ARTHRITIS & RHEUMATISM Vol. 58, No. 11, November 2008, pp 3299–3308 DOI 10.1002/art.24034 © 2008, American College of Rheumatology Pharmacokinetics of Oral Methotrexate in Patients With Rheumatoid Arthritis Judith M. Dalrymple, Lisa K. Stamp, John L. O’Donnell, Peter T. Chapman, Mei Zhang, and Murray L. Barclay Objective. There is evidence supporting a therapeutic range for methotrexate polyglutamate (MTXGlu) concentrations in the treatment of rheumatoid arthritis (RA). Knowledge of the pharmacokinetics of MTXGlu1–5 is required for optimal timing of blood sampling. The aim of this study was to determine the time to steady state and the half-life of accumulation of red blood cell (RBC) MTXGlu1–5 in patients with RA commencing oral MTX, and the time for RBC MTXGlu1–5 to become undetectable and the half-life of elimination of RBC MTXGlu1–5 in patients ceasing treatment with oral MTX. Methods. Ten patients beginning treatment and 10 patients stopping treatment with low-dose oral MTX were recruited. Blood samples were initially collected weekly, with gradual extension to monthly collection over the study period. RBC MTXGlu1–5 concentrations were assayed by high-performance liquid chromatography. Results were analyzed using a first-order exponential method. Results. The median times to reach steady state in RBCs (defined as 90% of the maximum concentration) were 6.2, 10.6, 41.2, 149, and 139.8 weeks, respectively, for MTXGlu1, MTXGlu2, MTXGlu3, MTXGlu4, and MTXGlu5. The median half-life of accumulation for RBC MTXGlu1–5 ranged from 1.9 weeks to 45.2 weeks. The median times for MTXGlus to become undetectable in RBCs were 4.5, 5.5, 10, 6, and 4 weeks, respectively, for MTXGlu1, MTXGlu2, MTXGlu3, MTXGlu4, and MTXGlu5. The median half-life of elimination for RBC MTXGlu1–5 ranged from 1.2 weeks to 4.3 weeks. Conclusion. There is wide interpatient variability of RBC MTXGlu1–5 accumulation and elimination in adults with RA. These data also suggest that after a dose change, >6 months are required for RBC MTXGlu1–5 to reach steady state. Such delays in achieving steady state suggest that more rapid dose escalation or subcutaneous administration from the outset should be considered. Methotrexate (MTX) is one of the most commonly used disease-modifying antirheumatic drugs (DMARDs). It is the first-line agent for the treatment of rheumatoid arthritis (RA) (1,2) and is suggested to be the “anchor” drug in all RA therapy (3,4). Joint damage occurs early in the course of RA (5,6) and once present is largely irreversible. Disease activity is an important predictor of the progression of joint damage (7,8) and the long-term requirement for joint-replacement surgery (9). Effective control of RA disease activity reduces radiographic progression of joint damage and improves physical function and quality of life (10–12). Thus, the primary goal of therapy is to achieve rapid, effective disease control to prevent the long-term damaging effects on joint structure and function. Serum concentrations of MTX fall rapidly following intravenous administration (13). However, MTX is transported intracellularly via the reduced folate carrier and is retained within cells long after it has been eliminated from serum. The parent drug contains 1 glutamate moiety and is also known as MTX glutamate 1 (MTXGlu1). Once inside red blood cells (RBCs), up to ACTRN: 012606000275561. Supported by the Health Research Council of New Zealand, the New Zealand Pharmacy and Education Research Foundation, Arthritis New Zealand, and the Canterbury Rheumatology and Immunology Research Trust. Judith M. Dalrymple, BPharm, Grad Dip Clin Pharm, Lisa K. Stamp, MBChB, FRACP, PhD, John L. O’Donnell, MBChB, FRACP, FRCPA, Peter T. Chapman, MBChB, MD, FRACP, Mei Zhang, PhD, Murray L. Barclay, MBChB, MD, FRACP: University of Otago, Christchurch, New Zealand. Address correspondence and reprint requests to Lisa K. Stamp, MBChB, FRACP, PhD, Department of Medicine, University of Otago, PO Box 4345, Christchurch 8140, New Zealand. E-mail: email@example.com. Submitted for publication March 26, 2008; accepted in revised form August 11, 2008. 3299 3300 4 additional glutamate moieties are added via the enzyme folylpolyglutamate synthetase (FPGS) (14). As a group, these products of intracellular glutamation are referred to as MTXGlun, where n represents the number of glutamate moieties. Gamma-glutamyl hydrolase removes terminal glutamate molecules, returning MTX to its monoglutamate form, which is rapidly transported out of the cell by multidrug-resistant proteins. MTXGlus can be measured inside RBCs, and concentrations within RBCs are thought to be representative of concentrations within other cells such as lymphocytes. The mechanism of action of MTX in RA is unclear, but it appears to act as a folate antagonist (15). Intracellular MTXGlun bind to dihydrofolate reductase and other folate pathway enzymes, thereby providing antiinflammatory effects. The dose of MTX required for treatment of individual patients with RA varies and is largely unpredictable. In patients with RA, both a trend toward a dose-response relationship (16,17) and no association between dose and response have been reported with oral MTX (18). Serum MTX concentrations have been shown to have no correlation with disease activity (19). However, recent studies