Advertisement
Research ArticleArticle

Intracellular CD3+ T Lymphocyte Teriflunomide Concentration Is Poorly Correlated with and Has Greater Variability Than Unbound Plasma Teriflunomide Concentration

Ashley M. Hopkins, Mahin Moghaddami, David J. R. Foster, Susanna M. Proudman, Richard N. Upton and Michael D. Wiese
Drug Metabolism and Disposition January 2017, 45 (1) 8-16; DOI: https://doi.org/10.1124/dmd.116.071985

Abstract

Leflunomide’s active metabolite teriflunomide inhibits dihydro-oroate dehydrogenase, an enzyme essential to proliferation of T lymphocytes. As teriflunomide must reach the target site to have this effect, this study assessed the distribution of teriflunomide into T lymphocytes, as intracellular concentrations may be a superior response biomarker to plasma concentrations. CD3 MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany) were used to extract CD3+ T cells from the peripheral blood of patients with rheumatoid arthritis who were taking a stable dose of leflunomide. Unbound plasma and intra-CD3+ T cell teriflunomide concentrations were quantified using liquid chromatography–mass spectrometry. Concentration (log transformed) and partition differences were assessed through paired Student t tests. Sixteen patients provided plasma steady-state teriflunomide samples, and eight provided a sample 6–12 weeks later. At time-point one, the geometric mean teriflunomide concentration (range) in CD3+ T cells was 18.12 μg/L (6.15–42.26 μg/L) compared with 69.75 μg/L (32.89–263.1 μg/L) unbound in plasma (P < 0.001). The mean partition coefficient (range) for unbound plasma teriflunomide into CD3+ T cells was 0.295 (0.092–0.632), which was significantly different from unity (P < 0.001). The median (range) change in teriflunomide concentration between the two time points was 14% (−10% to 40%) in unbound plasma and −29% (−69 to 138%) for CD3+ T cells. Because teriflunomide concentrations in CD3+ T cells were lower and displayed a higher intraindividual variability than the unbound plasma concentrations, its applicability as a therapeutic drug-monitoring marker may be limited.

Introduction

Leflunomide is a disease-modifying antirheumatic drug (DMARD) with comparable efficacy to methotrexate, the gold standard DMARD used in the treatment of rheumatoid arthritis (RA) (Alcorn et al., 2009). The therapeutic effects of leflunomide are mediated via its active metabolite teriflunomide (Breedveld and Dayer, 2000), which at therapeutic doses inhibits the mitochondrial enzyme dihydro-oroate dehydrogenase (DHODH), which is essential for de novo pyrimidine synthesis in activated T lymphocytes (Breedveld and Dayer, 2000). However, response to therapy is variable, and it may take up to 6 months for efficacy to be recognized, possibly due to the long half-life (∼15 days) of teriflunomide (Rozman, 2002). Furthermore, 20–40% of patients discontinue therapy due to toxicity (Aletaha et al., 2003; Alcorn et al., 2009). It has been hypothesized that this variability in response is related to the highly variable teriflunomide plasma concentrations, which range from 15 to 98 µg/mL and 3 to 150 µg/mL in individuals taking 10 and 20 mg doses, respectively (Schmidt et al., 2003; Van Roon et al., 2005; Bohanec Grabar, 2009). Although a total teriflunomide plasma concentration–effect relationship has been demonstrated, reported concentration thresholds for efficacy vary, ranging from 16 to 50 µg/mL (Chan et al., 2005; Van Roon et al., 2005; Bohanec Grabar, 2009), and, as such, target concentration interventions (TCI) are not used clinically.

Nonetheless, a superior target metric may enable the TCI strategy to be used to improve response to leflunomide. Because teriflunomide is extensively bound to plasma proteins (>99%), modest variability in the fraction unbound can lead to large variability in unbound teriflunomide concentrations, and unbound concentrations may be more closely related to response (Rakhila et al., 2011). Because the primary site of teriflunomide’s therapeutic effect is within T lymphocytes, intracellular concentrations may also be more predictive of response. Given the significant advances in the sensitivity of mass-spectrometry techniques over the past decade, quantification of intracellular concentrations appears feasible. As such, the aims of this study were as follows:

  1. Develop a method to quantify teriflunomide concentrations within peripheral blood (PB) CD3+ T lymphocytes (CD3+ T cells) and nontarget reference cells [i.e., red blood cells (RBC) and CD3− mononuclear cells (CD3− MNC)].

  2. Determine the partition coefficient of teriflunomide into CD3+ T cells and the nontarget cells with respect to unbound plasma teriflunomide concentrations.

  3. Determine the between-subject variability (BSV) and within-subject variability (WSV) of teriflunomide concentrations within CD3+ T cells, nontarget cells, and plasma.

  4. Determine the BSV and WSV of the partition coefficient of unbound plasma teriflunomide into CD3+ T cells and nontarget cells.

The overall hypothesis was that this information would lead to large-scale studies relating intracellular T lymphocyte teriflunomide concentrations to efficacy and toxicity, and ultimately allow a model-based analysis of the potential for intracellular T lymphocyte teriflunomide concentrations as a TCI.

Materials and Methods

Chemicals and Reagents

Teriflunomide (A77 1726) and D4-teriflunomide (A77 1726-D4; isotopic purity: 99%) were obtained from Toronto Research Chemical (Toronto, Canada) and stored in methanol (analytical grade; Mallinckrodt Chemicals, Phillipsburg, NJ) at −80°C. Water was Milli-RQ grade (EMD Millipore, Billerica, Massachusetts). Acetonitrile (analytical grade), ammonium acetate, zinc sulfate (ZnSO4), disodium phosphate (Na2HPO4), and sodium chloride (NaCl) were obtained from Merck (Darmstadt, Germany). Drug-free (blank) human plasma was prepared from PB, collected from healthy volunteers, and stored at −80°C. Lymphoprep was purchased from Axis-Shield (Oslo, Norway), whereas trypan blue solution (0.4%, sterile filtered) was from Sigma-Aldrich (Irvine, Scotland). CD3 MicroBeads human (lyophilized), MS MACS separation columns, and a MiniMACS magnetic separator were purchased from Miltenyi Biotec (Bergisch Gladbach, Germany). Endotoxin-free 1× phosphate-buffered saline (PBS) and fetal calf serum were obtained from the Media Production Unit (Institute of Medical and Veterinary Science, Adelaide, Australia). Wash buffer was prepared with PBS containing 2% fetal calf serum.

