Modifying Drugs

478 Annals ofthe Rheumatic Diseases 1996; 55: 478-481

Urinary glucaric acid excretion in rheumatoid Ann Rheum Dis: first published as 10.1136/ard.55.7.478 on 1 July 1996. Downloaded from arthritis: influence of disease activity and disease modifying drugs

R Addyman, C Beyeler, C Astbury, H A Bird

Abstract of drug toxicity as much as possible, especially Objective-To examine if a correlation in view of the polypharmacy that is common. exists between cytochrome P-450 enzyme Against the background of the subject's induction and disease activity in patients genetic constitution, drug metabolism- with rheumatoid arthritis (RA), measuring involving different enzyme systems of bio- urinary excretion of D-glucaric acid (GA) transformation-is affected by many variables as an index ofphase II drug metabolism. such as age, gender, diet, smoking, other Methods-Patients with RA were treated environmental factors, and drugs,' 2 and also with sulphasalazine, sodium aurothio- by hepatic and extrahepatic diseases such as malate, or D-penicillamine in standard diabetes mellitus, thyroid disease, cardiopul- dose regimens, for 24 weeks. Patients monary dysfunction, and malignancy. Phase I with ankylosing spondylitis (AS) or non- drug metabolism involves the mixed function inflammatory arthritis (NIA) acted as oxidase enzyme, cytochrome P-450, and is controls. The urinary GA:creatinine ratio considered to be the principal drug metabol- was measured at 0, 12, and 24 weeks of ising enzyme system. Located predominantly treatment. on the smooth endoplasmic reticulum, it Results-Patients with RA had a slightly catalyses the oxidation of a wide variety of greater urinary GA:creatinine ratio than compounds and it can be induced or inhibited patients with AS or NIA at baseline; this by a wide range of chemically unrelated increased during treatment with disease substances, including drugs.2 This is followed modifying antirheumatic drugs (DMARDs). by a change in drug efficacy and toxicity, Sulphasalazine treatment had a greater especially for drugs with a small therapeutic effect on GA excretion than sodium ratio and with long term application, necessi- aurothiomalate or D-penicillamine; this tating an alteration in the dosage schedule. http://ard.bmj.com/ difference was statistically significant States of inflammation also may affect the between weeks 0 and 12 (p = 0-01). Gamma rate of drug metabolism. In man, it has been glutamyltranspeptidase concentration shown that fever induced by the steroid showed a weak correlation with GA excre- pyrogen etiocholanolone or injected endotoxin tion between weeks 0 and 12 (p = 0.03), but is followed by impaired hydroxylation of all other measurements of changes in antipyrine, conjugation of bromosulphthalein, disease activity (plasma viscosity, C and glucuronidation of salicylamide. Acute on October 10, 2021 by guest. Protected copyright. reactive protein, platelets, and articular viral infections and influenza vaccination can index) were found not to correlate with impair cytochrome P-450 dependent oxidative GA excretion between weeks 0-12 or 0-24. drug metabolism leading to toxicity of theo- Conclusion-The increased excretion of phylline, phenytoin, and warfarin. In animal GA in patients with RA receiving DMARD models with adjuvant arthritis, a marked treatment is probably the result of an in- impairment of drug metabolism in vitro and in Clinical Pharmacology direct effect on hepatic metabolism bearing vivo has been shown.35 Unit (Rheumatism no relationship to disease activity. There is a need to explore other hepatic Research), Chapel Allerton Hospital, metabolic systems including phase II meta- Leeds, United (Ann Rheum Dis 1996; 55: 478-481) bolism-which involves the conjugation of Kingdom oxidised drugs with an ionisable compound, R Addyman Rheumatoid arthritis C Astbury (RA) is the most such as glucuronide or sulphate-as these H A Bird common inflammatory rheumatic disease of systems may act as markers of disease activity Department of unknown aetiology, characterised by persistent or predict disease response in RA. One such Rheumatology, synovitis and complicated by local destruction urine marker of phase II metabolism is University Hospital, of bone and cartilage and various systemic D-glucaric acid (GA), already shown to be CH-3010 Berne, Switzerland manifestations. Its treatment, usually over altered in patients with RA.6 Sugars, including C Beyeler long periods, comprises compounds pro- glucose and galactose, are broken down by Correspondence to: viding symptomatic relief (analgesics, non- a catabolic pathway producing GA from Dr H A Bird, steroidal anti-inflammatory drugs (NSAIDs)) D-glucuronic acid, which can be measured in Clinical Pharmacology Unit, Chapel Allerton Hospital, and disease modifying anti-rheumatic drugs urine. There is also evidence that this step in Chapeltown Road, (DMARDs) with a later onset effect (sodium the pathway may act as an indicator of Leeds LS7 4SA, United Kingdom. aurothiomalate, D-penicillamine, sulphasala- enhanced metabolism induced by drugs. Accepted for publication zine, and others). The aim of treatment is to Drugs affecting the pathway in either man or 27 February 1996 improve disease activity and to reduce the risk animals include phenobarbitone, amidopyrine, Urinary glucaric acid excretion in rheumatoid arthritis 479

