Ethylene glycol poisoning: toxicokinetic and analytical factors affecting laboratory diagnosis.
In cases of severe ethylene glycol poisoning, early diagnosis and aggressive therapeutic intervention are essential for a favorable clinical outcome. Although therapy may be initiated before the serum concentration of ethylene glycol is measured, rapid and accurate identification of ethylene glycol or its metabolites is necessary for definitive diagnosis. Conversely, reliable screening tests for ethylene glycol poisoning can eliminate this diagnosis from the list of possible causes of increased anion gap metabolic acidosis in a comatose patient. Rapid exclusion of the diagnosis of ethylene glycol poisoning prevents a patient from receiving invasive, inappropriate treatment.
This Case Conference details the challenges that clinicians and clinical laboratorians confront in diagnosing ethylene glycol intoxication. The patients in cases 1 and 2 related a clear history of ethylene glycol ingestion and manifested several of the classical signs and symptoms of poisoning. Prompt initiation of appropriate therapy prevented neurological, cardiac, and renal sequelae. The patient in case 3 presented with a confusing picture of unexplained metabolic acidosis and renal failure. He repeatedly denied ethylene glycol ingestion, and results of serum ethylene glycol measurements were repeatedly negative; ethylene glycol poisoning was diagnosed only after renal biopsy. This case illustrates the impact of toxicokinetic variables on ethylene glycol detection and other laboratory parameters. The patient in case 4 also denied ethylene glycol ingestion, but presented with signs and symptoms consistent with ethylene glycol poisoning, including a positive laboratory test result for serum ethylene glycol; however, this result proved to be the result of interference of serum components in the ethylene glycol enzymatic assay. This case introduces the effect of analytical variables on the laboratory diagnosis of ethylene glycol poisoning.
A 27-year-old man with a history of ethanol and cocaine abuse presented to the emergency room claiming to have ingested 2.25 L (0.5 gal) of Prestone antifreeze ~3 h earlier in a suicide attempt. He had also consumed an unreported amount of cocaine (crack) and cut his left forearm and neck. He was somnolent with a depressed affect but responded to questions appropriately. Physical examination revealed stable vital signs and superficial lacerations of his neck and left forearm. Laboratory values on admission are given in Table 1. A toxicology screen revealed cocaine, lidocaine, and nicotine in his urine. Based on the history and physical examination, a diagnosis of ethylene glycol poisoning was made. The serum ethylene glycol concentration, determined by gas chromatography, was 995 mg/L (16 mmol/L) and confirmed the diagnosis. Treatment included gastric lavage, charcoal, and intravenous ethanol. The serum ethanol concentration was maintained between 17 and 22 mmol/L (76-103 mg/dL). The serum ethylene glycol concentration decreased progressively from 995 mg/L to below the limit of detection (<100 mg/L) after 48 h. Serum creatinine concentrations were within the reference range throughout the hospitalization.
A 58-year-old man presented to the emergency room after a suicide attempt. He reported ingesting about "20 ounces [~56 mL] of half-strength" antifreeze, cutting his wrists with a dull knife, and falling down a flight of stairs. One empty and one half-full 1-gal (4.5-L) containers of Prestone antifreeze, reportedly half-strength, were found near the patient. In the emergency room, the patient was confused but otherwise neurologically intact. Physical examination revealed stable vital signs and superficial lacerations of both wrists. Laboratory values on admission are given in Table 1 and include a metabolic acidosis with increased anion gap and an increased osmolal gap. A diagnosis of ethylene glycol poisoning was made on the basis of history, physical examination, and confirmatory laboratory testing. The initial toxicology screen revealed ethylene glycol, caffeine, and nicotine. The serum ethylene glycol concentration on admission, 7910 mg/L (127 mmol/L), was determined by an enzymatic reaction that utilizes glycerol dehydrogenase from Enterobacter aerogenes (Boehringer Mannheim Biochemicals) on the Hitachi 704 automated analyzer. Treatment included charcoal, intravenous ethanol infusion (1 g/kg loading dose, 100 mg/kg per hour maintenance dose), elective intubation for airway protection, and emergency hemodialysis. The serum ethanol concentration was maintained between 17 and 42 mmol/L (77-194 mg/dL). The serum ethylene glycol concentration progressively decreased from 7910 mg/L (127 mmol/L) to 150 mg/L (2 mmol/L) after 28 h of therapy (including 16 h of hemodialysis). Serum creatinine concentrations were within the reference range, and arterial pH was never <7.21 throughout the hospital stay. The patient recovered completely within 3-5 days without discernible neurological, renal, or cardiac sequelae.
An 18-year-old man with a history of asthma presented to the emergency room complaining of nausea, vomiting, diffuse abdominal pain, and malaise over the previous 2 days. His vital signs were blood pressure 150/100, pulse 84, and rectal temperature 37.9 [degrees]C (100.7 [degrees]F). He was lethargic but arousable and oriented to person, time, and place. Medications included theophylline (Theodur) and epinephrine (Primatene Mist; 5.5 g/L). Laboratory values on admission are given in Table 1. Results of the urine and serum toxicology screens were negative. The patient was admitted for metabolic acidosis and renal failure. Renal function deteriorated and serum creatinine peaked at 1052 [micro]mol/L (11.9 mg/dL) on the fifth hospital day. Ultrasound revealed enlarged edematous echogenic kidneys, compatible with acute renal failure. Results of an arteriogram for renal vasculitis and a skin biopsy for HenochSchonlein purpura were negative. Renal biopsy revealed acute tubular necrosis and deposition of calcium oxalate crystals, supporting a diagnosis of ethylene glycol poisoning (Fig. 1). The patient denied ethylene glycol or antifreeze ingestion and suicide attempts. Ethylene glycol was not detected in stored serum samples from admission or taken on the fifth hospital day. Treatment with hemodialysis for 3 weeks resulted in complete recovery of renal function. Twenty-four months after this admission, the patient demonstrated no clinical or laboratory evidence of renal dysfunction [serum creatinine, 97 [micro]mol/L (1.1 mg/ dL); blood urea nitrogen, 4.3 mmol/L (12 mg/dL)].
