Sandra E. McConkey, DVM, PhD, DACVP; Alastair Cribb, DVM, PhD, FCAHS
Acetaminophen (APAP) is a commonly used human analgesic and antipyretic. Acetaminophen is the most frequent drug overdose reported to human poison control centers in the United States and Britain.(1,2) Veterinary poison control centers also report numerous cases of APAP toxicity in dogs and cats. In humans and most laboratory species, APAP toxicity is associated with hepatotoxicity characterized by centrolobular necrosis with clinical signs of abdominal pain, icterus and vomiting occurring at 48 hours post intoxication.(3,4) Dogs and cats are unique in that they primarily develop methemoglobinemia and hemolytic anemia with clinical signs of cyanosis, dyspnea, facial edema, depression and vomiting occurring at 2-4 hours post intoxication.(6,8) Methemoglobinemia can be > 45% within the first 2-4 hours of intoxication.(5) The initial intravascular hemolysis is mild and many animals continue to have a HCT within the normal reference range for the first 24 hours. Marked oxidative hemolysis characterized by Heinz bodies can occur at 48-72 hours post intoxication.(6,7) Dogs can develop hepatotoxicity due to centrolobular necrosis if they survive the initial hematotoxicity. Cats can develop elevated ALP and ALT activities, but these appear to be associated with generalized hepatocellular degeneration rather than centrolobular necrosis.(9) The feline and canine toxic APAP doses are 60 and 200 mg/kg respectively.(5,10)
Metabolism of Acetaminophen
There are significant species differences in acetaminophen metabolism, but the general pathways are similar. In most species, acetaminophen is metabolized predominantly by sulfation and glucuronidation in the liver followed by renal excretion of the conjugates.(4,5,11) See Figure 1. A small percentage of APAP is oxidized by cytochrome P450 (CYP) enzymes to the reactive metabolite N-acetyl-p-benzoquinoneimine (NAPQI).(11) At therapeutic doses of APAP, NAPQI binds to glutathione (GSH) and is then excreted in the urine as cysteine and mercapturic acid metabolites. At toxic doses, sulfate and glucuronosyl transferases become saturated and increased NAPQI production occurs.(3) If GSH is depleted to < 20% of its usual concentration, NAPQI binds covalently to cysteine groups on other hepatocellular proteins leading to cell death.(12) Species such as humans and hamsters are highly sensitive to APAP-induced hepatotoxicity while other species such as rats are more resistant. The difference in species sensitivity to APAP hepatotoxicity is attributed primarily to variation in NAPQI production. Species that are highly sensitive to APAP-induced hepatotoxicity produce more NAPQI and subsequently show greater covalent binding to hepatocellular proteins.(3,13) Species other than cats and dogs do not develop hematotoxicity regardless of how high the dosage of APAP, except for individuals with glucose-6-phosphate dehydrogenase deficiency (G6PD).
Click on the image to see a larger view.
|Figure 1. Metabolism of acetaminophen.|
Acetaminophen Toxicity in Dogs and Cats
Why is there a different toxic response to APAP toxicity in dogs and cats? The metabolite responsible for the APAP-induced hematotoxicity in dogs and cats has never been directly demonstrated. In veterinary literature, the reaction is often attributed to NAPQI.(14) It has been postulated that cats may produce more NAPQI than other species as they have limited APAP glucuronidation due to a feline pseudogene for uridine-diphosphate (UDG)-glucuronosyl transferase 1A6 (UGT1A6).(15) Feline metabolism of APAP is primarily by sulfation (92%), with minimal glucuronidation (1.3%) and oxidation (4.7%) whereas human APAP metabolism is primarily by glucuronidation (47-62%) with lesser amounts of sulfation (25-36%) and oxidation (5-8%).(5) It has been proposed that the red cell sensitivity in cats is due to high numbers of exposed sulfhydryl groups on feline hemoglobin predisposing feline hemoglobin to oxidation and subsequent Heinz body formation.(14) Cats have eight sulfhydryl groups per hemoglobin molecule as opposed to two in humans.(16,17)
While the limited glucuronidation and predisposition of feline erythrocytes to oxidation may contribute to the exquisite sensitivity of cats to APAP hematotoxicity, there would also appear to be other contributing factors. Dogs also develop APAP-induced hematotoxicity. Unlike cats, glucuronidation is the predominant biotransformation pathway (76%) in canine APAP metabolism.(5) In addition, canine hemoglobin has four reactive sulfhydryl groups per molecule and is less sensitive to oxidation than feline hemoglobin.(16)
There are also flaws in the hypothesis that NAPQI is the reactive metabolite responsible for APAP-induced hematotoxicity. First, NAPQI is not likely produced in erythrocytes as red blood cells lack CYP enzymes. Second, NAPQI is not released into the blood from hepatocytes. Minute quantities of cysteine-bound NAPQI (which is no longer chemically reactive) can be detected in the blood immediately after APAP overdose, but the majority of protein and cysteine-bound NAPQI does not enter the blood until significant hepatocellular degeneration or necrosis occurs 6-12 hours post intoxication.(12,18,19) Clinical signs of methemoglobinemia in dogs and cats typically occur within 2-4 hours of intoxication.(6)
Third, NAPQI doesn't have the chemical characteristics necessary to redox cycle effectively with oxyhemoglobin. Many chemicals known to induce methemoglobinemia do so by co-oxidation and redox cycling with oxyhemoglobin. Examples include dapsone hydroxylamine, phenylhydroxylamine and sulfamethoxazole hydroxylamine. In co-oxidation, an electron is lost from both the chemical and the Fe2+ in the oxyhemoglobin resulting in the production of a nitroso compound, methemoglobin and reactive oxygen species such as H2O2. The co-oxidation reaction occurs repeatedly due to redox cycling with repeated reduction of the nitroso containing metabolite by GSH followed by repeated co-oxidation, until the chemical is removed by alternative biotransformation pathways. Covalent binding of NAPQI with GSH or hepatocellular proteins produces stable conjugates, effectively preventing NAPQI from redox cycling.
