Effects of intravenous pulse methylprednisolone in neuromyelitis optica during the acute phase

Abstract

Background

Neuromyelitis optica spectrum disorder (NMOSD) is an anti‐aquaporin 4 (anti‐AQP4) autoantibodies‐mediated idiopathic inflammatory demyelinating disease of the central nervous system. While intravenous pulse methylprednisolone (IVMP) is the recommended initial treatment option for acute onset NMOSD, its therapeutic mechanism remains unclear. We hypothesized that IVMP would reduce the expression of pro‐inflammatory factors and increase the resolution of inflammation in patients with NMOSD.

Methods

Mendelian randomization (MR) analysis was used to screen meaningful inflammatory and resolution factors for inclusion. Three MR methods with inverse variance weighting (IVW) were primarily used to identify positive results. Interleukin (IL)‐10, IL‐1β, IL‐6, C‐X‐C motif chemokine ligand 12 (CXCL12), and tumor necrosis factor‐related apoptosis‐inducing ligand (TRAIL) were screened from 41 inflammatory factors, and resolvin D1 (RvD1), maresin 1 (MaR1), and lipoxin A4 (LXA4) were screened from 6 resolution markers for inclusion. Subsequently, 12 patients with NMOSD were enrolled and treated with IVMP. Serum levels of the aforementioned inflammatory and resolution markers were measured by enzyme‐linked immunosorbent assay before and after IVMP treatment.

Results

High levels of TRAIL, CXCL12, and IL‐1β were associated with an increased risk of NMOSD (TRAIL: odds ratio [OR], 1.582; 95% confidence interval [CI], 1.003–2.495; CXCL12: OR, 3.610; 95% CI, 1.011–12.889; IL‐1β: OR, 4.500; 95% CI, 1.129–17.927). High levels of RvD1, MaR1, and LXA4 were associated with a reduced risk of NMOSD (RvD1: OR, 0.725; 95% CI, 0.538–0.976; MaR1: OR, 0.985; 95% CI, 0.970–0.999; LXA4: OR, 0.849; 95% CI, 0.727–0.993). Among patients with NMOSD, serum levels of IL‐6, CXCL12, and TRAIL significantly decreased following IVMP treatment, compared with pretreatment levels, while levels of IL‐1β, LXA4, and MaR1 significantly increased after IVMP treatment (p < 0.05). A significant positive correlation was observed between CXCL12 levels and Expanded Disability Status Scale (EDSS) scores (r = 0.451, p < 0.05).

Conclusion

Several systemic inflammatory regulators associated with the pathogenesis of NMOSD were identified. The protective roles of LXA4 and MaR1 may be indispensable components of glucocorticoid treatment. Therefore, the use of resolution markers may be a potential strategy for improving central nervous system injury in individuals with NMOSD.

Background

Neuromyelitis optica spectrum disorder (NMOSD) is an autoimmune disease that affects the central nervous system (CNS) through astrocytopathic effects. ¹ This condition is characterized by recurrent episodes of inflammation in the brain and spinal cord, with optic nerve and spinal cord inflammation being the key diagnostic features. ² Most NMOSD cases (90%) show a recurrent disease course, with intervals between clinical events ranging from days to decades. These attacks frequently result in moderate to severe functional impairments. ³ In a substantial proportion of individuals with NMOSD (60%–80%), ⁴ the presence of a specific antibody targeting astrocyte (AS) aquaporin‐4 (AQP4) is a crucial distinguishing feature between NMOSD and multiple sclerosis (MS). ⁴ Disease progression in NMOSD is dependent on the complement system and cytotoxicity facilitated by antibodies. ⁵ Autoantibodies bind to AQP4 produced by ASs, initiating inflammatory mechanisms that facilitate astrocyte demise. ⁶

Prednisolones are commonly prescribed for the management of diverse inflammatory disorders and autoimmune conditions owing to their rapid and potent anti‐inflammatory effects. To mitigate the severity of NMOSD attacks and restore neurological function, the standard initial treatment approach typically involves the administration of high‐dose intravenous methylprednisolone (IVMP) at a daily dose of 1 g for a duration of 3–7 consecutive days. ⁷ Presently, IVMP is recommended as the primary therapeutic option for acute onset NMOSD, both during the initial attack and during relapse. Additional research is required to elucidate the pathological mechanisms underlying the acute onset of NMOSD and to evaluate the efficacy of therapeutic interventions.

