SECTION II: PATHOPHYSIOLOGIC BASIS: ORGAN SYSTEMS

CASE STUDY 3

ELECTROPHYSIOLOGIC PRINCIPLES

The clinical tool most commonly used to assess cardiac function is the surface electrocardiogram (ECG). The ECG records the sum of the electrical changes occurring within the myocardium. The electrophysiologic basis of cardiac function and the ECG are complex and are subject to alteration by numerous xenobiotics. Ion currents flowing through various ion channels are responsible for cardiac function. Electrophysiologic studies identify the functional types of membrane receptors and ion channels. Molecular genetic studies identify the gene coding for the key cardiac ion channels and elucidate the structural and physiologic relationships that lead to the toxic effects of many xenobiotics. These channels are critical for maintenance of the intracellular ion concentrations necessary for action potential development, impulse conduction throughout the heart, and myocyte contraction. This chapter will first review the individual ion channels and their currents, and then summarize their contribution and effects on the ECG.

ION CHANNELS OF THE MYOCARDIAL CELL MEMBRANE

Sodium Channels

The voltage-sensitive sodium channels are responsible for the initiation of depolarization of the myocardial membrane. All currently identified voltage-sensitive channels, including the sodium and calcium channels, have structures similar to the potassium channel assembly. The sodium channel gene encodes a single protein that contains 4 functional domains (D I to D IV). Each of these domains has the 6 membrane-spanning regions characteristic of the voltage-gated potassium channel and is structurally similar to the α subunit of the potassium channel. The single, large α subunit of the sodium channel assembles with regulatory β subunits to form the functional unit of the sodium channel (Fig. 15–1A). The best characterized of the sodium channels, the SCN5A gene–encoded α channel, is inactivated by xenobiotic interactions between the D III and the D IV domains to physically block the inner mouth of the sodium channel pore.²⁹ Seven specific receptor sites are currently identified on the α subunit, with different xenobiotics binding at the specific sites: tetrodotoxin, saxitoxin, μ-conotoxin (site 1); aconitine, batrachotoxin, grayanotoxin, veratridine (site 2); α scorpion toxins, sea anemone toxins (site 3); β scorpion toxins, Chinese bird spider toxin (site 4); brevetoxins, ciguatoxins (site 5); δ-conotoxins (site 6); pyrethroids (site 7), and local anesthetics and related antidysrhythmics and antiepileptics binding at overlapping receptor sites.^14,57,61,71

Potassium Channels

Ion channels that change their conductance of current with changes in the transmembrane voltage potential are called rectifying channels. The voltage-sensitive potassium channels are categorized based on their speed of activation and their voltage response. They include the “delayed rectifier” potassium currents, particularly the I_Kr (rapidly activating) and the I_Ks (slowly activating) channels.⁴³ Potassium channels form large multimolecular complexes with regulatory kinases, phosphatases, anchoring proteins, accessory subunits, and with each other.

The various voltage-gated potassium channels share an underlying structural similarity. The α subunit is a protein molecule with 6 membrane-spanning α-helical regions, 1 to 6 (Fig. 15–1B). The pore domain is located between the regions five and six of the α subunit. Region 4 is the voltage sensor region.^53,64 Apamin, found in apitoxin obtained from bee venom, is a potent blocker of a specific subset of Ca²⁺-activated K⁺ channels conferring atrial antidysrhythmic effects. The scorpion toxins scyllatoxin and tamapin block similar K⁺ channels although they are structurally dissimilar.⁵⁷ There are more than 70 genes encoding K⁺ channels in the human genome.³ Human Ether- à-go-go Related Gene (hERG) encodes the α subunit that assembles with β subunit proteins to form the I_Kr potassium channel. The C-terminus region of the α subunit encoded by hERG has a cyclic nucleotide binding domain and an N-terminus region similar to domains involved in signal transduction in cells.⁶⁴

Many xenobiotics interact with the hERG-encoded subunit of the potassium channel to reduce the current through the I_Kr channel and prolong the action potential duration. The hERG α subunit of the channel is particularly susceptible to xenobiotic-induced interactions because of 2 important differences from the other channels. First, region 6 of the hERG channel has aromatic side chains on the inner cavity pore that can bind to aromatic xenobiotics, for example, cisapride, haloperidol, and terfenadine.⁵⁵ Additionally, the inner cavity and entrance of the hERG channel is larger than the other voltage-gated potassium channels. This larger pore can accommodate larger xenobiotics that get trapped within the pore when the channel closes.^52,53,64

Calcium Channels

Calcium channel conductivity across the myocardial cell membrane is critical for the maintenance of the appropriate duration of cell membrane depolarization and for initiation of cellular contraction. The best characterized of the calcium channels are the slow (L-type), the fast (T-type), and the ryanodine receptor calcium channel. They are more prominently involved in cardiac contractility and discussed in Chapter 16.

ION CHANNELS AND MYOCARDIAL CELL ACTION POTENTIAL

An understanding of the basic electrophysiology of the myocardial cell is essential to understand the toxicity of xenobiotics and to plan appropriate therapy. Figure 15–2A shows the typical action potential of myocardial cell depolarization, the electrolyte fluxes responsible for the action potential, and the resulting ECG complex (Fig. 15–2B). The action potentials of the contractile and the conductive cells are also shown.

The action potential is divided into 5 phases: phase 0, depolarization; phase 1, overshoot; phase 2, plateau; phase 3, repolarization; and phase 4, resting. Phase 0 begins when the cell is excited either by a stimulus from an adjoining cell or by spontaneous depolarization of a pacemaker cell. The stimulus causes selective voltage-gated fast sodium channels (I_Na+) to open, resulting in rapid depolarization of the membrane. At the end of phase 0, the voltage-sensitive sodium channels close and a transient outward potassium current (I_To) occurs, resulting in a partial repolarization of the membrane—this constitutes phase 1.

During phase 2 (plateau phase), the inward depolarizing calcium currents are largely balanced by the outward repolarizing potassium currents. Voltage-sensitive calcium channels that open allow Ca²⁺ movement down the 5,000–10,000-fold concentration gradient into the cell. These channels are categorized based on their conductance (fast or slow) and their sensitivity to voltage changes.⁵⁷ The calcium currents (mostly the “long-lasting” current) gradually decrease as the channels inactivate. Simultaneously, the outward potassium “delayed rectifier” currents, particularly the I_Ks (slowly activating) current increases, terminating the plateau phase of the action potential and initiating cellular repolarization (phase 3). Other, smaller, outward potassium currents (not shown in Fig. 15–2A) play a lesser role in the duration of the action potential and development of phase 3, including I_Kur (ultrarapid), I_Kr (rapidly activating), I_K-Ach (acetylcholine-dependent), and I_K-ATP (adenosine triphosphate–dependent) currents.

In phase 3, the rapid repolarization phase, the cell membrane repolarizes as a result of the voltage-gated currents, the delayed rectifier currents (I_Kr I_Ks), and to a lesser extent the other K⁺ currents, the Na⁺,Ca²⁺ exchanger current, and the sodium/potassium pump (active transport via Na⁺,K⁺-ATPase.

Phase 4 is the resting state for much of the myocardium, except the pacemaker cells, and it corresponds to diastole in the cardiac cycle. During phases 3 and 4, active transport of Na⁺, K⁺, and Ca²⁺ against their electrochemical gradients returns the myocyte to its baseline resting state. The transmembrane electrochemical gradient is maintained during the resting state by a Ca²⁺,Na⁺ exchange mechanism and by ATP-dependent pumps in the membrane that together move Ca²⁺ out of the cells.⁵⁷

In the pacemaker cells during phase 4 of the action potential, gradual electrical depolarization of the membrane occurs due to potassium currents called the I_kf for “funny” or the I_Kh for “hyperpolarization-activated” current.^6,16 The membrane potential gradually increases in these pacemaker cells until the threshold potential is reached, the fast inward sodium channels open, and the I_Na current initiates the next phase 0. This electrical impulse is then propagated via the His-Purkinje conducting system of the heart.

In addition to its role in myocardial contractility, Ca²⁺ influx is also important in pacemaker function. Although spontaneous pacemaker cell depolarization has traditionally been ascribed to inward cation current through “pacemaker channels,”¹⁶ more recent research suggests that it is actually driven by rhythmic release of calcium from the sarcoplasmic reticulum.^36,37 Regardless, Ca²⁺ flux plays an important role in the spontaneous depolarization (phase 4) of the action potential in the sinoatrial (SA) node.³⁸ The rate of pacemaker cell depolarization is enhanced by β-adrenergic stimulation through phosphorylation of proteins on the sarcoplasmic reticulum and by a phosphorylation-independent action of cAMP at the pacemaker channels.⁴² Depolarization of cells in the SA node spreads to surrounding atrial cells where it triggers the opening of fast sodium channels and causes impulse propagation. Calcium flux also allows normal propagation of electrical impulses via the specialized myocardial conduction tissues in the atrioventricular (AV) node.

During phases 0 to 2, the cell cannot be depolarized again with another stimulus; the cell is absolutely refractory. During the latter half of phase 3, as the calcium channels recover from their inactivated to their resting state, an electrical stimulus of sufficient magnitude will cause another depolarization; the cell is said to be relatively refractory. During phase 4, the cell is no longer refractory and any appropriate stimulus that reaches the threshold level will cause depolarization.

ELECTROCARDIOGRAPHY

The ECG measures the sum of all electrical activity in the heart. It is used extensively in medicine and its interpretation is widely understood by physicians of nearly all disciplines. It is an invaluable diagnostic tool for patients with cardiovascular complaints. However, it is also a valuable source of information in poisoned patients and has the potential to enhance and direct their care. Although it seems obvious that an ECG would be required following exposure to a medication used for cardiovascular indications, many medications with no overt cardiovascular effects at therapeutic doses are cardiotoxic in overdose, for example loperamide. Furthermore, in patients with unknown exposures, the ECG can suggest specific xenobiotics or demonstrate electrolyte abnormalities, long before blood is drawn. For example, oropharyngeal or dermal burns in a patient whose ECG has evidence of hyperkalemia or hypocalcemia suggests exposure to hydrofluoric acid (Chap. 104).⁶⁹ Alternatively, a patient manifesting signs of the opioid toxidrome with runs of torsade de pointes might have ingested methadone (Chap. 36).²¹ QT interval prolongation is a clue to the etiology of an overdose with an atypical antipsychotic such as quetiapine (Chap. 67). The ECG is used to predict complications of poisoning, such as seizures following a cyclic antidepressant overdose (Chap. 68). Therefore, an ECG should be examined early in the initial evaluation of most poisoned patients.

History of the ECG

In the 1900s, Willem Einthoven graphically displayed the electrical activity of the heart and named the different waves—P, QRS, and T. He called this tracing an “elektrokardiogramme,” and was awarded a Nobel Prize in 1924 for his efforts. The acronym EKG, still employed by some authors, was derived from Einthoven’s spelling. The acronym ECG, which is consistent with our current spelling of electrocardiogram, is used throughout this text.

Since this initial description, both the normal electrophysiology of the heart and the pharmacologic effects of various xenobiotics on the ECG have been described. Despite the large number, diversity, and complexity of the various cardiac toxins, there are a limited number of electrocardiographic manifestations.

Basic Electrophysiology of an ECG

Simplistically, a positive or upward deflection on the oscilloscope is generated when an electrical force moves toward an electrical sensor or electrode, and a downward deflection occurs if the force moves away. An ECG represents the sum of movement of all electrical forces in the heart in relation to the surface electrode, and the height above baseline represents the magnitude of the force (Fig. 15–2B). Only during depolarization or repolarization does the ECG tracing leave the isoelectric baseline, because it is only during these periods that measurable net currents are flowing in the heart. During the other periods, mechanical effects are occurring in the myocardium, but large amounts of current are not flowing.

Leads

Although the reading from a single ECG lead can provide valuable information, to visualize the heart in a nearly 3-dimensional perspective, multiple leads must be assessed simultaneously. Given the cylindrical nature of both the heart and thorax, at any given moment some of these leads will record positive voltage and others negative. The lead placement that was described and refined by Einthoven forms the basis for the bipolar or limb leads, described as I, II, and III (Fig. 15–3). The Einthoven triangle is an equilateral triangle formed by the sum of these leads. Unipolar limb leads and precordial leads were subsequently added to the standard ECG. Unipolar leads were created when the limb leads were connected to a common point where the sum of the potentials from leads I, II, and III was zero. The currently used unipolar augmented (a) leads (aV_R, aV_L, and aV_F) are based on these unipolar leads (Fig. 15–4). The voltage potential of these unipolar, “augmented leads” is small; thus, it is amplified by incorporating the voltage change of the other 2 augmented leads. Together, leads I, II, III, aV_R, aV_L, and aV_F form the hexaxial reference system that is used to calculate the electrical axis of the heart in the frontal plane. The precordial leads, called V₁ through V₆, are also unipolar measurements of the change in electric potential measured from a central point to the sixth anterior and left lateral chest positions (Fig. 15–5). If V₂ is placed over the right ventricle, part of the initial positive ventricular deflection (QRS complex) reflects right ventricular activation, with electrical forces moving toward the electrode. The majority of the subsequent terminal negative deflection reflects activation of other muscle tissue (septum, left ventricular wall) when the electrical forces are moving away from the electrode. Recordings from each of these 12 leads (I, II, III, aV_R, aV_L, aV_F, V_1–6) evaluate the heart from 2 different planes in 12 different positions, yielding a 3-dimensional electrical “picture” of the heart, with respect aV_R, aV_L, aV_F, to time and voltage.

A continuous cardiac monitor often relies on recordings from one of 2 bipolar leads: a modified left chest lead (MCL₁) or a lead II. The recording from an MCL₁, in which the positive electrode is in the V₁ position, is similar in appearance to a V₁ recording on a 12-lead ECG. This lead visualizes ventricular activity well; however, lead II shows atrial activity (ie, the P wave) much more clearly. Right ventricular precordial leads (V₁, V_3-6R) are sensitive and specific for determining the presence of right ventricular myocardial infarctions, although specific applications to poisoning are not reported.

Various Intervals and Waves

The ECG tracing has specific nomenclature to define the characteristic patterns. Waves refer to positive or negative deflections from baseline, such as the P, T, or U waves. A segment is defined as the distance between 2 waves, such as the ST segment, and an interval measures the duration of a wave plus a segment, such as QT or PR interval. Complexes are a group of waves without intervals or segments between them (QRS complex). Electrophysiologically, the P wave and PR interval on the ECG tracing represent the depolarization of the atria. The QRS complex represents the depolarization of the ventricles. The plateau is depicted by the ST segment and repolarization is visualized as the T wave and the QT interval (QT). The U wave, when present, generally represents an afterdepolarization (Fig. 15–6).

The P Wave

The P wave is the initial deflection on the ECG that occurs with the initiation of each new cardiac cycle.

Electrophysiology

The early, middle, and late portions of the P wave are represented sequentially by the electrical potential initiated by the sinus node. The impulse is propagated directly through the right atrial muscle, producing contraction. The impulse is also propagated by specialized conduction tissue across the interatrial septum, to produce contraction of the left atrium. Additionally, internodal pathways rapidly conduct the impulse to the AV node. The electrical excitation of the sinus node differs from that of the ventricular myocardium in that current is mediated primarily by Ca²⁺ influx via slow T-type calcium channels, following hyperpolarization by the I_kf, mixed sodium-potassium current (f for “funny”). The “funny” current is also called the I_Kh (h for “hyperpolarization-activated”). The vagus nerve and b-adrenergic stimulation also affect conduction across nodal tissues.⁶

Abnormal P Wave

Clinically, abnormalities of the P wave occur with xenobiotics that depress automaticity of the sinus node, causing sinus arrest and nodal or ventricular escape rhythms (β-adrenergic antagonists, calcium channel blockers). The P wave is absent in rhythms with sinus arrest, such as occurs with xenobiotics that increase vagal tone such as cardioactive steroids and cholinergics. A notched P wave suggests delayed conduction across the atrial septum and is characteristic of quinidine poisoning or atrial enlargement (Fig. 15–7). P waves decrease in amplitude as hyperkalemia becomes more severe until they become indistinguishable from the baseline (Chap. 12).

PR Interval

The PR interval is measured from the beginning of the P wave to the beginning of the QRS complex (normal is 120–200 milliseconds {ms}).

Electrophysiology

Despite rapid conduction by specialized conduction tissue from the SA to the AV node, the AV node delays transmission of the impulse into the ventricles, ostensibly to allow for complete atrial emptying. Thus, the PR interval represents the interval between the onset of atrial depolarization and the onset of ventricular depolarization. Children usually have more rapid conduction and a shortened PR interval, and older adults generally have a prolonged PR interval. The segment between the end of the P wave and the beginning of the QRS complex (the PQ segment) reflects atrial contraction and is usually isoelectric. Atrial repolarization coincides with the Q wave, but the ECG evidence, or atrial T waves, are obscured by the QRS complex.

Abnormal PR Interval

Xenobiotics that decrease interatrial or AV nodal conduction cause marked prolongation of the PR segment until such conduction completely ceases. At this point, the P wave no longer relates to the QRS complex; this is AV dissociation, or complete heart block. Some xenobiotics suppress AV nodal conduction by blocking calcium channels in nodal cells, as does magnesium, β-adrenergic antagonists, or muscarinics through enhanced vagal tone. Although the therapeutic concentrations of digoxin, as well as early cardioactive steroid poisoning, cause PR prolongation through vagal tone, direct electrophysiologic effects account for the bradycardia of poisoning (Chap. 62 and Antidotes in Depth: A22).

