Background Previous work  reports a differential diagnostic solution for patients with undifferentiated arthritis (UA) based on synovial gene expression. This diagnostic model uses a previously established signature of 96 genes  and 3 clinical variables (skin psoriasis, rheumatoid factors, hand arthritis), achieving 93.5% sensitivity and 86.8% specificity for the diagnosis of rheumatoid arthritis among UA patients. The signature covers 7 metabolic pathways involved in the pathogenesis of rheumatic conditions.
Objectives To show the feasibility and relevance of a routine diagnostic test based on clinical and transcriptomic data from synovial biopsies, relying on a mathematical model, a central laboratory and an online software.
Methods Kits for sample transportation, containing vials of RNA-preserving solution (RNAlater, Thermo Fisher Scientific) were sent to clinicians who harvested synovial tissue (≤0.3cm3) via needle-arthroscopy or an ultrasound-guided approach from the patient's joint (mostly knees, a few MCPs). Samples were shipped on an individual basis to a central laboratory (BE) using next day transportation at ambient temperature. No data about the patient or the doctor were provided to the laboratory aside from a sample ID.
RNA was extracted with RNEasy mini kit (Qiagen). Purity and concentration were checked with NanoDrop (Thermo Fisher Scientific). Retro-transcription was performed, followed by a qPCR (TLDA, 7900HT, Applied Biosystem). Four samples were analysed on each 384-well plate. The lab operator uploaded the PCR data on a dedicated secure web platform. In parallel, clinicians filled the 3 clinical values on that web platform. Based on both sources of information the software computed the differential diagnosis and the expression level of the 7 metabolic pathways, to feed doctor's therapeutic choices. PDF reports were then automatically sent via email to the doctor in charge of each patient. The software also managed all administrative aspects in order to get a fully functional portal.
Results Synovial biopsies of 21 patients underwent the whole process. 19 (90.5%) showed satisfactory RNA quality. For the remaining 2, back up samples (harvested at the same time as the initial ones) could be analysed. One other sample (4.8%) could not be analysed due to insufficient RNA concentration. Gene expression values, diagnostic and pathways outputs were thus obtained for 20 samples (95.2%). The correlation between diagnosis and pathways expression levels was consistent with descriptions found in the literature [3,4] (Figure)
Conclusions The routine implementation of a multi-gene RT-qPCR diagnostic test based on synovial biopsies in a real-life setting is successful. The operations put in place allowed a hassle-free management for the clinicians, thereby increasing the availability and reproducibility of the test. Early results with ultrasound-guided biopsy seem promising but require additional confirmation.
Lauwerys B & al. Heterogeneity of synovial molecular patterns in patients with arthritis. PLoS ONE, 2015.
Lauwerys B & al. WO2008104608. Method for the determination and the classification of rheumatic conditions. 2008.
Choy ES & Panayi GS. Cytokine pathways and joint inflammation in rheumatoid arthritis. New England Journal of Medicine, 2001.
Sweeney SE & Firestein GS. Rheumatoid arthritis: regulation of synovial inflammation. The international journal of biochemistry & cell biology, 2004.
Disclosure of Interest None declared