The pathogenesis of NMOSD is affected by inflammatory processes. In individuals with NMOSD, both peripheral and CNS inflammation is mediated by a diverse array of cytokines and chemokines, with particular emphasis on those linked to T helper (Th)17 and Th2 cell responses. ⁸ During clinical relapse, patients with NMOSD exhibit elevated levels of interleukin (IL)‐17, IL‐21, IL‐6, IL‐22, and IL‐8 in both serum and cerebrospinal fluid (CSF), ⁹ , ¹⁰ , ¹¹ , ¹² which demonstrate positive correlations with spinal cord lesion severity. ¹³

In addition to inflammatory agents, lipids play a role in regulating inflammation. Based on recent studies, specialized pro‐resolving lipid mediators (SPMs), also known as lipid mediators, play a crucial role in the termination of inflammation. This phenomenon is commonly referred to as the “resolution of inflammation”. ¹⁴ Typically, an inherent self‐regulating inflammatory response occurs wherein activities that promote resolution are sufficient to counterbalance the inflammatory reaction. Thirteen distinct classes of SPMs have been discovered, including lipoxins (LXs), resolvins (Rvs), protectins (NPDs), and maresins (MaRs). ¹⁵ Our study postulates that IVMP treatment can effectively mitigate the acute phase of symptoms in patients with NMOSD by impeding the progression of inflammation and facilitating its resolution.

The primary objective of this investigation was to examine changes in inflammatory and resolution markers during the development of NMOSD by analyzing fluctuations in serum cytokines and SPM levels among patients with NMOSD before and after IVMP therapy. However, because of the extensive range of inflammatory and resolution factors, comprehensively examining all types of factors was challenging. Mendelian randomization (MR) analysis was performed to investigate the association between inflammatory factors, resolution factors, and NMOSD. Subsequently, a subset of 41 inflammatory factors and 6 resolution factors were identified, and those exhibiting a noteworthy influence on the risk of NMOSD were chosen for further investigation. The study design is illustrated in Figure 1.

Figure 1.

Flow chart for the experimental design.

Materials and Methods

Patient inclusion and exclusion criteria

At the Department of Neurology, First Hospital of Jilin University, China, the sera of 12 patients diagnosed with NMOSD according to the 2015 Wingerchuk et al. criteria ¹⁶ were collected. All patients were examined during the onset of NMOSD and tested positive for serum AQP4 antibodies. The serum of all patients was analyzed using cell‐based assay (CBA) methods. The serum AQP4 antibody titers of each patient at admission are shown in Table 1. The exclusion criteria for patients were as follows: (1) prior administration of an intravenous steroid within the past 30 days, (2) presence of a severe or uncontrolled infection, (3) hemodynamic instability, (4) contraindications to hormone therapy, and (5) other autoimmune diseases.

Table 1.

Serum AQP4 antibody titer at admission and EDSS score of patients before and after IVMP treatment.

	Serum AQP4 antibody titer at admission	EDSS score before IVMP treatment	EDSS score after IVMP treatment
Patient 1	1:10	1.5	0
Patient 2	1:1000	5.5	4
Patient 3	>1:1000	8	6.5
Patient 4	1:32	7	6.5
Patient 5	>1:1000	9	7.5
Patient 6	1:32	6	5
Patient 7	1:100	8	6
Patient 8	1:100	2	1
Patient 9	1:10	5.5	4
Patient 10	1:32	5.5	3
Patient 11	1:1000	6	3
Patient 12	1:32	5	2

Treatment protocol

The patient received IVMP based on the Chinese Guidelines for Diagnosis and Treatment of Optic Neuromyelitis Spectrum Diseases (2021 edition). The treatment protocol consisted of administering 1000 mg/day IVMP for 5 days, followed by 500 mg/day IVMP for 3 days. This dose was then reduced to 240 mg/day IVMP for the next 2 days, and finally, 120 mg/day of IVMP was administered for an additional 2 days. Oral hormone therapy was continued outside the hospital.

Patient sample collection

Samples from patients diagnosed with NMOSD were obtained at two distinct time points: (1) upon admission to the hospital prior to steroid treatment during an episode and (2) after completing a regimen of IVMP at a reduced dosage of 120 mg/day. The serum samples were kept at a temperature of −80°C until they were analyzed. The clinical records of patients with NMOSD included sex, age, Expanded Disability Status Scale (EDSS), CSF IgG, IgG index, and CSF routine tests. Approval from the Ethics Committee of the First Hospital of Jilin University, Changchun, China, was obtained (2023‐KS‐310). Written informed consent was obtained from all patients. All the patients were evaluated by a physician using the Neurostatus EDSS.

Mendelian randomization study to screen for inflammatory and resolution factors

To determine the causal connections between the systemic regulators of inflammation and NMOSD with AQP4‐IgG positivity, a comprehensive MR study was conducted in both directions. Three fundamental assumptions of MR were validated. First, exposure demonstrated a robust association with single nucleotide polymorphisms (SNPs). ¹⁷ Furthermore, SNPs remain unaffected by confounding variables that impact the association between exposure and outcomes. Third, the association between SNPs and outcomes was solely mediated by exposure. The STROBE‐MR guidelines offer a framework for reporting MR study. ¹⁸

Data resources

The summarized genome‐wide association study (GWAS) data for 41 systemic inflammatory regulators, 6 resolution factors, and NMOSD are available in public databases. This study utilized information on 41 systemic inflammatory regulators involving a total of 8293 Finnish participants from three separate population cohorts: the Young Finns Study on Cardiovascular Risk, FINRISK1997, and FINRISK2002. ¹⁹ The summarized GWAS data for 6 resolution factors were obtained from an association study of 8406 participants of the Atherosclerosis Risk in Communities Study, which identified loci associated with eicosanoid levels. ²⁰ The GWAS data for various antibody subtypes of NMOSD were sourced from a whole‐genome sequencing study. The study included a combined total of 132 individuals with AQP4‐IgG‐positive NMOSD and 1244 individuals serving as controls. ²¹ Table 2 provides a comprehensive explanation of the condensed GWAS data. The complete GWAS summary statistics are available at https://www.ebi.ac.uk/gwas/. On the website, the GWAS ID may be entered into the search box for database details.