Study Population

Subjects attending the Early Arthritis Clinic at the Royal Adelaide Hospital were eligible for inclusion in this study. All participants were aged ≥18 years with DMARD-naive RA, according to revised American College of Rheumatology (ACR) Criteria (Arnett et al., 1988). RA was treated according to a previously published treat-to-target protocol, in which newly diagnosed patients were initiated on triple therapy (methotrexate, sulfasalazine, and hydroxychloroquine) (Proudman et al., 2007). In those failing to respond to optimal dosing of triple therapy, leflunomide was added. Inclusion criteria for this study included participants on a stable dose of leflunomide for the previous 8 weeks (to ensure steady-state plasma teriflunomide concentrations). Participants who remained on the same leflunomide dose at their next clinic visit (6–12 weeks later) were invited to provide a second blood sample. Patients gave informed written consent for inclusion in the study, and ethics approval was obtained from the Human Research Ethics Committee of the Royal Adelaide Hospital and the University of South Australia.

Sample Collections and Processing

At two consecutive clinical visits (T1 and T2), participants provided predose PB samples (∼40 mL), which were collected in heparinized vacutainers. A fraction of the PB sample (2 mL) was centrifuged at 10,000 rpm for 15 minutes at room temperature, after which plasma was removed and stored at −80°C for later assay. The volume of remaining PB was recorded and diluted 1:1 in sterile PBS. This was then layered on a Lymphoprep density gradient and centrifuged at 1717 rpm for 20 minutes at room temperature. The MNC and RBC layers were collected and stored on ice.

The RBC interface was washed with PBS and centrifuged at 3135 rpm for 10 minutes at 4°C; the supernatant was aspirated; and RBC were stored at −80°C for later assay. The MNC interface was washed with wash buffer and centrifuged at 1200 rpm for 10 minutes at 4°C, after which the supernatant was aspirated and the number of MNCs was counted.

CD3+ T cells were isolated from the harvested MNC by positive magnetic selection using CD3 MicroBeads (Miltenyi Biotec), according to the manufacturer’s instructions. Briefly, the harvested MNCs were resuspended in wash buffer, filtered through a nylon cell strainer (40 μm pore size; BD Biosciences, Franklin Lakes, New Jersey), and centrifuged at 1200 rpm for 10 minutes at 4°C. The supernatant was aspirated, and the pellet was resuspended in 80 μL wash buffer plus 20 μL CD3 MicroBeads per 107 total cells. Cells were mixed well and incubated for 15 minutes at 4°C, and then washed with 1–2 mL wash buffer per 107 total cells and centrifuged at 1200 rpm for 10 minutes at 4°C. The supernatant was aspirated, and the pellet was resuspended in 500 μL wash buffer. Within a magnetic field, the cell suspension was applied to MS MACS separation column and rinsed with 3× 500 μL wash buffer. CD3− MNC passed through the column and were collected, counted, centrifuged, and stored in 200 μL water at −80°C for later assay. The separation column was removed from the magnetic field and loaded with 1 mL wash buffer, and isolated CD3+ T cells were collected, counted, centrifuged, and stored in 200 μL water at −80°C.

LCMS-8050 Triple Quadrupole Liquid Chromatograph Mass Spectrometer (LC-MS/MS) Method

Separation and quantification of teriflunomide were performed using a modified method of Rakhila et al. (2011), and used a reversed-phase column [Kinetex 2.6 µm C18, 100A, 50X2.1 mm (Phenomenex, Torrance, California)] with a C18 pre-column cartridge [SecurityGuard Ultra 2.1 mm (Phenomenex)]. Separation of teriflunomide and D4-teriflunomide was achieved using two LC-30AS liquid chromatographs, a SIL-30AC autosampler, a CBM-20A communications bus module, a CTO-20AC column oven, a DGU-20A5R degassing unit (Shimadzu, Nakagyo-ku, Kyoto, Japan), and a reverse-phase gradient method. Mobile phase A was composed of water–acetonitrile–1 M ammonium acetate in water at 95:5:0.05 v/v/v, and mobile phase B was composed of water–acetonitrile–1 M ammonium acetate in water at 5:95:0.05 v/v/v. The flow rate was 0.5 mL/min. Initially, mobile-phase composition was maintained at 15% of mobile phase B from 0 to 0.2 minute, then from 0.2 to 1.0 minute mobile-phase B was increased via a linear gradient to 80%, and maintained at 80% until 1.5 minutes, after which the percentage of mobile phase B was reduced to 15% via a linear gradient from 1.5 to 2.5 minutes, which was maintained until 3.5 minutes, followed by the next injection. Detection occurred with the use of a LCMS-8050 triple quadrupole MS (Shimadzu). Quantitation was achieved in negative ion mode with a dwell time of 147 milliseconds. Q1/Q3 transition was 269.1/160.0 amu and 273.1/164.0 amu for teriflunomide and D4-teriflunomide, respectively. Calibration curves were obtained using a linear regression algorithm using the peak area ratio of teriflunomide to d4-teriflunomide, with a weighting factor of 1/C2. Calibration curves were deemed acceptable if the mean accuracy and percentage relative S.D. of the triplicate calibration standards were within ±15%, or within ±20% at the lower limit of quantification (LLOQ).

Plasma Sample Preparation.