progesterone, thiopentone, organochloride Control subjects comprised patients with pesticides, diphenytoin, rifampicin, and the ankylosing spondylitis (AS) (n = 13)

anti-inflammatory, phenylbutazone.7 8 (according to the modified New York Ann Rheum Dis: first published as 10.1136/ard.55.7.478 on 1 July 1996. Downloaded from Other factors clearly influence the system. criteria'6), four of whom were receiving enteric Urinary excretion of GA varies between men coated sulphasalazine before they entered the and women (53 (SD 10) ,umoV24 h and 41 (8) study. A second control group comprised ,umolV24 h, respectively) though it is not patients with non-inflammatory arthritis (NIA) influenced by the menstrual cycle.9 Excretion (n = 16), characterised by normal plasma is increased twofold in pregnancy and 3-5-fold viscosity and serum concentration ofC reactive in diabetes mellitus, but decreases in neoplastic protein (CRP), including patients with non- disease.'0 Urinary output of GA is also articular rheumatism (n = 7), osteoarthritis increased in acute viral hepatitis. Fasting (n = 5), cervical spondylosis (n = 2), low back reduces excretion twofold within two days,8 pain (n = 1), and Sudeck's atrophy (n = 1), or while chronic cardiac insufficiency, burns, and a combination of these conditions. folic acid status all influence metabolism. Patients attended a special clinic at weeks 0, There is a correlation between GA and total 12, and 24. A variety of clinical assessments liver concentration of cytochrome P-450." 12 were performed, including those of the Ritchie Clearly, GA excretion is also influenced by articular index and early morning stiffness. renal function; coefficients of variation as high Haematological assessments included haemo- as 37% and 27% have been described.9' globin and platelet count, and biochemical Adjustment of estimates of GA excretion by assessments of disease activity included plasma expressing it as a proportion of creatinine viscosity, CRP and gamma glutamyl output in a 24 hour urine sample allows the transpeptidase (GGT) concentrations. At each best option of correcting for incomplete visit, the patient provided a 24 hour urine speci- collection and, possibly, renal function.'4 men for estimation of GA:creatinine ratio. We have investigated GA excretion in The volume of each 24 hour urine sample patients with RA treated with three different was measured and aliquots frozen and stored DMARDs in order to determine the influence, at -20°C until required for analysis. Urinary if any, of different drug treatments and the glucaric acid was determined by the method extent to which excretion correlates with of Simmons et al: 16 a spectrophotometric conventional indices of rheumatoid disease assay involving inhibition of P-glucuronidase/ activity. It also remains a possibility that phenophthalein glucuronide hydrolysis by enhanced GA excretion in RA, if confirmed, D-glucaro-1,4-lactone, the latter being the may be brought under control with drug product