Determinants of red blood cell methotrexate polyglutamate concentrations in rheumatoid arthritis patients receiving long-term methotrexate treatment.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 60, No. 8, August 2009, pp 2248–2256 DOI 10.1002/art.24653 © 2009, American College of Rheumatology Determinants of Red Blood Cell Methotrexate Polyglutamate Concentrations in Rheumatoid Arthritis Patients Receiving Long-Term Methotrexate Treatment Lisa K. Stamp,1 John L. O’Donnell,2 Peter T. Chapman,3 Mei Zhang,3 Christopher Frampton,1 Jill James,3 and Murray L. Barclay1 Objective. Methotrexate (MTX) is the most commonly used disease-modifying antirheumatic drug (DMARD) in the management of rheumatoid arthritis (RA). MTX is transported into cells, where additional glutamate moieties are added and it is retained as MTX polyglutamates (MTXGlu [referred to as a group as MTXGlun]). There is large interpatient variability in MTXGlun concentrations. This study was undertaken to determine nongenetic factors that influence red blood cell (RBC) MTXGlun concentrations in patients receiving long-term stable low-dose oral MTX. Methods. One hundred ninety-two patients receiving long-term oral MTX for the treatment of RA were recruited. Trough MTXGlun concentrations were measured by high-performance liquid chromatography. Univariate analysis was performed to determine variables influencing MTXGlun concentrations. Backward stepwise multivariate regression analysis was done to determine variables that affect individual MTXGlun concentrations; variables with P values of <0.1 in the univariate analysis for any MTXGlun were included. Results. Univariate analysis revealed that increased age, lower estimated glomerular filtration rate (GFR), higher MTX dosage, longer duration of MTX treatment, and use of prednisone were associated with significantly higher MTXGlun concentrations. Smokers had significantly lower concentrations of MTXGlu3, MTXGlu3–5, and MTXGlu1–5. Sex, rheumatoid factor and anti–cyclic citrullinated peptide status, RBC folate level, and body mass index had no significant effect on MTXGlun levels. Concomitant use of other DMARDs was associated with lower MTXGlu2 levels, and treatment with nonsteroidal antiinflammatory drugs was associated with lower MTXGlu3 and MTXGlu1–5 concentrations. Multivariate regression analysis revealed that age, MTX dosage, and estimated GFR were the major determinants of MTXGlun concentrations. Conclusion. Large interpatient variability in MTXGlun concentrations can be explained, at least in part, by a combination of factors, particularly age, MTX dosage, and renal function. There are complex interactions between smoking, RBC folate levels, and concentrations of MTXGlun. Methotrexate (MTX) is one of the most commonly used drugs in the management of rheumatic diseases. It is the anchor drug in the treatment of rheumatoid arthritis (RA) and is the basis of most disease-modifying antirheumatic drug (DMARD) combination therapies used in RA. MTX appears to act as a folate antagonist, although its exact mechanism of action remains unclear despite its widespread use (1). Serum MTX concentrations fall rapidly following intravenous administration (2). However, MTX is transported into red blood cells (RBCs), white blood cells, hepatocytes, and synoviocytes, mainly via the reduced folate carrier (RFC). Once inside the cell, MTX has glutamate groups added by folylpolyglutamate synthetase (FPGS) and is retained as MTX polyglutamates (MTXGlu) in the cell, where it exerts its actions. Gamma glutamyl hydrolase ANZCTR: 012606000275561. Supported by the Health Research Council of New Zealand and Arthritis New Zealand. 1 Lisa K. Stamp, FRACP, PhD, Christopher Frampton, PhD, Murray L. Barclay, MBChB, FRACP, MD: University of Otago, Christchurch, New Zealand; 2John L. O’Donnell, MBChB, FRACP, FRCPA: Canterbury Health Laboratories, Christchurch, New Zealand; 3Peter T. Chapman, FRACP, MD, Mei Zhang, PhD, Jill James, BN: Christchurch Hospital, Christchurch, New Zealand. Address correspondence and reprint requests to Lisa K. Stamp, FRACP, PhD, Department of Medicine, University of Otago, PO Box 4345, Christchurch 8140, New Zealand. E-mail: Lisa.firstname.lastname@example.org. Submitted for publication December 16, 2008; accepted in revised form April 7, 2009. 2248 DETERMINANTS OF MTX POLYGLUTAMATES removes terminal MTX polyglutamates, returning MTX to its monoglutamate form, which is rapidly transported out of the cell by multidrug resistance–associated proteins. Thus, intracellular MTXGlu concentrations are related to the balance of activity between these two enzymes. MTXGlu concentrations have been reported to correlate with clinical response, but not toxicity, in RA (3–5). Concentrations of either the longer-chain polyglutamates (MTXGlu3 or MTXGlu3–5), which are more stable, or total polyglutamate (MTXGlu1–5) were measured in those studies. These findings need to be confirmed in other cohorts, and factors that influence MTXGlu concentrations are an important consideration in interpreting the findings. We have previously shown that there