Sex differences in remission rates over 24 weeks among three different biological treatments compared to conventional therapy in patients with early rheumatoid arthritis (NORD-STAR): a post-hoc analysis of a randomised controlled trial

Kristina Lend, Ronald F. van Vollenhoven, Jon Lampa, Merete Lund Hetland, Espen A. Haavardsholm, Dan Nordström, Michael Nurmohamed, Bjorn Gudbjornsson, Anna Rudin, Mikkel Østergaard, Till Uhlig, Gerdur Grondal, Kim Hørslev-Petersen, Marte S. Heiberg, Tuulikki Sokka-Isler, Frieda A. Koopman, Jos W. R. Twisk, Irene van der Horst-Bruinsma

Research output: Contribution to journalArticleAcademicpeer-review

1 Citation (Scopus)


Background: Rheumatoid arthritis is a chronic inflammatory disease with a well-recognised female preponderance. In this post-hoc analysis of the NORD-STAR trial, we aimed to examine sex differences in remission rates with three different biological treatments combined with methotrexate versus active conventional treatment over 24 weeks, in patients with early rheumatoid arthritis. Methods: NORD-STAR was a multicentre, investigator-initiated, assessor-blinded, phase 4, randomised, controlled trial of early rheumatoid arthritis, done in Denmark, Finland, Iceland, Norway, Sweden, and the Netherlands. Newly diagnosed patients, naive to disease-modifying antirheumatic drugs, aged 18 years or older with early rheumatoid arthritis and with a symptom duration less than 24 months were randomly assigned (1:1:1:1) to receive active conventional treatment, certolizumab-pegol, abatacept, or tocilizumab. Sex was reported in case report forms by study physicians or by study nurses. Data on gender were not collected. Remission outcomes were analysed with logistic generalised estimating equations (GEE), using a logit link and exchangeable correlation matrix. The model included treatment, time, sex, and the relevant interactions. For this post-hoc analysis, the co-primary outcomes were differences in Clinical Disease Activity Index (CDAI) remission (CDAI score ≤2·8) between sexes over time and at week 24, assessed with interaction terms (men vs women within each treatment comparison) and using active conventional treatment as the reference. We present adjusted average marginal differences in remission rates (risk differences) with 95% CIs. Findings: Between Dec 14, 2012, and Dec 11, 2018, 812 patients were enrolled and randomly assigned; 217 received active conventional treatment, 203 received certolizumab-pegol, 204 received abatacept, and 188 received tocilizumab. All 812 patients were included in this analysis; 561 (69%) were women and 251 (31%) were men. Observed CDAI remission rates at 24 weeks were numerically higher among men than among women despite comparable disease activity at baseline (55% vs 50% with active conventional treatment, 57% vs 52% with certolizumab-pegol, 65% vs 51% with abatacept, and 61% vs 40% with tocilizumab). In the adjusted analysis, with active conventional treatment as the reference, the only significant difference between men and women was in the tocilizumab group (pinteraction=0·015); men in the tocilizumab group had a higher probability of CDAI remission, on average over time, than did men in the active conventional treatment group (0·12; 95% CI 0·00 to 0·23), whereas women in the tocilizumab group had a lower probability of remission than did women in the active conventional treatment group (–0·05, 95% CI –0·13 to 0·02). Interpretation: Numerically higher remission rates were observed in men than in women in all four treatment groups at week 24, suggesting that this generalised sex difference is not related to the treatment. The difference between men and women was significantly greater with tocilizumab, an interleukin (IL)-6 inhibitor, than with active conventional treatment, suggesting a possible additional sex-based effect specific for IL-6 blockade. Funding: None.
Original languageEnglish
Pages (from-to)e688-e698
JournalThe Lancet Rheumatology
Issue number10
Publication statusPublished - 1 Oct 2022

Cite this