Background
Behcet’s disease (BD) is a multi-system inflammatory disorder, characterized by recurrent oral ulcers, genital ulcers, skin lesions, and ocular involvement of unknown etiology. Notably, the clinical manifestation of orogenital ulcers and skin lesions including erythema nodosum could be mimicked by tuberculosis (TB). Additionally, active TB (ATB) is frequently observed in patients with BD, and those symptoms were relieved after anti-TB therapy [
1,
2]. BD and TB are also closely related in pathogenesis.
M.tuberculosis may act as a trigger of BD through molecular mimicking. On the other hand, defective cell-mediated immunity in BD patients may increase the susceptibility of TB [
3], and the use of glucocorticoids, immunosuppressants or biologic agents may increase the risk of reactivation of TB [
4]. Therefore, it is crucial and challenging to identify TB in BD patients.
Tuberculin skin test (TST), a routine screening test for latent TB (LTB) infection, has limited diagnostic value in BD patients due to cross-reactivity with the bacillus Calmette-Guérin (BCG) vaccine and nontuberculous mycobacteria as well as pathergy reaction [
5]. In additions, the autoimmune disorder or immunosuppressive status could lead to false-negative test results [
6]. Interferon-γ release assays (IGRAs) are powerful assays for detection of TB [
7], which detect the mycobacterial-antigen-specific interferon-γ-releasing T-cells in vitro
. Given IGRAs do not cross-react with BCG vaccination or nontuberculous mycobacteria infection, IGRAs show higher specificity than TST [
8]. IGRAs, including T-SPOT.TB, have been widely used to diagnose TB in autoimmune diseases, such as systemic lupus erythematosus (SLE) [
8] and rheumatoid arthritis (RA) [
9]. However, the diagnostic value of IGRAs for TB in BD patients remains unclear. To address this point, we retrospectively reviewed a cohort of BD patients tested with T-SPOT.TB and explored the diagnostic value of T-SPOT.TB for ATB in patients with BD.
Discussion
TB is an infectious disease primarily involving lungs and is endemic in developing countries. China has the world’s second largest TB prevalence, comprising around 9% of the world total according to the 2018 World Health Organization (WHO) TB report [
17]. Patients with rheumatic diseases at high risk of TB infection, due to the immune dysregulation and the adverse effects of immunosuppressive agents [
18‐
20].
M.tuberculosis culture, the gold standard of TB diagnosis usually takes 2–4 weeks. Therefore, a rapid diagnostic method of TB remains challenging. TST is a valuable screening test for TB infection [
21], but the diagnostic value of TST in BD patients is compromised by the widely usage of the BCG vaccine in Chinese population and the immune dysfunction in BD patients [
22].
In recent years, IGRAs emerges as accurate diagnostic assays for screening of TB infection. IGRAs specifically detect the presence of specific effector T cells which response to
M.tuberculosis-specific antigens [
23]. Dai Y. et al. systematically evaluated the diagnostic accuracy of T-SPOT.TB for detecting ATB in China, and found that the pooled sensitivity, specificity, PLR and NLR of T-SPOT.TB for the diagnosis of ATB were 88% (95%CI:86–91%), 89% (95%CI: 86–92%), 8.86 (95%CI: 5.42–14.46), and 0.13 (95% CI: 0.10–0.17), respectively [
24]. And The ROC area under the curve (AUC) was 0.9548 (95% CI: 0.9323–0.9773) [
24]. Consistently, Jiang B, et al. find that the sensitivity and specificity of T-SPOT.TB assays were 92.86 and 93.64%, in patients with rheumatic disease, which were higher than TST [
25].