is sequential accumulation of the polyglutamates, starting with MTXGlu1 and MTXGlu2, in RA patients who have begun MTX therapy (6). Given that uncertainty remains about which is the best MTXGlu measure for clinical monitoring, we elected to examine the individual polyglutamates as well as MTXGlu1–5 and MTXGlu3–5. The longer-chain MTX polyglutamates have much greater potency of inhibition of key enzymes in the folate pathway than the shorter-chain MTX polyglutamates (7,8). We and others have reported wide interpatient variability in MTXGlu1–5 concentrations in patients receiving low-dose oral MTX for RA (4,6). Most attention has focused on genetic polymorphisms within the folate pathway to explain such variation, while less attention has been paid to other, nongenetic factors that may affect RBC MTXGlu concentrations. Physicians are generally aware of common factors, such as age and renal function, that may affect drug concentration, although there are no current data on their effect on RBC MTXGlu levels. Interactions with drugs commonly coadministered with MTX, such as nonsteroidal antiinflammatory drugs (NSAIDs), corticosteroids, and other DMARDs, may influence the efficacy and adverse effects of MTX, but their effects on RBC MTXGlu levels have not been reported previously. Knowledge of the factors that influence RBC MTXGlu concentrations may have implications for dosing strategies and for interpretation of the results of studies examining the relationship between MTXGlu levels and efficacy/toxicity in RA. The aim of this study was to determine which nongenetic factors determine RBC MTXGlu concentrations in patients receiving long-term stable low-dose oral MTX. 2249 PATIENTS AND METHODS Ethics approval was obtained from the Upper South B Regional Ethics Committee, New Zealand. Written informed consent was obtained from each patient. Patients and study protocol. This was a cross-sectional study undertaken at a single center in Christchurch, New Zealand. Patients ⱖ18 years of age with RA that fulfilled the criteria of the American Rheumatism Association (9) were recruited. Patients were required to have been taking oral MTX for at least 3 months, with a stable dosage for at least 1 month prior to study entry. A preference was given to patients receiving MTX monotherapy. Concomitant therapy with NSAIDs and prednisone was allowed. A change in MTX dosage or introduction of another DMARD or oral prednisone within the previous month precluded enrollment in the study. Standard demographic data were collected. Estimated creatinine clearance (glomerular filtration rate [GFR]) was calculated from serum creatinine via the Modification of Diet in Renal Disease (MDRD) study equation (10). MTXGlu terminology and measurement. MTX, the parent drug, contains 1 glutamate moiety and is referred to below as MTXGlu1. MTXGlu1 and the products of intracellular glutamation (MTXGlu2, MTXGlu3, MTXGlu4, or MTXGlu5) are collectively referred to as MTX polyglutamates (MTXGlun). MTXGlu3–5 and MTXGlu1–5 were not measured directly, but calculated as the sum of each RBC MTXGlun, where n refers to the number of glutamate groups. Trough RBC MTXGlun concentrations were measured by high-performance liquid chromatography as previously described (6). Results were normalized to an RBC count of 8 ⫻ 1012 cells, so that results were comparable and were not confounded by changes in RBC counts between individuals. All samples were analyzed in duplicate, and the mean concentration of each RBC MTXGlun from each sample was used. Statistical analysis. Univariate associations between the MTXGlun concentrations and demographic and clinical features were analyzed using independent t-tests for binary variables and Pearson’s correlation coefficient for continuous variables. Variables included in this analysis were age, sex, smoking, duration of RA, rheumatoid nodules, radiographic erosions, MTX dosage, duration of MTX therapy, use of other DMARDs, use of NSAIDs, use of prednisone, body mass index, rheumatoid factor (RF) status, anti–cyclic citrullinated peptide (anti-CCP) status, estimated GFR, and RBC folate level. Variables showing some association (P ⬍ 0.1) with any individual MTXGlun concentration from these analyses were entered into backward stepwise linear regression analyses; binary independent variables were entered into these models as 0 or 1. For the final multivariate models, 2-tailed P values less than 0.05 were considered significant. To confirm that slight non-normality in the concentrations did not influence the univariate results, analyses were repeated using the nonparametric Mann-Whitney U test and Spearman’s correlation coefficient. These analyses yielded exactly the same levels of significance (P ⬍ 0.1 or not) for all associations, and hence, yielded the same variables that were subsequently assessed in the multivariate linear regression analyses. 