Table of Contents
Warfarin and Coumarin Anticoagulants
Warfarin and other coumarins including those found in rodenticides. This includes long acting coumarins:
Warfarin, Brodifacoum, Difenacoum, Bromadiolone, Chlorophacinone, Diphacinone, Flocoumafen, Pindone, Valone, Coumatetralyl
Coumarins are anticoagulants. Some coumarins used in rodenticides are extremely long acting. It should be noted that some rodenticide products contain both short-acting warfarin and longer-acting superwarfarins. The major clinical problem is preventing overcorrection in patients who have an ongoing need for anticoagulation. Treatment of overanticoagulated patients is best achieved by careful titration of low doses of intravenous vitamin K against serial international normalised ratios (INR). Fresh frozen plasma is indicated when there is active bleeding. Long-acting anticoagulant rodenticides (LAAR) are produced by substituting phenyl groups on warfarin and are 100 times more potent than short-acting warfarin.
Coumarin use in the first trimester may cause foetal warfarin syndrome. After the first trimester, exposure increases the risk of CNS defects, possibly due to foetal haemorrhage.
The single acutely toxic oral dose of warfarin is thought to be 0.5 mg/kg and above. Long acting coumarins may require supplementary Vitamin K for a prolonged time (weeks to months).
There are a number of scenarios that must be considered for anticoagulant poisoning:
- Over-anticoagulation in patients on therapeutic warfarin
- Acute warfarin overdose
- Long-acting oral anticoagulant overdose (eg. brodifacoum)
- Unintentional poisoning with rodenticides in children
MECHANISM OF TOXIC EFFECTS
Coumarin anticoagulants inhibit recycling of vitamin K from its (inactive) metabolite, vitamin K-2,3-epoxide. Without reactivation the body is dependent upon exogenous supply of the Vitamin K that is required for converting clotting factor precursors to the clotting factors II, VII, IX, and X.
Warfarin inhibits the g-carboxylation of several glutamine residues in factors II, VII, IX, X, as well as protein C. The blockade of this produces inactive clotting factors. The protein carboxylation is coupled with the oxidative deactivation of vitamin K. The anticoagulant prevents the reductive metabolism of the inactive vitamin K epoxide back to its hydroquinone form.
There is considerable delay in the anticoagulant effect following the administration of warfarin. The levels of factor VII and protein C fall first. Factor VII is of limited relevance to intravascular hypercoagulability, so in the first 24 hours there is a slightly hypercoagulable state secondary to the reduction in protein C. The peak hypoprothrombinaemic effect takes about 1‑3 days. The time to this is affected to some extent by the loading dose of warfarin, up to 0.75 mg/kg, but above this dose the speed of onset is independent to the initial dose.
Anticoagulant effect depends upon the half-life of the clotting factors.
- 6-8 hours (factor VII)
- 6-8 hours (protein C)
- 24 hours (factor IX)
- 40 hours (factor X)
- 60 hours (factor II)
Patients who are already anti-coagulated (therapeutic INR) have at most 25% of normal clotting factor activity and thus increased sensitivity to small increments in warfarin concentration.
KINETICS IN OVERDOSE
Completely absorbed after oral administration. May also be after inhalation. Warfarin is significantly absorbed through the skin, either in liquid or powder forms.
Plasma protein binding 99%.
Metabolism - Elimination
Hepatic P450 mediated metabolism, metabolites inactive. No significant renal elimination. Significant enterohepatic circulation. Half life varies from 15 to 70 h for warfarin to weeks for some rodenticides. Half life of superwarfarins (Brodifacoum) is around 10-70 days.
Seek evidence for or confirmation of visible or occult bleeding, including assessment of circulatory status, and any evidence of airway obstruction secondary to haemorrhage. Inspect accessible mucosal surfaces and skin for signs of haemorrhage, and test for haematuria +/- faecal occult blood.
All signs are secondary to the anticoagulant effects and depend upon the site of bleeding. Morbidity is highest with cerebral bleeding.
In patients who are not already anticoagulated, changes in the Prothrombin Time/International Normalised Ratio may begin after 12 hours but are more usually are observed 24 - 36 hours post ingestion. Patients who are already anticoagulated may show changes within 4 hours.
Abdominal pain (NB retroperitoneal bleeding), vomiting (including haematemesis), gastrointestinal bleeding.
Central nervous system effects
Intracranial haemorrhage and blood in the cerebrospinal fluid may occur. Symptoms range from headache to coma, focal signs depend upon the site of bleeding.
Hypotension may result from haemorrhage.
Haemorrhage in the upper airways may be manifest as pain, dysphonia, dysphagia, dyspnoea, haemoptysis and excessive bronchial secretions. Alveolar haemorrhage has been reported, as has haemothorax.
