium channel blocker. In the heart, this produces negative inotropy in isolated preparations. However it results in positive inotropy in healthy human volunteers, presumably because of decreased systemic and pulmonary arterial vascular resistance.^37,103 Additionally, limiting calcium outflow from the sarcoplasmic reticulum and restricting outward potassium movement appear to underlie the role of magnesium in reduction of sinus node rate firing, increased atrioventricular (AV) node refractoriness, and suppression of ectopy.^37,103 The role of magnesium in calcium channel blocker, activation of adenylate cyclase, and increased production of endothelial vasodilator prostacyclin and cyclic adenosine monophosphate (cAMP) lead to vascular relaxation in peripheral and intracranial blood vessels.^6,12,128 In the pulmonary system, the inhibitory actions of magnesium on smooth muscle lead to a weak bronchodilating effect.⁵⁸

In the nervous system, the glutamate N-methyl-D-aspartate (NMDA) receptor mediates excitatory neurotransmission. The NMDA receptor undergoes voltage-dependent blockade following magnesium administration with a high permeability for calcium in those receptors expressing NR1 and NR2 although not NR3 subunits.¹⁸ Magnesium used as an NMDA antagonist has demonstrated neuroprotective and antinociceptive effects in clinical trials.^3,133 Magnesium competitively blocks presynaptic calcium entry to decrease or abolish acetylcholine release at motor endplates and postsynaptically decreases sensitivity to acetylcholine.^48,77 Magnesium similarly blocks N- (and L-) type calcium channels to inhibit catecholamine release from adrenergic nerves.¹¹⁰ Effective serum concentrations for the prevention or treatment of eclamptic seizures range from 4 to 7 mEq/L (2.0–3.5 mmol/L).⁹¹

ROLE IN XENOBIOTIC-INDUCED HYPOMAGNESEMIA

Table 12–10 lists common xenobiotic causes of hypomagnesemia. These include xenobiotics that cause GI or renal losses (laxatives, alcohol, and diuretics), chemotherapeutics (cisplatin, human epidermal growth factor receptor antagonists), antimicrobials (aminoglycosides, amphotericin B, pentamidine), immunosuppressants (cyclosporine and tacrolimus), and proton pump inhibitors. In animal models, magnesium administration is additionally protective against cisplatin-induced nephrotoxicity while improving chemotherapeutic efficacy.^78,104 Receptor activator of nuclear factor κ-B ligand (RANKL) inhibitors used to treat osteoporosis and bone metastases also place patients at risk for hypomagnesemia.⁷⁹ The choice of oral or IV repletion should depend on the degree of hypomagnesemia and the clinical scenario, with IV treatment used for severe cases.

ROLE IN POISONING BY CARDIAC TOXINS

Digoxin inhibits the magnesium-dependent sodium-potassium adenosine triphosphatase (Na⁺-K⁺,-ATPase). Antidotal use of MgSO₄ in cardiovascular poisoning dates back to 1935, when it was used to mitigate digoxin-associated dysrhythmias in 15 patients.¹³⁹ In 1950, Szekely reported that 3 to 4 g of MgSO₄ interrupted digoxin-associated bidirectional tachycardia or extrasystolic bigeminy in six patients.¹¹⁵ In controlled studies of cardioactive steroids toxicity in dogs^71,126 and monkeys,¹²⁷ cardiotoxicity (ie, ventricular dysrhythmias and death) was significantly more pronounced and occurred at lower concentrations in magnesium-deficient animals. In another dog experiment, MgSO₄ reversed cardioactive steroids–induced myocardial potassium egress (a measure of sodium-Na⁺,K⁺-ATPase inhibition).⁸⁹ In humans, MgSO₄ abolished digoxin-associated extrasystoles in all but 1 of 14 patients.⁴² Magnesium sulfate eliminated digoxin-associated tachydysrhythmias in all 7 patients who were also determined to have decreased intracellular [magnesium] despite normal serum [magnesium] and [digoxin].²⁵

In seven patients taking digoxin for chronic Chagasic cardiomyopathy, all of those with magnesium deficiency as determined by muscle biopsy had cardiac dysrhythmias (eg, ventricular tachycardia {VT}, AV dissociation), some despite normal serum magnesium.¹⁷ Twenty-one hospitalized patients with low intracellular magnesium concentrations as determined by monocyte analysis or serum hypomagnesemia, experienced classic digoxin toxicity (eg, ventricular tachycardia, multiple premature or multifocal ventricular beats, supraventricular tachycardia with AV nodal block) at relatively low serum digoxin concentrations (three of 2 nmol/L).¹³⁵ Three of 21 patients with digoxin toxicity had low intracellular magnesium concentrations but normal serum concentrations. In human case reports, magnesium has stabilized refractory digoxin-associated ventricular dysrhythmias when digoxin-specific antibody fragments were unavailable.^46,68 Magnesium augmented the efficacy of Digifab in reversing marinobufagenin-induced Na⁺,K⁺-ATPase inhibition in humans with pregnancy-induced hypertension and elevated marinobufagenin concentrations, providing an experimental link for magnesium efficacy in preeclampsia.¹³⁷ On the basis of these data, it is reasonable to administer magnesium to patients with tachydysrhythmias evidence of digoxin poisoning, even in the setting of normal digoxin concentrations.

Magnesium sulfate has been used sporadically to treat patients with other specific cardiovascular poisonings. In murine models of amitriptyline poisoning, MgSO₄ was effective and superior to lidocaine in converting VTs and synergistic with norepinephrine in reducing occurrence of dysrhythmias.^72,73 Case reports demonstrate similar benefit of magnesium in converting refractory ventricular dysrhythmias in cyclic antidepressant overdose.^23,49,74,108 In one randomized trial of 72 patients with cyclic antidepressant poisoning, MgSO₄ 1 g every 6 hours in addition to standard care with bicarbonate infusion decreased intensive care unit (ICU) duration and decreased the mortality rate from 33% to 14%.⁴¹ In a review of ventricular dysrhythmias associated with aconite poisoning, MgSO₄ terminated dysrhythmias in two of nine patients, although its precise role was unclear.²⁹

Magnesium use in patients with cardiac dysrhythmias has demonstrated efficacy and safety. Large doses of MgSO₄ (7–12 g) were initially explored to suppress multifocal atrial tachycardia.⁶² As an adjunct, MgSO₄ improved rate control and rhythm conversion in patients with atrial fibrillation primarily treated with digoxin.³² A separate study found that twice as much IV digoxin was required for rate control of atrial fibrillation in hypomagnesemic patients (<1.5 mEq/L).³⁴ Intravenous MgSO₄ (3±2 g initially) enhanced the efficacy of dofetilide to chemically cardiovert 160 patients with atrial fibrillation and flutter, a 107% increased odd of success.²⁶ From a safety perspective, in 20 patients who received ibutilide for atrial fibrillation and flutter, 2 g of MgSO₄ eliminated the 29% increase in QT interval induced by ibutilide.⁵⁷ In a trial of 476 patients with atrial fibrillation and flutter, IV infusion of 5 g of MgSO₄ for 1 hour before the ibutilide administration followed by an additional IV infusion of 5 g of magnesium increased the rate of conversion to sinus rhythm from 67.3% to 76.5% and decreased the incidence of ventricular dysrhythmias (including torsade de pointes) from 7.4% to 1.2%.⁹⁵ In a separate trial of 229 consecutive patients who received ibutilide, MgSO₄ (1–4 g) was associated with a dose-responsive 78% increased odds of successful chemical conversion.¹¹⁷ Magnesium sulfate (0.037 g/kg body weight followed by 0.025 g/kg/h) eliminated the need to administer amiodarone (for heart rate >110 or non–sinus rhythm) in 16 of 29 patients critically ill patients with new-onset atrial fibrillation who would have otherwise received it per protocol.¹¹³ An oral dose of only 250 to 500 mg/day MgSO₄ did not improve the efficacy of conversion of atrial fibrillation with sotalol.⁴⁷ However, in a study of 34 patients, oral magnesium (500 mg) reduced the QTc interval prolongation induced by sotalol and dofetilid.⁸⁶ Although one meta-analysis was unable to demonstrate a benefit of prophylactic MgSO₄ in preventing postoperative atrial fibrillation in cardiac surgery,²⁷ subsequent evaluations have found a distinct dose–response benefit when magnesium was added to cardioplegic solutions.⁵⁰ These data suggest a benefit of magnesium as a primary antidysrhythmic and the ability of magnesium to mitigate cardiotoxicity from other antidysrhythmics.

Torsade de pointes is a subset of polymorphic VT that often self-terminates and occurs in the setting of a prolonged QT interval.¹⁴ Not all polymorphic VT is torsade de pointes. This is a critical distinction in interpreting the literature. Previously unrecognized blockade of the cardiac rapidly activating delayed potassium rectifier current (hERG/KCNH2 gene) has led to multiple drug withdrawals and led the FDA to mandate hERG assays and electrocardiogram trials since 2005 to evaluate “QT liability” and the prodysrhythmic cardiovascular risk of new chemical entities.⁸⁰ Many drugs possess QT liability, which can be exacerbated in overdose or occur with coingestants that modify metabolism. The QT interval has limitations as a surrogate marker for torsade de pointes, and toxins may be torsadogenic through mechanisms other than directly blocking hERG, such as by interfering with channel synthesis or surface expression (eg, pentamidine).⁸⁰ Furthermore, QT interval prolongation in poisoning is likely multifactorial.

The 2015 American Heart Association (AHA) guidelines update for emergency cardiovascular care did not readdress the use of magnesium in cardiac arrest with polymorphic VT.⁸¹ These guidelines noted that routine use of magnesium for ventricular fibrillation (VF) or polymorphic VT was not recommended in adult patients on the basis of randomized trials showing no benefit of magnesium in patients with out-of-hospital cardiac arrest receiving cardiopulmonary resuscitation,⁴³ in-hospital cardiac arrest,¹¹⁸ and refractory or recurrent VF.^5,55,81 These guidelines apply to a different population than those encountered in poisoning. The prehospital trial⁴³ did not mention polymorphic VT (or torsade de pointes); the in-hospital cardiac arrest trial¹¹⁸ specifically excluded patients with clinical indications for magnesium and torsade de pointes, the remaining trials^5,55,81 evaluated patients with VF (not with torsade de pointes or polymorphic ventricular tachycardia), and all failed to ascribe or associate a case of dysrhythmia or arrest to a toxicologic cause. The guidelines acknowledged that the most common cause of polymorphic VT in these cases was myocardial ischemia.⁸¹ The importance of this distinction between ischemic versus xenobiotic-induced QT interval prolongation was highlighted by a study in which IV MgSO₄ (3.75 g over 3 minutes) in nine patients shortened QTc interval prolongation induced by amiodarone and quinidine (despite normal electrolytes and magnesium concentrations) but did not do so in those with a prolonged QTc interval because of ischemic coronary disease.⁵³ Low serum magnesium is associated with an increased risk of sudden cardiac death.⁶⁷ The AHA guidelines recommend “consideration of” IV MgSO₄ when polymorphic VT was secondary to a long QT interval (ie, torsade de pointes).⁸¹

Magnesium sulfate was first reported to treat three patients with torsade de pointes induced by quinidine and amiodarone, procainamide, and imipramine in 1984.¹²³ A subsequent case series of 6 and 12 patients demonstrated similar benefit for torsade de pointes, as well as a benefit in patients with intractable ventricular tachydysrhythmias despite normomagnesemia.^62,97,122 MgSO₄ also eliminated torsade de pointes in pediatric patients with congenital and acquired long QT syndrome.^59,60 Case reports variously describe the success of MgSO₄ in resolving QRS complex and QTc interval abnormalities in poisonings by escitalopram,¹⁰⁹ flecainide,¹²⁹ antipsychotics,⁹⁰ and arsenic trioxide.³⁸ There is significant divergence of opinion among toxicologists about what constitutes a prolonged or corrected QT interval, QT interval thresholds that require intervention (if any), and magnesium dosing (if any) in cases of drug-induced QT interval prolongation.⁹² Other than where described in the clinical trials earlier, magnesium does not typically affect the QT interval, making ascertainment of efficacy difficult. Based on the limited available data, it is reasonable to ensure eumagnesemia in patients with prolonged QT interval and administer magnesium in patients with progressive QT interval prolongation due to poisoning. After an episode of torsade de pointes, it is recommended to administer IV MgSO₄.^81,119

ROLE IN ETHANOL DISORDERS AND THIAMINE DEFICIENCY

Chronic alcohol use is associated with hypomagnesemia. In one series of 105 patients with heavy alcohol use, magnesium deficiency was present in 48%.³⁶ Magnesium deficiency correlated with worse cognitive test scores, even in the setting of normal thiamine pyrophosphate concentrations.³⁶ A smaller study found magnesium deficiency in 53% of patients admitted for detoxification.³⁵ In another study of 62 active drinkers, hypomagnesemia was present in 23% and directly correlated with QT interval prolongation.⁸⁷ Magnesium is an essential cofactor of the thiamine-dependent enzymes transketolase and pyruvate dehydrogenase, and magnesium increases the activity of a third critical thiamine-dependent enzyme α-ketoglutarate dehydrogenase.⁹⁴ In people with alcoholism and nutritionally deficient patients, Wernicke’s encephalopathy and thiamine deficiency remained refractory despite thiamine repletion until magnesium deficiency was addressed.^28,39,121 More than 4 decades ago, murine studies confirmed early human observations of the critical role of adequate magnesium in thiamine efficacy. Magnesium deficiency results in a loss of thiamine from tissues and limits the response to exogenous thiamine; magnesium supplementation augments the response to thiamine repletion in thiamine-deficient rats.¹³⁸ Furthermore, thiamine pyrophosphokinase, the enzyme that activates thiamine to thiamine pyrophosphate (thiamine diphosphate), requires magnesium as a cofactor.¹⁰⁶ In a controlled trial of 36 patients with chronic alcohol use disorders, those given MgSO₄ in addition to thiamine demonstrated significantly greater trans­ketolase activity.⁹⁶

