Superwarfarin (Long-Acting Anticoagulant Rodenticides) Poisoning: from Pathophysiology to Laboratory-Guided Clinical Management

Abstract

Superwarfarins are long-acting anticoagulant rodenticides developed from warfarin. The mechanism of action is by inhibition of vitamin K epoxide reductase, resulting in the inability of the body to recycle vitamin K. Deficiency of vitamin K thereafter leads to inability for the body to synthesise vitamin K-dependent coagulation factors, factor II, VII, IX, and X, leading to prolonged prothrombin time. Due to the bulky aromatic sidechains, superwarfarins have a much longer half-life when compared to warfarin, and exposure to superwarfarins results in a prolonged period of anticoagulation which can result in clinical bleeding. Diagnosis is straightforward in patients with known history of superwarfarin exposure but has proved difficult for patients who did not report superwarfarin intake. Superwarfarin poisoning should therefore be suspected in all patients with unexplained prolongation of prothrombin time, and can be confirmed by their detection in serum. Treatment for superwarfarin poisoning includes rapid correction of factor deficiencies with either 4-factor prothrombin complex concentrate or fresh frozen plasma in patients with active bleeding, and high dose vitamin K therapy given multiple times per day for a prolonged period of weeks to months.

Introduction

Superwarfarins are long-acting, vitamin K antagonist anticoagulant rodenticides derived from warfarin, in turn a synthetic anticoagulant developed from dicoumarol.¹ Compared with warfarin, superwarfarins have very long half-lives and are much more potent in terms of their ability to induce coagulopathy.^2–6 Exposure to superwarfarin can be overt or covert. Patients can present with a history of exposure without symptoms, some can present with bleeding symptoms without a clear history of exposure. Mortalities have been reported.^7–10 The significant morbidity and mortality from superwarfarin can be contrasted with the highly treatable nature of superwarfarin poisoning with factor replacement and vitamin K therapy, available in most parts of the world.^{2,4,6,11–13}

The diagnosis of superwarfarin poisoning in patients without a clear history of exposure is difficult. Examples include victims of homicide attempts,^14,15 the unfortunate users of brodifacoum-laced synthetic cannabinoids in the recent Illinois outbreak,^9,16 and a substantial number of patients exposed to superwarfarins of yet uncertain origin in China and Hong Kong.^15,17 Worse still, patients with bleeding symptoms such as haematuria, haemoperitoneum, and limb haematoma often are initially cared for by non-toxicologists unfamiliar with superwarfarin poisoning, with the resultant effect that diagnosis was delayed and the management was suboptimal.^14,17,18 While detection of superwarfarins in serum confirms the poisoning, this service is not widely available in most parts of the world.^7,13,16

The aim of the present mini-review is therefore to raise the awareness of superwarfarin poisoning among clinicians, clinical biochemists and chemical pathologists, such that this highly treatable condition with significant potential for morbidity and mortality does not remain undiagnosed.

The Vitamin K Cycle

Vitamin K is a collective term describing two classes of naphthoquinone derivatives: phylloquinones and menaquinones; the former are plant in origin, whereas the latter are bacterial in origin (Figure 1). Vitamin K is a fat-soluble vitamin and participates in biochemical processes as electron carriers; in human, vitamin K is a co-factor of gamma-glutamyl carboxylase, which is necessary for post-translational gamma-carboxylation of certain glutamate residues in Gla domain-containing proteins.^19,20 The hydroquinone form of vitamin K serves in the process as electron donor, and is oxidised to the epoxide form.²¹ The epoxide form is cycled back to the hydroquinone by the action of vitamin K epoxide reductase.²² This cycle between the hydroquinone form and the epoxide form is known as the vitamin K cycle (Figure 2).

Figure 1.

Structure of major vitamin K classes: (a) phylloquinones (e.g. Vitamin K1, n=3), and (b) menaquinones (Vitamin K2, n≥1; Vitamin K3, n=0).

Figure 2.

The vitamin K cycle (VKORC1: vitamin K epoxide reductase; GGCX: gamma-glutamyl carboxylase).

