POS0530 MOLECULAR SIGNATURE IN SUSTAINED CLINICAL REMISSION INDUCED BY TOCILIZUMAB IN PATIENTS WITH RHEUMATOID ARTHRITIS

Abstract

BackgroundClinical remission is a clinical goal in the treatment of rheumatoid arthritis (RA). Sustained, biologics-free and true remission is an unachieved goal of the “treat-to-target” approach in most patients, and the determinants for achievement are still unclear. In our recent prospective study using multiomics analysis, we proposed that a molecular signature in peripheral whole blood can be a predictor for subsequent disease activity or activities of daily living.1 We also showed that tocilizumab (TCZ) induced deep clinical remission associated with gene expression in peripheral CD4+ T cells.2ObjectivesTo consolidate and expand our hypothesis, we investigated the significance of molecular signatures in sustained remission in a larger scale cohort.MethodsTo build and validate the diagnostic model, we collected 73 peripheral blood samples from 30 patients with active RA, 30 patients in clinical remission induced by TCZ and 13 healthy controls. We then collected another 23 samples at a point before TCZ was halted due to sustained clinical remission. In total, 96 samples were analyzed by a multiomics platform, which included RNA sequencing and comprehensive proteomics.ResultsWe first developed an optimized partial least-squares regression (PLSR) model using data from 5,436 genes and 255 proteins extracted in our previous model.1 The odds ratio in the model clearly reflected the clinical state with high fidelity (Figure 1). In that study, TCZ induced nearly half of the patients with clinical remission into molecular remission, with an odds ratio of less than zero. To clarify the characteristics of the molecular signature at sustained clinical remission under TCZ continuation, 23 samples were applied to the model. The odds ratio was largely the same as that for clinical remission. Next, we investigated the association with disease flare after cessation of TCZ. At some points before cessation, the median odds ratio in patients who experienced disease flare after stopping TCZ tended to be higher than that in patients with sustained remission after stopping TCZ in the transcriptomics model but not in the proteomics model. Thirty-five differentially expressed genes were identified between the two groups under the conditions of a >1.5-fold change and P-value<0.05.Figure 1.Odds ratio in the partial least-squares regression model using transcriptomics (A) and proteomics (B) data from rheumatoid arthritis and healthy control groupsConclusionOur larger scale study validated the idea in our previous study that TCZ induces molecular remission. A certain substantial gap associated with prognosis after quitting TCZ may exist as a molecular signature of sustained clinical remission induced by TCZ. These multiomics data sets enable us to understand sustained clinical remission at a molecular level.