Rheumatology 1999;38:841–847 Relationship between soluble markers of immune activation and bone turnover in post-menopausal women with rheumatoid arthritis

Objective. Regarding interactions between pro-inflammatory cytokines and bone metabolism in rheumatoid arthritis (RA), we investigated the relationship between the serum levels of interleukin (IL)-1b, IL-6, soluble interleukin-2 receptor (sIL-2r), C-reactive protein (CRP), the vitamin D metabolites 25-hydroxyvitamin D3 (25OHD3) and 1,25-dihydroxyvitamin D3 [1,25(OH)2D3 ], intact parathyroid hormone (iPTH) as well as serum and urinary parameters of bone turnover in 74 post-menopausal women with RA. Results. SIL-2r correlated negatively with 1,25(OH )2D3 (P< 0.01), whereas IL-6 showed a positive correlation with urinary excretion of deoxypyridinoline–collagen cross-links (P< 0.01). 23 (P< 0.01) and iPTH (P< 0.01) were negatively related to CRP, whereas the urinary excretion of pyridinoline (P< 0.01) and deoxypyridinoline (P< 0.01)–collagen cross-links showed a positive correlation with CRP. 23 (P< 0.01) and iPTH (P< 0.05) were positively related to bone alkaline phosphatase as a marker of osteoblast function. Conclusion. Our data indicate that IL-6 is a critical determinant of increased bone resorption in post-menopausal RA women with high disease activity and that serum levels of 1,25(OH)2D3 are inversely related to T-cell activation. K  : Rheumatoid arthritis, Osteoporosis, Pro-inflammatory cytokines, Vitamin D, PTH. Systemic osteoporosis is a frequent and serious comtumour necrosis factor alpha (TNF-a) and IL-6 have been shown to act on osteoblasts and osteoclasts in vitro plication of rheumatoid arthritis (RA). Various pathogenetic mechanisms, such as chronic glucocorticoid and in vivo, resulting in an increase of bone resorption [8–14]. This bone-resorptive effect of pro-inflammatory treatment and immobilization, contribute to systemic bone loss in this disease [1]. On the other hand, the cytokines may influence parathyroid hormone (PTH ) secretion and 1,25(OH)2D3 synthesis. Furthermore, chronic inflammation process itself has been shown to be an important risk factor in the development of because the vitamin D receptor is expressed on activated T cells [6, 7], the activation state of T cells may be a systemic osteoporosis in RA [2–4]. However, the mediator mechanisms and the link between the local arthritic critical determinant of the 1,25(OH)2D3 serum level. Thus, interactions between secretion and metabolism of process and systemic bone loss are still unclear. In a previous study, we observed a decrease in serum calcium-regulating hormones and pro-inflammatory cytokines certainly have a role in the regulation of bone levels of 1,25-dihydroxyvitamin D3 [1,25(OH)2D3 ] and an increase in parameters of bone resorption (pyridinoturnover in chronic inflammatory diseases in general and especially in RA. line and deoxypyridinoline–collagen cross-links) with increasing disease activity in an unselected group of RA Therefore, the purpose of our study was to investigate whether there are relationships between serum levels of patients [5]. Therefore, we suggested that a disease activity-dependent decrease in 1,25(OH)2D3 may be an pro-inflammatory cytokines, soluble markers of T-cell activation, calcium-regulating hormones and boneimportant mediator mechanism between local inflammation and systemic bone loss in RA. 1,25(OH 23 has turnover parameters in RA patients which may be of particular importance regarding the pathogenesis of immunomodulatory functions, including the suppression of activated T cells and the modulation of the secretion osteoporosis in RA. of pro-inflammatory cytokines [6, 7]. On the other hand, the pro-inflammatory cytokines interleukin-1 (IL-1), Patients and methods Patients Submitted 28 September 1998; revised version accepted 6 April 1999. Because of the well-known influences of sex and Correspondence to: P. Oelzner, Department of Internal Medicine menopausal state on bone metabolism and secretion of IV, Friedrich-Schiller University of Jena, Erlanger Allee 101, 07740 Jena, Germany. pro-inflammatory cytokines, only post-menopausal RA © 1999 British Society for Rheumatology 841 P. Oelzner et al. 842 women without hormone replacement therapy were (Pyd/crea) and as the Dpyd/creatinine