QRS Complex

The QRS complex is the second and typically largest deflection on the ECG. The normal QRS duration in adults varies between 60 and 100 ms. A QRS duration that is prolonged from the patient’s baseline in the setting of poisoning indicates toxicity even if the QRS is within the normal range. The normal QRS complex axis in the frontal plane lies between –30° and 90°, although most individuals have an axis between 30° and 75°. This axis will vary with the weight and age of the patient. Alterations in myocardial function will also alter the electrical axis of the heart.

Electrophysiology

The QRS complex reflects the electrical forces generated by ventricular depolarization mediated primarily by Na⁺ influx into the myocardial cells. Although under normal conditions both ventricles depolarize nearly simultaneously, the greater mass of the left ventricle causes it to contribute the majority of the electrical forces. The QRS complex is primarily positive in leads I and aV_L on the surface ECG recording because under normal conditions the depolarization vector is directed at 60° and is thus moving toward these leads.

The simultaneous and rapid depolarization of the ventricles results in a very short period of electrical activity recorded on the electrocardiogram. Mechanical systole lasts well past the end of the QRS complex and is maintained by continued depolarization during the plateau phase of the action potential. The return to, and maintenance of, the baseline, or isoelectric potential, simply represents that the entire heart is depolarized and there is no significant net flow of current during this period.

The axis of the terminal 40 ms (0.04 seconds) of the QRS complex represents the late stages of ventricular depolarization and generally follows the direction of the overall axis. This axis is determined by examining the last small box (0.04 seconds or 40 ms) of the QRS complex on the ECG. A leftward shift of the terminal 40 ms axis of the QRS complex correlates with toxicity from xenobiotics with Na⁺-channel–blocking effects (Chap. 68).

Abnormal QRS Complex

In the presence of a bundle-branch block, the 2 ventricles depolarize sequentially rather than concurrently. Although, conceptually, conduction through either the left or right bundle may be affected, many xenobiotics preferentially affect the right bundle (Fig. 15–8). The specific reasons for this effect are unclear, but it is likely related to differing refractory periods of the tissues. This effect typically results in the left ventricle depolarizing slightly more rapidly than the right. The consequence on the ECG is both a widening of the QRS complex and the appearance of the right ventricular electrical forces that were previously obscured by those of the left ventricle. These changes are typically the result of the effects of a xenobiotic that blocks fast Na⁺ channels. Implicated xenobiotics include amantadine,⁵⁶ buproprion,²² carbamazepine,³⁰ cocaine,¹ cyclic antidepressants,^11,12 diphenhydramine,⁵⁸ lamotrigine,⁴⁵ phenothiazines, quinidine and other type IA and IC antidysrhythmics,¹⁷ and topiramate. In the setting of cyclic antidepressant poisoning, a prolonged QRS duration has both prognostic and therapeutic value (Chap. 68).^8,12,25 In one prospective analysis of ECGs in cyclic antidepressant poisoned patients, the maximal limb lead QRS duration was prognostic of seizures (0% if <100 ms; 30% if greater) and ventricular dysrhythmias (0% if <160 ms; 50% if greater).¹²

The terminal 40-ms axis of the QRS complex also contains information regarding the likelihood, but not the extent, of poisoning by Na⁺ channel blockers. In a patient poisoned by a Na⁺ channel blocker, the terminal portion of the QRS has a rightward deviation greater than 120°. The common abnormalities include an R wave (positive deflection) in lead aV_R and an S wave (negative deflection) in leads I and aV_L.³³ The combination of a rightward axis shift in the terminal 40 ms of the QRS complex (Fig. 15–9) with a prolonged QT interval and a sinus tachycardia is highly specific and sensitive for cyclic antidepressant poisoning.⁴⁴ Another study suggests that although ECG changes, like a prolonged QRS duration, are better at predicting severe outcomes than the cyclic antidepressant concentration, neither is very accurate.⁸ One retrospective study suggests that an absolute height of the terminal portion of aV_R that is greater than 3 mm predicted seizures or dysrhythmias in cyclic antidepressant–poisoned patients.³² However, in infants younger than 6 months, a rightward deviation of the terminal 40-ms QRS axis is physiologic and not predictive of cyclic antidepressant toxicity.¹¹

An apparent increase in QRS complex duration and morphology, which is an elevation or distortion of the J point called a J wave or an Osborn wave (Fig. 29–3) is a common finding in patients with hypothermia.^47,65 Hypermagnesemia is also associated with prolongation of the QRS complex duration. A slight narrowing of the QRS complex may occur with hypomagnesemia. Significant hyperkalemia also causes prolongation and distortion of the QRS complex.

ST Segment

The ST segment is defined as the distance between the end of the QRS complex and the beginning of the T wave.

Electrophysiology

The ST segment reflects the period of time between depolarization and the start of repolarization, or the plateau phase of the action potential. During this period, no major currents flow within the myocardium, which explains why under normal circumstances the ST segment is isoelectric. Although both the degree of displacement from the baseline and the length of this segment are important, the ST segment duration is usually measured by its effects on the QT interval duration (see “QT Interval”).

Abnormal ST Segment

Displacement of the ST segment from its baseline typically characterizes myocardial ischemia or infarction (Fig. 15–10). The subsequent appearance of a Q wave is diagnostic of myocardial infarction. The ECG patterns of these entities reflect the different underlying electrophysiologic states of the heart. Ischemic regions are highly unstable and produce currents of injury because of inadequate repolarization, which is related to lack of energy substrate to power the Na⁺,K⁺-ATPase pump. Myocardial infarction represents the loss of electrical activity from the necrotic, inactive ventricular tissue, allowing the contralateral ventricular forces to be predominant on the ECG. Patients who are poisoned by xenobiotics that cause vasoconstriction, such as cocaine (Chap. 75), other α-adrenergic agonists, or the ergot alkaloids, are particularly prone to develop focal myocardial ischemia and infarction. The specific ECG manifestations help to identify the region of injury and are correlated with an arterial flow pattern: inferior (leads II, III, aV_F; right coronary artery); anterior (leads I, aV_L; left anterior descending artery); or lateral (leads aV_L, V_5–6; circumflex branch). However, any poisoning that results in profound hypotension or hypoxia also results in ECG changes of ischemia and injury. In these patients, the injury is more global, involving more than one arterial distribution. Diffuse myocardial damage may not be identifiable on the ECG because of global, symmetric electrical abnormalities. Under these circumstances, the diagnosis is established by other noninvasive testing, such as by echocardiogram or by finding elevated concentrations of serum markers for myocardial injury such as troponin.

Many young, healthy patients have ST segment abnormalities that mimic those of myocardial infarction. The most common normal variant is termed “early repolarization” or “J point elevation,” and is identified as diffusely elevated, upwardly concave ST segments, located in the precordial leads and typically with corresponding T waves of large amplitude (Fig. 15–11).¹³ The J point is located at the beginning of the ST segment just after the QRS complex. Because this ECG variant is common in patients with cocaine-associated chest pain (Chap. 75),²⁴ its recognition is critical to instituting appropriate therapy.

The Brugada electrocardiographic pattern (Fig. 15–12) is characterized by terminal positivity of the QRS complex and ST segment elevation in the right precordial leads. The Brugada pattern is found in some patients with mutations of the gene that codes for the α subunit of the sodium channel. These patients are at risk for sudden death, and a similar ECG pattern often occurs in patients who are poisoned by sodium-channel–blocking xenobiotics, including cyclic antidepressants, cocaine,²³ class IA (procainamide) antidysrhythmics and class IC (flecainide, encainide) antidysrhythmics.³⁴ In cyclic antidepressant–poisoned patients, this pattern is associated with an increased risk of hypotension, but not sudden death or dysrhythmias.¹⁰ This pattern is also associated with lithium toxicity.⁷⁰

Sagging ST segments, inverted T waves, and normal or shortened QT intervals are characteristic effects of cardioactive steroids, such as digoxin. These repolarization abnormalities are sometimes identified by their similar appearance to “Salvador Dali’s mustache” (Fig. 15–13). As a group, these findings, along with PR prolongation, are commonly described as the “digitalis effect.” They are found in patients with therapeutic digoxin concentrations and in patients with cardioactive steroid poisoning. As the serum concentration or, more precisely, the tissue concentration increases, clinical and electrocardiographic manifestations of toxicity will appear, which include profound bradycardia, AV nodal blockade, or ventricular dysrhythmias.

Changes in the ST segment duration are frequently caused by abnormalities in the serum Ca²⁺ concentration. Hypercalcemia causes shortening of the ST segment through enhanced Ca²⁺ influx during the plateau phase of the cardiac cycle, speeding the onset of repolarization. For practical purposes, this effect is more commonly identified by shortening of the QT duration (Fig. 15–14). In patients with hypercalcemia, the morphology and duration of the QRS complex and T and P waves remain unchanged. Xenobiotic-induced hypercalcemia results from exposure to antacids (ie, milk alkali syndrome), diuretics (eg, hydrochlorothiazide), cholecalciferol (vitamin D), vitamin A, and other retinoids. Hypocalcemia causes prolongation of the ST segment and QT interval.

T Wave

The T wave is the third deflection that occurs on the ECG.

Electrophysiology

The T wave represents ventricular repolarization, during which time current is again flowing, although at a cellular level in the opposite direction from that during depolarization. Cardiac repolarization on the larger level generally follows the same pattern as depolarization and thus the deflection is usually in the same direction as the QRS complex. During repolarization, the intracellular potential of the cardiac myocyte becomes more negative as a result of a net loss of positive charge because of the increasing outward flow of potassium ions. As repolarization progresses, the voltage-dependent ion channels “reset” themselves as the intracellular potential falls past their set points. Thus, the initial part of the T wave represents the absolute refractory period of the heart, because at this time there are an insufficient number of reset voltage-dependent calcium channels to allow an impulse to cause a contraction. The latter part of the T wave represents the relative refractory period of the heart, during which time a sufficient number of these calcium channels are available to open with an aberrant depolarization and initiate a contraction.

Abnormal T Wave

Isolated peaked T waves are usually evidence of early hyperkalemia.³⁹ Hyperkalemia initially causes tall, tented T waves with normal QRS and QT intervals, and a normal P wave (Fig. 15–15A). As the serum potassium concentration acutely rises to 6.5 to 8 mEq/L, the P wave diminishes in amplitude and the PR and QRS intervals prolong. Progressive widening of the QRS complex causes it to merge with the ST segment and T wave, forming a “sine wave” (Fig. 15–15B). ECG manifestations of hyperkalemia occur following chronic exposure to numerous xenobiotics, including potassium-sparing diuretics, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, or potassium supplements (Chap. 12). Fluoride, arsine, and cardioactive steroid poisoning can cause acute hyperkalemia, but the latter rarely produces hyperkalemic ECG changes. Peaked T waves also occur following myocardial ischemia and are often confused with early repolarization effects (ST segment). Consequently, the ability to properly identify electrolyte abnormalities by electrocardiography is often limited.

Hypokalemia typically reduces the amplitude of the T wave and, ultimately, causes the appearance of prominent U waves (Fig. 15–15A). Its effects on the ECG are manifestations of altered myocardial repolarization. Lithium similarly affects myocardial ion fluxes and causes reversible changes on the ECG that may mimic mild hypokalemia, although documentation of low cellular potassium concentrations is lacking.^59,68 Patients chronically poisoned with lithium have more T-wave abnormalities (typically flattening) than do those who are poisoned acutely, but these abnormalities are rarely of clinical significance.⁶⁰

QT Interval

The QT interval is measured from the beginning of the QRS complex to the end of the T wave. The QT interval normally varies because of biologic diurnal effects and autonomic tone, age, gender, technical issues with the environment or with processing and acquiring the ECG, and observer variability.^2,40,41,49 The bipolar limb lead with the largest T wave should therefore be used for this measurement.

The QT interval is normally prolonged at slower heart rates and shortens as the heart rate increases. This is especially important since many of the xenobiotics that affect the QT interval also affect the heart rate. As the normal QT also varies with the heart rate, numerous formulas and tables are available to obtain the corrected QT, known as the QTc. Using the QTc allows the determination of the appropriateness of the QT independent of the heart rate.

With a rate of 50 to 90 beats/min, the commonly used Bazett formula (QTc ms = QT (ms)/) is adequate for determining a rate-corrected QT interval (QTc). If the RR interval is calculated as 60 beats/min, 99% of men have a QTc <450 ms and 99% of women have a QTc < 460 ms.⁴² A QTc interval >500 ms weakly correlates with an increased risk of developing ventricular dysrhythmias. However, at higher heart rates, a normal patient will have an inaccurately calculated “prolonged” QTc interval using the Bazett formula. Studies suggest that medications such as bupropion prolong the QT interval when the “increase” in the QTc may be only a result of the increased heart rate.²⁶ A variety of formulas and corrections are proposed to attempt to identify normal QT intervals on ECGs at higher heart rates,^7,35 including the Fridericia formula (QTc = QT/) and the Framingham linear regression analysis (QT_LC = QT + 0.154 (1 – RR)).⁵⁴ Which correction formula is optimal remains unknown, and the FDA requests that “presentation of data with a Fridericia correction is likely to be appropriate in most situations, but other methods could be more appropriate” when performing a “thorough QT study” for any drug with prodysrhythmic potential. A QT nomogram that plots QT interval duration versus heart rate may better predict the risk for lethal dysrhythmia.¹⁵ The conventional use of QTc as a marker for dysrhythmia risk has a high false positive rate. The greatest discrepancy occurs during slow heart rates, 30 to 60 beats/min, when patients are at the highest risk for torsade de pointes. The nomogram that plots absolute QT interval duration versus heart rate is more sensitive. It is important to consider the known risk of the xenobiotic involved and whether the patient has an underlying abnormal QT interval (Fig. 15–16).^15,66 There is also evidence that the normal QT interval is shortened in adolescent males and varies substantially by age and gender.⁴⁹

With slow heart rates, a prominent U wave can obscure the terminal portion of the T wave, and with fast heart rates the subsequent P wave can obscure the terminal portion of the T wave. In these patients, estimate the QTc by following the downslope of the T wave.

QT interval measurements from the computerized ECG algorithms are less accurate than careful manual determinations of the interval. In August 2000, a panel of experts convened to address these issues and suggested that the QT interval should be measured manually in one of the limb leads that best shows the end T wave; the QT interval should be measured and averaged over 3 to 5 beats; and large U waves should be included in the QT interval measurement if they merge into the T wave and obscure the end of the T wave.⁵ However, a subsequent study of 334 health care professionals found that only 60% of the physicians were able to correctly measure a sample QT interval on the survey, although nearly all indicated correctly that the measurement should be from the beginning of the QRS to the end of the T wave.³¹

Electrophysiology

The QT represents the entire duration of ventricular systole and thus is composed of several electrophysiologic periods. Prolongation of the QT interval generally corresponds to a lengthening of the duration of phase 2 or 3 of the action potential. Although as noted above, depolarization abnormalities can affect the QT. These are uncommon, and the plateau phase and repolarization are primarily reflected by the QT. Variations in the speed of the paper,¹⁹ T wave morphology, irregular baseline, and the presence of U waves makes this determination difficult.⁵

Abnormal QT Interval

A prolonged QT reflects more time that the heart is “vulnerable” to the initiation of ventricular dysrhythmias (Table 15–1, Fig. 15–17). This occurs because although some myocardial fibers are refractory during this time period, others are not (ie, relative refractory period). Early afterdepolarizations (EADs) occur in patients with a prolonged repolarization time (Table 15–2). An EAD occurs when a myocardial cell spontaneously depolarizes before its repolarization is complete (Fig. 15–18). If this depolarization is of sufficient magnitude, it captures and initiates a premature ventricular contraction, which itself initiates ventricular tachycardia, ventricular fibrillation, or torsade de pointes. There are 2 types of EADs based on whether they occur during phase 2 (type 1) or phase 3 (type 2) of the cardiac action potential. Early afterdepolarizations are primarily due to reactivation of the voltage-gated L-type calcium channels. The importance of a pathologic rise in the late Na⁺ channel current (I_Na-L) in the genesis of EADs was studied using a sea anemone toxin (ATX II). ATX II specifically blocks I_Na-L with minimal effects on other ion currents, resulting in decreased EADs.²⁸ Early afterdepolarizations are also suppressed by magnesium.⁹

Xenobiotics that cause Na⁺ channel blockade (Chaps. 64 and 68), prolong the QT duration by slowing cellular depolarization during phase 0. Thus, the QT duration lengthens because of a prolongation of the QRS complex duration, and the ST segment duration remains near normal. Xenobiotics that cause potassium channel blockade similarly prolong the QT, but through prolongation of the plateau and repolarization phases. This specifically prolongs the ST segment duration. Although at a cellular level these xenobiotics are antidysrhythmic, the multicellular effects are prodysrhythmic. The highly selective serotonin reuptake inhibitors citalopram/escitalopram cause QT prolongation due to the potassium and calcium channel–blocking effects of the metabolite didesmethyl-citalopram. In a large retrospective cohort study, users of both typical antipsychotics (thioridazine and haloperidol) and atypical antipsychotics (clozapine, quetiapine, olanzapine, risperidone) had a risk of sudden cardiac death that was twice that of nonusers of antipsychotics. Xenobiotic-induced QT prolongation and the subsequent risk of dysrhythmias is the postulated etiology.⁵⁰ Loperamide has recently been recognized as a cause of QT prolongation and life-threatening ventricular tachycardia via inhibition of cardiac hERG channels.^18,63

Hypocalcemia of sufficient severity produces a prolonged QT interval and is caused by a number of xenobiotics, including fluoride, calcitonin, phosphates, and mithramycin (Table 12–9). Hypokalemia alone does not usually prolong the QT. Arsenic poisoning cause prolongation of the QT and results in torsade de pointes. The mechanism is unknown, although increases in cardiac Ca²⁺ currents and reduction in surface expression of the cardiac potassium channel hERG are postulated.²⁰

QT Dispersion

The QT interval varies in duration from lead to lead, reflecting a dispersion or variability in regional myocardial repolarization. Both QT and QTc dispersion can be calculated. The normal QT and QTc interval dispersions are 30 to 50 ms and 40 to 60 ms, respectively. The conduction characteristics vary regionally throughout the heart. For example, the subendocardial cells have a longer action potential duration than do epicardial cells; this is called dispersion of repolarization and is normal. This is important to allow the heart to contract and relax in an appropriate manner even though the impulse takes time to travel through the full thickness of the myocardial wall, from endocardium to epicardium. The subpopulations of the various ion channels (primarily potassium channels in the setting of repolarization) differ in character and density and account for this variation.⁴³ Ischemia and xenobiotics preferentially affect certain layers of the myocardium and alter—generally increasing—the regional heterogeneity of repolarization. M cells, located in the mid-myocardium, are very sensitive to the effects of xenobiotics that increase the repolarization heterogeneity.²⁷ This is reflected on the ECG as an increase in QT dispersion and a prolongation of the vulnerable period. This prolonged vulnerable phase is associated with occurrences of ventricular dysrhythmias. A measured QT dispersion greater than 80 ms after myocardial infarction was associated with ventricular tachycardia with a sensitivity rate of 73% and a specificity rate of 86%.⁴⁸ This heterogeneity is also correlated with both the efficacy and prodysrhythmic potential of therapy.⁶² However, the overall assessment of QT dispersion (from a standard 12-lead ECG) has not gained popularity as a useful clinical tool.