Bleeding is the most common sign of anticoagulant exposure. Prothrombin Time/International Normalised Ratio changes may begin within 12 hours but are more usually are observed 24 - 36 hours post ingestion. All signs are secondary to the anticoagulant effects.
Other manifestations of bleeding include bruising, haematoma, epistaxis (bleeding nose), bleeding gums, haematuria, vaginal bleeding, haematospermia and subconjunctival bleeding.
DETERMINATION OF SEVERITY
The single acutely toxic oral dose of warfarin is thought to be 0.5 mg/kg and above. Doses of 0.5 mg/kg and above warrant hospital observation and monitoring. The lowest quoted one-off fatal dose is 6.667 mg/kg.
From approximate theoretical calculations, one-off doses ranging between 0.075 - 0.48 mg/kg in adults have been estimated as likely to cause a prothrombin ratio (approximating the INR) of no higher than 1.5 - 2.5; which is a therapeutic anticoagulant effect (i.e. for patients on therapeutic anticoagulants target INR values are 2 - 3 and bleeding becomes a risk at levels> 5. Thus, doses much above this range are of concern (some disturbance of clotting factor synthesis can be seen at lower concentrations, but these are not usually clinically significant or even reflected in any changed PT).
Chronic anticoagulant exposure when the same total amount is given in divided doses over several days is more hazardous than as a single one-off dose. Persons already on anticoagulant therapy are at increased risk from a given amount.
An activated partial thromboplastin time (aPTT) greater than twice normal may be useful as a predictor of increased vitamin K requirements.
- Monitor for coagulation disturbance by measuring the prothrombin time (PT, PCT) and/or INR (International Normalised Ratio)
- Also standard haematology screen and ideally the aPTT
- Do PT and/or INR estimations on admission if patient is already on anticoagulants (i.e. acute-on-chronic)
A baseline PT/INR has also been recommended for one-off intentional ingestions where the dose is unknown. Monitoring frequency is controversial but depressed INR can occasionally be detected by 12 hours, and often by 24 hours and checks at this time should be made for all ingestions over 0.5 mg/kg. Thereafter frequency of monitoring should depend on laboratory and clinical findings, typically every 24 hours until any abnormality is reversed. In patients already anticoagulated, significant trends in the INR could theoretically be apparent within 4 hours.
NOTE: Individual factor assays are more sensitive tests than the PT or INR.
In superwarfarin poisoning, brodifacoum levels may be useful in monitoring long-term therapy and determining duration of vitamin K treatment. Several studies suggest that vitamin K therapy is not required after brodifacoum levels drop below 10ng/mL.
Any medical cause of a coagulopathy, toxicological causes include snake bite and any cause of hepatic failure.
Single dose activated charcoal for acute ingestions. Repeat doses of activated charcoal is likely to reduce the half-life and could be considered for large overdoses. Cholestyramine has been shown to reduce half life presumably by interrupting enterohepatic circulation
Assessment for treatment
Treatment requires an assessment of the risk of anticoagulant complications balanced against the risk of thromboembolism (in those patients already anticoagulated). The 2 main interventions are fresh frozen plasma (FFP) and vitamin K.
For long-acting coumarins, it is our practice to withhold Vitamin K until there is evidence of anticoagulation that would indicate the need for more prolonged treatment and monitoring. If the INR is normal at 48 hours then further treatment or investigation is unnecessary. If the INR is abnormal but less than 2.0 at 48 hours then a further INR should be done at 96 hours. If the patient becomes anticoagulated then initial treatment is with adequate doses of IV vitamin K (see below) followed, in the case of long-acting coumarins, by oral vitamin K in the maintenance period.
In the relatively rare scenario of severe haemorrhage, rapid correction of coagulopathy with FFP should be undertaken. This should be followed by an appropriate dose of Vitamin K.
See dose table for initial treatment aimed at rapid reversal of anticoagulant effect, using INR as a guide: it is most applicable to patients who have an ongoing need for anticoagulation.
More commonly, the patient presents with a history of ingestion and/or an asymptomatic coagulopathy.
No current anticoagulant treatment, no coagulopathy
In paediatric, accidental possible ingestion, the risk is likely to be so low that neither treatment nor INR is required. In adults, where the possible ingestion is likely to be deliberate give Vitamin K 10 mg, INR may be done at 24 hours (but probably not required). Paediatric ingestion of under one box of rodenticide is thought to not require hospitalisation in one large study (Ingels et al.)
Even in a definite ingestion of a long acting coumarin in children, poison centre data suggest that the risk is so low that no treatment or investigation is required. in short acting coumarins there are less data but neither treatment nor monitoring is required for ingestions of less than 0.5 mg/kg of warfarin. In adults, give Vitamin K 10 mg and check INR in 24 hours.
