Toxic Levels of Acetaminophen in Children Questioned To the Editor: In the case report and commentary “An Unusual Case of Acute Abdomen” (January 2006), Dr Mendoza and colleagues describe a possible case of subacute acetaminophen toxicity, although other causes and contributing factors, such as the role of clindamycin, are not excluded.
The authors assert that “a single dose of 120 to 150 mg/kg [of acetaminophen] is considered hepatotoxic in children.” This range is taken from the consensus report of the American Academy of Pediatrics Committee on Drugs,1 which cites 2 dubious references to support this lower range of concern.2,3
The first case describes an 11-month-old infant who received 420 mg/kg a day for 1.25 days (a total of 525 mg/kg) and had an acetaminophen concentration of 240 µg/mL at 11 hours after the last acetaminophen dose.2 It is not surprising that this child developed hepatotoxicity, but this case does not provide evidence for the unusually low level of toxicity of 120 mg/kg for acetaminophen.
The second case relies solely on the unconfirmed self-report of an 11-year-old girl who estimated taking 120 mg/kg of the drug over a period of 30 hours, for headache.3 She developed mild hepatotoxicity, with a peak aspartate aminotransferase level of only 454 IU/L. This single case, which is subject to recall bias by a young patient, is insufficient to compel the universal application of a 120-mg/kg dose as the toxic level criterion.
Young children are more resistant to acetaminophen toxicity than adults on a mg/kg basis.4-6 Although the dose limit for adolescents and adults is generally accepted to be 150 mg/kg, the dose of concern for children is much higher (250 mg/kg for infants <1 year of age and 200 mg/kg for toddlers aged <2 years).4-6
Michael E. Mullins, MD, FACEP Division of Emergency Medicine Washington University School of Medicine St. Louis, Mo
1. American Academy of Pediatrics. Committee on Drugs. Acetaminophen toxicity in children. Pediatrics. 2001;108: 1020-1024.
2. Henretig FM, Selbst SM, Forrest C, et al. Repeated acetaminophen overdosing causing hepatotoxicity in children. Clinical reports and literature review. Clin Pediatr (Phila). 1989;28:525-528.
4. Tenenbein M. Acetaminophen: the 150 mg/kg myth. J Toxicol Clin Toxicol. 2004;42:145-148.
5. Bond GR. Reduced toxicity of acetaminophen in children: it’s the liver. J Toxicol Clin Toxicol. 2004;42:149-152.
6. Tenenbein M. Why young children are resistant to acetaminophen poisoning. J Pediatr. 2000;137:891-892.
Dr Calello Replies: I welcome Dr Mullins’ contribution to the discussion of the case report. I believe, however, that his response may lead to some confusion about a few important issues.
First, there is a significant difference between acute single overdose and repeated supratherapeutic dosing. Although Dr Mullins takes exception to the authors’ statement that “a single dose of 120 to 150 mg/kg is toxic to young children,” the authors do go on to say that these guidelines do not apply in the unpredictable pharmacokinetics of chronic overdose, as was the case in this patient. Furthermore, the report he cites, by Dr Henretig and colleagues, also refers to chronic overdose.1 Although an absolute dose limit of concern in chronic overdose is not well established, the report by Dr Henretig and colleagues describes a case in which hepatotoxicity occurred at a dose close to the low level of toxicity of 120 mg/kg, at 175 mg/kg.1 This has also been suggested in a larger case series of 47 children, several of whom developed severe or fatal hepatotoxicity at acetaminophen doses as low as 60 mg/kg daily.2
I agree with the assertion that children tolerate acute overdoses better than adult patients. This has been proposed to be the result of a number of mechanisms, such as increased activity of alternative metabolic pathways and/or a proportionally larger liver per body size than in adults.3,4 The absolute implications of this, however, with respect to toxic dose and serum levels, remain unclear. Some authorities and Poison Control Centers nationwide continue to use the 150-mg/kg dose as a criterion for referral of pediatric patients to a health care facility.5 A consensus guideline has very recently been published, recommending revising this referral criterion upward to a 200-mg/kg dose if exposure is less than 24 hours, but the efficacy of this approach has not been proven in controlled studies. The guideline also recommends emergency department referral at lower doses when exposure spans more than 1 day.6
Finally, although this is an interesting discussion, a dispute over the precise levels of exposure that pose concern may obscure the major “take-home message” of the article. Chronic overdose with therapeutic intent is a major, and often-overlooked, cause of acetaminophen poisoning in children that can have severe clinical consequences. General physicians and pediatricians need to be aware of this phenomenon; they should counsel their patients to avoid overzealous antipyretic therapy of febrile illnesses and consider the diagnosis of inadvertent chronic acetaminophen toxicity early in children presenting with acute hepatic injury. Certainly, any question that arises regarding significant acute or chronic acetaminophen ingestion should prompt a request for consultation from the regional Poison Control Center.