A minor APAP metabolite that can co-oxidize and redox cycle with oxyhemoglobin is para-aminophenol (PAP). See Figure 2. PAP is associated with the methemoglobinemia of aniline dyes.(20,21) It is produced by deacetylation of APAP by hepatic microsomal carboxyesterases and has been demonstrated to be formed in cats, rats and mice.(22,23,24,25) Conjugated PAP and archadonic acid are now believed to bind to cannabinoid receptors in the central nervous system resulting in the analgesia of APAP.(26)
Click on the image to see a larger view.
|Figure 2. Co-oxidation of para-aminophenol and oxyhemoglobin.|
In rats and mice, >70% of PAP is removed by biliary excretion of GSH and N-acetyl conjugates.(24,25) N-acetylation of drugs is catalyzed by two closely related enzymes: N-acetyltransferase 1 and N-acetyltransferase 2 (NAT1 and NAT2; EC22.214.171.124).(27) Most species have 2-3 NAT enzymes but cats only have NAT1 and dogs have no NAT enzymes.(28,29) Thus dogs and cats are deficient in their ability to N-acetylate arylamine-containing compounds. Cats and dogs may subsequently be exposed to more systemic PAP during APAP toxicity and may not remove PAP from redox cycling in erythrocytes as efficiently as other species, thus prolonging methemoglobin formation.
We therefore hypothesized that PAP, not APAP or NAPQI, would induce methemoglobinemia in vitro. Moreover, we hypothesized that there would be species differences in susceptibility to the methemoglobin forming potential of PAP due to deficient N-acetylation in dogs and cats.
Our laboratory compared the in vitro induction of methemoglobin by APAP, NAPQI and PAP. PAP was the only chemical that induced methemoglobin formation. Greater than 50% methemoglobin was produced in canine and feline erythrocytes by 500 µM PAP versus < 3% methemoglobin by 5000 µM APAP and 500 µM NAPQI following 60 minutes incubation. Moreover there was a significant difference in the sensitivity of canine and feline erythrocytes (P<0.001) to the induction of methemoglobin by PAP compared to mouse and rat red blood cells (67%, 61%, 28% and 27% respectively). PAP methemoglobin-induction was compared in intact and lysed erythrocytes. In rats and mice, there was significantly greater methemoglobin generation in lysed erythrocytes (P<0.05), reflecting the loss of reducing and detoxification systems by lysed cells. In contrast, canine and feline erythrocytes had significantly greater methemoglobin in intact cells (P<0.05), indicating that loss of protective mechanisms was outweighed by the loss of a predisposing factor in intact cells in these species. This is supportive of prolonged PAP redox cycling.
There was also significantly greater methemoglobin induction in red blood cells in vitro of NAT1/NAT2 knockout mice than wildtype mice (35% and 28% respectively, P<0.05), supporting the hypothesis that deficient NAT activity may contribute to a species sensitivity to PAP induction of methemoglobin. However, the difference in methemoglobin formation in the knock-out mice was relatively mild, suggesting that there are other contributing factors.
Hepatotoxic doses of APAP in vivo in wildtype and NAT1/NAT2 knockout mice produced no methemoglobinemia, confirming that other factors in addition to N-acetyltransferase deficiency are required to predispose to APAP-induced hematotoxicity. Exposure to hepatotoxic doses of PAP did produce significantly greater methemoglobinemia in knockout mice than wildtype mice (P=0.008) primarily due to greater methemoglobin generation in the knockout females.
