Background Psoriatic skin disease is a burdensome, sometimes painful, dermatologic condition which was reported to occur as an adverse event (AE) during TNF-inhibitor (TNFi) treatment of rheumatoid arthritis (RA). Single case reports revealed the occurrence of psoriasis also during treatment with non-TNFi, but the magnitude under those agents remains unclear.
Objectives To compare incidence rates of psoriasis in RA under treatment with different biologic and conventional synthetic (b/cs)DMARDs and to investigate risk factors.
Methods We used data of 12,722 patients (53,585 patient years (py)) enrolled with the start of a b/csDMARD in the German biologics register RABBIT. Patients were required to have no psoriasis at baseline and at least one follow-up. All psoriatic events (PsE) reported until 30 April 2016 were selected and assigned to treatments administered within 3 months prior to the event. Crude incidence rates (IR) of PsE were calculated per 1,000py. Cox regression was applied to investigate risk factors for the occurrence of PsE with and without inverse probability weights (IPW) to adjust for confounding by indication.
Results 96 PsE were reported, with only 6 of them categorized as being serious. The median time between enrollment in the cohort and onset of psoriasis was 19 months (IQR:11–45 months). 21 of all PsE (22%) were palmoplantar manifestations of which 9 were reported as pustular type.
Compared to csDMARD treatment with a crude IR of 0.44/1,000py (95% CI 0.2;0.9), the IRs found under TNFi (IR 2.99 (95% CI 2.3;3.8)) and abatacept (IR 3.99 (95% CI 1.7;7.9)) were significantly higher. In patients treated with rituximab (IR 1.8 (95% CI 0.8;3.4)) or tocilizumab (IR 0.7 (95% CI 0.1; 2.0)) IRs for PsE were not significantly different from csDMARD patients. Across TNFi, the IR varied insignificantly.
Adjusted regression analysis showed higher risk for PsE with TNFi, abatacept and rituximab (graph). Female sex (adjusted hazard ratio (HR) 1.8 (1.0;3.3)) and being rheumatoid factor negative (HR 1.6 (1.0;2.6)) were additional significant risk factors. Smoking (HR 1.6 (1.0; 2.5)), age (HR 1.0 (0.98;1.01)), glucocorticoids per 5 mg/d increase (HR 1.1 (1.0;1.2)), and prior (≤6months) skin infections (HR 2.2 (0.5;9.7)) were not significantly associated. Replacing glucocorticoids with DAS28 did not show differing results. Adjustment with IPW attenuated the effect of rheumatoid factor (p=0.4) but smoking was significantly associated with a higher risk (p<0.01).
Conclusions This is the first analysis comparing the incidence of psoriasis under biologics with different modes of action within one cohort. Our results confirmed a higher risk for TNFi 1 and showed a similar result for abatacept. A lower but still significant increased risk was found for rituximab, whereas there was no difference for tocilizumab compared to csDMARDs. New onset psoriasis is a rare and most often non-serious event. The number needed to harm is 334 patients treated with TNFi for one year to observe one PsE.
Acknowledgements RABBIT is supported by a joint, unconditional grant from AbbVie, BMS, Celltrion, MSD, Pfizer, Roche, Samsung and UCB.
Hernandez et al., Arthritis Care Res 2013; 65:2024–31.
Disclosure of Interest A. Strangfeld Speakers bureau: BMS, MSD, Pfizer, Roche, Sanofi-Aventis, L. Baganz: None declared, A. Richter Consultant for: Pfizer, B. Manger Consultant for: Abbvie, BMS, MSD, Pfizer, Roche, UCB, G.-R. Burmester Consultant for: AbbVie, BMS, MSD, Pfizer, Roche, UCB, C. Eisterhues: None declared, S. Wassenberg Consultant for: AbbVie, Pfizer, Novartis, Janssen, Roche-Chugai, Celltrion, BMS, Fuji, Speakers bureau: AbbVie, Celgene, Novartis, Pfizer, MSD, Lilly, Janssen, UCB, A. Zink Speakers bureau: AbbVie, BMS, MSD, Pfizer, Roche, UCB, J. Listing Consultant for: Sandoz, Pfizer