Monitoring of serum TWEAK levels guides glucocorticoid dosages in the treatment of systemic lupus erythematosus

Accurate assessment of systemic lupus erythematosus (SLE) disease activity is critical. Currently existing indices or measures for assessment are either expensive, intricate, or inaccurate. The novel indices with higher sensitivity and specificity have become one of the aims of the investigators. This study was designed to explore the relationship between serum tumor necrosis factor-like weak inducer of apoptosis (TWEAK) and systemic lupus erythematosus disease activity index (SLEDAI) as well as its role in guiding glucocorticoid dosages in the treatment of SLE. Of 59 patients with SLE, 20 patients with subacute cutaneous lupus erythematosus (SCLE), 13 patients with discoid lupus erythematosus (DLE) and 32 healthy volunteers, soluble TWEAK level was determined in both serum and urine. Monomeric C-reactive protein, antinuclear antibody, interleukin 6, complements, erythrocyte sedimentation rate, and white blood cells were measured in serum samples. Moreover, SLEDAI-2K was used for evaluating the disease. Finally, methylprednisolone was administrated orally to SLE patients with the doses depending on serum TWEAK levels.

Serum TWEAK is a useful biomarker reflecting SLE disease, and monitoring of serum TWEAK levels can improve the outcomes of glucocorticoid therapy for patients with SLE.

Background
Systemic lupus erythematosus (SLE) is a severe autoimmune disease. Currently, systemic administration of glucocorticoids is one of the common therapies for patients with SLE in China. The disease flares are commonly observed in SLE patients undergoing various pharmacotherapies, including glucocorticoid administration. Accurate assessment of lupus flares is critical but problematic in clinical trials. Until now, different indices or measures have been developed for the assessment of SLE disease activity [1][2][3]. These indices are divided into two categories: the global score systems and individual organ/system assessment scales [4]. The former category includes the European Consensus Lupus Activity Measurements, Systemic Lupus Activity Measure, and Systemic Lupus Erythematosus Disease Activity Index (SLEDAI), which provide an overall measure of SLE disease activity. The latter category assesses disease activity in single organs, such as the British Isles Lupus Assessment Group Index. Additionally, the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index score can measure chronic damage in SLE and has a prognostic value in clinical trials [4]. SLEDAI-2K is an updated version of SLEDAI, and has been considered among the global scoring systems to have the highest predictability of treatment change [3,4]. However, most indices should be applied based on the number of collected clinical manifestations and laboratory results, which may be both time consuming and costly. Moreover, the global score systems and individual organ/system assessment scales may be insufficient in predicting important regional damage such as lupus nephritis (LN) and may be affected by systemic inflammation, respectively. Therefore, the novel indices with higher sensitivity and specificity have become one of the aims of the investigators.
Tumor necrosis factor-like weak inducer of apoptosis (TWEAK) is a proinflammatory cytokine mainly synthesized by macrophages and monocytes [5]. TWEAK acts by binding to its receptor fibroblast growth factor-inducible 14 (Fn14), subsequently inducing the production of downstream cytokines such as interferon gamma-induced protein 10 (IP-10), monocyte chemotactic protein-1 (MCP-1), and regulated upon activation normal T cell expressed and secreted (RANTES) [6]. TWEAK/Fn14 activation participates in inflammatory responses in injured tissue [7][8][9]. Actually, TWEAK/Fn14 signaling is prominent in various damage in patients with SLE, including cutaneous lupus erythematosus [10,11], LN [12,13], neuropsychiatric disease [14,15], and vasculitis [16]. Soluble TWEAK exhibits higher levels in sera of patients with SLE and in urine of patients with LN [16]. Recently, we found that TWEAK/Fn14 activation damages the glomerular filtration barrier and increases filtration of anti-dsDNA IgG in MRL/lpr lupus-prone mice, which is suggested to participate in the pathogenesis of LN [13,17]. Moreover, inhibition of the TWEAK/Fn14 pathway can ameliorate SCLE as well as LN or nephrotoxic serum nephritis in murine models [11][12][13]. Obviously, the serum or urine levels of TWEAK correlate closely with tissue inflammation and damage in SLE [10][11][13][14][15][16].  Table 3. There were 39 LN patients and 20 non-LN SLE patients in all SLE patients. Non-LN SLE patients were defined as patients who had SLE but no signs of kidney involvement recently and previously. LN patients were cooperatively diagnosed by nephrologists as SLE patients with kidney involvement based on clinical manifestation and kidney biopsies. Patients whose proteinuria levels exceeded 300mg/day or have urinary casts, hematuria, and leucocyturia in urine examination were suggested to seek care from a nephrologist to have kidney biopsy surgery to make further confirmation of the existence of LN and classification. Among LN patients enrolled in this study, only 11 patients performed renal biopsy (Table 1). Serum and urine samples were collected and then detected for relevant parameters. Patients with SLE were divided randomly into non-monitoring (n = 20) and TWEAK-monitoring (n = 25) groups.
Methylprednisolone tablets (Pfizer, Italia) were administered orally according to the serum levels of TWEAK in the TWEAK-monitoring group. The methylprednisolone doses were less than 0.2 mg/kg/day, 0.2 to 0.4 mg/kg/day and 0.4 to 0.8 mg/kg/day when serum TWEAK levels were less than 150 ng/ml, 150 to 500 ng/ml, 500 to 1000 ng/ ml respectively. Patients need to be hospitalized when the serum TWEAK level is greater than 1000 ng/ml.
In the non-monitoring group, the methylprednisolone doses are determined according to the assessment of patients. The demographic characteristics of the non-monitoring and TWEAK-monitoring groups patients are presented in Table 4. This clinical trial was randomized, non-placebo controlled, and unblinded. These treatments lasted for 3 to 12 months. Serum and urine samples were collected before (month 0) and after (months 1 to 12) treatment. An online SLEDAI-2K calculator (http://www.s2k-ri-50.com) was used for evaluating the disease. 22