Table 2.

Details of the GWAS included in this study.

Cytokines	Abbreviation	Sample information	Number
Cutaneous T‐cell attracting (CCL27)	CTACK	3631 European	GCST004420
Beta nerve growth factor	βNGF	3531 European	GCST004421
Vascular endothelial growth factor	VEGF	7118 European	GCST004422
Macrophage migration inhibitory factor (glycosylation‐inhibiting factor)	MIF	3494 European	GCST004423
TNF‐related apoptosis‐inducing ligand	TRAIL	8186 European	GCST004424
Tumor necrosis factor‐beta	TNFβ	1559 European	GCST004425
Tumor necrosis factor‐alpha	TNFα	3454 European	GCST004426
Stromal cell‐derived factor‐1 alpha (CXCL12)	SDF1α	5998 European	GCST004427
Stem cell growth factor‐beta	SCGFβ	3682 European	GCST004428
Stem cell factor	SCF	8290 European	GCST004429
Interleukin‐16	IL‐16	3483 European	GCST004430
Regulated on activation, normal T‐cell expressed and secreted (CCL5)	RANTES	3421 European	GCST004431
Platelet derived growth factor BB	PDGFbb	8293 European	GCST004432
Macrophage inflammatory protein‐1β (CCL4)	MIP1β	8243 European	GCST004433
Macrophage inflammatory protein‐1α (CCL3)	MIP1α	3522 European	GCST004434
Monokine induced by interferon‐gamma (CXCL9)	MIG	3685 European	GCST004435
Macrophage colony‐stimulating factor	MCSF	840 European	GCST004436
Monocyte‐specific chemokine 3 (CCL7)	MCP3	843 European	GCST004437
Monocyte chemotactic protein‐1 (CCL2)	MCP1	8293 European	GCST004438
Interleukin‐12p70	IL‐12p70	8270 European	GCST004439
Interferon gamma‐induced protein 10 (CXCL10)	IP10	3685 European	GCST004440
Interleukin‐18	IL‐18	3636 European	GCST004441
Interleukin‐17	IL‐17	7760 European	GCST004442
Interleukin‐13	IL‐13	3557 European	GCST004443
Interleukin‐10	IL‐10	7681 European	GCST004444
Interleukin‐8 (CXCL8)	IL‐8	3526 European	GCST004445
Interleukin‐6	IL‐6	8189 European	GCST004446
Interleukin‐1 receptor antagonist	IL1ra	3638 European	GCST004447
Interleukin‐1‐beta	IL‐1β	3309 European	GCST004448
Hepatocyte growth factor	HGF	8292 European	GCST004449
Interleukin‐9	IL‐9	3634 European	GCST004450
Interleukin‐7	IL‐7	3409 European	GCST004451
Interleukin‐5	IL‐5	3364 European	GCST004452
Interleukin‐4	IL‐4	8124 European	GCST004453
Interleukin‐2 receptor, alpha subunit	IL2rα	3677 European	GCST004454
Interleukin‐2	IL‐2	3475 European	GCST004455
Interferon‐gamma	IFN‐γ	7701 European	GCST004456
Growth regulated oncogene‐α (CXCL1)	GROα	3505 European	GCST004457
Granulocyte colony‐stimulating factor	GCSF	7904 European	GCST004458
Basic fibroblast growth factor	bFGF	7565 European	GCST004459
Eotaxin (CCL11)	Eotaxin	8153 European	GCST004460
AQP4‐postive NMOSD	–	1276 European	GCST006938
Resolvin D4	RvD4	6496 European	GCST90274668
Lipoxin A4	LXA4	6496 European	GCST90274599
Maresin 1	MaR1	6496 European	GCST90274459
Leukotriene B4	LTB4	6496 European	GCST90274533
Leukotriene D4	LTD4	6496 European	GCST90274561
Protectin D1	PD1	6496 European	GCST90274612