U Wave

The U wave is a small deflection that occurs after the T wave and usually with a similar orientation. Distinguishing a U wave from a notched T wave is difficult. The apices of a notched T wave are usually <150 ms apart, and the peaks of a TU complex are >150 ms apart.

Electrophysiology

U waves occur when there is fluctuation in the membrane potential following myocardial repolarization. Prominent U waves are generally representative of an underlying electrophysiologic abnormality, although they may be physiologic. Physiologic U waves are caused by repolarization of the Purkinje fibers, or they correspond to late repolarization of myocardial cells in the midmyocardium, and are implicated in the initiation of cardiac dysrhythmias.⁵

Abnormal U Wave

Abnormal U waves are typically caused by spontaneous afterdepolarization of membrane potential that occurs in situations where repolarization is prolonged (Fig. 15–18). An EAD occurs in situations where the prolonged repolarization period allows calcium channels (which are both time and voltage dependent) to close and spontaneously reopen because they may close at a membrane potential that is above their threshold potential for opening. M cells in the mid-myocardium are particularly sensitive to xenobiotics causing prolonged repolarization and EADs. In this situation, the opening of the calcium channels produces a slight membrane depolarization identified as a U wave (Fig. 15–19). A delayed afterdepolarization (DAD) occurs during phase 4 of the action potential when the myocyte is overloaded with Ca²⁺, as in the setting of cardioactive steroid toxicity. The excess intracellular Ca²⁺ can trigger the ryanodine receptors on the myocyte sarcoplasmic reticulum to release Ca²⁺, causing slight depolarization that is erroneously recognized as a U wave. If the afterdepolarization is of sufficient magnitude to reach the threshold, the cell can depolarize and initiate a premature ventricular contraction. Transient U-wave inversion can also be caused by myocardial ischemia or left ventricular overload as occurs in systemic hypertension.

Abnormal QU Interval

The QU interval is the distance between the end of the Q wave and the end of the U wave. Differentiation between the QU and the QT intervals is difficult if the T and U waves are superimposed (Fig. 15–19). When hypomagnesemia coexists with hypokalemia, as is usually the case, QU prolongation and torsade de pointes can occur.

CARDIAC DYSRHYTHMIAS AND CONDUCTION ABNORMALITIES

Xenobiotics produce adverse effects on the electrical activity of the heart, often by acting directly on the myocardial cells. Because metabolic abnormalities (especially acidemia, hypotension, hypoxia, and electrolyte abnormalities) further exacerbate the toxicity, or can actually be the sole cause of the cardiovascular abnormalities, correction of metabolic abnormalities must be a high priority in the treatment of patients with cardiovascular manifestations of poisoning. The terminal phase of any serious poisoning includes nonspecific hemodynamic abnormalities and cardiac dysrhythmias. However, many xenobiotics directly or indirectly affect cardiac rhythm or conduction, often through effects on the cardiac ion channels.

The distinction between xenobiotics that cause a rapid rate and those that cause a slow rate on the ECG is somewhat artificial, because many can do both. For example, patients poisoned by cyclic antidepressants almost always develop sinus tachycardia, but most who die will have a wide complex bradycardia immediately prior to death. In either case, abnormalities in the pattern or rate on the ECG can provide the clinician with immediate information about a patient’s cardiovascular status. ECG disturbances in many poisoned patients are categorized in more than one manner (abnormal pattern, fast rate, slow rate). In any case, when ECG abnormalities are detected, appropriate interpretation, evaluation, and therapy must be rapidly performed.

Xenobiotics that directly cause dysrhythmias or cardiac conduction abnormalities usually affect the myocardial cell membrane. Other xenobiotics that modify ion channels alter the transmembrane potentials within myocytes and may result in the spontaneous generation of an abnormal rhythm.

Mechanisms of Dysrhythmia Initiation and Propagation

Dysrhythmias are related to one or more of 3 mechanisms: abnormal spontaneous depolarization (enhanced automaticity), afterdepolarization (triggered automaticity), and reentry.⁴³ In normal myocardium, spontaneous, phase 4 depolarization occurs most rapidly in the sinus node, the normal pacemaker for the heart. Speeding or slowing the rate of phase 4 depolarization of the pacemaker cell results in sinus tachycardia or sinus bradycardia, respectively. However, xenobiotics can also speed the depolarization of other myocardial cells with pacemaker potential allowing them to overtake the sinus node as the primary pacemaker. This mechanism, called increased automaticity, accounts for many of the dysrhythmias that occur with cardioactive steroid and β-adrenergic agonist poisoning.

Afterdepolarizations, mentioned above, typically occur during phase 3 or 4 of the action potential and may reach the threshold potential, causing the fast Na⁺ channels to open and initiate an action potential. EADs account for the “trigger beats” that initiate episodes of ventricular tachycardia, commonly torsade de pointes (TdP), when the action potential is prolonged, as discussed below. DADs are typically associated with cardioactive steroid toxicity⁴³ and excessive sympathetic stimulation.

Most afterdepolarizations do not propagate rapidly throughout the myocardium and do not generate ectopic beats. However, because the normal dispersion of repolarization is increased by certain xenobiotics, ectopic beats (eg, atrial premature contraction or ventricular premature contraction) may propagate abnormally within the myocardium. Ectopy is the ECG manifestation of myocardial depolarization initiated from a site other than the sinus node. Ectopy may be lifesaving under circumstances in which the atrial rhythm cannot be conducted to the ventricles (ie, “escape rhythm”), as during high-degree AV blockade induced by cardioactive steroids (Chap. 62). Alternatively, ectopy may lead to dramatic alterations in the physiologic function of the heart or deteriorate into lethal ventricular dysrhythmias.

Occasionally, because of the altered regional repolarization (ie, increased dispersion of repolarization), an impulse reaches a branch point with a partial block (ie, relatively refractory) to conduction in one of the branches (Fig. 15–20). The impulse is carried through only one of the branches and then spreads through the myocardial cells. After a short delay, the impulse reaches the distal end of the previously blocked pathway. By this time, the region is no longer refractory and conducts the impulse in a retrograde fashion. The impulse continues in a continuous loop circuit, depolarizing the heart with each passage; this process is called reentry. Reentrant mechanisms appear to be responsible for the majority of the malignant tachydysrhythmias attributable to poisoning.

Tachydysrhythmias

Both supraventricular and ventricular tachydysrhythmias occur in poisoned patients (Table 15–3). Sinus tachycardia is the most common rhythm disturbance that occurs in poisoned patients. Parasympatholytic xenobiotics, such as atropine, raise the heart rate to its innate rate by eliminating the inhibitory tonic vagal influence. However, more rapid rates require direct myocardial stimulatory effects, generally mediated by β-adrenergic agonism. For example, catecholamine excess, as occurs in patients with cocaine use, psychomotor agitation, or fever, may cause sinus tachycardia with rates faster than 150 beats/min. Ventricular dysrhythmias frequently accompany hypotension, hypoxia, acidemia, electrolyte abnormalities, and other metabolic derangements that are present in poisoned patients or are a direct effect of the xenobiotic.

The intrinsic pacemaker cells of the heart undergo spontaneous depolarization and reach threshold at a predictable rate. Under normal circumstances, the sinus node is the most rapidly firing pacemaker cell of the heart; and as a result it controls the heart rate. Other potential pacemakers exist in the heart, but their rate of spontaneous depolarization is considerably slower than that of the sinus node. Thus, they are reset during depolarization of the myocardium and they never spontaneously reach threshold. Xenobiotics, such as sympathomimetics, which speed the rate of rise of phase 4, or diastolic depolarization, speed the rate of firing of the pacemaker cells. As long as the sinus node is preferentially affected, it maintains the pacemaker activity of the heart. If the firing rate of another intrinsic pacemaker exceeds that of the sinus node, ectopic rhythms develop.

Certain xenobiotics are highly associated with ventricular tachydysrhythmias following poisoning. Those that increase the adrenergic tone on the heart, either directly, or indirectly, cause ventricular dysrhythmias. Whether a result of excessive circulating catecholamines observed with cocaine and sympathomimetics, myocardial sensitization secondary to halogenated hydrocarbons or thyroid hormone, or increased second messenger activity secondary to methylxanthines, the extreme inotropic and chronotropic effects cause dysrhythmias. Altered repolarization, increased intracellular Ca²⁺ concentrations, or myocardial ischemia often cause dysrhythmias. Additionally, xenobiotics that produce focal myocardial ischemia, such as cocaine, ephedrine, or ergots often lead to malignant ventricular dysrhythmias. Finally, an uncommon cause of xenobiotic-induced ventricular dysrhythmias is persistent activation of sodium channels, such as following aconitine poisoning. Although not all prolonged QRS complex tachydysrhythmias are ventricular in origin, making this assumption is generally considered to be prudent. For example, in a patient poisoned with cyclic antidepressants or cocaine, the differentiation of aberrantly conducted sinus tachycardia (common) from ventricular tachycardia (rare) is important, but difficult. Although guidelines for determining the origin of a wide complex tachydysrhythmia exist, they are imperfect, difficult to apply, and unstudied in poisoned patients.^4,67

Bidirectional ventricular tachycardia is associated with severe cardioactive steroid toxicity and results from alterations of intraventricular conduction, junctional tachycardia with aberrant intraventricular conduction, or, on rare occasions, alternating ventricular pacemakers (Fig. 15–21). Aconitine, usually obtained from traditional Chinese or other alternative therapies that contain plants of the Aconitum spp. (such as Aconitum napellus {monkshood}), cause bidirectional ventricular tachycardia (Chaps. 43 and 118).

Tachydysrhythmia Associated With a Prolonged QT Interval: Torsade de Pointes

Xenobiotics that alter myocardial repolarization and prolong the QT interval predispose to the development of afterdepolarization-induced contractions during the relative refractory period (R on T phenomena), which initiates ventricular tachycardia. If torsade de pointes (TdP) is noted, this is undoubtedly the mechanism, and the QT interval should be carefully assessed and appropriate treatment initiated (Fig. 15–22).

Ventricular tachycardia, including TdP, is usually a reentrant-type rhythm. The presence of a prolonged QT interval on the ECG indicates the possible existence of conditions within the myocardium that favor occurrence of reentry dysrhythmias, as discussed above. The long action potential duration resulting from prolongation in the duration of phase 2 or 3 increases the occurrence of EADs. These, in combination with an increased dispersion of repolarization, increase the risk for reentrant dysrhythmias, particularly TdP.

Many xenobiotics interact with cardiac membrane ion channels and increase the risk of TdP. Most of these xenobiotics interact with the hERG-encoded subunit of the potassium channel to reduce the current through the I_Kr channel and prolong the action potential duration. The hERG subunit of the channel is particularly susceptible to xenobiotic interactions because of the larger inner cavity with aromatic-binding domains.^51,512 Acquired QT interval prolongation and TdP from xenobiotics occur most often with class IA and IC antidysrhythmics, the cyclic antidepressants, and the antipsychotics. Although class IC antidysrhythmics (such as encainide and flecainide) cause greater QT interval prolongation, class IA antidysrhythmics (such as quinidine and procainamide) are responsible for more reported cases. This is probably a result of the relatively infrequent use of the class IC antidysrhythmics, due paradoxically, to concerns about the higher risk of prodysrhythmic effects. Class IB antidysrhythmics, such as lidocaine, have no significant effect on potassium channels and the QT interval, and do not cause TdP. Acquired QT interval prolongation and TdP also commonly result from metabolic and electrolyte abnormalities, particularly hypocalcemia, hypomagnesemia, and hypokalemia (Chaps. 12 and 57).

Bradydysrhythmias

Bradycardia, heart block, and asystole are the terminal events following fatal ingestions of many xenobiotics, but some xenobiotics tend to cause sinus bradycardia (Table 16–2) and conduction abnormalities (Table 15–4) early in the course of toxicity. Sinus bradycardia with an otherwise normal electrocardiogram is characteristic of xenobiotics that reduce central nervous system outflow (Chap. 16). Xenobiotics that cause CNS sedation, such as the sedative-hypnotics, most opioids, and α₂-adrenergic receptor agonist (“centrally acting”) antihypertensives, and imidazolines will usually decrease sympathetic outflow to the heart and produce a heart rate in the range of 40 to 60 beats/min. Xenobiotics that directly affect ion flux across myocardial cell membranes cause abnormalities in AV nodal conduction. β-adrenergic antagonists, calcium channel blockers, and cardioactive steroids (Chaps. 59, 60, and 62) are the leading causes of sinus bradycardia and conduction disturbances. Indirect metabolic effects are also contributory, such as severe hyperkalemia (which may accompany any acidosis), causing a wide complex, sinusoidal bradycardic rhythm.

Xenobiotics with direct depressant effects on the cardiac pacemaker are most likely to produce bradycardia. The ECG manifestations of calcium channel blocker and β-adrenergic antagonist overdoses are challenging to distinguish. In general, both decrease dromotropy (conduction), although the specific pharmacologic actions of the drugs differ even within the class. For example, most members of the dihydropyridine subclass of calcium channel blockers do not have any antidromotropic effect, whereas verapamil and diltiazem routinely produce PR prolongation. Similarly, although most β-adrenergic antagonists produce sinus bradycardia and first-degree heart block, certain members of this group, such as propranolol, prolong the QRS complex through their Na⁺ channel–blocking abilities (Chap. 59). Others, such as sotalol, which has properties of the class III antidysrhythmics, block myocardial potassium channels and prolong the QT interval duration. Bradycardia associated with cardioactive steroids is typically accompanied by electrocardiographic signs of “digitalis effect,” including PR prolongation and ST segment depression (Chap. 62).

CONDUCTION ABNORMALITIES AND AV NODAL BLOCK

The cardiac toxicity of some xenobiotics results from their effects on the propagation of the electrical impulse through the conduction system of the heart. The ECG abnormalities produced are a result of effects on the AV node, producing first-, second-, or third-degree (complete) heart block, or on the His-Purkinje system, producing intraventricular conduction delays such as bundle-branch blocks. The effects of xenobiotics on myocardial conduction are often mediated through interactions with the sodium or potassium membrane channels. For example, xenobiotics that affect the fast inward I_Na currents (such as the type I antidysrhythmics and cyclic antidepressants) prolong the action potential duration, slow ventricular myocyte depolarization, and slow intraventricular conduction. This produces widening of the QRS complex and prolongation of the QT interval on the ECG. Table 15–4 lists some of the xenobiotics that cause conduction abnormalities. Many of the antidysrhythmics derive their clinical benefit from their ability to alter sodium and potassium channel function and slow conduction through the myocardium. Xenobiotics that depress phase 0 (the inward I_Na currents) produce slowing of conduction and widening of the QRS complex. Xenobiotics that prolong depolarization and repolarization (phase 2 or 3 of the action potential) produce prolongation of the QT interval on the ECG. The classes of the antidysrhythmics, their effects on the ion channels and on the action potential, and the resulting ECG abnormalities, are shown in Table 57–1 and discussed in detail in Chapters 16 and 57.

PEDIATRIC ECG

Normal Pediatric ECG

The normal pediatric ECG differs in many ways from the normal adult ECG. The resting heart rate of infants and children is substantially faster than that of adults and, in general, conduction is faster. In a term infant, the right ventricle is substantially larger than the left, and the ECG demonstrates prominent R waves in the right precordium and deep S waves in the left lateral precordium.⁴⁶ This can be misinterpreted as cyclic antidepressant toxicity.¹¹ An adult ratio of left to right ventricular size is usually reached by the age of 6 months. In infants, Q waves commonly exist in the inferior and lateral precordial leads, but are abnormal in leads I and aV_L. The T waves are the most notable difference between pediatric and adult ECGs. The T waves in the right precordial leads in children are deeply inverted until 7 years and sometimes older, in which case it is called a persistent juvenile T-wave pattern.