Diane P. Calello, MD The Children’s Hospital of Philadelphia The Poison Control Center Philadelphia, Pa
1. Henretig FM, Selbst SM, Forrest C, et al. Repeated acetaminophen overdosing causing hepatotoxicity in children. Clinical reports and literature review. Clin Pediatr (Phila). 1989;28:525-528.
2. Heubi JE, Barbacci MB, Zimmerman HJ. Therapeutic misadventures with acetaminophen: hepatoxicity after multiple doses in children. J Pediatr. 1998;132:22-27.
3. Tenenbein M. Acetaminophen: the 150 mg/kg myth. J Toxicol Clin Toxicol. 2004;42:145-148.
4. Bond GR. Reduced toxicity of acetaminophen in children: it’s the liver. J Toxicol Clin Toxicol. 2004;42: 149-152.
5. Bizovi KE, Smilkstein MJ. Acetaminophen. In: Goldfrank LG, Flomenbaum NE, Lewin NA, et al, eds. Goldfrank’s Toxicologic Emergencies. 7th ed. New York, NY: McGraw-Hill; 2002:480-506.
6. Dart RC, Erdman AR, Olson KR, et al, for the American Association of Poison Control Centers. Acetaminophen poisoning: an evidence-based consensus guideline for out-of-hospital management. Clin Toxicol (Phila). 2006;44:1-18.
Editor’s note: Dr Calello wrote the commentary for the case report. The authors declined the opportunity to reply.
Measuring QTc and Olanzapine Safety
To the Editor: I read with interest the fascinating case report “Long QT Syndrome After a Motor Vehicle Incident” by Dr Crown and colleagues (February 2006). It is important to mention, however, several points regarding the QT interval. First, it is well-established that approximately 2.5% of the healthy population has a prolonged QT interval (classically defined as >440 ms), and there are many cofactors that affect the QT interval. As mentioned by Dr Crown and colleagues, heart rate is the best known factor demonstrating a direct relationship to the QT interval. More than 30 formulas are available for evaluating the relationship between the QT interval and the patient’s heart rate, as reflected by the corrected QT (QTc) value. Each has specific limitations, however.
The basic equation describing this relationship is QTc = QT/RRa. The alpha exponent adjusts for the variation in heart rate. At least 8 studies have proposed different values for alpha, ranging from 0.25 to 0.603. The best known value is 0.5, proposed by Dr Henry Bazett in 1920.1 It remains a mystery why the Bazett formula has taken such a strong hold on the medical field, since it was derived empirically by prestatistical methods using a small patient population. Mathematically speaking, at an alpha of 0.5 and a heart rate of 60 beats/min, the QT equals the QTc. Consequently, extremes of heart rate in this model produce falsely elevated QTc values (heart rate >60 beats/min) and falsely reduced QTc values (heart rate <60 beats/min). The often-quoted normal QT interval duration mentioned in most studies is 440 ms. This was probably derived using the Bazett formula in patients whose heart rate was more than 60 beats/min. Therefore, it is conceivable that the normal QTc in the general population is less than 440 ms.
