Background There has been increasing interest on the effect of smoking in ankylosing spondylitis (AS). Studies suggest smoking is associated with higher incidence of AS, worse disease activity, structural damage, functional status and quality of life.1,2 Rheumatoid arthritis patients who smoke are reported to have reduced response to anti-TNF therapy (TNFi),3 although this has not been explored in AS.
Objectives To study the effect of smoking on response to TNFi in AS.
Methods AS patients at Aintree University Hospital, Liverpool, UK were entered into an audit database upon commencing TNFi from 2003 to 2014. Demographics, smoking status and pack years were recorded. Follow-up were aimed at 3 and 6 months, dependent on clinic availability and patient cancellation. Response was defined as reduction of BASDAI by 2 or 50% and spinal pain visual analogue scale (VAS) by 2. Survival analysis was performed using time to first recorded response. <20 pack years was considered light and ≥20 heavy smoking. All analyses were adjusted for age at starting TNFi.
Results 118 patients fulfilling the modified New York criteria for AS were included. 87 (73.7%) were male and 44/55 (80%) were HLA B27 positive. Mean (±sd) pre-TNFi BASDAI was 7.4±1.4 and VAS 7.8±1.5. Median time to first follow-up was 19 weeks (IQR 14, 28). Initial TNFi exposures were adalimumab 39.0%, etanercept 37.3%, golimumab 22.9% and infliximab 0.9%. 39.8% were current, 15.3% previous and 44.9% non-smokers. Of those that ever smoked 47.7% were heavy and 52.3% light smokers. Pre-TNFi disease activity and time to follow-up did not differ by smoking categories above or by gender.
Cox proportional hazard analysis showed trends toward ever-smoking being associated with reduced response (HRadj 0.70; 95%CI 0.46, 1.06), which was significant in males (HRadj 0.52; CI 0.32, 0.85). Figure 1 shows the survival estimate for male ever-smokers compared to non-smokers. 82.8% of light/non-smokers responded to TNFi compared to 61.3% in heavy smokers (P=0.015).
Conclusions In this small preliminary study, ever-smoking in males was associated with delayed response to TNFi. However this real life cohort had a high rate of non-attendance. Response could only be assessed at follow-up appointments which impact on time to response. Results suggest that ever-smokers may require longer follow-up in assessing treatment response. This hypothesis needs to be tested in larger cohorts with more frequent follow-up.
Chung HY et al. Smokers in early axial spondyloarthritis have earlier disease onset, more disease activity, inflammation and damage, and poorer function and health-related quality of life: results from the DESIR cohort. ARD 2012;71:809-16.
Videm V et al. Current smoking is associated with incident ankylosing spondylitis - the HUNT population-based Norwegian health study. J Rheumatol 2014;41:2041-8.
Abhishek A et al. Anti-TNF-alpha agents are less effective for the treatment of rheumatoid arthritis in current smokers. J. Clin Rheumat 2010;16:15-8.
Disclosure of Interest None declared