1 Rheumatology 2002;41: Does active treatment of rheumatoid arthritis limit disease-associated bone loss? A. L. Dolan, C. Moniz 1, H. Abraha 1 and P. Pitt 2 Department of Rheumatology, Queen Elizabeth Hospital, London SE18, 1 Department of Clinical Biochemistry, Kings College Hospital, London SE5 and 2 Department of Rheumatology, Bromley Hospital, Kent, UK Abstract Objective. Generalized bone loss in rheumatoid arthritis (RA) is multi-factorial, with the inflammatory disease itself thought to contribute to bone loss. To study the extent to which control of disease activity affects bone turnover in RA and whether treatment with diseasemodifying anti-rheumatic drugs (DMARDs) reduces bone turnover and loss of bone mass, we measured bone density and biochemical markers of bone resorption in a group of patients with active RA starting on DMARDS. Methods. Patients with active RA were enrolled on starting a new DMARD. Patients were mobile and none took steroids or any treatment for osteoporosis. Clinical and laboratory measures of disease activity were made at 3-monthly intervals and an index of disease activity (DAS) calculated. Bone density was assessed at 0, 1 and 2 yr (Hologic QDR 4500c). Urinary deoxypyridinoline (D-PYR) and pyridinoline (PYR) were measured by ELISA at 0, 3, 6, 9 and 12 months. Results. Forty patients were enrolled, mean age 59.5 (range 31 76), 26 female, 14 male, 25 had established RA, 15 had RA for <2 yr. Baseline D-PYR was elevated (8.4"4.55 nmolummol creatinine) and correlated with ESR (r=0.6, P<0.01) and DAS (r=0.4, P<0.05). On treatment ESR and DAS fell by 38.5 and 29.3%, respectively. D-PYR was reduced by 12.3% by 9 months (P<0.01). Spearman rank order correlation showed ESR to be the most significant determinant of D-PYR over 1 yr (r=0.43, P<0.001). Serial bone density was available on 21 patients. There was no significant change in BMD over the 2 yr. The change in DAS over 0 3 months showed an inverse relationship with the percent change in spine over 1 yr (r=20.5, P=0.05). The change in D-PYR over 0 3 months was not closely related to the change in BMD at hip or spine at 1 yr. Conclusion. Disease activity is a significant determinant of bone turnover in RA. Bone resorption markers fall on treatment of RA with DMARDs and no change in BMD was demonstrated at 2 yr. This study suggests the need to control disease activity in RA in order to prevent systemic bone loss. KEY WORDS: Rheumatoid arthritis, Inflammation, Osteoporosis, Bone turnover, Bone markers. Rheumatoid arthritis (RA) is associated with a high prevalence of bone loss resulting in juxta-articular and generalized osteoporosis w1, 2x. Generalized bone loss may be influenced by immobility, the inflammatory process and treatments such as steroids, whilst juxtaarticular loss is probably due to local release of inflammatory agents such as cytokines from the rheumatoid synovium and articular immobility w1x. The consequence is osteoporotic fractures of hip, vertebrae and peripheral bones w3x resulting in considerable cost in both acute orthopaedic management, rehabilitation and increased dependence. Submitted 2 April 2001; revised version accepted 25 March Correspondence to: A. L. Dolan, Department of Rheumatology, Queen Elizabeth Hospital, Stadium Road, London SE18 4QH, UK. Current practice is to introduce treatment early in RA. UK guidelines recommend the use of diseasemodifying anti-rheumatic drugs (DMARDs) to reduce inflammatory synovitis and control disease progression w4x. In early RA intervention, using DMARD, has been shown to improve functional outcome and retard bone loss. Emery w5x suggests that bone density assessment using dual energy X-ray absorption (DEXA) scans provides an outcome measure for assessing the consequences of active disease and the response to treatment. However, these measurements are expensive and not widely available, necessitating referral elsewhere and patient inconvenience. It is now possible to assess bone turnover using a range of biochemical markers to reflect bone resorption and formation w6x. Pyridinoline (PYR) and 1047 ß 2002 British Society for Rheumatology
