Dimethyl fumarate for patients with neuromyelitis optica spectrum disorder

We read with interest the report by Yamout et al. ¹ and the clinical commentary by Kira ² describing two patients with neuromyelitis optica spectrum disorder (NMOSD) and severe relapses during the first 3 months on dimethyl fumarate (DMF) treatment.

NMOSD may present as an aggressive autoimmune disease with an annualized relapse rate (ARR) of 2.0-3.0. Distinct histopathological NMOSD characteristics include astrocytic damage and often pronounced necrosis. In this context, just one relapse, especially if it involves the medulla oblongata, can lead to respiratory failure and death. Such severe relapses have been reported even for NMOSD patients on rituximab, the most effective disease-modifying drug (DMD), which according to most reports reduces NMOSD ARR to 0.3-0.5. ³

Regarding the immunopathology of NMOSD, although aquaporin-4 antibodies have been proven to be disease mediating in more than 70% of the cases, antigen-specific T cells are required for their generation in the periphery as well as for the disruption of blood–brain barrier and orchestration of the immune attack in the central nervous system. Furthermore, compared with multiple sclerosis (MS), NMOSD patients have elevated interleukin IL-6 and IL-17 in the cerebrospinal fluid and display higher proportions of circulating T helper 17 (Th17) and IL-17-producing CD8⁺ T cells. ⁴ Therefore, NMO is also considered a paradigm for a Th17-driven autoimmune disease.

The reason why many Th1 and/or Th17 effective DMDs, which have been developed for relapsing remitting multiple sclerosis are ineffective and probably exacerbate NMOSD can only be postulated. Natalizumab induces an intravascular increase of circulating B-cell precursors, which might be detrimental in the case of NMOSD. ⁵ Fingolimod could modulate local S1P receptors leading to enhancement of blood–brain barrier permeability and astrocyte activation. ⁶ Alemtuzumab could induce an imbalance of T regulatory cells and IL-21 receptor–positive cells with subsequent effects on B cells. ⁷

However, the case of DMF could be different. Firstly, in a spontaneous Th17-mediated chronic animal model of opticospinal experimental autoimmune encephalomyelitis (OSE-EAE), which resembles in many aspects the human NMOSD model, low dose fumaric acid esters reduced clinical myelitis signs, demyelination, and macrophage infiltration. ⁸ Apart from its actions on T cells, especially the shift of Th1/Th17 cells toward Th2 cells and the reduction of neutrophil adhesion on endothelial cells through the hydroxycarboxylic acid receptor (HCAR2), many reports propose a significant B-cell-mediated effect.

According to a recent report, 3 months after DMF treatment initiation, total B-cell counts decreased due to losses of mostly circulating mature/differentiated memory B cells. This effect on mature B-cell survival and induction of apoptotic cell death was also confirmed in vitro. Furthermore, DMF treatment reduces B-cell expression of granulocyte-macrophage colony-stimulating factor (GM-CSF), IL-6 and tumor necrosis factor (TNF)-a as well as phosphorylation of STAT5/6 and nuclear factor (NF)-kB, resulting in a significant anti-inflammatory shift of B-cell profiles. ⁹ In another recent study, stable patients showed significantly lower CD4⁺, CD8⁺ T cells as well as CD19⁺ B cells compared to clinically active patients under DMF treatment. Reduced CD8⁺ T cells and CD19⁺ B cells 6 months after DMF start may allow prediction of the treatment response in the first year. ¹⁰

It has to be pointed out that these immunomodulatory effects were measured at least 3 months after treatment initiation. This correlates with results from phase III clinical trials, which report that DMF significantly reduced the cumulative risk of relapse after 10 weeks on treatment and that the effects on relapse rates further increased during the second trimester. The reduction in Gd⁺ lesion activity was evident by 8 weeks and was statistically significant by 12 weeks after initiation of DMF treatment.^11,12

The aggressive relapses described in these two case reports occurred in the first 3 months of treatment and treatment “escalation” on DMF was performed during definite clinical activity. Based on the above-mentioned immunological mechanisms, the reported deterioration of the patients in this case report may rather be attributed to the aggressive nature of NMOSD and the lack of effect of DMF in the first 3 months. In our center, we have treated patients with NMOSD disorders with DMF after de-escalation from Rituximab. In patients with high disease activity, we had to return to Rituximab within 6–9 months (unpublished observations).

Concluding, surely knowledge is limited when patients with NMOSD are treated with DMF. DMF is clearly not potent enough to control NMOSD aggressive forms as escalation or even de-escalation treatment. However, we believe that the data arguing for an exacerbation of NMOSD during dimethyl treatment are at this time point scarce and therefore not reliable enough. Further immunological studies dissecting the immunological mechanisms of DMF and its effects on B cells will help us understand its potential role in NMOSD.