Falsely elevated salicylate levels due to diflunisal overdose

The doumai

ofEmergencyMedmx? Vol 5, pp 499-503, 1987

Printedanthe USA CopyrIght ic 1987 PergamonJo~mnlsLtd ??

FALSELY ELEVATED SALICYLATE LEVELS DUE TO DIFLUNISAL OVERDOSE Kurt R. Duffens, MD,* Martin J. Smilkstein, M&t Howard A. Bessen, MD,* Barry H. Rumack, MD§ *Department of Emergency Medrcrne. Denver General Hosprtal, Denver, Colorado, TRocky Mountarn Poison and Drug Center. Denver General Hosprtal, Department of Surgery, Sectron of Trauma. and Emergency Medrcrne. Unrversity of Colorado Health Scrences Center, Denver, Colorado; *Department of Emergency Medicine, Harbor-UCLA Medrcal Center, Torrance, Californra, and SDepartment of Pediatrics, Unrversrty of Colorado Health Scrences Center, Director, Rocky Mountain Poison and Drug Center. Denver General Hosprtal Repnnt address. Barry H. Rumack, MD, Director, Rocky Mountain Poison Center, 645 Bannock Street, Denver, CO 80204

0 Abstract-A 33-year-old man ingested 14 g of diflunisal (Dolobid@, a NSAID), 3 propoxypbene-acetaminopben tablets, and ethanol. Although salicylate ingestion was specifically denied, toxicology screening revealed markedly elevated salicylate levels (176 mg/dL, 100 mg/ dL, 41 mg/dL at 4’/2, 10 and 24 hours postingestion). Blood ethanol level was 450 mg/dL. There was no respiratory alkalosis, metabolic acidosis, or increased anion gap to suggest salicylate toxicity. The patient recovered fully. Comparison testing of the salicylate assay technique used in the case (TDx@ fluorescence polarization immunoassay) and another commonly used technique (Trinder calorimetric assay) revealed marked cross-reactivity be-

0 Keywords - diflunisal; salicylate; false-positive; drug-screening

Dr Smilkstein was recipient of a McNeil Fellowship in Clinical Toxicology at the time of this report. Abstract presented at the 1986 American Association of Poison Control Centers, American Association of Clinical Toxicology, American Board of Medical Toxicology, Canada Association of Poison Control Centers Annual Scientific Meeting held in Santa Fe, NM, September 25-30, 1986.

=YZIEEY ~

tween salicylate and diflunisal. Samples of known diflunisal concentration were tested using the TDx@ salicylate assay. Percent crossreactivity [(reported salicylate concentration/ actual diflunisal concentration) x 1001 was found to be 130 to 152.8% over the concentration range tested. Cross-reactivity between diflunisal and salicylate using the Trinder calorimetric assay was 20 to 52.5% over the concentration range tested. Diflunisal is not metabolized to salicylate and thus this crossreactivity represents a significant drug-laboratory interaction. Salicylate levels measured by these methods appear to be unreliable when diflunisal is present.

Introduction

With the proliferation and popularity of nonsteroidal antiinflammatory drugs (NSAIDs) the potential for overdose of these medications is increasing. Due to

Toxicology -one of the most critical and challenging areas confronting the emergency department staff-is coordinated by Kenneth Kulig, MD, of the Rocky Mountain Poison Center

RECEIVED:8 January 1987; ACCEPTED:19 February 1987 499

0736-4679/87 $3.00 + .OO

500

K. I?. Duffens, M. J. Smilkstein, H. A.

limited clinical experience, the appropriate role of toxicologic screening tests in NSAID overdose has yet to be determined. NSAIDs are not routinely assayed, and the possibility of their interference with routine screening tests has not been clarified. The following is a case of a polydrug overdose involving diflunisal (2’,4’ difluoro-4-hydroxy-3-biphenyl-carboxylic acid (Dolobid@)) in which toxicologic screening yielded a falsely elevated salicylate level due to the presence of diflunisal. Case Report

A 33-year-old man was brought to a community hospital approximately 45 min after ingesting drugs and alcohol in a suicide attempt. According to the patient, he ingested 28, 500-mg diflunisal (Dolobid@) tablets, 3 propoxyphene-acetaminophen (Darvocet@) tablets, and a large amount of alcohol. Thirty minutes after the ingestion, the patient called the paramedics. On arrival at the emergency department, he was awake and alert but had slurred speech. Vital signs were: blood pressure, 120/90 mm Hg; pulse, 100 beats per minute; respirations, 20/min. Physical examination was unremarkable except for alcohol on the breath and the presence of a short leg cast. The hematocrit was 48.1% with a WBC count of 7,800 per mm3. Serum electrolytes (in mEq/L) were: sodium 144, potassium 3.3, chloride 110, and bicarbonate 24. The BUN was 8 mg/dL and the glucose was 114 mg/dL. An arterial blood gas (on unknown FIO,) showed a p0, of 11.5 mm Hg, pC0, of 37 mm Hg, and a pH of 7.43. Blood ethanol level was 450 mg/dL. He was given 60 cc of syrup of ipecac, vomited shortly thereafter (the records did not note whether pill fragments were obtained), and was transferred to the Harbor-UCLA Medical Center for further care. On arrival (approximately four

