Toxicity After Repeated Acetaminophen Doses in Adults

Martin Smilkstein*
Oregon Poison Center and Department of Emergency Medicine
Oregon Health Sciences University
Portland, OR

Int J Med Toxicol 1998; 1(3): 19

Improving patient care through discovery, recognition and dissemination of valid information is a challenge achieved only by rigorous efforts at critical and statistical analysis. Medical toxicologists are expected to have the greatest ability and responsibility to take on this challenge within their area of interest. If so, they must take responsibility for the sorry state of information about toxicity from repeated acetaminophen (APAP) dosing and for the rapid spread of misconceptions which have resulted from unsubstantiated published conclusions.

On this issue, arguably the most important issue in medical toxicology today, the medical toxicology community has largely failed to contribute valid information and failed to limit or discredit invalid information. Rather than an evidence-based, logical progression toward the truth, unsupported notions have given rise to a polarization of opinions which has impaired meaningful analysis. This commentary seeks to stimulate and re-focus discussion of this issue among medical toxicologists by presenting one perspective on currently available information and by offering some “jumping off” concepts for dealing with practical issues of patient management. It is intended neither as a comprehensive review nor as an evidence-based conclusion to the discussion, but rather as a new starting point for dialogue and critical analysis.

Although repeated dosing in adults is the topic of this discussion, many of the same concepts apply to pediatric reports. (1) In either case, to avoid inappropriate generalizations between “apples and oranges” it is useful to break down the topic into a series of questions of increasing complexity.

Question #1: Can APAP toxicity result from repeated dosing?

Yes. Repeated doses, each smaller than that needed to cause toxicity after a single dose, may cause toxicity. There are abundant published (2,3) and anecdotal cases, and as long as threshold values for doses, dosing intervals, and duration of dosing are not included in the question, there is universal agreement on the answer.

Question #2: Can we reliably determine who is at risk on the basis of the dosing history?

No. Conceptually, repeated APAP dosing can be envisioned as a curve of APAP concentration ([APAP]) rising (absorption) and falling (elimination) with a superimposed curve of hepatic glutathione (GSH) falling (consumption) and rising (repletion). If toxicity risk begins when GSH is depleted to some critical threshold, (4) then these curves illustrate that a risk-predictive model needs to include not only APAP dose amounts, but also dosing interval and dosing duration.

A huge dose may be safe if enough time is allowed for GSH repletion prior to the next dose. Much smaller doses may be problematic if the subsequent doses follow too soon to allow GSH repletion. Conceivably, certain combinations of dose and interval might result in a very slow net decline in GSH, such that toxicity would only occur if dosing were continued for very long periods. Within any period of repeated dosing there will also be small or large variations in individual doses and dosing intervals, further reducing predictability. Finally, interindividual differences in absorption, glucuronidation and sulfation capacity, P450 metabolism, nutritional status, alcohol and medication use, and GSH metabolism make definition of reproducible toxic dosing parameters impossible.

With reliable information, efforts to establish “the toxic dose” would be difficult; with the available information it is all but impossible. Many “therapeutic misadventures” turn out to be occult intentional overdoses; histories of massive accidental overdosing may evolve into histories of recommended dosing (or vice versa) through unintentional or intentional actions of health care providers, patients, family members, or attorneys with a variety of motivations.(5) Such dosing information has been published in several major reports without evidence of credible verification. (1,6,7)

Nothing published thus far includes a reasonable effort at dose validation, a reasonable effort at case controls, and a reasonable effort at appropriate case definition. As a result, nothing published this far sheds a great deal of light on the subject of the dose-toxicity relationship. This fact and the myriad dosing and metabolic variations described above should tell us that there is little to be gained from trying to establish “the toxic dose”.

Instead, if we accept that there is certain toxicity at some “high-end” dosing pattern, certain safety at some “low-end” pattern, some risk threshold in between and that this threshold cannot be precisely defined, we can move on to a more practical approach to the analysis.

Question #3. Is there a risk of toxicity from APAP taken according to current recommendations? 

