Superwarfarin exposure or poisoning is a growing public health problem throughout the world and follows their widespread use and availability as rodenticides to overcome the resistance to warfarin. Their danger to humans lies in their high potency and long half-life. Affected patients often present with severely deranged clotting that necessitates the administration of blood products followed by large doses of vitamin K over weeks or months. Particular problems for both diagnosis and treatment monitoring are the variety of available rodenticides, their different properties, and the low concentrations at which they antagonize vitamin K function. Our Centre serves as a National and International Referral Centre for the diagnosis of anticoagulant poisoning. Cases referred include suspected self-poisoning, malicious poisoning, unintentional ingestion and occupational exposure. Unintentional poisoning of children constitutes a large group; a US surveillance system reported 23,447 exposures in a 4-year period. Because usual chromatographic techniques have proved both insensitive and inadequate for unambiguous characterization, we have developed new methods using liquid chromatography-electrospray ionisation-tandem mass spectrometry (LC-ESI-MS/MS) to identify and quantify vitamin K antagonists in different biological matrices. Anticoagulants of interest included warfarin, acenocoumarin, coumatetralyl, difenacoum, brodifacoum, bromadialone and flocoumafen. Each compound was first analysed by ESI-MS/MS (Applied Biosystems Q TRAP) to identify their molecular ion and fragmentation patterns. Preliminary results showed that not all superwarfarins ionised well in the positive mode but all gave a fairly strong [M-H] precursor ion in the negative mode. Therefore the negative ionisation mode was chosen for subsequent fragmentation of the molecular ions in the collision cell of the triple quadrupole to give characteristic patterns for each compound. From these fragmentation analyses, the most abundant ions were selected and the multiple reaction monitoring (MRM) facility used for the unequivocal identification of each superwarfarin. Plasma samples were extracted by a previously validated technique used in our usual LC warfarin assay and acenocoumarin was used as an internal standard for quantification. Capillary LC of the superwarfarins was done in the reverse phase mode at 7μl/min with a runtime of 25 minutes. The limit of detection for the anticoagulants was much lower than for traditional HPLC and ranged from 8–20 pg. One recent referral was a case of suspected poisoning with an unknown vitamin K antagonist in two Chinese infant siblings (aged 18 and 31 months). Both infants presented with a severe deficiency of vitamin K-dependent clotting factors of unknown aetiology and were treated by FFP and regular high dose vitamin K. Descarboxyprothrombin and vitamin K1 epoxide were elevated, both indicative of a vitamin K antagonist. LC-ESI-MS/MS showed that the superwarfarin bromadiolone was present in the plasma of both babies at levels (≤ 1.2 mg/L) too low for determination by LC alone. Subsequent monitoring was carried out until antagonist levels were undetectable. This same LC-ESI-MS/MS method has been used for analyses of archived plasma samples from other problematic cases of suspected superwarfarin poisoning and has enabled the identification of the culprit vitamin K antagonists in all cases. We conclude that LC-ESI MS/MS is the method of choice for investigating these difficult cases.

Author notes

Corresponding author