Salicylate and Other Nonsteroidal Anti-Inflammatory Drug Poisoning

Salicylate and Other Nonsteroidal Anti-Inflammatory Drug Poisoning

Marco L.A. Sivilotti

Christopher H. Linden

Nonsteroidal anti-inflammatory drugs (NSAIDs) include aspirin, related salicylates (Table 131.1), and a variety of other drugs (e.g., ibuprofen, indomethacin, phenylbutazone, and ketorolac), which modulate inflammation by inhibiting cyclooxygenase (COX). In clinical use for 100 years, aspirin still enjoys widespread popularity in the adult population, both by self-medication and by physician-recommended usage.

While the institution of child-resistant packaging and concerns about Reye’s syndrome resulted in a dramatic decline in pediatric overdose, aspirin remains a leading cause of death due to pharmaceutical overdose [1,2,3]. Reducing the amount of aspirin available over the counter was associated with a fewer overdose deaths in the United Kingdom [4]. Nevertheless, vigilance remains necessary because chronic salicylate intoxication, particularly in the elderly, is commonly unrecognized or mistaken for other conditions, such as dehydration, dementia, sepsis, and multiorgan failure. In contrast, most other NSAIDs have a substantially greater safety margin than aspirin in overdose. Although availability without prescription has resulted in increased use and frequency of overdose, significant acute toxicity is uncommon [1,5,6].


All NSAIDs have analgesic and antipyretic as well as anti-inflammatory activity. These effects are due to inhibition of COX, also known as prostaglandin G/H synthase, the enzyme responsible for the conversion of arachidonic acid to prostaglandins and thromboxanes [7,8]. The analgesic dose of most NSAIDs is approximately one-half the anti-inflammatory dose. For some NSAIDs, such as aspirin, ibuprofen, and fenoprofen, this gap is larger, whereas the converse is true for sulindac and piroxicam [9]. Antipyretic effects appear to be due to decreased pyrogen production peripherally as well as to a central hypothalamic effect. The existence of central nervous system (CNS) sites of action mediating analgesic activity has been postulated [10].

Two isoforms of COX have been characterized: COX-1, constitutionally present in platelets, endothelium, gastric mucosa, and the kidneys; and COX-2, induced by a variety of inflammatory mediators (e.g., cytokines, endotoxin, growth factors, hormones, and tumor promoters) but suppressed by glucocorticoids [8,11]. The anti-inflammatory and analgesic properties of NSAIDs appear to be primarily due to the inhibition of COX-2. Their adverse effects on gastric mucosa (e.g., hemorrhage, ulceration, and perforation) and kidney function (e.g., decreased renal blood flow and glomerular filtration rate), and their effects on platelet function appear to be mediated primarily by COX-1, but COX-2 inhibition may also be involved [8,12,13].

NSAIDs can be classified on the basis of their selectivity for COX-2. In particular, the coxibs rofecoxib, valdecoxib, and celecoxib were developed specifically for their COX-2 selectivity and the promise of improved safety. However, an increased risk of thrombotic events, primarily myocardial infarction and stroke, was identified in clinical trials and led to regulatory restrictions on the selective COX-2 inhibitors [14,15]. These adverse cardiovascular effects appear to be due to a relative excess of COX-1–generated thromboxane A2, which is vasoconstrictive and platelet-activating (i.e., prothrombic), and a relative lack of COX-2–generated prostaglandin I2 (prostacyclin), which is vasodilatory and platelet inhibitory (i.e., antithrombotic) [8,14]. It is important to note that traditional NSAIDS diclofenac, meloxicam, and nabumetone exhibit partial COX-2 selectivity, and that other traditional, nonselective NSAIDs may also contribute to adverse cardiovascular events. Thus, selectivity is relative, and all NSAIDs inhibit both COX isoforms in a dose-dependent manner.

Inhibition of COX-1 may result in increased lipoxygenation of arachidonic acid to leukotrienes. This alternate metabolic pathway seems to be responsible for the sometimes fatal allergic reactions to NSAIDs especially prevalent in adults with asthma
and nasal polyps [16,17]. The expression or upregulation of COX-2 may be involved in the pathogenesis of Alzheimer’s disease and some cancers (e.g., colon).

