Tumour necrosis factor blocking agents such as infliximab have proved to be effective in patients with ankylosing spondylitis as up to 60–70% of the patients meet the 20% response criteria of assessment in ankylosing spondylitis (ASAS).^1,2 However, it cannot be explained why 30% of patients fail to respond and develop adverse reactions.

In rheumatoid arthritis, inefficacy to infliximab was associated with low serum trough infliximab levels and the presence of antibodies to infliximab (ATI).³

This study was designed to identify whether infliximab levels and ATI predict clinical inefficacy and adverse events in ankylosing spondylitis.

Eight patients with active ankylosing spondylitis (fulfilling the 1984 modified New York Criteria⁴) were treated according to the international ASAS consensus statement,⁵ with infliximab 5 mg/kg given intravenously at baseline, weeks 2, 6, and 12, and every 6 weeks thereafter. Sera were collected at 12 and 24 weeks before infusion.

At every visit, questionnaires (eg, Bath Ankylosing Spondylitis Disease Activity Index) to assess ASAS 20% response were obtained and routine laboratory tests were performed. These data were correlated with disease activity (ASAS 20% response), serum trough infliximab levels and antibody levels.

All patients were men, with a median (range) age of 47 (24–52) years, and were human lymphocyte antigen B27 positive, with a median (range) disease duration of 11 (1–28) years (table 1). Patient 1 was concomitantly treated with 15 mg methotrexate weekly and patient 3 was treated with cyclosporine and sulfasalazine.

Most patients responded well to infliximab with a considerable decline in Bath Ankylosing Spondylitis Disease Activity Index, erythrocyte sedimentation rate and C reactive protein, high serum trough levels of infliximab and no development of ATI. However, two non-responders did not show detectable serum trough infliximab levels and developed ATI after, respectively, 12 and 24 weeks. Patient 3 did not respond to treatment at all, whereas patient 5 met the ASAS 20% response criteria but had an increase in erythrocyte sedimentation rate and C reactive protein levels. Both patients developed an infusion reaction to infliximab.

In this study on eight patients with ankylosing spondylitis, a correlation between efficacy of infliximab and high levels of serum trough infliximab was shown. In 25% of these patients with ankylosing spondylitis ATI developed within 24 weeks in association with undetectable serum trough infliximab levels, inefficacy of infliximab and infusion reactions. The number of patients, however, is too small to draw definite conclusions, but interestingly, these data point in the same direction as described previously in rheumatoid arthritis.³ Lower serum trough infliximab levels could be explained by enhanced clearance because of immune complex formation between anti-infliximab antibodies and infliximab. To prevent ATI formation that might inhibit the efficacy of infliximab, it might be helpful to increase the dosage of infliximab (as occurs in treatment of rheumatoid arthritis with infliximab), to shorten the interval between infliximab infusions (as is currently the strategy in Crohn’s disease) or to provide coadministration of other immunosuppressives (such as methotrexate). These data should be confirmed in a larger group of patients with ankylosing spondylitis to develop a more patient-specific treatment, which might predict the inefficacy of infliximab at an early stage and might prevent adverse reactions.