ratio (Dpyd/crea), respectively. included in our study. We investigated 74 postmenopausal women with RA aged between 49 and 80 yr Statistical analysis (mean 64.8 yr), and with a disease duration ranging from 6 months to 44 yr (mean 8.6 yr). All patients The data were analysed statistically using the SPSS for fulfilled at least four of the 1987 revised ACR criteria Windows Statistical Program. For correlation analysis, for RA [15]. With respect to the renal synthesis of we used the Spearman correlation coefficient. To com23 , only patients with normal serum creatinpare the parameters of the various patient groups, the ine levels were included in the study. Mann–Whitney U-test was used. Differences of P< 0.05 Fifty-one patients aged between 50 and 79 yr (mean were considered significant. 64.8 yr), and with a disease duration between 6 months and 44 yr (mean 9.1 yr), were receiving glucocorticoid treatment (5–15 mg prednisolone/day), whereas 23 Results patients aged between 49 and 80 yr (mean 64.9 yr), and Influence of therapy on parameters of disease activity with a disease duration between 6 months and 39 yr including serum cytokine levels and parameters of bone (mean 7.4 yr), were not. None of the patients were turnover administered i.m., intra-articular or i.v. steroids during the last 3 months before investigation. Twenty women According to therapy at the time of investigation, the post-menopausal women with RA were divided into were receiving disease-modifying anti-rheumatic drugs (DMARDs) (11 methotrexate, four azathioprine, four four groups: patients receiving neither glucocorticoids nor DMARDs, patients with glucocorticoid treatment sulphasalazine and one hydroxychloroquine). but without DMARDs, patients treated with glucocorLaboratory tests and analytical techniques ticoids and DMARDs and, as a subgroup of these, patients who received methotrexate (MTX ) (Table 2). The following parameters were investigated in the sera of patients: IL-1b, IL-6, soluble interleukin-2 receptor The ESR was lower in patients receiving glucocorticoids but no DMARDs (P< 0.05), in patients treated with (sIL-2r), C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), albumin, the vitamin D3 glucocorticoids and DMARDs (P< 0.01), and in the subgroup with MTX therapy (P< 0.01) compared to metabolites 25-hydroxyvitamin D3 (25OHD3) and 23 , intact PTH (iPTH), bone alkaline phospatients receiving neither glucocorticoids nor DMARDs. Patients treated with MTX had a lower ESR (P< 0.05) phatase (bALP), osteocalcin (OC), tartrate-resistant acid phosphatase (TRAP), calcium (Ca), calcium corand a higher serum albumin level (P< 0.05) than those treated with glucocorticoids only. sIL-2r was signifirected for albumin (Ca-alb.), and phosphate (PO4). Fasting venous blood samples were drawn between 7 cantly lower in patients treated with glucocorticoids than in those receiving neither glucocorticoids nor and 9 a.m. For the determination of vitamin D metabolites, iPTH and cytokines, serum aliquots were frozen DMARDs (P< 0.01). In patients treated with glucocorticoids alone (P< 0.05) and in those receiving glucocorand stored at −70°C until analysed. The measuring procedure, producer and intra-assay/interassay coeffiticoids and DMARDs (P< 0.01), we found a higher bALP activity than in patients receiving neither glucocients of variation of calcium-regulating hormones, cytokines and cytokine receptors are listed in Table 1. corticoids nor DMARDs. The serum concentration of PO4 was slightly lower in patients treated with MTX Furthermore, the excretion of pyridinoline (Pyd) and deoxypyridinoline (Dpyd) cross-links, as well as the than in those receiving neither glucocorticoids nor DMARDs (P< 0.05). The other parameters of bone calcium/creatinine ratio (Ca/crea) in the morning urine and the daily urinary calcium excretion, were measured. turnover, especially serum levels of vitamin D metabolites and iPTH, and the markers of bone resorption Pyd and Dpyd were determined in urine by a reversedphase HPLC method, described in detail by Müller et al. (Pyd/crea, Dpyd/crea and TRAP), were not different between the four groups. [16 ], and expressed as the Pyd/creatinine ratio T 1. Measuring procedure, producer and intra-assay/interassay coefficients of variation of calcium-regulating hormones, cytokines and