Table 131.1 Salicylate Preparations

Compound Common/trade names Percentage salicylate
Acetylsalicylic acid Aspirin 75
Bismuth subsalicylate In Pepto-Bismol 37
Choline salicylate Arthropan 56
Choline and magnesium salicylate Trilisate 76
Difluorophenyl salicylic acid
Diflunisal Dolobid a
Homomenthyl salicylate In sunscreens 51
Magnesium salicylate Doan’s Caplets, Magan 90
Methyl salicylate Oil of wintergreen 89
Salicylic acid In topical keratolytics 100
Salicylsalicylic acid Salsalate, Disalcid 96
Sodium salicylate Pabalate 84
Trolamine salicylate Aspercreme 48
aNot hydrolyzed to salicylic acid but may cause screening tests for salicylate to be falsely positive.

Aspirin (acetylsalicylic acid) is unique in that it acetylates a serine residue near the active site of COX, thereby irreversibly inhibiting its catalytic function. In contrast, the inhibition of COX by other NSAIDs is reversible and transient. This difference in activity is most notable in platelets, in which thromboxane A2 is essential for normal function [18]. Even in low doses (80 mg), aspirin inhibits platelet aggregation and prolongs the bleeding time for up to 1 week (pending the production of new platelets), whereas other NSAIDs do not have clinically significant platelet effects [19].

In high doses, aspirin and other salicylates also inhibit the hepatic synthesis of clotting factor VII and, to some degree, factors IX and X, thereby prolonging the prothrombin time. This effect appears to be due to interference with the activity of vitamin K and can be reversed by administration of phytonadione (vitamin K1). In contrast, other NSAIDs have insignificant effects on clotting-factor synthesis [19].


Salicylates are available in oral, rectal, and topical formulations. Enteric-coated and sustained-release aspirin tablets are also marketed. Aspirin preparations frequently contain other drugs such as anticholinergics, antihistamines, barbiturates, caffeine, decongestants, muscle relaxants, and opioids. The recommended pediatric dose of aspirin is 10 to 20 mg per kg of body weight every 6 hours, up to 60 mg per kg per day; for adults, the recommended dose is 1,000 mg initially, followed by 650 mg every 4 hours for anti-inflammatory effect. Therapeutic doses of other salicylate salts are similar but depend on their salicylate content (see Table 131.1) and formulation. After a single oral dose of aspirin, therapeutic effects begin within 30 minutes, peak in 1 to 2 hours, and last approximately 4 hours.

Being a weak acid (pKa, 3.5), aspirin is predominantly nonionized at gastric pH and, therefore, theoretically well absorbed in the stomach. However, gastric acidity reduces the solubility of aspirin, thereby slowing the dissolution of tablets. Hence, despite its higher pH, most absorption actually occurs in the small intestine, probably because of its much larger surface area. Peak serum salicylate levels of 10 to 20 mg per dL (0.7 to 1.4 mmol per L) occur 1 to 2 hours after ingestion of a single therapeutic dose. Levels up to 30 mg per dL can occur with long-term therapy and may be necessary for maximal anti-inflammatory effects in some patients. Absorption is delayed or prolonged after ingestion of enteric-coated or sustained-release preparations and suppository use [20]. With overdose, slow pill dissolution, and delayed gastric emptying due to aspirin-induced pylorospasm may lead to absorption continuing for 24 hours or longer after ingestion [21].

During absorption, aspirin is rapidly hydrolyzed by plasma esterases to its active metabolite, salicylic acid. At physiologic pH, salicylic acid (pKa: 3.0) is more than 99.9% ionized to salicylate, which, in contrast to nonionized salicylic acid, diffuses poorly across cell membranes. The drug may become sequestered preferentially in inflamed tissue due to this pH-dependent ionization.

The apparent volume of distribution of salicylate at pH 7.4 is only 0.15 L per kg, in part due to its extensive protein binding. Only free (i.e., unbound) salicylate is pharmacologically active. However, salicylate is unique in that its apparent volume of distribution is not constant. High drug levels (e.g., as a result of chronic therapeutic dosing or acute overdose), low albumin levels, and the presence of other drugs that bind to albumin increase the amount and fraction of free drug [22]. When this occurs, the apparent volume of distribution may increase to 0.60 L per kg [23]. Acidemia, as a consequence of either concomitant illness or severe poisoning, may additionally increase the fraction of nonionized, diffusible drug, promote its tissue penetration, and increase the apparent volume of distribution even more.

