Maxime Verhoeven

85 TCZ vs. TCZ+MTX; prevention of radiographic progression in RA Early RA Overall, in early RA patients, TCZ-monotherapy resulted in less prevention of radiographic progression (total SvdH scores) than TCZ+MTX combination therapy (RR 0.96, 95%CI 0.90 – 1.03). However, these effects were modified by baseline joint damage (TCZ vs. TCZ+MTX [p<0.01]), and DAS28 (TCZ vs. TCZ+MTX [p=0.04]), see Supplementary Table S1. In subgroups, the advantage of TCZ+MTX over TCZ on total SvdH scores disappeared with a high-level of baseline joint damage (RR for preventing progression of TCZ versus TCZ+MTX 1.02, 95%CI 0.87 – 1.18, versus RR 0.91, 95%CI 0.81 – 1.02 in the subgroup with a low-level of baseline damage) or with a low-level of DAS28 at baseline (RR 1.04, 95%CI 0.93 – 1.17 versus RR 0.92, 95%CI 0.83 – 1.03 in the subgroup with a high-level of baseline DAS28), see Figure 1 and Supplementary Table S2. Translating these results to absolute differences in the chance of preventing radiographic progression in the subgroups (see for calculation details footnote 1), resulted in an absolute risk difference between TCZ versus TCZ+MTX of 8% in patients with a low-level of baseline joint damage, versus only 1% in the subgroup with a high-level of baseline joint damage. In the subgroup with a low-level of disease activity at baseline, this was 3% versus 7% in the subgroup with a high-level of disease activity at baseline, see supplementary Table S3. Figure 1 Relative chance of preventing radiographic progression in early RA.* Number of patients in low-level baseline joint damage subgroup; 160 in TCZ vs. 166 in TCZ+MTX, number of patients in high-level baseline joint damage subgroup; 129 in TCZ vs. 132 in TCZ+MTX. Number of patients in low-level baseline DAS28 subgroup; 149 in TCZ vs. 145 in TCZ+MTX, number of patients in high-level baseline DAS28 subgroup; 146 in TCZ vs. 153 in TCZ+MTX. *Relative chances (95%CI) are based on stratified analyses, controlling for age, gender and DAS28 at baseline. A RR above 1 is associated with less radiographic progression for TCZ. Low/high levels of baseline joint damage (SvdH score ≤1/>1) or disease activity (DAS28 ≤6.37/>6.37) were based on their respective median values in the data. RA= rheumatoid arthritis; SvdH= Sharp van der Heijde; TCZ= tocilizumab; MTX= methotrexate; RR= relative chance; % Ref= proportion of patients with no progression (i.e., based on raw data) in the reference group, i.e., TCZ+ MTX group. 5