Unbound teriflunomide was separated from bound using Rapid Equilibrium Dialysis plates, according to the manufacturer’s instructions (ThermoFisher, Waltham, MA), in which 400 µL plasma (adjusted to pH 7.4) was dialyzed against isotonic PBS solution (pH 7.4) for 6 hours. Bound and total teriflunomide were diluted 1:200 in PBS, and unbound fraction was neat. Calibration standards were prepared using blank freeze-thawed plasma diluted 1:200 with PBS. To 100 μL standard dilution, 10 μL D4-teriflunomide (100 μg/L) and increasing amounts of teriflunomide (1000 μg/L) were added to prepare solutions containing 0, 5, 10, 15, 20, 40, 100, 200, 300, 400, and 500 μg/L. Samples were made to 160 μL with methanol, to which ZnSO4 (100 μL, 0.2 M) and acetonitrile (100 μL) were added to precipitate proteins. Tubes were capped, vortexed for 10 seconds, and centrifuged at 4000 rpm for 10 minutes at 4°C, and the clear supernatant was removed. Quality controls (QC; 15, 80, and 400 μg/L) and test samples were treated in the same manner, except that methanol was added to the test samples instead of the teriflunomide stock solution. For all samples and calibration standards, 1 μL was injected onto the high-performance liquid chromatography (HPLC) column for analysis. Injection of each standard and QC was made in triplicate.

RBC Sample Preparation

Calibration standards were prepared using blank freeze-thawed packed RBC diluted 1:10 with water. To 100 μL dilution, 10 μL D4-teriflunomide (100 μg/L) and increasing amounts of teriflunomide (1000 μg/L) were added to prepare solutions containing 0, 5, 10, 15, 20, 40, 100, 200, 300, 400, and 500 μg/L. Samples were made to 160 μL with methanol, to which water (200 μL), ZnSO4 (300 μL, 0.2 M), and acetonitrile (300 μL) were added to precipitate proteins. Tubes were capped, vortexed for 10 seconds, and centrifuged at 4000 rpm for 10 minutes at 4°C, and the clear supernatant was removed. QCs (15, 80, and 400 μg/L) and test samples were treated in the same manner, except that methanol was added to the test samples instead of the teriflunomide stock solution. For analysis, 0.5 μL supernatant was injected onto the HPLC column. Injection of each standard and QC was made in triplicate.

Mononuclear Cell Sample Preparation

Calibration standards were prepared by using 200 μL thawed wash buffer, to which 10 μL D4-teriflunomide (20 μg/L) and increasing amounts of teriflunomide (20 μg/L) were added to prepare solutions containing 0, 5, 10, 20, 25, 50, 100, and 200 pg/sample. Samples were made to 310 μL with methanol, to which 300 μL acetonitrile was added. This solution was then vacuum dried at 1400 rpm and 60°C (Concentrator Plus; Eppendorf, Hamburg, Germany). The dried sample was reconstituted in 55 μL acetonitrile, 30 μL methanol, and 115 μL water, and vortexed for 15 minutes. The sample was then centrifuged at 4000 rpm for 10 minutes at 4°C, and the clear supernatant was removed. QC (20, 100, and 800 pg/sample) and test samples were treated in the same manner, except that methanol was added instead of teriflunomide stock solution, and wash buffer was exchanged for freeze-thawed CD3+ T cells or CD3− MNC samples. For analysis, 5 μL CD3+ T cell supernatant and 2 μL CD3− MNC supernatant were injected onto the HPLC column. Each run included a standard curve, and injection of each standard and QC was made in triplicate.

Statistical Analysis

For the present study, intracellular concentrations (pg/106 cells) of teriflunomide in CD3+ T cells and CD3− MNC were transformed into μg/L values, by considering that the volume of 106 CD3+ T cells and CD3− MNC was 1.76 × 10−7 L and 1.87 × 10−7 L, respectively (Chapman et al., 1981; Bazzoli et al., 2011). The partition of teriflunomide into cell matrices was calculated by dividing CD3+ T cell, CD3− MNC, and RBC teriflunomide concentrations by the unbound plasma teriflunomide concentration. The difference between the geometric means of log-transformed teriflunomide concentrations in CD3+ T cells, CD3− MNC, RBC, and plasma (unbound and total) between the two visits was assessed via a paired Student t test. A paired Student t test was also used to compare the difference of geometric means between log-transformed CD3+ T cell teriflunomide concentrations and unbound teriflunomide concentrations at T1. The mean of the partition coefficient of teriflunomide from unbound plasma to CD3+ T cells at T1 was compared against a unity of 1 via a paired Student t test.

The R Software Version 3.1.1 (R Core Team, 2015) was used for data manipulation, graphical output, and statistical analysis using the ggplot2, doBy, reshape, and GGally packages (Wickham, 2007, 2009; Højsgaard et al., 2014) (http://cran.r-project.org/web/packages/GGally/index.html). Linear mixed-effect modeling with the lme4 package (http://cran.r-project.org/web/packages/lme4/index.html) was used to assess the between and within individual variance of log-transformed CD3+ T cell, CD3− MNC, RBC, unbound and total teriflunomide concentrations, and unbound to CD3+ T cell, unbound to CD3− MNC, and unbound to RBC teriflunomide partition. Subject identification number (ID) was accounted for as a random effect, in which the S.D. of the intercept represents BSV and the S.D. of the residuals represents WSV (Winter, 2013).

Results

Patient Demographics.

Seventeen participants were recruited, nine of whom provided a second matched dose steady-state sample. One participant was excluded as the teriflunomide concentration within CD3+ T cells for the first sample was below the LLOQ, and for the second sample the teriflunomide concentrations within CD3+ T cells, CD3− MNC and RBC, and the total and unbound plasma concentration were below the LLOQ. The characteristics of the remaining 16 participants are described in Table 1.

View this table:
TABLE 1

Patient demographics

IQR presents the first and third quartiles.

The median (range) volume of whole blood was 30 mL (17.5–32.5 mL) for the 24 samples. The median (range) number of CD3+ T cells and CD3− MNC isolated from these samples was 9.30 × 106 cells (2.80 × 106–2.62 × 107 cells) and 5.78 × 107cells (3.00 × 107–8.10 × 107 cells), respectively.

LC-MS/MS Method.