Separately, magnesium was evaluated for its role in mitigating the consequences of alcohol withdrawal and neuronal hyperactivity. Hypomagnesemia is associated with higher structured scores of alcohol withdrawal.¹⁰¹ The degree of hypomagnesemia in patients with alcoholism correlates with stroboscopic-inducible seizure (photoconvulsion) thresholds, which itself correlates with incidence of spontaneous seizures and delirium tremens.¹³² Unfortunately, there are few appropriately designed studies, and they are often limited in their clinical scope. In a randomized, parallel group, double-blind trial of 119 patients with mild to moderate alcohol withdrawal designed to evaluate biochemical parameters, 8 weeks of oral daily magnesium (500 mg) sped recovery of aspartate aminotransferase (AST).¹⁰⁰ In a randomized, placebo-controlled trial of 59 consecutive patients with alcoholic liver disease, 2 days of 729 mg (30 mmol)/day of IV MgSO₄ followed by 6 to 7 weeks of 304 mg (12.5 mmol)/day of oral magnesium (as magnesium oxide) did not improve muscle mass or strength beyond placebo but did improve serum magnesium concentrations and was associated with higher muscle magnesium concentrations in subsequent analysis.¹ The study was confounded by a failure to control for spironolactone treatment, which is magnesium-sparing. In 49 patients with alcoholism randomized in a double-blind fashion to receive either 364 mg (15 mmol) oral magnesium daily (as magnesium-lactate-citrate tablets) or matching placebo for 6 weeks, magnesium supplementation improved aminotransferases, maximal muscle hand grip strength, and sense of general well-being.⁵² An early review of treatment of delirium tremens before the routine use of benzodiazepines found that MgSO₄ administration decreased the duration of tremor and visual hallucinations.⁴⁵ However, in one of the few studies evaluating acute withdrawal, a double-blind, placebo-controlled trial of 100 patients, four doses of IM MgSO₄ (2 g every 6 hours) did not alter chlordiazepoxide use or the incidence of diaphoresis, tremor, vomiting, hallucinations, or withdrawal severity.¹³⁰ The study was underpowered to detect a difference in grand mal seizures and delirium tremens. This is in contrast to the findings of a retrospective review of 781 patients with alcoholism being treated for detoxification, which documented a lower seizure incidence in the patients treated with magnesium.³¹ In a separate controlled study of eight patients with acute withdrawal, magnesium repletion either completely abolished inducible photoconvulsive responses or raised photoconvulsion thresholds.¹³² Given the significant prevalence of magnesium and thiamine deficiency in individuals with alcoholism and magnesium’s critical role for thiamine efficacy and its alteration of seizure thresholds in controlled human experimental studies, it is reasonable to actively screen for magnesium deficiency and administer magnesium to patients with hypomagnesemia, those with moderate to severe alcoholic withdrawal (including withdrawal seizures or associated with elevated aminotransferases), and those at risk for Wernicke’s encephalopathy.

The importance of magnesium to thiamine pyrophosphate function is also reflected in the guidance to administer magnesium and thiamine as adjunctive treatments for ethylene glycol poisoning.¹⁰ Metabolism of the ethylene glycol metabolite glyoxylic acid to 2-hydroxy-3-oxoadipate requires thiamine pyrophosphate and is facilitated by magnesium administration.⁸⁸ In one in vitro study, the lack of magnesium decreased glyoxylic acid metabolism to 2-hydroxy-3-oxoadipate by 60%.⁶⁶ Separately, glyoxylic acid covalently inactivates thiamine pyrophosphate, and thus the pyruvate decarboxylase complex, making ensuring sufficient thiamine of critical importance.⁴⁴ As a relatively benign intervention, it is reasonable to administer magnesium as adjunctive therapy to ethylene glycol–poisoned patients with normal kidney function while pursuing definitive management (eg, hemodialysis, fomepizole) as reviewed in Chap. 106.

ROLE IN PESTICIDE POISONING

Because magnesium diminishes calcium channel mediated synaptic acetylcholine release, it was hypothesized that MgSO₄ might ameliorate organic phosphorus (OP) effects at the neuromuscular junction and elsewhere.⁴⁰ Magnesium sulfate blunted paraoxon-induced tachycardia, hypertension, and hemoconcentration in mini pigs.^98,99 In four OP-poisoned patients, MgSO₄ administration produced no appreciable clinical or electrophysiological change when given 9 days into poisoning; when given 2 to 3 days into care, electrophysiological alterations consistent with conversion of an OP depolarizing blockade to a presynaptic blockade occurred, with no clinical change.¹¹² A rat model of sarin poisoning highlighted the distinct importance of distinguishing tonic-clonic convulsions from ongoing seizure. Although magnesium attenuated visible convulsions, sustained seizure activity and brain injury continued to occur.⁶⁵ In a small controlled trial of 11 patients given 4 g/day of IV MgSO₄ compared with 34 patients given usual care (multiple-dose activated charcoal, atropine, and oximes) the requirement for atropine and oximes did not differ, but the duration of hospitalization decreased, and the mortality rate decreased from 15% to 0%.⁹³ In a subsequent phase 2 trial, 40 patients received MgSO₄ 4, 8,12, or 16 g in a 4:1 treatment:control pseudorandomization.¹¹ Compared with six patients (60%) who died in the total control group, 3 of 16 (19%), 2 of 8 (25%), 1 of 8 (13%), and 0 of 8 (0%) patients died in the 4-g, 8-g, 12-g, and 16-g groups, respectively. A double-blind, prospective randomized clinical trial of 100 patients with OP poisoning found that 4 g of MgSO₄ IV reduced atropine requirements over 30 minutes and reduced the need for intubation and ICU stay but did not alter the mortality rate.¹²⁵ Separately, the cardiovascular complications of organophosphorus compounds are well described. Severe rhythm and conductivity disorders resulted in cardiac arrest and death in 29 patients poisoned with organic phosphorus compounds.⁸⁵ In a review of 168 organ, QT interval prolongation, ST segment and T wave anomalies were present in 134 (80%), and frequent dysrhythmias appeared in 56 patients (33%).^69,70 A separate study confirmed these findings and reported QT interval prolongation in 14 of 15 patients (93%) and malignant dysrhythmias in 6 of 15 (40%).⁸³ Magnesium sulfate (dose unspecified) was used to suppress ventricular ectopy is these cases.⁷⁰ Given the available data, it would be reasonable to provide MgSO₄ in organic phosphorus compound poisoning.

Metal phosphides (eg, aluminum and zinc), used as rodenticides and fumigants, have become a common cause of self-poisoning in certain countries (Chap. 108). For example, phosphine was found in 619 of 674 cases (92%) of suicidal poisonings in Tehran.⁷⁵ Magnesium concentrations in the setting of poisoning are quite variable. The lack of specific antidotes and the significant toxicity of phosphides led to the search for other interventions. Administration of magnesium in the form of the novel pharmacophore porphylleren–MC16 reversed aluminum phosphide–induced bradycardia, hypotension, energy depletion, and lipid peroxidation in mice.⁹ One human trial found decreased markers of oxidative stress in aluminum phosphide–poisoned patients given MgSO₄.²² Human studies conflict as to the efficacy of magnesium administration; some demonstrate efficacy and other no benefit, leaving magnesium’s role unclear.^8,102 Given the absence of any meaningful alternatives to supportive care in phosphide poisoning, it is reasonable to administer MgSO₄. In barium poisoning, oral MgSO₄ is administered to precipitate insoluble barium sulfate to preclude absorption and is reviewed in Chap. 107.

ROLE IN HYDROFLUORIC ACID AND FLUORIDE-RELEASING XENOBIOTICS

In addition to binding calcium, fluoride (Chap. 104) also binds to potassium and magnesium. Calcium administration (Antidotes in Depth: 32) remains a mainstay of treatment for hydrofluoric acid (HF) exposures. In murine studies, oral MgCl₂ administration for decontamination successfully increased the LD₅₀ of sodium fluoride, but oral MgSO₄ was unsuccessful in increasing survival in another pretreatment HF ingestion model.^56,76,84 Intravenous MgCl₂ increased the lethal dose and delayed death in a porcine HF ingestion model.²⁴ Clinical experience has demonstrated that relatively small skin surface HF burns (3.5% or less of concentration HF) in humans can lead to profound hypomagnesemia (in addition to hypocalcemia), which, if left unaddressed and not aggressively reversed (in addition to calcium administration), can contribute to death.^30,116,134 Survival and mitigation of lethal dysrhythmias in HF or sodium fluoride ingestion both typically require large amounts of magnesium and calcium.^2,114 It is reasonable to administer magnesium in cases of systemic toxicity from fluoride-releasing xenobiotics.

ADVERSE EFFECTS AND SAFETY ISSUES

The side effects of MgSO₄ administration were all well described in a series of 100 parenteral administrations in 1941. These included cardiovascular effects of cutaneous vasodilation; conduction disturbances such as bradycardia, hypotension, and cardiac arrest; respiratory depression; GI complaints of nausea, vomiting, and thirst; and neurologic sequelae of restlessness, confusion, a sense of impending doom, CNS depression, and paralysis.¹³¹ The usefulness of surveillance of patellar reflexes in anticipating respiratory depression was also recognized.¹³¹ Hypocalcemia and hyperkalemia are also reported with MgSO₄ treatment for preeclampsia.⁶¹ Patients with acute or chronic kidney disease or low urine flow rates are at increased risk for hypermagnesemia.^82,136 Frequent serum magnesium concentrations should be obtained. Magnesium sulfate inhibits platelet function in vitro and in vivo, which is reported to prolong bleeding times.^51,54

Case series and reports abound of iatrogenic hypermagnesemia through a variety of mechanisms. These include free-flowing (“runaway”) or misprogrammed pump infusions; confounding with IV bags of other infusing solutions (eg, oxytocin, lactated ringers); mislabeled infusions; 10-fold administration errors; and concentration, dilution, and admixture errors.^19,111 In the event of hypermagnesemia or hypocalcemia leading to significant symptoms (eg, respiratory muscle paralysis, hypotension, or dysrhythmia), IV calcium (Antidotes in Depth: 32) is used to counteract the effects.

PREGNANCY AND LACTATION

In 2013, the status of magnesium was changed from Pregnancy Class A to D (positive evidence of human fetal risk, but the potential benefits are acceptable in certain situations despite its risks). An FDA safety alert cautioned that continuous MgSO₄ administration to pregnant women for longer than 5 to 7 days was associated with fetal hypocalcemia, osteopenia, skeletal abnormalities, and fractures.¹²⁴ Magnesium sulfate was associated with neonatal cord serum concentrations of 70% to 100% of maternal concentrations but did not lower neonatal Apgar scores.¹⁵ Magnesium sulfate was not associated with congenital malformations.¹⁵

Magnesium distributes into breast milk during parenteral magnesium administration. Breast milk values in patients with preeclampsia given MgSO₄ were similar to those of control participants by 48 hours.¹⁵ Although caution should be exercised when MgSO₄ is administered to a nursing mother, the American Academy of Pediatrics considers MgSO₄ to be compatible with breastfeeding.⁷

DOSING AND ADMINISTRATION

Magnesium dosing varies significantly. In the setting of magnesium deficiency, several grams of MgSO₄ provided over the first day in divided doses of 1 to 2 g are often required for repletion, with frequent serum determinations to guide therapy. For torsade de pointes in adults, 2 g (25–50 mg/kg in children up to 2 g) of MgSO₄ intravenously is recommended.¹¹⁹ The caveat for the potential need for repeated doses may be missed. In perspective, 2 to 5 g or more of MgSO₄ was typically used in the cardiac conversion and dysrhythmia trials. A dose of 2 g in adults followed by an infusion or intermittent doses of 0.5 to 1 g/h to sustain serum (and intracellular) concentrations appears more rational. Again by comparison, 4 g of IV MgSO₄ is recommended to initiate therapy for severe preeclampsia or eclampsia for maternal and fetal neuroprotection, with maintenance doses typically of 1 to 2 g/h.^33,82,91 Empiric IV MgSO₄ administration for a prolonged QT interval is controversial; approximately 50% of toxicologists use a QTc threshold of 500 ms, although this was scenario dependent.⁹² Eumagnesemia (and normal potassium) should be ensured in patients with prolonged QT interval, and given the lack of consensus, it would be reasonable to administer magnesium in patients with progressive QT interval prolongation from poisoning. The recommended dose mirrors that for torsade de pointes: 2 g of MgSO₄ in adults followed by infusion or intermittent doses of 0.5 to 1 g/h with endpoints guided by the specific clinical circumstances. The rate of administration depends on the clinical scenario. For example, in a seizing patient with eclampsia, a dose of 4 g MgSO₄ is rapidly provided over 5 minutes.⁸² Use in life-threatening ventricular dysrhythmia would predispose to a similar rapid rate, recalling that care should prioritize electrical therapy (cardioversion or defibrillation) in unstable rhythms. In nonemergent scenarios, MgSO₄ is administered at a rate not to exceed 150 mg/min. If used in organic phosphorus compound poisoning, MgSO₄ doses mirroring the clinical trials (at least 4 g/day) are recommended. Magnesium requirements in fluoride poisoning are significant and guided by serum and clinical assessment. Magnesium sulfate should be diluted to a less than or equal to 20% concentration for IV administration.