2 1048 A. L. Dolan et al. deoxypyridinoline (D-PYR) assess collagen breakdown. These cross-links form between the two collagen fibrils and on breakdown of collagen are released unchanged in the urine. PYR is the main cross-link between type II collagen fibres in cartilage and although PYR is also the most abundant cross-link in bone, D-PYR is bone specific being derived only from type 1 collagen. D-PYR has been validated as a useful marker for bone resorption and is elevated in high turnover states such as Paget s and post-menopausal osteoporosis w6x. Crosssectional studies have shown resorption markers to be elevated in active RA w7 9x. PYR levels correlated with other cartilage markers and with ESR, suggesting levels reflect joint damage and joint inflammation. D-PYR from bone also correlated with disease activity and may thus reflect diffuse bone loss. Gough w10x found PYR elevation to correlate with other markers of disease activity and bone density loss at the femoral neck. Biochemical markers may thus provide an alternative indication of change in bone turnover in response to treatment. Previous studies of bone markers in RA have been cross-sectional and in small numbers w7, 8x. Such studies do not allow assessment of whether markers provide an indication of the ability of DMARD treatment to alter bone turnover. This prospective study design considers a homogeneous group of RA patients, with active disease and known clinical characteristics, as they start conventional monotherapy for RA, without the confounding factors of immobility or steroid use, to assess the effect of disease control on bone markers and density. Our aim was to evaluate whether bone turnover markers could be used at earlier time points in the treatment to evaluate on-going bone loss and hence potential loss of BMD in the future. Methods Patients Forty patients who fulfilled the ACR criteria for RA were enrolled. Patients were assessed by a single investigator at three District General Hospitals in South East London, after approval of the protocol by local medical Ethical Committees. Formal consent was obtained in all patients. Patients with established RA were identified at the time the treating clinician felt disease activity merited change to a new DMARD, and patients with early disease enrolled at the time they started DMARDs for the first time. All patients had active disease demonstrated by three out of six or more tender joints, three or more swollen joints, 45 min morning stiffness and ESR >25 mmhg. Patients with physical incapacity (Steinbrocker III and IV) were excluded from the study. Other exclusion criteria were previous or present treatment with steroids or other treatment known to affect bone metabolism, such as bisphosphonates or calcium and vitamin D, or a concurrent disease which might affect bone metabolism. Peri-menopausal females within 5 yr of menopause were excluded, as this is a time of rapid change in bone turnover. No patients were taking HRT. At baseline patients were commenced on a DMARD considered appropriate by the treating physician. Seventeen were put on methotrexate, 10 on gold, 13 on sulphasalazine. Assessment Disease activity was assessed at 0, 3, 6, 12, 18 and 24 months by a visual analogue scale for pain on a 100 mm scale (VAS pain), patient s global assessment as a visual analogue score, early morning stiffness and Health Assessment Questionnaire (HAQ). A 28-point joint count was performed for number of tender and swollen joints. A disease activity score (DAS 28) w11x was calculated. Full blood count, ESR and liver function were measured at the assessments and as required by DMARD monitoring protocols. Imaging BMD of the spine and hip was assessed using a dual energy X-ray absorptiometer on a Hologic QDR 4500c at 0, 12 and 24 months. Results are expressed in gucm 2 and Z-scores. The long-term precision on this scanner over 7 yr was 1.12% for PA spine BMD and 2.21% for femoral neck BMD w12x. The results were compared with a normal database that had previously been shown not to differ from a south-east England population. As this study was conducted in the context of a service clinic, for logistical reasons, not all patients attended for three serial scans. Biochemical markers All patients had fasting second morning urine and blood samples taken before starting DMARDs and at 3, 6, 9, 12, 18 and 24 months. D-PYR and PYR cross-links were measured using ELISA assays. Results were expressed in nmolummol creatinine. Intra- and interassay coefficients of variation were less than 7 and 10%, respectively. Statistical analysis Results are expressed as mean and standard deviation. BMD and bone marker results were shown to be normally distributed by the Kolmogorov Smirnov test and were thus compared with baseline values using paired Student s t-tests. Continuous variables were compared by Pearson s correlation coefficient. The pre-determined analysis was to examine the relationship of disease activity (ESR, DAS, HAQ, VAS pain, patient assessment) with bone turnover markers at baseline and to explore the effect on these factors of a single intervention, i.e. treatment. The effect of early change in markers was related to BMD change at 1 yr to see if markers could be used to predict subsequent bone change in a clinical setting. A power calculation suggested 83% power to show change in BMD greater than the error of the measure.
