Cats and Rat Poison: Forensic LC–MS/MS Tracks Five Years of Rodenticide Deaths

Cats across South Korea are dying from rodenticide poisoning that often goes undetected, and a new forensic study has now mapped the scale of the problem with unprecedented precision. Researchers at the Animal and Plant Quarantine Agency developed and validated a liquid chromatography–tandem mass spectrometry method capable of simultaneously detecting eight anticoagulant rodenticides in feline gastric tissue and gastric contents, then applied it to 747 forensic cat necropsy cases submitted between January 2020 and August 2025. The results, published in Veterinary Medicine and Science, offer the most comprehensive picture yet of how these persistent poisons move through the environments cats inhabit—and reveal the detection of a rodenticide that should not have been there at all.

Anticoagulant rodenticides kill by sabotage. They inhibit the synthesis of vitamin K–dependent clotting factors II, VII, IX and X, progressively disabling the blood’s ability to coagulate until internal haemorrhage proves fatal. First-generation compounds such as warfarin, chlorophacinone, coumatetralyl and diphacinone are comparatively weak and short-lived, requiring repeated ingestion to deliver a lethal dose. Second-generation agents—bromadiolone, brodifacoum, flocoumafen and difethialone—are far more potent and persist in the body for extended periods, meaning a single exposure can kill. That same persistence makes them a hazard well beyond their intended targets, allowing them to bioaccumulate in non-target species, including companion animals and wildlife, long after the bait has been laid.

Diagnosing this poisoning in cats is notoriously difficult. Exposure is frequently unwitnessed, clinical signs such as pale mucous membranes, subcutaneous bleeding, pulmonary or gastrointestinal haemorrhage may be delayed for days, and owners often fail to recognise the early warning signs. Traditional coagulation tests such as prothrombin time and activated partial thromboplastin time, the mainstays of clinical practice, are indirect and nonspecific: they can be prolonged by liver disease, inflammation or malnutrition, and may still fall within reference ranges early after exposure. They cannot identify which compound the animal ingested, or how much. In South Korea, forensic instrumental analysis for companion animals has lagged behind human toxicology, hampered by the absence of standardized methods and limited case-based analytical studies.

The new method addresses that gap. The research team, led by HyunYoung Chae and JeongWoo Kang, targeted eight rodenticides—brodifacoum, bromadiolone, chlorophacinone, coumatetralyl, difenacoum, difethialone, flocoumafen and warfarin—and combined a modified AOAC QuEChERS extraction with automated solid-phase extraction using ISOLUTE AOAC Waxed cartridges on a Biotage Extrahera system. This approach, built on a clean-up strategy first described by Peres and colleagues in 2019, uses acetonitrile extraction with acidified conditions followed by cartridge-based removal of matrix-derived interferences such as proteins, lipids and pigments. The automation component matters as much as the chemistry: standardised, machine-driven sample preparation minimises variability between analysts, a persistent weakness in forensic laboratories handling high caseloads.

At the analytical core sits a hybrid quadrupole–linear ion trap mass spectrometer, the SCIEX QTRAP 5500+, operated in negative electrospray ionisation with multiple reaction monitoring. The clever part is the dual acquisition strategy. Within a single 15-minute run, the instrument performs sensitive MRM quantification while simultaneously triggering information-dependent acquisition of enhanced product ion spectra whenever precursor ions exceed a threshold signal. These EPI spectra are then matched against reference standards for structural confirmation of fragment ions, dramatically reducing the risk of false positives from structurally related compounds—a critical safeguard when legal consequences, including suspected malicious poisoning, may hang on the result. Chromatographic separation on a YMC-Triart C18 column cleanly resolved all eight compounds within the 15-minute window, with retention times ranging from 8.79 minutes for warfarin to 11.61 minutes for difethialone.

Validation performance was rigorous. Calibration curves spanning 1 to 100 nanograms per millilitre yielded coefficients of determination between 0.990 and 0.997 across all analytes. Limits of detection ranged from 0.05 to 0.1 micrograms per kilogram, and limits of quantification from 0.1 to 0.5 micrograms per kilogram—sensitive enough to catch trace residues. Recovery rates fell between 70.3 and 105.6 percent at spiked concentrations of 10, 20 and 50 micrograms per kilogram, comfortably inside the internationally accepted 70 to 120 percent criterion, while relative standard deviations of 2.4 to 8.6 percent confirmed repeatability well below the 15 percent threshold. Selectivity testing on blank feline gastric tissue showed no interfering peaks. The laboratory subsequently passed an external proficiency test and earned ISO/IEC 17025 accreditation under the Korea Laboratory Accreditation Scheme in October 2025.