Choosing genetic instrumental variables

SNPs must meet three fundamental assumptions to be considered instrumental variables (IVs) in MR. In the primary analysis, systemic inflammatory regulators were treated as exposure variables, while the outcome variable was NMOSD. The number of SNPs that meet the criteria is very restricted when selecting a significance threshold of p < 5 × 10⁻⁸. A less stringent threshold of p < 5 × 10⁻⁶ was employed to increase the number of SNPs associated with systemic inflammatory regulators or NMOSD. Subsequently, SNPs that exhibited a significant correlation with the outcome using the Steiger Test were excluded. ²² Moreover, autonomous SNPs were detected by implementing a linkage disequilibrium (LD) threshold of r ² = 0.001 and a distance of 10,000 kb. ²³ To identify SNPs associated with confounding factors, the selected SNPs were validated using the PhenoScanner database. ²⁴ The strength of the IVs ²⁵ was assessed by examining the F‐statistic with a threshold of >10. ²⁶

MR statistical analyses

Three methods were used for the two‐sample MR analyses: inverse variance weighted (IVW), MR‐Egger, and weighted median. The main MR analysis employed IVW, with the other two approaches used for the supplementary analyses. ²⁷ To assess heterogeneity, Cochran's Q test was performed, and the pleiotropic impact was evaluated using the p‐value for the intercept in MR‐Egger regression analysis. Furthermore, Mendelian Randomization Pleiotropy RESidual Sum and Outlier (MR‐PRESSO) was used to identify and remove horizontal pleiotropic outliers. ²⁸ Furthermore, a leave‐one‐out sensitivity analysis was performed to detect potentially influential SNPs, which ensured the reliability and coherence of the estimates of causal effects. Analyses were performed using the TwoSampleMR (v0.5.6) and MR‐PRESSO (v1.0) packages in R (v4.2.1). Our study selected eight factors, namely IL‐1β, C‐X‐C motif chemokine ligand 12 (CXCL12), tumor necrosis factor‐related apoptosis‐inducing ligand (TRAIL), IL‐6, IL‐10, resolvin D1 (RvD1), maresin 1 (MaR1), and lipoxin A4 (LXA4), based on the results of MR analysis out of a total of 41 inflammatory factors and 6 resolution factors.

Measurement of cytokine and SPM levels

Commercially available quantitative sandwich enzyme‐linked immunosorbent assay (ELISA) kits were used to quantify the serum levels of cytokines, chemokines, matrix metalloproteinases, and their inhibitors following the manufacturer's instructions. The cytokines consisted of hu‐IL‐1β (JL13662, Jianglaibio), hu‐IL‐6 (JL14113, Jianglaibio), hu‐CXCL12 (JL12957, Jianglaibio), hu‐IL‐10 (JL19246, Jianglaibio), and hu‐TRAIL (JL13691, Jianglaibio). The SPMs used were hu‐maresin 1 (501150; Cayman Chemical Company), hu‐resolvin D1 (500380; Cayman Chemical Company), and hu‐lipoxin A4 (590410; Cayman Chemical Company).

Quantification was performed using microtiter strips precoated with monoclonal antibodies against cytokines/SPMs. Following the addition of standards or samples and subsequent removal of any unbound molecules through washing, a second enzyme or antibody conjugated with biotin was introduced into the wells. After washing, the wells were treated with a chromogenic substrate for either the enzyme or streptavidin‐peroxidase, followed by addition of a chromogenic substrate for biotin. After adding a solution to stop the reaction, the color intensity was measured. Sample cytokine/SPM levels were extrapolated using a standard curve. All samples were analyzed without dilution or duplication. A BD LSRFortessa (BD Biosciences) was used to analyze the fluorescence intensity. The coefficient of variation (CV) values for ELISA were obtained from Cayman.

Disease severity assessment

The patients' symptoms were evaluated according to their clinical manifestations and magnetic resonance imaging (MRI) features. The EDSS was assessed according to Kurtzke's standard ²⁹ based on medical journal records at the First Hospital of Jilin University.

Statistical analysis

Statistical Product and Service Solutions (SPSS) version 22 (IBM, USA) was used to analyze the data. ³⁰ The patient group underwent statistical analyses using the paired‐sample t‐test and Wilcoxon paired‐sample test. The difference (d‐value) between the paired samples was calculated, and the normality of the d‐value was assessed using the Shapiro–Wilk normality test. Normally distributed data were analyzed using a paired‐sample t‐test. The Wilcoxon paired‐sample test was used for data with a non‐normal distribution. Statistical analyses were performed in SPSS with multiple comparison correction using false discovery rate (FDR), with an FDR threshold of ≤0.05. Statistical significance was attributed to values with a p‐value of <0.05. Plot outcomes were visualized using the R package ggplot2 (version 3.2.1). ³¹

Results

Patient information

In total, 12 patients with NMOSD who met the inclusion criteria were recruited. The sex ratio (male: female) of the patients was 1:5. The mean patient age was 41.67 years. The basic clinical and laboratory information of the patients is shown in Table 3. All patients were newly diagnosed, responded well to IVMP treatment and showed significant improvement in clinical symptoms. No immunosuppressants or biologically targeted drugs were used during the acute phase of treatment. The EDSS scores of all patients before and after treatment were provided in Table 1.

Table 3.

Clinical and laboratory patient information.