ofurinary GA hydrolysis at 100°C and treatment and might provide an indication pH 3-6. Urinary creatinine was measured by a either of hepatic function or of enzyme standard spectrophotometric assay. The GA: induction in patients with this disease. creatinine ratio was calculated (intra-assay precision 3-29%, interassay precision 9-98%). http://ard.bmj.com/ Statistical treatment of data included Patients and methods Spearman rank correlation, Wilcoxon signed Fifty seven patients with RA (12 men and 45 rank test, Mann-Whitney U test and Kruskal- women: age range 29-80 years (mean 59-8 Wallis analysis of variance as appropriate. All (SD 12-5) years)) were recruited to the study. tests were performed using Unistat IV statis- Twenty nine patients were recruited to serve as tical software on a personal computer. controls: 12 men and one woman with AS (age on October 10, 2021 by guest. Protected copyright. range 28-55 years (mean 45-4 (8-8) years)), and five men and 11 women with NIA (age Results range 40-71 years (mean 56-6 (10-8) years)). Among the 57 patients with RA, 19 received Patients with RA met the revised criteria of sulphasalazine, 16 received D-pencillamine, the American Rheumatism Association'5 and and 22 received sodium aurothiomalate after had disease activity sufficient to justify the adequate periods of washout of other DMARDs. introduction of DMARDs. They were allo- In addition, six of the patients were each cated to treatment with enteric coated sulpha- receiving concomitant treatment with one of salazine 0.5 g/day for two weeks, 1 g/day for the following potential cytochrome P-450 two weeks, 1-5 g/day for four weeks, and 2-0 enzyme inducing drugs: hormone replace- g/day thereafter), D-penicillamine (125 mg/ ment therapy, norgestrel, levonorgestrel, tam- day for two weeks, 250 mg/day for two weeks, oxifen, quinalbarbitone, phenylbutazone. Five 375 mg/day for two weeks, and 500 mg/day patients were receiving cimetidine, a known thereafter), or sodium aurothiomalate by intra- inhibitor of cytochrome P-450. Six patients muscular injection (10 mg test dose followed had coronary heart disease and two had by 50 mg/week for 20 weeks and 50 mg/month diabetes mellitus. GA excretion was neither thereafter). Where necessary, because of incresed nor decreased at baseline in any of toxicity, dosage was reduced. NSAIDs and all these patients. other treatments were continued at the same The mean duration of disease activity in dose throughout the 24 week study period patients taking sulphasalazine, D-penicilla- wherever possible. Particular attention was mine, and gold was 12-9, 15-5, and 9-4 years, paid to those drugs likely to influence the respectively (medians 8, 15, and 8-5 years, cytochrome P-450 system. Where change was respectively). Of these three drugs, penicilla- necessary on grounds of toxicity, this was mine was the one that subsequently showed carefully documented. least effect on GA:creatinine excretion ratio. 480 Addyman, Beyeler, Astbury, Bird

