**Background:** Bracken fern (Pteridium aquilinum) is one of the five most abundant plant species globally and contains ptaquiloside (PTA), a potent natural toxin with carcinogenic, mutagenic, and mitogenic properties in livestock and laboratory animals. Epidemiological studies have linked bracken fern ingestion to esophageal and stomach cancer in humans. Despite these risks, bracken fern is intentionally consumed in traditional foods in Korea, Japan, China, and Brazil. Human exposure can also occur indirectly through contaminated animal products such as meat and milk when livestock graze on bracken fern. PTA is chemically unstable, degrading to pterosin B under hydrolysis, heat, or enzymatic conditions, making accurate quantification challenging. Previous analytical methods were limited to one or two matrix types, and no method had been developed combining QuEChERS (quick, easy, cheap, effective, rugged, and safe) extraction with LC-MS/MS for direct PTA quantification across diverse food matrices. This study aimed to develop and validate such a method and apply it to monitor PTA in commercially available bracken fern, meat, and dairy products in Korea for the first time.
**Methods:** PTA standard (purity >98%) was used. Sample preparation employed a modular QuEChERS approach: 10 g homogenized sample was extracted with 10 mL acetonitrile, followed by addition of EN extraction salts (4 g MgSO₄, 1 g NaCl, 1 g sodium citrate, 0.5 g sodium hydrogen citrate sesquihydrate). After centrifugation, dispersive solid-phase extraction (d-SPE) cleanup was performed using two modules—Module 1 (25 mg PSA, 2.5 mg GCB, 150 mg MgSO₄) for pigment-rich matrices (bracken fern), and Module 2 (25 mg PSA, 25 mg C18EC, 150 mg MgSO₄) for fat-dominant matrices (meat, dairy). The final extract was diluted 10-fold with 0.1% formic acid in water. LC-MS/MS analysis used a Vanquish UHPLC–TSQ Altis system with a Kinetex C18 column (100 mm × 2.1 mm, 2.6 µm) at 35°C. Mobile phase consisted of water with 0.1% formic acid (A) and acetonitrile with 0.1% formic acid (B) at 0.5 mL/min over a 5-minute gradient. Detection was performed in MRM mode with positive electrospray ionization. The precursor ion was m/z 399 [M+H]⁺, with quantifier ion m/z 181 and qualifier ions m/z 277 and 381. Method validation assessed selectivity, linearity, LOD, LOQ, accuracy, precision, and recovery. Matrix-matched calibration used PTA-free bracken fern matrix at six concentrations (0.1–50 µg/kg). Cross-validation was performed at two independent institutions. The method was applied to 26 commercial samples (18 bracken fern products, 5 dairy products, 3 beef products) collected from Korean local and online markets. Dietary exposure was estimated using Korea National Health and Nutrition Examination Survey (KNHANES) data with an average body weight of 57.6 kg, using lower bound (LB) and upper bound (UB) approaches.
**Key Results:** The calibration curve showed excellent linearity (r² = 0.9979). LOD and LOQ were 0.03 µg/kg and 0.09 µg/kg, respectively. Intraday accuracy ranged from 83.5% to 98.5% with precision (CV) of 0.4–4.9%; interday accuracy ranged from 83.8% to 97.5% with precision of 1.5–7.2%. Extraction recovery rates were 71.2% at 0.1 µg/kg, 93.6% at 20 µg/kg, and 92.6% at 50 µg/kg. Cross-validation at two institutions showed accuracy of 94.1–113.7% (precision 1.1–5.2%) and 88.3–101.0% (precision 1.0–2.8%), both meeting AOAC guidelines. Matrix effect evaluation showed no significant differences: peak area differences compared to standard were −4.0% for bracken fern, −4.1% for meat, and +5.3% for milk. Slope CV across matrices was 1.7%. Among 26 commercial samples, PTA was detected below LOQ in one dried bracken fern sample (sample ID B, Korea) and at 0.1 µg/kg in one free-range beef sample from New Zealand (sample ID Z). All other samples were not detected (N.D.). Estimated dietary exposure using LB/UB approaches ranged from 0.0 to 1.1 × 10⁻⁶ µg/kg b.w./day for vegetables, 0.0 to 3.0 × 10⁻⁵ µg/kg b.w./day for dairy products, and 2.5 × 10⁻⁵ to 2.5 × 10⁻⁵ µg/kg b.w./day for meat.
**Clinical Implications:** This study provides the first validated QuEChERS-LC-MS/MS method capable of quantifying PTA across multiple food matrices (bracken fern, meat, dairy) using a single matrix-matched calibration approach. The method is rapid (5-minute run time), sensitive (LOQ 0.09 µg/kg), and reproducible. Application to commercial Korean market samples revealed that most products are PTA-free, likely due to pre-treatment processes such as blanching or boiling. However, detection of PTA in one beef sample (0.1 µg/kg) from grass-fed, free-range cattle in New Zealand confirms that indirect human exposure via animal products is possible. The authors note that PTA can accumulate in the body and cause cancer, so even low levels warrant attention. Dietary exposure estimates were provided, but formal risk assessment was not possible due to insufficient toxicity data. The study emphasizes the need for continuous food monitoring, expanded surveillance of meat and dairy products, and additional toxicity studies to enable comprehensive risk assessment for PTA in the food chain.