The LLOQ for CD3+ T cells, CD3− MNC, RBC, and the total and unbound plasma teriflunomide concentrations were 10 pg/sample, 10 pg/sample, 100 μg/L, 1 mg/L, and 5 ug/L, respectively. The mean accuracy (percentage relative S.D.) of the triplicate calibration standards of the CD3+ T cell teriflunomide concentration curve at 10, 25, and 200 pg/sample was 97.9% (11.2%), 106.9% (7.2%), and 88.2% (0.65%), respectively. The CD3+ T cell teriflunomide concentration curve extended to the lowest LLOQ (10 pg/sample) of the investigated metrics, and all other curves (i.e., from the different matrices) performed equally well or better.

Intracellular Teriflunomide Concentrations.

For the 16 samples at T1, CD3+ T cell, CD3− MNC, RBC, unbound, and total plasma teriflunomide concentrations had a median (range; %CV) of 19.57 μg/L (6.15–42.26 μg/L; 49.03%), 21.42 μg/L (4.45–41.79 μg/L; 51.50%), 1463.9 μg/L (799.5–5039.4 μg/L; 60.84%), 67.06 μg/L (32.89–263.10 μg/L; 70.05%), and 27.58 mg/L (15.02–115.03 mg/L; 70.03%), respectively (Fig. 1). T1 CD3+ T cell teriflunomide concentrations were significantly lower than unbound plasma teriflunomide concentrations (P < 0.001). Geometric mean CD3+ T cell, CD3− MNC, RBC, unbound, and total teriflunomide concentrations were not significantly different between T1 and T2 (P > 0.05).

Fig. 1.
Fig. 1.

Box plot of CD3+ T cell, CD3− MNC, RBC, unbound, and total teriflunomide concentrations at T1 and T2.

The CD3+ T cell, CD3− MNC, RBC, unbound, and total plasma teriflunomide concentrations were described by the linear mixed-effect model represented below, in which subject ID is a random effect representing the BSV, and ε quantifies the unknown random effects (residuals), which in part represent the WSV (Winter, 2013):Embedded Image(1)For unbound and total teriflunomide concentrations, the BSV was greater than the WSV (Table 2). For the RBC teriflunomide concentrations, the BSV was comparable to the WSV. For CD3+ T cells and CD3− MNC teriflunomide concentrations, the BSV had a trend toward being less than the WSV, although there was overlap of the 90% confidence intervals of the S.D. estimates for the intercept and residuals (Table 2).

View this table:
TABLE 2

Linear mixed-effect modeling results

The S.D. estimate of the intercept and residuals, respectively, represents the BSV and WSV for the log-transformed teriflunomide concentration metrics and partition coefficient. The 90% confidence intervals (90% CI) of the S.D. estimates have also been represented.

Considering the T1 samples, the correlation between the unbound concentration of teriflunomide in plasma and the total plasma and RBC teriflunomide concentrations was high (r = 0.92 and 0.67, respectively) (Fig. 2). However, the correlation of unbound teriflunomide concentrations with the teriflunomide concentration in CD3+ T cells and CD3− MNC was noticeably lower (i.e., r < 0.5) (Fig. 2).

Fig. 2.
Fig. 2.

Correlation matrix of CD3+ T cell, CD3− MNC, RBC, unbound, and total teriflunomide concentrations at T1. (CD3 positive represents CD3+ T cells, and CD3 negative represents CD3− MNCs).

The median (range) partition coefficient of unbound teriflunomide to CD3+ T cells, CD3− MNC, and RBC at T1 was 0.243 (0.092–0.632), 0.310 (0.035–0.907), and 23.079 (9.639–37.248), respectively (Fig. 3). At T1, the coefficient representing the partition of unbound plasma teriflunomide to CD3+ T cells was significantly lower than unity (P < 0.001).

Fig. 3.
Fig. 3.

Box plot of the partition coefficients of unbound teriflunomide to CD3+ T cells, CD3− MNC, and RBC at T1.

The partition coefficient for unbound teriflunomide concentrations into CD3+ T cells, CD3− MNC, or RBC was described by the linear mixed-effect model represented below (Winter, 2013):Embedded Image(2)The BSV of the coefficient representing the partition of unbound plasma teriflunomide to CD3+ T cells, CD3− MNC, or RBC had a trend toward being less than the WSV (Table 2).

Within the eight individuals who provided two matched dose samples of CD3+ T cells, CD3− MNC, RBC, unbound, and total plasma, the median change (range) in teriflunomide concentration was −29% (−69%–138%), −15% (−73%–133%), −6% (−45%–27%), 14% (−10%–40%), and 1% (−22%–57%), respectively (Fig. 4). Furthermore, the median change (range) in the partition coefficient for unbound plasma teriflunomide into CD3+ T cells, CD3− MNC, and RBC was −42% (−77%–166%), −19% (−76%–71%), and −21% (−51%–15%), respectively (Fig. 5).

Fig. 4.
Fig. 4.

Matched dose steady-state CD3+ T cell, CD3− MNC, RBC, unbound, and total teriflunomide concentrations.

Fig. 5.
Fig. 5.

Matched dose partition coefficients for unbound teriflunomide concentrations to CD3+ T cells, CD3− MNC, and RBC.

Discussion

In RA patients treated with leflunomide, the concentration of teriflunomide within CD3+ T cells, CD3− MNC, and RBC was determined for the first time. The partition of teriflunomide into CD3+ T cells and CD3− MNC was limited, whereas accumulation occurred within RBC. The correlation between unbound plasma teriflunomide concentration and intra-CD3+ T cell and CD3− MNC concentrations was poor.

The site of action of most small-molecule medicines is within cells, and, although plasma concentrations have been associated with response to therapy, they are merely surrogates for drug concentration at the target site, which, in the case of leflunomide, is mitochondrial DHODH exposure to teriflunomide within CD3+ T cells (Dollery, 2013). Adequate surrogate markers are preferred over measuring drug directly at the target site, as, despite recent advances in mass-spectrometric techniques, there are substantial time, expense, and technical difficulty in measuring intracellular drug concentrations. However, total teriflunomide plasma concentrations, although demonstrating a concentration–effect relationship, are not used clinically as a TCI as concentration thresholds have varied significantly between studies (Chan et al., 2005; Van Roon et al., 2005). This is most likely a reflection of the higher WSV of teriflunomide concentrations within CD3+ T cells that was demonstrated in the present study, and the poor correlation with free plasma teriflunomide concentrations. The most useful markers for TCI generally require low intraindividual variability, so the higher WSV of teriflunomide concentrations in CD3+ T cells in this study does not support its use as a biomarker.