Bessen, B. H. Rumack

hours after the ingestion), he was lethargic but easily arousable. The same ingestion history was obtained. The patient complained of mild dizziness and tinnitus, but denied any other symptoms. His temperature was 97.3”F; pulse rate, 120 beats per minute; respirations, 24/min; and blood pressure was 160/100 mm Hg. The physical exam revealed marked diaphoresis but was otherwise unremarkable. The patient was given 60 g of activated charcoal and 300 cc of magnesium citrate by nasogastric tube. Repeat electrolytes (mEq/L) showed a sodium of 143, potassium of 3.7, chloride of 105, and bicarbonate of 23. An arterial blood gas (drawn with the patient receiving oxygen at an unknown rate by nasal cannula) showed a p0, of 110 mm Hg, pC0, of 37, and pH of 7.41; calculated bicarbonate was 23 mEq/L. The ECG showed only sinus tachycardia, and the chest x-ray study was normal. Toxicologic testing revealed a serum acetaminophen level of 7.6 pg/mL, and a serum salicylate level (TDx@ Salicylate, Abbott Laboratories, Chicago, IL) of 1,760 pg/mL (176 mg/dL) 4*/z hours postingestion. After the salicylate level returned, the patient was given intravenous saline and sodium bicarbonate to achieve an alkaline diuresis. Over the next several hours, he remained tachycardic (120-140 beats per minute) and slightly tachypneic (18-28 breaths per minute). Blood pressure declined to 130170 mm Hg and remained stable thereafter. Arterial blood gases six hours after ingestion, drawn with the patient receiving oxygen by nasal cannula at 4 liters per minute, showed a p0, of 198, Torn pC0, of 41 Torr, pH of 7.41, and a bicarbonate of 26 mEq/L. Serum salicylate level ten hours after the ingestion was 1,004 pg/mL (100.4 mg/dL). Serum electrolytes were measured two hours later and revealed a bicarbonate concentration of 32 mEq/L with an anion gap (Na+ (Cl- + HC03-)) of 10 mEq/L. Two subse-

Falsely Elevated Salicylate Levels

quent determinations of serum electrolytes yielded similar results. The patient was observed for 24 hours and remained stable. His tachycardia and lethargy gradually resolved, no further symptoms developed, and he was discharged after psychiatric consultation. Salicylate level drawn approximately 24 hours after the ingestion was 410 pg/mL (4 1 mg/dL).

Discussion This patient’s salicylate levels of 176 mg/ dL, 100 mg/dL, and 41 mg/dL at 4’/2, 10, and 24 hours after ingestion, respectively, would be expected to be associated with symptoms of severe salicylate poisoning.’ Whereas the patient experienced some early tinnitus and tachypnea, these were not constant or persistent, and no other signs of salicylism were evident. Serial arterial blood gas and electrolyte determinations failed to show a respiratory alkalosis, metabolic acidosis, or an increased anion gap at any time in his clinical course. The absence of these signs is inconsistent with significant salicylate intoxicati0n.l By history, this patient had ingested no aspirin, but admitted to an ingestion of 14 grams of diflunisal. Possible theoretical explanations for the detection of elevated salicylate levels in this case could include inaccurate history, metabolism of diflunisal to salicylate, or a false-positive salicylate assay due to the presence of a crossreacting substance. Previous studies demonstrate that although diflunisal is a salicylic acid derivative (Figure 1) there is no known metabolism to salicylate. Diflunisal undergoes hepatic conjugation and is excreted primarily in the urine as ester and ether glucuronide conjugates. In addition, a small amount is excreted unchanged, but no other metabolites have been identified.2 The clinical course of this patient was

501

consistent with diflunisal toxicity. Although some of the signs and symptoms were likely related to ethanol intoxication, lethargy, tachycardia, tachypnea, and tinnitus have all been attributed to diflunisal toxicity.3 The findings in this case appear to be due to cross-reactivity between ditlunisal and salicylate with the TDx@ salicylate assay. The TDx@ assay utilizes a flourescence polarization immunoassay to measure salicylate concentration.J,5 Salicylate has been previously noted to interfere with fluorescence assays for diflunisal,? and the first reports of false-positive salicylate assays due to diflunisal interference appeared recently.h,: Following this case, samples of known diflunisal concentration (containing no salicylate) were assayed using the TDx~ salicylate assay. These results are shown in Table 1, and demonstrate marked cross[(reported reactivity. The cross-reactivity salicylate concentration/actual diflunisal concentration) x 1001 of diflunisal using this assay (130-152.8%) is far greater than for other related substances. Acetylsalicylit acid and salicyluric acid have crossreactivities of only 1.3 and 2.3%, respectively, using the TDx@ salicylate assay.’ Acetaminophen and propoxyphene, also ingested in this case, show less than 0.1% cross-reactivity.5 As a result of our study, the TDx@ product manual has been changed to reflect these findings. Another commonly used salicylate assay is the Trinder calorimetric method.8 Table 2 shows the results of calorimetric assay of several known concentrations of diflunisal. Significant cross-reactivity (20-52.5%) was seen with this assay as well. Diflunisal assays are not widely available. Techniques that are likely to be specific for diflunisal include high-pressure liquid chromatography and gas chromatography. The cost and complexity of these techniques make them unavailable in most clinical settings. The reliability of extrapo-

502

K. R. Duffens,

M. J. Smilkstein,

H. A. Bessen,

B. H. Rumack

COOH OH

Diflunisal

Acetylsalicylic Acid (Aspirin)

Figure 1. Structures of diflunisal, aspirin (acetylsalicylic

acid), and salicylic acid.

