OP0084 DIGITAL SPATIAL PROFILING REVEALS DISTINCT SYNOVIAL TISSUE MACROPHAGE TRANSCRIPTOMIC SIGNATURE OF SUSTAINED REMISSION IN RHEUMATOID ARTHRITIS PATIENTS AT RISK OF DISEASE FLARE AFTER TREATMENT CESSATION

Abstract

BackgroundSustained remission is the treatment goal for Rheumatoid Arthritis (RA) and once achieved patients are eligible to treatment tapering or discontinuation. However, this exposes patients to the occurrence of unpredictable disease flare, and to date there are no definitive predictive biomarkers of flare for RA in remission that could be used in clinical practice.ObjectivesTo assess the impact of clinical classification of remission on synovial tissue (ST) features of RA in sustained remission and to identify predictive biomarkers of disease flare.Methods200 RA in sustained clinical (102 RA with DAS<1.6 and 98 RA fulfilling Boolean remission criteria for at least 9 months, respectively) and ultrasound (US) remission (PD negative) under Methotrexate with or without biological-Disease Modifying Anti-Rheumatic Drugs (bDMARDs) were enrolled and underwent to US guided ST biopsy. 373 naive RA were included as comparison. For each patient, synovitis degree was determined using a H&E-based semiquantitative score1. Some ST samples of remission RA were used for synovial tissue macrophage (STMs)(CD206/MerTK) FACS phenotyping and digital spatial profiling (GeoMx DSP, Nanostring) to quantitate transcript abundance of CD68pos cells in 138 spatially distinct ST regions of interest (ROI). After study entry, RA were randomly assigned to tapering/discontinuation (TAP/DISC) (tapering c- or b-DMARD treatment for 6 months and discontinuing c- or bDMARD afterwards) or maintaining the same therapeutic scheme (CONT). Each RA was followed every 3 months to assess flare rate after treatment modifications for 24 months.ResultsRegardless of either DAS- or Boolean-defined, remission patients had significantly lower KSS than naive RA (p<0.0001 for both). However, ST of RA in Boolean remission had lower KSS (p<0.0001) and was enriched in CD206posMerTKpos STMs (p=0.0012) as compared to DAS-defined remission RA. 73(36.5%) RA experienced a disease flare regardless of the treatment change during 24 months follow-up. Stratifying RA in remission based on remission definition and treatment group, DAS-defined remission RA who had a disease flare within at least 6 months follow-up had, at study entry, significantly higher KSS (p<0.0001) than RA who maintained a sustained remission, regardless of the treatment change (CONT:p=0.0027 and TAP/DISC:p=0.0011). Logistic regression analysis revealed that baseline KSS≥3 [AUC:0.748(95%CI:0.649-0.846)p<0.0001] was an independent predictive factor of disease flare [OR:6.9(95%CI:2.82-16.81)] within 24 months follow-up in DAS-defined remission RA. Conversely, RA in Boolean remission did not differ for KSS at study entry in both the CONT (p>0.05) and the TAP/DISC (p>0.05) group in relation to disease flare. However, considering STMs phenotype, RA in Boolean remission in the TAP/DISC group who had low levels of CD206posMerTKpos (<38.1%), experienced more likely a disease flare compared to RA in the CONT subgroup with CD206posMerTKpos≥38.1% (p=0.0014). Logistic regression analysis confirmed that, before treatment change, STMs phenotype (CD206posMerTKpos <38.1%) in RA in remission is an independent predictor of disease flare [OR:6.25(95%CI:1.33-29.43)] within 24 months. Finally, DSP analysis using CD68 morphology marker, revealed that lining and sublining layer CD68pos spatial transcriptomics distinguished, at baseline, remission RA who flared after treatment modification from those who did not.ConclusionDisease flare is a common event in RA in sustained remission after treatment modification. KSS and STMs phenotype identified by flow cytometry or by tissue spatial transcriptomic can identify RA in remission at higher risk of flare after treatment modification. Thus, spatial transcriptomic with defined panel of markers on histological biopsy tissues could be a way forward in predicting disease flare.