Determination of oxycodone and its major metabolites noroxycodone and oxymorphone by ultra-high-performance liquid chromatography tandem mass spectrometry in plasma and urine: application to real cases

Chemicals and reagents

Oxycodone and oxycodone-d₃, oxymorphone and oxymorphone-d₃, and noroxycodone and noroxycodone-d₃ were supplied by Cerilliant Corporation (Round Rock, TX, USA). β-Glucuronidase (type HP-2 from Helix Pomatia), ultrapure water, and all other reagents of analytical grade were purchased from Sigma-Aldrich (Milan, Italy).

Biological samples

Plasma and urine samples were obtained as discharge material by six hospitalized patients under treatment with different dosages and formulations (Depalgos^® and Targin^®) of oxycodone, whose biological samples were collected for routine controls. Information on sex and age of the patients, oxycodone formulation and dosage, co-medication, and type of pain to be treated was available for all the patients (Table 1). The patients gave signed consent to the use of their discharged biological samples, which were anonymized with correspondent clinical information in a database before analysis.

Blood was collected in Vacutainers^® EDTA plasma tubes and immediately centrifuged after collection, and the obtained plasma was aliquoted into 0.5-mL plastic tubes. Both plasma and urine samples were frozen at –20°C until analysis. No preservatives were added to specimens.

Drug-free plasma and urine samples were obtained from 20 different healthy donors, analyzed during method validation to exclude any source of chromatographic interferences, and mixed to obtain a homogeneous pool of blank samples to be used for calibration standards and quality control (QC) samples.

Standard solutions, calibrators, and controls

Stock solutions of each analyte (1 mg/mL) were combined and diluted with methanol to set working solutions (5.0, 1.0, and 0.1 μg/mL) and stored at −20°C until analysis. The internal standards (ISs) working solution was used at a concentration of 1.0 μg/mL.

Appropriate amount of working solutions were used to spike pre-checked drug-free plasma and urine samples to prepare five different standards for calibration curve. QC samples at three concentrations (low, medium, high) spanning the linear dynamic ranges of the calibration curves for all the matrices of interest were also daily prepared to be included in each analytical batch to check validation parameters (e.g. accuracy, precision, analytical recovery).

Sample preparation

Aliquots of 0.5 mL plasma and urine were added with appropriate amount of ISs working solution (100 ng for urine and 50 ng for plasma) and diluted with 0.5 mL phosphate buffer, pH 8.4. After 30 s of vortexing, analytes were extracted twice with 1.5-mL aliquots of a chloroform/isopropanol (9:1, v/v) mixture.

After centrifugation at 1500 g for 5 min, the organic phase was evaporated to dryness under a stream of nitrogen and re-dissolved in 100 μL mobile phase (1% acetic acid in water).

An aliquot of urine samples, added with 50 μL of 10,000 units/mL β-glucuronidase, was incubated overnight before extraction, to check for eventual presence of glucuronated metabolites.

UHPLC-MS/MS analysis

Oxycodone, oxymorphone, noroxycodone were measured in plasma and urine samples by UHPLC-MS/MS. The chromatographic separation was achieved with a UPLC Acuity HSS C18 column (2.1×150 mm, 1.8 μm) using a linear gradient elution with two solvents: acetic acid 1% in water (solvent A) and methanol (solvent B). Solvent A was maintained at 85% for the first 1.50 min. It was decreased to 40% from 1.50 to 5.0 min, then increased to 85% from 5.0 to 5.1 min and held at 85% from 5.1 to 10.00 min for re-equilibration. The flow rate was kept constant at 0.35 mL/min during the analysis. The separated analytes were detected with a triple quadrupole mass spectrometer operated in multiple reaction monitoring (MRM) mode via positive electrospray ionization (ESI). The applied ESI conditions were the following: capillary voltage, 1.0 kV; desolvation temperature, 600°C; source temperature, 150°C; cone gas flow rate, 50 L/h; desolvation gas flow rate, 1000 L/h; collision gas flow rate, 0.12 mL/min. MRM transitions, cone voltage, and collision energy as well as retention time (RT) were established for each analyte as listed in Table 2.