The patient in case 4 was reported previously . Briefly, he was a 23-year-old man with a 1-year history of dilated cardiomyopathy and pulmonary hypertension who experienced shortness of breath, productive cough, hemoptysis, nausea, and vomiting over the 2 days before admission. Medications included captopril, digoxin, furosemide, and cimetidine. On physical examination, he was jaundiced, in mild respiratory distress, febrile [37.8 [degrees]C (100 [degrees]F)], tachycardic (126/min), and tachypneic (26/min). Results of physical examination were remarkable for bilateral pulmonary rales and mild hepatomegaly. On the first hospital day, he became increasingly anxious, restless, and tachypneic with subxiphoid pain radiating to his back. Pertinent laboratory values are listed in Table 1. Urinalysis revealed proteinuria (1 g/L), hematuria (packed), and "unidentified" crystals, which were retrospectively interpreted as being urates. Serum ethylene glycol was 90 mg/L (1.5 mmol/L), as measured by the enzymatic reaction described above, utilizing glycerol dehydrogenase and the Hitachi 704 analyzer. However, ethylene glycol was not detected by gas chromatography.
[FIGURE 1 OMITTED]
These cases expose the challenges encountered during the evaluation of ethylene glycol intoxication. The history obtained from patients regarding their current illness is often the most valuable diagnostic tool. In cases 1 and 2, the intoxication was recognized immediately because the patients clearly related the details of their suicide attempts. In contrast, the diagnosis was delayed in case 3 because the patient did not report ethylene glycol ingestion. Although family members suspected he had attempted suicide, he repeatedly denied intentional ethylene glycol consumption during the course of the hospitalization and several months later during an emergency room visit for an unrelated problem. No evidence of intentional or accidental poisoning was discovered. The diagnosis of ethylene glycol poisoning was not established until a renal biopsy was performed to investigate the cause of acute renal failure.
Denial of ethylene glycol ingestion occurs because the patient or a third party either is concealing intentional poisoning or is unaware of consumption. In the US, a child undergoing an extensive medical evaluation for a possible inherited genetic disease had, in fact, been poisoned with ethylene glycol by a care giver . In Germany, a man unknowingly drank contaminated water from a radiator for 6 weeks and consumed enough ethylene glycol to result in intoxication and renal failure . More commonly, a patient is unable to relate his history because of profound neurological depression.
In the absence of an accurate history, the alternative metabolic derangements considered for a comatose patient include diabetic ketoacidosis, alcoholic ketoacidosis, renal failure, and ingestion of methanol or other toxic compounds. In these cases, additional laboratory information may facilitate an accurate diagnosis.
TEMPORAL ASPECTS OF CLINICAL PRESENTATION AND PATIENT VARIABILITY
The clinical presentation of ethylene glycol poisoning is classically described in three stages. Neurological manifestations are apparent within 0.5-12 h of ingestion and include inebriation in the absence of detectable ethanol on breath or in blood, nausea, vomiting, nystagmus, papilledema, depressed reflexes, convulsions, and coma. Cardiopulmonary manifestations may be observed 12-24 h after ingestion and may include tachypnea, tachycardia, hypertension, pulmonary edema, and congestive heart failure. Renal complications are generally a late feature, occurring 24-72 h after ingestion, and consist of flank pain, costovertebral angle tenderness, oliguria, and renal failure.
The severity of each stage and the progression from one stage to the next depend on the amount of ethylene glycol ingested as well as the timing of medical intervention. The most serious clinical features observed in ethylene glycol poisoning are due not to the parent compound but to the metabolites. Ethylene glycol per se causes only minimal inebriation, which resembles ethanol intoxication [1,2]. Because metabolism of the parent compound is required for toxicity, the latent period reflects the time required for the toxic metabolites to accumulate. In cases 1 and 2, the only apparent manifestation of ethylene glycol poisoning was mild neurological depression, because the patients were evaluated within hours of the ingestion. In striking contrast, substantial renal insufficiency caused by the metabolites of ethylene glycol was manifest in case 3, because the poisoning was not detected until late in the clinical course, most likely several days after ingestion.
Although these stages provide a useful theoretical framework for the clinical description of ethylene glycol poisoning, the onset and progression of this condition are not always straightforward or predictable. These stages may be confluent, one stage may predominate, or one or more of the stages may not be clinically apparent. Often, a patient is discovered comatose, experiencing both respiratory distress and acute renal insufficiency . Variable symptom profiles are found in the literature of case reports [8-11]. Renal failure is the most frequently reported manifestation of ethylene glycol ingestion; however, as in cases 1 and 2, prompt treatment may prevent crystalluria and renal insufficiency [12-14].
Although the symptoms usually completely resolve soon after appropriate treatment for the acute episode, sequelae of ethylene glycol poisoning have included prolonged renal failure requiring dialysis for months, residual kidney damage, and cranial nerve deficits manifesting late in the clinical course and lasting as long as several months [15-18]. Complete recovery from ethylene glycol poisoning with aggressive treatment, however, has been reported even after severe encephalopathy  or profound acidemia with a serum pH of 6.46 
TOXICOKINETIC EVALUATION OF ETHYLENE GLYCOL ELIMINATION
Because of the potential severity of ethylene glycol poisoning, kinetic variables for ethylene glycol disposition in humans have been measured only in poisoned individuals (Table 2). Ethylene glycol is rapidly absorbed from the gastrointestinal tract, and symptoms of poisoning may be experienced within 30 min of ingestion. Percutaneous absorption of ethylene glycol has not been reported, but topical burn preparations containing propylene glycol or diethylene glycol have produced considerable toxicity in burn patients [20,21]. Ethylene glycol is rapidly metabolized by alcohol dehydrogenase and other hepatic enzymes to glycoaldehyde and organic acids (Fig. 2). The elimination half-life of ethylene glycol is increased at least fivefold in the presence of ethanol because both compounds compete for the active site of alcohol dehydrogenase. This enzyme has a much greater affinity for ethanol than for ethylene glycol or methanol, and concentrations of ethanol as low as 11 mmol/L (50 mg/dL) saturate the enzyme . To ensure competitive inhibition of alcohol dehydrogenase activity, however, the serum concentration of ethanol should be maintained >22 mmol/L (>100 mg/dL) during hemodialysis [22,23]. Hemodialysis rapidly removes both ethylene glycol and its toxic metabolites, particularly glycolate and oxalate [24-27]. Hemodialysis clearance is greater than renal clearance, but functioning kidneys contribute to the removal of ethylene glycol from the blood .