To date the most substantial work on intracellular drug concentrations has been carried out in HIV sufferers. The concentrations of zidovudine, lamivudine, tenofovir, carbovir, and their active metabolites have been explored in MNC, with the intent of identifying the reasons for suboptimal viral suppression (Hawkins et al., 2005; Bazzoli et al., 2011; Wang et al., 2011). In these studies, the intracellular pharmacokinetics have been shown to be more variable than plasma pharmacokinetics, the cause of which has not been fully elucidated (Wang et al., 2011). RBC concentrations of methotrexate polyglutamates have also shown some potential as a suitable efficacy marker in RA (Mohamed et al., 2015). Similar to teriflunomide, methotrexate acts within lymphocytes; however, intralymphocyte concentrations have not been explored due to limitations with analytical sensitivity (Mohamed et al., 2015). Over the past decade, analytical methods have improved (Dollery, 2013), enabling measurement of intra-CD3+ T cell teriflunomide and the subsequent finding of a poor correlation between teriflunomide concentrations in CD3+ T cells and RBC. This highlights the potential dissimilarity between surrogate cell types and target cells, although this may vary from drug to drug according to pharmacochemical properties.

Only unbound unionized drug can passively diffuse into and out of cells. Given teriflunomide is a weak acid (pKa ∼5.48) (Wiese et al., 2013), and plasma pH is tightly controlled between 7.38 and 7.42 (Atherton, 2003), approximately 99% (range, 98.8%–98.9%) of unbound teriflunomide will be ionized within the plasma (Po and Senozan, 2001). Despite this, plasma pH may account for some of the WSV of intracellular concentrations, as the permeability of teriflunomide in Caco-2-TC7 cells was decreased by approximately 50% with an increase in pH from 6.5 to 7.4 (http://www.accessdata.fda.gov/drugsatfda_docs/nda/2012/202992Orig1s000ClinpharmR.pdf), whereas predictions indicate that a realistic plasma pH change from 7.38 to 7.42 would halve unbound unionized teriflunomide concentrations from 0.2% to 0.1%. Intracellular pH (pHi) is also important, and it can vary between cell types and even within cell structures (Casey et al., 2010; Chen et al., 2013). Typically, RBC pHi (7.06–7.27) is lower than extracellular pH, and peripheral blood lymphocytes maintain a pHi of approximately 7.17 (Roos and Boron, 1981; Deutsch et al., 1982). Typically, ion trapping occurs when differences between intracellular and extracellular pH result in more ionized intracellular drug. As the pHi of cells assessed in this work was expected to be slightly lower than the plasma pH, a minor increase in the unionized percentage of teriflunomide would be expected within cells. Thus, ion trapping, although not directly measured, would appear unlikely to describe the observed intracellular/unbound plasma partition coefficients.

The duration that cells are exposed to unbound unionized teriflunomide is likely to affect whether equilibrium between intracellular and plasma concentrations is reached. RBC have a life span between 70 and 140 days (Franco, 2012), whereas CD3+ lymphocytes have complex maturation profiles and life cycles, depending upon cell type and underlying disease status (Reinherz and Schlossman, 1980; Hellerstein et al., 1999; Appay et al., 2008; Ladell et al., 2008). RBC turnover is relatively constant within an individual; therefore, RBC teriflunomide concentrations will reach a relatively consistent pseudo-steady state. In contrast, circulating T lymphocyte numbers can vary between, and even during a day (Miyawaki et al., 1984). T lymphocytes are also central to the pathogenesis of RA, which has characteristic fluctuations in disease activity, and the resultant effects on cell circulation are currently poorly defined (De Boer et al., 2003; Iikuni et al., 2007; Straub and Cutolo, 2007). Typically, RA increases the turnover rate of lymphocytes, before ultimately resulting in a deterioration of the immune system and T lymphocyte repertoire exhaustion, yet naive or memory CD4+ T lymphocyte numbers are maintained (Koetz et al., 2000; De Boer et al., 2003). It follows that exposure time of CD3+ T cells to unbound unionized teriflunomide may fluctuate between and within individuals. Thus, if the time to reach equilibrium between unbound unionized plasma and intracellular teriflunomide concentrations is longer than the turnover rate of cells, the reported partition coefficient may be influenced by the turnover rate. Consequently, the time to reach equilibrium between plasma and intracellular teriflunomide concentrations should be investigated, particularly in cell subtypes that perish rapidly and at varying plasma pHs.

Different teriflunomide concentrations within CD3+ T cells, CD3− MNC, and RBC may reflect differences in cellular physiologies between individuals and within different cell types (including lymphocyte subpopulations). For example, there is differing expression of influx and efflux transporters (Chaudhary et al., 1992; Albermann et al., 2005; Prasad et al., 2013), such as the ATP-binding cassette subfamily G member 2, which has been associated with altered teriflunomide plasma concentrations (Kim et al., 2011; Wiese et al., 2012). Varying expression of influx and efflux transporters may also affect the time to reach steady-state intracellular teriflunomide concentrations. Furthermore, mature erythrocytes are rich in hemoglobin, yet barren of organelles (e.g., mitochondria and nucleus), which can express and synthesize transporters (Wohlrab, 2009). Similar to the high plasma protein binding of teriflunomide, RBC may accumulate teriflunomide through intracellular protein binding, whereas efflux transportation is likely to be lower than within lymphocytes. Conversely, there is potential for differing transporter expression between circulating CD4+, CD8+, CD16+, and NK T lymphocytes (among others) (Boldt et al., 2014), which adds another degree of variability to the concentration of teriflunomide within CD3+ T cells. Therefore, teriflunomide concentrations within T lymphocyte subpopulations should be investigated, as this may reduce the variability and provide a superior efficacy biomarker.