No (practically speaking). This question has important implications. What do we advise patients in general? What about alcoholics, those with other chronic illnesses or taking other medications? Should they avoid APAP? Should they take less? It is on this question that critical thinking and academic rigor have most noticeably vanished, hidden in the cloud created by worse-than-imperfect information. If, as suggested in some reports, recommended therapeutic dosing of APAP commonly leads to consequential liver injury, liver failure and death, the cllinical and economic implications are profound. Because of its importance, one would expect that the APAP dose would receive the utmost scrutiny and skepticism in these studies, as should be the case for any critical variable in any study. Instead, the conclusion that toxicity is not unusual after recommended doses has appeared in reputable journals on the basis of data failing to meet any acceptable standards of validation or verification. (1,6,7) The “snowball” effect on subsequent published reports and on how clinicians evaluate or fail to evaluate patients is evident. Despite lack of substantiation, it is now likely that many practitioners, public, and press; nearly all plantiffs attorneys; and perhaps many medical toxicologists accept that toxicity after recommended doses of APAP is a common enough problem that dosing recommendations should change. To look at this complex issue, a series of sub-questions need to be addressed.

Question #3a: Are recommended doses a concern in normal, healthy adults?

No. Both literature review (3) and epidemiology are reassuring. Literature review tells us that reported cases of serious toxicity after recommended dosing essentially do not exist without alcohol abuse or other risk factors. Epidemiology tells us that even if there were a small number of reported cases, it would not change the answer to this question. Consider that there are 200,000,000 adults in the United States alone, and that APAP is used extensively worldwide. Multiply this by the 20 or so years of popular APAP use and it is remarkable that there are not many reports of toxicity merely on the basis of chance.

Question #3b: Are recommended APAP doses a concern in alcoholics?

No (almost certainly). On the basis of marked over representation in case reports and series, (3,6) and consistent with indirect evidence, (8-11) it is likely that alcoholics are predisposed to APAP toxicity after repeated dosing. This conclusion is primarily based on the apparently higher incidence of toxicity after excessive dosing in alcoholics compared to others. (2,3,6) Despite the growing number of reports of APAP toxicity in alcoholics after recommended doses, (6,7) reasons to be skeptical go beyond the aforementioned inadequacy of valid dosing information.

Even if published dosing histories could be validated, linking them to actual APAP-induced toxicity has been complicated by failure to include either adequate case controls or adequate case definitions. Simply using the presence of a common drug like APAP or the presence of severe hepatotoxicity is insufficient to identify APAP-induced toxicity. One must also look at the prevalence of APAP use among those without hepatotoxicity and look at the prevalence of severe hepatotoxicity in those not using APAP; neither has been done. Selection bias occurs in both directions; attribution of toxicity to APAP or attribution of toxicity to ethanol alone without adequate validation of either. These limitations may account for why one large series may have no cases (2,12) of toxicity after appropriate dosing and another notes a significant APAP risk to public health. (7)

For the sake of argument, even if one accepts that all reports of low-dose APAP toxicity in alcohol users are true, epidemiologic considerations still suggest that no change in routine dosing recommendations is warranted for alcoholics. Estimates of alcohol use and alcoholism suggest, in the US alone, that there are at least 5,000,000 to 10,000,000 alcoholics and twice that number of "problem drinkers". In the published reports, and in the courtroom, it has been claimed that even modest use of ethanol, (6) or less constitutes an amount of alcohol use that creates risk from recommended doses of APAP. If so, applying this to the worldwide population during more than 20 years of APAP use, with several APAP uses per year in some, the number of potentially risky APAP dosing episodes must number in the many, many billions. If every reported case were true, and if these were only small fraction of the actual cases, it would still suggest an incidence so low that any large hospital could expect to treat all of their alcoholic patients with APAP for decades without ever seeing a case of resultant toxicity.

Expressing the risk in these terms is important. It suggests that the risk is probably less than that of being killed by lightning, and less than that of getting cancer from a lifetime of eating peanut butter. (14) It does not suggest that it is completely impossible; only that there is no meaningful chance that any individual alcoholic that we treat will develop hepatotoxicity if recommendations are followed. To avoid APAP or change to lower dosing limits on the basis of these data is simply unfounded.

Question #3c: Are recommended APAP doses a concern in anyone?

