A majority of the patients registered in BIOBADASER were diagnosed with RA (n = 4,006; 68.46%). SpA accounts for 26.04% (n = 1,524), AS 43% (n = 657), and PsA 37% (n = 570). The characteristics of patients are summarised in Table .
Characteristics of patients with spondylarthritis and rheumatoid arthritis in BIOBADASER and medications used by diagnosis
Compared with patients with RA, patients with SpA were younger (p < 0.001) and more frequently male (p < 0.001). Adalimumab was rarely used in SpA (15 treatments [0.90%] versus 577 [12.72%] in RA; p < 0.001).
The total exposure to TNF antagonists was 2,430 patient-years for SpA and 7,865 for RA. A detailed exposure rate by biologic and diagnosis is shown in Table . As expected, patients with RA not only had larger exposure in terms of patient-years, but also had used different biologic treatments more frequently. Adalimumab contributed little to the exposure to biologics in our population.
Exposure to the different biologic response modifiers depending on diagnosis in BIOBADASER
Survival of TNF antagonists in SpA is significantly greater than in RA at 1, 2, and 3 years (Table ), and the difference seems even larger with prolonged exposures (Figure ). As shown in Table and Figure , there are no significant differences in drug survival in the different types of SpA, although the group of "other SpA," which includes reactive arthritis, juvenile SpA, and chronic seronegative oligoarthritis, has lower drug survival (yet due to the small size of the group did not reach statistical significance [n = 42]).
Drug survival at one, two, and three years by diagnosis
Survival of tumor necrosis factor antagonists in patients with spondylarthritis (SpA) and rheumatoid arthritis
The better survival in SpA is expressed by a hazard ratio (HR) for discontinuation of 0.66 (95% confidence interval [CI], 0.58–0.76) compared with RA. There are unevenly distributed factors in SpA and RA (Table ) that had an impact on discontinuation: being older than 60 (HR = 1.21 [95% CI, 1.08–1.36]), being female (HR = 1.27 [95% CI, 1.13–1.43]), and using infliximab (HR = 1.53 [95% CI, 1.31–1.78]). After adjustment for these three factors, the HR for discontinuation in SpA was 0.66 [95% CI, 0.57–0.76] compared with RA. Survival of infliximab as first medication in SpA was significantly better than in RA (p = 0.0065). Similarly, survival of etanercept as first medication in SpA was significantly better than in RA (p = 0.0439). There were no differences however, when survival of infliximab or etanercept as second medication in RA was compared with their survival as second drug in SpA.
The distribution and the causes for discontinuation do not differ significantly in SpA and RA (pchi-square
= 0.131). They are AE (45.4% versus 48.7%), lack of efficacy (34.6% versus 36.0%), and others (19.9% versus 15.3%). The type of infection was also distributed equally in SpA and RA. Herpes zoster virus and Mycobacterium tuberculosis
were the leading identified microorganisms causing infection. Recommendations for screening and treatment of latent tuberculosis were put in place in March 2002 [35
], and their impact has been reported elsewhere [37
The HR for discontinuation due to AEs does not differ significantly among the types of SpA. On the contrary, the HR for discontinuation due to lack of efficacy is statistically greater for chronic seronegative arthritis than for the other types of SpA (HR = 6.7 [95% CI, 1.6–28.2]). Compared with patients with RA, patients with SpA treated with TNF antagonists have fewer AEs (17% versus 26%, respectively, had one or more AEs) that occur at a lower rate (13 events [95% CI, 11–14] in SpA versus 17 events per 1,000 patient-years in RA [95% CI, 16–18]). The incidence risk ratioSpA versus RA is 0.78 [95% CI, 0.68–0.89]). The HR of presenting an AE in SpA compared with RA is 0.69 [95% CI, 0.61–0.78]. However, as confirmed by the confidence intervals in Table , there were no differences in the rate of any particular AE. For the purpose of this study, discontinuation for AEs was considered when infusion or injection treatment after the event was not received within fourfold of the expected therapeutic interval. Under this definition, treatment was discontinued in 553 of 1,388 (39.8%) AEs in RA and in 123 of 329 (37.4%) in SpA (p = 0.412). Furthermore, the proportion of AEs that ended in hospitalisation or a prolonged hospital stay was 26.4% in RA and 21.3% in SpA (p = 0.056). Rate of AEs associated with death was 3.9% in RA and 1.5% in SpA (p = 0.034).
Types of adverse events occurring in patients with spondylarthritis and rheumatoid arthritis treated with biologics
After adjustment for being older than 60 years (HR = 1.51 [95% CI, 1.37–1.67]), having a disease duration longer than three years prior to therapy with TNF antagonist (HR = 1.29 [95% CI, 1.12–1.48]), and using infliximab (HR = 2.52 [95% CI, 2.11–2.99]), the HR for an AE in SpA versus RA was 0.80 [95% CI, 0.70–0.91]. Adjustment for age as a continuous variable did not produce different results than adjustment for age as a dichotomized variable (older than 60 years).