2250 STAMP ET AL RESULTS Demographic characteristics of the patients. Two hundred patients were recruited between October 2005 and February 2008. Eight patients were excluded because of incomplete assay data. Of the 192 patients included in the final analysis, 72.9% were female, the mean age was 60.5 years (range 18–84), and the mean duration of RA was 10.5 years (range 0.25–53). Sixtythree percent of the patients had radiographic erosions, 25.5% had rheumatoid nodules, 80.7% were RF positive, and 76.4% were anti-CCP positive. The median weekly dosage of MTX was 15.6 mg (range 5–25), and all but 1 patient received folic acid 5 mg/week (taken 3–4 days after MTX was taken). Patients had been taking MTX for a median of 4.5 years (range 0.25–19) prior to study entry, and had been receiving MTX at the studyentry dosage for a median of 12 months (range 1–240). The mean estimated GFR was 76.4 ml/minute/1.72m2 (range 37–118). RBC MTXGlun profile. MTXGlu3 was found to be the predominant polyglutamate, accounting for 35.8% of the total RBC MTXGlun. MTXGlu1 accounted for 21.7%, MTXGlu2 for 20.5%, MTXGlu4 for 14.5%, and MTXGlu5 for 7.5%. MTXGlu3 levels were strongly correlated with MTXGlu3–5 levels (r ⫽ 0.93, P ⬍ 0.0001). There was wide between-patient variability in levels of each MTXGlun (Table 1). Strong association of MTX dosage with higher concentrations of long-chain RBC MTXGlun. There was a trend toward correlation between MTX dosage and MTXGlu1 levels (r ⫽ ⫺0.12, P ⫽ 0.09), but MTX dosage was not correlated with MTXGlu2 concentrations (r ⫽ 0.15, P ⫽ 0.83). However, MTX dosage was significantly correlated with the long-chain polyglutamates RBC MTXGlu3 (r ⫽ 0.45, P ⬍ 0.0001), MTXGlu4 (r ⫽ 0.47, P ⬍ 0.0001), MTXGlu5 (r ⫽ 0.32, P ⬍ 0.0001), MTXGlu1–5 (r ⫽ 0.33, P ⬍ 0.0001), and MTXGlu3–5 (r ⫽ 0.47, P ⬍ 0.0001) (Figure 1). Longer duration of MTX therapy was associated with significantly higher Table 1. RBC MTXGlun concentrations in the 192 patients with RA* MTXGlu1 MTXGlu2 MTXGlu3 MTXGlu4 MTXGlu5 30.32 ⫾ 15.43 (5.8–86.9) 28.69 ⫾ 11.24 (6.7–76.2) 49.91 ⫾ 24.04 (3.8–139.5) 20.02 ⫾ 18.01 (0–113.6) 10.52 ⫾ 10.09 (0–60.9) * Values are the mean ⫾ SD (range) nmoles/8 ⫻ 1012 red blood cells (RBCs). MTXGlun ⫽ methotrexate polyglutamate; RA ⫽ rheumatoid arthritis. Figure 1. Mean red blood cell (RBC) concentration of the indicated methotrexate polyglutamates (MTXGlu), by weekly dosage of MTX. concentrations of RBC MTXGlu 4 , MTXGlu 5 , MTXGlu3–5, and MTXGlu1–5 (Table 2). Effect of patient variables on RBC MTXGlun concentrations. As expected, increasing age and lower estimated GFR were associated with higher RBC MTXGlun concentrations. There was a trend toward higher RBC MTXGlu1 concentrations in female patients as compared with male patients, but this did not reach statistical significance (mean ⫾ SD 31.5 ⫾ 16.2 nmoles/ 8 ⫻ 1012 RBCs versus 27.0 ⫾ 12.7 nmoles/8 ⫻ 1012 RBCs; P ⫽ 0.07) (Table 3). Body mass index had no DETERMINANTS OF MTX POLYGLUTAMATES 2251 Table 2. Univariate analysis of factors potentially associated with RBC MTXGlun concentrations* Variable MTXGlu1 MTXGlu2 MTXGlu3 MTXGlu4 MTXGlu5 MTXGlu3–5 MTXGlu1–5 MTX dosage Duration of MTX treatment Age Estimated GFR BMI RA duration RBC folate level RBC folate level adjusted for smoking ⫺0.12 0.15 0.21‡ ⫺0.32† ⫺0.02 0.14 ⫺0.13 0.14§ 0.02 0.02 0.13 ⫺0.15§ ⫺0.11 0.04 ⫺0.01 0.03 0.45† 0.14 0.28† ⫺0.19‡ ⫺0.12 0.09 ⫺0.07 0.04 0.47† 0.19‡ 0.32† ⫺0.21‡ ⫺0.05 0.12 0.09 0.07 0.32† 0.15§ 0.22‡ ⫺0.14§ ⫺0.10 0.11 0.12 0.10 0.47† 0.18§ 0.32† ⫺0.21‡ ⫺0.10 0.13 0.09 0.06 0.33† 0.21‡ 0.32† ⫺0.31† ⫺0.11 0.16§ ⫺0.04 0.01 * Data are presented as r values. GFR ⫽ glomerular filtration rate; BMI ⫽ body mass index (see Table 1 for other definitions). † P ⬍ 0.0001. ‡ P ⬍ 0.01. § P ⬍ 0.05. significant effect on the concentration of any RBC MTXGlun (Table 2). Effect of disease variables on RBC MTXGlun concentrations. Duration of disease, RF status, CCP status, presence of nodules, and presence of radiographic erosions had no significant effect on MTXGlun concentrations. Effect of NSAIDs, other DMARDs, and steroids on RBC MTXGlun concentrations. Eighty-six of the 192 patients (44.8%) were receiving NSAIDs. Twenty (10.4%) were receiving another DMARD (4 sulfasalazine, 9 hydroxychloroquine, 5 sulfasalazine and hydroxychloroquine, 1 leflunomide, and 1 leflunomide and hydroxychloroquine). Fifty-nine patients (30.7%) were receiving oral prednisone, at a mean daily dosage of 5.8 mg (range 1–20). Concomitant use of prednisone was associated with significantly higher RBC concentrations of MTXGlu 2 , MTXGlu 3 , MTXGlu 4 , MTXGlu 1–5 , and MTXGlu3–5. While there was a trend toward lower RBC MTXGlun concentrations with concomitant use of any other DMARD, this reached statistical significance for MTXGlu2 only. Concomitant use of NSAIDs was associated with lower MTXGlu3 and MTXGlu1–5 concentrations (Table 3). Association of smoking with lower RBC MTXGlu3 concentrations. There were 157 current nonsmokers and 35 current smokers in the study population. RBC concentrations of MTXGlu 3 , MTXGlu 3–5 , and MTXGlu1–5 were all significantly lower in smokers compared with nonsmokers, with no significant difference in Table 3. Effect of prednisone, other DMARDs, NSAIDs, sex, and smoking on RBC MTXGlun concentrations* Prednisone Yes (n ⫽ 59) No (n ⫽ 133) P Other DMARDs