Uveitis is a sight-threatening intraocular inflammation, characterized by a heterogeneous clinical presentation. This condition results from a complex interaction among multiple genetic, immunity, environmental, and epigenetic factors. The diagnosis of uveitis is often challenging because many diverse diseases with similar clinical features can cause this condition. These diseases are a common cause of blindness, collectively accounting for 10–25% of the current legally defined blindness worldwide [1
]. They have a profound negative effect on the quality of life. Behҫet’s disease (BD), sarcoidosis, and Vogt-Koyanagi-Harada disease (VKH) are the three most common non-infectious diseases with uveitis [3
]. These highly sight-threatening diseases develop in genetically susceptible individuals. The uveitis seen in these diseases is characterized by an exaggerated intraocular immune response against antigens interacting with CD4 T cells (Th1 and Th17 cells) and antigen-presenting cells. This immune response ultimately leads to alterations in blood-retinal barrier function, retinitis, choroiditis, and tissue damage [7
]. The etiology of these kinds of uveitis remains a conundrum, and there is no established diagnostic biomarker. They are mostly diagnosed according to clinical symptoms and the clinical criteria shared by other types of uveitis. Diagnostic biomarkers specific for BD or VKH have not yet been identified, and lumbar puncture is a relatively invasive procedure and may not be practical or suitable for the diagnosis of VKH. Previous studies have assessed whether candidate biomarkers, including angiotensin (ACE) and soluble interleukin (IL) -2 receptor, can diagnose ocular sarcoidosis [12
]. The results of these studies indicate that when these markers are used alone, the sensitivity and specificity of detection are insufficient to reliably identify sarcoidosis patients with uveitis. Clinical confusion can lead to the diagnosis being missed or delayed, which increases the risk of impaired vision. Thus, diagnosing BD, sarcoidosis, or VKH patients is often challenging for non-uveitis specialists. Furthermore, 80–90% of BD or VKH patients do not completely meet the clinical criteria and are described to have an “incomplete” or “probable” case of the disease, which contributes to clinical confusion and delayed diagnosis [14
]. Treatment with systemic steroids is effective for sarcoidosis and VKH, but not BD. Since the therapeutic strategies for the three diseases are different, accurate diagnostic biomarkers are essential to improve clinical outcomes, avoid unnecessary therapies, and suppress retinal damage as well as systemic manifestations. Given the high pathological and therapeutic difference among these three diseases, identification of metabolomic biomarkers specifically expressed in the sera of corresponding patients might greatly facilitate diagnosis. Metabolomics is an -omics approach to quantify metabolites in biological samples. These small molecules frequently reflect the pathology of various diseases, especially during inflammation. Therefore, simultaneous assessment of various pathways with multiple multivariate analyses have been used to characterize the metabolic statuses of patients and diagnose the corresponding diseases accordingly [16
]. The metabolites in the vitreous fluid detected using nuclear magnetic resonance (NMR) along with a partial least squared-discrimination analysis (PLS-DA) can discriminate between lens-induced uveitis and chronic uveitis [17
]. However, invasive sampling of aqueous humor is required. Conversely, blood is one of the biofluids that can easily be sampled, enabling frequent testing. Previously, gas chromatography-mass spectrometry with a PLS-DA has been used to discriminate BD patients from healthy controls (HCs) [18
]. NMR with principal component analysis (PCA) has been used to analyse serum samples to discriminate sarcoidosis patients from HCs [19
]. Plasma samples collected from VKH patients have been analysed using liquid chromatography (LC)-MS with PCA and orthogonal partial least square discriminant analysis (OPLS-DA) to differentiate these patients from HCs [20
]. All these studies have aimed to distinguish patients with a particular type of uveitis from HCs, whereas methods to differentiate different types of uveitis patients from each other have not been investigated.
Here, the serum metabolites of BD, sarcoidosis, and VKH patients were compared to identify disease-specific diagnostic biomarkers. We quantified hydrophilic metabolites using LC-time-of-flight (TOF)-MS and evaluated the discrimination abilities of these quantified data.