FORMULATION AND ACQUISITION

A variety of MgSO₄ formulations are available for parenteral administration: 4% (40 mg/mL) or 8% (80 mg/mL) in sterile water volumes ranging from 100 mL to 1 L. Magnesium sulfate injection, 50% in water (500 mg/mL), is typically supplied as 2 or 10 mL. Magnesium sulfate is also formulated in 5% dextrose in concentrations ranging from 1% (10 mg/mL) to 2% (20 mg/mL) and in similar volume ranges.

Magnesium in much smaller concentrations (eg, 0.3 mg/mL) is found as a component of multiple or mixed electrolyte or parenteral nutritional infusion formulations (eg, in combination with acetate, chloride, potassium, and sodium). These would not be anticipated to be sufficient for antidotal ­purposes. Of note, dialysate solutions containing magnesium present the potential for medication error and should not be confused with those intended for IV infusion.

SUMMARY

• Magnesium sulfate is used to treat xenobiotic-associated hypomagnesemia and as an adjunctive therapy to treat cardiovascular toxins, fluoride toxicity, and pesticide poisoning.

• Ectopy in patients taking cardioactive steroids can be mitigated by MgSO₄ while awaiting definitive therapy (digoxin Fab fragments).

• Magnesium sulfate therapy and correction of electrolytes are mainstays of therapy to preclude recurrent torsade de pointes, although optimal dosing remains to be defined.

• The decision to intervene with MgSO₄ in isolated QT interval prolongation is complex. Repletion of hypomagnesemia should occur. An initial dose followed by an infusion is reasonable because cellular exchange of magnesium is slow compared with more rapid kidney excretion.

• Hypomagnesemia should be assessed and treated in patients with ­alcohol use disorders.

REFERENCES

HISTORY AND EPIDEMIOLOGY

Antithrombotics have numerous clinical applications, including in the treatment of coronary artery disease, cerebrovascular events, hypercoagulable states, deep vein thrombosis (DVT), and pulmonary embolism (PE). The antithrombotics are a diverse group of xenobiotics that are widely studied and constantly in the process of therapeutic evolution.

The origins and discovery of antithrombotics are extraordinary.^6,29,156,259 The discovery of modern-day oral anticoagulants originated after investigations of a hemorrhagic disorder in Wisconsin cattle in the early 20th century that resulted from the ingestion of spoiled sweet clover silage. The hemorrhagic agent, eventually identified as bishydroxycoumarin, would be the precursor to its synthetic congener warfarin (named after the Wisconsin Alumni Research Foundation). Warfarin was rapidly marketed as both a medicine and a rodenticide. “Superwarfarins” were subsequently developed for the rat population that had developed increasing genetic resistance to warfarin.

The origins of the anticoagulant heparin are equally fascinating. A medical student initially attempting to study ether soluble procoagulants derived from porcine intestines serendipitously found that, over time, these apparent “procoagulants” actually prevented normal blood coagulation. The phospholipid anticoagulant responsible for this effect would later be identified as a variant form of heparin. Shortly thereafter, the water-soluble mucopolysaccharide termed heparin (because of its abundance in the liver) was ­discovered. Unfractionated heparin (UFH) is a mixture of polysaccharide chains with varying molecular weights. After the identification of the active pentasaccharide segment of heparin in the 1970s, multiple low-molecular-weight heparins (LMWHs) were isolated, and synthetic forms were created.

Hirudin, a 65–amino acid polypeptide, was produced by the salivary glands of the medicinal leech (Hirudo medicinalis).²⁵⁶ Antistasin and antistasinlike proteins are naturally secreted by the Mexican leech, Haementeria officinalis, and the earthworm.^106,370 These xenobiotics have not been used therapeutically; however, they inspired the development of the synthetic factor inhibitors, such as direct thrombin inhibitors (DTIs) and factor Xa inhibitors.

In the late 19th century, human urine was noted to have proteolytic activity with specificity for fibrin. A substance found to be an activator of endogenous plasminogen leading to the consumption of fibrin, fibrinogen, and other coagulation proteins was isolated and purified and given the name urokinase. Streptokinase, a protein produced by β-hemolytic streptococci, tissue plasminogen activator (t-PA), and other synthetic thrombolytics were later discovered. Although known to exist for many years, ancrod, a purified derivative of Malayan pit viper, only recently gained therapeutic attention as a naturally occurring antithrombotic.

In the early 20th century, the antithrombotic properties of aspirin were noted in anecdotal reports of patients having a predisposition to bleeding while taking aspirin. Clinicians also noted lower rates of myocardial infarction (MI), and studies to elucidate the effects of aspirin on coagulation soon followed with further research exposing the role of platelet aggregation in thrombosis.²⁵¹ These discoveries led to the development of the antiplatelet xenobiotics.

The diversity of these antithrombotics led to ever-increasing use in many fields of medicine. Although warfarin is the most common oral anticoagulant in use today because of its utility in patients with cerebrovascular disease, cardiac dysrhythmias, and thromboembolic disease, the emergence of newly developed oral antithrombotics has changed the landscape of modern anticoagulation. During the period from 2011 to 2015, the American ­Association of Poison Control Centers has listed anticoagulants in the top 25 categories associated with the largest number of fatalities (Chap. 130). The total number of cases of reported antithrombotic exposures to the American Association of Poison Control Centers was 96,498 with 130 deaths during that 5-year period (Chap. 130). During that time period, new antithrombotics such as apixaban, edoxaban, and ticagrelor joined the market. In 2011, dabigatran and warfarin led the US Food and Drug Administration (FDA) Safety Information and Adverse Event Reporting Program’s list of adverse drug events, with 3,781 reports of serious adverse events associated with dabigatran, including 542 patient deaths. By comparison, warfarin alone accounted for 1,106 reports with 72 deaths.²⁵⁵ Since then, the FDA is continually reviewing dabigatran and initiated its own study that demonstrated lower risks of death, stroke, and intracranial hemorrhage but higher risks of gastrointestinal (GI) bleeding in patients anticoagulated with dabigatran compared with warfarin.³⁷¹

PHYSIOLOGY

Balance Between Coagulation and Anticoagulation

An understanding of the normal function of the coagulation pathways is essential to appreciate the etiology of a coagulopathy. This section summarizes the critical steps of the coagulation cascade. For additional details, readers are referred to Chap. 20 and several reviews.^{135,252,257,304}

Coagulation consists of a series of events that prevent blood loss and assist in the restoration of blood vessel integrity. Although the traditional understanding of the events that occur in the coagulation cascade,^91,231 as discussed later, adequately describe in vitro events, the current understanding emphasizes some distinct differences that occur in vivo.^135,257,304 Despite these differences, an understanding of the traditional model is most useful for interpreting the results of diagnostic tests of coagulation.

Within the cascade, coagulation factors exist as inert precursors and are transformed into enzymes when activated. Activation of the cascade occurs through one of two distinct pathways, the intrinsic and extrinsic systems (Fig. 58–1).^91,231 After being activated, these enzymes catalyze a series of reactions that ultimately converge to generate thrombin with the subsequent ­formation of a fibrin clot.

The intrinsic pathway is activated by the complexation of factor XII (Hageman factor) with high-molecular-weight kininogen (HMWK) and prekallikrein or vascular subendothelial collagen. This results in sequential activation of factor XII, active kallikrein, active factors IX to XI, and prothrombin (factor II) (Fig. 58–1). Prothrombin is converted to thrombin in the presence of factor V, calcium, and phospholipid. The integrity of this system is usually evaluated by determining the partial thromboplastin time (PTT).

In the extrinsic or tissue factor–dependent pathway, a complex is formed between factor VII, calcium, and tissue factor, which is released after injury. A calcium- and lipid-dependent complex is then created between factors VII and X. The factor VII–X complex subsequently converts prothrombin to thrombin, which promotes the formation of fibrin from fibrinogen (Fig. 58–1). The integrity of this pathway is usually assessed by determining the prothrombin time (PT or international normalized ratio {INR}¹²⁰). The distinction between PT and INR is discussed in Chap. 20.

Activation of factor X provides the important link between the intrinsic and extrinsic coagulation pathways. Additional evidence that tissue factors can activate both factors IX and X suggests that there are more interrelations between the two pathways.²⁷⁵ Furthermore, cell surfaces facilitate the process of clotting. Platelets are also known to interact with proteins of the coagulation cascade through surface receptors for factors V, VIII, IX, and X.^142,265,339 As a final step, factor XIII assists in the cross-linking of fibrin to form a stable thrombus.

Antithrombin (AT), and proteins C, S, and Z serve as inhibitors, maintaining the homeostasis that is required to prevent spontaneous clotting and keep blood fluid. Protein C, when aided by protein S, inactivates two plasma factors, V and VIII.^46,73,135 Protein Z is a glycoprotein (GP) molecule that forms a complex with the protein Z–dependent protease inhibitor (ZPI) which, in turn, inhibits the activated factor X (Xa).³⁸⁰ Antithrombin complexes with all the serine protease coagulation factors (factor Xa, factor IXa, and contact factors, including XIIa, kallikrein, and HMWK), except factor VII.^46,135,304

Thrombolytics such as streptokinase, urokinase, anistreplase, and recombinant tissue plasminogen activator (rt-PA) enhance the normal processes that lead to clot degradation.²⁵⁷

Thrombosis is initiated when exposed endothelium or released tissue factors lead to platelet adherence and aggregation, the formation of thrombin, and cross-linking of fibrinogen to form fibrin strands.^135,257,304 This results in a hemostatic plug or thrombus formation. Thrombus formation, in turn, leads to generation of plasmin from plasminogen, which causes fibrinolysis and eventual dissolution of the hemostatic plug.^74,75 Thus, the fibrinolytic system is a natural balance against unregulated coagulation. Thrombolytic therapy increases fibrinolytic activity by accelerating the conversion of plasminogen to plasmin, which actively degrades fibrin.^74,75 After the administration of thrombolytics, a drug-induced coagulopathy ensues, and fibrin degradation products (FDPs) are elevated ­secondary to the rapid turnover of clot.

DEVELOPMENT OF COAGULOPATHY

Impaired coagulation results from decreased production or enhanced consumption of coagulation factors, the presence of inhibitors of coagulation, activation of the fibrinolytic system, or abnormalities in platelet number or function. Platelets are involved in the initial phases of clotting after blood vessel injury by assisting in the formation of the fibrin plug.

Decreased production of coagulation factors results from congenital and acquired etiologies. Although congenital disorders of factor VIII (hemophilia A), factor IX (Christmas factor Hemophilia B), factor XI, and factor XII (Hageman factor) are all reported, their overall incidence is still quite low. Clinical conditions that result in acquired factor deficiencies are much more common and result from either a decrease in synthesis or activation. Factors II, V, VII, and X are entirely synthesized in the liver,^135,257,304 making hepatic dysfunction a common cause of acquired coagulopathy. In addition, factors II, VII, IX, and X also require postsynthetic modification by an enzyme that uses vitamin K as a cofactor,^353,358,359 such that vitamin K deficiency (from malnutrition, changes in gut flora secondary to xenobiotics, or malabsorption), and inhibition of vitamin K cycling (from warfarin) are capable of impairing coagulation.

Excessive consumption of coagulation factors usually results from massive activation of the coagulation cascade. Massive activation occurs during severe bleeding or disseminated intravascular coagulation (DIC). The latter results from infection, such as sepsis, and from conditions that introduce tissue factor into the blood, such as neoplasms, snake envenomations, stagnant blood flow, diffuse endothelial injury secondary to hyperthermia, ruptured aortic aneurysm, or aortic dissection. The hallmark of a consumptive coagulopathy is a depressed concentration of fibrinogen with an elevation of FDPs. This combination suggests the rapid turnover of fibrin in the coagulation process. In the other coagulopathic conditions, the failure to activate the coagulation cascade is associated with normal or high fibrin concentrations and low FDPs because of limited clot formation.

