Background In inflammatory diseases, including ankylosing spondylitis (AS), systemic inflammation induces secondary dyslipidaemia with lower total cholesterol (TC) and lower high density cholesterol (HDL-C) levels. Effective anti-inflammatory treatment with tumor necrosis factor (TNF) alpha-blocking therapy has been shown to increase lipid levels, which as a result, may affect cardiovascular (CV) risk. It is still unclear whether lipid changes following TNF-alpha blocking therapy are due to suppressed inflammation, or due to a specific effect of TNF-alpha blocking therapy.

Objectives We investigated the effects of changing inflammation levels during treatment with TNF-alpha blocking therapy on the lipid profile in AS patients.

Methods 230 consecutive AS patients with an indication for TNF-alpha blocking therapy with etanercept or adalimumab were enrolled. Data was collected at baseline and after 52 weeks of treatment. Serum C-reactive protein (CRP) was measured at each visit. High inflammatory status was defined as CRP≥10mg/L. Non-fasting lipid samples were collected at baseline and at 52 weeks.

Results CRP decreased significantly during treatment from 8 (3-22) to 2 (1-6) mg/l (p<0.01). TC, HDL-C and low density lipoprotein cholesterol (LDL-C) increased significantly with 4.6%, 3.7%, and 4.3% respectively. Apolipoprotein A-1 increased with 5.3%, while apolipoprotein B did not change. The TC/HDL-C ratio was not significantly changed after 52 weeks of TNF-alpha blocking therapy.

Regression analyses yielded an inverse association between changes in CRP and changes in TC (+0.104 mmol/l per 10mg/l reduction in CRP) and HDL-C (+0.024mmol/l per 10mg/l reduction in CRP) but not TC/HDL-C ratio. Significant changes in TC (+8.2%) and HDL-C (+8.3%) levels were only seen in patients whom CRP levels decreased during treatment from ≥10 mmol/l to <10mmol/l, but again, the TC/HDL-C ratio did not change.

Conclusions TNF-alpha blocking therapy is associated with a modest, but broadly parallel increase in TC, LDL-C, and HDL-C that might affect CV risk. Also, our data show, for the first time, that lipid changes following TNF-alpha blocking therapy are mostly due to suppressing inflammation and not to a specific TNF-alpha blocking therapy effect. Finally, consistent with previous findings, our data illustrate that the TC/HDL-C ratio is not appreciably altered by TNF-alpha blocking therapy and is therefore currently the most appropriate marker to determine CV risk in patients with an inflammatory condition.

Disclosure of Interest None declared