Article Text

Download PDFPDF

Bone mineral density improvement in spondyloarthropathy after treatment with etanercept
  1. H Marzo-Ortega1,
  2. D McGonagle1,2,
  3. G Haugeberg1,3,
  4. M J Green1,
  5. S P Stewart4,
  6. P Emery1
  1. 1Rheumatology and Rehabilitation Research Unit, The Leeds General Infirmary, Great George Street, Leeds LS1 3EX, UK
  2. 2Department of Rheumatology, Calderdale General Hospital, Salterhebble, Halifax HX3 0PW, UK
  3. 3Oslo City Department of Rheumatology, Diakonhjemmet Hospital, PO Box 23, Vinderen, Oslo N-0318, Norway
  4. 4Academic Unit of Medical Physics, The University of Leeds, Leeds LS1 3EX, UK
  1. Correspondence to:
    Dr H Marzo-Ortega;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

New bone formation with spinal fusion is the hallmark of ankylosing spondylitis (AS) and the related spondyloarthropathies (SpA), although concomitant osteoporosis is also a major problem both in early and established disease and correlates with disease activity.1 We have previously reported the efficacy of etanercept in patients with active and resistant spinal and peripheral SpA.2 Our aim in this study was to investigate whether suppression of inflammation with etanercept prevents bone loss in patients with AS and SpA.


Ten patients with active, resistant spinal and peripheral SpA were treated with a six month course of etanercept 25 mg subcutaneously twice weekly, as previously reported.2 Diagnoses in this group were: AS (n=7), Crohn’s spondylitis (n=2), and undifferentiated SpA (n=1). Results were compared with those for a second group of patients with equivalent disease activity, but shorter disease duration treated conventionally (table 1). The diagnoses of the second group were: undifferentiated SpA (n=6), reactive arthritis (n=3), psoriatic arthritis (n=1). Their treatment during this time was with non-steroidal anti-inflammatory drugs (n=8) and sulfasalazine (n=2). In addition, one patient received an intramuscular injection of corticosteroid at baseline and another patient two intra-articular steroid injections in two peripheral joints (wrist and knee). Hip (femoral neck and total hip) and spine (L2-L4) bone mineral density (BMD) was measured at baseline and after 24 weeks by one technician using the same dual x ray absorptiometry equipment (Lunar Expert, Madison, Wisconsin). Short term in vivo precision was 1.4% at the total hip, 2.9% at the femoral neck, and 2.4% at the spine. The long term spine phantom precision for the whole study period was 0.8%. All analyses were performed with SPSS (Statistical Package for Social Sciences) program 9.0 (SPSS, Chicago, IL). Paired and independent two tailed Student’s t test and Pearson’s χ2 test were used when appropriate.

Table 1

Demographic and disease characteristics at baseline and at six month follow up in both groups. Unless otherwise indicated, values are the mean (SD)

After six months all measures of disease activity were significantly reduced in the patients treated with etanercept but not in the controls (table 1). Figure 1 shows that the mean BMD increased at the hip and spine during this period in the first group, but decreased in the second (total hip: +1.6% v −1.3% (p=0.03); femoral neck: +0.2% v −1.5% (p=0.34); spine L2–4: +1.1% v −1.4% (p=0.19).

Figure 1

Percentage change in BMD in patients with SpA treated with etanercept compared with controls over 24 weeks.


Bone loss secondary to active inflammation is a serious long term complication of AS and SpA and can occur early in disease.3 Here we have shown that disease measurement variables were reduced significantly during the study only in the patients treated with anti-tumour necrosis (anti-TNF) factor, and this was associated with a prompt gain in BMD. Although this is a small and non-homogeneous group we believe that both groups are representative of the wide spectrum encompassed within the SpA. Also, the study was not randomised and the patients in the conventional treatment group had a shorter disease duration but, nevertheless, they were contemporaneous and of similar disease activity. Furthermore, previous studies of longer disease duration1 have shown an equivalent loss of BMD to that of the patients in this group. These findings suggest that adequate suppression of inflammation in AS and SpA is likely to improve BMD in both early and established disease.

Etanercept may improve BMD by removing excess cytokine from both osteoclasts and osteoblasts. It has been shown in rheumatoid arthritis that TNFα blockade also reduces interleukin 1 and interleukin 6, both of which are known to be potent pro-resorptive cytokines acting via the osteoprotegerin/RANKL pathway.4 In addition, it has recently been shown, that the differentiation, activation, and survival of osteoclasts are facilitated by TNFα.5 TNFα also inhibits the differentiation of osteoblasts from progenitor cells and retards bone formation by differentiated cells. The absence of supportive data on conventional disease modifying antirheumatic drugs in the prevention of osteoporosis in AS and SpA may relate to their inability to adequately suppress TNFα production, although the primary pathogenic pathway of osteoporosis in SpA is still to be determined.

In summary, this report shows that TNF blockade with etanercept may improve BMD in subjects with active and resistant SpA. These findings have implications for the long term treatment of patients with SpA with TNF blocking agents and need to be confirmed in larger cohorts in randomised controlled trials.


Dr McGonagle’s work is sponsored by the Medical Research Council (MRC). Professor Paul Emery is an Arthritis Research Campaign (ARC) professor in rheumatology.