Probably not. In light of the enormous use of APAP and the paucity of evidence for low dose toxicity, if risk of such toxicity exists it is unlikely to be defined by any single common trait such as ethanol use. It is far more conceivable that unusual characteristics or combinations of characteristics could result in risk. So much noise has been made about alcohol-APAP interaction that little consideration has been given to other possible risks. Unlike ethanol use, with numbers large enough to provide epidemiologic reassurance, it is unknown whether outliers exist with risk factors such that APAP avoidance or dose reduction is rational. Examples of suggested factors to consider alone, in combination, or in combination with ethanol include use of antituberculous medications (15-17) or anticonvulsants, (18) starvation, (12) cachexia, HIV disease, and concomitant infectious or alcoholic hepatitis. Genetically-determined unusual APAP metabolic profiles, infections with known and newly recognized agents, and a variety of medication interactions also deserve scrutiny on theoretical grounds. (19,20) At present, none of these factors has been shown to warrant routine APAP avoidance or dose reduction.

Question #4: What should we tell our patients?

Follow directions. Using Tylenol as an example, current adult APAP dosing recommendations state that patients should take APAP up to 1 gm every four hours to a maximum of 4 gm in 24 hours, or as directed by a physician.(21) In addition, patients are instructed to discontinue dosing after 3 days for fever or 10 days for pain, unless directed to do otherwise by a physician. As noted, if recommended doses ever result in toxicity, it must be vanishingly rare. Therefore, simply reinforcing that excessive dosing is not beneficial and may be harmful, and reminding patients to be sure that other APAP-containing products are not concurrently used would be expected to eliminate nearly all cases of toxicity.

Question #5: What should be done to detect toxicity?

Advice to simply follow directions is, of course, not good enough in medical practice. An appropriate approach must protect those who intentionally or unintentionally fail to follow recommendations, and outliers with recognized or unrecognized extraordinary risk factors. If attempting to control the dose, dosing interval and dosing duration proves inadequate to prevent toxicity, the next line of defense is early recognition of toxicity, followed by discontinuation of APAP, and in most cases of suspected toxicity, treatment with NAC. The importance of these seemingly obvious steps may be under appreciated.

Unlike the acute overdose setting in which the entire dose has been ingested prior to care and, obviously, prior to toxicity, many repeated-dosing cases continue dosing after contact with a health care provider, and perhaps more importantly continue dosing after there is evidence of toxicity. Days of nausea, vomiting and other symptoms have often been evident. At time of recognition, advanced fulminant liver failure is commonly seen, suggesting substantial time since the onset of injury. Other clinical features, such as unusually high AST and AST/ALT ratios (6) may suggest that the pattern of toxicity is atypical in these cases; continued dosing after toxicity is one of several causal possibilities. Once toxicity is suspected, current evidence supports the use of NAC since apparent benefit has been shown even when serious hepatotoxicity is already evident. (22,23) Together, these observations offer hope that earlier recognition and treatment might substantially lessen morbidity and mortality.

Focusing attention and study on prevention and surveillance rather than on dosing amount is a fundamental shift which needs to be debated and studied. If repeated dose APAP toxicity is largely an issue of dose-related toxicity followed by inevitable consequences, then diminished and selective use of APAP would be appropriate in order to prevent even rare tragedies. If instead, patient instruction, surveillance, and treatment effectively prevents consequential injury, then no change in APAP dosing is warranted. Although not proven, current information suggests that an information, surveillance and early treatment strategy is more justified. (24)

Question #6: Where do we go from here?

Hopefully, the ensuing dialogue will lead to focused study which will include, at least, the following: a valid case definition of repeat dose APAP toxicity, appropriate practical standards for attempting dose verification, suggestions for appropriate case controls, large scale prospective monitoring of at risk patients, consideration of how preexisting or concurrent conditions might mimic toxicity or alter the course and manifestations of toxicity, and renewed efforts to establish and validate a marker for which precedes injury but indicates risk threshold.

* Dr. Smilkstein has served as a scientific and legal consultant for McNeil Consumer Products.