There are many vitamin K-dependent proteins in humans and these include coagulation factors II (prothrombin), VII, IX, and X, as well as the anticoagulants protein C, protein S, and protein Z.²³ In the absence of vitamin K hydroquinone, coagulation factors and endogenous anticoagulants produced do not undergo post-translational modification and lack properly functioning Gla domains in these factors.^19,24 Functional Gla domains are necessary for binding of calcium ion and stabilisation of the final tertiary structure of the proteins.^20,25,26 Despite affecting both coagulation factors and endogenous anticoagulants, the overall effect of vitamin K is that of a procoagulant, the very reason it is named vitamin K (for koagulation, the Scandinavian and German spelling of coagulation).²⁷ Recent evidence cast doubt as to whether properly functioning Gla domains are necessary for the action of prothrombin, and suggested that the pathophysiological mechanism for vitamin K deficiency or antagonism lies in the other factors (VII, IX, and X).²⁸

Apart from the coagulation-related proteins, vitamin K deficiency has been implicated as a factor in osteoporosis, vascular calcification, and insulin resistance, the effects of which are mediated through the Gla proteins osteocalcin, matrix Gla protein, and the signalling protein gas6 (growth arrest specific gene 6).^29–31

The Development of Warfarin and Superwarfarins

The discovery of warfarin and its predecessor dicoumarol comes from the Karl Link’s investigation of haemorrhagic disease that occurred when cattle consumed sweet clover hay that was infected by moulds (the haemorrhagic sweet clover disease).^32,33 With the help of Wisconsin Alumni Research Foundation, Link and colleagues were able to crystallise and identify the chemical cause of the haemorrhagic sweet clover disease, subsequently known as dicoumarol.^34,35 Dicoumarol was subsequently released into clinical use as an anticoagulant.³⁶ Subsequently, more than 150 analogues of dicoumarol were synthesised in an attempt to create a better rodenticide (or ‘better mousetraps’, in the words of Karl Link).³³ Compound 42, one of the synthesised analogues, later known as warfarin, was developed as a result.^1,33,37 The hydroxycoumarins warfarin, coumachlor, coumafuryl, and coumatetralyl, together with the indandiones diphacinone and chlorophacinone, collectively known as the first generation anticoagulant rodenticides, were developed in the 1950s and early 1960s. These anticoagulant rodenticides dominated the market of rodent control world-wide in the subsequent years.³⁸

This market dominance did not last long with heritable warfarin resistance identified in many species of rats and mice.³⁹ In response to this resistance, superwarfarins were developed.^40,41 This led to the development of difenacoum, brodifacoum, and bromadiolone, and these compounds were found to be useful against warfarin-resistant rodents.^41–43 These superwarfarins found rapid market acceptance and is reflected in the number of registered products.⁴⁴ The Table lists commonly encountered anticoagulant rodenticides with their respective molecular formula and molecular weight.

Table.

Commonly encountered anticoagulant rodenticides. Compounds which are not long-acting are marked with an asterisk.

Class	Compound	Molecular formula	Molecular weight
4-hydroxycoumarins	Warfarin*	C19H16O4	308.34
	Coumatetralyl*	C19H16O3	292.33
	Bromadiolone	C30H23BrO4	527.41
	Brodifacoum	C31H23BrO3	523.43
	Difenacoum	C31H24O3	444.52
	Difethialone	C31H23BrO2S	539.49
	Flocoumafen	C33H25F3O4	542.54
Indandiones	Pindone*	C14H14O3	230.26
	Chlorophacinone	C23H15ClO3	374.83
	Diphacinone	C23H16O3	340.38

