|
Abstract INTRODUCTION:
Fentanyl remains a significant public health challenge in the United States, with its widespread use contributing to ongoing morbidity and mortality. While urine fentanyl concentrations are not typically used to guide individual clinical decision-making, they offer valuable insights into population-level exposures and substance use patterns. There are no studies published in the literature that evaluate fentanyl, norfentanyl, and xylazine concentrations in urine across different patient populations. In this study, we developed and validated an in-house LC-MS/MS method to quantify fentanyl, norfentanyl, and xylazine in urine samples.
METHODS:
A total of 280 samples were analyzed, from 5 distinct patient groups, patients at intake for opioid replacement therapy (OTOP, N=143), patients for which comprehensive drug testing was ordered by their clinician (CDHR, N=37), patients receiving care in the emergency department (ED, N=37), patients for which a urine drug screen was ordered from an outpatient clinic (OPC, N=28), and patients for which a urine drug screen was ordered from a specialty care clinic e.g., HIV, TB, Addiction medicine) (SCC, N=31). The samples were collected using a targeted sampling approach over a one-year period from February 2024 to January 2025. This study used remnant patient samples and was approved by the UCSF IRB committee.
Urine samples were prepared using a dilution method. Specifically, 50 μL of urine was diluted 1:10 with the appropriate diluent. The prepared samples were then subjected to analysis. Liquid chromatographic separation was performed using a Kinetex 2.6 µm Phenyl-Hexyl column (50 x 4.6 mm) with a 4.25-minute gradient. The gradient started at 40% mobile phase B (MpB) and held for 0.75 minutes, then ramped to 90% MpB by 1.25 minutes and held until 2.25 minutes. The system was equilibrated from 2.25 to 4.25 minutes at 40% MpB. The flow rate was 0.600 mL/min. The column temperature was held at 40°C. Fentanyl, norfentanyl, and xylazine were detected in positive ion mode using multiple reaction monitoring (MRM), with transitions optimized for each analyte. The MRM transitions were as follows: fentanyl (337.174 > 188.137, 337.174 > 105.04), norfentanyl (233.086 > 84.094, 233.086 > 56.073), and xylazine (221.112 > 164.056, 221.113 > 90.023). Internal standards for all three analytes were also included, with optimized transitions incorporated into the method. The method followed a validation plan that incorporated both CLSI 62 A guidelines as well as FDA bioanalytical guidelines. The linear model used was linear with 1/x weighting with a corresponding R2 of .991 for xylazine, .995 for fentanyl, and .952 for norfentanyl. The lower limit of the measuring interval was determined by having a CV <20%, for fentanyl this corresponds to .5 ng/ml, norfentanyl, 2 ng/ml, and xylazine 1 ng/ml. For all analytes, inter- and intra-assay precision was <15% CV for a high, mid, and low QC material. A dilution was implemented into the method to expand the analytical range, of an additional 10x in addition to the original 10x dilution for the method. Matrix effects, carryover, and stability were evaluated and adhered to the guidelines.
Data were analyzed using R. Statistical comparisons were performed with Wilcoxon rank-sum and Kolmogorov-Smirnov (KS) tests. For group comparisons, Kruskal-Wallis testing was followed by Dunn's test for multiple comparisons to account for non-normal distributions between the patient groups.
RESULTS:
Statistical analysis revealed that OTOP patients had significantly higher fentanyl concentrations (Median = 1.49 µg/mL; Range = 0.008–37.02 µg/mL) and norfentanyl concentrations (Median = 5.50 µg/mL; Range = 0.10–96.22 µg/mL) compared to all other groups (Wilcoxon p < 1.18e-12). In comparison, median fentanyl concentrations in other groups were lower: CDHR (Median = 0.31 µg/mL; Range = 0.001–10.14 µg/mL), ED (Median = 0.02 µg/mL; Range = 0.0005–2.37 µg/mL), OPC (Median = 0.13 µg/mL; Range = 0.0005–4.08 µg/mL), and SCC (Median = 0.09 µg/mL; Range = 0.0005–4.79 µg/mL). Norfentanyl concentrations followed a similar trend, with CDHR (Median = 2.24 µg/mL; Range = 0.14–17.18 µg/mL), ED (Median = 0.25 µg/mL; Range = 0.002–9.15 µg/mL), Outpatient Clinics (Median = 0.31 µg/mL; Range = 0.005–20.80 µg/mL), and Special Clinics (Median = 0.75 µg/mL; Range = 0.005–16.57 µg/mL). Xylazine was detected across all groups but remained at consistently low concentrations. The highest xylazine concentrations were observed in OTOP (Median = 0.03 µg/mL; Range = 0.00–1.43 µg/mL), with lower values in CDHR (Median = 0.01 µg/mL; Range = 0.00–0.20 µg/mL), ED (Median = 0.01 µg/mL; Range = 0.00–0.08 µg/mL), Outpatient Clinics (Median = 0.01 µg/mL; Range = 0.00–0.02 µg/mL), and Special Clinics (Median = 0.00 µg/mL; Range = 0.00–0.01 µg/mL).
Additionally, analysis of the fentanyl/norfentanyl ratio revealed significant differences between groups, with the OTOP population displaying notably higher ratios compared to other groups. The percentage of fentanyl associated with xylazine also varied between groups, with OTOP showing higher percentages compared to the other groups. Despite these differences, Pearson’s correlation analysis between the fentanyl/norfentanyl ratio and the percentage of fentanyl associated with xylazine yielded a weak, non-significant relationship (cor = -0.038, p = 0.530), similar results were obtained when looking at total fentanyl (fentanyl + Norfentanyl) (cor = -0.082, p = 0.1762) instead of the ratio suggesting no clear linear association between these two measures.
CONCLUSION:
looking at the toxicology profile of patients across different departments revealed no significant correlation between the fentanyl/norfentanyl ratio and xylazine levels, indicating independent factors influencing these measures such as variability in the amount of xylazine added to the fentanyl drug supply overtime. |
= Discovery stage. (53.14%, 2025)
= Translation stage. (22.33%, 2025)
= Clinically available. (24.53%, 2025)