In this study, concentrations were explored within individuals on stable doses of leflunomide, and we assumed that, because teriflunomide has a plasma half-life of approximately 2 weeks (mean ∼15.7 days) (Rozman, 2002), steady-state concentrations in cell subtypes would also be reached with 8 weeks of stable dosing. Furthermore, the relationship between teriflunomide concentrations in any matrices with either efficacy or toxicity was not investigated. The isolation of CD3+ T cells is tedious and time consuming, with variations in temperature, washing procedure, and time likely to affect extraction and leaching. Although a strict protocol was used to minimize these affects, the process most likely adds variability to measured concentrations (Wang et al., 2011).

In this study, the sample size was sufficient to determine that CD3+ T cell teriflunomide concentrations are lower than unbound plasma teriflunomide concentrations, which has not been reported previously. Linear mixed-effect modeling was used to assess the between- and within-individual variance of CD3+ T cell, CD3− MNC, RBC, unbound, and total teriflunomide concentrations. These analyses indicated that the BSV of unbound and total teriflunomide concentrations was greater than the WSV, which aligns with previous literature (Schmidt et al., 2003; Van Roon et al., 2005; Bohanec Grabar, 2009; Rakhila et al., 2011). However, this was in contrast with the results for the CD3+ T cell, CD3− MNC teriflunomide concentrations, in which there was a trend for the BSV to be less than the WSV. These results will need to be explored further in a larger population analyses; however, at this stage, these preliminary results do not support the use of CD3+ T cell teriflunomide concentrations as a useful marker for TCI. In this study, it was also assumed that the CD3+ T cell, CD3− MNC, RBC, unbound, and total teriflunomide concentrations, which are physiologic variables, would be log-normally distributed for a population, and that the partition coefficient of free teriflunomide to CD3+ T cell, CD3− MNC, and RBC teriflunomide would be normally distributed for a population.

For this new field of investigation, it would be desirable to conduct a prospective analysis with a rich data set to analyses both intraday and daily intracellular teriflunomide concentrations, in which the impact of food, concomitant drugs, other environmental factors and transporters is as yet largely unknown. A large prospective analysis that includes patients on stable doses of leflunomide should also be conducted, as this may capture the time to reach intracellular steady state as well as a larger range of CD3+ T cell teriflunomide concentrations. Such a prospective study may also assist the investigation of the relationship between CD3+ T cell teriflunomide concentration and toxicity or efficacy. It would be of particular interest to identify whether CD3+ T cell teriflunomide concentrations are more closely associated with efficacy compared with total concentrations, which to date have demonstrated highly variable thresholds (Chan et al., 2005; Van Roon et al., 2005; Bohanec Grabar, 2009). Because teriflunomide’s precise location of action is mitochondrial DHODH (Breedveld and Dayer, 2000), determining teriflunomide concentrations within T lymphocyte mitochondria may also be important for future research, although organelle isolation and analytical sensitivity make such a determination technically difficult at this time.

In conclusion, teriflunomide concentrations within CD3+ T cells displayed higher intraindividual variability and were lower than unbound plasma teriflunomide concentrations. Furthermore, the correlation of CD3+ T cells and RBC teriflunomide concentrations was low, highlighting that, although abundant, RBC appear to be a poor surrogate for drug at the target site. CD3+ T cell subpopulation analysis may be important, as correlation with intracellular CD3− MNC teriflunomide concentrations was low, highlighting the potential impact of different cellular physiologies, pHi, and circulation profiles.

Acknowledgments

We thank Leah McWilliams, Llewellyn Spargo, and Professor Michael James for assistance with data and sample collection.

Authorship Contributions

Participated in research design: Hopkins, Proudman, Wiese.

Conducted experiments: Hopkins, Moghaddami, Wiese.

Contributed new reagents or analytic tools: Hopkins, Moghaddami, Foster, Proudman, Upton, Wiese.

Performed data analysis: Hopkins, Foster, Upton, Wiese.

Wrote or contributed to the writing of the manuscript: Hopkins, Moghaddami, Foster, Proudman, Upton, Wiese.

Footnotes

Abbreviations

BSV
between-subject variability
DHODH
dihydro-oroate dehydrogenase
DMARD
disease-modifying antirheumatic drug
HPLC
high-performance liquid chromatography
ID
identification number
LLOQ
lower limit of quantification
MNC
mononuclear cell
PB
peripheral blood
PBS
phosphate-buffered saline
pHi
intracellular pH
QC
quality control
RA
rheumatoid arthritis
RBC
red blood cell
TCI
target concentration interventions
WSV
within-subject variability