  1. Heubi JE, Barbacci MB, Zimmerman HJ: Therapeutic misadventures with acetaminophen: Hepatoxicity after multiple doses in children. J Pediatrics 1998;132:22-27.
  2. Makin AJ, Wendon J, Williams R: A 7-year experience of severe acetaminophen-induced hepatotoxicity (1987-1993). Gastroenterology 1995;109:1907-1916.
  3. Prescott LF: Adverse reactions and interactions. In Paracetamol (acetaminophen). A Critical Bibliographic Review, London, Taylor & Francis, 1996, pp 353-400.
  4. Mitchell JR, Jollow DJ, Potter WZ, Gillette JR, Brodie BB: Acetaminophen-induced hepatic necrosis. IV. Protective role of glutathione. J Pharmacol Exp Ther 1973;187:211-217.
  5. Litovitz TL, Smilkstein MJ, Felberg L, Klein-Schwartz W, Berlin R, Morgan J: 1996 Annual Report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med 1997;15:447-500 (case 184).
  6. Zimmerman HJ, Maddrey WC: Acetaminophen (paracetamol) hepatotoxicity with regular intake of alcohol: analysis of instances of therapeutic misadventure. Hepatology 1995;22:767-773.
  7. Schiodt FV, Rochling FA, Casey DL, Lee WM: Acetaminophen toxicity in an urban hospital. N Engl J Med 1997;337:1112-1117.
  8. Thummel KE, Slattery JT, Nelson SD, Lee CA, Pearson PG: Effect of ethanol on hepatotoxicity of acetaminophen in mice and on reactive metabolite formation by mouse and human liver microsomes. Toxicol Applied Pharmacol 1989;100:391-397.
  9. Tredger JM, Smith HM, Read RB, Williams R: Effects of ethanol ingestion on the metabolism of a hepatotoxic dose or paracetamol in mice. Xenobiotica 1986;16:661-670.
  10. Raucy JL, Sker JML, Lieber CS, Black M: Acetaminophen activation by human liver cytochromes P-450 IIE1 and P-450 IA2. Arch Biochem Biophysics 1989;271:270-283.
  11. Lauterburg BH, Velez ME: Glutathione deficiency in alcoholics: risk factor for paracetamol hepatotoxicity. Gut 1988;29:1153-1157.
  12. Whitcomb DC, Block GD: Association of acetaminophen hepatotoxicity with fasting and ethanol use. JAMA 1994;272:1845-1850.
  13. Wilson R, Crouch EAC: Risk assessment and comparisons: an introduction. Science 1987;236:267-270.
  14. Epstein MM, Nelson SD Slattery JT, Kalhorn TF, Wall RA, Wright JM: Inhibition of the metabolism of paracetamol by isoniazid. Br J Clin Pharmacol 1991;31:139-142.
  15. Murphy R, Swartz R, Watkins PB: Severe acetaminophen toxicity in a patient receiving isoniazid. Ann Intern Med 1990;113:799-800.
  16. Zand R, Nelson SD, Slattery JT, et al: Inhibition and induction of cytochrome P4502E1- catalyzed oxidation by isoniazid in humans. Clin Pharmacol Ther 1993;54:142-149.
  17. Bray GP, Harrison PM, O'Grady JG, Tredger JM, Williams R: Long-term anticonvulsant therapy worsens outcome in paracetamol-induced fulminant hepatic failure. Hum ExpToxicol 1992;11:265-270.
  18. Kearns GL, Leeder JS, Wasserman GS: Acetaminophen overdose with therapeutic intent. J Pediatrics 1998;132:5-8.
  19. Prescott LF: Factors influencing paracetamol metabolism. In Paracetamol (acetaminophen). A Critical Bibliographic Review, London, Taylor & Francis, 1996, pp 103-106.
  20. McNeil Consumer Products: Package label, Extra Strength Tylenol. McNeil-PPC, Fort Washington, PA.
  21. Keays R, Harrison PM, Wendon JA, et al: Intravenous acetylcysteine in paracetamol induced fulminant hepatic failure: a prospective controlled trial. Br Med J 1991;303:1026-1029.
  22. Wendon JA, Harrison PM, Keays R, Williams R: Cerebral blood flow and metabolism in fulminant liver failure. Hepatology 1994;19:1407-1413.
  23. Smilkstein MJ: Acetaminophen. In Goldfrank LR, Flomenbaum NA, Lewin NE, Howland MA, Weisman RS, Hoffman RS, eds, Goldfrank's Toxicologic Emergencies (6th ed). Appleton & Lange, Norwalk, CT, 1998, in press.


Int J Med Toxicol 1998; 1(3): 19

This article is located at http://www.ijmt.net/ijmt/1_3/1_3_19.html

Quick Survey
Please rate this article:
1 2 3 4 5
(5 is best)

IJMT Home | Current Issue | Past Issues | Search | Technical Support | Send Comments to ACMTNet

Copyright 1999-2003, American College of Medical Toxicology.