Yes (n ⫽ 20) No (n ⫽ 172) P NSAIDs Yes (n ⫽ 86) No (n ⫽ 106) P Female Yes (n ⫽ 140) No (n ⫽ 52) P Smoker Yes (n ⫽ 157) No (n ⫽ 35) P MTXGlu1 MTXGlu2 MTXGlu3 MTXGlu4 MTXGlu5 MTXGlu3–5 MTXGlu1–5 32.3 ⫾ 16.9 29.4 ⫾ 15.0 0.24 31.3 ⫾ 11.7 27.5 ⫾ 10.8 0.03 57.4 ⫾ 23.7 46.6 ⫾ 23.5 0.004 24.4 ⫾ 20.8 18.1 ⫾ 16.3 0.023 10.4 ⫾ 9.9 10.6 ⫾ 10.1 0.93 92.2 ⫾ 46.4 752 ⫾ 45.9 0.02 155.8 ⫾ 55.2 132.2 ⫾ 58.2 0.009 23.9 ⫾ 9.9 31.1 ⫾ 15.8 0.052 23.8 ⫾ 8.8 29.2 ⫾ 11.4 0.042 41.1 ⫾ 23.4 50.9 ⫾ 23.9 0.084 17.9 ⫾ 16.3 20.3 ⫾ 18.2 0.58 12.6 ⫾ 11.8 10.3 ⫾ 9.9 0.34 71.3 ⫾ 43.9 81.5 ⫾ 46.9 0.36 119.2 ⫾ 48.3 141.8 ⫾ 58.9 0.1 29.5 ⫾ 15.3 31.0 ⫾ 15.6 0.49 27.5 ⫾ 10.8 29.6 ⫾ 11.6 0.19 45.9 ⫾ 20.6 53.1 ⫾ 26.1 0.042 17.3 ⫾ 13.7 22.2 ⫾ 10.7 0.059 9.9 ⫾ 9.4 10.9 ⫾ 10.6 0.48 73.3 ⫾ 51.9 86.3 ⫾ 51.9 0.054 130.2 ⫾ 50.9 146.9 ⫾ 62.7 0.048 31.5 ⫾ 16.2 27.0 ⫾ 12.7 0.07 29.3 ⫾ 11.5 27.1 ⫾ 10.5 0.23 50.6 ⫾ 24.6 48.2 ⫾ 22.6 0.54 20.1 ⫾ 18.4 19.9 ⫾ 16.9 0.95 10.6 ⫾ 10.5 10.2 ⫾ 9.1 0.80 81.2 ⫾ 47.5 78.2 ⫾ 44.4 0.69 142.1 ⫾ 59.2 132.4 ⫾ 55.4 0.31 27.8 ⫾ 13.9 30.9 ⫾ 15.7 0.28 25.4 ⫾ 9.8 29.4 ⫾ 11.4 0.058 41.4 ⫾ 20.8 51.8 ⫾ 24.4 0.02 14.7 ⫾ 10.8 21.2 ⫾ 19.1 0.054 8.9 ⫾ 9.9 10.8 ⫾ 10.1 0.31 65.1 ⫾ 35.9 83.9 ⫾ 48.1 0.03 118.3 ⫾ 41.7 144.2 ⫾ 60.4 0.017 * Values are the mean ⫾ SD nmoles/8 ⫻ 1012 RBCs. DMARDs ⫽ disease-modifying antirheumatic drugs; NSAIDs ⫽ nonsteroidal antiinflammatory drugs (see Table 1 for other definitions). 2252 STAMP ET AL Table 4. Multivariate analysis of determinants of RBC MTXGlun concentrations* Variable Age Sex Smoking Any other DMARD Prednisone RBC folate Estimated GFR MTX dosage Duration of MTX treatment RA duration NSAIDs MTXGlu1 MTXGlu2 MTXGlu3 MTXGlu4 MTXGlu5 MTXGlu3–5 MTXGlu1–5 0.006 0.001 ⬍0.0001 ⬍0.0001 0.002 0.092 0.082 0.09 0.096 0.043 0.079 0.001 ⬍0.001 0.041 0.03 0.047 ⬍0.0001 0.092 ⬍0.0001 0.042 0.054 ⬍0.0001 0.06 ⬍0.0001 0.001 ⬍0.0001 0.036 0.033 * Variables with P values of less than 0.1 in the univariate model were included in the multivariate model. Data are presented as P values; all P values that were less than 0.1 in the multivariate model are shown. P values less than 0.05 were considered significant. DMARD ⫽ disease-modifying antirheumatic drug; GFR ⫽ glomerular filtration rate; NSAIDs ⫽ nonsteroidal antiinflammatory drugs (see Table 1 for other definitions). concentrations of MTXGlu1, MTXGlu2, MTXGlu4, and MTXGlu5 (Table 3). Even after adjustment for MTX dosage, smokers had significantly lower concentrations of MTXGlu3 (r ⫽ 0.45, P ⫽ 0.014), MTXGlu4 (r ⫽ 0.38, P ⫽ 0.05), MTXGlu3–5 (r ⫽ 0.45, P ⫽ 0.023), and MTXGlu1–5 (r ⫽ 0.45, P ⫽ 0.015). Effect of RBC folate on RBC MTXGlun concentrations. No significant association between RBC folate concentrations and RBC MTXGlun concentrations was observed (Table 2). Smoking was associated with lower RBC folate concentrations (mean ⫾ SD 628.6 ⫾ 44.1 nmoles/liter, versus 750.4 ⫾ 24.1 nmoles/liter in nonsmokers; P ⫽ 0.012). After adjustment for the effect of smoking, RBC folate levels were significantly positively associated with RBC concentrations of MTXGlu1 only (Table 2). Multivariate analysis of factors determining RBC MTXGlun concentrations. Backward stepwise multivariate regression analysis was undertaken to identify variables that were associated with individual RBC MTXGlun concentrations. Variables with a P value of ⬍0.1 for any RBC MTXGlun in the univariate analysis were assessed. This resulted in inclusion of the variables age, sex, smoking, use of other DMARDs, use of NSAIDs, use of prednisone, estimated GFR, MTX dosage, RBC folate level, duration of MTX therapy, and duration of RA. MTX dosage and age were most strongly associated with long-chain RBC MTXGlun concentrations. As expected, lower estimated GFR had a significant effect on all RBC MTXGlun except MTXGlu4, MTXGlu5, and MTXGlu3–5. NSAID use was significant only for MTXGlu3 and MTXGlu3–5, prednisone use was associated only with MTXGlu5, and RBC folate was associated only with MTXGlu1. There was no significant association between RBC MTXGlun concentrations and duration of RA, concomitant use of other DMARDs, or sex (Table 4). Age, MTX dosage, and estimated GFR accounted for 11% of the variability of RBC MTXGlu1, 2% of the variability of MTXGlu2, 28% of the variability of MTXGlu3, 23% of the variability of MTXGlu4, 16% of the variability of MTXGlu5, 30% of the variability of MTXGlu3–5, and 26% of the variability of MTXGlu1–5. DISCUSSION Despite its widespread use in a variety of rheumatic diseases, the exact mechanism of action of MTX remains unclear. The majority of pharmacokinetic studies with MTX have investigated serum concentrations only, which decrease rapidly after drug administration (2). MTX is taken into cells, where it is retained as MTX polyglutamates, which have a longer half-life. MTX polyglutamates are the biologically active component, and it has been suggested that they may have a role in therapeutic drug monitoring in RA (3). The potential relationship between MTXGlun concentrations and efficacy and/or toxicity needs to be confirmed, and variables affecting MTXGlun concentrations are of relevance in interpreting such