Toxicokinetic evaluation of ethylene glycol elimination during hemodialysis and ethanol infusion was performed in case 2 (Fig. 3). From the estimation of the dose provided by the patient and the assumption that the peak concentration was 7910 mg/L (127 mmol/L), the volume of distribution ([V.sub.d]) was calculated as 0.5 L/kg, a value in good agreement with previously reported values for VD (Table 2). Utilizing the noncompartmental pharmacokinetic model with extravascular administration of the historical dose (Wirmonlin software package; Scientific Consulting Inc., Apex, NC), we determined the elimination half-life as 3.79 h, based on linear regression of 14 terminal arterial concentration values. The calculated half-life largely reflects the half-life during dialysis in the presence of a serum ethanol concentration of 17-42 mmol/L (77-194 mg/dL). During the two short intervals when ethanol infusion was maintained but hemodialysis was not performed (Fig. 3), the slope of the line is more shallow, reflecting the prolongation of the half-life of ethylene glycol in the presence of ethanol (Table 2). The overall clearance was 75 mL/min, beginning from the time the patient entered the emergency room. The overall clearance is the combined effect of the dialyzer clearance and the body's own clearance (through kidney excretion and liver metabolism) until saturating concentrations of ethanol are reached. If the first 6 h before dialysis are not included in the calculation, the clearance is 113 mL/min. We also calculated the removal of ethylene glycol by the dialyzer, referred to in some studies as dialyzer clearance, by using the formula [Cl.sub.H], = [[Q.sub.IN] ([CI.sub.N - C.sub.OUT])] / [C.sub.IN], where [Q.sub.IN] is blood flow into the dialyzer (measured with a flow sensor), and [C.sub.IN] and [C.sub.OUT] are the ethylene glycol concentrations measured in the arterial blood entering and the venous blood leaving the dialyzer, respectively. For a dialysis blood flow of 350 mL/min, dialyzer clearance in case 2 was calculated as 228 mL/min, a value slightly higher than reported previously, possibly reflecting the higher blood flow rate used for this patient [14,18,23,26].
TEMPORAL ASPECTS OF LABORATORY DIAGNOSIS
The evolving laboratory profile in cases of ethylene glycol poisoning reflects the metabolism of ethylene glycol, the accumulation of organic acids, and the timing of medical intervention. In case 2, the initial serum ethylene glycol concentration was 7910 mg/L, among the highest reported values for an intoxicated patient who survived with therapy. Death, however, has been reported in patients with virtually undetectable serum ethylene glycol concentrations . This discrepancy underscores the importance of prompt diagnosis and treatment of intoxication. The poor correlation between serum ethylene glycol concentration and clinical outcome is the result of the rapid clearance of the parent drug and conversion to the severely toxic metabolites. The "window" period during which the parent compound ethylene glycol can be detected may be relatively narrow. In case 3, this time had elapsed, and the only evidence of ethylene glycol poisoning was the deposition of calcium oxalate crystals in the kidney. At a given time after ingestion, the concentrations of the remaining ethylene glycol and the accumulated acidic metabolites affect the magnitude of the osmolal gap and anion gap, respectively. An osmolal gap may not be apparent late in the course of poisoning, whereas an anion gap may not be evident early in the clinical course [18,29,30]. These observations reflect the relative amounts of the compounds that are the source of the gaps.
The anion gap is the difference between the sum of the measured cations and the sum of the measured anions [([Na.sup.+] + [K.sup.+]) - (HC[O.sup.-.sub.3] + [Cl.sup.-])]. Most laboratories routinely measure only sodium and potassium, which account for ~95% of the extracellular cations, and chloride and bicarbonate, which account for ~85% of the extracellular anions. Because the sum of the measured cations does not equal the sum of the measured anions in healthy individuals, the "normal" anion gap is 12-16 mmol/L. The production of unmeasured organic acids will increase the anion gap. Glycolic acid accounts for as much as 96% of the anionic gap in patients poisoned with ethylene glycol [24,25]. The osmolal gap is also an estimate of unmeasured constituents in the serum. In healthy individuals, serum osmolality is determined by the concentration of sodium, urea nitrogen, and glucose and is approximated by the following formula :
Calculated plasma osmolality, mOsm/kg [H.sub.2]O =
[1.86 X sodium (mmol/L) + urea nitrogen (mmol/L) + glucose (mmol/L)]/0.93
or, in traditional units, 11.86 X sodium (mmol/L) + [urea nitrogen (mg/dL)/2.8] + [glucose (mg/dL)/18]}/0.93
Serum osmolality is 270-290 mOsm/kg [H.sub.2O] in a healthy individual. The difference between measured and calculated osmolality, the osmolal gap, results from the presence of other solutes in serum, which are not considered in the above formula. An increase in the osmolal gap, generally considered important when >10 or 15 mOsm/kg [H.sub.2]O, suggests the presence of low-molecularmass substances that achieve appreciable serum concentration such as ethylene glycol, methanol, ethanol, and acetone [31-33]. Consequently, osmolality should be measured by freezing point depression, because the vapor pressure method will underestimate volatile alcohols .
SENSITIVITY OF OSMOLAL AND ANION GAPS IN ETHYLENE GLYCOL POISONING
As ethylene glycol is being metabolized or removed by dialysis, its contribution to the osmolal gap diminishes because the accumulating acidic, negatively charged metabolites do not contribute to the osmolal gap . These anions are counterbalanced by sodium and are taken into consideration in the formula for calculating the serum osmolality . Nonionic metabolites of ethylene glycol may contribute to the osmolal gap; however, the concentration of the parent compound accounts for most of the osmolal gap in cases of ethylene glycol poisoning .
