07.01 A multi-parameter response prediction model for rituximab in rheumatoid arthritis

    loading  Checking for direct PDF access through Ovid



Elevated expression of type I IFN response genes (IRGs) was previously described to be associated with a poor response to rituximab in rheumatoid arthritis (RA). In the present study, we aimed to validate this association and assessed the predictive performance upon combination of the predictive IRG gene set with clinical parameters.

Materials and methods

In two independent cohorts of 93 (Cohort I) and 133 (Cohort II) rituximab-starting RA patients, baseline peripheral blood expression of eight selected IRGs was determined, and averaged into an IFN score. Individual predictive performance of the IFN score and clinical parameters was assessed by logistic regression. A multivariate prediction model was developed using a forward stepwise selection procedure. Patients with a decrease in disease activity score (ΔDAS28)≥1.8 after 6 months of therapy were considered responders.


A higher IFN score was observed in RTX non-responders compared to RTX responders from both Cohort I and II, but this difference was most pronounced in patients who did not use prednisone (PREDN-), as described before. Univariate analysis showed that baseline DAS28, IFN score and DMARD use were associated with non-response to rituximab, whereas positivity for IgM-RF and ACPA was associated with a good response to RTX. The multivariate model consisted of DAS28, IFN score and DMARD use, which showed an area under the curve (AUC) of 0.82. Validation of this multivariate model in Cohort II revealed a comparable AUC in PREDN- patients (0.78), but the AUC in PREDN+ patients was significantly lower (0.63), which seemed due to effect modification of the IFN score by prednisone. Multivariate analysis of the PREDN+ subgroup of Cohort II revealed a model containing DAS28 and positivity for IgM-RF and ACPA, which showed an AUC of 0.75.


The combination of predictive parameters, including IFN score, provided a promising model for the prediction of non-response to rituximab, with possibilities for optimisation via definition of the exact interfering effect of prednisone on IFN score.

Related Topics

    loading  Loading Related Articles