After single therapeutic doses, salicylate is metabolized in the liver to the inactive metabolites salicyluric acid (the glycine conjugate; 75% of the dose), salicyl phenolic glucuronide (10%), salicyl acyl glucuronide (5%), and gentisic acid (less than 1%). The remaining 10% of the dose is excreted unchanged in the urine. When serum concentrations exceed 20 mg per dL, the two main pathways of metabolism become saturated, and elimination changes from first order (i.e., proportional to the serum level) to zero order (constant), as described by Michaelis–Menton kinetics. Hence, the apparent half-life of salicylate is 2 to 3 hours after a single therapeutic dose, 6 to 12 hours with chronic therapeutic dosing (i.e., serum levels of 20 to 30 mg per dL), and 20 to 40 hours with overdose (i.e., when levels exceed 30 mg per dL) [24]. Because of saturable metabolism, a small increase in the daily dose can lead to a large increase in serum drug levels, with the potential
for unintentional poisoning [25]. Depletion of glycine stores may reduce the capacity of the salicyluric acid pathway and further slow elimination in overdose [26].

Renal excretion of salicylate becomes the most important route of elimination when hepatic transformation becomes saturated. The rate of excretion is determined by the glomerular filtration, active proximal tubular secretion of salicylate, and passive distal tubular reabsorption of salicylic acid. Alkalinization of the urine decreases the passive reabsorption of salicylic acid by converting it to ionized, nondiffusible salicylate and thereby increases drug excretion. Similarly, increasing the rate of urine flow increases drug clearance by increasing the glomerular filtration and decreasing the distal tubular reabsorption of salicylic acid (by diluting its concentration in the tubular lumen). Combined alkalinization and diuresis can augment the renal elimination of salicylate by 20-fold or more [27,28]. Conversely, dehydration and aciduria perhaps due to preexisting illness or to salicylate poisoning itself decrease salicylate excretion, and increase the duration of toxicity once it develops.

Salicylates readily cross the placenta and enter breast milk. Salicylate elimination in the fetus or infant may be prolonged because of immature metabolic pathways and renal function [29]. It may also be prolonged in patients with liver or renal disease.

The pathophysiology of salicylate poisoning is multifactorial [30,31,32,33,34,35]. Initially and in mild poisoning, direct stimulation of the respiratory center in the medulla by toxic salicylate concentrations results in a respiratory alkalosis, unless blunted by concomitant ingestion of CNS depressants [31]. Direct stimulation of the medullary chemoreceptor zone and irritant effects on the gastrointestinal tract are responsible for nausea and vomiting. Exaggerated antipyretic effects involving the hypothalamus may cause vasodilation and sweating [36]. Dehydration results from gastrointestinal, skin, and insensible fluid losses. The osmotic diuresis that occurs as bicarbonate is excreted in response to alkalemia also contributes to dehydration. Sodium and potassium depletion result from excretion of these electrolytes along with bicarbonate (in exchange for hydrogen ion reabsorption). A functional hypocalcemia (decreased ionized calcium) may accompany alkalemia and cause or contribute to cardiac arrhythmias, tetany, and seizures.

Subsequently, in moderate poisoning, the accumulation of salicylate in cells causes uncoupling of mitochondrial oxidative phosphorylation, inhibition of the Krebs cycle, inhibition of amino acid metabolism, and stimulation of gluconeogenesis, glycolysis, and lipid metabolism [37,38]. These derangements result in increased but ineffective metabolism, with increased glucose, lipid, and oxygen consumption and increased amino acid, carbon dioxide, glucose, ketoacid, lactic acid, and pyruvic acid production. High serum levels of organic acids contribute to an increased anion-gap metabolic acidosis, and the renal excretion of these acids results in aciduria. However, increased carbon dioxide production further stimulates the respiratory center, and the respiratory alkalosis persists, resulting in alkalemia with paradoxical aciduria. An osmotic diuresis further accentuates fluid and electrolyte losses.