The contribution of ethylene glycol or ethanol to the osmolal gap can also be calculated: each 16 mmol/L (100 mg/dL) increment in ethylene glycol concentration contributes ~16 mOsm/kg [H.sub.2]O, and each 22 mmol/L (100 mg/dL) of ethanol contributes 22 mOsm/kg [H.sub.2]O to the osmolal gap. The corrected osmolal gap, or the residual osmolality, can be used to monitor patients who have simultaneously ingested ethylene glycol and ethanol or patients who are receiving ethanol infusion for ethylene glycol poisoning . In case 1, the serum osmolal gap of 16 mOsm/kg was probably entirely attributable to ethylene glycol because the initial serum concentration of ethylene glycol, 995 mg/L, would contribute ~16 mOsm/kg [H.sub.2]O to the osmolal gap.
Although metabolism of ethylene glycol diminishes the osmolal gap, the generation of unmeasured acidic metabolites of ethylene glycol augments the anion gap [18,30]. In two series of individuals intoxicated with ethylene glycol, metabolic acidosis was initially apparent in 50% and 86% of cases [35,36]. In case 3, the increased anion gap most probably reflected the delay in diagnosis and the accumulation of acidic metabolites. In case 1, however, the anion gap was within the reference range, most probably because the intoxication was detected and treated promptly; acidic metabolites had not reached appreciable concentration. Aside from the temporal dependence of the increased anion gap on the metabolism of the parent compound, an anion gap may not be present in cases of ethylene glycol poisoning for other reasons. Simultaneous ingestion of ethanol will competitively inhibit the metabolism of ethylene glycol and delay the appearance of an anion gap . Simultaneous ingestion of bromide masks the anion gap because bromide is not distinguished from chloride in some assays . Finally, simultaneous ingestion of lithium carbonate may conceal the anion gap by providing additional bicarbonate .
SPECIFICITY OF ANION AND OSMOLAL GAPS
For these reasons, neither osmolal nor anion gaps are universally present in cases of ethylene glycol poisoning, and their absence cannot be used to rule out toxic alcohol or ethylene glycol ingestion [29,33]. Conversely, the simultaneous presence of metabolic acidosis with an increased anion and osmolal gap, although highly suggestive of ethylene glycol or methanol poisoning, is not specific for these intoxications [40,41]. Anion and osmolal gaps may be present in other clinical settings such as diabetic ketoacidosis, alcoholic ketoacidosis, chronic renal failure, multiple organ failure, and critical illness [40,42-48]. In diabetic ketoacidosis, for example, the osmolal gap is due primarily to the accumulation of acetone, which may approach concentrations of 13 mmol/L, and the anion gap metabolic acidosis is due primarily to the accumulation of acetoacetate and [beta]-hydroxybutyrate [43,44]. Although the anion gap may be profoundly increased in diabetic ketoacidosis, the osmolal gap is not usually greater than 20-25 mOsm/kg [H.sub.2]O [40,44]. In case 4, both metabolic acidosis with an increased anion gap and an increased osmolal gap were present. Ethylene glycol poisoning was initially considered but later discounted after gas chromatography failed to confirm the results of the screening test for ethylene glycol. The increased anion and osmolal gaps in this case were probably the result of multiple organ failure.
SENSITIVITY AND SPECIFICITY OF CALCIUM OXALOSIS
Calcium oxalate crystalluria and deposition of these crystals in the kidney, brain, or other organs are distinctive laboratory features in ethylene glycol poisoning. In case 3, the findings of renal tubular necrosis and deposition of calcium oxalate crystals by renal biopsy (Fig. 1) supported the diagnosis. Chelation of calcium by the oxalate deposited in the kidneys and other organs may explain the hypocalcemia that is often observed in cases of ethylene glycol poisoning. Another clinical setting characterized by calcium oxalate nephrolithiasis is primary hyperoxaluria, a rare inherited metabolic disorder associated with early-onset renal failure and death. Healthy individuals, especially those with dietary excesses of foods rich in oxalate such as tomatoes, garlic, spinach, rhubarb, cocoa, and tea, may also exhibit calcium oxalate crystalluria without associated renal insufficiency.
Calcium oxalate crystals in urine are pleomorphic, variegated, and birefringent when viewed through polarized light. In cases of ethylene glycol poisoning, calcium oxalate may be excreted not only as dihydrate crystals, which are envelope-shaped (dipyramidal, octahedral), but also as monohydrate crystals, which are needle-shaped (spindle or prism shaped) (Fig. 4). Other forms of calcium oxalate include dumbbell, ovoid, and elliptical crystals. If the results of more-definitive laboratory tests are not available, the detection of calcium oxalate crystalluria, particularly the monohydrate form, provides supportive evidence for the diagnosis of ethylene glycol poisoning [30,49-51]. Because the monohydrate form may be the only form seen early or at any time during the course of the episode, familiarity with the microscopic features of calcium oxalate monohydrate crystals is important.
[FIGURE 4 OMITTED]
In the early 1980s, identification of the pleomorphic nature of calcium oxalate crystalluria prompted the recommendation that physicians recognize the monohydrate calcium oxalate crystals to facilitate rapid diagnosis of ethylene glycol ingestion . In the 1990s, however, the monohydrate form was still considered "unusual" or was likely to be misidentified as hippuric acid crystals in cases of ethylene glycol poisoning . The medical literature has been confused by morphological descriptions of crystals that resemble those of hippuric acid as well as by theoretical arguments supporting their formation [25,30]. X-ray diffraction, however, definitively identifies the needle-shaped crystals as calcium oxalate monohydrate and not hippuric acid [30,50,51]. Renewed emphasis has recently been placed on the need to increase proficiency during microscopic analysis of urine to recognize calcium oxalate monohydrate as well as dihydrate crystals in cases of suspected ethylene glycol poisoning . Because urinalysis is rapid and easy, repeated urine microscopy is a potentially useful adjunct in the differential diagnosis of an anion gap metabolic acidosis of unknown origin [18,50]. Other frequently reported findings on urinalysis in cases of ethylene glycol poisoning include low specific gravity, proteinuria, and microscopic hematuria.