Hypothermia is a well-known cause of DIC leading to death.³⁶⁵ Canine studies show decreased fibrinogen, factor II, and factor VII concentrations when cooled to 68°F (20°C) while a simultaneous rise in factor V concentration occurs.¹⁴⁰ In vitro studies show increased coagulation time, clot formation time, and maximum clot strength in hypothermic whole-blood samples from healthy volunteers.^103,311 In addition, thrombocytopenia occurs secondary to sequestration in the spleen, liver, and splanchnic circulation.^301,355 Infants with hypothermia have an increased risk for intracranial and pulmonary hemorrhage.⁶²

Inhibitors of the coagulation cascade (circulating anticoagulants) are of two types: immunoglobulins to coagulation factors or antibodies to phospholipid membrane surfaces. Immunoglobulins develop without obvious cause, are part of a systemic autoimmune disorder, or result from repeated transfusions with exogenous factors (as occurs in patients with hemophilia).^162,209,330 Antibodies to factors V, VII to XI, and XIII are described.^35,330 The clinical syndromes associated with antibody inhibitors are similar to those associated with deficiencies of the particular coagulation factors involved. Antiphospholipid antibodies are directed against phospholipid membrane surfaces and β₂-GP I, also known as apolipoprotein H. This protein is essential in preventing hemostasis by inhibiting adenosine diphosphate (ADP)–induced platelet aggregation, inhibiting activation of the intrinsic pathway of the coagulation cascade, and inhibiting both platelet-mediated factor Xa and factor VII activation. Therefore, destroying β₂-GP I creates a prothrombotic clinical status.^324,325,334 Paradoxically, this antibody creates a prolonged PTT because the antibody also binds to most phospholipid-containing PTT reagents. This reduction in amount of active reagent results in a falsely elevated PTT.^162,209

GENERAL MANAGEMENT

Gastrointestinal decontamination should be performed on patients who are believed to have potentially significant life-threatening ingestions unless they already present with significant bleeding. For patients who present after a few hours of ingestion, gastric emptying is not indicated (Chap. 5). In general, a single dose of activated charcoal (AC) decreases absorption of some anthrombotics and should be given unless contraindicated.^272,377,391 Oral cholestyramine enhances warfarin elimination,²⁹⁹ but no studies are available that compare these two therapies or evaluate the role of combined AC and cholestyramine therapy. Therefore, early administration of AC after ingestion is the preferred form of GI decontamination, even for patients who overdose on warfarin. In addition to general supportive measures, the patient should be placed in a supervised medical and psychiatric environment that offers protection against external or self-induced trauma and permits observation for the onset of coagulopathy.

Blood transfusion is required for any patient with a history of blood loss or active bleeding who is hemodynamically unstable, has impaired oxygen transport, or is expected to become unstable. Although a transfusion of packed red blood cells (PRBCs) is ideal for replacing lost blood, it cannot correct a coagulopathy, so patients will continue to bleed. Massive transfusion protocols (MTPs) aim to correct coagulopathy with an array of blood products, electrolytes, and antifibrinolytics. If available, it is reasonable to use an MTP for unstable patients. If an MTP is not available, whole blood is reasonable in severe cases because it contains many components, including platelets, white blood cells, and non–vitamin K–dependent factors.

VITAMIN K ANTAGONISTS

Warfarin and “Warfarinlike” Anticoagulants

The vitamin K antagonist (VKA) anticoagulants can be divided into two groups: (1) hydroxycoumarins, including warfarin, difenacoum, coumafuryl, fumasol, prolin, ethyl biscoumacetate, phenprocoumon, dicumarol, bishydroxycoumarin, and acenocoumarin, and (2) indanediones, including chlorophacinone, diphacinone, diphenadione, phenindione, and anisindione. Regardless of the classification, their mechanism of action involves inhibition of the vitamin K cycle. Vitamin K is a cofactor in the postribosomal synthesis of clotting factors II, VII, IX, and X (Fig. 58–2). The vitamin K–sensitive enzymatic step that occurs in the liver involves the γ-carboxylation of 10 or more glutamic acid residues at the amino terminal end of the precursor proteins to form a unique amino acid γ-carboxyglutamate.^{104,353,358,359} These amino acids chelate calcium in vivo, which allows the binding of the four vitamin K–dependent clotting factors to phospholipid membranes during activation of the coagulation cascade.⁴⁰⁰

Vitamin K is inactive until it is reduced from its quinone form to a quinol (or hydroquinone) form in hepatic microsomes. This reduction of vitamin K must precede the carboxylation of the precursor factors. The carboxylation activity is coupled to an epoxidase activity for vitamin K, whereby vitamin K is oxidized simultaneously to vitamin K 2,3-epoxide (Fig. 58–2).^358,399 This inactive form of the vitamin is converted back to the active form by two successive reductions.^{104,233,277,395} In the first step, an epoxide reductase (known as vitamin K 2,3-epoxide reductase) uses reduced nicotinamide adenine dinucleotide (NADH) as a cofactor to convert vitamin K 2,3-epoxide to a quinone form.^270,271,358 Subsequently, the quinone is reduced to the active vitamin K quinol form (Antidotes in Depth: A18).

Warfarin is a racemic mixture of R warfarin and S warfarin enantiomers. In humans, S warfarin is approximately 2.7 to 3.8 times more potent than R warfarin.³⁷ Warfarin and all warfarinlike compounds inhibit the activity of vitamin K 2,3-epoxide reductase, as is demonstrated by the observation of elevated concentrations of vitamin K 2,3-epoxide.^69,403 Vitamin K quinone reductase is also inhibited by warfarin and its related compounds (Fig. 58–2).^104,121 This reduction in the cyclic activation of vitamin K subsequently inhibits the formation of activated clotting factors.

Pharmacology of Vitamin K Antagonists

Oral warfarin is well absorbed, and peak serum concentrations occur approximately 3 hours after administration.³⁵⁷ Because only the free warfarin is therapeutically active, concurrent administration of xenobiotics that alter the concentration of free warfarin, either by competing for binding to albumin or by inhibiting warfarin metabolism, markedly influences the anticoagulant effect.^24,126,357 The pharmacologic response to warfarin is a polygenic trait with approximately 30 genes contributing to its therapeutic effects.¹⁹⁵ Readers are referred to a review of xenobiotics and foods that potentiate warfarin’s effects.³⁹⁸ Although vitamin K regeneration is altered almost immediately, the anticoagulant effect of warfarin and other warfarinlike anticoagulants is delayed until the existing stores of vitamin K are depleted and the active coagulation factors are removed from circulation. Because vitamin K turnover is rapid, this effect is largely dependent on factor half-life (t_1/2), with factor VII (t_1/2 ∼5 hours) depleted most rapidly.¹²⁶ For a prolongation of the INR to occur, factor concentrations must fall to approximately 25% to 30% of normal values.⁷² Assuming complete inhibition of the vitamin K cycle, this suggests that most patients require at least 15 hours (three ­factor VII half-lives) before the effect of warfarin is evident on the INR.¹²³ In fact, because complete inhibition does not occur, the onset of coagulation is delayed even further.

Because the half-life of warfarin in humans is 35 hours, its duration of action is documented to be as long as 5 days.^50,357 On average, it takes approximately 6 days of warfarin administration to reach a steady-state anticoagulant effect.

R warfarin is metabolized by enzymes CYP1A2 and CYP3A4, and S warfarin is metabolized by CYP2C9. Whereas R warfarin is metabolized by side-chain reduction to secondary alcohols that are subsequently excreted by the kidney, S warfarin is metabolized by hydroxylation to 7-hydroxy warfarin, which is excreted into the bile.³⁵⁷ The elimination of S warfarin is more rapid than that of R warfarin.⁴⁹

Dosing of warfarin and other VKAs is problematic in many patients. In one study, genetic polymorphisms of the vitamin K epoxide reductase complex 1 (VKORC1) and CYP2C9 genes were the strongest predictors of interindividual variability in the anticoagulant effect of warfarin.¹⁹⁵ Pharmacogenomic research with complex xenobiotics, such as warfarin, improves safety of treatment and predicts or prevents interactions with other ­xenobiotics. Although the FDA has approved a commercially available test to identify variants within these genes,¹⁹³ current guidelines do not advocate genetic testing to guide VKA dosing.¹⁵⁴

Within the coumarin group are two 4-hydroxycoumarin derivatives, ­difenacoum and brodifacoum, that differ from warfarin by their longer, higher molecular weight polycyclic hydrocarbon side chains. Together with chlorophacinone, an indandione derivative, they are known as superwarfarins or long-acting VKAs.

Long-acting VKAs were designed to be effective rodenticides in warfarin-resistant rodents.²³⁰ Their mechanism of action is identical to that of the traditional warfarinlike anticoagulants, as demonstrated by increased concentrations of vitamin K 2,3-epoxide after administration.^{48,51,56,210,277} The ability of these xenobiotics to perform as superior rodenticides is attributed to their high lipid solubility and concentration in the liver.^210,230,277 They also saturate hepatic enzymes at very low concentrations, as demonstrated by zero-order elimination after overdose.⁵⁶ These factors make them about 100 times more potent than warfarin on a molar basis.^210,230,277 In addition, they have a longer duration of action than the traditional warfarins.^210,230,277 For example, to obtain 100% lethality in a mouse, more than 21 days of feeding with a warfarin-containing rodenticide (0.025% anticoagulant by weight of bait) is required.²³⁰ Similar efficacy is achieved with a single ingestion of brodifacoum (0.005% anticoagulant by weight of bait).²³⁰ Furthermore, more concentrated liquid formulations, such as brodifacoum (0.5% anticoagulant), are available through illegal vendors and are implicated in prolonged coagulopathy in humans.¹³⁸ It should also be noted that although ingestion of these xenobiotics is the most common route of exposure and subsequent cause of toxicity, dermal absorption of liquid preparations occurs, also resulting in a coagulopathy.³⁴⁴

Clinical Manifestations

Although intentional ingestions of warfarin-containing products are uncommon, adverse drug events resulting in excessive anticoagulation and bleeding frequently occur. The risk of bleeding during VKA therapy depends on a myriad of factors, including the intensity of anticoagulation, patient characteristics, and comorbid conditions. Clearly, the most serious complication of excessive anticoagulation is intracranial bleeding, which is reported to occur in as many as 2% of patients on long term therapy, which is an 8- to 10-fold increase in risk compared with patients who are not anticoagulated.^125,126,405 The overall risk of major bleeding in patients can be estimated using tools such as the Hypertension, Abnormal Renal/Liver Function, Stroke, Bleeding History or Predisposition, Labile INR, Elderly, Drugs/Alcohol Concomitantly (HAS-BLED) score, which was prospectively validated in patients with atrial fibrillation.^225,285 Bleeding ranges between 1 and 12.5 episodes per 100 patient-years, with a higher risk in patients with multiple risk factors.²⁸⁵ Another study showed that patients older than the age of 80 years have a cumulative incidence of major bleeding at 13.1 per 100 person-years with major bleeding defined as bleeding at serious sites (eg, intracranial, retroperitoneal, intraspinal, or pericardial) or bleeding that results in blood transfusion or death. Patients between the ages of 65 and 80 years have a lower risk of 4.7 major episodes of bleeding per 100 person-years.¹⁷⁶ Major bleeding complications are associated with a fatality rate as high as 77%.²³⁹ In patients with intracranial bleeding a decreased level of consciousness and an increased size of hematoma are predictors of poor prognosis.⁴¹⁹ Somewhat surprisingly, the degree of INR elevation was not associated with worse outcome.⁴¹⁹

Many cases of intentional overdose of long-acting VKAs in humans are reported. The clinical courses of these patients are characterized by a severe coagulopathy that last weeks to months, often accompanied by consequential blood loss. Most patients do not seek medical care until bruising or bleeding is evident,^{19,56,57,119,170,194} which often occurs many days after ingestion. The most common sites of bleeding are the GI and genitourinary tracts. In one study describing 12 patients with surreptitious ingestion of oral anticoagulants, nine were health care professionals.²⁷¹ These patients presented with bruising, hematuria, hematochezia, and menorrhagia. Bleeding into the neck with resultant airway compromise is a rare but life-threatening complication.⁴³

Patients with unintentional ingestions must be distinguished from those with intentional ingestions because the former demonstrate a low likelihood of developing a coagulopathy and have rare morbidity or mortality. Most patients (usually children) are entirely asymptomatic and have a normal coagulation profile after an acute unintentional exposure. Typical warfarinlike rodenticides contain only small concentrations of anticoagulant, 0.005% (or 5 mg of brodifacoum per 100 g of product). A 10-kg child would require an initial dose of 10 g of rodenticide (8 pellets, each of which is approximately 25 mm in size). These quantities are far greater than those that occur in typical “tastes.” Thus, single small unintentional ingestions of warfarin-containing rodenticides pose a minimal threat to normal patients.¹⁹⁰ Prolongation of the INR is unlikely with a single small ingestion of a long-acting VKA rodenticide. Clinically significant anticoagulation is even rarer. In a combined pediatric case series, prolongation of the INR occurred in only 8 of 142 children (5.6%) reported with single small ingestions of long-acting VKAs.^{28,189,190,342} Only one child in this group was reported to have “abnormal prolonged bleeding,” but this required no medical attention.³⁴² In a single case report, a 36-month-old child developed a coagulopathy manifested by epistaxis and hematuria, with anticoagulation persisting for more than 100 days after a presumed, but unwitnessed, single unintentional ingestion of brodifacoum.³⁶⁶ Clinically significant coagulopathy can result, however, after small repeated ingestions. Two children reportedly became poisoned by repeated ingestions of a long-acting anticoagulant. One child presented with a neck hematoma that compromised his airway and the other with a hemarthrosis.¹⁵¹ Similarly, a 7-year-old girl required multiple hospitalizations over a 20-month period after repeated unintentional ingestions of brodifacoum.³² Finally, a 24-month-old child who presented with unexplained bruising and a PT greater than 125 seconds was the victim of brodifacoum poisoning secondary to Munchausen syndrome by proxy, which is currently called pediatric ­condition falsification (Chap. 31).¹⁵

Laboratory Assessment

Although warfarin concentrations are useful to confirm the diagnosis in unknown cases and to study drug kinetics,^155,268 the routine use of simple and inexpensive measures such as INR determination seems more appropriate (Table 58–1). When blood loss is evident, serial determinations of hemoglobin concentration are indicated.