In severe poisoning, progressive dehydration and impaired cellular metabolism cause multisystem organ dysfunction. Metabolic acidosis with acidemia becomes the dominant acid–base disturbance. Respiratory acidosis, lactic acidosis, and impaired renal excretion of organic acids due to dehydration and acute tubular necrosis contribute to the acidemia. Acidemia increases the fraction of nonionized salicylate in serum, thereby promoting its tissue penetration and toxicity, and rapid clinical deterioration may ensue with increasing brain salicylate levels. Impaired cellular metabolism can cause increased capillary permeability [39] leading to cerebral edema and noncardiogenic pulmonary edema or acute respiratory distress syndrome. Coma and seizures may result from impaired cellular metabolism, cardiovascular depression, cerebral edema, acidemia, hypoglycemia, and acute white matter damage due to myelin disintegration and activation of glial caspase-3 [41,42]. Respiratory alkalosis may be replaced by respiratory acidosis if coma or seizures cause respiratory depression. Tissue hypoxia resulting from pulmonary edema, impaired perfusion, or seizures may lead to anaerobic metabolism and concomitant lactic acidosis.

Hemorrhagic diathesis may result from increased capillary fragility, decreased platelet adhesiveness, thrombocytopenia, and coagulopathy secondary to liver dysfunction. It occurs primarily in patients with chronic poisoning.

Other Nonsteroidal Anti-inflammatory Drugs

Despite their structural diversity, the pharmacokinetics of traditional NSAIDs are quite similar. Like aspirin, they are weak acids, with pKa ranging from 3.5 to 5.6 and pH-dependent ionization being the major determinant of tissue distribution and sequestration. They are rapidly absorbed after ingestion, have small volumes of distribution (0.08 to 0.20 L per kg), and are 90% to 99% protein bound (principally to albumin). Most have half-lives of less than 8 hours, with low non–flow-dependent hepatic clearance, primarily by the CYP2C subfamily of cytochrome P450 enzymes, to inactive metabolites that are then conjugated, mostly with glucuronic acid, and excreted in the urine. Sulindac is one exception in that its sulfide metabolite is the active form of the drug and has a half-life of 16 hours [42]. Nabumetone is also a prodrug, and its active metabolite, 6-methoxy-2-naphthylacetic acid, has a half-life of more than 20 hours (and even longer in the elderly) [43]. Phenylbutazone, oxyphenbutazone, and piroxicam are notable for half-lives of longer than 30 hours. Diflunisal, like aspirin, has a dose-dependent half-life of 5 to 20 hours. Indomethacin, sulindac, etodolac, piroxicam, carprofen, and meloxicam undergo enterohepatic recirculation [42,44,45]. Small amounts (less than 10%) of nonsalicylate NSAIDs are excreted unchanged in the urine, limiting the effect of urine pH on clearance. The coxibs are nonacidic drugs [13], highly protein bound and primarily metabolized in the liver.

In contrast to salicylates, the metabolism of most nonsalicylate NSAIDs is not saturable or prolonged in overdose, and elimination follows first-order kinetics. An exception is phenylbutazone, whose elimination may follow Michaelis–Menton kinetics.

Toxic effects of NSAIDs appear to be primarily due to exaggerated pharmacologic effects, with gastric irritation and renal dysfunction resulting from the inhibition of prostaglandin synthesis [46]. In contrast to salicylate poisoning, the acidosis that sometimes occurs with large overdoses of these agents appears to be due to high levels of parent drug and metabolites rather than to disruption of metabolism [47]. Mechanisms responsible for their CNS toxicity remain to be defined.

Clinical Toxicity


Salicylate poisoning may occur with acute as well as chronic overdose [30,31,32,33,34,35,48,49,50,51,52,53,54,55]. It most commonly results from ingestion, but poisoning due to topical use [56] and rectal self-administration [57] has been reported. The ingestion of topical preparations of methyl salicylate (oil of wintergreen, also present in Chinese propriety medicines) can result in
rapid-onset poisoning, due to its concentration, rapid absorption kinetics, and higher lipid solubility [58]. Infants may become poisoned by ingesting the breast milk of women chronically taking therapeutic doses of salicylate [59]. Intrauterine fetal demise resulting from poisoning during pregnancy [60] and neonatal poisoning resulting from the transplacental diffusion of therapeutic doses of salicylate taken before delivery [61] have also been described. Delays to presentation, diagnosis, and chronicity each increase the severity and mortality [50,62,63] and with severe poisoning, the fatality rate may be as high as 50% [51,55].

