Activation of disease during therapy with alemtuzumab in 3 patients with multiple sclerosis

Discussion

The authors of one recent case series in The Lancet Neurology³ presenting also with rapid clinical deterioration 6 months after the initial course of alemtuzumab hypothesized a B-cell-mediated pathogenesis of disease activation. Due to the striking similarity of our reported cases, we discuss the facts favoring this hypothesis but also present alternative explanations. First, the pharmacodynamics of alemtuzumab support a B-cell-mediated pathogenesis in our cases: alemtuzumab leads to a CD52-selective depletion of B and T lymphocytes.⁵ Reconstitution of T cells is slow, with CD4+ T cells and CD8+ T cells reaching 9% and 13% of baseline levels after 3 months and 36% and 71% after 12 months. Repopulation of B lymphocytes, on the other hand, usually occurs more rapidly, with a return of the total B cell count to normal levels after 3 months and to above normal values (165% of baseline) after 12 months.⁶ In one of our patients, details regarding the lymphocyte count were available at different points in time. In line with the literature,⁶ CD4+ T cells and CD8+ T cells were slow to reconstitute, reaching 23% and 19% by month 6. However, the total B cell count (baseline 0.28 × 10⁹/L) had risen to 0.72*10⁹/L (257%) after 6 months, and accordingly was more than twice as high as the B cell count reported in the literature⁶ by month 6 (figure 2). The reconstituting B cell pool is initially dominated by mature naive cells,⁶ possibly responsible for the B-cell-mediated autoimmunity in alemtuzumab treatment (e.g., idiopathic thrombocytopenic purpura, Goodpasture syndrome, thyroid disease).¹ Yet it is currently not understood why B-cell-mediated autoimmunity does not lead to an exacerbation of MS. B cell counts at the time of disease activation were not provided in the 2 other recent case series on disease activity during immunotherapy with alemtuzumab.^3,4 However, in one series, the authors state that in 4 patients lymphocytes reconstituted more rapidly than expected (3 months),⁴ which may be a sign of overshooting B cell response. Also, the gadolinium ring-enhancing cerebral lesions, which are seen in other B-cell-mediated diseases involving the CNS, may be further evidence of a B-cell-mediated pathogenesis⁷ and were also reported or depicted in the case series.^3,4 However, ring-enhancing lesions may not be exclusive to B-cell-mediated disease⁸; therefore their value as a biomarker may be limited. Finally, clinical observations support a B-cell-mediated pathogenesis. In one case, plasmapheresis led to a significant reduction of symptoms in the acute phase, which suggests that humoral factors may have played a crucial role in the pathophysiology; interestingly, plasmapheresis in the acute phase of deterioration was also applied in 1 of the 2 cases published in The Lancet Neurology and led to marked clinical improvement.³ In all cases, B-cell-depleting therapy with rituximab was initiated and stabilized the course of the disease, which was also true for 2 patients in 1 of the 2 other case series.³ However, it remains unclear if a further early alemtuzumab course would have been another possible treatment option, as it also results in B cell depletion and according to the literature some patients need more than 2 courses to achieve disease stability.⁹

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 B cell count during alemtuzumab treatment

B cell count at baseline (before alemtuzumab), at month 6 (onset of disease activity), at month 7 (after first course of rituximab), and at month 11 (during therapy with rituximab) in patient 2 (dark, dots); B cell count after 6 months (onset of disease activity) in patient 3 (green, square); and B cell count (±SD) at baseline and months 3, 6, 9, and 12 as reported in the literature⁶ (red, triangle).

Although these observations may support a B-cell-mediated pathogenesis, disease activation could also be explained by a compartment problem, seeing that all patients had received fingolimod in the past. A substantial number of lymphocytes could remain hidden from the intravascular therapeutic effects of alemtuzumab as a result of selective lymphoid sequestration and may later egress from lymph nodes and initiate rebound inflammatory activity, as suggested by a case series on disease activation after switching from fingolimod to alemtuzumab.⁴ Interestingly, all of our patients received fingolimod prior to alemtuzumab, although only one patient switched directly from fingolimod to alemtuzumab (patient 3) and treatment with fingolimod was stopped on day 1 in patient 1. Similarly, 10 out of the 11 published cases^3,4 had also received fingolimod prior to alemtuzumab, undermining this hypothesis.

Also, alemtuzumab has been shown to exacerbate neuromyelitis optica spectrum disorder in single patients. This finding was interpreted as indicative for a suppression of the adaptive immunity and possible activation of monocytes by alemtuzumab, reflecting on the fact that CD52 is also expressed at lower levels on monocytes as well as natural killer cells and macrophages.¹⁰ Alemtuzumab might thus have further effects on the interactions between the adaptive and innate immunity as well as interaction between B and T cell lineages than known so far, which might explain the observed disease activation.

When discussing treatment options with the patient, practitioners should consider not only nonresponse, but also the potential reactivation of disease treated with alemtuzumab. Identifying patients at risk of disease activation using reliable biomarkers (i.e., lymphocyte phenotyping and specific MRI features) remains a major challenge. Finally, the pathogenesis of disease activation after alemtuzumab needs further study, including identification of specific targets of treatment in this subgroup of patients at risk of disease activation.