With regards to drug therapy, QT intervals of less than 470 ms in women and less than 450 ms in men are probably well tolerated and do not require medication adjustment.2 More recently, a number of studies have demonstrated that the alpha exponent is likely person-specific and is probably consistent from year to year. This phenomenon will make it difficult to generate an accurate method for calculating the QTc in the general population, particularly if there is drug-induced prolongation. To add to the complexity of the QT interval, a hysteresis exists between the ever-so-intertwined QT interval and heart rate. The QT interval is highly dependent on heart rate, but changes in the QT interval are not instantaneous with changes in heart rate, and it may take several minutes for the QT interval to reach a steady state after a change in heart rate. The dependence of the QT interval on heart rate is implied by QT:heart rate. We expect most people to reach a steady-state QT interval within 1 to 2 minutes following a change in heart rate. Patients with a delay in reaching that steady state (ie, a delayed ability for QT: heart rate adaptation) may be at a greater risk for torsades de pointes or other dysrhythmias. It is recommended that the heart rate be stable for several minutes before attempting to calculate the QT interval.
One final note. The majority of drugs implicated in causing drug-induced QT prolongation exert their effect on the delayed rectifier potassium current (Ikr) channel. Considering the number of drugs implicated in prolonging the QT interval, the overall clinical incidence of torsades de pointes is probably low. In addition, not all drugs that block Ikr have the same proarrhythmic potential. The authors list several medications known to prolong the QT interval; one of these medications—olanzapine—is listed in error. Unlike other antipsychotic medications, olanzapine does not contribute significantly to QT prolongation.3,4 Because of the common association of QT prolongation with antipsychotic medications, olanzapine is often recommended in patients with a known long QT syndrome or those using other medications that may prolong the QT interval.
G. Patrick Daubert, MD Children’s Hospital of Michigan Regional Poison Control Center Wayne State University School of Medicine Detroit, Mich
1. Bazett HC. An analysis of the time-relationships of electrocardiograms. Heart. 1920;7:353-370.
2. Al-Khatib SM, LaPointe NM, Kramer JM, et al. What clinicians should know about the QT interval [published correction appears in JAMA. 2003;290:1318]. JAMA. 2003;289:2120-2127.
3. Czekalla J, Kollack-Walker S, Beasley CM Jr. Cardiac safety parameters of olanzapine: comparison with other atypical and typical antipsychotics. J Clin Psychiatry. 2001;62(suppl 2): 35-40.
4. Isbister GK, Whyte IM, Smith AJ. Olanzapine overdose [letter]. Anaesthesia. 2001;56:400-401.
The Authors Reply: We appreciate the input of Dr Daubert concerning our recent case report. In particular, he mentions a current popular medication, olanzapine, which is listed in Table 1 among drugs to be wary of in the setting of QT interval problems.
Olanzapine initially was discussed in an article published in 2003 as being a possible problem for high-risk patients.1 Dr Daubert is correct that the current online search on PubMed leads to a site that does not consider olanzapine as a risk in any group of patients with torsades de pointes or QT interval problems. We therefore accept this updated medical information and welcome his bringing it to everyone’s attention. The other antipsychotics have far more risk, and olanzapine may be the drug of choice in situations requiring this category of medication.
We also welcome Dr Daubert’s recommendation for caution in measuring QT intervals and agree that it is difficult to accurately measure this phenomenon. His comment about the need for the heart rate to be “stable for several minutes before attempting to calculate the QT interval” is most valuable.
Loren A. Crown, MD University of Tennessee Covington, Tenn
Joseph J. Flagge, MD Palmetto General Hospital Hialeah, Fla
Dale Criner, MD University of Tennessee
1. Al-Khatib SM, LaPointe NM, Kramer JM, et al. What clinicians should know about the QT interval [published correction appears in JAMA. 2003;290:1318]. JAMA. 2003;289:2120-2127.
Search
Resources