3 Bone turnover in treated RA 1049 TABLE 1. Change in disease markers with treatment Months D-PYR (nmolummol creatinine) (S.D.) 8.4 (4.5) 7.24 (3.4) 7.2 (3.8) a 6.83 (2.5) b 7.36 (3.3) PYR (nmolummol creatinine) (S.D.) 56.4 (64) (43) (33) 37.6 (30) 34.2 (20) a DAS 5.5 (1.12) 4.66 (1.2) b 4.33 (1.3) b 3.38 (1.2) b 3.89 (1) b Early morning stiffness (86.3) 55.0 (66.7) c 43.5 (64.2) c (75.2) c 29 (42.2) c HAQ (12.2) 9.08 (5.8) 8.69 (6) b 7.48 (6.1) b 7.47 (5.6) a VAS pain 51.4 (19.3) 39.4 (20) c 35.2 (24.3) c 33.7 (23) b (20.2) c Patient global assessment 54.8 (21.5) (22) b 34.8 (23.7) c 36.6 (24.6) a 33.5 (25) a ESR (mmhg) 43.6 (27) 30.3 (19) a 25.5 (25) a 24.9 (18) b 26.8 (18) b a P<0.05; b P<0.01; c P<0.001 with respect to baseline. Results Forty patients were enrolled, mean age 59.9 (range 31 76), 26 females, 24 male. Twenty-five had established RA, 15 had RA for <2 yr and had not previously used a DMARD. Changes in measures of disease activity and bone markers are shown in Table 1. Mean baseline ESR was 43.6"27.4 mmhg, mean baseline DAS 5.5"1.1. As expected, baseline DAS was closely correlated with ESR (r=0.67, P<0.0001), since DAS includes ESR. As a result of treatment of the active RA with DMARDs both ESR and DAS fell and at 1 yr mean ESR was 26.8"18 mmhg and DAS 3.89"1, (P<0.01) (Table 1). HAQ improved significantly by 6 months and patients reported significant improvement in pain (VAS), early morning stiffness and patient assessment by 3 months, demonstrating the efficacy of the DMARD treatment. Bone markers Serial bone marker results were available on 40 patients. Baseline D-PYR was elevated (8.4"4.55 nmolummol creatinine) in the RA patient group compared with normal age- and sex-matched controls (4.91"1.1) P< D-PYR was significantly reduced by 6 months, 7.2"3.8 nmolummol creatinine (P<0.05), and at 9 months, 6.83"2.5 nmolummol creatinine (P<0.01), with respect to baseline values. At baseline, D-PYR levels correlated with ESR (r=0.6, P<0.001) and with DAS (r=0.4, P<0.05). HAQ, reflecting functional state, showed no relationship with bone resorption over time (r=0.007). Spearman rank order correlation showed ESR to be the most significant determinant of bone resorption (D-PYR) over 1 yr (r=0.43, P<0.001). PYR levels were raised at baseline and fell over 1 yr, but not by a significant degree (Table 1). Baseline PYR and D-PYR were only weakly correlated (r=0.64), reflecting their different sites of origin and a discordant TABLE 2. Change in BMD with treatment Months DEXA spine 0.98 (0.16) 0.97 (0.2) 0.97 (0.19) DEXA NOF 0.79 (0.16) 0.79 (0.16) ( 0.16) a P<0.05, b P<0.01, c P<0.001 with respect to baseline. effect of disease on bone and cartilage. The relationship of PYR to ESR was not significant (r=0.26). Imaging Serial bone density data was available on 21 patients (Table 2). There were eight patients with early rheumatoid and 13 patients with RA for longer than 2 yr prior to the study, who had DEXA scans. There was no difference in sub-analyses of these groups and therefore in view of the small numbers the data was analysed as a whole. At baseline, the spine Z-score for the group was ("1.57) and the neck of femur Z-score ("1.35). Overall, there was no significant change in BMD at either site over the 2-yr follow-up (Table 2). Eleven out of 21 patients showed an increase in spine BMD from baseline by 1 yr. There was an inverse relationship between the change in DAS over 0 3 months and the percent change in spine over 1 yr (r=20.52, P=0.05). There was no relationship between D-PYR at baseline and change in spine or hip BMD by 1 or 2 yr. The change in D-PYR over 0 3 months was not closely linked to the change in BMD at hip (r=20.25) or spine (r=20.36) at 1 yr, so this cannot be a useful predictor of bone mass in RA over short time periods. Discussion Long-term observational studies have been pessimistic as to the outcome of treatment in RA, suggesting that DMARDs fail to limit erosive damage and do not prevent consequent functional disability w13x. Such results are quoted as the reason for adopting aggressive combination therapies w14, 15x. However, these long-term studies inevitably reflect a time when DMARDs were not traditionally introduced until erosions were present and before the early use of drugs such as methotrexate. More recent studies have shown that currently used DMARDs such as methotrexate and leflunomide w16x can inhibit erosions within a short time, both in early and established disease. As such, the future outcome for RA may be more optimistic. Our study shows no change in bone mineral density in hip or spine over 2 yr of active treatment, suggesting that current monotherapy regimes may limit bone loss. Aggressive treatment in early disease has also been shown to limit disease progression w17x and reduce bone
