Rheumatoid arthritis (RA) is a common chronic autoimmune disease that can affect multiple joints [1, 2]. It is characterized by symmetrical joint pain, swelling and stiffness accompanied by progressive joint destruction and disability [3]. The pathogenesis of RA is still not fully elucidated, and the existing treatments are not yet able to completely cure RA but only control inflammation and delay progression [4]. Therefore, timely and accurate diagnosis and treatment can reduce irreversible joint injury and disability in RA patients [5, 6], which is of great significance for the survival and prognosis of RA patients.

Currently, the clinical diagnosis of RA is mainly based on the patient’s clinical symptoms, X-ray findings and classical laboratory indicators. However, at the early stage, the clinical representations of RA are relatively diverse. The traditional laboratory diagnostic indicators have many limitations in clinical practice, which are prone to missed diagnosis and misdiagnosis, thus causing patients to miss the best opportunity for treatment. The classification criteria of RA are based on the 2010 American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) classification criteria [7]. The specific biomarkers for the detection of RA provided in the 2010 ACR criteria are anti-cyclic citrullinated peptide antibodies (anti-CCP) and rheumatoid factor (RF) [8, 9]. Clinically, RA patients can be divided into serologically positive (SP) RA patients (RF+ and/or anti-CCP+) and serologically negative (SN) RA patients (RF− and anti-CCP−) based on these two indicators. Moreover, when anti-CCP and RF are both negative, more than ten joints must be affected to be considered RA according to the 2010 ACR criteria. Therefore, it is urgent to find potential diagnostic tools for SNRA patients to improve the accuracy of diagnosis. This will significantly reduce the missed diagnosis rate and misdiagnosis rate of RA, which is of great significance for the timely clinical diagnosis and prognosis of RA patients.

In our previous studies, we focus on the role of transcription factor YY1, Th17 cell differentiation, inflammatory factor IL-6, matrix protein Cyr61, red blood cell distribution width, etc., in RA pathogenesis [10,11,12,13] and the laboratory diagnosis of RA. Previous studies have indicated [14] that peripheral blood circulating microRNAs miR-22-3p and let-7a-5p have high diagnostic potential in RA. We realized that molecular diagnostic markers may be a good alternative or complementary for traditional serological diagnostic markers. Therefore, the diagnostic value of molecular biomarkers in RA has received much attention.

Long noncoding RNA (LncRNA) is longer than 200 nucleotides noncoding RNAs [15] and cannot encode proteins [16,17,18], and it is well known that lncRNAs regulate gene expression mainly through various interactions with DNA, RNA and proteins [19, 20]. Therefore, lncRNAs are involved in various critical regulatory processes, such as X-chromosome silencing, chromatin modification, transcriptional activation interference and post-transcriptional modification [21]. In addition, lncRNAs are widely distributed in a variety of bodily fluids and have been shown to be quite stable in plasma, which may serve as biomarkers for various diseases [22]. Meanwhile, from detection perspective, a single quantitative real-time PCR (qRT-PCR) could simultaneously detect multiple lncRNAs. These studies support the innate advantages of lncRNAs in establishing a combined diagnosis of multiple indicators. In clinical studies, researchers have found that lncRNAs are independent risk factors for a variety of diseases [23, 24], which suggests that lncRNAs have the enormous potential to replace or supplement conventional diagnostic markers. It has been found that circulating lncRNASNHG11 in peripheral blood can distinguish between precancerous lesions and early tumors of colorectal cancer. The combination of lncRNAs ZFAS1, SNHG11, LINC00909 and LINC00654 showed a good diagnostic effect in the colorectal cancer (AUC = 0.937) [25]. However, compared with other diseases, there are few studies on lncRNAs as biological diagnostic markers for RA. Therefore, a more in-depth study is necessary. Studies have demonstrated that lncRNAs have been shown to be implicated in disease progression of RA, and a variety of lncRNAs are abnormally expressed in synovial cells [26], peripheral blood mononuclear cells and T cells [27,28,29], which provides theoretical support for lncRNAs as diagnostic markers of RA.

At present, a single assessment indicator often fails to meet current clinical needs, and the development and construction of a multi-indicator combined diagnostic model has gradually become a new trend in disease diagnosis research [30, 31]. The multi-index combined diagnostic model can diagnose or predict certain diseases through multiple clinical indicators or characteristics and provide clinicians with more accurate and reliable clinical diagnosis tools. Therefore, in this study, based on the expression of lncRNA, clinical data and laboratory indicators of patients were also collected to build a diagnostic model.

In this study, we collected peripheral blood cells from RA patients and healthy donors for eukaryotic long noncoding RNA sequencing, and then, the results of lncRNA sequencing were analyzed by bioinformatics, and verified by quantitative real-time PCR (qRT-PCR). The results indicated that lncRNA PTPRE increased in SNRA patients compared with healthy donors (HD). To improve its diagnostic capability in SNRA, we constructed a SNRA diagnostic model based on PTPRE expression, neutrophil count and RDW after logistic regression analysis.