suggest a correlation between RBC MTXGlun concentrations and disease activity (18,20,21). The first of these studies in 65 patients with RA showed that RBC MTXGlu1–5 concentrations were significantly higher in responders (mean ⫾ SD 60.7 ⫾ 18.9 nmoles/liter) and partial responders (50.8 ⫾ 23.3 nmoles/liter) to treatment than in nonresponders (21.5 ⫾ 10.5 nmoles/liter) (20). The initial study by Dervieux and colleagues (18) of 108 patients was then extended to 226 patients (21), and patients with RBC MTXGlu3 levels ⬍60 nmoles/liter were more likely to have a poor clinical response. In addition, in a small group of 23 patients commencing MTX treatment, a higher RBC MTXGlu3 concentration at 3 months was associated with a greater chance of better disease control at 6 months (22). If MTXGlun drug concentrations are relevant to the clinical response, then the time required for adequate drug concentrations to be achieved in patients starting treatment with MTX is important. In New Zealand, treatment of patients with RA generally begins with oral MTX at a dosage of 7.5–10 mg/week, with dose escalations made according to the clinical response. In comparison, patients with inflammatory bowel disease are started on a dosage of 25 mg/week, administered subcutaneously. With the slower dosing strategy, valuable time to achieve adequate disease control may be lost depending on the time required for adequate drug DALRYMPLE ET AL concentrations to be achieved. Currently, the accumulation and elimination rates of RBC MTX polyglutamates are poorly defined. There are no data describing the rate of accumulation of individual MTXGlun in human RBCs, and only limited data are available describing the rate of elimination of individual RBC MTXGlun in children (23). The aims of this study were to determine the time to steady state of RBC MTXGlu1–5 in adult patients with RA starting treatment with oral low-dose MTX and the length of time required for MTXGlu1–5 to become undetectable in the RBCs of adult patients with RA who are stopping treatment with low-dose oral MTX. Finally, we aimed to determine the half-life of accumulation and elimination of RBC MTXGlu1–5 in adult patients with RA. PATIENTS AND METHODS Ethical approval was obtained from the Upper South B Regional Ethics Committee. Patients ages 18 years and older with RA, as defined by the American College of Rheumatology (ACR; formerly, the American Rheumatism Association) 1987 revised criteria for the classification of RA (24), who began or stopped treatment with oral MTX between October 2005 and October 2007 were invited to participate. Patients stopping MTX were included only if they had been taking MTX for at least 3 months, with a stable dosage for at least 1 month. Baseline demographic data were collected for all patients, including ethnicity, sex, age, height and weight, duration of RA, presence of rheumatoid factor, radiographic erosions and rheumatoid nodules, and serum creatinine. An estimated glomerular filtration rate from serum creatinine was calculated according to the equation described by the Modification of Diet in Renal Disease Study Group (25). Patients were seen at weeks 8, 16, and 24, and a record of each patient’s current drug therapy was obtained. In patients starting treatment, standard MTX toxicity monitoring was undertaken, according to the ACR recommendations. Blood sampling strategy. Blood samples for RBC MTXGlu1–5 assays were obtained within 36 hours prior to administration of the weekly MTX dose (i.e., a trough concentration). In patients starting MTX treatment, samples were collected weekly until week 8, then fortnightly until the MTX dose had been titrated to the clinically effective dose. Thereafter, blood samples were drawn every 4 weeks until 24 weeks after reaching the maintenance dose, or until the patient withdrew from the study. In those patients stopping MTX, samples were collected weekly for 8 weeks, then fortnightly for 8 weeks, and then every 4 weeks for 24–32 weeks after treatment with MTX ceased. High-performance liquid chromatography (HPLC) of RBC MTXGlun. A 5-ml whole-blood sample was collected in an EDTA-coated tube and centrifuged to isolate the RBCs. The RBCs were washed twice in 2 volumes of saline and centrifuged at 1,250g for 5 minutes. The washed RBCs were MTX PHARMACOKINETICS counted to normalize MTXGlu1–5 concentrations to 8 ⫻ 1012 RBCs, so that results were comparable and not confounded by changes in the RBC count between patients and between visits. The concentrations of MTXGlun in RBCs were analyzed by HPLC with fluorescence detection, using a modification of a previously described method (26,27). All samples were analyzed in duplicate, and the mean concentration of each RBC MTXGlun concentration from each sample was used in the analysis. MTXGlu terminology. MTX, the parent drug, contains 1 glutamate moiety and will be referred to as MTXGlu1. The products of intracellular glutamation are referred to as MTXGlu2–5. The sum of concentrations of MTXGlu1 and its glutamated intracellular metabolites (RBC MTXGlu2–5) will be referred to as RBC MTXtotal. The MTXtotal concentration was not measured directly but was calculated as the sum of each RBC MTXGlun concentration, where n refers to the number of glutamate groups. Pharmacokinetics analysis. MTXGlun concentrations