Current anticoagulant treatment
Assess thromboembolic risk in all patients. Patients who are at high or moderate risk will require alternative anticoagulation if overcorrected.
Acute ingestion on chronic treatment
INR therapeutic: Repeat INR 4th hourly, if rising then give Vitamin K.
INR above therapeutic: Give an appropriate dose of Vitamin K and repeat INR.
Coagulopathy on chronic treatment
Give an appropriate dose of Vitamin K and repeat INR Assess for precipitant causes such as drug interactions or intercurrent illness.
The predictive value for the risk of haemorrhage of the time the INR is above various values has been established in many studies looking at chronic anticoagulant therapy. Implicit in such an approach is the time spent with a prolonged INR. Such studies cannot be directly generalised to the situation of acute poisoning. However, the risk of thromboembolism, especially for high risk patients, is less time dependent.
Vitamin K1 (phytomenadione) is indicated in all symptomatic cases of bleeding or with a reliable history of a frankly toxic dose. In other cases, its use depends on the results of PT/INR investigation at possibly 12 hours but especially later tests, which should be done at 24 and 48 hours. Rapid reversal of anticoagulant effect using parenteral vitamin K1 should not be necessary unless the INR is greater than 6.0.
It should be noted that the only effective Vitamin K analogue is K1. Other vitamin K analogues such as menadione (vitamin K3) should NOT be used for treatment of anticoagulant overdose. They are NOT effective (do not have the right length side chains) and reliance on treatment with these other analogues can result in serious consequences.
In acute on chronic (i.e. maintenance) patients, where Vitamin K1 use must be cautious to avoid overcorrection, or where there is life threatening bleeding, administration of fresh frozen plasma (FFP) is preferred. In these cases, check PT or INR 6 hours after administration of FFP.
See dose table for initial treatment for rapid reversal of anticoagulant effect, using INR as a guide: it is most applicable to patients who have an ongoing need for anticoagulation.
Vitamin K can be administered IV, SC, IM and orally. However, IV administration is the route of choice for rapid reversal of anticoagulant effect when it should be given as a slow IV injection (<1 mg/minute).
Oral vitamin K could be considered where there are minor changes in INR, but may not be necessary. In such cases, review after 24 or 48 hours is indicated. Oral administration may also be required to maintain a satisfactory INR after initial stabilisation of long-acting superwarfarin poisoning.
Oral preparations available in Australia which the authors have used successfully in the treatment of superwarfarin poisoning include:
- Nutrition Care Vitamin K1 10Mg Capsules 100
- KMav Vit K1 10mg or 50mg Tablets (this is a veterinary product)
Parenteral (i.e. IV, SC, IM) doses up to 25 mg (and rarely even 50 mg) have been administered. The turnover rate of vitamin K is at least 30% per hour, thus repeat dosing is often required. There is also a delay in the onset of effective action of Vitamin K, with significant increased concentrations of clotting factors II, VII, IX, X only appearing 6 - 8 hours post-administration.
Vitamin K dose
|2.0 - 5.0||No||Rapid reversal not indicated|
|5.0 - 9.5||No||Vitamin K 0.5 mg slow IV|
|9.5 - 20.0||No||Vitamin K 2 - 5 mg slow IV . Check INR every 6 hours|
|>20.0||None or any||Vitamin K 10 mg slow IV Check INR every 6 hours|
|Any||Life threatening||Fresh frozen plasma or factor concentrates Vitamin K 10 mg IV. INR every 6 hours|
Rarely, intravenous administration of vitamin K can produce severe, shock-like reactions. Such a reaction is generally as a result of too rapid administration (> 1 mg/minute). Symptoms include flushed face, sweating, feeling of chest constriction, cyanosis and peripheral vascular collapse. This is appears to be less likely with the new formulations Very rarely, local venous irritation, or phlebitis, may occur.
Fresh whole blood or plasma (including fresh frozen plasma) can be used to replace clotting factors where severe haemorrhage is present, or if emergency surgery is required. Exchange transfusion is a further option. It has been recommended particularly for patients on pre-existing maintenance anticoagulants for therapeutic reasons, because of the potential danger of overcorrection with excess vitamin K and an increased risk of clotting. In such patients vitamin K should not generally be given unless anticoagulation is severe, and doses should aim to return INR to therapeutic, not normal concentrations.
In the presence of evidence of capillary damage (which can occur at higher doses) vitamin C has been advocated.
The site and nature of complications define prognosis. Patients who have had a coagulopathy should have their INR repeated until it is stabilised. Coagulopathies due to LAARs may require monitoring (and vitamin K treatment) for months.
Warfarin-induced skin necrosis and venous limb gangrene in the setting of heparin-induced thrombocytopenia. Srinivasan AF, Rice L, Bartholomew JR, Rangaswamy C, La Perna L, Thompson JE, Murphy S, Baker KR.Arch Intern Med. 2004 Jan 12;164(1):66-70.