Validation protocol

Validation protocol applied in the present study included linearity, limits of detection (LOD), and quantification (LOQ), precision, accuracy, selectivity, carryover, matrix effect, recovery, and process efficiency, as elsewhere reported [23]. Validation parameters were calculated from five different daily replicates of QC samples along five subsequent working days. Over-the-curve samples (drug-free samples spiked with concentration of analytes five or 10 times higher than the highest calibration point) were tested for calibration curve fitting, precision, and accuracy once they were properly diluted.

Recovery, matrix effects, and process efficiency were calculated using the experimental protocol suggested by Matuszewski et al. [24]. Set 1 were five replicates of QC material prepared in the mobile phase. Sets 2 and 3 were five replicates of blank matrix samples fortified with QC material after and before extraction, respectively. Matrix effects were determined by dividing mean peak areas of set 2 by set 1 multiplied by 100. Recovery was determined by comparing the mean peak areas of analytes under investigation obtained in set 3 to those in set 2 multiplied by 100. Process efficiency expressed as the ratio of the mean peak area of an analyte spiked before extraction (set 3) to the mean peak area of the same analyte standards (set 1) multiplied by 100. The effect of three freeze-thaw cycles (storage at –20°C) on the compounds stability in urine and plasma was evaluated by repeated analysis (n=3) of QC samples. In addition, mid-term stability test was performed for real samples stored at –20°C. Three replicates of two samples were analyzed once a month during a 6-month period. The stability was expressed as a percentage of the initial concentration (first analyzed batch) of the analytes both in QC and real samples.

UHPLC-MS/MS method validation

Chromatograms obtained after the extraction of spiked plasma sample and spiked urine sample are shown in Figures 1A and 2A. Separation of analytes was obtained in less than 5 min and column re-equilibration in 10 min.

Figure 1:

UHPLC-MS/MS chromatogram of an extract of (A) drug-free plasma spiked with 50.0 ng/mL oxycodone, oxymorphone, and noroxycodone; (B) drug-free plasma sample; (C) plasma sample containing 62.0 ng/mL oxycodone, 0.9 ng/mL oxymorphone, and 32.4 ng/mL noroxycodone.

Figure 2:

UHPLC-MS/MS chromatogram of an extract of (A) drug-free urine spiked with 100.0 ng/mL oxycodone, oxymorphone, and noroxycodone; (B) drug-free urine sample; (C) urine sample containing 444.0 ng/mL oxycodone, 20.6 ng/mL oxymorphone, and 1234.8 ng/mL noroxycodone.

Linear calibration curves for all the analytes under investigation in urine and plasma samples showed determination coefficients (r²) equal or higher than 0.990 up to 500 ng/mL plasma and urine samples. LOD and LOQ values calculated for each analyte were adequate for the purpose of the present study (Table 3). The intra- and inter-assay precision (measured as coefficient of variation, CV%) and accuracy (measured as % error) values were always lower than 13% and mean absolute analytical recoveries obtained for the three different QC samples were always above 86% (Table 4). The sensitivity and the analytical recovery obtained by the here described method were both significantly better than those obtained in previously published method using HPLC-MS/MS [18].

No additional peaks due to endogenous substances that could have interfered with the detection of the compounds of interest were observed in drug-free samples (Figures 1B and 2B). Blank samples injected after the highest point of the calibration curve did not present any traces of carryover.

No significant ion suppression/enhancement (less than 10% analytical signal suppression due to matrix effect) occurred during chromatographic runs.

Regarding the freeze-thaw stability assays for QC samples, no significant degradation was observed after any of the three freeze-thaw cycles; the differences in concentration compared to the initial concentration were lower that 10%. Similar results (differences always lower than 10%) were obtained in case of mid-term stability test, performed re-analyzing replicates of two real plasma and urine samples once a month for 6-month period, assuring the validity of stored samples analysis.