	Before IVMP (n = 12)	After IVMP (n = 12)
EDSS	5.75 ± 2.24	4.04 ± 2.35
Serum AQP4 antibody titers	All positive	8 patients positive, 2 negative
Serum CRP (mg/L)	8.12 ± 1.24	6.17 ± 1.57
ESR (mm/h)	32.31 ± 3.74	12.21 ± 4.58
CSF WBC (106/L)	8.47 ± 1.48	4.56 ± 2.36
CSF IgG (mg/L)	55.35 ± 3.74	35.46 ± 5.82

CRP, C reactive protein; CSF, cerebral spinal fluid; EDSS, Expanded Disability Status Scale; ESR, erythrocyte sedimentation rate; WBC, white blood cells.

Effects of systemic inflammatory regulators on the risk of AQP4‐positive NMOSD

With AQP4‐positive NMOSD serving as the outcome, 41 systemic inflammatory factors and 6 resolution factors were considered as exposures. Details of the IVs are presented in Table S1. Using IVW methods, the risks of NMOSD increased as the levels of TRAIL, CXCL12, and IL‐1β decreased (TRAIL: odds ratio [OR], 1.582; 95% confidence interval [CI], 1.003–2.495; p = 0.048; CXCL12: OR, 3.610; 95% CI, 1.011–12.889; p = 0.048; IL‐1β: OR, 4.500; 95% CI, 1.129–17.927; p = 0.033). The levels of three inflammatory dissipating factors, RvD1, MaR1, and LXA4, were associated with a reduced risk of NMOSD (RvD1: OR, 0.725; 95% CI, 0.538–0.976; p = 0.034; MaR1: OR, 0.985; 95% CI, 0.970–0.999; p = 0.042; LXA4: OR, 0.849; 95% CI, 0.727–0.993; p = 0.040). The MR‐Egger intercept test did not detect potential horizontal pleiotropy for TRAIL, CXCL12, IL‐1β, RvD1, MaR1, and LXA4 (p = 0.666, p = 0.418, p = 0.463, p = 0.125, p = 0.750, and p = 0.844, respectively). Moreover, the Q values derived from the MR‐Egger and IVW analyses indicated the absence of significant heterogeneity (all p‐values >0.05). Findings of the MR‐PRESSO Global test showed no abnormal variation or horizontal pleiotropy (p = 0.835, p = 0.669, p = 0.714, p = 0.416, p = 0.454, and p = 0.487, respectively). A sensitivity analysis conducted using leave‐one‐out studies showed that the individual studies had no impact. Table 4 displays favorable outcomes. Table S1 also contains a summary of the findings from other MR results, heterogeneity analysis, and pleiotropy analysis.

Table 4.

Results of the MR study testing causal association between systemic inflammatory regulators and risk of NMOSD.

Inflammatory factors	Method	Beta	OR (95% CI)	p	p for heterogeneity test	p for MR‐Egger intercept	p for MR‐PRESSO (0 outliers)
CXCL12	IVW	1.284	3.610 (1.011–12.889)	0.048	0.633	0.418	0.669
	MR Egger	2.111	8.257 (0.883–77.228)	0.123	0.617
	WM	1.276	3.581 (0.587–21.841)	0.167
TRAIL	IVW	0.459	1.582 (1.003–2.495)	0.048	0.739	0.666	0.835
	MR Egger	0.534	1.706 (0.972–2.995)	0.100	0.668
	WM	0.444	1.559 (0.884–2.751)	0.125
IL‐1β	IVW	1.504	4.500 (1.129–17.927)	0.033	0.666	0.463	0.714
	MR Egger	0.447	1.563 (0.107–22.928)	0.775	0.684
	WM	1.504	4.499 (0.771–26.252)	0.095
RvD1	IVW	−0.322	0.725 (0.538–0.976)	0.034	0.321	0.125	0.416
	MR Egger	−0.676	0.509 (0.307–0.843)	0.022	0.442
	WM	−0.363	0.695 (0.472–1.024)	0.066
LXA4	IVW	−0.016	0.985 (0.97–0.999)	0.042	0.442	0.750	0.454
	MR Egger	−0.011	0.989 (0.959–1.021)	0.51	0.362
	WM	−0.011	0.989 (0.969–1.01)	0.298
MaR1	IVW	−0.163	0.849 (0.727–0.993)	0.04	0.498	0.844	0.487
	MR Egger	−0.242	0.785 (0.36–1.715)	0.558	0.413
	WM	−0.146	0.864 (0.695–1.075)	0.189

Modifications in resolution indicators of individuals with NMOSD undergoing IVMP therapy

The analysis involved examining the serum levels of RvD1, MaR1, and LXA4 to assess the impact of IVMP treatment on the resolution functions of patients with NMOSD. The primary resolution indicator concentrations of the 24 plasma samples were determined using ELISA. In patients with NMOSD, the serum levels of LXA4 significantly increased after IVMP treatment, compared with those before IVMP treatment (p < 0.05) (Fig. 2A,B). In patients with NMOSD, the serum levels of MaR1 significantly increased after IVMP treatment, compared with those before IVMP treatment (p < 0.05) (Fig. 2C,D). Despite the growing inclination toward treatment, no notable disparities were observed in RvD1 levels before and after 12 days of IVMP treatment (p = 0.27) (Fig. 2E,F).