were graphed and analyzed using GraphPad Prism 4 software (San Diego, CA). A first-order exponential model was fitted to the data to calculate a half-life of accumulation, starting from the time of the final dose adjustment or from the start of therapy if there was no dose adjustment. Using the accumulation rate constant in the first-order exponential model, as calculated in Prism 4, the time for each RBC MTXGlun concentration to reach 90% of the modeled maximum concentration was calculated, and this was reported as the time to reach steady state. In patients stopping MTX, the time for each RBC MTXGlun concentration to become undetectable in RBCs for individual patients was reported directly from the raw data as the time when the concentration was below the limit of quantification (LOQ) for the assay. A median time for each RBC MTXGlun concentration to reach undetectable concentrations was then calculated. MTXGlun concentrations were fitted to a first-order exponential model in GraphPad Prism 4 to calculate the terminal half-life of elimination for each MTXGlun concentration. The model was applied using the data points from the time when a steady decline in the concentrations could be observed. Several patients had very low concentrations of MTXGlu4–5 from the time they ceased MTX treatment. To increase the ability to calculate the elimination half-life, raw values for MTXGlu1–5 were used, even when such values were below the LOQ. The half-lives of accumulation and elimination for MTXtotal were calculated using a first-order exponential method, as described for the individual MTXGlun concentrations. For each data-fitting analysis described above, the coefficient of determination (R2) was calculated using GraphPad Prism 4 software for the nonlinear regression aspect of the modeling. When the R2 value was ⬍0.7, the results were excluded from the values for calculation of the median half-lives of accumulation, because it was considered that these regression analysis results were too inaccurate to be meaningful. RESULTS Demographics. The demographic details for the 10 patients starting MTX treatment and the 10 patients 3301 Table 1. Baseline demographic characteristics of patients starting oral MTX and patients stopping oral MTX* Characteristic Female sex New Zealand European ethnicity Age, median (range) years Height, median (range) cm Weight, median (range) kg Estimated creatinine clearance, median (range) ml/minute Duration of RA, median (range) months Rheumatoid factor positive Radiographic erosions Rheumatoid nodules Starting MTX (n ⫽ 10) Stopping MTX (n ⫽ 10) 60 90 60 100 60 (42–72) 173.0 (158–182) 76.5 (50–110) 73.0 (54–84) 59 (43–68) 168.5 (158–186) 70.0 (46–88) 78.0 (57–93) 34.0 (2–216) 132.0 (3.5–480) 90 60 10 60 70 50 * Except where indicated otherwise, values are the percent. MTX ⫽ methotrexate; RA ⫽ rheumatoid arthritis. stopping MTX treatment are shown in Table 1. Of the 10 patients starting MTX, 4 were receiving another DMARD (2 were receiving sulfasalazine, and 2 were receiving sulfasalazine plus hydroxychloroquine). Eight patients were receiving a nonsteroidal antiinflammatory drug, and 5 were receiving prednisone. Of the 10 patients stopping MTX, 5 were receiving another DMARD (4 were receiving leflunomide, and 1 was receiving sulfasalazine). Eight patients were receiving a nonsteroidal antiinflammatory drug, and 3 were receiving prednisone. Time for RBC MTXGlun to become detectable in patients starting MTX. Patients commenced treatment with oral MTX at a median dosage of 10 mg/week (range 5–10 mg/week). The MTX dosage was titrated according to the clinical response to a median dosage of 15 mg/week (range 10–20 mg/week). The final dose of MTX was 10 mg in 2 patients, 12.5 mg in 2 patients, 15 mg in 3 patients, and 20 mg in 3 patients. The dosage of MTX was stable for a median of 28 weeks (range 18–32 weeks). Patients remained in the study for a median of 40 weeks (range 24–48 weeks). All patients received folic acid at a dosage of 5 mg/week, 3–4 days after the dose of MTX was administered. In 9 of 10 patients, MTXGlu1 was detectable in RBCs 1–2 weeks following administration of the first dose of MTX. MTXGlu2 was detectable in RBCs at a median of 2 weeks (range 1–4 weeks) after commencement of MTX treatment. In 9 patients, MTXGlu3 was detectable in RBCs at a median of 3 weeks (range 1–5 weeks) after commencing MTX. In 9 patients, concentrations of RBC MTXGlu4 were detectable after a median of 8 weeks (range 1–28 weeks). One patient had 3302 DALRYMPLE ET AL Figure 1. Concentration-time profiles for each red blood cell methotrexate polyglutamate moiety (RBC MTXGlun) and MTXtotal (the sum of concentrations of MTXGlu1 and its glutamated intracellular metabolites) for a patient with no dose changes. The line is fitted to a nonlinear exponential accumulation (first-order) model as calculated using GraphPad Prism 4 software. Broken lines indicate the limit of quantification of the assay. no detectable RBC MTXGlu5 at any time during the first 40 weeks of MTX therapy. In the remaining 9 patients, MTXGlu5 was detectable after a median of 7 weeks (range 1–28 weeks). Accumulation of RBC MTXGlun in patients starting MTX. Figure 1 shows a representative concentration-time profile for MTXGlun in a patient commencing MTX treatment who had