Abnormal Pediatric ECG

Although congenital heart disease is the most common cause of ECG abnormalities in children, electrolyte disorders and xenobiotics also cause changes in electrophysiology that are reflected on the ECG. Abnormalities that are useful markers on the adult ECG are not always as useful in children. For example, in older children, a retrospective chart review of 37 children diagnosed with tricyclic antidepressant overdose and 35 controls (<11 years) found interpatient variability, unrelated to age, so great that a rightward deviation of the terminal 40-ms QRS axis could not distinguish between poisoned and healthy children.¹¹

SUMMARY

• Electrocardiography is one of the few widely available diagnostic procedures that reveals immediate, useful clinical information.

• A thorough mechanistic understanding of the etiologies of the various waves and intervals will help provide an early clue to the etiology of a patient’s toxicity.

• QRS prolongation is characteristic of sodium channel blockade.

• QT prolongation is characteristic of potassium channel blockade.

REFERENCES

INTRODUCTION

Adequate tissue perfusion depends on maintenance of volume status, vascular resistance, cardiac contractility, and cardiac rhythm. All of these components of the hemodynamic system are vulnerable to the effects of xenobiotics. Cardiovascular toxicity is manifested by the development of hemodynamic instability, heart failure, cardiac conduction abnormalities, or dysrhythmias. The presence of a specific pattern of cardiovascular anomalies (toxicologic syndrome or “toxidrome”) often suggests a particular class or type of xenobiotic.

In addition, an alteration in hemodynamic functioning is frequently the indirect result of metabolic abnormalities. Poisoning with a xenobiotic often leads to development of acid–base disturbances, hypoxia, or electrolyte abnormalities with secondary hemodynamic changes. In these patients, supportive care with ventilation, oxygenation, and fluid and electrolyte repletion can dramatically improve the cardiovascular status.

PHYSIOLOGY OF THE CARDIOVASCULAR SYSTEM

Maintaining cardiac contractility, heart rate and rhythm, and vascular resistance requires complex modulation of the cardiac and vascular systems. Xenobiotics cause hemodynamic abnormalities as a result of direct effects on the myocardial cells, the cardiac conduction system, or on the arteriolar smooth muscle cells. These effects are frequently mediated by interactions with cellular ion channels or cell membrane neurohormonal receptors. These complex cellular systems provide multiple sites for xenobiotics to demonstrate their toxicologic effects. Xenobiotics and xenobiotic metabolites interact with the cellular receptors, intracellular signal mechanisms, effector enzymes, or intracellular organelles.

Toxic effects of xenobiotics can be due to direct poisoning from excessive amounts of a xenobiotic that follow an overdose. Slower accumulation of the xenobiotic or active metabolites (due to alterations in metabolism) also leads to adverse effects. Additionally, the toxic effects of a xenobiotic can be altered by the characteristics of the host subject. Underlying medical conditions, presence of other xenobiotics, electrolyte abnormalities, concurrent acid–base, and hydration status contribute to the potential adverse hemodynamic effects of a xenobiotic. Even with a small concentration of a xenobiotic, hemodynamic toxicity may occur due entirely to underlying genetic differences in the cellular receptors or the intracellular signal transducers in the particular patient.

This complex interaction between the xenobiotic and patient’s physiology and genetic diversity is exemplified by the Brugada syndrome. This congenital cardiac channelopathy (Chaps. 15 and 57) predisposes to sudden cardiac death due to polymorphic ventricular tachycardia or ventricular fibrillation. Brugada syndrome is characterized by an atypical right bundle branch pattern with a characteristic cove-shaped ST segment elevation in leads V1 to V3 of the electrocardiogram (ECG) in the absence or structural heart disease, ischemia, or electrolyte disturbances).^10,83 This typical type 1 Brugada ECG pattern is shown in Fig. 15-12. However, this distinctive ECG pattern can be covert³⁰ and only unmasked by sleep, fever, bradycardia, or by xenobiotics such as vagotonic medications or class I antidysrhythmics (sodium channel blockers).^4,59 The reason for this variable and dynamic response to xenobiotics is the heterogeneous genetic basis of the disorder. Mutations in each of 10 different genes have been linked to the Brugada syndrome; more than 300 different mutations have been identified in the SCN5A gene alone which encodes the α subunit of the cardiac sodium channel. Brugada syndrome is also associated with mutations in other cardiac ion channels and with mutations of the glycerol-3-phosphate 1-like (GDP1L) gene, which interacts with the cardiac sodium channel subunits at the cell membrane.⁵ The complexity of the potential xenobiotic interactions and variable potential for toxicity has lead to the establishment of a Web page (www.brugadadrugs.org) to provide up-to-date classification of xenobiotics into those to avoid, those to preferably avoid, and those to potentially use for treatment.^11,59

AUTONOMIC NERVOUS SYSTEM AND HEMODYNAMICS

In addition to the voltage-dependent ion channels, the cell membrane contains channels that open in response to receptor binding of neurotransmitters or neurohormones.^60,61 The hemodynamic effects of many xenobiotics are mediated by interactions with membrane receptors and by changes in the autonomic nervous system. The autonomic nervous system is functionally divided into the sympathetic (ie, adrenergic) and parasympathetic (ie, cholinergic) systems. The 2 systems, which share certain common features, function semi-independently of each other. The sympathetic nervous system is primarily responsible for the maintenance of arteriolar tone and cardiac inotropy and chronotropy through the release of norepinephrine and epinephrine. The parasympathetic nervous system opposes the sympathetic nervous system and reduces overall cardiovascular response such as through vagal nerve innervation of the heart, resulting in reduced heart rate. Through complex feedback, the 2 systems provide the balance needed for existence under changing external conditions.

Adrenergic Receptors

Cellular Physiology of Adrenergic Receptors

The effects of adrenergic xenobiotics on the cell are primarily mediated through a secondary messenger system of cyclic adenosine monophosphate (cAMP). The intracellular cAMP concentration is regulated by the membrane interaction of 3 components: the adrenergic receptor, a G-protein complex, and adenylate cyclase, the (enzyme that synthesizes cAMP in the cell.)^13,27,76,77 These receptors are described in detail in Chapter 13.

The G proteins serve as “signal transducers” between the receptor molecule in the cell membrane and the effector enzyme, adenylate cyclase, in the cytosol. The G proteins consist of 3 subunits: α, β, and γ.^17,54,55 The α subunit of a G protein complex binds to the adrenergic receptor in the cell membrane and to the adenylate cyclase enzyme. The G protein complex exists in several isomeric forms, depending on their interactions with adenylate cyclase. These forms, G_s, G_i, and G_q, have different functions in the regulation of cellular activity. The G_s protein complexes contain α_s subunits that stimulate adenylate cyclase when “activated” by adrenergic receptor interaction. These G_s complexes are primarily responsible for the stimulatory activity of β-adrenergic agonist agents. β₁- and β₂-adrenergic receptors interact primarily with β_s subunits in stimulatory G_s protein complexes. The α_i subunits of G_s proteins inhibit the activity of adenylate cyclase. Some β₂-adrenergic receptors and the α₂-adrenergic receptors interact with inhibitory G_i proteins to decrease the activity of adenylate cyclase. A third form, G_q, interacts with the α₁-adrenergic receptors but does not interact directly with adenylate cyclase. Instead, the G_q interacts with phospholipase C to mediate the cell response to α₁-adrenergic stimulation.

The G protein complex is composed of the cellular receptor and the 3 subunits, α, β, and γ, which are involved in the cellular response to catecholamines. In the absence of a catecholamine at the receptor site, the receptor protein is bound to the β and β-γ–dimer of the G protein, and guanosine diphosphate (GDP) is bound to the α subunit. Catecholamine binding to the receptor causes a conformational change in the α subunit; GDP dissociates and guanosine triphosphate (GTP) binds to the α subunit. The α subunit (with GTP bound) then dissociates from the receptor and from the β-γ–dimer. This “activated” α subunit then interacts with adenylate cyclase or other effector enzymes. Interaction of the α_s subunit with adenylate cyclase increases the activity of the enzyme, resulting in a rapid increase in the intracellular cAMP (Fig. 16–1).^13,17,45,54

Cyclic AMP acts as a secondary messenger in the cell. Cyclic AMP interacts with protein kinase A (PKA) and other cAMP-dependent protein kinases to increase their protein phosphorylating activity.⁴⁴ In the absence of cAMP, PKA is a tetramer of 2 regulatory and 2 catalytic subunits. Cyclic AMP binds to the regulatory subunits to release the active enzymatic units from the tetramer (Fig. 16–1). The activated protein kinases then transfer phosphate groups from ATP to serine (as well as to threonine and tyrosine amino acid groups) on enzymes that are involved in intracellular regulation and activities. Phosphorylation increases or decreases the activity of specific enzymes, and specific protein kinases are highly selective in the proteins that they phosphorylate.^73,74

Protein kinase A phosphorylates a variety of cellular proteins involved in Ca²⁺ regulation, including the voltage-sensitive Ca²⁺ channel, phospholamban, and troponin,^31,32,72 which are involved in the regulation and control of cellular muscle fiber contraction. Phosphorylation of the L-type calcium channel increases the entry of Ca²⁺ ions into the cell during membrane depolarization.⁵⁸ Phosphorylation of phospholamban decreases its ability to inhibit the calcium ATPase pump on the sarcoplasmic reticulum (SR). Decreased inhibition of the calcium ATPase pump on the SR increases the efficiency of Ca²⁺ storage in the SR, which enhances both the cellular contractility as the Ca²⁺ is released into the cell cytosol and the relaxation of muscle fibers as the Ca²⁺ is pumped back into the SR.⁵⁸

Physiologic Effects of Adrenergic Receptor Subclasses

The existence of 2 types of adrenergic receptors, α and β, was first proposed in 1948 to explain both the excitatory and the inhibitory effects of catecholamines on different smooth muscle tissue.¹ The α receptor was subsequently subdivided into α₁ and α₂ when norepinephrine and other α-adrenergic agonists were found to inhibit the release of additional norepinephrine from neurons into the synapse. The α₁-adrenergic receptors are located on postsynaptic cells outside the central nervous system, primarily on blood vessels, and mediate arteriolar constriction. The “autoregulatory” α₂-adrenergic receptors are primarily located on the presynaptic neuronal membrane and, when stimulated, decrease release of additional norepinephrine into the synapse. Additionally, some α₂-adrenergic receptors are also found on the postsynaptic membrane in the central nervous system. Activation of these postsynaptic α₂-receptors in the cardiovascular control center in the medulla and elsewhere in the central nervous system decreases sympathetic outflow from the brain. Thus, α₂-adrenergic agonists generally decrease peripheral vascular resistance, decrease heart rate, and decrease blood pressure. The α₁- and α₂-adrenergic receptors also interact with circulating catecholamines and other sympathomimetics. The effects of sympathomimetics vary in the different organ systems as a result of differences in the adrenergic receptors and in the cellular responses to the receptor interactions.

The β-adrenergic receptors are subclassified into 3 subtypes: β₁, β₂, and β₃ (Table 16–1). The most prevalent β-adrenergic subtype in the heart is β₁ (80%), although β₂ (20%) and β₃ (few) receptors are also present.^9,20,26,60 Stimulation of the β₁-adrenergic receptors increases heart rate, contractility, conduction velocity, and automaticity. The β₂-adrenergic receptors primarily cause relaxation of smooth muscle with resulting bronchodilation and arteriolar dilation. The β₃ receptors are located primarily on adipocytes, where they play a role in lipolysis and thermogenesis.¹⁴ With normal aging and in the setting of heart failure, a decline in β-adrenergic responsiveness referred to as “β-adrenergic desensitization” occurs. This process is believed to be due to a decrease in β₁-adrenergic receptor density in cardiac myocytes, alterations in signaling of G-proteins and kinase activity, and an increase in baseline circulating plasma norepinephrine and epinephrine levels.^22,81

The β₁- and β₂-receptors interact with G_s proteins and stimulate the adenylate cyclase enzyme. Differences in the resultant clinical effects are primarily related to the location and number of the different receptors in different tissues and to differences in the specificity of the tissue protein kinases activated by cAMP. Stimulation of the β₁-adrenergic receptor results in increased heart rate and increased contractility. β₂-Adrenergic receptor stimulation causes relaxation, as opposed to contraction, of smooth muscle. Because both β-adrenergic receptor subtypes interact with stimulatory G_s proteins, their clinical effects would appear to be paradoxical. However, there are 2 primary reasons for their different effects when G_s complexes are stimulated by β₁- or β₂-adrenergic agonists. First, PKA is not a single enzyme, but a group of related enzymes variably expressed in different tissues.^8,35,56 The actions and the substrates of the varied protein kinase enzymes differs between β₁- and β₂-adrenergic responsive tissues. Second, whereas β₁-adrenergic stimulation results in cAMP-mediated effects throughout the cytoplasm, β₂-adrenergic stimulation is compartmentalized within the cell. The effect of β₂-adrenergic stimulation of G_s-type receptors is primarily localized phosphorylation of the L-type calcium channels, increasing their activity.^16,39,84,85 Additionally, some β₂-adrenergic receptors are also coupled to G_i-type receptors that inhibit adenylate cyclase and prevent the diffuse cytoplasmic increases in cAMP.^69,70,85 Also, β₂-adrenergic receptor stimulation does not result in phosphorylation of phospholamban⁴² or troponins.¹⁶

The α₂-adrenergic receptor interacts with a G_i protein that has an inhibitory interaction with adenylate cyclase. Binding of α₂-adrenergic agonists to the receptor inhibits (not stimulates) adenylate cyclase and decreases intracellular cAMP.

The α₁-adrenergic receptors also are associated with G proteins. However, rather than being associated with G_s proteins and adenylate cyclase, the α₁-adrenergic receptors are associated with G_q proteins that are linked to phospholipase C.⁷⁵ Agonist binding to the receptor activates the hydrolysis of phosphatidyl inositol 4,5-bisphosphate (PIP₂) to 1,2-diacylglycerol (DAG) and inositol triphosphate (IP₃).²⁹ The IP₃ acts as an intracellular messenger, binds to receptors on the SR, and initiates the release of calcium ion.⁶ 1,2-Diacylglycerol activates protein kinase C, which phosphorylates slow Ca²⁺ channels and other intracellular proteins, and increases the influx of Ca²⁺ ion into the cell (Fig. 16–2).^62,68,87

Many xenobiotics interact with G-protein membrane receptors and alter the intracellular cAMP or Ca²⁺ concentration. β-Adrenergic antagonist overdose results in decreased stimulation of adenylate cyclase by G_s proteins, decreased production of cAMP, decreased activation of the cAMP-dependent kinases, and decreased Ca²⁺ release (Chap. 59). Similarly, by different mechanisms, Ca²⁺ channel blocker overdose results in decreased cytoplasmic [Ca²⁺] concentration (Chap. 60).

Glucagon receptors, which resemble β-adrenergic receptors, are coupled to the same G_s proteins as β-adrenergic receptors and thus stimulate adenylate cyclase activity.^28,86 The ability of glucagon to increase cAMP is further enhanced by its inhibitory activity on phosphodiesterase (preventing cAMP breakdown).^21,51 Phosphodiesterase inhibitors, such as amrinone, milrinone, and enoximone, exert at least some of their inotropic activity by preventing the degradation of cAMP and enhancing calcium cycling.^46,79,82 In a canine model of propranolol poisoning, amrinone significantly increased inotropy, stroke volume, and cardiac output.⁴⁶

INTRACELLULAR CALCIUM, CALCIUM CHANNELS, AND MYOCYTE CONTRACTILITY

The contraction and relaxation cycle of the myocyte is controlled by the flux of Ca²⁺ into and out of the SR into the cytoplasm of the cell.^18,43,63 Only a small proportion of the Ca²⁺ involved in myofibril contraction actually enters the cell through the exterior cell membrane during the action potential and membrane depolarization. Most of the Ca²⁺ is actually released from the SR of the cell invaginations of the myocyte membrane known as T-tubules, which place L-type Ca²⁺ channels in close approximation to calcium-release channels (ryanodine receptors {RyR}) on the sarcoplasmic reticulum. The local increase in Ca²⁺ concentration that follows the opening of a single L-type calcium channel triggers the opening of the associated RyR channels, resulting in a large release of Ca²⁺ from the SR.¹⁵ Myocytes contain tens of thousands of couplons, clusters of L-type calcium channels and RyR channels. The Ca²⁺ released from one couplon is not sufficient to trigger firing of neighboring couplons. Therefore, myocyte contraction requires synchronized release of Ca²⁺ from numerous couplons throughout the myocyte. The cell membrane depolarization synchronizes opening of L-type channels and subsequent Ca²⁺ release from the sarcoplasmic reticulum.^19,58 This phenomenon of Ca²⁺-induced Ca²⁺ release results in a rapid increase in the intracellular [Ca²⁺] and initiates a rapid myosin and actin interaction.¹⁸

At the conclusion of cellular contraction, SR-associated Ca²⁺-adenosine triphosphatase (ATPase) pumps return the cytosolic Ca²⁺-ATPase into the SR. This SR-associated Ca²⁺-ATPase pump is regulated by phospholamban, a cellular protein. When phospholamban is bound to the Ca²⁺-ATPase pump, the activity of the pump is decreased and less Ca²⁺-ATPase is stored in the SR. Phosphorylation of phospholamban decreases its affinity for binding to the Ca²⁺-ATPase pump. Dissociation of the phosphorylated phospholamban increases the activity of the Ca²⁺-ATPase pump. β-Adrenergic stimulation increases protein kinase activity and leads to phosphorylation of phospholamban, dissociation of the phosphorylated phospholamban from the pump, and an increase in the total SR Ca²⁺ stores.^24,25 The increased activity of the SR associated Ca²⁺-ATPase pump enhances the contractility and increases the rate of relaxation of the myocytes.