LABORATORY MEASUREMENT OF ETHYLENE GLYCOL METABOLITES
Examination of urine for calcium oxalate crystals is the most widely accessible laboratory technique for detecting a metabolite of ethylene glycol. Because the toxicity of ethylene glycol depends on its oxidation to organic acids, several other methods have been utilized to measure these products. Glycolic acid, the predominant metabolite, has been measured by HPLC and gas chromatography, by a colorimetric method, and by isotachophoresis [27,52-54]. The colorimetric method utilizes sulfuric acid and chromotropic acid and is relatively specific for glycolic acid . In contrast, isotachophoresis can measure the four major acidic metabolites of ethylene glycol simultaneously . Isotachophoresis is an electrophoretic technique that orders and concentrates substances of intermediate effective mobilities between an ion of high effective mobility and one of much lower effective mobility. Sample components ultimately separate into adjacent zones that migrate at the same velocity. Although a laboratory test for glycolic acid would have been a useful diagnostic adjunct in case 3, none is currently available at most medical centers or reference laboratories.
ANALYTICAL VARIABLES AFFECTING LABORATORY DIAGNOSIS
Gas chromatography. The method of choice for measuring ethylene glycol is gas chromatography, with flame ionization detection of ethylene glycol itself or of a derivative. Commonly, ethylene glycol is analyzed as the boronic ester derivative by using packed or capillary columns. Underivatized ethylene glycol is difficult to analyze because of poor chromatographic behavior and the poor detection limit of flame ionization detectors. However, direct injection of ethylene glycol on a wide-bore capillary column has been described [55,56] Advantages of these methods include elimination of the derivatization step, resolution of diethylene glycol and other diols as well as other polar drugs, and extended analytical life of the Nukol column.
An appropriate internal standard, such as 1,3-propanediol or 1,2-butanediol, must be used with gas-chromatographic analysis. Recently, the use of propylene glycol as an internal standard has been discouraged [57-59]. The inclusion of propylene glycol in some intravenous pharmaceutical preparations contributes to the concentration of the internal standard, thus resulting in underestimation of the ethylene glycol concentration in the serum sample. Moreover, propylene glycol itself may be responsible for clinical toxicity, and its presence in a serum sample may be masked when this compound is used as an internal standard.
Although gas chromatography is the "gold standard" for detecting ethylene glycol, diagnostic inaccuracy with this method has occurred. In one notorious case, propionic acid was mistakenly identified as ethylene glycol by gas chromatography . Consequently, a mother was falsely accused of poisoning her infant son who, in fact, had an inherited metabolic disease, methylmalonic acidemia. This error occurred in two independent laboratories. Accurate interpretation of the retention times of the compounds identified in the serum sample might have prevented the tragic consequences of this inaccurate diagnosis. Identification based on retention time alone has led to confusion of 2,3-butanediol as well as propionic acid with ethylene glycol [60,61]; thus, if the presence of ethylene glycol is suspected in a sample, confirmation by mass spectrometry is recommended. Moreover, methanol-like products generated by the oxidation-reduction derivatization procedure in sera from ketoacidotic diabetic patients may be misinterpreted as evidence of ethylene glycol poisoning .
Enzymatic assay. The screening assay utilizes glycerol dehydrogenase purified from E. aerogenes [63-65]. This enzyme catalyzes the oxidation of ethylene glycol, producing NADH, which is measured spectrophotometrically. No cross-reaction is observed with ethanol, methanol, n-propanol, isopropanol, acetaldehyde, lactate, glyoxal, glycolic acid, glyoxylic acid, or oxalic acid. Interfering substances include glycoaldehyde and glycerol, which compete with ethylene glycol for the active site of the enzyme. Although glycoaldehyde is one of the ethylene glycol metabolites, its short half-life makes it unlikely to interfere with the assay. The interference from glycerol is not significant in most cases and may be disregarded on the basis of expected serum concentrations of glycerol (<0.5 mmol/L). However, critically ill patients may have an increased serum concentration of free glycerol, most frequently related to intravenous infusion of glycerolcontaining medications . Conceivably, glycerol could reach a concentration in serum sufficient to interfere in the enzymatic assay for ethylene glycol in this clinical setting.
Another interference in the screening test for ethylene glycol was delineated in case 4 . Ethylene glycol poisoning was considered in this patient because the toxicology screen detected ethylene glycol and several of the key diagnostic features were present: cardiorespiratory compromise, increased anion gap metabolic acidosis, increased osmolal gap, renal insufficiency, and crystalluria. The positive result obtained in the enzymatic screening test, however, was not confirmed by gas chromatography. The source of the interference in the enzymatic assay for ethylene glycol in this case was markedly increased concentrations of serum L-lactate dehydrogenase (LD) and lactic acid. In patients with increased serum lactate and LD, extraneous production of NADH from the oxidation of lactate to pyruvate catalyzed by LD may interfere with the assay, resulting in falsely positive values for ethylene glycol. This interference has also been reported in the enzymatic assay for ethanol, which likewise measures NADH as its endpoint .
TREATMENT OF ETHYLENE GLYCOL POISONING
Goals of treatment in cases of ethylene glycol intoxication include reducing the load of ingested ethylene glycol, correcting the metabolic acidosis of early toxicity, preventing additional metabolism of ethylene glycol, and removing the parent compound as well as toxic metabolites from the circulation. If ingestion is recognized early, ethylene glycol may be removed from the gastrointestinal tract by inducing emesis, administering activated charcoal, or performing gastric lavage. Ethanol is administered as a preferential substrate for alcohol dehydrogenase, thereby competitively inhibiting metabolism of ethylene glycol to its toxic metabolites and allowing excretion of the unmetabolized parent compound. Ethanol therapy is most effective if instituted early, given the half-life of ethylene glycol (~3 h).
