Background After failure of a first anti-TNF agent (aTNF), clinicians may choose to prescribe an alternative aTNF or switch to a ‘non-aTNF biologic antirheumatic agent’ (non-aTNF-BI: abatacept, tocilizumab or rituximab (RTX)). In this setting, it has been previously shown that non-aTNF-BIO have a better overall drug retention compared to alternative aTNF,(1) but no direct comparison of their respective effectiveness and safety has been made.
Objectives To compare drug retention rates of non-aTNF-BIO with alternative aTNF, as groups, prescribed in second or third intention and to compare specifically drug discontinuation due to ineffectiveness or adverse events (AEs). Secondly, analyze these outcomes for the individual non-aTNF-BIO biologic agents.
Methods This longitudinal, population-based cohort study (SCQM-RA) analyzed all patients treated with an alternative biological, after a 1st inadequate response to an aTNF agent (ineffectiveness in 50%, AE in 33%). The primary end point is drug discontinuation rate for (i) ineffectiveness and (ii) intolerance (AEs). Time to discontinuation was defined as the time between drug initiation and last administration plus 1 dispensation interval. Because RTX may be administered ‘on demand’ with flexible intervals, we defined RTX discontinuation as a major alteration of the antirheumatic regimen after RTX initiation or the time of last infusion plus six months, whatever happened first. Drug discontinuation was analyzed using a Cox proportional hazards model, adjusting for potential confounders and correcting for multiple failures within the same individual (Anderson-Gill).
Results We identified 2391 biologic treatment courses in aTNF inadequate responders (1,199 with an alternative aTNF and 1,192 with a non-aTNF-BIO) and 1203 treatment discontinuations. A 2nd biological was administrated 1507 times, a 3rd - 587, a 4th or 5th - 297 times, contributing a total of 3554 patient-years on biological agents. No major differences existed in baseline characteristics, except for a larger number of previous biologic failures (median of 2 versus 1, p < 0.001) and slightly older ages, longer disease durations in the non-aTNF-BIO group.
The overall drug discontinuation was higher in the aTNF group than in the non-aTNF-BIO group (Hazard Ratio (HR) for aTNF: 1.96 [95% CI: 1.71-2.26]). Drug discontinuation for ineffectiveness (N=710, HR for aTNF: 1.74 [95% CI: 1.46 – 2.08]) and for AEs (N=279, HR for aTNF: 2.09 [95% CI: 1.58 – 2.77]) were also higher with alternative aTNF than with non-aTNF-BIOs. While a trend to lower drug discontinuation was observed for all agents within the non-aTNF-BIO group, the effect was significantly stronger with RTX (HR for ineffectiveness: 0.44 [99% CI: 0.33 – 0.60); HR for AEs: 0.34 [99% CI: 0.20 – 0.57]), which may be partially explained by a completely different approach of administration (‘on demand’ with variable intervals, med 11.5 months [IQR: 8.6 – 16.2]).
Conclusions In patients having experienced an inadequate response to a previous aTNF agent, biologic agents with a different mode of action appear to have significantly lower drug discontinuation rates than alternative aTNF agents, for both ineffectiveness and for AEs.
Ann Rheum Dis. 2012;71:997-9
Disclosure of Interest S. Martin-du-Pan: None Declared, D. Neto Grant/research support from: Unrestricted research grant from BMS, P. Zufferey: None Declared, A. Ciurea: None Declared, H. Ziswiler: None Declared, C. Gabay: None Declared, A. Finckh Grant/research support from: Unrestricted research grant from BMS