In summary, current evidence suggests that PAP is most likely the causative metabolite in APAP-induced methemoglobinemia in cats and dogs. While N-acetyltransferase deficiency likely plays an important role in determining the species sensitivity to APAP hematotoxicity, susceptibility in dogs and cats is clearly multifactorial. This is typical of pharmacogenetic syndromes in general. In very few cases is susceptibility determined by a single enzyme difference. The lack of glucuronidation in cats may contribute to the especially high feline sensitivity to APAP but, not, as previously thought, by resulting in increased NAPQI production, but instead by increasing deacetylation of APAP to PAP. In most species PAP is primarily detoxified by N-acetylation but dogs and cats are deficient in N-acetylation activity. This may result in greater quantities of PAP being released systemically from the liver as well as prolonged redox cycling of PAP and oxyhemoglobin in red blood cells with subsequently greater methemoglobin production. The low methemoglobin reductase activity in canine and feline erythrocytes that has been reported elsewhere (30) limits the ability of these species to efficiently reduce the methemoglobin produced by prolonged co-oxidation. This may very well be the second defect required to create the red blood cell sensitivity in cats and dogs.
The majority of APAP associated hemolysis is due to oxidation and formation of Heinz bodies. Chemicals that co-oxidate with oxyhemoglobin cause Heinz body hemolytic anemia by binding of the oxidized compound to Hgb cysteine groups once cytosolic GSH is exhausted. Free radicals can also contribute to the oxidation if reducing enzymes such as catalase are overwhelmed. Feline erythrocytes may be especially predisposed because of their greater number of sulfhydryl groups.
The current treatment for APAP toxicity in dogs and cats is primarily N-acetylcysteine and supportive therapy.(31) Cimetidine has been advocated by some authors as adjunct therapy. This is based on early studies in laboratory animals and humans that demonstrated that cimetidine, an inhibitor of some CYP enzymes, decreased the production of NAPQI. This would be of benefit to species that develop centrolobular necrosis due to NAPQI however cats have not been shown to develop centrolobular necrosis. Cimetidine is no longer used in treatment of human APAP toxicity as subsequent studies have failed to demonstrate any effect on NAPQI production at therapeutic concentrations. Cimetidine has been shown to decrease glucuronidation and NAT activity in rats.(32,33) Our studies indicate a mild, but significant inhibitory effect by cimetidine on feline NAT activity (P<0.05). Thus cimetidine should not be recommended as an adjunct therapy for APAP toxicosis in cats and dogs. Its use has not been shown to be effective, and may in fact inhibit beneficial biotransformation pathways.
1. Bialas MC, et al. J Accid Emerg Med 1998;15:13.
2. Watson WA, et al. Am J Emerg Med 2004;22:335.
3. Davis DC, et al. Life Sci 1974;14:2099.
4. Prescott LF, et al. Lancet 1971;1:519.
5. Savides MC, et al. Toxicol Appl Pharmacol 1984;74:26.
6. Aronson LR. J Vet Emerg Crit Care 1996;6:65.
7. Gaunt SD, et al. Am J Vet Res 1981;42:1982.
8. MacNaughton SM. CanVet J 2003;44:142.
9. St Omer VV, et al. J AmVet MedAssoc 1980;176:911.
10. Hjelle JJ, et al. 1986 J AmVet Med Assoc 1986;188:742.
11. Davis M, et al. Q J Med 1976;45:181.
12. Pumford NR, et al. Biochem Pharmacol 1990;40:573.
13. Tee LB, et al. Biochem Pharmacol 1987;36:1041.
14. Taylor NS, et al. Compend Contin Educ Pract Vet 2003;12:160.
15. Court MH, et al. Biochem Pharmacol 1997;53:1041.
16. Snow NS. Biochem J 1962;84:360.
17. Riggs AF, et al. J Gen Physiol 1956;39:585.
18. Muldrew KL, et al. Drug Metab Dispos 2002;30:446.
19. Gillette JR, et al. Adv Exp Med Biol 1981;36:931.
20. Bernheim F, et al. J Pharmacol Exp Ther 1937;61:311.
21. Harrison JH, et al. J Pharmacol ExpTher 1986;238:1045.
22. Newton JF, et al. Toxicol Appl Pharmacol 1982;65:336.
23. Welsh RM, et al. Biochem Pharmacol 1966;15:521.
24. Song H, et al. J Biochem Mol Toxicol. 2001;15:34.
25. Klos C, et al. Toxicol Appl Pharmacol 1992;115:98.
26. Bertolini A, et al. CNS Drug Rev 2006;12:250.
27. Grant DM, et al. FEBS Lett 1989;244:203.
28. Trepanier LA, et al. Pharmacogenetics 1998;8:169.
29. Trepanier LA, et al Biochem Pharmacol 1997;54:73.
30. Stolk JM, et al. Biochem.Pharmacol. 1966;5:343.
31. Rumbeiha WK, et al. Am J Vet Res 1995;56:1529.
32. Svensson CK, et al. Drug Metab Dispos 1992;20:74.
33. Emery S, et al. Biochem Pharmacol 1985;34:1415.