As in case 1, ethanol therapy may be sufficient to prevent renal failure and effect complete clinical recovery [12,14,68]. Additional measures, however, are often initiated simultaneously before renal failure supervenes. Hemodialysis removes both ethylene glycol and its toxic metabolites, particularly glycolate and oxalate, as well as ethanol [23-25]. Consequently, ethanol administration should be maintained during hemodialysis. In addition to hemodialysis, chronic alcohol abuse and activated charcoal may decrease serum ethanol concentration so that loading doses must be adjusted to maintain serum alcohol concentration >22 mmol/L (>100 mg/dL). Ethanol therapy is associated with additional neurological depression, and frequent monitoring during hemodialysis is necessary to maintain ethanol serum concentrations at appropriate values. As an alternative to ethanol therapy for ethylene glycol poisoning, other potent and specific inhibitors of alcohol dehydrogenase, e.g., 4-methylpyrazole, have been used to treat ethylene glycol poisoning in humans [69,70].
In conclusion, early diagnosis and treatment of ethylene glycol poisoning can prevent substantial morbidity and mortality. The diagnosis may be straightforward, as in cases 1 and 2, if the patient or a third party relates a clear history of ethylene glycol poisoning. If the poisoning is not detected until late in the clinical course, as in case 3, toxicokinetic variables affecting measurement of serum ethylene glycol and other variables such as the anion gap and osmolal gap may obscure the laboratory diagnosis of suspected ethylene glycol poisoning. Finally, as in case 4, analytical variables may affect the results of laboratory assays for ethylene glycol.
Received May 13, 1997; revision accepted September 8, 1997.
[1.] Turk J, Morrell L, Avioli LV. Ethylene glycol intoxication. Arch Intern Med 1986;146:1601-3.
[2.] Jacobsen D, McMartin KE. Methanol and ethylene glycol poisonings. Mechanism of toxicity, clinical course, diagnosis and treatment [Review]. Med Toxicol 1986;1:309-34.
[3.] Litovitz TL, Smilkstein L, Felberg L, Klein-Schwartz W, Berlin R, Morgan JL. 1996 annual report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med 1997;15:447-500.
[4.] Eder AF, Dowdy YG, Gardiner JM, Wolf BA, Shaw LM. Elevated serum lactate and lactate dehydrogenase interfere in an enzymatic assay for ethylene glycol [Tech Brief]. Clin Chem 1996;42: 1489-91.
[5.] Woolf AD, Wynshaw-Boris A, Rinaldo P, Levy HL. Intentional infantile ethylene glycol poisoning presenting as an inherited metabolic disorder. J Pediatr 1992;120:421-4.
[6.] Kaiser W, Steinmauer HG, Biesenbach G, Janko 0, Zazgornik J. Chronic ethylene glycol poisoning. Dtsch Med Wochenschr 1993; 118:622-6. In German.
[7.] Linnanvuo-Laitinen M, Huttunen K. Ethylene glycol intoxication. Clin Toxicol 1986;24:167-74.
[8.] Denning DW, Berendt A, Chia Y, Morgan SH. Myocarditis complicating ethylene glycol poisoning in the absence of neurological features. Postgrad Med J 1988;64:867-70.
[9.] Hasiec T, Tynecka-Turowska M. Acute ethylene glycol poisoning with the course resembling the development of intracranial hematoma. Neurol Neurochir Pol 1991;25:501-3. In Polish.
[10.] Catchings TT, Beamer WC, Lundy L, Prough DS. Adult respiratory distress syndrome secondary to ethylene glycol ingestion. Ann Emerg Med 1985;14:594-6.
[11.] Levy RI. Renal failure secondary to ethylene glycol intoxication. J Am Med Assoc 1960;173:1210-3.
[12.] Haupt MC, Zull DN, Adams SL. Massive ethylene glycol poisoning without evidence of crystalluria: a case for early intervention. J Emerg Med 1988;6:295-300.
[13.] Blakeley KR, Rinner SE, Knochel JP. Survival of ethylene glycol poisoning with profound acidemia [Letter]. N Engl J Med 1993; 328:515-6.
[14.] Underwood F, Bennett WM. Ethylene glycol intoxication. Prevention of renal failure by aggressive management. J Am Med Assoc 1973;226:1453-4.
[15.] Collins JM, Hennes DM, Holzgang CR, Gourley RT, Porter GA. Recovery after prolonged oliguria due to ethylene glycol intoxication. Arch Intern Med 1970;125:1059-62.
[16.] Gutman RA, Hamon CB, Striker GE. Recovery after prolonged oliguria. Arch Intern Med 1970;126:914-5.
[17.] Spillane L, Roberts JR, Meyer AE. Multiple cranial nerve deficits after ethylene glycol poisoning. Ann Emerg Med 1991;20:20810.
[18.] Jacobsen D, Hewlett TP, Webb R, Brown ST. Ethylene glycol intoxication: evaluation of kinetics and crystalluria. Am J Med 1988;84:145-51.
[19.] Steinke W, Arendt G, Mull M, Reiners K, Toyka KV. Good recovery after sublethal ethylene glycol intoxication: serial EEG and CT findings. J Neurol 1989;236:170-3.
[20.] Cantarell MC, Fort J, Camps J, Sans M, Piera L. Acute intoxication due to topical application of diethylene glycol [Letter]. Ann Intern Med 1987;106:478-9.
[21.] Fligner CL, Jack R, Twiggs GA, Raisys VA. Hyperosmolality induced by propylene glycol. A complication of silver sulfadiazine therapy. JAMA 1985;253:1606-9.
[22.] Peterson CD, Collins AJ, Keane WF. Ethanol for ethylene glycol poisoning [Letter]. N Engl J Med 1981;305:977.
[23.] Peterson CD, Collins AJ, Himes JM, Keane WF. Ethylene glycol poisoning: pharmacokinetics during therapy with ethanol and hemodialysis. N Engl J Med 1981;304:21-3.
[24.] Jacobsen D, Ovrebo S, Ostborg J, Sejerstd OM. Glycolate causes the acidosis in ethylene glycol poisoning and is effectively removed by hemodialysis. Acta Med Scand 1984;216:409-16.
[25.] Gabow PA, Clay K, Sullivan JB, Lepoof R. Organic acids in ethylene glycol intoxication. Ann Intern Med 1986;105:16-20.