References

  1. ↵
    1. Albermann N,
    2. Schmitz-Winnenthal FH,
    3. Z’graggen K,
    4. Volk C,
    5. Hoffmann MM,
    6. Haefeli WE, and
    7. Weiss J
    (2005) Expression of the drug transporters MDR1/ABCB1, MRP1/ABCC1, MRP2/ABCC2, BCRP/ABCG2, and PXR in peripheral blood mononuclear cells and their relationship with the expression in intestine and liver. Biochem Pharmacol 70:949–958.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Alcorn N,
    2. Saunders S, and
    3. Madhok R
    (2009) Benefit-risk assessment of leflunomide: an appraisal of leflunomide in rheumatoid arthritis 10 years after licensing. Drug Saf 32:1123–1134.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Aletaha D,
    2. Stamm T,
    3. Kapral T,
    4. Eberl G,
    5. Grisar J,
    6. Machold KP, and
    7. Smolen JS
    (2003) Survival and effectiveness of leflunomide compared with methotrexate and sulfasalazine in rheumatoid arthritis: a matched observational study. Ann Rheum Dis 62:944–951.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. Appay V,
    2. van Lier RAW,
    3. Sallusto F, and
    4. Roederer M
    (2008) Phenotype and function of human T lymphocyte subsets: consensus and issues. Cytometry A 73:975–983.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Arnett FC,
    2. Edworthy SM,
    3. Bloch DA,
    4. McShane DJ,
    5. Fries JF,
    6. Cooper NS,
    7. Healey LA,
    8. Kaplan SR,
    9. Liang MH,
    10. Luthra HS,
    11. et al.
    (1988) The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 31:315–324.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Atherton JC
    (2003) Acid-base balance: maintenance of plasma pH. Anaesth Intensive Care Med 4:419–422.
    OpenUrl
  7. ↵
    1. Bazzoli C,
    2. Bénech H,
    3. Rey E,
    4. Retout S,
    5. Salmon D,
    6. Duval X,
    7. Tréluyer J-M,
    8. Mentré F, and
    9. COPHAR2-ANRS 111 Study Group
    (2011) Joint population pharmacokinetic analysis of zidovudine, lamivudine, and their active intracellular metabolites in HIV patients. Antimicrob Agents Chemother 55:3423–3431.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Bohanec Grabar P,
    2. Grabnar I,
    3. Rozman B,
    4. Logar D,
    5. Tomsic M,
    6. Suput D,
    7. Trdan T,
    8. Peterlin Masic L,
    9. Mrhar A, and
    10. Dolzan V
    (2009) Investigation of the influence of CYP1A2 and CYP2C19 genetic polymorphism on 2-Cyano-3-hydroxy-N-[4-(trifluoromethyl)phenyl]-2-butenamide (A77 1726) pharmacokinetics in leflunomide-treated patients with rheumatoid arthritis. Drug Metab Dispos 37:2061–2068.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Boldt A,
    2. Borte S,
    3. Fricke S,
    4. Kentouche K,
    5. Emmrich F,
    6. Borte M,
    7. Kahlenberg F, and
    8. Sack U
    (2014) Eight-color immunophenotyping of T-, B-, and NK-cell subpopulations for characterization of chronic immunodeficiencies. Cytometry B Clin Cytom 86:191–206.
    OpenUrl
  10. ↵
    1. Breedveld FC and
    2. Dayer JM
    (2000) Leflunomide: mode of action in the treatment of rheumatoid arthritis. Ann Rheum Dis 59:841–849.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Casey JR,
    2. Grinstein S, and
    3. Orlowski J
    (2010) Sensors and regulators of intracellular pH. Nat Rev Mol Cell Biol 11:50–61.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Chan V,
    2. Charles BG, and
    3. Tett SE
    (2005) Population pharmacokinetics and association between A77 1726 plasma concentrations and disease activity measures following administration of leflunomide to people with rheumatoid arthritis. Br J Clin Pharmacol 60:257–264.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Chapman EH,
    2. Kurec AS, and
    3. Davey FR
    (1981) Cell volumes of normal and malignant mononuclear cells. J Clin Pathol 34:1083–1090.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Chaudhary PM,
    2. Mechetner EB, and
    3. Roninson IB
    (1992) Expression and activity of the multidrug resistance P-glycoprotein in human peripheral blood lymphocytes. Blood 80:2735–2739.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Chen S,
    2. Hong Y,
    3. Liu Y,
    4. Liu J,
    5. Leung CW,
    6. Li M,
    7. Kwok RT,
    8. Zhao E,
    9. Lam JW,
    10. Yu Y,
    11. et al.
    (2013) Full-range intracellular pH sensing by an aggregation-induced emission-active two-channel ratiometric fluorogen. J Am Chem Soc 135:4926–4929.
    OpenUrl
  16. ↵
    1. De Boer RJ,
    2. Mohri H,
    3. Ho DD, and
    4. Perelson AS
    (2003) Turnover rates of B cells, T cells, and NK cells in simian immunodeficiency virus-infected and uninfected rhesus macaques. J Immunol 170:2479–2487.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Deutsch C,
    2. Taylor JS, and
    3. Wilson DF
    (1982) Regulation of intracellular pH by human peripheral blood lymphocytes as measured by 19F NMR. Proc Natl Acad Sci USA 79:7944–7948.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Dollery CT
    (2013) Intracellular drug concentrations. Clin Pharmacol Ther 93:263–266.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Franco RS
    (2012) Measurement of red cell lifespan and aging. Transfus Med Hemother 39:302–307.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Hawkins T,
    2. Veikley W,
    3. St Claire RL 3rd.,
    4. Guyer B,
    5. Clark N, and
    6. Kearney BP
    (2005) Intracellular pharmacokinetics of tenofovir diphosphate, carbovir triphosphate, and lamivudine triphosphate in patients receiving triple-nucleoside regimens. J Acquir Immune Defic Syndr 39:406–411.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Hellerstein M,
    2. Hanley MB,
    3. Cesar D,
    4. Siler S,
    5. Papageorgopoulos C,
    6. Wieder E,
    7. Schmidt D,
    8. Hoh R,
    9. Neese R,
    10. Macallan D,
    11. et al.
    (1999) Directly measured kinetics of circulating T lymphocytes in normal and HIV-1-infected humans. Nat Med 5:83–89.
    OpenUrlCrossRefPubMed
  22. ↵
    Højsgaard S, Halekoh U, Robison-Cox J, Wright K, and Leidi AA (2014) doBy: doBy - Groupwise summary statistics, LSmeans, general linear contrasts, various utilities: CRAN.R-project.org.
  23. ↵
    1. Iikuni N,
    2. Nakajima A,
    3. Inoue E,
    4. Tanaka E,
    5. Okamoto H,
    6. Hara M,
    7. Tomatsu T,