Table 1. Diflunisal Cross-Reactivity Salicylate Assay* Actual [Diflunisal] WW 19.0 95.0 190.0

Salicylic Acid

with TDx@

Reported [Salicylate] (~Lg~mL)

% Cross-Reactivityi

24.7 126.4 290.3

130.0 133.0 152.8

*Personal communication, Todd A. Price, PhD, Abbott Laboratories. t(Reported salicylate concentration/Actual diflunisal concentration) x 100

lating diflunisal levels after overdose from either the TDx@ or calorimetric salicylate results is unknown. With the TDx@ assay cross-reactivity varied from 130% (diflunisal= 19 pg/mL) to 152.8% (diflunisal= 190 pg/mL) in our report. Sarma et al6 reported 230% cross-reactivity (diflunisal = 74 and 113 pg/mL). Dalrymple et al7 report cross-reactivities of > 200,246, and 3 11% (diflunisal = 400, 200, and 100 pg/mL, respectively). The variability of these results demonstrate that further testing over a wide range of concentrations will be needed to determine if TDx@ salicylate values can ultimately be used to predict actual diflunisal levels. Much of the variation in results may be due to a change from rabbit to sheep TDx@ antiserum. Preliminary work in this report using rabbit antiserum yielded considerably higher cross-reactivity. Reported cross-reactivity using colorimetric techniques shows less variation. With the Trinder method our results varied from 20% (diflunisal= 50 pg/mL) to 49, 52.5, and 45% (diflunisal 100, 200,

and 400 pg/mL respectively). Dalrymple et al7 found 61, 52, and 49% (diflunisal 100, 200, and 400 pg/mL) cross-reactivity with the Trinder method. Cross-reactivity using the DuPont aca@ has been reported to be 63 and 59% (diflunisal 74 and 113 pg/mL)6 and 78, 61, and 50% (diflunisal 100, 200, and 400 pg/mL).7 Again, further testing will be needed to assess use of these methods to predict actual diflunisal levels. Summary

In summary, a significant cross-reactivity exists between diflunisal and salicylate using the TDx@ fluorescence polarization immunoassay or the Trinder calorimetric method. Other reports demonstrate crossreactivity using the DuPont aca@ system. Many medical laboratories utilize these popular methods for salicylate detection and measurement. If diflunisal is present, these methods may yield misleading data. In the presence of known diflunisal inges-

503

Falsely Elevated Salicylate Levels

Table 2. Diflunisal Cross-Reactivity Calorimetric Assay’

Actual [Diflunisal] WW 50.0 100.0 200.0 400.0

Reported [Salicylate] WmL)

with Trinder

% Cross-Reactivity 20.0 49.0 52.5 45.0

10.0 49.0 105.0 180.0

‘In-house modification of Trinder calorimetric assay. Personal communication, Analytitox” , Denver, CO.

or when signs and symptoms do not fit with elevated salicylate levels, results of these assays must be interpreted with caution. In all cases, treatment should be guided by the clinical status of the patient. tion,

Acknowledgment

-The authors thank Todd A. Price, PHD of Abbott Laboratories, and the staff of Analytitox” for performing the assays reported here. The authors also thank Lisa Wolfe for manuscriot preparation.

REFERENCES I. Temple AR: Acute

and chronic effects of aspirin toxicity and their treatment. Arch Intern Med 1981; 141:364-369. 2. Tocco DJ, Breault GO, Zacchei AC, et al: Physiological disposition and metabolism of 5-(2’,4’-difluorophenyl) salicylic acid, a new salicylate. Drug

Metab Dispos 1975; 3:453-466. 3. Court H, Volans GN: Poisoning after overdose with non-steroidal anti-inflammatory drugs. Adv Drug React AC Pois Rev 1984; 3: l-21. 4. Theory of measurement. TDP System Operation Manual. Abbott Laboratories, Chicago, IL, 1984. 5. Toxicologic assays - TDx~ Salicylate. TDfl Sys-

fern Operation Manual.

Abbott Laboratories, Chicago, IL, 1985. 6. Sarma L, Wang SH, DellaFera S: Diflunisal significantly interferes with salicylate measurements by FPIA-TDx@ and UV-VIS aca method, letter. Clin Chem 1985; 31:1922-1923. 7. Dalrymple RW, Stearns FM: Diflunisal interferes with determination of salicylate by the Trinder, Abbott TDx@ and DuPont aca@ method, letter.

C/in Chem 1986; 32:230. 8. Trinder P: Rapid determination of salicylate in biological fluids. Biochem J 1954; 57:301-m 303.