Long non-coding RNA signatures to distinguish fibromyalgia syndrome from rheumatic diseases

点击翻译按钮获取中文摘要

No other group of diseases encompasses a greater pathophysiology than do the rheumatic diseases. Spanning multiple organ systems, clinical decisions often rely on coordinated efforts from primary care providers and rheumatologists to rule in or rule out differential diagnoses when treating inflammatory conditions such as rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). RA is a symmetric, inflammatory, peripheral polyarthritis leading to deformity of joints via erosion of bone and surrounding cartilage. SLE can affect virtually any organ leading to fatigue, fever, myalgia, weight change and complications associated with renal, central nervous system, and hematologic systems can be life-threatening. RA and SLE are diagnosed through clinical judgment after excluding alternative diagnoses. In the case of both diseases, individual laboratory tests are effective only in a portion of the disease population. Across these analyses, the sensitivity and specificity for these laboratory measurements may have high specificity to rule in SLE but lack sensitivity as these diagnostic markers can be found in other disorders. As a physician colleague pointed out, “it is difficult to diagnose a negative”. Diagnostic approaches for both RA and SLE often rely on multiple, independent laboratory tests combined with clinical observation. Distinguishing between these diseases is important since treatment procedures for these diseases are different. Time is a factor in diagnosis of these diseases and tools are required to facilitate earlier diagnosis as treatment for autoimmune diseases are highly effective and early initiation of therapy leads to the best outcomes. Misdiagnosis of these conditions is not uncommon. Another common disease seen by rheumatologists is fibromyalgia syndrome (FMS). FMS is a common cause of widespread musculoskeletal pain that affects tendons, ligaments, and muscle. FMS is difficult to diagnose and treat and a critical clinical point is that FMS is not explained by another rheumatic or systemic disorder. Thus, FMS is a diagnosis of exclusion once other etiologies have been considered and excluded. RA and SLE are two diseases that must be eliminated from the differential diagnosis. Given the complicated diagnostic process these patients are often forced to endure, recent studies have also suggested that healthcare dollars are saved post-diagnosis and patient outcomes improve. To date, there is no laboratory test that can determine presence or absence of these three conditions from a single blood sample. The question of whether or not disease classifiers capable of providing clinically useful information could be built based upon disease-specific expression levels of mRNAs in whole blood has been a subject of research for several years. Long non-coding RNAs (lncRNA) are recently discovered regulatory RNA molecules that do not code for proteins but influence a vast array of biological processes. It is also thought that lncRNAs drive biologic complexity observed in vertebrates that may also be reflected by the greater array of complex idiopathic diseases that humans develop. As such, our data obtained in the phase 1 portion of this work, support the notion that disease-associated lncRNAs exhibit far greater differences in expression than disease-associated mRNAs. In this application, we propose to explore the hypothesis that lncRNAs are better biomarkers of human disease than mRNAs. Here, we will focus on FMS and the rheumatic diseases as disease categories and have identified and validated FMS and rheumatic disease-associated associated differentially expressed lncRNAs. Study of lncRNAs in human autoimmune disease is in its infancy and exploration of lncRNAs as biomarkers of autoimmune disease has not been previously addressed. We propose to determine expression levels of target lncRNAs in blood obtained from larger cohorts of subjects that include 1) subjects with fibromyalgia syndrome, 2) healthy controls, 3) rheumatoid arthritis, 4) systemic lupus erythematosus, and 5) peripheral autoimmune disease controls obtained from various sites in the U.S. and Europe to establish a wide geographic distribution and to identify optimum machine learning classifiers to distinguish fibromyalgia syndrome and rheumatic diseases from healthy and disease control cohorts with greatest overall accuracy.