In this study, we evaluated T-SPOT.TB and TST for the diagnosis of ATB in BD patients. To our knowledge, this is the first study that reports the diagnostic value of T-SPOT.TB for TB in BD patients. We reported the high sensitivity (88.9%) and specificity (74.8%) of T-SPOT.TB for the diagnosis of ATB in BD patients. Compared to Dai Y’s study [
24], our study showed similar results of sensitivity and NLR of the T-SPOT.TB test for the diagnosis of ATB while the lower specificity and PLR. The different study populations may explain this difference. Using SFCs> 70/10
6 PBMC as the cutoff, the specificity and PLR are improved for diagnosing ATB in BD patients. Although, neither T-SPOT.TB nor TST can clearly distinguish between LTB infection, ATB or history of TB infection [
23,
26], we found the median number of SFCs in BD-ATB group was higher than those in BD-LTB infection group and previous TB group, which suggested that higher SFCs in T-SPOT.TB test might indicate ATB. The numbers of IFN-γ producing T cells decreased during anti-tuberculous treatment, but IGRAs results remained positive at the end of treatment in most patients, which suggest a limited role of IGRAs in monitoring treatment efficacy [
27]. Meanwhile, particular attention should be paid to ATB with negative IGRAs results, since those patients had poor outcomes, potentially due to delayed treatment [
28].
It has been reported that the pooled sensitivity of T-SPOT.TB was higher than that of TST for the diagnosis of ATB in culture-confirmed or non-confirmed TB [
29]. A discrepancy between T-SPOT.TB and TST test was noted in our study. The serial testing could increase the specificity of T-SPOT.TB for ATB diagnosis in BD patients when combined with TST, but the result didn’t reach statistical differences due to a limited number of cases. Further studies were required to explore the diagnostic efficiency between T-SPOT.TB and TST.
LTB infection is defined as the sustained immune response to
Mycobacterium tuberculosis antigens with no evidence of clinically ATB. Given the risk of LTB infection reactivation increases in patients receiving immunosuppressive therapy or with immune dysfunction [
30], it is important to identify LTB infection in BD patients. However, currently, there is no gold standard test for diagnosing LTB infection available. Several studies [
31‐
33] suggested that IGRAs, including QuantiFERON-TB Gold In-Tube (QFT-GIT) and T-SPOT.TB, and TST, are all acceptable for screening LTB infection. In immune-mediated inflammatory diseases, the sensitivity and specificity of IRGAs in the diagnosis of LTB infection is higher than those in TST [
34‐
36]. Also, T-SPOT.TB has been reported with higher sensitivity in detecting LTB infection than QFT-GIT and TST [
32,
37], and Vassilopoulos et al. reported that the positive rates of QFT-GIT and T-SPOT.TB were 21 and 25%, respectively [
38] . In our study, patients who had positive T -SPOT.TB but did not have ATB or history of TB infection were considered as LTB infection. The percentage of LTB infection in BD patients in our study was similar to that in other studies [
32,
38]. A notable concern of IGRA in BD patients is the false positive or false negative rates of the IGRA test. Firstly, immunosuppressive agents or glucocorticoid might suppress T cell response to TB antigens and cause false negative IGRA test in BD patients. Nevertheless, a meta-analysis of IGRA in rheumatic patients showed that neither steroid nor disease-modifying anti-rheumatic drugs (DMARDs) significantly affect positive IGRA results [
39]. Secondly, in BD patients with high disease activity, the systemic inflammation in BD potentially promotes T cell response via high-level proinflammatory cytokines. However, T cell response detected in IGRA is derived from the T cells with specific TCR recognizing tuberculosis-specific peptide, and systemic inflammation would not promote antigen-specific T cell response in the absence of specific peptide. Consistently, the SFCs of the negative control was less than 10 spots in our patients, which was comparable to those in healthy controls. Therefore, positive IGRA is strongly suggestive for latent TB in BD patients.
Our study has several limitations. First, our study is a retrospective analysis of a case series, which may have incomplete data of TST. Second, since all BD patients were enrolled from a single-center with relatively small sample sizes, and our center is a national referred center for complicated rheumatic diseases, potential selection bias of BD patients is possible. Third, it is challenging to identify active TB in BD patients based on clinical features since systemic inflammation is a shared feature of both BD and TB. The majority of our ATB patients were diagnosed according to clinical criteria rather than M.tuberculosis bacteriologically confirmed, which might introduce a high risk of bias. Fourth, given no gold standard of LTB infection is available, the precision of the diagnosis of LTB infection by T-SPOT.TB could not be assessed in this study. Nevertheless, this is the first study on the diagnostic value of T-SPOT.TB for TB in BD patients. A large-scale, multi-center study enrolled bacteriologically confirmed TB patients are warranted to confirm our findings.