Enzyme-linked immunosorbent assay
The TWEAK levels in sera or urine were quantitated with a sandwich enzyme-linked immunosorbent assay (ELISA) method. Briefly, the capture antibody (Cat # DY1090, R&D Systems Inc., Minneapolis, MN, USA) was coated to 96 well microplates, followed by blocking with 2% bovine serum albumin in phosphate buffer saline. The serum or urine samples (1:100 diluted) and streptavidin-horseradish peroxidase conjugated detection antibody (Cat # DY1090, R&D Systems Inc.) were added to plates in order. Finally, substrate (H 2 O 2 and tetramethylbenzidine mixture) and stop (2N sulfuric acid) solutions were added accordingly. The optical densities were determined using a microplate reader set to 450 nm (Thermo Fisher Scientific, Waltham, MA, USA).
Anti-dsDNA ELISA was performed as previously described. 17 Anti-nuclear antibody (ANA) was detected by an indirect immunofluorescence kit that has Hep-2 slides as substrate (Oumeng Inc., Beijing, China). Urine albumin was determined by using a sandwich ELISA kit (Elabscience Inc., Wuhan, China).

Determination of interleukin 6, complements, and erythrocyte sedimentation rate
Interleukin (IL) 6 was determined by using a chemiluminescent immunoassay kit (Siemens, Marburg, Germany). C3 and C4 were determined by using scattering turbidimetric assay kits (Siemens). Erythrocyte sedimentation rate (ESR) was determined by using an automatic ESR machine (Vital Diagnostics, Forli, Italy). Protocols were provided by the manufacturers.

Statistical analysis
The STATA 10.0 software package (StataCorp, College Station, TX) was used for the analysis of experimental data. Data were expressed as the means ± standard error of the mean. For comparison of more than two groups, analysis of variance was used, followed by the Bonferroni test or the Student t test that compared two groups for statistical difference. The chi-square test was used for the comparison of genders. Linear regression was used to analyze the relationship between two parameters. Differences were considered significant at p < 0.05.

