Background/Purpose: The influence of anti-TNF therapy on radiographic progression in ankylosing spondylitis (AS) is not well established. We studied this effect on radiographic progression in AS patients.

Methods: Three-hundred-and-seventy-six patients were included with at least 2 sets of radiograph at least 1.5 years apart (range 1.5 to 9 years) from a multicenter longitudinal cohort of 1600 AS patients.  Patients with total spinal ankylosis at baseline were excluded. Patients were assessed at least annually on a protocol format. Time-averaged BASDAI, ESR and CRP, NSAID and anti-TNF indices were calculated. Full recommended dose of NSAID/anti-TNF taken for the entire period between x-rays were assigned a score of 100. Total anti-TNF exposure was calculated from the anti-TNF index and the years of exposure. Patients with mSASSS increase at a rate of ≥ 1 unit/year were considered progressors. T-test, Chi-square and logistic regression were performed. For multivariate analysis, variables were selected from those that were significant in the univariate analysis. Variables were separated in different models based on the correlation matrix.

Results: The mean age of patients was 40.8 ± 13.8 years (75% males and 92% HLA-B27 positive) and the mean disease duration was 16.5 ± 12.9 years. Among patients enrolled, 60% had never smoked and 17% stopped smoking in the period between radiographs. No baseline radiographic abnormality was seen in 44% patients and 35% showed progression ≥ 1 mSASSS unit/year.

In the univariate analysis, males progressed faster (1.2 vs 0.8 mSASSS unit/year) with an odds ratio (OR) of 1.9 (95% CI: 1.09-3.39; p=0.02). Smokers had an odds of progression of 1.6 (95% CI: 1.02-2.54; p=0.04). Other significant variables included age of onset, baseline and time-averaged CRP and ESR and cumulative dose of anti-TNF received (Table 1A). The NSAID index was not significant in the univariate analysis. In multivariate analysis (Table 1B), the following variables remained significant: Gender (β=3.7;p=0.003), baseline ESR (β=1.03;p=0.001), baseline mSASSS (β=1.05;p=2×10^-5) and cumulative anti-TNF exposure (β=0.99;p=0.006). To estimate the effect with any exposure to biologics, cumulative anti-TNF exposure was substituted with the dichotomous variable: history of anti-TNF exposure. Anti-TNF use was protective against progression with an OR of 0.4 (p=0.006). The other variables remained significant as in the previous model.

Conclusion: Inflammation is associated with radiographic progression. Anti-TNF therapy slows the rate of radiographic progression in AS.

Table 1: Univariate and Multivariate Regression Analysis for radiographic progression

A. Univariate Regression
Covariate	p value	Odds Ratio	Confidence Interval
Male Gender	0.02	1.93	1.10 – 3.39
Age of Onset	0.004	1.03	1.01 – 1.06
Disease Duration	2 x 10-5	1.04	1.02 – 1.06
Smoking History	0.04	1.61	1.02 – 2.54
Baseline CRP	0.004	1.03	1.01 – 1.05
Time-Averaged CRP	0.002	1.05	1.02 – 1.08
Baseline ESR	0.0001	1.03	1.01 – 1.04
Time-Averaged ESR	0.006	1.02	1.01 – 1.04
Baseline mSASSS	1 X 10-11	1.07	1.05 – 1.09
Total Biologic Exposure	0.04	0.99	0.99 – 1.00
B. Multivariate Regression
Model 1 Variables included: Gender, Disease Duration, Baseline ESR, Baseline mSASSS & Total Anti-TNF Exposure
Male Gender	0.003	3.70	1.57 – 8.70
Baseline ESR	0.001	1.03	1.01 – 1.04
Baseline mSASSS	2x 10-5	1.05	1.03 – 1.08
Total Anti-TNF Exposure	0.006	0.998	0.997 – 0.999
Model 2 Variables included: Gender, Age of Onset, Baseline ESR, Baseline mSASSS & Total Anti-TNF Exposure
Male Gender	0.01	2.7	1.3 – 5.9
Baseline ESR	0.0004	1.03	1.01 – 1.04
Baseline mSASSS	1 x 10-8	1.06	1.040 – 1.09
Total Anti-TNF Exposure	0.04	0.99	0.99 – 1.000
Model 3 Variables included: Gender, Disease Duration, Baseline ESR, Baseline mSASSS & Anti-TNF (Y/N)
Male Gender	0.01	2.6	1.2 – 5.4
Baseline ESR	5 X 10-5	1.03	1.01 – 1.04
Baseline mSASSS	3x 10-9	1.07	1.04 – 1.09
Anti-TNF Use (Yes/No)	0.006	0.44	0.24– 0.79
Age of onset and Disease duration were strongly correlated in the matrix (R=0.98) and so they were included in separate models. Only one inflammatory parameter (Baseline ESR) was used for the same reason. Model 3 was used to calculate the Odds of progression in patients who took anti-TNF at some time compared to those with no exposure.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/anti-tnf-therapy-slows-radiographic-progression-of-ankylosing-spondylitis/