no dose changes for the duration of the study. The dose of MTX was unchanged during the study in only 2 patients. The other 8 patients had at least 1 dose alteration during the study period. Figure 2 shows a representative concentrationtime profile for a patient commencing MTX in whom the dose was changed during the study period. Some of the concentration-time data fit the firstorder exponential model poorly, in which case GraphPad Prism 4 calculated an R2 value ⬍0.7 or the R2 value could not be calculated using a first-order exponential model. There were 10 such sets of concentration-time data among the total of 50. These results were then excluded from the calculation of the median times to reach steady state and half-lives of accumulation. The medians of the time to reach steady state for the remaining MTXGlun concentration data are presented in Table 2. The estimated median half-life of accumulation of RBC MTXtotal was 8.3 weeks (range 2.0–18.8 weeks). Following administration of a stable dose of MTX, the median time until 90% of the maximum steady-state concentration was reached was 27.5 weeks (range 6.6– 62.0 weeks). The mean rates of accumulation and final MTXGlun concentrations were no different in those patients receiving sulfasalazine compared with those who were not receiving sulfasalazine. Effect of age and renal function on accumulation of RBC MTXGlun. Increasing age and impaired renal function did not correlate significantly with higher final concentrations for individual RBC MTXGlun or MTXGlutotal in RBCs. This analysis was undertaken using concentrations corrected for the individual participant’s final dose and weight, to help eliminate confounding factors. Variability and proportions of RBC MTXGlu1–5 in patients starting MTX. At the last visit, the range of RBC MTXtotal concentrations in patients starting MTX treatment was ⬃4-fold (90.9–351.5 nmoles/8 ⫻ 1012 RBCs). At the final visit, mean concentrations of MTXGlu1 accounted for 25% of MTXtotal, with MTXGlu2 accounting for 21%, MTXGlu3 accounting for 37%, MTXGlu4 accounting for 11%, and MTXGlu5 account- MTX PHARMACOKINETICS 3303 Figure 2. Concentration-time profiles of RBC MTXGlun and MTXtotal for a patient who commenced MTX treatment at a dosage of 10 mg/week (week 0), with an increase in the dosage to 15 mg/week at week 8 and 20 mg/week at week 12 (dose changes are indicated by the dotted lines). The line is fitted to a nonlinear exponential accumulation (first-order) model as calculated using GraphPad Prism 4 software from the time at which the patient received the final stable dose. See Figure 1 for definitions. ing for 6%. These proportions are similar to those reported in other studies described in the literature (23,26). RBC MTXGlun elimination in patients stopping MTX treatment. At the time of discontinuing MTX, the median dosage was 12.5 mg/week (range 7.5–20 mg/ week). Patients had been receiving a stable dosage of MTX for a median of 17 months (range 1.5–120 months). MTX was discontinued for a variety of reasons, Table 2. Estimated accumulation half-life and time to reach steady state after the final stable dose of MTX* MTXGlu1 MTXGlu2 MTXGlu3 MTXGlu4 MTXGlu5 No. of patients Accumulation half-life, weeks 7† 10 10 6‡ 4§ 1.9 (0.0–4.2) 3.2 (2.1–23.4) 12.5 (6.0–20.2) 45.2 (4.9–252.1) 42.4 (4.7–80) Time to reach steady state, weeks 6.2 (0.0–13.9) 10.6 (7.0–77.2) 41.2 (19.8–66.7) 149 (16.2–831.6) 139.8 (15.5–264.0) * Values are the median (range). Steady state was defined as 90% of the maximum concentration. MTXGlu ⫽ methotrexate polyglutamate. † Three patients were excluded because R2 ⬍ 0.7. ‡ Four patients were excluded because R2 ⬍ 0.7. § Four patients were excluded because R2 ⬍ 0.7, 1 was excluded because MTXGlu5 was undetectable, and 1 was excluded because the data did not fit a first-order exponential model. including adverse effects, lack of efficacy, quiescent disease, and concurrent infection. A representative concentration-time profile from 1 patient stopping MTX is shown in Figure 3. In all 10 patients, MTXGlu 1 became undetectable before MTXGlu2–5 became undetectable. With the exception of MTXGlu1, there was generally a lag time before a steady decline was observed in the MTXGlun concentrations. This lag time generally increased as the number of glutamate moieties increased. Although there was variation in the pattern of MTXGlu2–5 loss from RBCs, MTXGlu3–5 persisted in circulating RBCs longer than MTXGlu1–2. Time for RBC MTXGlun to become undetectable and half-life of elimination. The median times until each RBC MTXGlun became undetectable in patients stopping MTX are shown in Table 3. The first time point used for modeling to calculate the half-life of elimination for each patient was taken as the time when a consistent decrease in the MTXGlun concentration was observed. This terminal elimination halflife was then calculated using the first-order exponential model in GraphPad Prism 4. Some of the data fit this model poorly. Where GraphPad Prism 4 calculated an R2 value ⬍0.7, these results were excluded 3304 DALRYMPLE ET AL Figure 3. RBC MTXGlun and MTXtotal concentration-time profiles for a participant stopping MTX. The line is fitted to a nonlinear elimination (first-order) model as calculated using GraphPad Prism 4 software. Broken lines indicate the limit