Serum TWEAK levels correlates positively with SLEDAI-2K in patients with systemic lupus erythematosus
Serum TWEAK levels were determined in patients and in healthy controls. It showed that the patients with SLE (383.0 ± 45.37 ng/ml) or SCLE (129.1 ± 25.73 ng/ml) had higher TWEAK levels than patients with DLE (78.38 ± 22.85 ng/ml) or healthy controls (49.28 ± 6.04 ng/ml) (p < 0.05) (Fig. 1A). The SLE patients also had higher TWEAK levels than patients with SCLE (p < 0.05), whereas there was no significant difference between patients with DLE and healthy controls (p > 0.05) (see Fig. 1A Fig. 2A).
By Spearman Rho test, the correlations between the serum parameters and SLEDAI-2K were analyzed in patients with SLE (Table 2). It showed that TWEAK, mCRP, ANA and ESR had positive correlations with SLEDAI-2K (p < 0.05). Moreover, TWEAK levels had the highest correlation coefficient among these parameters. The other parameters including IL-6, C3, C4, anti-dsDNA, and white blood cell had no significant correlations with SLEDAI-2K score (p > 0.05) ( Table 2).

Urine TWEAK levels reflect renal damage in patients with lupus nephritis
Urine TWEAK levels were determined in patients and healthy controls, revealing that the patients with SLE (226.72 ± 40.71 ng/ml) had higher TWEAK levels than patients with DLE (55.50 ± 13.74 ng/ml) or healthy controls (54.97 ± 16.94 ng/ml) (p < 0.05), whereas there was no significant differences between patients with other different types (SCLE or DLE) and healthy controls (p > 0.05) ( Table. 3 The relationships between these two parameters (urine TWEAK and serum anti-dsDNA IgG) and SLEDAI-2K were further analyzed by linear regression. It showed that urine TWEAK (r 2 = 0.360, p = 0.030) and serum anti-dsDNA IgG (r 2 = 0.296, p = 0.036) exhibited a positive correlation with proteinuria levels (see Fig. 3B and C).

TWEAK-based glucocorticoid therapy is associated with fewer lupus flares
To further explore the guidance role of serum TWEAK, methylprednisolone was administered orally to SLE patients with the doses depending on the TWEAK levels. We observed clinical manifestation such as fever, oral ulcer, erythema, arthralgia, alopecia, neuropsychiatry, and LN (proteinuria, urinary casts, hematuria, and leucocyturia). During the sixth to twelfth months, the TWEAK-monitoring group had lower SLEDAI-2K scores as compared with the non-monitoring group (see Fig. 4A). Accordingly, the average doses of glucocorticoid were lower in the TWEAK-monitoring group (Fig. 4B). Moreover, the TWEAKmonitoring group had fewer flares of certain manifestations during months 10 to 12, including cutaneous erythema and arthralgia (see Fig. 4C).

Discussion
In this study we demonstrated that serum TWEAK levels are higher in patients with SLE or It was previously demonstrated that the serum TWEAK level in SLE patients is significantly higher compared with healthy donors, and fluctuates with the MCP-1 and IP-10 levels in sera , whereas the patients with SCLE or DLE were not discussed [19]. We found that TWEAK, as well as downstream proinflammatory cytokines including RANTES, MCP-1, and IP-10, is also highly expressed in skin lesions of CLE patients [10]. Glucocorticoids constitute the basis of SLE therapy. On the contrary, steroids are responsible for many of the most severe comorbidities that threaten patients with lupus and may be a key factor contributing to infection, which is the common cause of death for patients with lupus [27]. In this study, methylprednisolone was administered orally to patients with SLE with the doses depending on serum TWEAK levels. We found that the average doses of glucocorticoid were lower in the TWEAK-monitoring group. Moreover, the TWEAK-monitoring group had lower SLEDAI-2K scores and fewer flares of certain manifestations. A flare or relapse of SLE can be defined as an increase in disease activity requiring more-intensive treatment [28]. These results strongly suggest that monitoring serum TWEAK levels may improve the outcomes of glucocorticoid therapy for patients with SLE.
A limitation in this study was that the clinical trial was randomized but not placebocontrolled and unblinded. Also, the number of patients in each group was as few as 20.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Open Access
This article is distributed under the terms of the Creative Commons

Competing interests
The authors declare that they have no competing interests.

Funding
This study was partially supported by the National Natural Science Foundation of China    There was no significant difference between two groups.
3 Mann-Whitney U test was used for the comparison of clinical parameters.