data. Wide interpatient variability in MTX polyglutamate concentrations has been observed. While some of this variability may be due to large between-patient variability in oral availability of MTX (11), other factors have received little attention. We have shown that a variety of factors affect individual RBC MTXGlun concentrations and that MTX dosage, renal function, and age are the most important determinants. MTX dosage did not correlate well with RBC MTXGlu1 or MTXGlu2 levels, but higher doses were DETERMINANTS OF MTX POLYGLUTAMATES associated with higher concentrations of the longerchain MTX polyglutamates. While RBC concentrations of MTXGlu1, and to a lesser extent MTXGlu2, may reach steady state over days to weeks, the longer-chain polyglutamates (MTXGlu3–5) take many weeks to months to reach steady state and to be eliminated after treatment cessation (6). Thus, any relationship with MTX dosage is likely to be most apparent with the more stable longer-chain MTX polyglutamates, as we have shown. Longer duration of therapy with MTX was associated with higher RBC MTXGlu4 and MTXGlu5 concentrations in the univariate analysis, although this did not persist in the multivariate model. We have previously demonstrated progressive accumulation of MTXGlu2–5, starting with MTXGlu2 and followed sequentially by MTXGlu3, MTXGlu4, and MTXGlu5 (6). This occurs at least in part because RBC MTXGlu2–5 are produced intracellularly and MTXGlu1 must be present before other polyglutamates can be produced. Furthermore, glutamation of MTXGlun is a slow reaction, contributing to both the length of time required to reach steady state and the relationship between duration of therapy and MTXGlu4 and MTXGlu5 concentrations. MTX is largely eliminated via the kidneys, and impaired renal function may therefore predispose to accumulation of MTX. Studies of serum MTX concentrations have demonstrated decreased clearance and increased elimination half-life of MTX with renal impairment (12). From data pooled from 11 clinical trials including 496 patients, an increase in adverse effects of MTX in patients with renal impairment was reported, with an ⬃4-fold increase in the odds of severe toxicity in these patients. Although renal function deteriorates with age, increased age was not associated with higher risk of side effects in that study (13). Taken together, these data have led to suggestions that the dosage of MTX should be reduced in patients with renal impairment. Our data are consistent with the findings of previous studies in that worsening renal function was associated with higher RBC MTXGlun concentrations in the univariate analysis. Furthermore, the expected association persisted for MTXGlu1, MTXGlu2, MTXGlu3, and MTXGlu1–5 in the multivariate analysis. The association of renal function with the shorter-chain polyglutamates suggests that delayed clearance allows more MTX to be taken up into the cells. However, the lack of association with longer-chain polyglutamates suggests that intracellular glutamation is the rate-limiting step for production. Although we did not use a gold standard measure of renal function (e.g., 2253 radionuclide GFR), the MDRD equation is one that is now in widespread operational use globally and yields an estimate that is thus pertinent to everyday clinical practice. Age has a number of effects on both pharmacokinetics and pharmacodynamics. In a small study comparing plasma MTX pharmacokinetics in patients ages 65–83 years and patients ages 21–45 years, the clearance of MTX was inversely proportional to age, an effect that was likely mediated by deteriorating renal function with age (14). As would be expected, we demonstrated an association of increasing age with higher concentrations of long-chain MTX polyglutamates. However, because age is strongly associated with renal function, it is difficult to determine the independent effects of age and renal function in this context. Disease duration has been reported to be one of the most important factors determining response to therapy in patients with RA (15). Disease appears to be more responsive to therapy early in its course, which has led to the concept of a “window of opportunity” for remission induction in patients with early RA. We did not observe any effect of disease duration on RBC MTXGlun concentrations. This supports previous suggestions that alterations in the biologic process of RA over time may render patients less responsive to treatment (16). MTX is most often used in combination with NSAIDs, corticosteroids, and other DMARDs. Case reports of MTX toxicity have implicated an interaction between NSAIDs and MTX in precipitating toxicity (17). There are a number of potential mechanisms for the interaction between NSAIDs and MTX, including NSAID-induced inhibition of prostaglandin production leading to decreased renal blood flow and estimated GFR, and competitive inhibition of active renal tubular secretion of MTX by NSAIDs. Several studies have shown no change in serum MTX kinetics with concomitant use of NSAIDs (18,19). NSAIDs have also been reported to decrease renal