The Toxicology of Superwarfarins

The two most commonly encountered superwarfarins in clinical poisoning are bromadiolone and brodifacoum.^45,46 Both bromadiolone and brodifacoum are warfarin/4-hydroxycoumarin derivatives.³⁸ In the broad sense of the term used clinically, superwarfarins also encompass long-acting anticoagulant rodenticides which are not warfarin/4-hydroxycoumarin derivatives, for example chlorophacinone and diphacinone, which are indandione derivatives.⁴⁷ These superwarfarins are all highly lipophilic with extremely long biological half-lives.⁴⁸ This can be seen by the high octanol:water partition coefficient (log P_oct/wat, a measurement of hydrophobicity) of bromadiolone (7.02) and brodifacoum (8.50) in comparison to that of warfarin (2.70) and dicoumarol (1.92).⁴⁹ They are also much more potent than warfarin: brodifacoum has an IC₅₀ of 0.15 μM towards rat microsomal vitamin K epoxide reductase, compared to 2.2 μM for warfarin.⁵⁰ The LD₅₀ of brodifacoum and bromadiolone are 0.26 and 1.13 mg/kg respectively, 715 times and 165 times lower than that of warfarin in albino Norway rats.³⁸

Superwarfarins such as bromadiolone and brodifacoum are bound in the liver and remain stable with hepatic half-life of more than 114 days for brodifacoum and ranges from 28 days to 318 days for bromadiolone across different animal species.^48,51,52 The plasma half-lives for brodifacoum and bromadiolone were determined to be 91.7 and 33.3 days respectively in rats in a recent study employing liquid chromatography-tandem mass spectrometry based method for quantitative analysis.⁴⁸ Readers are referred to the excellent review by Horak et al. for a comprehensive review on the toxicokinetics of warfarin and superwarfarins in animals.⁵ In stark contrast with the vast amount of clinical data available for warfarin,⁵³ human data concerning the pharmacokinetics of superwarfarin remained scarce, with only isolated case reports detailing individual cases of bromadiolone and brodifacoum poisoning.^2,3,54–58 The elimination-phase half-life for brodifacoum has been reported to be between 15 and 30 days,^3,58 whereas those of bromadiolone have been reported to be between 10 and 24 days.^2,57 Enterohepatic recirculation has been suggested as a factor contributing to the very long half-life of superwarfarins.⁵¹

Warfarin and superwarfarins act by inhibiting the enzyme vitamin K epoxide reductase, and make it impossible for the body to recycle vitamin K.^21,59,60 The enzyme kinetics of vitamin K epoxide reductase following incubation with warfarin have been reported as non-competitive.⁶¹ This orthodox concept has since been challenged by recent findings that warfarin competes with vitamin K binding⁶² as well as trapping the vitamin K epoxide reductase in an intermediate state.⁶³ The exact mechanism by which warfarin and superwarfarins inhibit vitamin K epoxide reductase remains unclear.

Following exposure to superwarfarins, the onset of measurable coagulopathy by prothrombin time depends upon the half-lives of vitamin K-dependent coagulation factors which range from 0.25 days for factor VII to 2.5 days for prothrombin.²³ Clinical bleeding as a result of the coagulopathy are later in time course and are most common between 3 and 9 days after an acute exposure.⁷

Epidemiology of Superwarfarin Poisoning

A review of over 300,000 cases reported to the National Poison Data System in the US showed that a large majority of reported superwarfarin poisoning are accidental (95.6%) and occur in children less than 6 years-of-age, with most cases of superwarfarin exposure due to oral exposure towards rodenticide baits, with commercial rodenticides commonly available in two forms: bait and liquid concentrate.^7,46 The bait form is usually at a lower concentration (e.g. bromadiolone at 0.005%) whereas liquid concentrate contains superwarfarin at a much higher concentration (e.g. bromadiolone at 0.5%).⁶⁴

Superwarfarin poisoning in adults, on the other hand, occurs with deliberate self-harm/suicide, accidental ingestion, occupational exposure with most cases being sporadic in nature.^{7,14,16,17,65,66} In contrast to the reports in the Western countries, case series from Asia showed a much larger proportion of superwarfarin poisoning occurring in adults, with many cases being suicide attempts, and a large group of patients in whom no possible source of superwarfarin exposure could be elucidated.^14,15,17,67

The use of superwarfarin as a poison for homicide or child abuse has been reported in the literature.^14,65 In 2011, a radio-controlled explosive device containing lead weights laced with brodifacoum was found and defused.^68,69 There are suggestions that similar explosive devices have been employed by Palestinian suicide bombers.⁷⁰ So far, there have been no reports of explosive injuries associated with superwarfarin poisoning.