For patients with an acute and significant long acting VKA overdose, daily INR evaluations for 2 days are adequate to identify most patients at risk for coagulopathy. Earlier detection through direct coagulation factor analysis is available but is typically reserved to trend patients with known coagulopathy for treatment purposes.^155,170 Concentrations of long-acting VKAs can now be measured, although they are usually only performed in reference laboratories.^199,268 We recommend that children with possibly significant exposures be followed up with at least a single INR at least 48 hours after the exposure. Even though significant toxicity from long-acting VKAs is rare, it should be recognized that the reported benign courses of exposures in children are somewhat misleading. Multiple retrospective studies suggest that children with unintentional acute exposures do not require any follow-up coagulation studies.^{256,260,279,332} However, this conclusion and approach to management is an unjustified attempt to decrease the cost of “unnecessary” coagulation studies. There are clearly insufficient data to justify this conclusion because many of these “exposed” children were never documented to have ingested long-acting VKAs (Chap. 130). We recommend that clinicians continue to manage these children as possibly significant exposures and that all children be followed up with at least a single INR at a minimum of 2 days after the exposure. A baseline INR is usually unnecessary but is reasonable if there is a suspicion of an antecedent ingestion. A baseline INR is also helpful when chronic exposure is suspected.

Treatment of Vitamin K Antagonist–Induced Coagulopathy

Life-threatening bleeding secondary to VKA toxicity should be immediately reversed with factor replacement followed by vitamin K₁ (Table 58–1). The American College of Chest Physicians (ACCP) and the FDA recommend factor replacement with four-factor prothrombin complex concentrate (PCC) as a first-line treatment for major bleeding in the setting of VKA-induced coagulopathy.¹⁷² This recommendation is reasonable given the ease of administration as compared with fresh-frozen plasma (FFP). Most countries use traditional four-factor PCC, which contains concentrated factors II, VII, IX, X, C, and S. Dosing of PCC is based on INR and body weight. The typical doses range between 25 and 50 units/kg with the largest dose based on a maximum weight of 100 kg. The PCC is contained in both 500-unit and 1,000-unit vials. However, the factor IX content of each vial is variable: 500-unit vial may contain between 400 to 620 units/vial, and a 1,000-unit vial may contain between 800 and 1,240 units/vial. Each individual vial states the actual factor IX content in each bottle, and the dosage should be based on the actual factor IX content in each vial.⁸⁸ Most forms of PCC, contain small amounts of heparin. These heparin-containing PCCs are contraindicated in patients with a history of heparin-induced thrombocytopenia (HIT) and heparin-induced thrombocytopenia and thrombosis syndrome (HITT). In those cases, factor eight inhibitor bypassing activity (FEIBA), an activated prothrombin complex concentrate (aPCC), is recommended. Activated aPCCs contain factors II, VII, IX, and X in inactive and activated forms. Factor eight inhibitor bypassing activity is typically used to treat patients with hemophilia A and B. Unfortunately, FEIBA has not been studied extensively in the setting of VKA-induced coagulopathy. Therefore, there is no standard dose for coagulopathy reversal. Typical doses for treating bleeding or for perioperative management range between 50 and 100 units/kg with the dose based on the amount of factor VIII inhibitor bypass activity.²³ Despite its increased cost, the rationale for using PCC instead of FFP includes less risk of infection transmission.¹⁷² In addition, small-volume factor replacement is preferable in patients at risk of volume overload, such as patients with heart failure, chronic kidney disease (CKD), or intracranial bleeding.² Unfortunately, there are only a few studies comparing PCC with FFP in reversing coagulopathy. These studies show that PCC safely produces complete INR reversal faster than FFP with more consistent factor IX replacement.^95,185,234 However, unequal factor replacement in these small studies favors more factor replacement in patients receiving PCC.³¹⁵ Fresh-frozen plasma is rich in active vitamin K–dependent coagulation factors and will reverse oral anticoagulant-induced coagulopathy in most patients. In general, approximately 15 mL/kg of FFP should be adequate to reverse any VKA-induced coagulopathy.⁸⁷ However, the specific factor quantities and volume of each unit may be varied, leading to an unpredictable response.²³⁴ In addition, delay to FFP administration is accentuated by requirements for blood type matching and thawing. Therefore, we recommend in acute major bleeding in the setting of VKA-induced coagulopathy the use of PCC as the preferred reversal agent.

Activated recombinant factor VII (rFVIIa) is approved for patients with hemophilia or various factor inhibitors and has successfully reversed warfarin toxicity.²⁸⁴ The safety of off-label rFVIIa to reverse VKA coagulopathy is unclear. Preliminary data using rFVIIa suggested its utility for bleeding secondary to warfarin-induced excessive anticoagulation and a case series showing beneficial effects in four patients with long acting VKA toxicity.⁴²⁰ However, adverse outcomes, such as arterial thrombotic events, occur at higher than acceptable rates. Initial concerns for thromboembolic adverse events were raised when review of the FDA’s Adverse Event Reporting System database found that the majority of these complications arose in patients who received rFVIIa for off-label purposes. Cerebrovascular accident, acute MI, PE, venous thrombosis, and clotted devices all occurred at higher rates in off-label applications.²⁶⁹ The Factor Seven for Acute Hemorrhagic Stroke Trial (FAST) showed that arterial thromboembolic adverse events increased in a dose-dependent fashion with the administration of rFVIIa.¹⁰² A subsequent meta-analysis confirmed that an increased rate of arterial thrombosis occurs in patients who receive rFVIIa.²¹⁹ The risk of continued bleeding versus the benefit of rapid reversal of coagulopathy with rFVIIa is unknown, and further experience with rFVIIa is necessary to determine its safety and efficacy in anticoagulant-induced bleeding. We do not recommend routine use of rFVIIa for reversing VKA-associated coagulopathy. If FFP, PCC, and FEIBA are not available, then it is reasonable that rFVIIa be given in life-threatening situations.¹ It should also be noted that if rFVIIa is used, assays based on PT are inaccurate and should be avoided when administering rFVIIa.¹⁹¹

Treatment with vitamin K₁ takes several hours to activate enough factors to reverse a patient’s coagulopathy,^234,278 and this delay is potentially fatal. Repetitive, large doses of vitamin K₁ (on the order of 60 mg/d or greater) are reported in some patients with massive ingestion.^155,271 If complete reversal of INR prolongation occurs or is desirable (as in most cases of life-threatening bleeding) and the underlying medical condition of the patient still requires some degree of anticoagulation, the patient can then receive anticoagulation with heparin after the bleeding is controlled and clinical stability restored. Heparin anticoagulation was used without apparent bleeding complications in 25% of patients in one cross-sectional study.⁴⁰²

Vitamin K₁ is preferable over the other forms of vitamin K; the other forms are ineffective^{183,262,270,373} are potentially toxic,¹⁷ and are unavailable in the United States. Parenteral administration of vitamin K₁ (phytonadione) is traditionally preferred as initial therapy by many authors, but success can also be achieved with early oral therapy, especially when the coagulopathy is not severe.⁵⁶ In most cases, the patient can be switched to oral vitamin K₁ for long-term care. Vitamin K₁ can be administered intramuscularly, subcutaneously, intradermally, or intravenously. Although intravenous (IV) therapy has the most rapid onset of action of all routes of delivery, its use as the sole therapeutic is still associated with a delay of several hours^278,395 and carries the added risk of anaphylactoid reactions.³⁰² The use of low doses and slow rate of administration reduces this risk³³⁵ (Antidotes in Depth: A18). In cases in which oral administration is undesirable, for example, with significant GI bleeding the subcutaneous (SC) route should not be used because absorption is erratic. In this case, we recommend the IV route of administration of vitamin K₁ because the benefits outweigh the risks.

For patients with non–life-threatening bleeding, the clinician must evaluate whether anticoagulation is required for long-term care. In patients not requiring chronic anticoagulation we recommend treating even small elevations of the INR with vitamin K₁ alone to prevent deterioration in coagulation status and reduce the risk of bleeding. In contrast to ingestions of warfarin, we recommend not to give prophylactic vitamin K₁ to asymptomatic patients with unintentional ingestions of long-acting warfarinlike anticoagulants because (1) if the patient develops a coagulopathy, it will last for weeks, and the one or two doses of vitamin K₁ given will not prevent complications; (2) a gradual decline in coagulation factors occurs over the first day of anticoagulation, so an individual would not be expected to develop a life-threatening coagulopathy in 1 or 2 days; and (3) after vitamin K₁ is administered, the onset of an INR abnormality will be delayed, which could impair the ability of the clinician to recognize a coagulation abnormality, possibly requiring the patient to undergo an unnecessarily prolonged observation period.

For patients requiring chronic anticoagulation we agree with the ACCP guidelines for the management of patients with elevated INRs (Table 58–2). It should also be noted that low-dose vitamin K is safe to administer to patients with mildly elevated INRs (4–10) to decrease the INR more ­rapidly; however, one study did not demonstrate decreased bleeding in the treatment group.⁸⁶ Furthermore, simply omitting warfarin doses is usually adequate for a patient without active bleeding who has an INR between 4 and 9.⁸⁶

It is often unclear why patients with consistent therapeutic dosing have seemingly random elevations in their INR. A case-control study identified the following risk factors associated with overanticoagulation from VKAs: previous medical history of increased INR, antibiotic therapy, fever, and concomitant use of amiodarone and proton pump inhibitors.⁶⁰ Clinicians should pay particular attention to patients with these conditions, and close monitoring of coagulation profiles should be performed.⁶⁰

Treatment of Long-Acting Vitamin K Antagonist Overdoses

In patients with unintentional, small ingestions of long-acting VKAs, the risk of coagulopathy is low. However, it takes days to develop coagulopathy. Administration of AC is recommended to prevent absorption in acute ingestions if no contraindication exists.^190,342

Treatment of a patient with a coagulopathy resulting from a long-acting anticoagulant overdose is essentially the same as the treatment of oral anticoagulant toxicity with certain exceptions. Although initial parenteral vitamin K₁ doses as high as 400 mg have been required for reversal,⁵⁷ daily oral vitamin K₁ requirements are often in the range of 50 to 200 mg. Recent experience in both animals and humans suggests that parenteral vitamin K₁ therapy is not required after initial stabilization (Antidotes in Depth: A18).^56,411

In one case report, a 2-year-old child with confirmed brodifacoum-associated coagulopathy developed anemia, epistaxis, ecchymosis and GI bleeding. She developed an anaphylactoid reaction after several doses of IV vitamin K₁. Patients typically tolerate IV vitamin K₁ well as long as it is administered appropriately. In stable patients, oral vitamin K is recommended. However, because oral vitamin K was not available at this hospital, the patient received five plasma exchange treatments, which resulted in normalization of her coagulation studies. In addition, repeat brodifacoum concentrations were undetectable after her third plasma exchange treatment.⁹⁴ Because there are other less invasive methods of correcting VKA-induced coagulopathy and lack of established experience with plasma exchange beyond this case report, we do not recommend plasma exchange therapy for VKA-induced coagulopathy.

Long-acting VKAs are metabolized by the cytochrome P450 system.^16,270 In a rat model, the duration of coagulopathy was shortened by administering phenobarbital, a CYP3A4 inducer.¹⁶ Although phenobarbital has never been systematically studied in humans, this approach was used by several authors in isolated human cases of long-acting anticoagulant toxicity.^{57,183,226,366,393} These anecdotal reports suggest some improvement with phenobarbital therapy, but the risk of sedation in a patient who might be prone to bleeding complications outweighs any purported benefit.

Patients with long-acting VKA overdose should be followed until their coagulation studies remain normal without treatment for several days. This usually requires daily or even twice-daily INR measurements until the INR is at the lower limit of the therapeutic range. Monitoring of serial INR measurements should allow for a gradual decrease in vitamin K₁ requirement over time. Periodic coagulation factor analysis (particularly factor VII), however, provides an early marker of toxicity resolution.¹⁷⁰ An anticoagulated patient will require weeks to months of close observation for both psychiatric and medical management. A critical superwarfarin concentration below which anticoagulation does not occur is not defined.⁵⁷ In one case report, brodifacoum was observed to follow zero-order elimination kinetics.⁵⁶

Nonbleeding Complications of Vitamin K Antagonists

Warfarin therapy is associated with three nonhemorrhagic lesions of the skin: urticaria,³²³ purple toe syndrome,¹²² and warfarin skin necrosis.^{77,193,202,242,384} Although warfarin skin necrosis was once thought to be a rare and idiosyncratic reaction,^193,202 more recent evidence suggests a link between this disorder and protein C deficiency.^202,384 Protein C synthesis is also dependent on vitamin K.⁷³ Patients who are homozygotes for protein C deficiency have an increased incidence of thrombosis and embolic events, such that they often require long-term anticoagulant therapy.⁷³ Because the half-life of protein C is shorter than that of many of the vitamin K–dependent coagulation factors, protein C concentrations fall rapidly during the first hours of warfarin therapy. This results in an imbalance that actually favors coagulation, and skin necrosis results because of microvascular thrombosis in dermal vessels.^242,384 Although warfarin skin necrosis is more common in patients with protein C deficiency, this disorder is also described in patients with protein S and AT deficiencies.⁷⁷ Unfortunately, these deficiencies are neither necessary nor sufficient to account for the incidence of warfarin necrosis.⁷⁷ If necrosis occurs, warfarin should be discontinued, and heparin should be initiated to decrease thrombosis of postcapillary venules. Some patients also require surgical debridement.³¹⁴

The purple toe syndrome, in contrast to warfarin-induced skin necrosis, is presumed to result from small atheroemboli that are no longer adherent to their plaques by clot (Fig. 17–13).