Cellular contraction occurs when myosin filaments interact with the actin-tropomyosin helix. A complex of troponins T, I, and C binds to the actin helix near the myosin-binding site and act as regulators of the interaction. Troponin T binds the regulatory complex to the actin helix, troponin I prevents myosin from accessing its binding site on the actin helix, and troponin C acts as a Ca²⁺ trigger to initiate contraction. When the intracellular [Ca²⁺] increases, 4 molecules of Ca²⁺ bind to troponin C and a conformational shift occurs in the troponin complex. Troponin I shifts away and the myosin-binding site is exposed. Myosin then binds to the exposed site and myofibril contraction occurs (Fig. 16–3).^36,37,64,65

Calcium transport through the cellular membrane ion channels is critical for normal cardiac muscle function and contractility and for maintenance of vascular smooth muscle tone. The physiologic response to Ca²⁺ channel blockers and to xenobiotics that interact with the α- or β-adrenergic receptors is mediated through changes in the intracellular Ca²⁺ concentration. Calcium channel blockers in current clinical use primarily block the L-type Ca²⁺ channel, although their specificity differs for the Ca²⁺ channels on the vascular smooth muscle cells versus on the myocardial cells. Dihydropyridine Ca²⁺ channel blockers preferentially act on peripheral vascular smooth muscle cells, whereas non-dihydropyridine Ca²⁺ channel blockers exert stronger central effects. This results in variable effects of the different Ca²⁺ channel blockers on the vascular tone and peripheral vascular resistance, and on the contractility and electrical activity of the myocardial cells.

Patients poisoned by Ca²⁺ channel blockers have less Ca²⁺ entry into the cell during cardiac membrane depolarization. Administration of exogenous Ca²⁺ increases the concentration gradient across the cell membrane, enhances flow through available Ca²⁺ channels, and restores the triggered response of the RyR-2 channels to release Ca²⁺ from the sarcoplasmic reticulum (Antidotes in Depth: A32). Because β-adrenergic antagonists have negative effects on Ca²⁺ handling by the SR, similar effects occur in the myocyte affected by β-adrenergic antagonists.

In the vascular smooth muscle, the cytosolic [Ca²⁺] maintains the basal contraction of the vascular muscle. Any decrease of Ca²⁺ influx results in relaxation and arterial vasodilation.⁵⁵ Any influx of Ca²⁺ binds calmodulin, and the resulting complex stimulates myosin light-chain kinase activity.⁴⁷ The myosin light-chain kinase phosphorylates myosin. The phosphorylated myosin has increased activity for binding to actin, which causes contraction (Fig. 60–1).^7,37

CARDIOVASCULAR EFFECTS OF XENOBIOTICS

Certain xenobiotics produce adverse effects on the cardiovascular system by acting on the myocardial cells or the autonomic nervous system to directly affect the heart rate, blood pressure, or cardiac contractility. Because metabolic abnormalities (especially acidemia, hypotension, hypoxia, and electrolyte abnormalities) can further exacerbate toxicity, or can be the sole cause of the cardiovascular abnormalities, correction of metabolic abnormalities must be a high priority in the treatment of patients with cardiovascular manifestations of poisoning. The terminal phase of any serious poisoning includes nonspecific hemodynamic abnormalities and cardiac dysrhythmias.

Heart Rate Abnormalities

Xenobiotics that directly cause dysrhythmias or cardiac conduction abnormalities usually affect the myocardial cell membrane. These abnormal rhythms, cardiac conduction abnormalities, and heart blocks are discussed in Chap. 15.

Bradycardia

Sinus bradycardia is probably the most common xenobiotic-induced bradycardia. Xenobiotics (Table 16–2) produce bradycardia through different mechanisms. The xenobiotic influences the central or peripheral nervous systems affecting rhythm generation or conduction in the heart. Most xenobiotics that cause central nervous system depression, such as the sedative-hypnotics, opioids, and α₂-adrenergic receptor agonist (“centrally acting”) antihypertensives, usually decrease sympathetic outflow to the heart and produce sinus bradycardia in the range of 40 to 60 beats/min. Digoxin, certain cholinergics, and α₁ agonists produce bradycardia and heart block through the enhancement of vagal tone. Sodium channel activators, such as aconitine, cause bradycardia due to intracellular Na⁺ overload with a resultant alteration in Ca²⁺ handling. The most profound bradycardia results from overdoses of xenobiotics that have direct depressant effects on the cardiac pacemakers.⁷¹ The inability to propagate an impulse within the cardiac conducting system produces bradycardia or even asystole. Examples of xenobiotics that produce pacemaker or conduction effects include Ca²⁺ channel blockers and β-adrenergic antagonists.

Bradycardia, heart block, and asystole frequently are terminal events in patients with massive overdose of many noncardiotoxic xenobiotics. These dysrhythmias occur as a result of direct effects on the myocardium or of indirect metabolic effects. For example, severe hyperkalemia or metabolic acidosis results in a wide complex, sinusoidal bradycardic rhythm.

Tachycardia

Sinus tachycardia is the most common xenobiotic induced tachycardia. Parasympatholytic drugs, such as atropine, raise the heart rate by elimination of the inhibitory vagal influence. However, more rapid rates result from direct myocardial stimulatory effects, generally mediated by β-adrenergic agonism. For example, catecholamine excess (eg, cocaine, psychomotor agitation, methylxanthines, sedative-hypnotic withdrawal) causes sinus tachycardia with rates faster than 150 beats/min. Tachycardia is also an indirect effect in response to hypotension, hypoxia, acidemia, fever, electrolyte abnormalities, and other metabolic derangements that present in poisoned patients (Table 16–3).

Decreased Cardiac Contractility and Congestive Heart Failure

Xenobiotics reduce cardiac contractility, with a resulting decrease in cardiac ejection fraction and cardiac output, a decrease in blood pressure, and development of congestive heart failure (CHF). Cardiogenic pulmonary edema generally occurs as a result of the direct negative effects of the xenobiotic on the contractility, or inotropy, of the heart, or through increases in the preload or afterload. Acute cardiogenic pulmonary edema, resulting from impaired left heart filling, which often results from decreased cardiac output, occurs in patients poisoned by a nondihydropyridine calcium channel blocker or β-adrenergic receptor antagonist. Other xenobiotics that exert direct depressant effects on cardiac contractility, generally through sodium channel blockade, include antihistamines, phenothiazines, antidysrhythmics, and local anesthetics. By slowing intraventricular conduction these xenobiotics reduce the ability of the heart to contract efficiently. Pulmonary edema can result from the fluid overload accompanying infusion of large quantities of sodium-containing xenobiotics (eg, sodium penicillin), the renal effects of medications such as nonsteroidal antiinflammatory drugs, or as a late consequence of xenobiotics that cause kidney failure (Chap. 27).

Xenobiotics can cause cardiomyopathy through chronic toxic effects directly on the myocardium or indirectly through effects on blood pressure or cardiac vasculature (Table 16–4). In most cases, the exact mechanism of toxicity is not known. However, free radical generation, nitric oxide formation, acetaldehyde production, myocardial ischemia, mechanical overload, nutritional deficiency, and persistent tachycardia are each implicated in the cellular toxicity of the various xenobiotics and development of cardiomyopathy.^{23,38,57,67,80}

Blood Pressure Abnormalities

Blood pressure is dependent upon cardiac and vascular function. (The blood pressure is directly related to the heart rate (HR), the stroke volume (SV), and the systemic vascular resistance (SVR): Blood pressure = HR × SV × SVR.) The systolic component of the blood pressure measurement is a reflection of the inotropic state of the myocardium, whereas the diastolic component reflects the vascular tone. Blood pressure is also expressed as the mean arterial pressure (MAP) during a single cardiac cycle. Because at normal heart rates the duration of diastole is approximately twice that of systole, the MAP is calculated as ({(2 × diastolic) + (systolic)}/3).

Systemic vascular resistance depends on vascular tone, vessel diameter, and flow properties of blood. Arteries are twice the thickness of veins and thus a higher resistance to flow and the ability to handle higher shear stress. The viscoelastic properties of large arteries dampen the pulasatile pressure and flow from cardiac output, which enables the microvasculature to steadily deliver nutrients and oxygen to tissues. Most of the pressure change in the vascular system occurs in precapillary arterioles. To maintain blood flow to vital organs over a wide range of pressures, the smooth muscle tone of these arterioles is directly linked to changes in blood pressure. Arteriolar resistance is also affected by vasoactive humoral and tissues factors such as bradykinin, vasopressin, circulating catecholamines, angiotensin II, and natriuretic peptides as well as metabolic effects such as blood oxygen and carbon dioxide content, pH, osmolarity, and K⁺. Additionally, release of nitric oxide and prostaglandin I₂ triggered by metabolic factors and/or shear stress on the endothelial surface of arterioles leads to local vasodilation and liberation of nitric oxide downstream, furthering vessel dilatation. Arterioles vary by territory and, given their different structure, function and react variably in different locations in the body.³⁴

Capillaries have very thin walls and are highly permeable to fluid. The low capillary pressure is important for maintenance of tissue fluid balance and nutrient and oxygen delivery to cells. The hydrostatic pressure within the early capillary section drives diffusion and filtration across the capillary beds, which is counterbalanced by the oncotic pressure. In the setting of circulatory shock, increased release of catecholamines causes reduced microvascular perfusion to nonessential organ systems in an attempt to maintain perfusion to vital organs. Tissue hypoxia and release of inflammatory mediators increase vascular permeability and ultimately decrease venous return and cardiac output. When blood pressure falls below the limits of autoregulation, cardiovascular instability occurs, leading first to decreased oxygen delivery and, if left uncorrected, to activation of an inflammatory cascade leading to cellular apoptosis and organ failure.^2,34,66

Hypertension Caused by Xenobiotics

Hypertension can be the result of an increase in either inotropy or vascular resistance or both. For example, stimulation of the α₁-adrenergic receptor causes hypertension through vasoconstriction, and stimulation of the β₁-adrenergic receptor causes hypertension through enhanced myocardial contractility (Table 16–5).

The hemodynamic results of a xenobiotic overdose depend on the specific xenobiotic ingested and the relative action on the various types of receptors. This suggests that, among β-adrenergic agonists, only those with a predominant β₁-adrenergic effect cause hypertension. Nonselective β-adrenergic agonists (those that agonize at both β₁ and β₂) produce β₁-mediated systolic hypertension (through inotropic effects) with β₂-mediated vascular vasodilation and diastolic hypotension. This results in a widened pulse pressure, which is the numerical difference between the systolic and diastolic pressures. Exposure to selective β₂-adrenergic agonists such as albuterol result in tachycardia and enhanced inotropy with a widened pulse pressure in a manner analogous to nonselective β-adrenergic agonists. The resulting blood pressure depends on the relative physiologic balance between inotropy and vasodilation.

Xenobiotics that interact only with the α₁-adrenergic receptor (such as phenylephrine) cause vasoconstriction and hypertension. Baroreceptors detect the increased blood pressure and signal the parasympathetic nervous system neurons of the vagus nerve to fire and slow the heart rate. In the absence of β-adrenergic stimulation, a “reflex” bradycardia results. Norepinephrine is an α₁-adrenergic agonist with additional β-adrenergic activity. Profound hypertension is the primary hemodynamic toxic effect due to the activity of norepinephrine as both a positive inotrope (β₁) and a vasoconstrictor (α₁). Reflex bradycardia does not occur as a result of stimulation of the cardiac β₁-adrenergic receptors.

Hypotension Caused by Xenobiotics

Typically, hypotension in adults is arbitrarily defined as a systolic blood pressure of less than 90 mm Hg or a MAP of less than 70 mm Hg. However, this is not an adequate clinical parameter. Young children and adults with a small body habitus often have a normal systolic pressure less than 90 mm Hg (Chap. 3). Patients with hypothermia have decreased metabolic demands, and a lower blood pressure is often “normal” for these patients. Most importantly, patients with long-standing hypertension often have inadequate tissue perfusion even with a MAP of more than 70 mm Hg. The cerebral arterioles normally constrict or dilate to maintain a relatively constant cerebrovascular blood flow despite changes in the peripheral blood pressure. Chronically hypertensive patients lose this autoregulatory response as a consequence of atherosclerotic disease, arteriolar hypertrophy, or arteriolar smooth muscle constriction. These narrowed arterioles typically require a higher peripheral blood pressure to properly perfuse the brain.

Clinically, hypotension is defined as blood pressure that is inadequate to perfuse tissues. The clinical assessment of tissue perfusion is based on the vital signs, skin color, capillary refill, mental status, urine output and concentration, and acid–base balance (eg, serum lactate concentration). However, if a xenobiotic directly affects one or more of these parameters, the clinical assessment of volume and hemodynamic status may be difficult.

Poor tissue perfusion results from hypovolemia, decreased peripheral vascular resistance, myocardial depression, or a dysrhythmia that reduces cardiac output. A single xenobiotic can exert several effects on the hemodynamic system, such as diltiazem, a calcium channel blocker that causes negative inotropy and vasodilation. Appropriate treatment of the hypotension requires an understanding of the pathophysiologic consequences of the xenobiotic and the resultant hemodynamic derangement (Table 16–6).

Hypotension in a poisoned patient also results from intravascular volume depletion. Intravascular volume decreases as a result of gastrointestinal, urinary, or insensible losses; or fluid may redistribute from the intravascular space into the intracellular, interstitial, pleural, or peritoneal spaces. Xenobiotics cause significant intravascular volume depletion through all of these mechanisms.

Hypotension is also caused by xenobiotics that affect venous tone. These xenobiotics increase venous capacitance, decrease the central venous pressure, and result in relative hypovolemia. The effects are mediated via central effects on the sympathetic nervous system or direct effects on the peripheral vasculature. Sedative hypnotics and central α₂-adrenergic agonists (eg, clonidine) decrease the central sympathetic outflow and result in hypotension. Other xenobiotics directly block peripheral α₁-adrenergic receptors or stimulate β₂-adrenergic receptors on the blood vessels to produce vascular smooth muscle relaxation, venodilation, and hypotension. A large number of xenobiotics are reported to cause hypotension; however, the hypotension often is not a direct action of the xenobiotic. Rather, the cause of hypotension is coexisting hypoxia, acidemia, anaphylaxis, volume depletion, or dysrhythmias.

Identification of the specific xenobiotic causing hypotension requires the integration of a detailed history, complete physical examination, and laboratory studies. Often the identification of the specific xenobiotic responsible for hypotension is based on other physical findings associated with the xenobiotic or recognition of a specific toxic syndrome.

ASSESSMENT OF HEMODYNAMIC STATUS IN THE POISONED PATIENT

Assessment of volume status can be particularly difficult in the poisoned patient because of functional alterations in the patient’s autonomic nervous system and the pharmacologic effects of the xenobiotic. For example, the usual signs of salt and water depletion, such as dry mucous membranes, dry skin, low blood pressure, tachycardia, narrowed pulse pressure, altered sensorium, and decreased urine output, can be mimicked by a number of xenobiotics, including cyclic antidepressants. Additionally, hypovolemic patients present with diaphoresis, flushed skin, hypertension, bradycardia, or increased urine output following exposure to a cholinergic xenobiotic such as an organic phosphorus compound.

The most accurate immediate assessment of volume status in a patient is the demonstration of increased cardiac output after a fluid challenge. If concern about providing a fluid bolus exists, providers can perform the passive leg-raising test to functionally give the patient a fluid bolus to assess the adequacy of fluid resuscitation without adding to whole-body volume status. The test is performed by having the patient sit upright in a semi-recumbent position at about 45 degrees. The head is then lowered to the recumbent position and the legs are raised to about 45 degrees. This transiently increases the circulatory volume by approximately 130 to 300 mL, and it results in changes in the arterial pressure and heart rate.^33,53 Several studies demonstrate that this simple maneuver is a good predictor of hemodynamic response to a fluid bolus especially in intensive care patients.^{12,40,41,47,49,50,52,78}

Other modes of hemodynamic monitoring have been proposed, including CVP, ultrasonography of the inferior vena (IVC), neck vein distension, and orthostatic vital sign monitoring. However, none of these modes have proven to be reliable or accurate markers of volume status and should not be relied upon when managing a critically ill poisoned patient. The only truly reliable means to assess the volume status of a patient and how a patient will respond to fluid is a trial bolus.

Lactate is a widely accepted diagnostic and prognostic marker of shock in medical and surgical patients due to trauma, sepsis, and cardiogenic shock. In circulatory failure, lactate is elevated as a result of a combination of increased production from tissue hypoxia, increased glucose metabolism/metabolic demands, and/or decreased clearance.³ In a retrospective case–control study serum lactate concentration had excellent prognostic utility to predict drug-overdose fatality.⁴⁸ However lactate as a marker of adequate microcirculatory perfusion and overall hemodynamic status is nonspecific. Elevated lactate can occur due to factors outside hemodynamics including seizure (eg, methylxanthine or cocaine toxicity), mitochondrial DNA changes (eg, nucleoside inhibitors), increased production (eg, propylene glycol), or decreased clearance (eg, metformin).