Real samples

The method described here was applied to real plasma and urine samples collected from patients under treatment with oxycodone to evaluate the robustness and reliability of the method and the eventual applicability to compliance study and TDM. Representative chromatograms obtained following the extraction of real samples containing oxycodone, oxymorphone, and noroxycodone are shown in Figures 1C and 2C, and Table 5 shows the concentrations of oxycodone and its major metabolites obtained in the six different patients together with time interval between drug administration and sample collection and related clinical outcomes.

When analyte concentration in urine samples was higher than those in the calibration curve, samples were re-processed once diluted 5 or 10× (over-curve samples).

Following oral administration, oxycodone was excreted in the urine primarily as noroxycodone (concentration range, 86.0–20829.8 ng/mL), secondarily as parent drug (concentration range, 21.7–1510.6 ng/mL) and in lower concentration as oxymorphone (concentration range, 18.0–115.4 ng/mL). With respect to substances glucuronidation, in accordance with the study of Fang et al. [18], oxycodone and oxymorphone were primarily excreted in their glucuronide form. Indeed, after urine hydrolysis, oxycodone and oxymorphone concentrations showed about 2× and 10× concentration increase, respectively. The latter did not apply to noroxycodone.

Since oxycodone is widely abused [18], [25] and several fatalities have been attributed to this substance [26], [27], [28], [29], [30], [31], [32], urine analysis could provide a valuable tool to determine the use of this substance.

In five out of the six patients, plasma oxycodone concentrations were found in the range of 11.2–62.0 ng/mL. This concentration range falls within the therapeutic range reported elsewhere (i.e. 5–100 ng/mL) [33]. Only one patient (patient 4) had a concentration of oxycodone below the therapeutic range (i.e. 2.1 ng/mL), suggesting non-compliance. Indeed, when questioned about possible non compliance the patient referred to have taken the drug 20 min before the blood collection in order not to fail the therapy.

The analgesic effects of opioids are related to their plasma concentration. Either mean effective concentration (MEC) or minimum effective analgesic concentration (MEAC) is used to assess concentration-effect relationships.

Oxycodone treatment was effective or partially effective in some of the patients here tested whereas no substantial benefit was observed in others (Table 5). However, with these few data correlation between plasma concentration and clinical outcomes could not be established.

Generally speaking, the relationship between plasma oxycodone concentration and the analgesic response depends on the patients age, state of health, medical condition and the eventual extent of previous opioid treatment [34]. Moreover, variability in analgesic efficacy could be explained by inter-subject variations in plasma levels of parent drug and its active metabolite eventually affected also by co-medications [19], and therefore, dosage must be titrated for optimal effect and avoidance of toxicity after appropriate TDM.

The MEAC of oxycodone to achieve analgesia widely varies among patients, especially among patients who have been previously treated with potent agonist opioids. Thus, patients need to be treated with individualized titration of dosage to reach the desired effect. Moreover, repeated dosing due to an increase in pain and/or development of tolerance, may also increase the MEAC of this substance.

In any case, these were just preliminary data obtained in a very low number of patients with the principal aim of validating a simple and fast UHPLC assay to be applied in clinical TDM of oxycodone treatment.

It is very important for clinical practice to evaluate the compliance to the prescribed dose and to prevent abuse phenomena especially in outpatients cases, taking into consideration that adverse reactions and fatal outcomes can occur. Therefore, it is important to have appropriate diagnostic tools, like the one here described, to manage the therapeutic use of oxycodone as well as to identify potential diversion of therapy and abuse. The main limitation of this study is the small sample size tested and the heterogeneity of the cases (different doses, formulations, co-medications). However, a research on the same topic is ongoing and a larger sample size is to be analyzed to assess how co-medication (CYP2D6 and CYP3A4 inhibitors), dose and formulation can affect the efficacy of the drug on pain relief.