Warfarin-related nephropathy (WRN) is a well-recognized form of nonhemorrhagic warfarin toxicity. In patients with CKD, an acute increase in the INR above 3 was associated with an increase in serum creatinine and accelerated reduction in kidney function. The etiology of the WRN is attributed to acute tubular injury and glomerular bleeding, demonstrated by the finding of red blood cell (RBCs) filling Bowman’s capsule and RBCs casts obstructing glomeruli on specimens from kidney biopsies.^54,55 Further studies that retrospectively reviewed kidney function in patients on long-term warfarin anticoagulation showed 33% of patients with CKD and 16.5% of patients without CKD developed WRN. Warfarin-related nephropathy is associated with a mortality rate of 31.1%, a significant increase in risk compared with 18.9% in patients without WRN.⁵⁴

An additional major nonhemorrhagic complication of warfarin therapy is warfarin embryopathy. Most warfarin-induced fetal abnormalities occur during weeks 6 to 12 of gestation, but central nervous system (CNS) and ocular abnormalities can develop at any time during gestation (Chap. 30).^158,354

ANTITHROMBIN AGONISTS

Heparin

Conventional heparin also known as unfractionated heparin C or unfractionated heparin (UFH) is a heterogeneous group of molecules within the class of glycosaminoglycans.¹⁸⁰ The heparin precursor molecule is composed of long chains of mucopolysaccharides, a polypeptide, and carbohydrates. The main carbohydrate components of heparin molecules include uronic acids and amino sugars in polysaccharide chains. Heparin for pharmaceutical use is extracted from bovine lung tissue and porcine intestines.³²⁸

Heparin inhibits thrombosis by accelerating the binding of AT to thrombin (activated factor II) and other serine proteases involved in coagulation.^232,308 Thus, factors IX to XII, kallikrein, and thrombin are inhibited. Heparin also affects plasminogen activator inhibitor, protein C inhibitor, and other components of coagulation. The therapeutic effect of heparin is usually measured through the activated PTT. The activated clotting time (ACT) is more useful for monitoring large therapeutic doses or in the overdose situation.²⁰¹

Low-molecular-weight heparins are 4,000- to 6,000-Da fractions obtained from conventional heparin.¹²⁸ As such, they share many of the pharmacologic and toxicologic properties of conventional heparin.⁴⁵ The various LMWHs (eg, nadroparin, enoxaparin, dalteparin) are prepared by different methods of depolymerization of heparin; consequently, they each differ to a certain extent regarding their pharmacokinetic properties and anticoagulant profiles. The major differences between LMWHs and conventional heparin are greater bioavailability, longer half-life, more predictable anticoagulation with fixed dosing, targeted activity against activated factor X, and less targeted activity against thrombin.^45,128 As a result of this targeted factor X activity, LMWHs have minimal effect on the activated PTT, thereby eliminating either the need for, or the usefulness of monitoring. They are therefore administered on a fixed dose schedule. However, in certain instances (eg, patients with CKD, pregnancy), monitoring of anti–factor Xa activity is performed to assess adequacy of anticoagulation and to prevent the risk of bleeding.¹⁵⁹ Controversy exists as to whether such testing is clinically necessary,⁴⁴ but studies have not shown significant benefit in determining risk of thromboembolic recurrence or consistent risk of major bleeding. We do not recommend the routine testing of anti–factor Xa activity.

Studies investigating LMWHs for the prevention of thromboembolic disease after hip surgery and trauma, in patients with stroke or DVT, in pregnancy, and in other conditions where anticoagulation with heparin would otherwise be indicated (eg, at the onset of oral anticoagulation therapy) are numerous. Low-molecular-weight heparins have a minimal risk in ­pregnancy²⁴⁶ because they do not cross the placenta,^124,356 and they are therefore preferred for the treatment or prophylaxis of thromboembolic disease in pregnancy.³⁰⁶ Most studies demonstrate a lower incidence of embolization; however, there is still a trend toward increased bleeding.^30,152,220 The PROTECT trial, which compared dalteparin and UFH in terms of rates of proximal leg DVT, PE, and major bleeding in critically ill patients, found that patients ­randomized to the dalteparin group had statistically fewer PE. This study found no statistically significant differences in proximal DVT and major bleeding rates.¹⁵³

Pharmacology

Because of the large size of heparin and negative charge, the molecule is unable to cross cellular membranes. These factors prevent oral administration, and heparin must be administered parenterally. After parenteral administration, heparin remains in the intravascular compartment, in part bound to globulins, fibrinogen, and low-density lipoproteins, resulting in a volume of distribution of 0.06 L/kg in humans.^118,273 Because of its rapid metabolism in the liver by a heparinase, heparin has a short duration of effect.²³² Although the half-life of elimination is dose dependent and ranges from 1 to 2.5 hours,^232,241,273 the duration of anticoagulant effect is usually reported as 1 to 3 hours.²³² Dosing errors or drug interactions with thrombolytics, antiplatelet drugs, or nonsteroidal antiinflammatory drugs increase the risk of bleeding.¹⁶¹ Low-molecular-weight heparins are nearly 90% bioavailable after SC administration and have an elimination half-life of 3 to 6 hours.¹³⁷ Anti–factor Xa activity peaks between 3 and 5 hours after dosing.¹³⁷ Low-molecular-weight heparins are renally eliminated and patients with stage 4 or 5 CKD are at increased risk of toxicity.³⁸⁶ Although there are insufficient data guiding therapeutic LMWH dosing in patients with severe CKD, some advocate dose reduction to decrease the risk of bleeding. However, no dosage regimens are provided.¹⁵⁴

Clinical Manifestations

Intentional overdoses with heparin are rare.²³⁸ Most reported cases involve iatrogenic toxicity in hospitalized patients.^{136,145,238,276,326} These cases involved the administration of large amounts of heparin as a consequence of misidentification of heparin vials, during the process of flushing IV lines, and secondary to IV pump malfunction. Significant bleeding complications occurred in several cases, including at least one fatality.¹³⁶ However, intentional overdoses of LMWHs are reported, although none of them were fatal.^58,254

Similar adverse effects to UFH are also reported with LMWHs and include epidural or spinal hematoma, intrahepatic bleeding,¹⁷³ abdominal wall hematomas,¹³ psoas hematoma after lumbar plexus block,¹⁹² and intracranial bleeding in patients with CNS malignancy.⁹⁸ Although these complications were all reported in patients who received the LMWH enoxaparin, there are no data to suggest differing toxicities among LMWHs.

Diagnostic Testing

For therapeutic anticoagulation, the effect of heparin is usually monitored with activated partial thromboplastin time (aPTT). Most hospitals have heparin nomograms that recommend heparin dose alterations based on aPTT, although these nomograms need to be individualized by each hospital because of the variation expected from the particular aPTT reagent and laboratory technology used. For patients undergoing cardiovascular procedures, the ACT is used to monitor them because they require higher dose heparin.¹³⁷ More recently, an anti-Xa assay specific for heparin was FDA approved. Heparin dosing is titrated based on the anti-Xa assay results, with therapeutic ranges higher than prophylactic ranges. Goal ranges also differ depending on whether LMWH or UFH is being used.

In patients with heparin resistance, in whom extremely high doses of heparin are required to accomplish a therapeutic aPTT, anti-Xa activity is recommended to guide heparin dosing. Despite lower heparin dosing and subtherapeutic aPTTs, patients monitored with anti-Xa activity had similar rates of recurrent thromboembolism and bleeding compared with the patients who were given high-dose heparin to maintain therapeutic aPTTs in a prospectively collected randomized control trial.²²¹

Low-molecular-weight heparins are usually administered at fixed doses for venous thromboembolism (VTE) prophylaxis or at weight-based doses for VTE treatment. Laboratory monitoring is not done unless patients are pregnant, have CKD, or are obese. In these cases, anti-Xa activity is measured, with target ranges varying based on agent and dosing regimen.¹³⁷ In patients without the aforementioned risk factors, monitoring and dose adjustment are of no benefit compared with fixed-dose regimens of LMWH.⁵ In addition, several studies show no correlation between anti-Xa activity and bleeding propensity.^18,215,389

Treatment

After stabilization of the airway and breathing and circulation is ensured, the physician should be prepared to replace blood loss and reverse the coagulopathy, if indicated. Because of the relatively short duration of action of heparin, observation alone is indicated if significant bleeding has not occurred. For a patient requiring anticoagulation, serial aPTT determinations will indicate when it is safe to resume therapy. If significant bleeding occurs, either removal of the heparin or reversal of its anticoagulant effect is indicated. Because heparin has a very small volume of distribution, it can be effectively removed by exchange transfusion.³²⁶ Although this technique has been used successfully in neonates, protamine has also safely been given to neonates without a history of fish allergy or previous exposure to protamine or protamine-containing insulin.²⁵³ Exchange transfusion to remove heparin should be performed in neonates with significant bleeding if protamine is contraindicated.

When severe bleeding occurs, protamine sulfate partially neutralizes UFH⁷ (Table 58–1). Protamine is a low-molecular-weight protein found in the sperm and testes of salmon, which forms ionic bonds with heparin and renders it devoid of anticoagulant activity.²³² One milligram of protamine sulfate injected intravenously neutralizes 100 units of UFH.²³² The dose of protamine should be calculated from the dose of heparin administered if known and assuming the approximate half-life of heparin to be 60 to 90 minutes; the amount of protamine should not exceed the amount of heparin expected to be found intravascularly at the time of infusion. As with other foreign proteins, protamine administration is associated with numerous adverse effects such as hypotension, bradycardia, and allergic reactions. Because approximately 0.2% of patients receiving protamine experience anaphylaxis, a complication that carries a 30% mortality rate, we recommend that protamine be reserved for patients with life-threatening bleeding (Antidotes in Depth: A19).¹⁷³ Clinicians should be aware that excess protamine administration does result in paradoxical anticoagulation, but this should not deter clinicians from administering protamine in life-threatening situations.

Because of the severe adverse effects associated with protamine, research has focused on safer methods to reverse heparin anticoagulation. These agents include heparinase,²⁴⁸ synthetic protamine variants,^387,388 and platelet factor 4 (PF4). These therapies are not widely available, and their efficacy and safety are not established.

If life-threatening bleeding occurs after LMWH administration we also recommend treating patients with protamine. Several studies show that protamine partially reverses LMWHs such as enoxaparin, dalteparin, and tinzaparin. In one case report of a 10-fold dosing error of enoxaparin, protamine effectively reversed the anticoagulant effects.⁴⁰⁴ In a series of 14 bleeding patients given protamine to reverse LMWH, bleeding ceased in two-thirds of the patients. These data, however, are complicated by repeat dosing of protamine as well as administration of other procoagulant antidotes such as clotting factors and vitamin K in a subset of patients. In addition, patients received protamine between 30 minutes and 48 hours after their last dose of LMWH.³⁷⁹ Approximately 1 mg protamine will neutralize 100 anti–factor Xa units, and 1 mg of enoxaparin will neutralize 100 anti–factor Xa units for up to 8 hours after LMWH administration.¹³⁷ A second dose of 0.5 mg protamine per 100 anti–factor Xa units is reasonable if bleeding continues. If more than 8 hours has elapsed, a smaller dose of protamine should be administered. The maximum dose of 50 mg should not be exceeded.¹³⁷ The appropriate dosages for protamine are described in detail in the Antidotes in Depth: A19. The newer experimental protamine variants appear to be effective against LMWHs but are not yet clinically available.^387,388 Interestingly, there is one case report of recombinant activated factor VII (rFVIIa) reversing the effects of LMWH in the setting of postoperative acute kidney injury (AKI),^96,249,264 and there is also a single case report demonstrating efficacy at reversing severe bleeding caused by enoxaparin.¹⁷⁵ Based on this limited reported experience, we do not recommend rFVIIA for LMWH-associated bleeding.