    8. Kamatani N, and
    9. Yamanaka H
    (2007) What’s in season for rheumatoid arthritis patients? Seasonal fluctuations in disease activity. Rheumatology 46:846–848.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Kim KA,
    2. Joo HJ, and
    3. Park JY
    (2011) Effect of ABCG2 genotypes on the pharmacokinetics of A771726, an active metabolite of prodrug leflunomide, and association of A771726 exposure with serum uric acid level. Eur J Clin Pharmacol 67:129–134.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Koetz K,
    2. Bryl E,
    3. Spickschen K,
    4. O’Fallon WM,
    5. Goronzy JJ, and
    6. Weyand CM
    (2000) T cell homeostasis in patients with rheumatoid arthritis. Proc Natl Acad Sci USA 97:9203–9208.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Ladell K,
    2. Hellerstein MK,
    3. Cesar D,
    4. Busch R,
    5. Boban D, and
    6. McCune JM
    (2008) Central memory CD8+ T cells appear to have a shorter lifespan and reduced abundance as a function of HIV disease progression. J Immunol 180:7907–7918.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Miyawaki T,
    2. Taga K,
    3. Nagaoki T,
    4. Seki H,
    5. Suzuki Y, and
    6. Taniguchi N
    (1984) Circadian changes of T lymphocyte subsets in human peripheral blood. Clin Exp Immunol 55:618–622.
    OpenUrlPubMed
  28. ↵
    1. Mohamed HJ,
    2. Sorich MJ,
    3. Kowalski SM,
    4. McKinnon R,
    5. Proudman SM,
    6. Cleland L, and
    7. Wiese MD
    (2015) The role and utility of measuring red blood cell methotrexate polyglutamate concentrations in inflammatory arthropathies: a systematic review. Eur J Clin Pharmacol 71:411–423.
    OpenUrl
  29. ↵
    1. Po HN and
    2. Senozan NM
    (2001) The Henderson-Hasselbalch equation: its history and limitations. J Chem Educ 78:1499–1503.
    OpenUrlCrossRef
  30. ↵
    1. Prasad B,
    2. Lai Y,
    3. Lin Y, and
    4. Unadkat JD
    (2013) Interindividual variability in the hepatic expression of the human breast cancer resistance protein (BCRP/ABCG2): effect of age, sex, and genotype. J Pharm Sci 102:787–793.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Proudman SM,
    2. Keen HI,
    3. Stamp LK,
    4. Lee ATY,
    5. Goldblatt F,
    6. Ayres OC,
    7. Rischmueller M,
    8. James MJ,
    9. Hill CL,
    10. Caughey GE,
    11. et al.
    (2007) Response-driven combination therapy with conventional disease-modifying antirheumatic drugs can achieve high response rates in early rheumatoid arthritis with minimal glucocorticoid and nonsteroidal anti-inflammatory drug use. Semin Arthritis Rheum 37:99–111.
    OpenUrlCrossRefPubMed
  32. ↵
    R Core Team (2015) R: A language and environment for statistical computing., in R Foundation for Statistical Computing, Vienna, Austria, URL http://www.R-project.org/.
  33. ↵
    1. Rakhila H,
    2. Rozek T,
    3. Hopkins A,
    4. Proudman S,
    5. Cleland L,
    6. James M, and
    7. Wiese M
    (2011) Quantitation of total and free teriflunomide (A77 1726) in human plasma by LC-MS/MS. J Pharm Biomed Anal 55:325–331.
    OpenUrlCrossRefPubMed
  34. ↵
    1. Reinherz EL and
    2. Schlossman SF
    (1980) The differentiation and function of human T lymphocytes. Cell 19:821–827.
    OpenUrlCrossRefPubMed
  35. ↵
    1. Roos A and
    2. Boron WF
    (1981) Intracellular pH. Physiol Rev 61:296–434.
    OpenUrlFREE Full Text
  36. ↵
    1. Rozman B
    (2002) Clinical pharmacokinetics of leflunomide. Clin Pharmacokinet 41:421–430.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Schmidt A,
    2. Schwind B,
    3. Gillich M,
    4. Brune K, and
    5. Hinz B
    (2003) Simultaneous determination of leflunomide and its active metabolite, A77 1726, in human plasma by high-performance liquid chromatography. Biomed Chromatogr 17:276–281.
    OpenUrlCrossRefPubMed
  38. ↵
    1. Straub RH and
    2. Cutolo M
    (2007) Circadian rhythms in rheumatoid arthritis: implications for pathophysiology and therapeutic management. Arthritis Rheum 56:399–408.
    OpenUrlCrossRefPubMed
  39. ↵
    1. van Roon EN,
    2. Jansen TLTA,
    3. van de Laar MAFJ,
    4. Janssen M,
    5. Yska JP,
    6. Keuper R,
    7. Houtman PM, and
    8. Brouwers JRBJ
    (2005) Therapeutic drug monitoring of A77 1726, the active metabolite of leflunomide: serum concentrations predict response to treatment in patients with rheumatoid arthritis. Ann Rheum Dis 64:569–574.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Wang L,
    2. Soon GH,
    3. Seng K-Y,
    4. Li J,
    5. Lee E,
    6. Yong E-L,
    7. Goh B-C,
    8. Flexner C, and
    9. Lee L
    (2011) Pharmacokinetic modeling of plasma and intracellular concentrations of raltegravir in healthy volunteers. Antimicrob Agents Chemother 55:4090–4095.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Wickham H
    (2007) Reshaping data with the {reshape} package. J Stat Softw 21:1–20.
    OpenUrl
  42. ↵
    Wickham H (2009) ggplot2: elegant graphics for data analysis: CRANR-projectorg, Springer, New York.
  43. ↵
    1. Wiese MD,
    2. Rowland A,
    3. Polasek TM,
    4. Sorich MJ, and
    5. O’Doherty C
    (2013) Pharmacokinetic evaluation of teriflunomide for the treatment of multiple sclerosis. Expert Opin Drug Metab Toxicol 9:1025–1035.
    OpenUrlCrossRefPubMed
  44. ↵
    1. Wiese MD,
    2. Schnabl M,
    3. O’Doherty C,
    4. Spargo LD,
    5. Sorich MJ,
    6. Cleland LG, and
    7. Proudman SM
    (2012) Polymorphisms in cytochrome P450 2C19 enzyme and cessation of leflunomide in patients with rheumatoid arthritis. Arthritis Res Ther 14:R163.
    OpenUrlPubMed
  45. ↵
    Winter B (2013) Linear models and linear mixed effects models in R with linguistic applications. arXiv preprint arXiv:13085499. Cornell University Library, New York. 42.
  46. ↵
    1. Wohlrab H
    (2009) Transport proteins (carriers) of mitochondria. IUBMB Life 61:40–46.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleArticle

CD3+ T Cell Teriflunomide Concentration Quantification

Ashley M. Hopkins, Mahin Moghaddami, David J. R. Foster, Susanna M. Proudman, Richard N. Upton and Michael D. Wiese
Drug Metabolism and Disposition January 1, 2017, 45 (1) 8-16; DOI: https://doi.org/10.1124/dmd.116.071985
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Advertisement