When the validated method was applied to all 747 feline forensic toxicology cases, anticoagulant rodenticides were detected in 19 cases, an overall detection rate of 2.5 percent. Coumatetralyl dominated, appearing alone in 14 cases and alongside flocoumafen in one more, with concentrations spanning 2.3 to 107.6 micrograms per kilogram and a mean of 42.6. Flocoumafen alone accounted for three cases, at concentrations from 15.9 to 80.3 micrograms per kilogram, and bromadiolone appeared once. Annual detections peaked at seven cases in 2022 and six in 2023. Necropsy findings aligned grimly with the mechanism of the poison: perinasal and perioral bleeding in six cases, extensive pulmonary haemorrhage in four, pulmonary congestion or hyperaemia in three, plus blood-tinged gastrointestinal contents, jaundice with petechiae, mesenteric haemorrhage and blood-tinged fluid accumulations in others. Most cases came from police submissions—17 of 19—and most carcasses arrived frozen.

One finding stands out as a regulatory red flag. Flocoumafen, a second-generation compound, was not listed as a registered active ingredient in the domestic rodenticide registration records maintained by South Korea’s National Institute of Environmental Research for the study period. Registered formulations were based primarily on coumatetralyl and bromadiolone, which fits the coumatetralyl-dominant pattern observed. But the presence of flocoumafen in four cats suggests exposure to products or sources operating outside the documented registration system—whether unregistered imports, misuse, or pathways the records simply do not capture. The authors argue this alone justifies broad multi-analyte screening that includes both registered and potentially unregistered compounds, rather than tests limited to whatever appears in the national registry.

The study also confronts a stubborn interpretive problem: knowing that a cat ingested a rodenticide does not reveal how. Cats encounter these poisons through primary exposure—directly eating bait or contaminated food—or secondary exposure, by hunting and eating poisoned rodents. As predators roaming urban, peri-urban and agricultural landscapes, cats face both routes. Yet in the nine AR-positive cases where gastric contents were available, gross examination identified no recognisable bait material or prey remains, leaving the exposure pathway and any question of deliberate poisoning unresolved. The authors stress that gastric findings indicate recent oral ingestion but cannot, by themselves, distinguish accidental from intentional exposure or primary from secondary routes; that determination requires integration with environmental and investigative information.

What the method delivers is compound-specific, molecular-level evidence of recent ingestion—something coagulation panels can never provide—and it does so with a runtime short enough for routine forensic workloads. Gastric tissue and contents proved practical matrices for cats, whose small body size and the competing demands of histopathology and bacteriology often limit liver availability. The team notes that gastric contents can remain positive even when liver samples test negative, reinforcing their complementary value. By pairing MRM screening with EPI spectral confirmation in a single automated workflow, the study establishes a template for veterinary forensic toxicology that could extend to other species and toxicants. As rodenticide use continues and the environments of companion animals increasingly overlap with treated areas, the authors call for continued toxicovigilance and systematic monitoring—because for cats, as this half-decade of forensic casework shows, the poison is often invisible until the bleeding starts.

Subject of Research: Forensic LC–MS/MS detection and monitoring of anticoagulant rodenticide poisoning in cats

Article Title: Forensic LC–MS/MS Monitoring of Anticoagulant Rodenticide Poisoning in Cats Over a Half‐Decade

Article References: Chae, H., Byun, J.-W., Yim, S.-J., Ku, B.-K., Kim, T.-W., & Kang, J. (2026). Forensic LC–MS/MS Monitoring of Anticoagulant Rodenticide Poisoning in Cats Over a Half‐Decade. Veterinary Medicine and Science, 12(5), Article e71209. https://doi.org/10.1002/vms3.71209

Image Credits: AI Generated

DOI: 10.1002/vms3.71209

Keywords: anticoagulant rodenticides, LC–MS/MS, veterinary forensics, cat poisoning, coumatetralyl, flocoumafen, bromadiolone, gastric contents, QTRAP mass spectrometry, QuEChERS extraction, South Korea, toxicovigilance