[26.] Cheng J-T, Beysolow TD, Kaul B, Weisman R, Feinfeld DA. Clearance of ethylene glycol by kidneys and hemodialysis. Clin Toxicol 1987;25:95-108.
[27.] Baselt RC, Cravey RH, eds. Disposition of toxic drugs and chemicals in man, 3rd ed. Chicago: Year Book Medical Publishers, 1989.
[28.] Litovitz TL, Felberg L, Soloway RA, Ford M, Geller R. 1994 annual report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med 1994:13:55197.
[29.] Steinhart B. Case report: severe ethylene glycol intoxication with normal osmolal gap-"a chilling thought." J Emerg Med 1990;8: 583-5.
[30.] Godolphin W, Meagher EP, Sanders HD, Frohlich J. Unusual calcium oxalate crystals in ethylene glycol poisoning. Clin Toxicol 1980;16:479-86.
[31.] Glasser L, Sternglanz PD, Combie J, Robinson A. Serum osmolality and its applicability to drug overdose. Am J Clin Pathol 1973;60:695-9.
[32.] Aabakken L, Johansen KS, Rydningen E-B, Bredesen JE, Ovrebo S, Jacobsen D. Osmolal and anion gaps in patients admitted to an emergency medical department. Hum Exp Toxicol 1994;13: 131-4.
[33.] Hoffman RS, Smilkstein MJ, Howland MA, Goldfrank LR. Osmol gaps revisited: normal values and limitations. Clin Toxicol 1993; 31:81-93.
[34.] Eisen TF, Lacouture PG, Woolf A. Serum osmolality in alcohol ingestions: differences in availability among laboratories of teaching hospital, nonteaching hospital, and commercial facilities. Am J Emerg Med 1989;7:256-9.
[35.] Karlson-Stiber C, Persson H. Ethylene glycol poisoning: experiences from an epidemic in Sweden. Clin Toxicol 1992;30:56574.
[36.] Moriarty RW. The spectrum of ethylene glycol poisoning. Clin Toxicol 1974;7:583-96.
[37.] DaRoza R, Henning RJ, Sunshine I, Sutheimer C. Acute ethylene glycol poisoning. Crit Care Med 1984;12:1003-5.
[38.] Heckerling PS. Ethylene glycol poisoning with a normal anion gap due to occult bromide intoxication. Ann Emerg Med 1987;16: 1384-6.
[39.] Leon M, Graeber C. Absence of high anion gap metabolic acidosis in severe ethylene glycol poisoning: a potential effect of simultaneous lithium carbonate ingestion. Am J Kidney Dis 1994;23: 313-6.
[40.] Jacobsen D, Bredesen JE, Eide I, Ostborg J. Anion and osmolal gaps in the diagnosis of methanol and ethylene glycol poisoning. Acta Med Scand 1982;212:17-20.
[41.] Cadnapaphomchai P, Taher S, Bhathena D, McDonald FD. Ethylene glycol poisoning: diagnosis based on high osmolal and anion gaps and crystalluria. Ann Emerg Med 1981;10:94-7.
[42.] Gabow PA, Kaehny WD, Fennessey PV, Goodman SI. Diagnostic importance of an increased serum anion gap. N Engl J Med 1980;303:854-8.
[43.] Sulway MJ, Malins JM. Acetone in diabetic ketoacidosis. Lancet 1970; ii 1:736-40.
[44.] Davidson DF. Excess osmolal gap in diabetic ketoacidosis explained. Clin Chem 1992;38:755-7.
[45.] Schelling JR, Howard RL, Winter SD, Linas SL. Increased osmolal gap in alcoholic ketoacidosis and lactic acidosis. Ann Intern Med 1990;113:580-2.
[46.] Fulop M, Bock J, Ben-Ezra J, Antony M, Danzig J, Gage JS. Plasma lactate and 3-hydroxybutyrate levels in patients with acute ethanol intoxication. Am J Med 1986;80:191-4.
[47.] Inaba H, Hirasawa H, Mizuguchi T. Serum osmolality gap in postoperative patients in intensive care. Lancet 1987;i:1331-5.
[48.] Sklar AH, Linas SL. The osmolal gap in renal failure. Ann Intern Med 1983;98:481-2.
[49.] Huhn KM, Rosenberg FM. Critical clue to ethylene glycol poisoning. Can Med Assoc J 1995;152:193-5.
[50.] Jacobsen D, Akesson I, Shefter E. Urinary calcium oxalate monohydrate crystals in ethylene glycol poisoning. Scand J Clin Lab Invest 1982;42:231-4.
[51.] Terlinsky AS, Grochowski J, Geoly KL, Stauch BS, Hefter L. Identification of atypical calcium oxalate crystalluria following ethylene glycol ingestion. Am J Clin Pathol 1981;76:223-6.
[52.] Fraser AD, MacNeil W. Colorimetric and gas chromatographic procedures for glycolic acid in serum: the major toxic metabolite of ethylene glycol. Clin Toxicol 1993;31:397-405.
[53.] Hewlett TP, McMartin KE. Ethylene glycol poisoning. The value of glycolic acid determinations for diagnosis and treatment. Clin Toxicol 1986;24:389-402.
[54.] Ovrebo S, Jacobsen D, Sejersted OM. Determination of ionic metabolites from ethylene glycol in human blood by isotachophoresis. J Chromatogr 1987;416:111-7.
[55.] Edinboro LE, Nanco CR, Soghioan DM, Poklis A. Determination of ethylene glycol in serum utilizing direct injection on a wide-bore capillary column. Ther Drug Monit 1993;15:220-3.
[56.] Livesey JF, Perkins SL, Tokessy NE, Maddock MJ. Simultaneous determination of alcohols and ethylene glycol in serum by packed or capillary-column gas chromatography. Clin Chem 1995;41: 300-5.
[57.] LeGatt DF, Tisdell RH. Ethylene glycol quantification: avoid propylene glycol as an internal standard [Letter]. Clin Chem 1990; 36:1860-1.