clearance; however, the effect is dose dependent (20). We observed a significant reduction in RBC MTXGlu3 concentrations in patients receiving concomitant NSAID treatment, which persisted in the multivariate analysis. Further investigation is needed to determine whether this is a clinically significant interaction, and its mechanism. There are potential interactions between MTX and other commonly administered DMARDs. For example, in vitro sulfasalazine is a potent noncompetitive inhibitor of RFC-mediated cellular uptake of MTX and folate (21). In a small study of MTX and hydroxychlo- 2254 roquine in 10 healthy volunteers, the plasma MTX area under the curve (AUC) was increased and the maximum plasma MTX concentration was decreased when MTX was administered with hydroxychloroquine (22). An interaction between MTX and leflunomide may be mediated via a breast cancer resistance protein, resulting in increased drug concentrations (23). The number of patients receiving any one particular DMARD in our study was too small to allow calculation of individual associations. Further studies in this area are warranted, to determine the nature of these drug interactions and their clinical significance. Surprisingly, data on the pharmacologic interactions between MTX and corticosteroids are very limited. Findings of in vitro studies with murine cell lines suggest that hydrocortisone and prednisone inhibit MTX cellular uptake and antagonize MTX antitumor activity in a dose-dependent manner (24,25). In a canine model of monarticular inflammation, the ratio of synovial fluid to serum MTX concentration and the ratio of synovial fluid MTX concentration in the inflamed knee versus that in the uninflamed knee in the same animal were significantly reduced with prednisone treatment (26). These results suggest that, regardless of the serum MTX concentration, less MTX will enter the inflamed synovium and synovial fluid during treatment with prednisone. More recently, in a small study of 33 RA patients receiving intramuscular MTX, Lafforgue et al demonstrated that long-term corticosteroid therapy was associated with significantly higher AUC and lower clearance of serum MTX (27). We have shown that patients receiving prednisone have increased RBC MTXGlun concentrations, although in the multivariate analysis this remained significant only for MTXGlu5. These data are consistent with those of Lafforgue and colleagues in that higher AUC and lower clearance may allow increased uptake of MTX into cells. We acknowledge that our finding of higher MTXGlu5 concentrations in patients receiving prednisone could be a chance finding and needs to be confirmed in other studies; in addition, the clinical relevance of such a finding needs to be determined. Smoking is now considered to be an important environmental risk factor for the development of RA. The relative risk of RA has been reported to be increased in current smokers (1.43 [95% confidence interval 1.16–1.75]) and previous smokers (1.47 [95% confidence interval 1.23–1.76]) compared with those who have never smoked (28). Evidence for compounding of STAMP ET AL risk with smoking and presence of the shared epitope has been reported (29). The association between smoking and disease outcomes has been less consistent. Smoking has been associated with higher radiographic scores (30,31), with no difference in radiographic scores or radiographic progression (32), and with a trend toward reduced radiographic progression (33). Smoking has also been associated with both higher and lower tender and swollen joint counts and overall levels of disease activity (31,32). More recently, Westhoff et al reported that within 3 years of RA onset, smokers use significantly more DMARD combinations or biologic agents than do nonsmokers (34). Similarly, smoking has been reported to be associated with decreased efficacy of infliximab (but not etanercept) in RA (35). In a clinical pharmacogenetic model predicting response to MTX, smoking was 1 of only 3 relevant clinical variables, the other being sex and RF status (36). We have shown that smoking is associated with significantly lower RBC MTXGlu3 concentrations. There are several potential mechanisms for the observed reduction of MTXGlu3 in smokers. Cigarette smoking has been reported to increase the basal metabolic rate in RA patients, which may influence drug metabolism. Given the structural similarity between MTX and folate, it is plausible that smoking will result in similar effects on MTX. Ingested folates are transported into cells by the RFC and folate receptor and compete with MTX as a substrate for polyglutamation by FPGS. Thus, high concentrations of intracellular folates may result in a decrease in MTX polyglutamation (37,38). In patients with psoriasis, use of folic acid 20 mg per week concomitantly with MTX resulted in lower RBC MTXGlun concentrations compared with those found in patients taking MTX alone, although the difference did not reach statistical significance (39). While we observed a trend toward lower RBC MTXGlu1 concentrations with higher RBC folate levels, this reached statistical