Figure 2.

(A, B) The serum levels of LXA4 showed a significant increase after IVMP treatment (p < 0.05). (C, D) The MaR1 levels showed a significant increase after IVMP treatment (p < 0.05). (E, F) No notable disparities were observed in RvD1 levels (p = 0.27). *p < 0.05; ns, no significance.

Changes in inflammatory markers of patients with NMOSD during IVMP treatment

MR test results were used to assess the impact of IVMP treatment on the inflammatory indicators of NMOSD, with an analysis conducted on the serum levels of TRAIL, CXCL12, and IL‐1β. Furthermore, the involvement of IL‐6 and IL‐10 is being increasingly elucidated in the development of NMOSD. ³² , ³³ These five inflammatory markers were selected for further investigation. All cytokines were within the linear detection range of the ELISA, and the inter‐ and intra‐assay precisions of all ELISA kits had CVs of <10%.

The serum levels of IL‐1β were significantly increased in patients with NMOSD after receiving IVMP treatment, compared with those before IVMP treatment (p < 0.05) (Fig. 3A,B). In these patients, the serum levels of IL‐6, CXCL12, and TRAIL were significantly decreased after IVMP treatment, compared with those before IVMP treatment (p < 0.05) (Fig. 3C–H). Despite the increasing trend, no notable disparities were observed in IL‐10 levels before and after 12 days of IVMP treatment (Fig. 4A,B).

Figure 3.

(A, B) The serum IL‐1β levels showed a significant increase in patients with NMOSD after receiving IVMP treatment (p < 0.05). (C, D) The serum levels of IL‐6 showed a significant decrease after IVMP treatment (p < 0.05). (E, F) The serum levels of CXCL12 showed a significant decrease after IVMP treatment (p < 0.05). (G, H) The serum levels of TRAIL showed a significant decrease after IVMP treatment (p < 0.05). *p < 0.05; **p < 0.01.

Figure 4.

(A, B) No notable disparities were observed in IL‐10 levels after 12 days of IVMP treatment. (C) Correlation between CXCL12 levels and EDSS scores in patients with NMOSD (r = 0.451, p < 0.05). *p < 0.05; ns, no significance.

Correlation between EDSS scores with levels of inflammatory and resolution markers

As NMOSD is a CNS inflammatory condition, our subsequent goal was to establish a connection between the levels of pro‐inflammatory cytokines and neurological impairment, as assessed using the EDSS score. The correlation between the levels of CXCL12 and the EDSS scores of patients with NMOSD (r = 0.451, p < 0.05) is shown in Figure 4C. No correlation was found between EDSS scores and the levels of other inflammatory and resolution factors.

We also analyzed the correlation between ΔEDSS (the difference of EDSS after and before IVMP treatment) and resolution markers in the treatment group after IVMP, which is shown in Figure 5. No correlation was found between ΔEDSS and resolution factor levels.

Figure 5.

Correlation between ΔEDSS and inflammation dissipation index in each group after IVMP treatment.

Correlation between levels of inflammatory and resolution markers

Correlation analysis revealed significant relationships between the inflammatory and resolution markers in patients with NMOSD. As demonstrated in Figure 6, CXCL12 was positively correlated with IL‐1β concentration (r = 0.636, p < 0.05), IL‐10 was negatively correlated with IL‐1β concentration (r = −0.466, p < 0.05), LXA4 was positively correlated with RvD1 concentration (r = 0.526, p < 0.05), and TRAIL was positively correlated with IL‐6 concentration (r = 0.877, p < 0.05).

Figure 6.

Correlations between levels of inflammatory markers and resolution markers in patients with NMOSD.

Discussion

Regulatory mechanisms effectively govern the dynamic process of inflammation, which is a pivotal event in the pathogenesis of NMOSD. Under normal circumstances, these mechanisms orchestrate the progression of the inflammatory response, including the identification, stimulation, defensive destruction, and resolution of inflammation and restoration. ³⁴

Currently, in the acute stage of NMOSD attacks, the recommended treatment involves immune suppression through high‐dose IVMP pulse therapy with or without oral tapering, which is regarded as the most efficacious approach. ³⁵ Although IVMP is generally well‐tolerated, some patients may experience adverse reactions or fail to derive any benefit from its administration. Glucocorticoids have been linked to various detrimental consequences, such as Cushing's syndrome. The present study aimed to assess the serum levels of resolution and inflammatory markers in patients with NMOSD before and after IVMP therapy.

To conduct our investigation, we initially employed a two‐way multivariate regression analysis to compare two samples. The use of MR techniques has the advantage of mitigating confounding effects resulting from unmeasured or unknown factors, as risk alleles from instrumental variables are independently associated with confounding factors. ³⁶ By analyzing extensive GWAS datasets, we identified a significant correlation between cytokine levels and NMOSD, suggesting a reciprocal predictive relationship between them.