4 1050 A. L. Dolan et al. turnover w18x, however, such early studies inevitably include patients with a naturally good prognosis. It is encouraging to see DMARDs alone can limit bone loss even in those with established disease and gives further support to the current strategy of changing DMARD to optimise disease control. The study design did not allow comparison with a group of patients given placebo for ethical reasons and hence it was not possible to follow the natural progression of bone turnover and bone density if treatment was not given. This study is relatively small and was performed during the course of routine follow-up of patients in a clinic; hence there was not complete serial BMD data for all those originally enrolled. A power calculation suggests that the failure to show a change in BMD in the 21 patients with all three serial scans is a valid conclusion rather than a type 2 error. However, there is a potential selection bias in that those failing to attend for all scans may be generally less compliant in other aspects of treatment and as such have a poorer prognosis. A larger study to address the issues raised in this clinical study would be warranted. Disease activity is a significant determinant of bone turnover in active RA. Gough w10x showed patients with active disease and higher cumulative CRP had greater bone loss, in early RA. In common with previous studies, we have found that bone resorption is increased in active RA and that disease activity, as shown by ESR and the DAS, correlates with markers of bone resorption (D-PYR). The novel finding in this study, is that in both established and early RA, intervention with DMARDs alone can itself reduce bone resorption, as demonstrated by the fall in bone markers seen here. There is an inverse relationship between early suppression of disease activity and preservation of BMD. ESR, over the 1 yr after commencing the current DMARD was the most significant determinant of bone resorption (D-PYR), confirming the importance of the inflammatory state in bone loss. The HAQ, reflecting functional state, showed no relationship with bone resorption over time. The effect of DMARDs in reducing bone loss is apparent in the bone markers by 6 months on treatment, long before densitometry could be expected to demonstrate a change. The mechanism by which the inflammatory cytokines from active rheumatoid joints affects bone turnover has recently been studied. Inflamed RA synovium releases inflammatory cytokines, such as IL1 and IL6, which can exert their effects on bone locally and throughout the skeleton w19, 20x, possibly by acting via complex signalling pathways, such as the newly identified osteoprotegerin pathways w21x. In normal trabecular bone the remodelling cycle is stimulated by the production of IL6 from osteoblasts, which stimulate osteoclast maturation, bringing about bone resorption w22x. In diseases with high levels of systemic inflammatory cytokines, the normal bone remodelling cycle may be over-ridden or uncoupled and systemic IL6 acts directly on osteoblasts. This is the case in myeloma w23x and has been suggested in polymyalgia rheumatica w24x. Elevation of bone resorption markers (PYR) without elevation of formation markers, suggests bone resorption is the dominant process in RA w18x. Early change in D-PYR in the first 3 months of DMARDs was not related to spine BMD at 1 yr. Therefore, it is unlikely that this marker can be used as a surrogate to predict future bone density in individuals. However, the observed fall in bone resorption overall is compatible with the preservation of BMD seen in this group. It is possible that the changes in bone resorption markers indicate reductions in bone turnover, which may influence qualities of bone microarchitecture, which are not measured in a bone density scan w25x. Clinical studies with raloxifene w26x and nasal calcitonin w27x indicate that bone density is not the final arbiter of future fracture risk, as these treatments prevent future fracture to a much greater degree than would be predicted by their effect on BMD. The concept of bone quality and turnover may be as important. If DMARDs reduce bone turnover, that in itself may be adequate to reduce the future fracture risk. Further studies are needed to address this hypothesis. Acknowledgement This study was supported by a grant from the National Osteoporosis Society. References 1. Sambrook PN. The skeleton in rheumatoid arthritis: common mechanisms for bone erosion and osteoporosis? J Rheumatol 2000;27: Joffe I, Epstein S. Osteoporosis associated with rheumatoid arthritis: pathogenesis and management. Semin Arthritis Rheum 1991;20: Spector TD, Hall GM, McCloskey EV, Kanis JA. Risk of vertebral fracture in women with rheumatoid arthritis. Br Med J 1993;306: Scott DL. Clinical guidelines for management. Baillieres Clin Rheumatol 1997;11: Emery P. Assessment of rheumatoid arthritis a clinician s viewpoint. J Rheumatol (Suppl.) 1994;42: Garnero P, Delmas PD. New developments in biochemical markers for osteoporosis. Calcif Tissue Int 1996;59 (Suppl. 1):S Seibel MJ, Duncan A, Robins SP. Urinary hydroxypyridinium crosslinks provide indices of cartilage and bone involvement in arthritic diseases. J Rheumatol 1989; 16: Spector TD, James IT, Hall GM, Thompson PW, Perrett D, Hart DJ. Increased levels of urinary collagen crosslinks in females with rheumatoid arthritis. Clin Rheumatol 1993;12: Gough AK, Peel NF, Eastell R, Holder RL, Lilley J, Emery P. Excretion of pyridinium crosslinks correlates with disease activity and appendicular bone loss in early rheumatoid arthritis. Ann Rheum Dis 1994;53: Gough AK, Lilley J, Eyre S, Holder RL, Emery P. Generalised bone loss in patients with early rheumatoid arthritis. Lancet 1994;344:23 7.