significance in the multivariate analysis only. These results suggest that uptake of MTX into cells may be affected by coadministration of folate. We acknowledge that this study has some weaknesses. The cross-sectional design does not allow us to conclusively identify trends in MTX polyglutamate concentrations and the factors that may relate to these. Because we investigated univariate associations between 16 independent variables and 7 dependent variables (MTXGlun), there were a considerable number of sta- DETERMINANTS OF MTX POLYGLUTAMATES tistical comparisons. To some extent the effects of this number of comparisons on the Type I error rate are mitigated by generation of multivariate models using backward stepwise regression. However, the models and conclusions from this study will require validation in other patient groups. We have shown that up to 30% of interindividual variability in MTXGlun concentrations can be explained by nongenetic factors. Previous studies have identified polymorphisms of the folate–purine–pyrimidine pathway that are associated with improved response to MTX (3). The combination of genetic and nongenetic factors may better explain interpatient variability in MTXGlun concentrations and clinical response to MTX. It has been suggested that MTXGlun concentrations may be used in therapeutic drug monitoring in RA. The value of such an approach requires confirmation with both crosssectional studies of patients with established MTX treatment and prospective studies of patients beginning treatment. In summary, large interpatient variability in RBC MTXGlun concentrations can be explained, at least in part, by several patient factors. Age, MTX dosage, and renal function are the most important factors. In addition, there are complex interactions between smoking, RBC folate levels, and MTX polyglutamate concentrations. Better understanding and greater awareness of these factors will help guide optimal dosing of MTX in patients with RA. 2255 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. ACKNOWLEDGMENT We gratefully acknowledge the assistance of Jan Ipenburg, Rheumatology Clinical Nurse Specialist, in assisting with patient data collection. 13. 14. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. 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 conception and design. Stamp, O’Donnell, Chapman, Frampton, Barclay. Acquisition of data. Stamp, O’Donnell, Chapman, Zhang, James. Analysis and interpretation of data. Stamp, O’Donnell, Chapman, Frampton, James, Barclay. REFERENCES 1. 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. 2. Tishler M, Caspi D, Graff E, Segal R, Peretz H, Yaron M. Synovial 15. 16. 17. 18. 19. 20. and serum levels of methotrexate during methotrexate therapy of rheumatoid arthritis. Br J Rheumatol 1989;28:422–3. 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. Dervieux T, Greenstein N, Kremer J. Pharmacogenomic and metabolic markers in the folate pathway and their association with methotrexate effects during dosage escalation in rheumatoid arthritis. Arthritis Rheum 2006;54:3095–103. Angelis-Stoforidis P, Vajda FJ, Christophidis N. Methotrexate polyglutamate levels in circulating erythrocytes and polymorphs correlate with clinical efficacy in rheumatoid arthritis. Clin Exp Rheumatol 1999;17:313–20. Dalrymple JM, Stamp LK, O’Donnell JL, Chapman PT, Zhang M, Barclay ML. Pharmacokinetics of oral methotrexate in patients with rheumatoid arthritis. Arthritis Rheum 2008;58:3299–308. Allegra CJ, Drake JC, Jolivet J, Chabner BA. Inhibition of phosphoribosylaminoimidazolecarboxamide transformylase by methotrexate and dihydrofolic acid polyglutamates. Proc Natl Acad Sci U S A 1985;82:4881–5. Baggott JE, Vaughn WH, Hudson BB. Inhibition of 5-aminoimidazole-4-carboxamide ribotide transformylase, adenosine deaminase and 5⬘-adenylate deaminase by polyglutamates of methotrexate and oxidized folates and by 5-aminoimidazole-4-carboxamide riboside and ribotide. Biochem J 1986;236:193–200. 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. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D, et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;130:461–70. Bannwarth B, Pehourcq F, Schaeverbeke T, Dehais J. Clinical pharmacokinetics of low-dose pulse methotrexate in rheumatoid arthritis. Clin Pharmacokinet 1996;30:194–210. Bressolle F, Bologna C, Kinowski J, Sany J, Combe B. Effects of moderate renal insufficiency on pharmacokinetics of methotrexate in rheumatoid arthritis patients. Ann Rheum Dis 1998;57:110–3. Rheumatoid Arthritis Clinical Trial Archive Group. The effect of age and renal function on the efficacy and toxicity of methotrexate in rheumatoid arthritis. J Rheumatol 1995;22:218–23. Bressolle F, Bologna C, Kinowski JM, Arcos B, Sany J, Combe B. Total and free methotrexate pharmacokinetics in elderly patients with rheumatoid arthritis: a comparison with young patients. J Rheumatol 1997;24:1903–9. Anderson JJ, Wells G, Verhoeven AC, Felson DT. Factors predicting response to treatment in rheumatoid arthritis: the importance of disease duration. Arthritis Rheum 2000;43:22–9. Emery