Determination of three metabolites of thromboxane A<sub>2</sub> and 8-iso-prostaglandin F<sub>2α</sub> in urine by ultra performance liquid chromatography-tandem mass spectrometry

Thromboxane A₂ (TXA₂), a prothrombotic factor that induces platelet aggregation and thrombosis, acts as a vasoconstrictor by activating TXA₂ receptors (TP receptors). TXA₂ is extremely unstable and metabolizes into three major metabolites: 2,3-dinor thromboxane B₂ (2,3-dinor-TXB₂), 11-dehydro TXB₂(11-dh-TXB₂), and 11-dehydro-2,3-dinor TXB₂(11-dh-2,3-dinor-TXB₂). 8-Iso-prostaglandin F_2α(8-iso-PGF_2α), a prostaglandin-like compound widely considered the best biomarker of oxidative stress, can also activate TP receptors. The accurate quantification of TXA₂ metabolites and 8-iso-PGF_2α is critical in cardiovascular disease (CVD). In this study, a method based on ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) was developed for the simultaneous determination of 2,3-dinor-TXB₂,11-dh-2,3-dinor-TXB₂, 11-dh-TXB₂, and 8-iso-PGF_2α in human urine. Urine samples were collected from healthy volunteers and patients with CT- or MRI-confirmed ischemic stroke occurring less than 48 h earlier, and cryopreserved at -80 ℃ within 1 h after collection. The urine samples were thawed at room temperature and acidified to pH 2.0-4.0 using hydrochloric acid. The supernatant was collected after centrifugation. A total of 1 mL of each urine sample was added with 100 μL of the internal standard working solution and mixed well. The samples were loaded onto a C18 SPE column (50 mg). The SPE cartridges were preconditioned with 500 μL of methanol and then equilibrated with 500 μL of water. After sample loading, the SPE cartridges were washed with 500 μL of water, 500 μL of 5% methanol aqueous solution containing 0.5% (v/v) ammonia, and 500 μL of 5% methanol aqueous solution containing 2% (v/v) formic acid. The cartridges were dried, and the analytes were eluted with 400 μL of methanol. The eluents were dried and subsequently reconstituted with 50 μL of 13% acetonitrile aqueous solution. After filtration through a filter membrane, the samples were analyzed on an ACQUITY UPLC^® BEH phenyl column (50 mm×2.1 mm, 1.7 μm) via gradient elution using 2 mmol/L ammonium acetate aqueous solution containing 0.002% (v/v) ammonia and acetonitrile as the mobile phases. The flow rate was 0.3 mL/min, and the column temperature was 40 ℃. The analytes were determined in negative electrospray ionization and multiple-reaction monitoring modes. The four target compounds showed satisfactory linearity within the relevant ranges, with linear correlation coefficients (<i>R</i>²) greater than 0.99. The limits of detection of the method were 0.02 ng/mL for 2,3-dinor-TXB₂ and 0.01 ng/mL for 11-dh-2,3-dinor-TXB₂, 11-dh-TXB₂, and 8-iso-PGF_2α. The limits of quantification were 0.1 ng/mL for 2,3-dinor-TXB₂ and 0.05 ng/mL for 11-dh-2,3-dinor-TXB₂, 11-dh-TXB₂, and 8-iso-PGF_2α. In actual urine, the recovery rates at the LOQ level were in the range of 91.48%-104.87%. The recovery rates at low, medium, and high levels were in the range of 92.95%-104.90%. The intra- and inter-day precisions were in the range of 2.79%-13.01% and 4.45%-13.67%, respectively. The relative error (RE) between the average peak area of the mixed matrix and the sum of the ratios of the pure solution and urine matrices was within ±20%. The samples were stable at 4 ℃ for 24 h and at -70 ℃ for 10 d. The developed method is the first to realize the simultaneous determination of 2,3-dinor-TXB₂, 11-dh-2,3-dinor-TXB₂, 11-dh-TXB₂, and 8-iso-PGF_2α in urine. The method was used to determine the concentrations of 2,3-dinor-TXB₂, 11-dh-2,3-dinor-TXB₂, 11-dh-TXB₂, and 8-iso-PGF_2α in healthy controls and patients with ischemic stroke, and the results were corrected using creatinine. Binary logistic regression analysis was used to construct the prediction model, and a receiver operating characteristic (ROC) curve was drawn to evaluate the clinical diagnostic ability of the method for the target compounds. TXA₂ was calculated as the sum of its three metabolites. The area under the curve (AUC) of TXA₂ was 0.849, and the method sensitivity and specificity were 69.2% and 92.3%, respectively. The AUC of 8-iso-PGF_2α was 0.775, and the method sensitivity and specificity were 84.6% and 76.9%, respectively. The proposed method has good clinical value and is expected to assist in the early screening and diagnosis of ischemic stroke.