If hypoperfusion and hypotension do not resolve after initial fluid resuscitation, an echocardiogram to assess cardiac output is a useful adjunct to guide further management. For example, a patient with a decreased ejection fraction and normal volume status, which occurs in a β-adrenergic antagonist overdose, can benefit from an inotrope such as dobutamine, whereas an α₁-adrenergic agonist such as phenylephrine is indicated for a patient with hypotension despite good cardiac contractility and adequate fluid resuscitation. Definitive care of the poisoned patient with hemodynamic compromise or a dysrhythmia begins with the recognition that a possible xenobiotic ingestion occurred. Infections, cardiovascular disease, and other metabolic disorders must always be considered; however, the toxic effects of xenobiotics must be included in the differential diagnosis. A variety of clinical clues, when present, should heighten the clinician’s suspicion that a xenobiotic effect is responsible for the hemodynamic or dysrhythmic problem (Table 16–7).

SUMMARY

• Xenobiotics interact with the heart or blood vessels to produce ­hypotension or hypertension, congestive heart failure, dysrhythmias (including bradycardias and tachycardias), or cardiac conduction delays.

• These toxic effects often occur through interactions with specific receptors or with the ion channels in the cell membrane.

• Disruption of the normal cellular regulation of metabolic processes or of the cellular ionic status leads to the cardiovascular and hemodynamic compromise.

• The occurrence of these abnormalities, individually or in combination, might suggest a particular xenobiotic or class of xenobiotics as etiologic (toxic syndrome) and might dictate initial treatment.

• By understanding both the pharmacology of the xenobiotic and the physiology of the heart vasculature autonomic nervous system appropriately tailored treatment can be delivered.

REFERENCES

CASE STUDY 4

INTRODUCTION

Dermatology is a specialty in which visual inspection allows for rapid diagnosis. A brief physical examination prior to a lengthy history is valuable because some of the classic skin diseases with obvious morphologies allow a “doorway diagnosis” to be established. The tools the physician needs are readily available: magnifying glass, glass slide (for diascopy to determine if a lesion is blanchable), adequate lighting, a flashlight, alcohol pad to remove scale or makeup, scalpel, and at times a Wood lamp. Universal precautions should always be used.

The ability to describe lesions accurately is an important skill, as is the ability to recognize specific patterns. These abilities aid clinicians in their approach to the patient with a cutaneous eruption both in developing a differential diagnosis and while communicating with other physicians. The classic dermatologic lesions are defined in Table 17–1.

The skin shields the internal organs from harmful xenobiotics in the environment and maintains internal organ integrity. The adult skin covers an average surface area of 2 m². Despite its outwardly simple structure and function, the skin is extraordinarily complex. The skin is affected by xenobiotic exposures that occur through many routes. Dermal exposures themselves are important as they account for approximately 7% of all human exposures reported to the American Association of Poison Control Centers (Chap. 130). The clinician must obtain essential information as to the dose, timing, route, and location of exposure. Knowledge of the physical and chemical properties of the xenobiotic can be used to make relevant predictions of adverse cutaneous reactions and whether the response will be local or systemic. The location of xenobiotic exposure determines the histologic morphology, the severity of the reaction pattern, and the overall clinical findings. It should be noted, however, that different xenobiotics produce clinically similar skin changes and conversely that an individual xenobiotic produces diverse cutaneous lesions.

SKIN ANATOMY AND PHYSIOLOGY

The skin has 3 main components that interconnect anatomically and interact functionally: the epidermis, the dermis, and the subcutis or hypodermis (Fig. 17–1). Some experts further categorize the components of the skin into 3 reactive units: The superficial reactive unit, which is composed of the epidermis, the dermal–epidermal junction, and the superficial or papillary dermis with its vascular system; the dermal reactive unit, which is composed of the reticular layer of the dermis and the dermal microvascular plexus; and the subcutaneous reactive unit, which consists of fat lobules and septae.³⁸ The primary physiologic role of the epidermis, the outermost layer of the skin, is to serve as a barrier, maintain fluid balance, and prevent infection. The degree of barrier function of the epidermis varies with the thickness of the epidermis, which ranges from 1.5 mm on the palms and soles to 0.1 mm on the eyelids. The epidermis is composed of 4 layers: the horny layer (stratum corneum), the granular layer (stratum granulosum), the spinous layer (stratum spinosum), and the basal layer (stratum germinativum), which overlies the basement membrane zone (Fig. 17–1). The keratinocyte, or squamous cell, which is an ectodermal derivative, comprises the majority of cells in the epidermis.

The stratum corneum, a semipermeable surface composed of differentiated keratinocytes, is predominantly responsible for the physical barrier function of the skin. Disruption or abnormal formation of the stratum corneum leads to inadequate function of this barrier, whether by disorders of proliferation or desquamation. For example, accelerated cornification leads to retained nuclei in the stratum corneum (parakeratosis) causing gaps in the stratum corneum, as in psoriasis, which impedes barrier function.³⁸ Alternatively, in some forms of ichthyosis there is decreased desquamation leading to epidermal retention that influences the barrier function of the stratum corneum.⁴ Barrier function is also partly maintained by the upper spinous and granular layers. In this layer, there are Odland bodies, also known as membrane-coating granules, lamellar granules, and keratinosomes. The contents of these organelles provide a barrier to water loss while mediating stratum corneum cell cohesion.¹⁹ The stratum corneum is covered by a surface film composed of sebum emulsified with sweat and breakdown products of keratinocytes.³³ This surface film functions as an external barrier to protect from the entry of bacteria, viruses, and fungi. The role of the surface film, however, is limited with regard to percutaneous absorption. The major barrier molecules to percutaneous absorption in the skin are lipids called ceramides.³³ Diseases characterized by dry skin, such as atopic dermatitis and psoriasis, are in part caused by decreased concentrations of ceramide in the stratum corneum, which allows increased xenobiotic penetration because of barrier degradation.³³ Similarly, hydrocarbon solvents, including alcohols, or detergents, commonly produce a “defatting dermatitis” by keratolysis or the dissolution of these surface lipids.

The cells of the basal layer control the renewal of the epidermis. The basal layer contains stem cells and transient amplifying cells, which are the proliferative cells resulting in new epidermal formation that occurs approximately every 28 days.³⁸ As the basal cells migrate toward the skin surface they flatten, lose their nuclei, develop keratohyalin granules, and eventually develop into keratinocytes of the stratum corneum. The basal layer of the epidermis is just above the basement membrane zone and is also populated by melanocytes and Langerhans cells in addition to basal keratinocytes. Melanocytes contain melanin, which is the major chromophore in the skin that protects the skin from ultraviolet radiation. Melanocytes are primarily responsible for producing skin pigmentation. Langerhans cells are bone marrow–derived dendritic cells with a primary role in immunosurveillance. These cells function in the recognition, uptake, processing, and presentation of antigens to previously sensitized T lymphocytes. In addition, Langerhans cells also carry antigens via dermal lymphatics to regional lymph nodes.

The basement membrane zone (BMZ) consists of 3 layers—the lamina lucida, the lamina densa, and the sublamina densa (which is composed of anchoring fibrils) — and separates the epidermis from the dermis (Fig. 17–1). It provides a site of attachment for basal keratinocytes and permits epidermal–dermal interaction. The BMZ is also of clinical significance as it is the target of genetic defects and autoimmune attack, leading to a variety of inherited and acquired cutaneous diseases.

The dermal–epidermal junction (DEJ) provides resistance against trauma, gives support to the overlying structures, organizes the cytoskeleton in the basal cells, and serves as a semipermeable barrier. The dermis, below the DEJ, contains the adnexal structures, blood vessels, and nerves. It is arranged into 2 major regions, the upper papillary dermis and the deeper reticular dermis. The dermis provides structural integrity and contains many important appendageal structures. The structural support is provided by both collagen and elastin fibers embedded in glycosaminoglycans, such as chondroitin A and hyaluronic acid. Collagen accounts for 70% of the dry weight of the skin, whereas elastic fibers comprise 1% to 2% of the skin’s dry weight. Several important cells, including fibroblasts, macrophages, and mast cells, are residents of the dermis, each with their own unique function. Traversing the dermis are venules, capillaries, arterioles, nerves, and glandular structures.

The arteriovenous framework of the skin is derived from a deep plexus of perforating vessels within the skeletal muscle and subcutaneous fat. From this deep plexus, smaller arterioles transverse upward to the junction of the reticular and papillary dermis, where they form the superficial plexus. Capillary venules form superficial vascular loops that ascend into and descend from the dermal papillae (Fig. 17–1). The communicating blood vessels provide channels through which xenobiotics exposed on the skin surface can be transported internally. This circulatory network provides nutrition for the tissue and is involved in temperature and blood pressure regulation, wound repair, and numerous immunologic events.¹² Parallel to the vasculature are cutaneous nerves, which serve the dual function of receiving sensory input and carrying sympathetically mediated autonomic stimuli that induce piloerection and sweating.²²

The apocrine glands consist of secretory coils and intradermal ducts ending in the follicular canal. The secretory coil is located in the subcutis and consists of a large lumen surrounded by columnar to cuboidal cells with eosinophilic cytoplasm.²² Apocrine glands, which are concentrated in select areas of the body such as the axillae, eyelids, external auditory meatus, areolae, and anogenital region, produce secretions that are rendered odoriferous by cutaneous bacterial flora.

The eccrine glands, in contrast, produce an isotonic to hypotonic secretion that is modified by the ducts and emerges on the skin surface as sweat. The eccrine unit consists of a secretory gland as well as intradermal and intraepidermal ducts. The coiled secretory gland is located in the area of the deep dermis and subcutis. These glands are innervated by postganglionic sympathetic fibers that use acetylcholine neurotransmission, explaining the clinical effects of anticholinergic xenobiotics. Xenobiotics that are concentrated in the sweat increase the intensity of the local skin reactions. Certain chemotherapeutics, such as cytarabine or bleomycin, directly damage the eccrine sweat glands, resulting in neutrophilic eccrine hidradenitis.⁶⁵

Sebaceous glands also reside in the dermis. They produce an oily, lipid-rich secretion that functions as an emollient for the hair and skin, and can be a reservoir of noxious environmental xenobiotics. Pilosebaceous follicles, which are present all over the body, consist of a hair shaft, hair follicle, sebaceous gland, sensory end organ, and erector pili. Certain halogenated aromatic chemicals, such as polychlorinated biphenyls (PCBs), dioxin, and 2,4-dichlorophenoxyacetic acid, are excreted in the sebum and cause hyperkeratosis of the follicular canal. This produces the syndrome, chloracne, which appears clinically like severe acne vulgaris but predominates in the malar, retroauricular, and mandibular regions of the head and neck and typically develops after several weeks of exposure (Fig. 17–2). Similar syndromes result from exposure to brominated and iodinated compounds, and are known as bromoderma and ioderma, respectively.⁷⁰

The subcutis serves to insulate, cushion, and allow for mobility of the overlying skin structures. Adipocytes represent the majority of cells found in this layer. Leptin, an adipose-derived hormone responsible for feedback of appetite and satiety signaling, is synthesized and regulates fat mass (adiposity) in this layer.

The hair follicle is divided into 3 portions, the hair bulb, infundibulum, and isthmus.⁵⁷ The deepest portion of the hair follicle contains the bulb with matrix cells. The matrix cells are highly mitotically active and often are the target of cytotoxic xenobiotics. The rate of growth and the type of hair are unique for different body sites. Hair growth proceeds through 3 distinct phases: the active prolonged growth phase (anagen phase) during which matrix cell mitotic activity is high; a short involutional phase (catagen phase), and a resting phase (telogen phase). The length of the anagen phase determines the final length of the hair and varies depending on site of the body. For example, hair on the scalp has the longest anagen phase ranging from 2 to 8 years with hair growth at a rate of 0.37 to 0.44 mm/day.³⁸ Understanding the phases of hair growth is important because hair growth can be used to identify clues regarding the timing of exposure and the mechanism of action of a particular xenobiotic.

The nail plate, which is often considered analogous to the hair, is also a continuously growing structure. Fingernails grow at average of 2 to 3 mm per month and toenails grow approximately 1 mm per month. The mitotically active cells of the nail matrix that produce the nail plate are subject to both traumatic and xenobiotic injury, which in turn affects the appearance and growth of the nail plate. Because nail growth is relatively stable, location of an abnormality in the plate can predict the timing of exposure, such as Mees lines (transverse white lines).

TOPICAL TOXICITY

Transdermal Xenobiotic Absorption

Although there is no active cutaneous uptake mechanism for xenobiotics, many undergo percutaneous absorption by passive diffusion. Lipid solubility, concentration gradient, molecular weight, and certain specific skin characteristics are important determinants of dermal absorption.^23,24,50,54 Absorption is determined to a great extent by the lipid solubility of the specific xenobiotic.^15,39 The pharmacokinetic profile of transdermally administered xenobiotics is markedly different than by the enteral or other parenteral routes.¹⁷ As with any other routes of administration, adverse effects are caused by excessive absorption following application or even with therapeutic use of a transdermal patch. Other xenobiotics, topically applied without a specific delivery device, including podophyllin, camphor, phenol, organic phosphorus compounds, ethanol, organochlorines, nitrates, and hexachlorophene, can lead to systemic morbidity and mortality (Special Considerations: SC3).

Direct Dermal Toxicity

Exposure to any of a myriad of industrial and environmental xenobiotics results in dermal “burns.” Although the majority of these xenobiotics injure the skin through chemical reactivity rather than thermal damage, the clinical appearances of the two are often identical. Injurious xenobiotics act as oxidizing or reducing agents, corrosives, protoplasmic poisons, desiccants, or vesicants. Often an injury initially appears to be mild or superficial with minimal erythema, blanching, or discoloration of the skin. Over the subsequent 24 to 36 hours, the injury progresses to extensive necrosis of the skin and subcutaneous tissue.

Both inorganic and organic acids are capable of penetrating and damaging the epidermis via protein denaturation and cytotoxicity; however, organic acids tend to be less irritating. The damaged tissue coagulates and forms a thick eschar, which limits the spread of the xenobiotic. The histopathologic finding following acid injury is termed coagulative necrosis.¹⁰ Inorganic acids that are frequently used in industry include hydrochloric and sulfuric acids which lead to the severe injury. The weakly acidic hydrofluoric (HF) acid, is used for the etching of glass, metal and stone. Hydrofluoric acid, because of its limited dissociation constant, is able to penetrate intact skin with subsequent penetration into deeper tissues. The fluoride ion is extremely cytotoxic, causing severe tissue damage, including bone destruction, by interfering with cellular enzymes. Severe pain is due to the capacity of fluoride ions to bind tissue calcium, thus affecting nerve conduction.⁶¹ Once in the dermis, the proton (H⁺) and fluoride ions (F^–) cause both acid-induced tissue necrosis and fluoride-induced toxicity (Chap. 104).⁵

Alkali exposures characteristically produce a liquefactive necrosis, which allows continued penetration of the corrosive. Consequently, cutaneous and subcutaneous injury following alkali exposure is typically more severe than after an acid exposure, with the exception of hydrofluoric acid. With alkali burns there are generally no vesicles, but rather necrotic skin due to the disruption of barrier lipids, including denaturation of proteins with subsequent fatty acid saponification. Common strong alkalis include sodium, ammonium, and potassium hydroxide; sodium and potassium carbonate; and calcium oxide. These are used primarily in the manufacture of bleaches, dyes, vitamins, pulp, paper, plastics, and soaps, and detergents. Alkali burns from wet cement, which has an initial pH of 10 to 12 that rises to 12 to 14 as the cement sets,³⁷ result from the liberation of calcium hydroxide.

Thermal damage also results from xenobiotic exposure. For example, the exothermic reaction generated by the wetting of elemental phosphorus or sodium results in a thermal burn.¹⁸ In these circumstances, the products of reactivity, phosphoric acid and sodium hydroxide, respectively, produce secondary chemical injury. Alternatively, skin exposure to a rapidly expanding gas, such as nitrous oxide from a whipped cream cartridge or compressed liquefied nitrogen, or to frozen substances, such as dry ice (CO₂), produce a freezing injury, or frostbite. Dermatologists routinely use liquid nitrogen to induce a cold injury that destroys precancerous lesions such as actinic keratoses.

Hydrocarbon-based solvents are liquids that are capable of dissolving non–water-soluble solutes.¹⁰ Although the most prominent effect is dermatitis due to loss of ceramides from the stratum corneum of the epidermis, prolonged exposure results in deeper dermal injury.