Table 131.2 Severity of Salicylate Poisoning

Severity grade Serum pH Underlying acid–base abnormality
Mild > 7.45 Respiratory alkalosis
Moderate 7.35–7.45 Combined respiratory alkalosis and metabolic acidosis
Severe < 7.35 Metabolic acidosis with or without respiratory acidosis

Regardless of whether poisoning is acute or chronic, it can be characterized as mild, moderate, or severe on the basis of the serum pH and underlying acid–base disturbance (Table 131.2). This approach was first described in the classic papers by Done [48,64], who also developed a nomogram that attempted to correlate the severity of poisoning with a timed salicylate level after acute ingestion. Although Done’s nomogram has subsequently been shown to have poor predictive value in acute poisoning [49] and is not applicable to chronic poisoning, to acute poisoning by enteric-coated aspirin and nonaspirin salicylates, or to patients with acidemia [62] his observation that the clinical severity of poisoning correlates with acid–base status remains undisputed.

Mild poisoning is characterized by alkalemia (serum pH greater than 7.45) and a pure respiratory alkalosis. It may develop 2 to 8 hours after acute ingestion of 150 to 300 mg per kg of aspirin [48,64] or any time during chronic therapy. Associated signs and symptoms include nausea, vomiting, abdominal pain, headache, tinnitus, tachypnea (or subtle hyperpnea), ataxia, dizziness, agitation, and lethargy. The anion gap (see Chapter 71) is normal until late in this stage, when compensatory renal bicarbonate excretion eventually lowers the serum bicarbonate level. Serum glucose, potassium, and sodium values may be high, low, or normal. Despite total body fluid and electrolyte depletion and clinical dehydration, laboratory evidence of dehydration (e.g., hemoconcentration, increased serum blood urea nitrogen [BUN] and creatinine, increased urine specific gravity) may be absent.

Moderate poisoning is characterized by a near normal serum pH (7.35 to 7.45) with an underlying metabolic acidosis as well as respiratory alkalosis. It can occur 4 to 12 hours after an acute overdose of 300 to 500 mg per kg of aspirin [48,64]. It may also occur in patients with chronic ingestion who delay seeking medical care for symptoms of mild poisoning and continue to take salicylate. Electrolyte analysis demonstrates a low serum bicarbonate value with an increased anion gap. Gastrointestinal and neurologic symptoms are more pronounced. There may be agitation, fever, asterixis, diaphoresis, deafness, pallor, confusion, slurred speech, disorientation, hallucinations, tachycardia, tachypnea, and orthostatic hypotension. Coma and seizures can also occur. Leukocytosis, thrombocytopenia, increased or decreased serum glucose and sodium values, hypokalemia, and increased serum BUN, creatinine, and ketones may be present.

Severe poisoning is defined by the presence of acidemia (serum pH less than 7.35) with underlying metabolic acidosis and respiratory alkalosis or acidosis and a high anion gap. It can occur 6 to 24 hours or more after the acute ingestion of more than 500 mg per kg of aspirin [48,64] or in unrecognized or untreated chronic poisoning. Severe dehydration and marked sinus tachycardia are often present. Other findings may include coma, seizures, papilledema, hypotension, dysrhythmias, congestive heart failure, oliguria, hypothermia or hyperthermia, rhabdomyolysis and multiple organ failure [55,63,65,66]. Laboratory abnormalities are similar to those seen in moderate poisoning but are more pronounced. Hypoglycemia, pulmonary edema and cerebral edema or hemorrhage may be present [54,67]. Asystole is the most common terminal dysrhythmia, but ventricular tachycardia and ventricular fibrillation can also occur [54,55,68,69]. When cardiac arrest occurs, death appears to be inevitable. Successful resuscitation in this situation has yet to be reported [69].

Only gold members can continue reading. Log In or Register to continue

Sep 5, 2016 | Posted by in CRITICAL CARE | Comments Off on Salicylate and Other Nonsteroidal Anti-Inflammatory Drug Poisoning
Premium Wordpress Themes by UFO Themes