of quantification. See Figure 1 for definitions. from the calculation of the median elimination halflife for the individual MTXGlun concentration, because it was considered that the fitted model results were too inaccurate for meaningful further analysis. There were 6 such sets of concentration-time data among a total of 50. The estimated median half-life of elimination of RBC MTXtotal was 3.1 weeks (range 0.94–4.1 weeks). Table 3. Time until each RBC MTXGlu concentration was undetectable, and elimination half-life* MTXGlu1 MTXGlu2 MTXGlu3 MTXGlu4 MTXGlu5 Time until undetectable, weeks No. of patients Elimination half-life, weeks No. of patients 4.5 (2–14) 5.5 (3–32) 10 (2–⬎21) 6 (2–⬎21) 4 (2–10) 10 10 10 9‡ 5¶ 1.2 2.3 4.3 2.7 2.1 9† 9† 10 6§ 4# * Values are the median (range). † One patient was excluded because R2 ⬍ 0.7. ‡ One patient was excluded because no red blood cell methotrexate polyglutamate 4 (RBC MTXGlu4) was detectable. § Three patients were excluded because R2 ⬍ 0.7, and 1 was excluded because no RBC MTXGlu4 was detectable. ¶ Five patients were excluded because no RBC MTXGlu5 was detectable. # One patient was excluded because R2 ⬍ 0.7, and 5 were excluded because no MTXGlu5 was detectable. The median time until RBC MTXtotal was undetectable was 15 weeks (range 3–⬎32 weeks) from the time MTX treatment was ceased. Effect of age and renal function on MTXGlun. Increasing age did not correlate significantly with higher initial concentrations of each MTXGlun or MTXtotal in RBCs. However, impaired renal function did correlate with higher initial concentrations of RBC MTXGlu2, RBC MTXGlu3, and RBC MTXGlutotal. This analysis was undertaken using concentrations corrected for the individual participant’s final dose and weight, to reduce confounding factors. Variability and proportions of RBC MTXGlu1–5 in patients stopping MTX treatment. The range of RBC MTXtotal concentrations at the first visit in patients stopping MTX was ⬃5-fold (48.5–242.3 nmoles/8 ⫻ 1012 RBCs). At the initial visit, the mean concentration of MTXGlu1 accounted for 26% of the MTXtotal, with MTXGlu2 accounting for 28%, MTXGlu3 accounting for 30%, MTXGlu4 accounting for 11%, and MTXGlu5 accounting for 5%. DISCUSSION Weekly oral administration of low-dose MTX is the mainstay of treatment for RA. Despite the wide- MTX PHARMACOKINETICS spread use of MTX, the understanding of its mechanism of action and pharmacokinetics remains limited. After ingestion, MTX is rapidly taken up into a variety of cells, including RBCs. Within the cell, MTXGlun bind to and inhibit several important enzymes, including dihydrofolate reductase (which leads to decreased DNA methylation), thymidylate synthase (which interferes with DNA synthesis), and 5-aminoimidazole-4-carboxamide ribonucleotide transformylase (which increases adenosine release into the circulation, ultimately inhibiting tumor necrosis factor ␣ and interleukin-1␤). Thus, MTX has several important antiinflammatory actions mediated through a variety of different pathways. The rapid intracellular uptake of MTX and short plasma half-life of the parent drug mean that plasma concentrations are unable to be used for therapeutic drug monitoring. However, RBCs are a readily accessible source for measuring intracellular MTXGlun concentrations. To date, there are limited data regarding the accumulation and elimination of MTXGlun. We have shown that after oral ingestion, the parent drug MTXGlu1 is detectable in circulating RBCs 1–2 weeks after administration of the first dose. The data presented herein suggest a progressive accumulation of MTXGlu2–5, starting with MTXGlu2 (which generally is present after 2 weeks), and subsequently MTXGlu3 (which is present after 3 weeks), MTXGlu4 (present after a median of 8 weeks), and finally MTXGlu5 (present after a median of 7 weeks). This sequential accumulation occurs at least in part because RBC MTXGlu2–5 are produced intracellularly and therefore require MTXGlu1 to be present before MTXGlu2 and the other polyglutamates can be produced. Our observation that production of MTXGlu2–5 is delayed and progressive may be explained partly by the fact that glutamation of MTXGlun is a slow reaction. Although the longer-chain polyglutamates (MTXGlu3–5) were detectable within 3–8 weeks, the period of time until steady state was achieved was significantly longer. From a clinical perspective, a response to MTX usually occurs within the first 6–12 weeks after initiation of therapy, with a maximal response observed at 6 months (28,29). This suggests that the longer-chain polyglutamates, which take longer to become detectable and reach steady state, may have a more important role in the clinical response compared with the shorter-chain polyglutamates. These data beg the question as to whether a more rapid attainment of steady state MTXGlun concentrations and a clinical response could be achieved with alternative dosing strategies. Such a strategy might include more rapid dose escalation or starting therapy with 3305 higher doses. However, high doses have lower associated oral availability (30). In addition, starting with higher oral doses may result in discontinuation due to adverse events such as nausea. Alternatively, MTX can be administered by the subcutaneous or intramuscular route, resulting in improved bioavailability (30). In a recent study, subcutaneous administration of MTX was associated with a significantly improved response at 24 weeks as compared with oral administration (31). It would be of interest to know whether the observed improved clinical response seen with subcutaneous administration occurred at an earlier time point compared with oral administration. The proportion of individual RBC MTXGlu concentrations at steady state in this study is similar to that in other studies, with MTXGlu3 accounting for the largest portion (30% of MTXGlutotal) and MTXGlu5 accounting for the smallest proportion (⬃5% of MTXGlutotal) (23,26,32). Interestingly, some patients starting MTX had no detectable RBC MTXGlu5 at any time during the study. Further larger studies are required to determine whether this lack of MTXGlu5 has any effect on the clinical response. Upon discontinuation of oral MTX, the loss of MTXGlu1 from RBCs is relatively rapid. In contrast, there is a lag time before initiation of the loss of the other MTXGlun (MTXGlu2–5). This is similar to the pattern of elimination observed in children with acute lymphoblastic leukemia who discontinued oral MTX (16.8–21.4 mg/m2/week) after 3 years of therapy (23). This delay before loss of MTXGlu2–5 may have 2 contributing factors. First, each MTXGlun may be present in immature RBCs (reticulocytes) that continue to enter the circulation after MTX dosing has ceased. Second, progressive metabolism from the parent drug to MTXGlu2, then MTXGlu3, then MTXGlu4, and finally MTXGlu5 may continue intracellularly even after MTX dosing has ceased. This predisposes to earlier loss of shorter-chain MTXGlu1–2 in comparison with MTXGlu3–5. It has been suggested that once MTX enters the RBC, a proportion does not efflux and persists for the life of the RBC. The human RBC has a lifespan in adults of ⬃110–120 days. Two studies of RBC MTXGlu1–5 concentrations suggest that the portion that persists in the RBC consists of MTXGlu3–5 (32,33). The initial elimination appears predominantly due to RBC loss of MTXGlu1 and MTXGlu2. This is followed by a slower elimination of MTXGlu3–5, suggesting that this phase in elimination is determined by the lifespan of RBCs plus the dilutional effect of new RBCs that do not contain MTX being released into the circulation (32,33). 3306 Schroder suggests that the concentrations of MTXGlu3–5 did not change in more mature RBCs, because the activity of intraerythrocytic enzymes declines as the cells age. However, there is no direct evidence showing that mature RBCs lack the enzyme FPGS or ␥-glutamyl hydrolase. RBC MTXGlu3 was often the last MTXGlun to be eliminated from RBCs. This may simply be because it starts at the highest concentration compared with the other MTXGlun rather than having different elimination kinetics. It may also be attributable to the rate of conversion from RBC MTXGlu3 to the longer-chain polyglutamates being slower than the rate of transformation from RBC MTXGlu1 to MTXGlu2 and subsequently to MTXGlu3. When MTXGlu1, MTXGlu2, and MTXGlu3 are compared, the estimated half-life of elimination of each increased with the greater number of glutamate residues. However, this pattern did not extend to MTXGlu4 and MTXGlu5. The sensitivity of the assay may have been a limiting factor in calculating an accurate elimination half-life for MTXGlu4 and MTXGlu5, which were not present in high concentrations in the majority of patients. Drugs that accumulate and are eliminated by simple first-order pharmacokinetic mechanisms usually accumulate and are eliminated at approximately the same rate. The data from this study suggest that the median rate of elimination of each RBC MTXGlun is more rapid than the median rate of accumulation. This may be related to the RBC lifespan. The gradual loss of RBCs from the circulation, combined with the ongoing production of new RBCs, provides a “dilutional effect” when MTX dosing is ceased. The RBC lifespan may also contribute to the delay in achieving steady state, although this is less clear from the data presented here. If it were only “young” RBCs that have intact mechanisms for metabolizing MTXGlu1 to MTXGlu2–5, this might explain the delay to steady state while waiting for the generations of young cells to move through to maturity. Previous studies have demonstrated an approximate half-life of elimination of RBC MTXtotal of 2 weeks to 11.3 weeks (23,32,34). In comparison, the half-life of elimination in our population ranged from 0.9 week to 13.3 weeks. The time for RBC MTXtotal concentrations to become undetectable has been reported to be within 15 weeks of stopping MTX (23,35). In comparison, we have shown that RBC MTXtotal concentrations became undetectable after anywhere from 3 weeks to more than 32 weeks. At least some of this difference might be attributable to differences in the DALRYMPLE ET AL populations studied. The previous studies involved either children, in whom one can expect pharmacokinetic differences due to age, or adults with solid tumors, in whom much higher dosing regimens are utilized. The long elimination period appears to be inconsistent with the period of time until a disease flare occurs after stopping MTX (frequently within 1 month) (36). The long duration of elimination has implications for discontinuation of MTX prior to