PRINCIPLES OF DERMAL DECONTAMINATION

On contact with xenobiotics, the skin should be thoroughly cleansed to prevent direct effects and systemic absorption. In general, water in copious amounts is the decontaminant of choice for skin irrigation. Soap should be used when adherent xenobiotics are involved. Following exposures to airborne xenobiotics, the mouth, nasal cavities, eyes, and ear canals should be irrigated with appropriate solutions such as water or a 0.9% NaCl solution. For nonambulatory patients, the decontamination process is conducted using special collection stretchers if available.⁹

There are a few situations in which water should not be used for skin decontamination. These situations include contamination with the reactive metallic forms of the alkali metals, sodium, potassium, lithium, cesium, and rubidium, which react with water to form strong bases. The dusts of pure magnesium, sulfur, strontium, titanium, uranium, yttrium, zinc, and zirconium will ignite or explode on contact with water. Following exposure to these metals, any residual metal should be removed mechanically with forceps, gauze, or towels and stored in mineral oil. Phenol, a colorless xenobiotic used in the manufacturing of plastics, paints, rubber, adhesives, and soap, has a tendency to thicken and become difficult to remove following exposure to water. Suggestions for phenol decontamination include alternating washing with water and polyethylene glycol (PEG 400) or 70% isopropanol for 1 minute each for a total of 15 minutes.⁴⁰ Our conclusions are that inexpensive readily available tepid water should be utilized (Special Considerations: SC2). Calcium oxide (quicklime) thickens and forms Ca(OH)₂ following exposure to water, which releases heat and causes cutaneous ulcerations, suggesting that mechanical removal as above is advised.

DERMATOLOGIC SIGNS OF SYSTEMIC DISEASES

Cyanosis

Normal cutaneous and mucosal pigmentation is caused by several factors, one of which is the visualization of the capillary beds through the translucent epidermis and dermis. Cyanosis manifests as a blue or violaceous appearance of the skin, mucous membranes, and nailbeds. It occurs when excessive concentrations of reduced hemoglobin (>5 g/dL) are present, as in hypoxia or polycythemia, or when oxidation of the iron moiety of heme to the ferric state (Fe³⁺) forms methemoglobin, which is deeply pigmented (Chap. 124). The presence of the more deeply colored hemoglobin moiety within the dermis results in cyanosis that is most pronounced on areas of thin skin such as the mucous membranes or underneath fingernails. In the differential diagnosis of skin discoloration is pseudochromhidrosis, also termed extrinsic apocrine chromhidrosis. The discoloration is a product of staining of the sweat by chromogenic bacteria including Corynebacterium, Malassezia furfur, and Bacillus spp; the latter 2 species have been known to cause blue discoloration of the skin. Several cases of blue pseudochromhidrosis due to topiramate are reported in the literature, and diagnosis is established by the ability of the clinician to wipe off the discoloration with a damp cotton swab (Fig. 17–3).¹¹

Xanthoderma

Xanthoderma is a yellow to yellow-orange macular discoloration of skin.²⁶ Xanthoderma is caused by xenobiotics such as carotenoids, which deposit in the stratum corneum, and causes carotenoderma. Carotenoids are lipid soluble and consist of α-carotene, β-carotene, β-cryptoxanthin, lycopene, lutein, and zeaxanthan, and serve as precursors of vitamin A (retinol). The carotenoids are excreted via sweat, sebum, urine, and GI secretions. Jaundice is typically a sign of hepatocellular failure or hemolysis and is caused by hyperbilirubinemia, either conjugated or unconjugated, deposits in the subcutaneous fat. Jaundice due to hyperbilirubinemia is often accompanied by other cutaneous stigmata including spider angiomas, telangiectasias, palmar erythema, and dilated superficial abdominal veins (caput medusae). True hyperbilirubinemia is differentiated from hypercarotenemia by the presence of scleral icterus in patients with hyperbilirubinemia. In addition, the cutaneous discoloration seen in hypercarotenemia can be removed by wiping the skin with an alcohol swab. Hypercarotenemia is reported among people who take carotene nutrient supplements (Fig. 17–4).⁵⁹ Lycopenemia, an entity similar to carotenemia, is caused by the excessive consumption of tomatoes, which contain lycopene. Additionally, topical exposure to dinitrophenol, picric acid, or stains from cigarette use produces localized yellow discoloration of the skin.

Pruritus

Pruritus is the poorly localized, unpleasant sensation that elicits a desire to scratch. The biologic purpose of pruritus is to provoke the removal of a pruritogen, a response likely to have originated when most pruritogens were parasites. Pruritus is a common manifestation of urticarial reactions, but at times it is of nonimmunologic origin. Pruritus is the most common dermatologic symptom and can arise from a primary dermatologic condition or is a symptom of an underlying systemic disease in an estimated 10% to 50% of patients.²⁹ Pruritus is also caused by topical exposure to the urticating hairs of Tarantula spiders, spines of the stinging nettle plant (Urtica spp), or via stimulation of substance P by capsaicin.²⁵ Virtually any xenobiotic can cause a cutaneous reaction that can be associated with pruritus, whether by inducing hepatotoxicity, cholestasis, phototoxicity, or histamine release (ie, neurologically mediated). ­Xenobiotics commonly implicated in neurally mediated itch include tramadol, codeine, cocaine, morphine, butorphanol, and methamphetamine.²⁹

Flushing

Vasodilation of the dermal arterioles leads to flushing, or transient reddening of the skin, commonly of the face, neck, and chest. Flushing occurs following autonomically mediated vasodilation, as occurs with stress, anger, or exposure to heat, or it can be induced by vasoactive xenobiotics. Xenobiotics that cause histamine release through a type I hypersensitivity reaction are the most frequent cause of xenobiotic-induced flush. Histamine poisoning produces flushing from the consumption of scombrotoxic fish (Chap. 39). Flushing after the consumption of ethanol is common in patients of Asian and Inuit descent and is similar to the reaction following ethanol consumption in patients exposed to disulfiram or similar xenobiotics (Chap. 78). The inability to efficiently metabolize acetaldehyde, the initial metabolite of ethanol, results in the characteristic syndrome of vomiting, headache, and flushing. Niacin causes flushing through an arachidonic acid–mediated pathway that is generally prevented by aspirin.^7,66 Vancomycin, if too rapidly infused, causes a transient bright red flushing, mediated by histamine and at times can be accompanied by hypotension. This reaction typically occurs during and immediately after the infusion, and is termed “red man syndrome.” Idiopathic flushing is managed with nonselective beta-adrenergic antagonists (nadolol, propranolol) or clonidine, while anxiolytics are beneficial if emotional distress or anxiety is determined to be causative. Other nontoxicologic causes of flushing including carcinoid syndrome, pheochromocytoma, mastocytosis, anaphylaxis, medullary carcinoma of the thyroid, pancreatic cancer, menopausal flushing, and renal carcinoma, are in the differential diagnosis of the flushed patient.²⁸

Sweating

Xenobiotic-induced diaphoresis is either part of a physiologic response to heat generation or is pharmacologically mediated following parasympathetic or sympathomimetic xenobiotic use. Because the postsynaptic receptor on the eccrine glands is muscarinic, most muscarinic agonists stimulate sweat production. Sweating occurs following exposure to cholinesterase inhibitors, such as organic phosphorus compounds, but also occurs with direct-acting muscarinic agonists such as pilocarpine. Alternatively, antimuscarinics, such as belladonna alkaloids or antihistamines, reduce sweating and produce dry skin. Certain xenobiotics have proven useful for the treatment of hyperhidrosis including the anticholinergics glycopyrrolate, propantheline bromide, and botulinum toxin. Botulinum toxin-A derived from Clostridium botulinum, which is FDA approved for the treatment of primary focal axillary hyperhidrosis, temporarily chemodenervates eccrine sweat glands at the neuroglandular junction via inhibition of presynaptic acetylcholine release.¹

Xenobiotic-Induced Dyspigmentation

Cutaneous pigmentary changes can result from the deposition of xenobiotics that are ingested and carried to the skin by the blood or that permeate the skin from topical applications. Many heavy metals are associated with dyspigmentation. Argyria, a slate-colored discoloration of the skin resulting from the systemic deposition of silver particles in the skin after excessive ingestion of colloidal silver, can be localized or widespread. The discoloration tends to be most prominent in areas exposed to sunlight, probably because silver stimulates melanocyte proliferation. Histologically, fine black granules are found in the basement membrane zone of the sweat glands, blood vessel walls, the dermoepidermal junction, and along the erector pili muscles (Chap. 98). Gold, which was historically used parenterally in the treatment of rheumatoid arthritis, caused a blue or slate-gray pigmentation, often periorbitally, known as chrysiasis. The pigmentation is also accentuated in sun-exposed areas but, unlike argyria, sun-protected areas do not histologically demonstrate gold. Also, melanin is not increased in the areas of hyperpigmentation. The hyperpigmentation is probably caused by the gold itself, but the cause of its distribution pattern remains unknown. Histologically, the gold is found within lysosomes of dermal macrophages and distributed in a perivascular and perieccrine pattern in the dermis. Bismuth produces a characteristic oral finding of the metallic deposition in the gums and tongue known as bismuth lines, as well as a blue-gray discoloration of the face, neck, and dorsal hands. Chronic arsenic exposure occurs following exposure to pesticides or contaminated well-water, which can cause cutaneous hyperpigmentation with a bronze hue, with areas of scattered hypopigmentation occurring between 1 and 20 years following exposure. Lead also deposits in the gums, causing the characteristic “lead lines,” which are the result of subepithelial deposition of lead granules. Intramuscular injection of iron stains the skin, resulting in pigmentation similar to that seen in tattoos, and iron storage disorders, known as hemochromatosis, and results in a bronze appearance of the skin.²¹

Medications are also often implicated in dyspigmentation. The tetracycline-class antibiotic minocycline is a highly lipid-soluble, yellow crystalline xenobiotic that turns black with oxidation. Minocycline-induced discoloration of the skin is at times accompanied by darkening of the nails, sclerae, oral mucosa, thyroid, bones, and teeth. Hyperpigmentation from minocycline is divided into 3 types depending on the color, anatomic distribution, and whether iron- or melanin-containing granules are found within the skin. Other medications associated with hyperpigmentation include amiodarone, zidovudine, bleomycin, and other chemotherapeutics, antimalarials, and psychotropics (chlorpromazine, thioridazine, imipramine, desipramine, amitriptyline).³⁰ Although not true dyspigmentation, as noted above topiramate is linked to blue pseudochromhidrosis.¹¹

XENOBIOTIC-INDUCED CUTANEOUS REACTIONS (DRUG REACTIONS)

The skin is one of the most common targets for adverse drug reactions.³ Drug eruptions occur in approximately 2% to 5% of inpatients and in greater than 1% of outpatients. Several cutaneous reaction patterns account for the majority of clinical presentations of xenobiotic-induced dermatotoxicity (Table 17–2). The following drug reactions will be discussed in detail: urticaria, erythema multiforme, Steven-Johnson syndrome and toxic epidermal necrolysis, fixed eruptions, and drug-induced hypersensitivity syndrome (formerly called “DRESS” syndrome).

Urticarial Drug Reactions

Urticarial drug reactions are characterized by transient, pruritic, edematous, pink papules, or wheals that arise in the dermis, which blanch on palpation and are frequently associated with central clearing. At times, the urticarial lesions are targetoid and mimic erythema multiforme. Approximately 40% of patients with urticaria experience angioedema and anaphylactoid reactions as well.² The reaction pattern is representative of a type I, or IgE-dependent, immune reaction and commonly occurs as part of clinical anaphylaxis or anaphylactoid (non–IgE-mediated) reactions. Widespread urticaria occurs following systemic absorption of an allergen or following a minimal localized exposure in patients highly sensitized to the allergen. Regardless of whether the eruption is localized or widespread, it occurs as a result of immunologic recognition of a putative antigen by IgE antibodies, thus triggering the immediate degranulation of mast cells, which are distributed along the dermal blood vessels and nerves. The release of histamine, complements C3a and C5a, and other vasoactive mediators results in extravasation of fluid from dermal capillaries as their endothelial cells contract. This produces the characteristic urticarial lesions described above. Activation of the nearby sensory neurons produces pruritus. Nonimmunologically mediated mast cell degranulation producing an identical urticarial syndrome also occurs following exposure to any xenobiotic.¹⁴

Erythema Multiforme

Historically, it was believed that erythema multiforme existed on a spectrum with Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) given overlapping clinical features and morphology. However, these entities were reclassified on the basis that most cases of erythema multiforme are believed to be triggered by viral infection (herpes simplex virus most commonly) and most cases of SJS/TEN are triggered by xenobiotics.⁵¹ Erythema multiforme is an acute self-limited disease characterized by target-shaped, erythematous macules and patches on the palms and soles, as well as the trunk and extremities (Fig. 17–5). The Nikolsky sign, defined as sloughing of the epidermis when direct pressure is exerted on the skin, is absent. Mucosal involvement is absent or mild in erythema multiforme minor and severe in erythema multiforme major. While less common than viral-induced erythema multiforme, xenobiotics such as sulfonamides, phenytoin, antihistamines, many antibiotics, rosewood, and urushiol also elicit erythema multiforme. Differentiating erythema multiforme from SJS/TEN, which can also present with targetoid lesions, can be difficult, especially in the case of bullous erythema multiforme. A skin biopsy revealing partial or full-thickness epidermal necrosis would favor a diagnosis of SJS/TEN rather than erythema multiforme.

Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis

Toxic epidermal necrolysis (TEN) and Stevens-Johnson syndrome (SJS) (Fig. 17–6) are considered to be related disorders that belong to a spectrum of increasingly severe skin eruptions.⁴⁵ Stevens-Johnson syndrome is defined by lesser than 10% body surface area epidermal detachment, SJS-TEN overlap 10% to 30% involvement, and TEN greater than 30% epidermal sloughing. Although on a spectrum, SJS has a mortality rate of 5%, which is far lower than the approximate 25% to 50% mortality rate for patients with TEN.^49,52 However, a more recent series of 40 patients with TEN cared for at academic burn units revealed a 10% mortality rate.³⁵ Toxic epidermal necrolysis is a rare, life-threatening dermatologic emergency whose incidence is estimated at 0.4 to 1.2 cases per 1 million persons. More than 220 xenobiotics are causally implicated in 80% to 95% of the TEN cases. The largest study examining medication triggers of TEN divided these medications into long-term (used for months to years) and short-term. Short-term xenobiotics most commonly implicated in the development of TEN included trimethoprim-sulfamethoxazole and other sulfonamide antibiotics, followed by cephalosporins, quinolones, and aminopenicillins.⁵³ With chronic medication use, the increased risk largely occurred during the first 2 months of treatment and was greatest for carbamazepine, phenobarbital, phenytoin, valproic acid, the oxicam NSAIDs, allopurinol, and corticosteroids.

Classically, the eruption of TEN is painful and occurs within 1 to 3 weeks after the exposure to the implicated xenobiotic(s). The eruption is preceded by malaise, headache, abrupt onset of fever, myalgia, arthralgia, nausea, vomiting, diarrhea, chest pain, or cough. One to 3 days later, signs begin in the mucous membranes including the eyes, mouth, nose, and genitals in 90% of cases.⁴⁹ A macular erythema then develops that subsequently becomes raised and morbilliform on the face, neck, and central trunk and finally the extremities. Individual lesions can appear targetoid because of their dusky centers and progress to bullae in the following 3 to 5 days involving the entire thickness of the epidermis. This necrosis and sloughing can also lead to loss of the fingernails. A Nikolsky sign, or sloughing of the epidermis with gentle manual pressure, is suggestive of TEN but not pathognomonic. A Nikolsky sign occurs in a variety of other dermatoses, including pemphigus vulgaris. If the diagnosis is suspected, a punch biopsy should be performed for immediate frozen section and the suspected triggering xenobiotic discontinued immediately. The histopathology typically shows partial- or full-thickness epidermal necrosis, with subepidermal bullae with a sparse infiltrate, and vacuolization with numerous dyskeratotic keratinocytes along the dermoepidermal junction adjacent to the necrotic epidermis.

The incidence of TEN is higher in patients with advanced HIV disease.^45,62 There is general agreement that the keratinocyte cell death in TEN is the result of apoptosis, which is suggested based on electronic microscopic studies with DNA fragmentation analysis.⁴⁵ Cytotoxic T lymphocytes are the main effector cells and experimental evidence points to involvement of the Fas-ligand (FasL) and perforin/granzyme pathways. There are several theories as to the pathogenesis of SJS/TEN. These include that a xenobiotic could induce upregulation of Fas Ligand leading to a death receptor–mediated apoptotic pathway. A xenobiotic might interact with MHC class I–expressing cells and cause drug-specific CD8⁺ cytotoxic T lymphocytes to accumulate within epidermal blisters, releasing perforin and granzyme B that kills keratinocytes. Finally it has been proposed that the xenobiotic triggers the activation of CD8⁺ T lymphocytes, NK cells and NKT cells to secrete granulysin, with keratinocyte death not requiring cell contact.⁴⁴ Serum Fas ligand concentrations are elevated up to 4 days prior to mucosal involvement in patients with SJS/TEN and have the potential to be useful clinically as an early predictor of these severe dermatologic diseases.⁴³ Serum granulysin, a proinflammatory cytolytic enzyme released by CD8+ T lymphocytes found in the blisters of TEN was demonstrated to be a potential early predictive marker of SJS/TEN.²⁰ A rapid immunochromatographic test that detects elevated serum granulysin (>10 ng/mL) in 15 minutes demonstrates promise in a small study in which its sensitivity was noted to be 80% and specificity 95.8% for differentiating SJS/TEN from ordinary exanthematous drug eruptions.²⁰ However, this test is not yet commercially available.

