Elevated sCD40L in Secondary Progressive Multiple Sclerosis in Comparison to Non-progressive Benign and Relapsing Remitting Multiple Sclerosis

Abstract

Background: The long-term prognosis of relapsing-remitting multiple sclerosis (RRMS) is usually unfavorable as most patients transition to secondary progressive multiple sclerosis (SPMS) with accumulative disability. A rare form of non-progressive multiple sclerosis (MS) also exists, known as benign MS (BMS or NPMS), which lacks disease progression defined as Expanded Disability Status Scale (EDSS) ≤3 after 15 years of disease onset without treatment.

Purpose: Our study aims to identify soluble plasma factors predicting disease progression in multiple sclerosis (MS).

Research design and study sample: We utilized Luminex multiplex to analyze plasma levels of 33 soluble factors, comparing 32 SPMS patients to age-, sex-, and disease duration-matched non-progressive BMS patients, as well as to RRMS patients and healthy controls.

Results: Plasma levels of EGF, sCD40L, MCP1/CCL2, fractalkine/CX3CL1, IL-13, Eotaxin, TNFβ/LTα, and IL-12p40 were significantly different between the various types of MS. Plasma sCD40L was significantly elevated in SPMS compared to BMS and RRMS. The combination of MCP1/CCL2 and sCD40L discriminated between RRMS and SPMS. MCP1/CCL2 was found to be the most effective classifier between BMS and RRMS, while BMS was most effectively distinguished from SPMS by the combination of sCD40L and IFNγ levels.

Conclusions: These differences may facilitate personalized precision medicine and aid in the discovery of new therapeutic targets for disease progression through the improvement of patient stratification.

Keywords: Benign MS; multiple sclerosis; relapsing remitting MS; secondary progressive multiple sclerosis; soluble CD40 ligand.

Figure 1.

Z-score heatmap comparing relative expression levels of different soluble factors detected in plasma among HC and 3 types of MS. Expression levels of 8 significantly different (P < .05) plasma soluble factors detected using Luminex and analyzed with Kruskal–Wallis test comparing HC, RRMS, BMS, and SPMS are displayed after Z-transformation. HC (n = 5), RRMS (n = 8), BMS (n = 12), and SPMS (n = 33). Hierarchical clustering of soluble factors and patient groups was completed in R using divisive clustering methodology (DIANA) from the cluster package.

Figure 2.

Correlation analysis of plasma cytokine levels against age, disease duration, and platelet count. Ages of all participants (A) and disease duration (B) of all MS patients in this study were analyzed using the Kruskal–Wallis test. Statistically significant P values between groups calculated using Dunn’s adjusted multiple comparison are shown above horizontal lines. Spearman rank correlation analysis was performed using the plasma levels of 8 most significantly different cytokines among different groups against age, disease duration, and circulating platelet count of corresponding individuals. Their correlation coefficient values are shown in scale of color as matrix (C). HC (n = 5), RRMS (n = 8), BMS (n = 12), and SPMS (n = 32).

Figure 3.

Plasma biomarkers discriminate SPMS from RRMS. (A) Column scatter graph of plasma concentrations of MCP1/CCL2 and sCD40L of RRMS and SPMS patients. P value using Mann–Whitney test are shown above the line. (B) ROC analysis using plasma concentrations of MCP1/CCL2 and sCD40L derived from RRMS and SPMS groups. (C) Multiple logistic regression analysis using plasma concentration of sCD40L and MCP1/CCL2 derived from patients of both RRMS and SPMS groups. RRMS (n = 8), SPMS (n = 30). ROC, receiver-operating characteristic; AUC, area under curve.

Figure 4.

Plasma levels of MCP1/CCL2 discriminate BMS from RRMS. Plasma concentration of MCP1/CCL2 derived from RRMS and BMS patients were compared along with their ROC plots. (A) Column scatter graph comparing plasma level of MCP1/CCL2 of RRMS group (n = 8) with BMS group (n = 12). P value using Mann–Whitney test are shown above the line. (B) ROC analysis using MCP/CCL2 to discriminate BMS from RRMS. ROC, receiver-operating characteristic; AUC, area under curve.

Figure 5.

Plasma levels of sCD40L and IFNγ discriminate SPMS from BMS. Plasma concentration of sCD40L and IFNγ in BMS and SPMS patients were compared along with their ROC plots. (A) Column scatter graph of plasma concentrations of sCD40L and IFNγ derived from BMS and SPMS patients. P value using Mann–Whitney test are shown above the line. (B) ROC analysis using plasma concentrations of sCD40L and IFNγ derived from BMS and SPMS groups. (C) Multiple logistic regression analysis combining plasma concentration of sCD40L and IFNγ derived from patients of both BMS and SPMS groups. n: BMS = 12, SPMS = 32. ROC, receiver-operating characteristic; AUC, area under curve.

Figure 6.

MCP1/CCL2 and sCD40L are significantly correlated with MS disease progression. Spearman correlation analysis was performed using plasma levels of MCP1/CCL2 (A) and sCD40L (B) against corresponding EDSS scores in all participants. r is the Spearman correlation coefficient. The solid line is the best fit line of linear regression with 95% CI (dotted line).

Figure 7.

Summary. RRMS (green) becomes SPMS (orange) over time with increased disability; BMS (blue) does not accumulate disability despite long duration of disease. Progressing to SPMS from RRMS (indicated with a half orange/half green arrow), sCD40L and MCP1 are increased. Comparing BMS to RRMS (indicated with a half blue/half green arrow), only MCP1 is increased. Although both BMS and SPMS have long disease duration, SPMS exhibits increased sCD40L and IFNγ compared to non-progressive BMS (indicated with a half orange/half blue arrow). Both MCP1 and sCD40L were found to be correlated with EDSS.