The recent association of superwarfarin poisoning with consumption of synthetic cannabinoids have renewed the worry of drugs of abuse-associated superwarfarin poisoning.^3,9,71 A major outbreak of synthetic cannabinoid-associated coagulopathy occurred in Illinois, US in 2018, with a total of over 300 persons being affected and a total of 7 deaths between March and October 2018.^9,13,16 In the case series by Kelkar et al., brodifacoum was identified as the culprit and was detected in all patients tested, whereas difenacoum, bromadiolone, and warfarin were additionally detected among the patients.¹⁶ Limited experimental data suggests that there is no pharmacodynamics interaction between brodifacoum and synthetic cannabinoids, though pharmacokinetic interactions have not been ruled out.⁷² It remains unclear as to whether the presence of superwarfarins in synthetic cannabinoids represents malicious adulteration or inadvertent contamination.

Clinical Presentation and Laboratory Testing

The signs and symptoms resulting from haemorrhage and bleeding tendencies are the most important clinical features in superwarfarin poisoning. Most patients with accidental superwarfarin exposure do not develop any symptoms.⁷ For patients with definite history of accidental superwarfarin exposure, a dose of 1 mg of active ingredient, equivalent to 20 g of bait at a concentration of 0.005%, has been suggested as a referral threshold for emergency department assessment.⁶

Patients with symptomatic superwarfarin exposure present with bleeding-related manifestations such as gross haematuria, mucosal bleeding, gastrointestinal bleeding, soft tissue bleeding, menorrhagia, and haemoptysis, and non-bleeding related symptoms and signs such as abdominal pain, flank pain, and headache.^{7,14,16,17,73}

Rarely, patients with superwarfarin exposure present with clinical features of both thrombosis and bleeding. A 34-year-old woman developed haematuria and later died of brain herniation secondary to thrombosis of superior longitudinal sinus after chlorophacinone poisoning.⁸ Similarly, cases of deep vein thrombosis together with haematuria have been reported in a patient with brodifacoum poisoning and in another patient with difethialone poisoning.^74,75 It has been postulated that the initial phase of superwarfarin poisoning is hypercoagulable, due to deficiencies of vitamin K-dependent endogenous anticoagulants Protein C and Protein S occurring before deficiencies of vitamin K-dependent coagulation factors, similar to that described with the use of warfarin.^8,75,76

Prothrombin time and international normalised ratio (INR) are the investigations of choice in identifying the coagulopathy caused by warfarin and superwarfarins.^6,77 Further testing on individual coagulation factor levels may identify deficiencies of vitamin K-dependent coagulation factors and may be useful to exclude alternative causes,⁷⁸ and testing for PIVKA-II and vitamin K epoxide may help to identify patients with vitamin K antagonism.⁷⁹ Quantitative analysis of superwarfarins is the test of choice for confirmation of superwarfarin poisoning.^2,7,12,80 The dangerous practice of using the presence of warfarin and its metabolites in urine to exclude superwarfarin exposure could not be recommended as patients who had co-exposure of superwarfarin(s) and warfarin have been reported in the literature ^16,17 and would not be identified by this workflow, and commercial preparations containing warfarin and brodifacoum as a combination rodenticide have been reported.⁸⁰

The difficulty of diagnosing superwarfarin poisoning lies in those who present without a history of superwarfarin exposure,^7,45 with one case series noting a median delay of 6 days to achieve optimal medical treatment for patients without exposure history, reflecting the lack of alertness in entertaining this important differential diagnosis in patients presenting with coagulopathies.¹⁷ Superwarfarin poisoning should be suspected in all patients with unexplained bleeding with prolonged prothrombin time.^3,7,12,67,70

Detection of Superwarfarin in Biological Matrices

The definitive test for superwarfarin poisoning is direct analysis of superwarfarins in serum.^2,7,12,80 Earlier publications used methods such as high performance thin layer chromatography (HPTLC),⁸¹ high performance liquid chromatography with fluorescence detection (HPLC-FD)⁸² and gas chromatography-mass spectrometry (GC-MS).⁸³ Nowadays, liquid chromatography-tandem mass spectrometry (LC-MS/MS) is the preferred technique in the modern clinical toxicology laboratory as it is more sensitive than HPLC-FD and less cumbersome in operation when compared to GC-MS.^54,84,85