Palareti G, Leali N, Coccheri S, Poggi M, Manotti C, D'Angelo A, Pengo V, Erba N, Moia M, Ciavarella N, Devoto G, Berrettini M, Musolesi S.Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy. Lancet. 1996 Aug 17;348(9025):423-8.
Isbister GK, Hackett LP, Whyte IM.Intentional warfarin overdose.Ther Drug Monit. 2003 Dec;25(6):715-22.
Hung A, Singh S, Tait RC, A prospective randomized study to determine the optimal dose of intravenous vitamin K in reversal of over-warfarinization. Br J Haematol 2000 Jun;109(3):537-9
Kumer K, Nwangwu PV. Clinical toxicology of warfarin from commercial rodenticides: symptoms, diagnosis and management. Clin Toxicol Consultant 1981;3: 23-7.
Smolinske SC, Scherger DL, Kearns PS, Wruk KM, Kulig KW, Rumack BH. Superwarfarin poisoning in children: a prospective study. Pediatrics 1989; 84: 490-4.
Kruse JA, Carlson RW. Fatal rodenticide poisoning with brodifacoum. Ann Emerg Med 1992; 21: 331-6.
Sheen SR, Spiller HA, Grossman D. Symptomatic brodifacoum ingestion requiring high-dose phytonadione therapy. Vet Hum Toxicol 1994; 36: 216-7.
Morgan B, Tomaszewski C. Spontaneous hemoperitoneum from brodifcoum overdose (abstract). J Toxicol Clin Toxicol 1995; 33: 549.
Helmuth RA, et al Fatal ingestion of a brodifacoum-containing rodenticide.Lab Med 1989; 20: 25-7.
Liebeschuetz S, Baddeley H, Thomas DJB. Haematomyelia following warfarin therapy. Br J Clin Pharm 1994; 48: 105-6.
Travis SF, Warfield W, Greenbaum BH, Molokisher M, Siegel JE. Clinical and laboratory observations: Spontaneous hemorrhage associated with accidental brodifacoum poisoning in a child. J Pediatr 1993; 122: 982-4.
Barnett VT, Bergmann F, Humphrey H, Chediak J Diffuse alveolar hemorrhage secondary to superwarfarin ingestion. Chest 1992; 102: 1301-2.
Boster SR, Bergin JJ. Upper airway obstruction complicating warfarin therapy - with a note on reversal of warfarin toxicity. Ann Emerg Med 1983; 12:711-5.
Hollinger BR, Pastoor TP. Case management and plasma half-life in a case of brodifacoum poisoning. Arch Intern Med 1993; 153: 1925-8.
La Rosa FG, Clarke SH, Lefkovitz JB. Brodifacoum intoxication with marijuana smoking. Arch Path Lab Med 1997; 121(1): 67-9.
Lipton RA, Klass EM. Human ingestion of a “superwarfarin” rodenticide resulting in a prolonged anticoagulant effect. JAMA 1984; 252: 3004-5.
Bachmann KA, Sullivan TJ. Dispositional and pharmacodynamic characteristics of brodifacoum in warfarin-sensitive rats. Pharmacology. 1983;27(5):281-8.
Weitzel JN, Sadowski JA, Furie BC, Moroose R, Kim H, Mount ME, Murphy MJ, Furie B. Surreptitious ingestion of a long-acting vitamin K antagonist/rodenticide, brodifacoum: clinical and metabolic studies of three cases.
Isbister GK, Dawson A, Isbister JP. Recommendations for the management of over-anticoagulation with warfarin. Emerg Med (Fremantle) 2001 Dec;13(4):469-71
Shepherd G, Klein-Schwartz W, Anderson BD. Acute, unintentional pediatric brodifacoum ingestions. Pediatr Emerg Care. 2002 Jun;18(3):174-8.
Buitenhuis HC, Soute BA, Vermeer C. Comparison of the vitamins K1, K2 and K3 as cofactors for the hepatic vitamin K-dependent carboxylase. Biochim Biophys Acta. 1990 May 16;1034(2):170-5.
Spahr JE, Maul JS, Rodgers GM. Superwarfarin poisoning: a report of two cases and review of the literature. Am J Hematol. 2007 Jul;82(7):656-60.
Bruno GR, Howland MA, McMeeking A, Hoffman RS. Long-acting anticoagulant overdose: brodifacoum kinetics and optimal vitamin K dosing. Ann Emerg Med. 2000 Sep;36(3):262-7.
Ingels M, Lai C, Tai W, Manning BH, Rangan C, Williams SR, Manoguerra AS, Albertson T, Clark RF. A prospective study of acute, unintentional, pediatric superwarfarin ingestions managed without decontamination. Ann Emerg Med. 2002 Jul;40(1):73-8.