NONBLEEDING COMPLICATIONS

Postoperative thrombocytopenia that occurs in the first 1 or 2 days after surgery usually results from platelet consumption. This early fall in platelet count tends to cause concern for a drug-induced thrombocytopenia called HIT because postoperative VTE prophylaxis with heparin is usually started simultaneously.³⁹² However, postoperative thrombocytopenia usually improves by the third postoperative day, distinguishing itself from HIT, which typically occurs between days 5 and 10 after heparin initiation. In patients who were previously treated with heparin, HIT-related events sometimes occur within 24 hours after reexposure.²²³

Heparin-induced thrombocytopenia affects up to 5% of patients receiving heparin.^223,392 Heparin stimulates platelets to release PF4, which subsequently complexes with heparin to provoke an IgG response, causing platelet aggregation and thrombocytopenia.^14,416 A more severe form of thrombocytopenia, HITT (formerly known as HIT-2 or the white clot syndrome), occurs in up to 55% of patients with untreated HIT.²²³ The antibodies against the heparin–PF4 complex activate platelets, which leads to platelet–fibrin thrombotic events.^14,416

Patients present with either hemorrhagic or thromboembolic complications. Low-molecular-weight heparin is also associated with thrombocytopenia (isolated HIT) and less frequently with HITT.²²³ Consequently, when HITT occurs, LMWH is contraindicated.²²³ Treatment of patients with HIT includes discontinuation of heparin or LMWH and immediate use of alternative anticoagulant such as lepirudin, argatroban, or danaparoid.^68,223 In addition to HIT and HITT, necrotizing skin lesions²⁸⁶ and hyperkalemia from aldosterone suppression²⁷⁴ also rarely occur in patients receiving heparin therapy. These patients should not receive heparin or LMWH again, not even in low doses to maintain venous patency.

Some additional complications of heparin use include osteoporosis, which mostly occurs in patients on long-term therapy with UFH.¹⁷⁴ A small percentage of these patients develop bone fractures if treated continuously for more than 3 months. Data for LMWHs are limited, and the incidence of osteoporosis is less compared with UFH.¹⁷⁴ In 2008, an outbreak of adverse events was linked to heparin contaminated with oversulfated chondroitin sulfate.²²⁷ The contaminated heparin, which was found in at least 10 countries, originated in China.³⁸ Many patients developed anaphylactoid-type reactions with at least 100 reported deaths.²²⁷

DIRECT THROMBIN INHIBITORS

Hirudin and its congeners (lepirudin and desirudin) are used in patients with acute coronary syndromes, the prevention of thromboembolic disease, and in patients with HITT.^36,322,329 Desirudin is at least as effective as UFH and without an increased risk of bleeding or thrombocytopenia. However, in the Global Use of Strategies to Open Occluded Coronary Arteries (GUSTO) Iib study of patients with unstable angina or non–Q wave MI, there was an increase in the number of blood transfusions in patients who received desirudin compared with those who received heparin.³³⁸ This increased risk of bleeding compared with heparin has caused desirudin to fall out of favor. Instead bivalirudin and argatroban tend to be used more often, especially in the setting of HITT. In fact, initial studies using bivalirudin during coronary artery angioplasty for unstable or postinfarction angina showed that it is a safe substitute for heparin with lower bleeding rates.³⁶ Unfortunately, all of these DTIs are short acting and require parenteral administration.

Because of the potential therapeutic limitations of warfarin (eg, dosing, risk of bleeding, narrow therapeutic window), direct oral anticoagulants were developed with directed activity against specific clotting factors. The proposed benefits of these medications include the convenience of fixed dosing and avoidance of close therapeutic monitoring. Ximelagatran was one of the first DTIs that was as effective as warfarin in the treatment of stroke prevention, nonvalvular atrial fibrillation, and DVT.¹⁶⁸ Ximelagatran had many advantages over warfarin, including rapidity of onset, fixed dosing, stable absorption, decreased risk of drug interactions, and lack of necessity for therapeutic monitoring.¹⁶⁸ However, in 2006, drug manufacturers abandoned ximelagatran after noticing a high rate of hepatic failure. Countries that had approved ximelagatran withdrew the medication from the market.

Subsequently, dabigatran was approved for systemic anticoagulation in patients with nonvalvular atrial fibrillation in the United States and many other countries. Later it was approved for the treatment and prophylaxis of VTE. The Randomized Evaluation of Long Term Anticoagulant Therapy (RE-LY) trial demonstrated that dabigatran administration was associated with lower rates of systemic embolic events, with similar rates of bleeding, compared with anticoagulation with warfarin.⁸¹ However, subsequent evaluation of the data acquired from the RE-LY trial demonstrates that patients 75 years of age and older an increased risk of extracranial bleeding when compared with those taking warfarin.¹¹⁰ In multiple noninferiority trials, dabigatran demonstrated similar efficacy to enoxaparin in reducing VTE without increasing bleeding risk after hip joint replacement.^114,115

Pharmacology

Thrombin has four separate binding sites, each of which is specific for ­substrate, inhibitor, or cofactors.³⁶⁸ Whereas bivalent DTIs, such as hirudin and bivalirudin, bind the active site and one of two exosites (binding sites outside of the active site), univalent DTIs, such as dabigatran, bind just the active site⁹⁷ (Fig. 58–3). By directly inhibiting thrombin, anticoagulation is possible without the need for AT.³²⁹ In addition, inhibiting thrombin also inhibits platelet activation because thrombin is a potent and direct-acting platelet activator. Unlike heparin, DTIs are able to enter clots and inhibit clot-bound thrombin because of their small size, offering the distinct advantage of restricting further thrombus formation.

Desirudin, bivalirudin, and argatroban are all administered parenterally. Desirudin is given as SC injections for VTE prophylaxis. Although the dosing for desirudin is 15 mg as a subcutaneous injection every 12 hours, this dose should be decreased or stopped in patients with stage 3 or greater CKD. In addition, patients previously treated with the hirudins develop antibodies, creating the potential for anaphylaxis with repeat dosing.¹³⁷ Bivalirudin is also an analogue of hirudin. It is used in patients with HITT who require cardiac catheterization or cardiopulmonary bypass surgery.¹³⁷ Argatroban binds noncovalently to the active site of thrombin to function as a competitive inhibitor.¹³⁷ It is metabolized by CYP3A4/5 in the liver and is particularly useful in patients with CKD.¹³⁷

Dabigatran is orally administered as dabigatran etexilate. This prodrug has no anticoagulant properties, and serum esterase coverts it to dabigatran, the active drug.³⁴⁷ At therapeutic doses, peak concentrations occur in 2 hours (Table 58–3). Approximately 35% of dabigatran is protein bound; 85% is eliminated renally, with 78% of it is eliminated within the first 24 hours. Its mean terminal half-life is approximately 8 to 12 hours.³⁷ Contraindications to dabigatran are active hemorrhage or a hypersensitivity reaction to dabigatran. The manufacturer recommends either dose adjustment or avoidance of concomitant administration of P-glycoprotein inhibitors in patients with renal insufficiency. It also recommended to avoid coadministration of rifampin, a P-glycoprotein inducer.³⁹ P-glycoprotein activity prevents enteric absorption of dabigatran, and rifampin augments this activity, resulting in subtherapeutic dabigatran serum concentrations.

Clinical Manifestations

Intentional overdoses with the DTIs are rare events. Although there are reports of patients who develop significant coagulopathy after unintentional ingestion of excess dabigatran,^64,261,409 the more common scenario is that patients become overanticoagulated because of improper dosing in patients with CKD or failure to adjust dosing in patients who develop AKI. In general, the parenterally administered medications have short elimination half-lives, and iatrogenic medication administration errors, if recognized, are often less risky than those that are longer acting.

Dabigatran, the newest of the DTIs, is orally administered with a longer duration of action. Since its approval in 2010, bleeding was rapidly recognized as a complication of therapy. Only recently has a specific monoclonal antibody reversal xenobiotic, idarucizumab (Praxbind), been approved by the FDA. In 2011, the FDA released an advisory announcing that a review of postmarketing reports of serious bleeding associated with dabigatran use was initiated. In addition, dabigatran leads the list of fatalities reported to the MedWatch system. There are many published case reports of patients bleeding while anticoagulated with dabigatran. Patients may have been subjected to increased risk of bleeding because of risk factors such as increased age or CKD. In New Zealand and Australia, a significant number of bleeding events, including intracranial bleeding, GI bleeding, hematuria, and hemoptysis, were reported in the period immediately after the approval of the medication. Up to 25% of the reports in this series involved errors in prescribing practices.¹⁶⁰ In particular, off-label use of dabigatran for anticoagulation in the setting of mechanical heart valves led to valve thrombosis.^70,293 Cardiac tamponade from hemopericardium, fatal epistaxis, serious GI bleeding, postoperative bleeding complications, and intracranial bleeding are all reported.^{26,65,107,340,367} In addition, clinicians have expressed difficulties in treating hemorrhage in trauma victims while anticoagulated with dabigatran.⁸⁵

Furthermore, several questions regarding other adverse effects related to dabigatran use have been raised. A meta-analysis comparing patients anticoagulated with dabigatran versus warfarin, enoxaparin, or placebo showed higher rates of MI or acute coronary syndrome.³⁷² Additionally, though clinical trials implied that discontinuation of dabigatran does not cause rebound thrombosis, the authors of anecdotal reports of thrombotic events after cessation of dabigatran anticoagulation have questioned whether thrombosis is caused by rebound or just sequelae from underlying hypercoagulability or illness.³⁶⁴

Laboratory Assessment

Monitoring the anticoagulant effect in the setting of DTI use is complex. For bivalirudin and argatroban, serial aPTT measurements are commonly used to estimate the degree of anticoagulation.¹³⁷ Admittedly, the aPTT is not a good test because the degree of anticoagulation does not follow a linear relationship. This is particularly emphasized with dabigatran, the most recently developed DTI. For example, the aPTT increases at higher dabigatran concentrations, but the relationship is nonlinear, with the aPTT plateauing at dabigatran concentrations greater than 200 ng/mL.^346,378 Furthermore, the PT and INR are usually elevated, but they do not correlate with the anticoagulant effect of dabigatran.^349,378 Dilute thrombin time (dTT) and ecarin clotting time (ECT) are proposed as more accurate reflections of the anticoagulation effect.³⁷⁸ Several assays are used to measure serum dabigatran effect using the dTT assay. In these assays, standards with known dabigatran concentrations are used. Unfortunately, none of these modalities are FDA approved.^105,317,348 Ecarin clotting time is a laboratory assay that uses ecarin, a derivative of saw-scaled viper venom, as the reagent to activate prothrombin. Although some hospitals are able to obtain a thrombin time (TT) in a clinically relevant time, the dTT and the ECT are usually unavailable.

Treatment of Direct Thrombin Inhibitor–Induced Coagulopathy

In the event of an acute ingestion of dabigatran, AC is indicated based on data from an in vitro model.³⁷⁷ It should be noted that there have been at least two cases of intentional dabigatran overdose that were managed via orogastric lavage and AC. The initial dabigatran concentrations were later found to be 970 ng/mL and 4,170 ng/mL, respectively.^261,410 Based on a small prospective study with dabigatran assays, the average peak concentration in patients receiving therapeutic dosing of dabigatran was 112.7 ± 66.6 ng/mL.³¹⁷ However, because both patients after intentional dabigatran overdoses cases did not have bleeding complications, they were observed until the resolution of coagulopathy. One patient was given FFP for an aPTT of 79s (normal, 23–37 s)and an INR of 3.3. The other patient was noted to have a mild coagulopathy with an aPTT of 48.8 s and an INR of 1.3.^261,410 An overdose of argatroban was successfully treated with FFP in a case report.⁴¹⁵ With the widespread dabigatran use, the first of the novel oral anticoagulants, the absence of a reversal agent became a potentially deadly complication. The manufacturer fast-tracked the development of idarucizumab (Praxbind), a monoclonal antibody targeting dabigatran.^374,375 In murine and human studies, idarucizumab administration results in normalization of a battery of functional clotting assays, including clotting time, aPTT, and TT.³²¹ In a prospective study containing 90 patients who received 5 g of IV idarucizumab for urgent reversal of dabigatran-induced coagulopathy, reversal of coagulopathy occurred in 100% of patients despite 18 deaths in the study population. Study investigators noted that hemostasis in bleeding patients was restored at a median of 11.4 hours after administration of idarucizumab. In addition, antidotal therapy restored ECT and dTT in 88% to 98% of the patients. In patients undergoing an emergent procedure, normal intraoperative hemostasis was noted in 85% of cases. There are a significant number of limitations with this particular study. It should be noted that in addition to idarucizumab, 56% of patients received some type of blood products, including PRBCs, FFP, platelets, cryoprecipitate, PCCs, or whole blood. Thrombotic events were noted in five patients. One thrombotic event (simultaneous DVT and PE) occurred within 72 hours of idarucizumab administration in a single patient. Other complications noted were left atrial thrombus at 9 days, DVT at 7 days, MI at 13 days, and ischemic stroke at 26 days. Unfortunately, there was no ­control group to determine if morbidity or mortality outcomes were changed by administration of idarucizumab. In addition, dabigatran concentrations in enrolled patients ranged between 5 and 3,600 ng/mL with a median of 132 and 114 ng/mL in the two study groups.²⁸⁷ Although the majority of patients in the RE-LY trial had dabigatran concentrations less than 1,000 ng/mL, the literature reports exceedingly high dabigatran concentrations in many patients, particularly in patients who have kidney injury or in patients who have intentionally overdosed on dabigatran. The standard dose of idarucizumab, 5 g, will be insufficient in reversing coagulopathy in patients with these extremely high dabigatran concentrations.