To refine the detection process, magnetic resonance analysis was employed to identify several robust candidate inflammatory markers. The MR results obtained in our study showed that TRAIL, CXCL12, and IL‐1β levels were positively associated with an increased risk of NMOSD, while RvD4, LXA4, and MaR1 levels were negatively associated with an increased risk of NMOSD. Consequently, these markers were included in the sandwich ELISA method, which effectively reduced both the duration of the experiment, and the overall costs associated with the research. Owing to the limitations of commercially available ELISA kits for testing resolution markers, levels of serum MaR1, LXA4, and RvD1 were measured. Following the administration of IVMP treatment, notable reductions were observed in the levels of IL‐6, CXCL12, and TRAIL, while increases were observed in the levels of IL‐1β, LXA4, and MaR1.

MaR1 induces inflammation resolution and modulates immune responses, thereby facilitating the efficient elimination of apoptotic neutropenia, reducing neutropenia infiltrations, and promoting regenerative processes. ³⁷ , ³⁸ Our previous research findings have shown that MaR1 can enhance cognitive function and alleviate the activation of pro‐inflammatory glial cells through the promotion of survival and phagocytosis, as well as the suppression of inflammation and apoptosis pathways. ³⁹ Additionally, reports have indicated that MaR1 intervention in a collagen‐induced arthritis model resulted in decreased joint inflammation and damage by increasing the proportion of Treg cells and reducing the proportion of Th17 cells. ³⁸ A recent study demonstrated that the administration of MaR1 to mice with autoimmune encephalomyelitis (EAE) resulted in the suppression of various pro‐inflammatory cytokines and the reduction in immune cell populations in both the spinal cord and blood. ⁴⁰ In this study, we investigated the effects of MaR1 on the resolution of inflammation during NMOSD progression and found that the administration of MaR1 could potentially be utilized as a strategy to enhance recovery from CNS injury in individuals with NMOSD. Both LXA4 and 15‐epi‐LXA4 have been identified as substances that enhance macrophage phagocytosis, thereby mitigating inflammatory responses and facilitating the resolution of inflammation. ⁴¹ In a mouse model, LXA4 effectively ameliorated cognitive impairments induced by surgery while concurrently reducing neuroinflammation and microglial activation in the hippocampus. ⁴² Furthermore, two additional studies demonstrated the protective role of LXA4 in preventing spinal cord damage by modulating the Akt/nuclear factor (erythroid‐derived 2)‐like 2/heme oxygenase‐1 signaling pathway. ⁴³ , ⁴⁴ In our study, we observed that serum LXA4 levels increased in patients following IVMP treatment, thereby providing additional evidence for the anti‐inflammatory properties of LXA4 in these patients. Consequently, we postulate that LXA4 operates through a similar mechanism to promote inflammatory remission in individuals with optic neuromyelitis.

Regulation of inflammation is significantly influenced by RvD1. In our previous study, we observed an inverse correlation between the concentrations of CSF AQP4‐IgG and the levels of RvD1 in the CSF of patients with NMOSD. The regulation of inflammation is significantly impacted by RvD1, indicating that a decrease in RvD1 synthesis may be attributed to AQP4‐IgG in the CSF. ³⁴ Although an ongoing investigation did not identify any significant changes in RvD1 levels after IVMP therapy, a noteworthy correlation was observed between RvD1 and LXA4 levels (r = 0.526, p < 0.05). These findings suggest that RvD1 exerts an inflammatory dissipation effect in patients with NMOSD.

The IL‐6 pathway has gained increasing recognition as a pathogenic mechanism in NMOSD, as it promotes T‐cell polarization and aberrant activation of B cells, ultimately leading to the production of pathogenic AQP4 antibodies by plasma blasts. Furthermore, the presence of IL‐6 has been shown to increase the permeability of the blood–brain barrier (BBB), facilitating the entry of AQP4 antibodies into the CNS, where they can bind to AQP4 water channels located on the astrocytic end‐feet. This consequently triggers activation of the complement cascade, resulting in astrocytic damage. ⁴⁵ Our results provide supplementary evidence regarding the role of IL‐6 in the pathogenesis of NMOSD.

The transmembrane protein TRAIL, a member of the tumor necrosis factor (TNF) superfamily, is expressed in multiple tissues and is capable of initiating apoptosis by interacting with death‐inducing receptors. ⁴⁶ The induction of neuronal apoptosis by TRAIL can result in neuroinflammation, and localized inhibition of the TRAIL pathway has been shown to mitigate disease severity. ⁴⁷ The interaction between DR4‐DR5 and TRAIL leads to brain cell death in patients with EAE. Associations have been observed between TRAIL levels and genetic variations in the TRAIL system, which have been linked to MS susceptibility and the response to INF‐β treatment. ⁴⁸ An in vivo model of Alzheimer's disease demonstrated that the inhibition of the TRAIL pathway can mitigate the neurotoxic effects induced by amyloid‐β (Aβ). ⁴⁹ In our study, we observed a significant positive correlation between IL‐6 and TRAIL levels (r = 0.877, p < 0.05). While these findings provide additional evidence supporting the involvement of TRAIL in NMOSD progression, further studies are required to elucidate the underlying pathogenic mechanisms.