surgery, conception, and in the face of concomitant infection. Although newly produced cells will not be exposed to MTX, existing cells will clearly continue to be affected by MTX for a considerable length of time. There was significant interpatient variability in the measured concentrations of RBC MTXGlun. This variation was not explained by age. A correlation between renal impairment and an increased RBC MTXGlun concentration is suggested, but further study in a larger population is required to substantiate this finding. There are some limitations of our study design, including the small sample size. The low number of participants means that these results cannot be extrapolated to all patients taking MTX, particularly given the variability displayed. The results also suggest that participants stopping MTX should have been included only after receiving a stable dosage of MTX for at least 6 months. Three of the participants included in this study had been receiving their current MTX dosage for fewer than 6 months. Furthermore, 2 patients had been receiving MTX therapy for fewer than 6 months in total. This limited our ability to calculate the rate of elimination and observe the time to undetectable concentrations of each RBC MTXGlun in some patients, because it is unlikely that concentrations were at steady state before ceasing MTX. The planned analysis was based on the expectation that first-order pharmacokinetics were likely to explain the accumulation and elimination of RBC MTXGlun. This appears reasonable for MTXGlu1. However, the concentration-time profiles for MTXGlu2–5 accumulation and elimination suggest that a more complex analysis is required. Although we acknowledge the limitations of this analysis, these data provide a substantial increase in our understanding of RBC MTXGlun accumulation and elimination. However, the variability in time until RBC MTXGlu3–5 concentrations reach steady state after initiation and dose adjustment may limit the clinical usefulness of RBC MTXGlu1–5 as a therapeutic monitoring tool in clinical practice. If RBC MTXGlun concentrations correlate with MTX efficacy, consideration of subcutaneous administration from the onset of treat- MTX PHARMACOKINETICS ment or rapid oral dose escalation to reduce the length of time to steady state may be required. 3307 11. ACKNOWLEDGMENTS We gratefully acknowledge the assistance of Jill James, Rheumatology Research Nurse, and Jan Ipenburg, Rheumatology Clinical Nurse Specialist, in assisting with patient data collection. AUTHOR CONTRIBUTIONS Dr. Stamp had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Dalrymple, Stamp, O’Donnell, Chapman, Barclay. Acquisition of data. Dalrymple, Stamp, O’Donnell. Analysis and interpretation of data. Dalrymple, Stamp, Chapman, Zhang, Barclay. Manuscript preparation. Dalrymple, Stamp, O’Donnell, Chapman, Zhang, Barclay. Statistical analysis. Dalrymple, Barclay. Development and validation of HPLC assay. Zhang. 12. 13. 14. 15. 16. 17. 18. REFERENCES 1. Luqmani R, Hennell S, Estrach C, Birrell F, Bosworth A, Davenport G, et al, on behalf of the British Society for Rheumatology and British Health Professionals in Rheumatology, Guidelines and Audit Working Group. British Society for Rheumatology and British Health Professionals in Rheumatology guideline for the management of rheumatoid arthritis (the first two years). Rheumatology (Oxford) 2006;45:1167–9. 2. American College of Rheumatology Subcommittee on Rheumatoid Arthritis Guidelines. Guidelines for the management of rheumatoid arthritis: 2002 update. Arthritis Rheum 2002;46: 328–46. 3. Pincus T, Sokka T. Should aggressive therapy for rheumatoid arthritis require early use of weekly low-dose methotrexate, as the first disease-modifying anti-rheumatic drug in most patients? [editorial]. Rheumatology (Oxford) 2006;45:497–9. 4. Pincus T, Yazici Y, Sokka T, Aletaha D, Smolen J. Methotrexate as the “anchor drug” for the treatment of rheumatoid arthritis. Clin Exp Rheumatol 2003;21(5 Suppl 31):179–85. 5. Plant M, Jones P, Saklatvala J, Ollier W, Dawes P. Patterns of radiological progression in early rheumatoid arthritis: results of an 8 year prospective study. J Rheumatol 1998;25:417–26. 6. Van der Heijde D. Joint erosions and patients with early rheumatoid arthritis. Br J Rheumatol 1995;34 Suppl 2:74–8. 7. Wolfe F, Sharp JT. Radiographic outcome of recent-onset rheumatoid arthritis: a 19-year study of radiographic progression. Arthritis Rheum 1998;41:1571–82. 8. Graudal N, Tarp U, Jurik A, Galloe A, Garred P, Milman N, et al. Inflammatory patterns in rheumatoid arthritis estimated by the number of swollen and tender joints, the erythrocyte sedimentation rate and hemoglobin: long-term course and association to radiographic progression. J Rheumatol 2000;27:47–57. 9. Wolfe F, Zwillich SH. The long-term outcomes of rheumatoid arthritis: a 23-year prospective, longitudinal study of total joint replacement and its predictors in 1,600 patients with rheumatoid arthritis. Arthritis Rheum 1998;41:1072–82. 10. Grigor C, Capell H, Stirling A, McMahon A, Lock P, Vallance R, et al. Effect of a treatment strategy of tight control for rheumatoid 19. 20. 21. 22. 23. 24. 25. 26. 