There are many case reports describing the use of idarucizumab in various clinical scenarios. It has been used in life-threatening bleeding, intentional overdose, and therapeutic accumulation secondary to kidney injury and in the preoperative or preprocedural setting.^{47,140,165,166,235,282,303,307,319,331,351,363} At least one case report showed continuation of life-threatening GI bleeding despite receiving recommended dosing of idarucizumab.⁴ One report described using idarucizumab in conjunction with hemodialysis to achieve normal coagulation function.²³⁵ It should be noted that several case reports demonstrated a rebound drug effect after administration of idarucizumab. One case described a patient with a massive dabigatran overdose with an initial dabigatran concentration of 3,337 ng/mL. After receiving idarucizumab, the patient’s dabigatran concentration fell to 513 ng/mL, with improvement in his coagulopathy. However, 7 hours after receiving idarucizumab, the patient’s²⁹⁴ dabigatran concentration rebounded to 1,126 ng/mL with corresponding increases in INR, PT, and aPTT.³⁰⁹ Dabigatran rebound after idarucizumab treatment is noted in other case reports.¹⁶⁹ Some advocate for idarucizumab administration before thrombolytic therapy for acute ischemic cerebrovascular accidents in patients anticoagulated with dabigatran. In these case reports, patients received 5 g of idarucizumab followed by t-PA without adverse consequence.^{31,139,187,263,320,327} Some clinicians routinely recommend idarucizumab treatment before thrombolytic therapy for acute ischemic stroke.^100,139 However, given the paucity of data and lack of a formal clinical trial, this should not be the standard of care as case reports cannot determine the efficacy and safety of this practice.

Before the FDA approval of idarucizumab in 2015, the manufacturer suggested several strategies to treat patients with significant bleeding while anticoagulated with dabigatran. Transfusion of RBCs and FFP in addition to supportive care were the mainstays of treatment (Table 58–1). Recombinant factor FVIIa, PCCs, and hemodialysis were also used, but these interventions are incompletely studied. In a murine study of dabigatran, collagenase-induced intracranial hematoma expansion was inhibited in a dose-dependent fashion by PCC with the best results at 100 units/kg, the equivalent of 200% factor replacement. In the same study, high-dose PCC decreased but did not normalize the tail vein bleeding time. Recombinant factor VIIa was ineffective in limiting intracranial hematoma volume, and FFP limited intracranial hematoma size in the mice given a lower dose of dabigatran.⁴¹⁸

However, conclusions based on data obtained from murine studies should be interpreted with caution when treating humans. First, the hemostatic therapies used in these murine experiments are human derived, and cross-species effects cannot be predicted. Second, these mice had prolongation of their tail vein bleeding times, but dabigatran at therapeutic doses does not change bleeding time in humans.¹⁸⁶ Another study done in healthy humans found that aPTT, endogenous thrombin potential lag time, TT, and ECT did not decrease with the administration of four-factor PCC.¹⁰⁸

One ex-vivo study showed that after a single dose of dabigatran in healthy volunteers, FEIBA decreased endogenous thrombin potential and lag time, and thrombin generation increased in a dose dependent fashion.²³⁷ In a murine study, aPCC reduced bleeding time at low doses. Unfortunately, at high doses, this effect was reduced.³⁷⁶

Hemodialysis for dabigatran removal remains a controversial intervention. A single study showed that after a subtherapeutic dose of dabigatran in patients with stage 5 CKD, the extraction ratios were up to 68% at 4 hours after the initiation of hemodialysis.³⁵⁰ Although this suggests that dabigatran is effectively removed from the serum, the greater implication of this finding is unknown. This study did not address the safety of hemodialysis in bleeding patients who are actively bleeding or the possibility of rebound concentrations after hemodialysis.^349,350 Several published case reports demonstrated significant drug rebound after hemodialysis in bleeding patients, up to 87% of the initial dabigatran concentration within 2 hours after the cessation of hemodialysis.^63,64,340 In one case, hemodialysis was not effective in stopping bleeding, and despite massive transfusion of RBCs, FFP, and platelets, deaths from exsanguination occurred. In some cases, a repeat hemodialysis session or continuous venovenous hemodiafiltration (CVVHD) was performed in anticipation of posthemodialysis rebound. Although they were successful in decreasing serum dabigatran concentrations, they did not normalize the aPTT or the TT.³⁴⁰

Although the safety and efficacy of these adjunctive therapies are unknown, they are reasonable in patients with serious bleeding if idarucizumab is not available or if a patient continues to have life-threatening hemorrhage despite idarucizumab administration. Repletion of blood volume and coagulation factors with PCC is reasonable. If maximal efforts at repletion of blood volume and coagulation factors are ineffective, then hemodialysis followed by CVVHD is reasonable if the patient can tolerate the procedure hemodynamically.

Ciraparantag is a “universal antidote” being developed to reverse the activity of a number of anticoagulants, including the oral DTIs, the factor Xa inhibitors, and the heparins.^9-11 Clinical trial results have yet to be published.

FACTOR Xa INHIBITORS

Rivaroxaban was the first orally active direct factor Xa inhibitor approved for VTE and stroke prophylaxis and treatment. Development of this class of drugs started in the 1980s after the discovery of antistasin, a naturally occurring factor Xa inhibitor found in leeches. After screening a library of more than 200,000 compounds, a pharmaceutical company was able to identify structures that inhibit factor Xa. After structure optimization, rivaroxaban was created for clinical trials.²⁸³

Numerous clinical trials investigated the efficacy and safety of rivaroxaban.^{22,113,114,188,203,369} Compared with warfarin, rivaroxaban prevented more strokes or systemic thromboembolic events and significantly reduced the number of intracranial hemorrhages in patients with nonvalvular atrial fibrillation.²⁸¹ The most recent published trial compared the use of rivaroxaban compared with placebo in patients with acute coronary syndrome. There was a significant risk reduction in death attributed to any cardiac cause or stroke. However, rivaroxaban-treated patients had statistically significant increased rates of major and intracranial hemorrhages.²⁴³

Apixaban is another oral factor Xa inhibitor that is approved by the FDA for VTE prophylaxis in patients with atrial fibrillation. Numerous studies showed benefit of apixaban in preventing VTE in specific populations while simultaneously lowering mortality or specific types of bleeding, such as intracranial hemorrhage.^{80,147,150,203-206}

In trials investigating the use of apixaban in patients with recent acute coronary syndrome, excess bleeding caused early cessation of trials; however, most patients were concurrently treated with dual antiplatelet therapy.^3,76

Edoxaban is the newest of the approved factor Xa inhibitors based on studies for VTE prophylaxis in patients with atrial fibrillation and for the treatment of VTE disease.^{66,71,111,143,144,171,296,396,414} Studies with edoxaban demonstrated promising results compared with enoxaparin in post–orthopedic surgery VTE prophylaxis, although it is not currently FDA approved for this indication.^129-132,297 The risk of hemorrhage compared with enoxaparin varies among the many studies.^{52,66,71,129-132,414} Compared with warfarin in patients with nonvavular atrial fibrillation, the ENGAGE AF-TIMI 48 trial demonstrated a lower risk of cardiovascular death and hemorrhage, including intracranial hemorrhages and fatal bleeding.^{111,143,144,396} However, a study demonstrated the addition of an antiplatelet agent, such as aspirin, to edoxaban therapy in patients with atrial fibrillation resulted in a higher risk of hemorrhage compared with patients who received warfarin in conjunction with a single antiplatelet agent.⁴¹²

Factor Xa inhibitor development continues. Betrixaban is undergoing clinical trials to determine its efficacy and safety in VTE prophylaxis.⁷⁹

Pharmacology

The factor Xa inhibitors are ideal anticoagulants because their site of action is the intersection of the intrinsic and extrinsic pathways, preventing thrombin activation.⁵⁹ These synthetic drugs reversibly inhibit factor Xa without any cofactor requirements. Theoretically, this class of anticoagulants is safer than the DTIs because they do not completely neutralize thrombin.

Rivaroxaban and apixaban selectively bind free and clot-bound factor Xa without inhibiting related serine proteases, including thrombin, trypsin, plasmin, or other activated clotting factors.¹¹⁶ In addition, rivaroxaban and apixaban inhibit tissue factor or collagen-induced thrombin formation. In vitro, the factor Xa inhibitors hinder tissue factor–induced platelet aggregation.^116,408 Rivaroxaban does not directly inhibit platelet aggregation, and concomitant aspirin use does not affect the pharmacokinetics or safety of rivaroxaban in studies of healthy humans.^59,116,196

Rivaroxaban has an oral bioavailability of approximately 80%.³⁹⁷ Inhibition of factor Xa peaks approximately 3 hours after administration of the rivaroxaban. This inhibition lasts for approximately 12 hours.¹⁹⁸ However, kidney or liver disease lengthens its duration of action (Table 58–3). Dosing varies by indication and kidney function. The manufacturer provides recommended dose adjustments for each indication based on kidney function.

Approximately one-third of rivaroxaban is eliminated unchanged by the kidneys, and one-third is metabolized to an inactive form and excreted by the kidney. The remaining one-third is metabolized by the CYP3A4-dependent and -independent pathways in the liver and then excreted fecally. Concomitant use of CYP3A4 inhibitors or P-glycoprotein inhibitors is contraindicated because of increased risk of drug accumulation of up to 160%.^164,397

In the future, rivaroxaban could be a suitable anticoagulant alternative in patients with HITT because in vitro studies do not show platelet aggregation or activation in the presence of HITT antibodies or release of PF4 from platelets.³⁹⁰ Future studies proving efficacy and safety are needed before these xenobiotics can be recommended for use in patients with HITT.

In preclinical studies, the oral bioavailability of apixaban is approximately 66%. Apixaban is primarily distributed to the blood compartment and is approximately 87% protein bound.¹¹⁶ Peak serum concentrations occur between 1 and 3 hours postingestion. There are multiple elimination pathways, suggesting that patients with either renal or hepatic impairment should be able to tolerate apixaban well. Approximately 25% of apixaban is excreted in the urine, and the majority of apixaban is excreted fecally between 24 to 48 hours after ingestion.^116,295 Simultaneous administration of strong CYP3A4 inhibitors is contraindicated with apixaban.

Apixaban is approved in the United States for stroke prevention in patients with nonvalvular atrial fibrillation, for VTE prophylaxis, and for VTE treatment.^80,150 Dosing is based on indication. The manufacturer recommends either decreasing the dose or avoiding coadministration of apixaban if strong inhibitors of CYP3A4 and P-glycoprotein are administered.⁵³

Edoxaban is approved for the treatment of nonvalvular atrial fibrillation and for the treatment of DVT and PE. The dosage varies by indication. For patients with atrial fibrillation, the dosage also depends on creatinine clearance. Edoxaban is not recommended for patients with creatinine clearance less than 95 mL/min. Patients with VTE disease should receive edoxaban after 5 to 10 days of treatment with a parenteral anticoagulant.^89,310 The manufacturer also notes that coadministration of rifampin, due to CYP3A4 induction, and coadministration of other anticoagulants are contraindicated.^89,247 The prescribing information provided by the manufacturers of edoxaban also recommend a maximum dose of 30 mg/day for patients with body weight less than 60 kg.⁸⁹

Edoxaban is orally administered and has a bioavailability of 61.8%.^240,250 P-glycoprotein appears to be the key factor as a transporter of edoxaban. Edoxaban metabolism is minor, and the compound is largely excreted in the urine and feces.²⁵⁰ Approximately 35% of an oral dose is eliminated renally, and 49% of the oral dose is excreted in the feces.²⁴⁰ Peak edoxaban concentrations occur 1 to 2 hours after ingestion, and the terminal half-life is 10 to 14 hours.⁸⁹

Clinical Manifestations

Hemorrhage is the most concerning consequence of the factor Xa inhibitors, even at therapeutic doses. However, hemorrhage does not always occur in overdose. There are reports of no hemorrhage despite acute overdoses of both rivaroxaban and apixaban. In a multi–poison control center study involving 223 rivaroxaban or apixaban exposures, 12 patients involved suicide attempts, and none of these patients developed hemorrhage, although confirmation of ingestion was not verified in all cases.³⁴⁵ A case report describes a 21-month-old girl who developed an INR of 3.5 approximately 12 hours after ingesting an unknown quantity of rivaroxaban. She did not have any hemorrhagic consequences.²⁴⁵ Another case report describes a 71-year-old man who intentionally ingested 1,940 mg of rivaroxaban, resulting in an INR of 7.2 drawn 2 hours after ingestion and rivaroxaban concentration of 160 ng/mL drawn 3 days after ingestion. He never developed bleeding, and he was monitored until his coagulopathy resolved. He did not receive any blood products or reversal agents.³⁰⁰ In another case, a 42-year-old man intentionally ingested multiple pharmaceuticals, including 1,400 mg of rivaroxaban. He never developed any bleeding, but given his INR of 2.4 drawn 5 hours after ingestion, he was given 1 g of tranexamic acid and 3,000 units of PCC. He was discharged 1 day after this ingestion without any bleeding or adverse events.²²⁴ There are many other cases reporting massive intentional factor Xa inhibitor overdose, resulting in abnormal coagulation studies or supratherapeutic concentrations.^20,212,213 More predictably, bleeding occurs at therapeutic doses. In the previously mentioned multi–poison control center study, all bleeding events occurred in patients on long-term anticoagulant therapy.³⁴⁵ Another case report describes a 58-year-old man taking 10 mg of rivaroxaban per day who developed prolonged rectal bleeding 31 days after initiating treatment for prophylaxis after hip replacement.⁴⁰