P, Salmon M. Early rheumatoid arthritis: time to aim for remission? Ann Rheum Dis 1995;54:944–7. Thyss A, Milano G, Kubar J, Namer M, Schneider M. Clinical and pharmacokinetic evidence of a life threatening interaction between methotrexate and ketoprofen. Lancet 1986:8475;256–8. Ahern M, Booth J, Loxton A, McCarthy P, Meffin P, Kevat S. Methotrexate kinetics in rheumatoid arthritis: is there an interaction with nonsteroidal antiinflammatory drugs? J Rheumatol 1988;15:1356–60. Stewart CF, Fleming RA, Arkin CR, Evans WE. Coadministration of naproxen and low-dose methotrexate in patients with rheumatoid arthritis. Clin Pharmacol Ther 1990;47:540–6. Kremer JM, Hamilton RA. The effects of nonsteroidal antiinflammatory drugs on methotrexate pharmacokinetics: impairment of 2256 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. renal clearance of MTX at weekly maintenance doses but not at 7.5mg. J Rheumatol 1995;22:2072–7. Jansen G, van der Heijden J, Oerlemans R, Lems WF, Ifergan I, Scheper RJ, et al. Sulfasalazine is a potent inhibitor of the reduced folate carrier: implications for combination therapies with methotrexate in rheumatoid arthritis. Arthritis Rheum 2004;50:2130–9. Carmichael SJ, Beal J, Day RO, Tett SE. Combination therapy with methotrexate and hydroxychloroquine for rheumatoid arthritis increases exposure to methotrexate. J Rheumatol 2002;29: 2077–83. Kis E, Nagy T, Jani M, Molnar E, Janossy J, Ujhelly O, et al. Leflunomide and A771726, its metabolite are high affinity substrates of BCRP: implications for drug resistance. Ann Rheum Dis 2009;68:1201–7. Zager RF, Frisby SA, Oliverio VT. The effects of antibiotics and cancer chemotherapeutic agents on the cellular transport and antitumor activity of methotrexate in L1210 murine leukaemia. Cancer Res 1973;33:1670–6. Bruckner HW, Schreiber C, Waxman S. Interaction of chemotherapeutic agents with methotrexate and 5-fluorouracil and its effects on de novo DNA synthesis. Cancer Res 1975;35:801–6. Stewart CF, Christensen ML, Evens RP, Cremer M, Evans WE. Influence of concomitant aspirin or prednisone on methotrexate synovial fluid concentration. J Pharmacol Exp Ther 1987;243: 131–7. Lafforgue P, Monjanel-Mouterde S, Durand A, Catalin J, Acquaviva P. Is there an interaction between low doses of corticosteroids and methotrexate in patients with rheumatoid arthritis? A pharmacokinetic study in 33 patients. J Rheumatol 1993;20:263–7. Costenbader KH, Feskanich D, Mandl LA, Karlson EW. Smoking intensity, duration and cessation, and the risk of rheumatoid arthritis in women. Am J Med 2006;119:503–11. Padyukov L, Silva C, Stolt P, Alfredsson L, Klareskog L, for the Epidemiological Investigation of Rheumatoid Arthritis Study Group. A gene–environment interaction between smoking and shared epitope genes in HLA–DR provides a high risk of seropositive rheumatoid arthritis. Arthritis Rheum 2004;50:3085–92. Saag KG, Cerhan JR, Kolluri S, Ohashi K, Hunninghake GW, STAMP ET AL 31. 32. 33. 34. 35. 36. 37. 38. 39. Schwartz DA. Cigarette smoking and rheumatoid arthritis severity. Ann Rheum Dis 1997;56:463–9. Papadopoulos NG, Alamanos Y, Voulgari PV, Epagelis EK, Tsifetaki N, Drosos AA. Does cigarette smoking influence disease expression, activity and severity in early rheumatoid arthritis patients? Clin Exp Rheumatol 2005;23:861–6. Harrison BJ, Silman AJ, Wiles NJ, Scott DG, Symmons DP. The association of cigarette smoking and disease outcome in patients with early inflammatory polyarthritis. Arthritis Rheum 2001;44: 323–30. Finckh A, Dehler S, Costenbader KH, Gabay C, on behalf of the Swiss Clinical Quality Clinical Management Project for RA. Cigarette smoking and radiographic progression in rheumatoid arthritis. Ann Rheum Dis 2007;66:1066–71. Westhoff G, Rau R, Zink A. Rheumatoid arthritis patients who smoke have a higher need for DMARDs and feel worse, but they do not have more joint damage than non-smokers of the same serological group. Rheumatology (Oxford) 2008;47:849–54. Hyrich KL, Watson KD, Silman AJ, Symmons DP. Predictors of response to anti-TNF-␣ therapy among patients with rheumatoid arthritis: results from the British Society for Rheumatology Biologics Register. Rheumatology (Oxford) 2006;45:1558–65. Wessels JA, van der Kooij SM, le Cessie S, Kievit W, Barerra P, Allaart CF, et al. A clinical pharmacogenetic model to predict the efficacy of methotrexate monotherapy in recent-onset rheumatoid arthritis. Arthritis Rheum 2007;56:1765–75. Jolivet J, Faucher F, Pinard M. Influence of intracellular folates on methotrexate metabolism and cytotoxicity. Biochem Pharmacol 1987;36:3310–2. Kennedy DG, van den Berg HW, Clarke R, Murphy RF. The effect of leucovorin on the synthesis of methotrexate poly-␥glutamates in the MCF-7 human breast cancer cell line. Biochem Pharmacol 1985;34:2897–903. Chladek J, Simkova M, Vaneckova J, Hroch M, Chladkova J, Martinkova J, et al. The effect of folic acid supplementation on the pharmacokinetics and pharmacodynamics of oral methotrexate during the remission-induction period of treatment for moderateto-severe plaque psoriasis. Eur J Clin Pharmacol 2008;64:347–55.