[58.] Apple FS, Googins M, Resen D. Propylene glycol interference in gas-chromatographic assay of ethylene glycol [Letter]. Clin Chem 1993;39:167.
[59.] Smith NB. Internal standards in gas-chromatographic analyses for ethylene glycol in serum [Letter]. Clin Chem 1993;39:2020.
[60.] Shoemaker JD, Lynch RE, Hoffmann JW, Sly WS. Misidentification of propionic acid as ethylene glycol in a patient with methylmalonic acidemia. J Pediatr 1992;120:417-21.
[61.] Jones AW, Nilsson L, Gladh SA, Karlsson K, Beck-Friis J. 2,3Butanediol in plasma from an alcoholic mistakenly identified as ethylene glycol by gas-chromatographic analysis. Clin Chem 1991: 37:1453-5.
[62.] Bjellerup P, Kallner A, Kollind M. GLC determination of serumethylene glycol, interferences in ketotic patients [Letter]. Clin Toxicol 1994;32:85-7.
[63.] Ryder KW, Glick MR, Jackson SA. Emergency screening for ethylene glycol in serum. Clin Chem 1986;32:1574-7.
[64.] Hansson P, Masson P. Simple enzymatic screening assay for ethylene glycol (ethan-1,2-diol) in serum. Clin Chim Acta 1989; 182:95-102.
[65.] Standefer J, Blackwell W. Enzymatic method for measuring ethylene glycol with a centrifugal analyzer. Clin Chem 1991;37: 1734-6.
[66.] Jessen RH, Dass CJ, Eckfeldt JH. Do enzymatic analyses of serum triglycerides really need blanking for free glycerol? Clin Chem 1990;36:1372-5.
[67.] Thede-Reynold K, Johnson GF. False positive ethanol results by EMIT [Abstract]. Clin Chem 1993;39:1143.
[68.] Brown CG, Trumbull D, Klein-Schwartz W, Walker JD. Ethylene glycol poisoning. Ann Emerg Med 1983;12:501-6.
[69.] Baud FJ, Galliot M, Astier A, Vu Bien DV, Gamier R, Likforman J, Bismuth C. Treatment of ethylene glycol poisoning with intravenous 4-methylpyrazole. 1988;319:97-100.
[70.] Harry P, Turcant A, Bouachour G, Houze P, Alquier P, Allain P. Efficacy of 4-methylpyrazole in ethylene glycol poisoning: clinical and toxicokinetic aspects. Hum Exp Toxicol 1994;13:61-4.
ANNE F. EDER, CINDY M. MCGRATH, YVONNE G. DOWDY, JOHN E. TOMASZEWSKI, FRANCES M. ROSENBERG, (1) ROBERT B. WILSON, BRYAN A. WOLF, and LESLIE M. SHAW *
Department of Pathology and Laboratory Medicine Hospital of the University of Pennsylvania, 3400 Spruce St., Philadelphia, PA 19103. (1) Departments of Pathology and Laboratory Medicine, St. Paul's Hospital and University of British Columbia, Vancouver, BC, Canada V6Z 1Y6.
*Corresponding author: fax (215) 662-7529; e-mail les shaw@PATHIA. MED.
Table 1. Laboratory findings on presentation. Case Test Ref. range 1 2 Serum ethylene glycol, 995 (16) 7910 (127) mg/L (mmol/L) Arterial blood gases: pH 7.35-7.45 7.32 7.23 [MATHEMATICAL EXPRESSION 35-45 34 16 NOT REPRODUCIBLE IN ASCII], mmHg HC[0.sub.3.sup.-], 22-26 17 7 mmol/L Anion gap, mmol/L 12-16 13 26 Lactate, mmol/L 0.7-2.1 N.R. N.R. Serum osmolality, mOsm/kg [H.sub.2]O Measured 270-290 304 434 Calculated 288 264 Gap <10 16 170 Creatinine, [micro]mol/L 44-150 106 80 Urea nitrogen, mmol/L 2.9-8.9 3.6 1.8 Urinalysis Few Ca ox. Rare Ca ox. crystals crystals Case Test 3 4 Serum ethylene glycol, N. D. 90 (1.5) mg/L (mmol/L) Arterial blood gases: pH 7.29 7.11 [MATHEMATICAL EXPRESSION 21 14 NOT REPRODUCIBLE IN ASCII], mmHg HC[0.sub.3.sup.-], 11 9 mmol/L Anion gap, mmol/L 20 29 Lactate, mmol/L 1.1 13.3 Serum osmolality, mOsm/kg [H.sub.2]O Measured N. D. 303 Calculated N. D. 277 Gap N. D. 26 Creatinine, [micro]mol/L 221 150 Urea nitrogen, mmol/L 6.8 11.4 Urinalysis Hematuria, rare, Proteinuria, amorphous hematuria, crystals urate crystals N.D., none detected; N.R., not reported; Ca ox., calcium oxalate. Table 2. Toxicokinetic parameters for ethylene glycol. (a) Lethal dose (b) 1.4-1.6 mL/kg [V.sub.d] 0.5-0.8 L/kg [t.sub.1/2] 3-8.6 h [t.sub.1/2] during ethanol infusion 17-18 h [t.sub.1/2] during hemodialysis 2.5-3.5 h Mean renal clearance (c) 0.75-27.5 mL/min Mean dialyzer clearance 156-210 mL/min (a) Data compiled from references 18, 23, 26, 27, and 33. (b) Although the lethal dose is considered 100 ml -for a 70-kg adult, survival has been reported after ingestions of 0.27-2 L. Also, death has been reported after ingestion of 30 mL. (c) Lower value obtained in patient with renal failure; upper value obtained in patient with normal renal function.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Case Conference|
|Author:||Eder, Anne F.; McGrath, Cindy M.; Dowdy, Yvonne G.; Tomaszewski, John E.; Rosenberg, Frances M.; Wil|
|Date:||Jan 1, 1998|
|Previous Article:||Neopterin concentrations in cerebrospinal fluid and serum as an aid in differentiating central nervous system and peripheral infections in children.|
|Next Article:||Direct molecular interaction of phenylalanine with DNA measured by reversed-phase HPLC.|