Published methods for multi-analyte analysis of superwarfarins in biological matrices with LC-MS/MS all utilised liquid-liquid extraction for sample extraction, with acetone, ethyl acetate, and acetonitrile being popular choices.^{48,54,84,86–91} In recent years, assays using newer sample extraction/pretreatment methods such as ultrasound-assisted liquid-liquid microextraction,⁹² phospholipid removal⁹³ and supported liquid extraction⁹⁴ have been reported. The use of C18-based chromatographic column, negative-mode electrospray ionisation, and detection using multiple reaction monitoring mode appears to be near-universal among publications with simultaneous determination of multiple superwarfarins by LC-MS/MS.^{48,54,84,86–94}

In the present authors’ laboratory, an in-house method based on liquid chromatography-tandem mass spectrometry is used to provide quantitative analysis of bromadiolone and brodifacoum and qualitative analysis of warfarin and superwarfarins in serum as well as qualitative analysis in urine.¹⁷ After addition of internal standards (bromadiolone-d₅, brodifacoum-d₄, and dicoumarol), serum specimens are subject to liquid-liquid extraction with 5% ethanol in ethyl acetate, followed by protein precipitation with trichloroacetic acid. The organic layer of the specimens is then dried, reconstituted and injected into a liquid chromatograph equipped with a C18 column and analytes are detected by negative electrospray ionisation-tandem mass spectrometry operated in multiple reaction monitoring (MRM) mode. The MRM chromatogram of this assay is shown in Figure 3.

Figure 3.

Multiple reaction monitoring (MRM) chromatogram of the in-house assay used in the authors’ laboratory. Legend: (1) warfarin metabolite, (2) coumafuryl, (3) warfarin, (4) coumatetralyl, (5) coumachlor, (6) pindone, (7) diphacinone, (8) bromadiolone, (9) chlorphacinone, (10) difenacoum, (11) flocoumafen, (12) brodifacoum, and (13) difethalone. Note the presence of diastereoisomer present in bromadiolone, difenacoum, brodifacoum, and difethialone.

Treatment for Superwarfarin Poisoning

The treatment of superwarfarin poisoning in a bleeding patient includes vitamin K1 therapy which addresses the underlying vitamin K antagonism, and early and rapid correction of coagulopathy to treat uncontrolled haemorrhage.^3,6,7,16,45 On the other hand, the guideline published by the American Association of Poison Control Centers suggested the evaluation of patients with superwarfarin exposure without clinical bleeding with testing for prothrombin time at presentation and after 48–72 hours exposure, and suggested against the administration of vitamin K1 before laboratory evaluation.⁶ Gastrointestinal decontamination is not considered to be helpful at least for accidental poisoning.⁹⁵ The use of multiple-dose activated charcoal to arrest enterohepatic circulation has not been successful.^6,96

For rapid correction of severe coagulopathy due to vitamin K antagonism, fresh frozen plasma and 4-factor prothrombin complex concentrate have been suggested,^{7,12,13,16,70} and used in a number of patients suffering from haemorrhage due to superwarfarin poisoning.^11,16,97,98 In a number of patients, recombinant activated factor VII was used.^14,99 There were no studies directly comparing the effectiveness of fresh frozen plasma versus 4-factor prothrombin complex concentrate. Analogy can be drawn with the reversal of warfarin effect. 4-factor prothrombin complex concentrate has been found to be non-inferior and superior to fresh frozen plasma, when both are co-administered with vitamin K1, in a phase IIIb, open-label, randomised trial: rapid INR reduction (INR ≤1.3 at 30 minutes post-infusion) was achieved in 55% of patients receiving the 4-factor prothrombin complex concentrate, versus 10% of patients in the plasma group.¹⁰⁰ On the other hand, in the recent case series by Kelkar et al., the rate of early responses were similar for patients who have been put on adequate oral vitamin K1 therapy, whether or not intravenous vitamin K1, fresh frozen plasma, or both were used.¹⁶