5 Bone turnover in treated RA van der Heijde DM, van t Hof M, van Riel PL, van de Putte LB. Development of a disease activity score based on judgment in clinical practice by rheumatologists. J Rheumatol 1993;20: Patel R, Blake GM, Rymer J, Fogelman I. Long-term precision of DXA scanning assessed over seven years in forty postmenopausal women. Osteoporos Int 2000; 11: Scott DL, Symmons DP, Coulton BL, Popert AJ. Longterm outcome of treating rheumatoid arthritis: results after 20 years. Lancet 1987;1: Proudman SM, Conaghan PG, Richardson C et al. Treatment of poor-prognosis early rheumatoid arthritis. A randomized study of treatment with methotrexate, cyclosporin A, and intraarticular corticosteroids compared with sulfasalazine alone. Arthritis Rheum 2000;43: Neva MH, Kauppi MJ, Kautiainen H et al. Combination drug therapy retards the development of rheumatoid atlantoaxial subluxations. FIN-RACo Trial Group. Arthritis Rheum 2000;43: Smolen JS, Kalden JR, Scott DL et al. Efficacy and safety of leflunomide compared with placebo and sulphasalazine in active rheumatoid arthritis: a double-blind, randomised, multicentre trial. European Leflunomide Study Group. Lancet 1999;353: Emery P. Therapeutic approaches for early rheumatoid arthritis. How early? How aggressive? Br J Rheumatol 1995;34(Suppl. 2): Gough A, Sambrook P, Devlin J et al. Osteoclastic activation is the principal mechanism leading to secondary osteoporosis in rheumatoid arthritis. J Rheumatol 1998; 25: Eastgate JA, Symons JA, Wood NC, Grinlinton FM, di Giovine FS, Duff GW. Correlation of plasma interleukin 1 levels with disease activity in rheumatoid arthritis. Lancet 1988;2: Gowen M, Wood DD, Ihrie EJ, McGuire MK, Russell RG. An interleukin 1 like factor stimulates bone resorption in vitro. Nature 1983;306: Hofbauer LC, Khosla S, Dunstan CR, Lacey DL, Boyle WJ, Riggs BL. The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J Bone Miner Res 2000;15: Manolagas SC, Jilka RL. Bone marrow, cytokines, and bone remodeling. Emerging insights into the pathophysiology of osteoporosis. N Engl J Med 1995;332: Kawano M, Hirano T, Matsuda T et al. Autocrine generation and requirement of BSF-2uIL-6 for human multiple myelomas. Nature 1988;332: Dolan AL, Moniz C, Dasgupta B et al. Effects of inflammation and treatment on bone turnover and bone mass in polymyalgia rheumatica. Arthritis Rheum 1997; 40: Riggs BL, Melton LJ 3rd. Bone turnover matters: the raloxifene treatment paradox of dramatic decreases in vertebral fractures without commensurate increases in bone density. J Bone Miner Res 2002;17: Ettinger B, Black DM, Mitlak BH et al. Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: results from a 3-year randomized clinical trial. Multiple Outcomes of Raloxifene Evaluation (MORE). J Am Med Assoc 1999; 282: Chesnut CH, Silverman S, Andriano K et al. A randomized trial of nasal spray salmon calcitonin in postmenopausal women with established osteoporosis: the Prevent Recurrence Of Osteoporotic Fractures study. PROOF Study Group. Am J Med 2000;109: