Osteoporosis is a common feature of ankylosing spondylitis (AS), and vertebral fractures are an increasingly recognised complication. A cumulative fracture prevalence of between 9.5% and 18% has been reported, with a six- to eightfold relative increased risk of vertebral fracture.^1–3 Osteoporosis and new bone formation (syndesmophytosis) suggest that disordered bone turnover has a role in disease pathogenesis in AS. Bisphosphonates accumulate at sites of increased bone turnover, and inhibit bone resorption by inducing osteoclast apoptosis,⁴ thereby improving bone density and reducing fracture rates.⁵ Pulse pamidronate has recently been used with clinical efficacy in the treatment of AS.^6,7 Biochemical markers of bone turnover have been used to monitor response to treatment in postmenopausal osteoporosis.⁸

The effect of bisphosphonate treatment on biochemical bone turnover markers has not previously been studied in AS. We aimed at studying the efficacy of pulse pamidronate treatment in severe AS, and at determining its effect on biochemical bone turnover markers.

METHODS AND RESULTS

Patients with severe AS were treated with 6 monthly intravenous pulses of pamidronate, receiving 30 mg for the first infusion and 60 mg subsequently. The Bath AS Disease Activity Index (BASDAI) and the Bath AS Metrology Index (BASMI) scores were recorded, and C reactive protein (CRP) and erythrocyte sedimentation rate (ESR) measured at each visit. Fifteen patients participated (13 male, mean age 44). Mean disease duration was 14.8 years. All patients were receiving non-steroidal anti-inflammatory drugs; two were also receiving sulfasalazine, one methotrexate, and one azathioprine. No patients were currently receiving an oral bisphosphonate. Four patients had a history of uveitis, two psoriasis, two inflammatory bowel disease, and eight peripheral arthritis.

Fasting blood samples were taken monthly for measurement of biochemical bone turnover markers, according to the manufacturers’ instructions. Degradation products of type I collagen C-terminal telopeptides were measured with the serum crosslaps enzyme linked immunosorbent assay (ELISA; Nordic Bioscience Diagnostics A/S, Herlev, Denmark). Serum bone GLA protein was measured using the N-MID osteocalcin ELISA kit (Nordic Bioscience Diagnostics A/S, Herlev, Denmark). Bone-specific alkaline phosphatase was measured with the Access Ostase assay (Beckman Coulter Inc, Fullerton, California, USA). Non-parametric analyses (Wilcoxon signed rank tests) on an intention to treat basis were used to analyse the data.

Three patients did not complete the study (arthralgia/headache, back pain/nausea, and myalgia, respectively). The mean total dose of pamidronate received was 277 mg. The initial median serum crosslaps was 1845.0 ng/ml. Mean (SD) population serum crosslaps concentrations are 506 (255) ng/l for postmenopausal women, 321 (155) ng/l for premenopausal women, and 332 (190) ng/l for men. The initial median ostase concentration was 12.4 μg/l. Mean (SD) population ostase concentrations are 12.3 (4.3) μg/l for men, 8.7 (2.9) μg/l for premenopausal women, and 13.2 (4.7) μg/l for postmenopausal women. The initial median osteocalcin concentration was 19.5 ng/ml. Mean (SD) population osteocalcin concentrations are 17.9 (6.5) ng/ml for premenopausal women, 28.4 (9.5) ng/ml for postmenopausal women, and 21.4 (9.1) ng/ml for men.

Median serum crosslaps fell from 1845.0 to 556.5 ngl/l (Z = −3.29, p = 0.001) (fig 1). Median serum osteocalcin fell from 19.5 to 16.2 ng/ml (Z = −2.34, p = 0.02). Median serum ostase fell from 12.4 to 9.6 μg/l (Z = −3.11, p = 0.02). The median BASDAI improved from 6.80 to 5.75 (Z = −1.98, p = 0.048), but there was no significant improvement in the median BASMI (initial 8.00 v final 7.50, Z = −1.64, p = 0.10). There were non-significant trends of reduction in median CRP (initial 36.4 v final 26.6 mg/l, Z = −1.89, p = 0.06) and median ESR (initial 38 v final 29 mm/1st h, Z = −0.71, p = 0.48). There were no significant correlations between clinical measures and bone turnover markers.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

 Median serum crosslaps values.

DISCUSSION

Pulse pamidronate treatment significantly reduced all three biochemical bone turnover markers. This was particularly marked for serum crosslaps, the bone resorptive marker, where there was a 69.8% relative reduction. Such marked reduction of the crosslaps concentration suggests that bisphosphonates have a role in the management of the osteoporosis of AS. Ultimately, studies assessing fracture rates in patients with AS receiving bisphosphonates would be of great interest. One possibility is that by reducing the rate of new bone formation, bisphosphonates may also reduce syndesmophyte formation. However, longer term controlled studies in patients with early disease are needed to examine this subject.

Pulse pamidronate treatment also had a small beneficial effect on disease activity as measured by the BASDAI, which has also been demonstrated in a recent randomised controlled study of pamidronate treatment in AS.⁷ The patients in our study had established, severe disease. All but one would have met recent ASAS criteria for the use of anti-tumour necrosis factor drugs in AS.⁹ Clinical studies in earlier, less severe disease are required to determine if bisphosphonates may be of benefit to these patients.