Previous experimental models of NMOSD have already shed light on the role of IL‐1β in the formation of NMOSD‐like lesions. ⁵⁰ The findings of the study suggest that the presence of IL‐1β in NMOSD lesions and its role in triggering the production and accumulation of complement factors, such as C1q, contribute to the infiltration of neutrophils and the disruption of the BBB in proximity to NMOSD lesions. ⁵¹ The release of IL‐1β stimulates neutrophils to release myeloperoxidase and elastase, resulting in the destruction of the BBB and the development of extensive myelopathy. Additionally, IL‐10, a potent cytokine known for its ability to regulate multiple anti‐inflammatory pathways, exhibits strong anti‐inflammatory properties. ⁵² Our study findings indicate a significant negative correlation between IL‐1β and IL‐10 concentrations (r = −0.466, p < 0.05), providing evidence that IL‐1β promotes the pathological production of NMOSD by inhibiting the secretion of the anti‐inflammatory factor IL‐10. Furthermore, the MR analysis suggests that elevated circulating IL‐1β levels are associated with an increased risk of NMOSD (OR, 4.500; 95% CI, 1.129–17.927; p = 0.033). However, we also observed an increase in serum IL‐1β levels in patients with NMOSD following IVMP treatment. The administration of IVMP has been hypothesized to lead to an increase in IL‐1β levels owing to the glucocorticoids' ability to enhance the circulating neutrophil count and inhibit neutrophil apoptosis. ⁵³ The observed elevation of IL‐1β levels may potentially explain the heightened presence of neutrophils in the serum of individuals with NMOSD. ⁵⁴ Additionally, CXCL12 exhibits chemotactic properties toward monocytes, neutrophils, and immature B‐cell progenitors. ⁵⁵ The infiltration of monocytes into brain tissue is considered a crucial factor in the progression and development of inflammatory reactions in NMOSD. The recruitment of these cells, which primarily relies on CXCL12, plays a crucial role in this process. The production of CXCL12 and other chemokines that attract dendritic cells in human astrocytes is induced by IL‐1β and TNF. ⁵⁶ In our study, we also observed a significant positive correlation between IL‐1β and CXCL12 levels (r = 0.636, p < 0.05). Given that IL‐1β can stimulate the production of CXCL12, the alteration in the consistency between IL‐1β and CXCL12 can be elucidated during NMOSD. The results of our meta‐regression analysis suggested that elevated circulating CXCL12 levels were significantly associated with an increased risk of NMOSD (OR, 3.610; 95% CI, 1.011–12.889; p = 0.048). Additionally, a positive correlation was observed between CXCL12 levels and EDSS scores in patients with NMOSD (correlation coefficient = 0.451, p < 0.05). However, certain findings also support the role of CXCL12 in remyelination. Specifically, CXCL12 is believed to contribute to inflammation by attracting CD34⁺ cells and naïve T cells. ⁵⁷ However, in demyelinating diseases, alteration of effector Th1 cell polarization by CXCL12 may lead to the suppression of EAE. ⁵⁸ Given its diverse effects on the immune and nervous systems, CXCL12 plays a dual role in diseases characterized by CNS inflammation.

Conclusion

In conclusion, this study identified several systemic inflammatory regulators associated with the pathogenesis of NMOSD. The protective roles of LXA4 and MaR1 may be indispensable components of glucocorticoid treatment. Therefore, the use of pro‐resolving mediators may be a potential strategy for treating CNS injury in individuals with NMOSD. However, additional research is required to verify the safety of LXA4 and MaR1 in the context of NMOSD therapy.

Author Contributions

Formal analysis and writing: Shengnan Wang. Experiments: Jianglong Wang, Mengru Xue, Yanqing Shao, Rui Wu, Jiacheng Liu, and Ke Luo. Review and editing: Mingqin Zhu.

Funding Information

The authors are grateful for the support from the grants from Jilin Provincial Department of Science and Technology (Nos. YDZJ202301ZYTS028 and 20240402014GH).

Conflicts of Interest

The authors declare that they have no competing interests.

Consent for Publication

Informed consent was obtained from all subjects. All participants agreed to participate in the study and provide blood samples.

Supporting information

Table S1. The association between genetically predicted systemic inflammatory regulators and NMOSD.

Funding Statement

This work was funded by Jilin Provincial Department of Science and Technology grants YDZJ202301ZYTS028 and 20240402014GH.

Data Availability Statement

Human data used in this study are publicly available. All databases were obtained from the GWAS catalog website (https://www.ebi.ac.uk/gwas/). The GWAS ID listed in Table 2 can be entered into the website to query and download the GWAS dataset used in this study. If doubtful, please consult with the corresponding author.