Because immediate removal of the inciting xenobiotic is critical to survival, patients with TEN related to a xenobiotic with a long half-life have a poorer prognosis, and these patients should be transferred to a burn or other specialized center for sterile wound care. Risk factors for mortality include older patient age, higher total surface area of involvement, and more pre-existing comorbidities.³⁵ In a recent study, only serum bicarbonate less than 20 mmol/L was found to portend hospital mortality in patients with TEN.⁶⁸ Porcine xenografts or human skin allografts including amniotic membrane transplantation are used and are widely accepted therapy.⁴⁸ A meta-analysis of 17 studies revealed a trend toward improved mortality with high-dose IVIG in adults and good prognosis in children; however, there is insufficient evidence to support a clinical benefit.²⁷ Patients with TEN develop metabolic abnormalities, sepsis, multiorgan failure, ­pulmonary emboli, and gastrointestinal hemorrhages and should be closely monitored. The major microbes leading to sepsis are Staphylococcus aureus and Pseudomonas aeruginosa. In a patient with SJS/TEN with ophthalmic involvement, early ophthalmologic consultation is necessary because blindness is a potential complication.

Mimickers of TEN include SJS, staphylococcal scalded skin syndrome, severe exanthematous drug eruptions, erythema multiforme-major, linear IgA dermatosis, paraneoplastic pemphigus, acute graft versus host disease, drug-induced pemphigoid and pemphigus vulgaris, and acute generalized exanthematous pustulosis. Discussion of some of these entities is beyond the scope of this chapter (Table 17–3).

Bullous Reactions (Blistering Reactions)

In addition to SJS and TEN, other bullous cutaneous reactions include drug-induced pseudoporphyria, fixed drug eruption, acute generalized exanthematous pustulosis, phototoxic drug eruptions, and drug-induced autoimmune blistering diseases. Xenobiotic-related cutaneous blistering reactions are clinically indistinguishable from autoimmune blistering diseases such as pemphigus vulgaris or bullous pemphigoid (Fig. 17–7). Certain topically applied xenobiotics such as the vesicant cantharidin derived from “blister beetles” in the Coleoptera order and Meloidae family are used in the treatment of molluscum and viral warts. In high concentrations, xenobiotics lead to necrosis of both skin and mucous membranes. Other systemic xenobiotics cause a similar reaction pattern mediated by the production of antibody directed against the cells at the dermal–epidermal junction (Table 17–3).

A number of medications, many of which contain a “thiol group” such as penicillamine and captopril, induce either pemphigus resembling pemphigus foliaceus, a superficial blistering disorder in which the blister is at the level of the stratum granulosum, or pemphigus vulgaris, in which blistering occurs above the basal layer of the epidermis (Fig. 17–1). Other xenobiotics, such as furosemide, penicillin, and sulfasalazine produce tense bullae that resemble bullous pemphigoid. Direct immunofluorescence studies demonstrate epidermal intracellular immunoglobulin deposits at the dermal–epidermal junction. The offending xenobiotic should be discontinued and the patient should be referred to a dermatologist to determine treatment. The reaction may persist for up to 6 months after the offending xenobiotic is withdrawn.

Fixed Drug Eruptions

Fixed drug eruption is another bullous drug eruption and is characterized by well-circumscribed erythematous to dusky violaceous patches which with central bullae or erosions and develops 1 to 2 weeks after first exposure to the drug. This reaction pattern is so named because re-exposures to the xenobiotic cause lesions in the same area, typically within 24 hours of exposure (Fig. 17–8). Typical locations include acral extremities, genitals, and intertriginous sites, and this process is confused with TEN if widely confluent as in a “generalized fixed drug eruption.” This reaction pattern is generally not life threatening and heals with residual postinflammatory hyperpigmentation. Bullous fixed-drug reactions result from exposure to diverse xenobiotics such as angiotensin-converting enzyme inhibitors and a multitude of antibiotics. As mentioned above, EM can have a bullous variant that can also be confused with SJS/TEN.

Coma Bullae

“Coma bullae” are tense bullae on normal-appearing skin that occur within 48 to 72 hours in comatose patients with sedative–hypnotic overdoses, particularly phenobarbital, or carbon monoxide, poisoning. They also occur in patients in coma from infectious, neurologic, or metabolic causes. Although these blisters are thought to result predominantly from pressure-induced epidermal necrosis, they occasionally occur in non–pressure-dependent areas, suggesting a systemic mechanism. Histologically, an intraepidermal or subepidermal blister is observed. There is accompanying eccrine duct and gland necrosis.

Drug-Induced Hypersensitivity Syndrome

The skin is linked with systemic immunologic diseases such that an alteration in the metabolism of certain xenobiotics leads to a hypersensitivity syndrome. The drug-induced hypersensitivity syndrome, formerly called Drug Reaction with Eosinophilia and Systemic Symptoms (DRESS), can be severe and potentially life threatening. The hypersensitivity syndrome is characterized by the triad of fever, skin eruption, and internal organ involvement.³² The frequency is estimated between 1 in 1,000 to 1 in 10,000 with antiepileptic or sulfonamide antibiotic exposures and usually presents within 2 to 6 weeks of the initial exposure. For antiepileptics, the inability to detoxify arene oxide metabolites is suggested to be a key factor; once a patient has a documented drug-induced hypersensitivity syndrome to one antiepileptic, it is important to note that cross-reactivity between phenytoin, carbamazepine, and phenobarbital is well documented, both in vivo and in vitro.⁴⁷ In the case of sulfonamides, acetylator phenotype and lymphocytes’ susceptibility to the metabolite hydroxylamine are risk factors for developing drug hypersensitivity syndrome. Further support for the role of genetic predisposition comes from data in Northern European populations in which the presence of the HLA-A*3101 allele significantly increases the risk of developing carbamazepine-induced hypersensitivity syndrome.³⁴ Fever and a cutaneous eruption are the most common symptoms. Accompanying malaise, pharyngitis, and cervical lymphadenopathy are frequently present. Atypical lymphocytes and eosinophilia occur initially. The exanthem is initially generalized and morbilliform, and conjunctivitis and angioedema occur (Fig. 17–9). Later the eruption becomes edematous, and facial edema, which is often present, is a hallmark of this syndrome. One-half of patients with drug induced hypersensitivity syndrome will have hepatitis, interstitial nephritis, vasculitis, CNS manifestations (including encephalitis, aseptic meningitis), interstitial pneumonitis, acute respiratory distress syndrome, and autoimmune hypothyroidism. Hepatic involvement can be fulminant and is the most common cause of deaths associated with this syndrome. Colitis with bloody diarrhea and abdominal pain are associated. In addition to the aromatic antiepileptics (phenobarbital, carbamazepine, and phenytoin), lamotrigine, allopurinol, sulfonamide antibiotics, dapsone and the protease inhibitor abacavir are implicated. Early withdrawal of the offending xenobiotic is crucial and treatment is generally supportive.^42,67 If cardiac or pulmonary involvement is present, systemic corticosteroids should be initiated. However, their benefit on outcome has not been demonstrated, and relapse may occur during tapering, necessitating long-term (several month) courses of therapy. The authors recommend that patients with drug-induced hypersensitivity syndrome be managed in conjunction with a dermatologist and that adequate follow-up is ensured.

Erythroderma

Erythroderma, also known as exfoliative dermatitis, is defined as a generalized redness and scaling of the skin. However, it does not represent one disease entity, but rather it is a severe clinical presentation of a variety of skin diseases including psoriasis, atopic dermatitis, drug reactions, or cutaneous T-cell lymphoma (CTCL). At times, the underlying etiology of erythroderma is never discovered and this is termed “idiopathic erythroderma.” The importance of this presentation is its association with systemic complications such as hypothermia, peripheral edema, and loss of fluid, electrolytes and albumin with subsequent tachycardia and cardiac failure. Many xenobiotics produce erythroderma (Table 17–2). Boric acid, when ingested, can cause systemic toxicity in addition to a bright red eruption (“lobster skin”), followed usually within 1 to 3 days by a generalized exfoliation.⁵⁵

Vasculitis

Xenobiotic-induced vasculitis (Fig. 17–10) comprises 10% to 15% of secondary cutaneous vasculitis. It generally occurs from 7 to 21 days after initial exposure to the xenobiotic or 3 days after rechallenge and is considered to be a secondary cause of cutaneous small vessel vasculitis (typically involving dermal postcapillary venules). Many xenobiotics are implicated as triggers of cutaneous vasculitis (Table 17–2).⁶⁰ Cutaneous vasculitis is characterized by purpuric, nonblanching macules that usually become raised and palpable. The purpura tends to occur predominantly on gravity-dependent areas, including the lower extremities, particularly the feet, ankles, and buttocks (Fig. 17–11). Sometimes the reaction pattern has edematous purpuric wheals (urticarial vasculitis), hemorrhagic bullae, or ulcerations. Histologic examination of affected skin shows a leukocytoclastic vasculitis, which is characterized by fibrin deposition in the vessel walls. There is a perivascular infiltrate with intact and fragmented neutrophils that appear as black dots, known as “nuclear dust” and extravasated red blood cells. Vasculitis can be limited to the skin, or can involve other organ systems, particularly the kidneys, joints, liver, lungs, and brain. The purpura results from the deposition of circulating immune complexes, which form as a result of hypersensitivity to a xenobiotic. Treatment consists of withdrawing the putative xenobiotic and initiating systemic corticosteroid therapy if systemic involvement is present. A syndrome of vasculitis, neutropenia, and retiform purpura occurs as a result of levamisole-adulterated cocaine (Fig. 17–12).¹³ The earlobe is a common site of purpuric lesions from levamisole, and it was estimated that up to 70% of the cocaine sold in some areas during an epidemic in the United States contained levamisole.^6,63,64

Purpura

Purpura is the multifocal extravasation of blood into the skin or mucous membranes (Fig. 17–12). Ecchymoses, therefore, are considered to be purpuric lesions. Chemotherapeutics that either diffusely suppress the bone marrow or specifically depress platelet counts below 30,000/mm³, predispose to purpuric macules. Xenobiotics that interfere with platelet aggregation, such as aspirin, clopidogrel, ticlopidine, valproic acid, and thrombolytics, cause purpura. Anticoagulants, such as heparin and warfarin, may also result in purpura (Chaps. 20 and 58).

Anticoagulant-Induced Skin Necrosis

Skin necrosis from warfarin, low-molecular-weight heparin, or unfractionated heparin usually begins 3 to 5 days after the initiation of treatment, which corresponds with the expected early decline of protein C function with warfarin (Fig. 17–13). The estimated risk is one in 10,000 persons. It is 4 times higher in women, especially if obese, with peaks in sixth to seventh decades of life. The necrosis is secondary to thrombus formation in vessels of the dermis and subcutaneous fat. Heparin-induced cutaneous necrosis results from antibodies that bind to complexes of heparin and platelet factor 4 and induce platelet aggregation and consumption. This causes bullae, ecchymosis, ulcers, and massive subcutaneous necrosis, usually in areas of abundant subcutaneous fat, such as the breasts, buttocks, abdomen, thighs, and calves. Heparin-induced necrosis is associated with protein C or S deficiency, anticardiolipin antibody syndrome, as well as factor V Leiden mutations.⁴⁶

Contact Dermatitis

When a xenobiotic comes in contact with the skin, it can result in either an allergic contact dermatitis (20% of cases) or more commonly an irritant contact dermatitis (80% of cases). Contact dermatitis is characterized by inflammation of the skin with spongiosis (intercellular edema) of the epidermis that results from the interaction of a xenobiotic with the skin. Well-demarcated erythematous vesicular or scaly patches or plaques, and at times bullae, are noted on areas in direct contact with the xenobiotic, whereas the remaining areas are spared.

Allergic contact dermatitis fits into the classic delayed hypersensitivity, or type IV, immunologic reaction. The development of this reaction requires prior sensitization to an allergen, which, in most cases, acts as a hapten by binding with an endogenous molecule that is then presented to an appropriate immunologic T cell. Upon reexposure, the hapten diffuses to the Langerhans cell, is chemically altered, bound to an HLA-DR, and the ­complex is expressed on the Langerhans cell surface. This complex interacts with primed T cells either in the skin or lymph nodes, causing the Langerhans cells to make interleukin-1 and the activated T cells to make interleukin-2 and interferon. This subsequently activates the keratinocytes to produce cytokines and eicosanoids that activate mast cells and macrophages, leading to an inflammatory response (Fig. 17–14).³¹

Many allergens are associated with contact dermatitis (­Table 17–2). Among the most common plant-derived sensitizers are urushiol (Toxicodendron species), sesquiterpene lactone (ragweed), and tuliposide A (tulip bulbs). Metals, particularly nickel, are commonly implicated in ­contact dermatitis and should be considered in patients with erythematous, vesicular, or scaly patches or plaques around the umbilicus from nickel buttons on pants, and on the ear lobes from earrings. Several industrial chemicals, such as the thiurams (rubber) and urea formaldehyde ­resins (plastics), account for the majority of occupational contact dermatitis. Medications, particularly topical medications such as neomycin, commonly cause contact dermatitis. Another important allergen is paraphenylenediamine, a black dye in permanent and semipermanent hair coloring, leather, fur, textiles, industrial rubber products, and black henna tattoos. According to the North American Contact Dermatitis Group, the frequency of sensitization has been found to be 5%.⁶⁹ The management of contact dermatitis varies based on the severity of the reaction, ranging from treatment with topical steroids to oral cyclosporine (Table 17–4). A thorough history in addition to patch testing (the gold standard) will often identify the culprit.

Irritant dermatitis, although clinically indistinguishable from allergic contact dermatitis, results from direct damage to the skin and does not require prior antigen sensitization. Still, the inflammatory response to the initial mild insult is the cause of the majority of the damage. Xenobiotics that cause an irritant dermatitis include acids, bases, solvents, and detergents, many of which, in their concentrated form or after prolonged exposure, can cause direct cellular injury. The specific site of damage varies with the chemical nature of the xenobiotic. Many xenobiotics affect the lipid membrane of the keratinocyte, whereas others diffuse through the membrane, injuring the lysosomes, mitochondria, or nuclear components. When the cell membrane is injured, phospholipases are activated and affect the release of arachidonic acid and the synthesis of eicosanoids. The second-messenger system is then activated, leading to the expression of genes and the synthesis of various cell surface molecules and cytokines. Interleukin-1 is secreted, which can activate T cells directly and indirectly by stimulation of granulocyte-macrophage colony–stimulating factor production. The treatment is similar to allergic contact dermatitis (Table 17–4).

Photosensitivity Reactions

Photosensitivity is caused by topical or systemic xenobiotics. Nonionizing radiation, particularly to ultraviolet A (UVA) (320–400 nm) and less often to ultraviolet B (UVB) (280–320 nm), are the wavelengths that commonly cause photosensitivity. There are generally 2 types of xenobiotic-related photosensitivity: phototoxic and photoallergic.⁴¹ Phototoxic reactions occur within 24 hours of the first exposure, usually within hours, and are dose-related. These reactions result from direct tissue injury caused by ultraviolet-induced activation of a phototoxic xenobiotic. The clinical findings include erythema, edema, and vesicles in a light-exposed distribution, and resemble a severe sunburn that lasts days to weeks, with patients complaining of burning and stinging (Fig. 17–15). A subtype of phototoxic reaction is phytophotodermatitis, in which linear streaks of erythema occur due to skin contact with furocoumarins from plants followed by exposure to sunlight (Table 17–2). Photoallergic reactions occur less commonly, occur following even small exposures, and resemble allergic contact dermatitis with lichenoid papules or an eczematous dermatitis on exposed areas and is often pruritic. These are type IV hypersensitivity reactions that develop in response to a xenobiotic that has been altered by absorption of nonionizing radiation, acting as a hapten and eliciting an immune response on first exposure. Only on recurrent exposure do the lesions develop. Studies indicate that benzophenone-3 (oxybenzone), often found in sunscreen, is the most common cause of photoallergic ­dermatitis.^8,16 Other common photoallergens include xenobiotics such as promethazine, NSAIDs, fragrances, and antibiotics. Photoallergic reactions can be diagnosed by the use of photopatch tests. Both phototoxic and photoallergic reactions are managed with symptomatic treatment, including topical or, if needed, systemic corticosteroids. Identification and avoidance of the triggering xenobiotic are crucial in addition to avoidance of sun exposure and wearing a broad-spectrum sunscreen (SPF 30 or above) that blocks both UVA and UVB preferably without para-aminobenzoic acid (PABA) and oxybenzone. Para-amino benzoic acid is a sensitizer for many patients and is rarely included in current sunscreen products.

Scleroderma-Like Reactions

A number of environmental xenobiotics are associated with localized or diffuse scleroderma-like reactions. Scleroderma refers to a tightened indurated surface change of the skin that typically occurs on the face, hands, forearms, and trunk and is 3 times more common in women. This can be accompanied by facial telangiectasias and Raynaud syndrome. Raynaud syndrome consists of skin color changes of white, blue, and red accompanied by intense pain with exposure to cold, and can cause acral ulcerations, if untreated. The fibrotic process usually does not remit with removal of the external stimulus, and specific autoantibodies are absent. The association of scleroderma-like reactions with polyvinyl chloride manufacture is likely related to exposure to vinyl chloride monomers. Similar reports of this syndrome are associated with exposure to trichloroethylene and perchlorethylene, which are structurally similar to vinyl chloride. Epoxy resins, silica, and organic solvents are implicated as environmental causes. Bleomycin, carbidopa, pentazocine, and taxanes are also causative.

In Spain, patients exposed to imported rapeseed oil mixed with an aniline additive and colorant developed widespread cutaneous sclerosis. This became known as the “toxic oil syndrome.” A similar syndrome following ingestion of contaminated L-tryptophan as a dietary supplement used as a sleeping aid resulted in the eosinophilia myalgia syndrome, which is characterized by myalgia, edema, arthralgias, alopecia, urticaria, mucinous yellow papules, and erythematous plaques.⁵⁶