15- The same trends were seen when the data o _5 were analysed in two groups according to gender, though the median GA:creatinine ratio Ann Rheum Dis: first published as 10.1136/ard.55.7.478 on 1 July 1996. Downloaded from co E 10-' was slightly smaller in men (6 5 (4 0-8 1) .E E T pLmol/mmol) than in women (8 8 (5 8-14 6) 0E5- Lmol/mmol) at baseline. The GA:creatinine ratio plotted against age .. showed good distribution in both RA and control groups, with no correlation between Week 0 Week 12 Week 24 age and GA:creatinine ratio in either group Figure I Urinary glucanrc acid (GA) :creatinine ratio in (Spearman rank correlation p = 034 and patients with rheumatoid arthnrtis (U), ankylo?sing p = 0-21, respectively). spondylitis (n), or non-inflammatory arthritiss(C:) Patients with RA who smoked had slightly median and interquartile range. smaller GA:creatinine ratios compared with non- smokers, though this did not reach statistical significance (Mann-Whitney U test, = 0 a) p 32). c 80 Table 1 shows the effect of DMARD 60- intervention on GA:creatinine ratio in patients C>. - with RA. There was a difference according to < the drug used to obtain remission. For sulpha- 20 salazine, the GA:creatinine ratio increased consistently between weeks 0 and 12 as a X -40 - response was obtained, and it remained increased thereafter. There was a more modest .ca) 2C-0 increase in groups of patients treated with Week 12 WeDek 24 sodium aurothiomalate and D-pencillamine, Figure 2 Change in glucaric acid (GA):creati over 24 weeks compared with baseline values, i7ninepratiot the maximum change again occurring between with rheumatoid arthritis (U), ankylosing spo?ndylitis weeks 0 and 12, at the time when clinical (0), or non-inflammato?y arthritis (D): meidian and improvement was usually most pronounced. interquartile range. An expected decrease in plasma viscosity was observed in the patients with RA over the At baseline, median GA excretion was 57 24 week period: from 1 83 (1 72-1-94) mPa at ,umolI24 h (interquartile range 41-877 imol/24 week 0, decreasing to 1-75 (1 69-1 89) mPa h) in RA, 68 (34-111) p.mol124 h ir1 AS, and and to 1 74 (1-64-1-83) mPa at weeks 12 and 64 (45-83) p.mol124 h in NIA (Kruskal-Wallis 24, respectively (Wilcoxon signed rank test: analysis of variance, p = 0 85). Figure 1 shows p = 0 003 and p = 0 0003, respectively). Early the median GA excretion corre for morning stiffness and articular index, the only ~cted http://ard.bmj.com/ creatinine excretion (GA:creatinine ratio) for other two clinical parameters measured, both all patient groups at weeks 0, 12, and 24, while decreased with disease modifying intervention, figure 2 shows the changes in GA: creatinine but this was not correlated with changes with ratio over 24 weeks in patients with RA, AS, changes in GA excretion. and NIA. Median GGT activity was 20 (14-33) U/! at Patients with RA had a slightly greater baseline and increased slightly at week 12, to urinary GA:creatinine ratio than those with AS 22 (12-32) U/1, followed by a statistically or NIA at baseline, though this failecd to reach significant decrease from baseline, to 19 (12-29) on October 10, 2021 by guest. Protected copyright. statistical significance (Kruskal-Walliis analysis U/l at week 24 (Wilcoxon signed rank test, of variance, p = 0 26). During DMA:RD treat- p = 0 0 1). There was a weak correlation ment, the GA:creatinine ratio increased until, between changes in GA:creatinine ratio and at weeks 12 and 24, patients with IRA had a changes in GGT for weeks (0- 12) (Spearman statistically significant greater GA: creatinine rank correlation, p = 0 03). ratio than those with AS or NIA (Kruskal- Clinical, biochemical, and haematological Wallis analysis of variance, p =O C)003 and markers of disease activity all showed a de- p = 0-002, respectively). crease upon DMARD intervention, but these decreases did not correlate with changes in the Effect ofdisease modiymng antirheumatic drug (DMARD) intervention on urnary glucanrc GA:creatinine ratio. acid (GA):creatinine ratio in patients with rheumatoid arthritis DMARD GA:creatinine ratio (,umol/mmol) tP Week 0 Week 12 Week 24 Week 0 Week 0 Discussion v v The aim of this study was to investigate Week 12 Week 24 whether a correlation exists between cyto- Sulphasalazine 7 1 14 3 14-6 chrome P-450 enzyme induction and disease (5-8-14 6) (11.3-16 2) (6-8-21-2) 0 061 0 125 activity in patients suffering from RA, using (n= 19) (n= 15) (n= 15) Gold 7.3 110 10-2 patients with AS as a secondary inflammatory (54-11-4) (8-9-16 6) (7 7-14.0) 0 040 0-314 polyarthritis and patients with NIA as control (n = 22) (n = 20) (n = 19) groups. Urinary GA was measured as an index D-Penicillamine 9-2 9.9 9.7 (62-15-5) (6 3-164) (8 1-16 1) 0-638 0-071 for phase II drug metabolism. (n = 16) (n = 14) (n = 12) Analysis of the results for those potentially tp 0-3648 0 2125 0.5198 confounding patients who were receiving Values are median (interquartile range). cytochrome P-450 enzyme inducing drugs tWilcoxon signed rank test; tKruskal-Wallis analysis of variance. (quinalbarbitone, levonorgestrel, tamoxifen, Urinary glucanic acid excretion in rheumatoid arth?itis 481

and phenylbutazone) and enzyme inhibiting 1 Park B K, Breckenridge A M. Clinical implications of enzyme induction and enzyme inhibition. Clin Pharmaco- drugs (cimetidine) revealed no increase or kinet 1981; 6: 1-24.

decrease in the GA:creatinine ratio at week 0. 2 Kupfer A, Preisig R. Inherited defects of hepatic drug Ann Rheum Dis: first published as 10.1136/ard.55.7.478 on 1 July 1996. Downloaded from metabolism. Semin Liver Dis 1983; 3: 314-54. Statistically significant findings were thought 3 Morton D M, Chatfield D H. The effects of adjuvant not to result from inaccuracies in the assay induced arthritis on the liver metabolism of drugs in rats. Biochem Pharmacol 1970; 19: 473-81. technique, as the precision of the assay was 4 Beck F J, Whitehouse M W. Effect of adjuvant disease in very good. Even though a large between patient rats on cyclophosphamide and isophosphamide metabolism. Biochem Pharmacol 1973; 22: 2453-68. variability was seen, the median baseline GA 5 Cawthorne M A, Palmer E D, Green J. Adjuvant-induced excretion values were well matched between arthritis and drug metabolizing enzymes. Biochem Pharmacol 1976; 25: 2683-8. patient groups and were within the ranges 6 McKenna F, Coombes G, Sandle L, Mather H. u-Glucaric previously published"7. It is well established acid in rheumatoid arthritis [abstract]. Br Jf Rheumatol 1992; 31 (suppl 2): 115. that increased GA excretion caused by foreign 7 Aarts E M. Evidence for the function of D-glucarcid acid as compounds nearly always indirectly accom- an indicator for drug induced enhanced metabolism through the glucuronic acid pathway in man. Biochem panies an enhancement of cytochrome P-450 Pharmacol 1965; 14: 359-63. reactions. However, there is no evidence in the 8 Sotaumi E A, Medzihradsky F, Eliasson G. Glucaric acid as an indicator of use of enzyme-inducing drugs. Clin literature that any of the DMARDs used here Pharmacol Ther 1974; 15: 417-23. for patient treatment induces or inhibits any of 9 Mowat A P. Developmental effects of liver D-glucuro- nolactone dehydrogenase levels on D-glucaric acid the cytochrome P-450 isoenzymes. Increased excretion in urine; hormal effects on i-glucaric acid GA excretion was therefore unlikely to be excretion in urine. JEndocrinol 1968; 42: 585-90. 10 Marsh C A, Carr J. Changes in enzyme activity related to caused by direct induction of cytochrome n-glucaric acid synthesis with age, pregnancy and P-450 as a result of DMARD treatment. malignancy. Clin Sci 1965; 28: 209-17. 11 Hunter J, Maxwell J D, Stewart D A, Williams R. Urinary The weak correlation seen with changes in i-glucaric acid excretion and total liver content of GGT and GA:creatinine ratio between week 0 cytochrome P450 in guinea pigs: relationship during enzyme induction and following inhibition of protein and week 12 may not be of practical sig- synthesis. Biochem Pharnacol 1973; 22: 743-7. nificance. However, the non-significance of 12 Lecamwasam D S, Franklin C, Turner P. Effect of phenobarbitone on hepatic drug-metabolising enzymes correlation between other markers of disease and urinary i-glucaric acid excretion in man. Br J Clin activity and change in GA excretion indicates Pharmacol 1975; 2: 257-62. 13 Gilbert J C, Scott A K, Galloway D B, Petric J C. that the increase in GA excretion between Ethosyximide: liver enzyme induction and n-glucaric acid weeks 0 and 12 is unlikely to reflect an excretion. BrJClin Pharmacol 1974; 1: 249-52. 14 HunterJ, MaxwellJ D, Carrella M, Stewart D A, Williams R. improvement in disease activity, but is more D-glucaric acid excretion as a test for hepatic enzyme possibly the result of increased production of induction in man. Lancet 1971; i: 572-5. 15 Arnett F C, Edworthy S M, Bloch D, et al. The American glucuronate via the degradation of haemo- Rheumatism Association 1987 revised criteria for the globin to urobilinogen, because of increased classification of rheumatoid arthritis. Arthritis Rheum 1988; 31: 315-24. liver metabolism. We conclude, therefore, that 16 van der Linden S, Valkenburg H A, Cats A. Evaluation of glucaric acid excretion cannot be used as a diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum satisfactory index for disease activity in RA. 1984; 27: 361-8. 17 Simmons C J, Davis M, Dordoni B, Williams R. Urinary We thank the Swiss National Foundation and the Arthritis and i-glucaric acid assay by an improved enzymatic http://ard.bmj.com/ Rheumatism Council (B0132) for financial support. We also procedure. Clin ChimActa 1974; 51: 47-51. thank Mrs J L Childs for secretarial help. on October 10, 2021 by guest. Protected copyright.