Vitamin K1 can be administered orally, subcutaneously, or intravenously. Again drawing from the evidence with warfarin reversal, a meta-analysis comparing the effectiveness of oral, subcutaneous, and intravenous administration concluded that oral and intravenous administration were equivalent, and both are superior to subcutaneous administration, for INR reduction at 24 hours after administration.¹⁰¹ On the other hand, intravenous vitamin K was found to be superior at 4 hours post-administration and equivalent at 24 hours post-administration, at reversing the anticoagulation of warfarin.¹⁰² For non-bleeding patients, the earlier onset of the reversal of anticoagulation must be weighed against the risk of anaphalactoid reaction to intravenous vitamin K.^103,104

There was no obvious consensus on the optimal dose and frequency of vitamin K1 for patients with superwarfarin poisoning in the literature. The optimal frequency of administration was suggested to be every 6–8 hours.⁴⁵ In the recent outbreak of synthetic cannabinoid-associated superwarfarin poisoning, patients were treated with vitamin K1 in doses of 100 mg/day, resulting in a monthly drug cost of US$24,000–37,000 per patient although these initially cited prices may be an overestimate due to short-term shortage of drug.^16,105–108 Such high cost of vitamin K was not seen in Hong Kong, and is not the case for example with the UK.¹⁰⁹

Currently, there are no established guidelines for termination of vitamin K1 therapy in patients with known superwarfarin poisoning. Termination of therapy based on a normal clotting profile obtained while the poisoned patient is still on vitamin K1 can be dangerous.¹¹⁰ A tapering approach with vitamin K1 therapy has been used in the past.⁶⁵ In many centres, where quantitative measurement of superwarfarins is not routinely available, an approach based on stopping vitamin K1 therapy followed by measurement of coagulation parameters at 48–72 hours have been employed.^13,14 While no evidence-based safe concentration levels have been defined for the superwarfarins, bromadiolone and brodifacoum levels of <10 ng/mL have been reported to be associated with normal coagulation and treatment can be terminated safely without exposing patients to the risk of bleeding.^3,4

Patients with superwarfarin exposure from an unknown source represent the most difficult group of patients to be treated and followed-up. The difficulty with patients with deliberate self-harm or suicidal intent is obvious and their poor adherence to vitamin K1 therapy is well-reported,^14,45 as is the risk of re-exposure.¹⁵ On the other hand, where the patients truly do not know about the source of superwarfarin that they are exposed to (which we called ‘hidden superwarfarin poisoning’ in our centre), it is inherently impossible for them to ‘avoid re-exposure’ other than ‘being more careful with what they eat’.

It is suggested that the quantitative analysis of superwarfarins in serum is the first step towards solving the problems surrounding the treatment and follow-up of these patients. This analytical approach takes away the guesswork from the differential diagnoses in a patient with vitamin K-dependent coagulopathy, enables the clinical toxicologist to ascertain re-exposure, and importantly, provides an objective end-point which can be relied upon for termination of treatment without exposing the patient to bleeding risks that stems from stopping vitamin K1.

Conclusions: A Glimpse into the Future

Over the past 20 years, there has been tremendous improvement in the diagnosis and management of superwarfarin poisoning, enabled by better understanding in the pathophysiology of vitamin K antagonism, the pharmacokinetics of vitamin K formulations, the adoption of clinical mass spectrometry for detection of superwarfarins, and the improved care for coagulopathic patient. Superwarfarins have withstood the test of time and the power of natural selection (of rodents) and will remain both in the fields of pest control and the practice of toxicology.

The great difficulty in the diagnosis of superwarfarin poisoning means that there is much that pathologists and scientists can do. It can be as simple as providing suggestions on further investigations based on related test results, for example, deficiency in vitamin K-dependent factors in a patient not on warfarin, or a negative urine toxicology in a coagulopathic patient. Last but not least, bearing in mind the limited availability of such testing services world-wide, we wish to reiterate the importance of providing a quantitative serum assay for superwarfarins to definitively establish the diagnosis, as well as to inform termination of vitamin K1 treatment in patients with superwarfarin poisoning.