27. arthritis (the TICORA study): a single blind randomised controlled trial. Lancet 2004;364:263–9. Sharp JT, Strand V, Leung H, Hurley F, Loew-Friedrich I, on behalf of the Leflunomide Rheumatoid Arthritis Investigators Group. Treatment with leflunomide slows radiographic progression of rheumatoid arthritis: results from three randomized controlled trials of leflunomide in patients with active rheumatoid arthritis. Arthritis Rheum 2000;43:495–505. Pettit AR, Weedon H, Ahern M, Zehntner S, Frazer IH, Slavotinek J, et al. Association of clinical, radiological and synovial immunopathological responses to anti-rheumatic treatment in rheumatoid arthritis. Rheumatology (Oxford) 2001;40:1243–55. Tishler M, Caspi D, Graff E, Segal R, Peretz H, Yaron M. Synovial and serum levels of methotrexate during methotrexate therapy of rheumatoid arthritis. Br J Rheumatol 1989;28:422–3. Kremer JM. Toward a better understanding of methotrexate [review]. Arthritis Rheum 2004;50:1370–82. Cutolo M, Sulli A, Pizzorni C, Seriolo B, Straub R. Anti-inflammatory mechanisms of methotrexate in rheumatoid arthritis. Ann Rheum Dis 2001;60:729–35. Furst D, Koehnke R, Burmeister L, Kohler J, Cargill I. Increasing methotrexate effect with increasing dose in the treatment of resistant rheumatoid arthritis. J Rheumatol 1989;16:313–20. Seidman P. Preliminary report: methotrexate: the relationship between dose and clinical effect. Br J Rheumatol 1993;32:751–3. Dervieux T, Furst D, Lein DO, Capps R, Smith K, Walsh M, et al. Polyglutamation of methotrexate with common polymorphisms in reduced folate carrier, aminoimidazole carboxamide ribonucleotide transformylase, and thymidylate synthase are associated with methotrexate effects in rheumatoid arthritis. Arthritis Rheum 2004;50:2766–74. Lafforgue P, Monjanel-Mouterde S, Durand A, Catalin J, Acquaviva P. Lack of correlation between pharmacokinetics and efficacy of low dose methotrexate in patients with rheumatoid arthritis. J Rheumatol 1995;22:844–9. Angelis-Stoforidis P, Vajda F, Christophidis N. Methotrexate polyglutamate levels in circulating erythrocytes and polymorphs correlate with clinical efficacy in rheumatoid arthritis. Clin Exp Rheumatol 1999;17:313–20. Dervieux T, Furst D, Lein DO, Capps R, Smith K, Caldwell J, et al. Pharmacogenetic and metabolite measurements are associated with clinical status in rheumatoid arthritis patients treated with methotrexate: results of a multicentred cross sectional observational study. Ann Rheum Dis 2005;64:1180–5. Kremer J, Lein DO, Meyer G, Barham R, Simpson L, Do J, et al. Measurement of erythrocyte methotrexate polyglutamates predicts response to methotrexate therapy in a dose escalation study with rheumatoid arthritis patients [abstract]. Arthritis Rheum 2004;50 Suppl 9:S181–2. Schroder H, Fogh K. Methotrexate and its polyglutamate derivatives in erythrocytes during and after weekly low-dose oral methotrexate therapy of children with acute lymphoblastic leukaemia. Cancer Chemother Pharmacol 1988;21:145–9. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. Levy AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D, for the Modification of Diet in Renal Disease Study Group. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;130:461–70. Dervieux T, Lein D, Marcelletti J, Pischel K, Smith K, Walsh M, et al. HPLC determination of erythrocyte methotrexate polyglutamates after low-dose methotrexate therapy in patients with rheumatoid arthritis. Clin Chem 2003;49:1632–41. Brooks A, Begg E, Zhang M, Frampton C, Barclay M. Red blood 3308 28. 29. 30. 31. cell methotrexate polyglutamate concentrations in inflammatory bowel disease. Ther Drug Monit 2007;29:619–25. Kremer JM, Lee JK. The safety and efficacy of the use of methotrexate in long-term therapy for rheumatoid arthritis. Arthritis Rheum 1986;29:822–31. Weinblatt ME, Trentham DE, Fraser PA, Holdsworth DE, Falchuk KR, Weissman BN, et al. Long-term prospective trial of low-dose methotrexate in rheumatoid arthritis. Arthritis Rheum 1988;31:167–75. Hoekstra M, Haagsma C, Neeg C, Proost J, Knuif A, van de Laar M. Splitting high dose oral methotrexate improves bioavailability: a pharmacokinetic study in patients with rheumatoid arthritis. J Rheumatol 2006;33:481–5. Braun J, Kastner P, Flaxenberg P, Wahrisch J, Hanke P, Demary W, et al, for the MC-MTX.6/RH Study Group. Comparison of the clinical efficacy and safety of subcutaneous versus oral administration of methotrexate in patients with active rheumatoid arthri- DALRYMPLE ET AL 32. 33. 34. 35. 36. tis: results of a six-month, multicenter, randomized, double-blind, controlled, phase IV trial. Arthritis Rheum 2008;58:73–81. Schroder H, Fogh K, Herlin T. In vivo decline of methotrexate and methotrexate polyglutamates in age-fractionated erythrocytes. Cancer Chemother Pharmacol 1988;21:150–5. Schroder H. Methotrexate pharmacokinetics in age-fractionated erythrocytes. Cancer Chemother Pharmacol 1986;18:203–7. Schalhorn A, Sauer H, Wilmanns W, Stupp-Poutot G. Pharmacokinetics of erythrocyte methotrexate after high-dose methotrexate. Cancer Chemother Pharmacol 1982;9:65–9. Schroder H, Clausen N, Ostergaard E, Pressler T. Pharmacokinetics of erythrocyte methotrexate in children with acute lymphoblastic leukaemia during maintenance treatment. Cancer Chemother Pharmacol 1986;16:190–3. Kremer J, Rynes R, Bartholomew L. Severe flare of rheumatoid arthritis after discontinuation of long-term methotrexate therapy. Am J Med 1987;82:781–6.