Vitiligo after alemtuzumab treatment

Results

In September 2016, a 31-year-old woman presented with depigmentation of her skin following therapy with alemtuzumab. She had been diagnosed with RRMS in 2004 with typical findings on brain MRI and positive oligoclonal bands in the CSF. Despite receiving different immunomodulatory treatments including IFN-β-1a and natalizumab, she demonstrated ongoing disease activity on MRI and several relapses with an EDSS progression to 3.5. In terms of compassionate use, a first course of alemtuzumab was given in June 2012. In March 2013, the patient presented with a worsening of preexisting right-sided sensorimotor hemiparesis responsive to IV methylprednisolone. Since alemtuzumab was retained from the market at this time, she received 2 courses of rituximab 1,000 mg in August 2013 and April 2014. However, she experienced 3 further relapses within 1 year and was therefore switched back to alemtuzumab, receiving the second course in October 2015. Since then, clinical and MRI disease course remained stable. Skin depigmentation started in September 2016 with circumscribed, symmetrical white macules and patches typical for vitiligo (figure, A), which slowly progressed, now involving 30% of the body surface. Skin biopsy confirmed diagnosis showing complete loss of epidermal pigmentation with absence of melanocytes at the basal layer and a subtle perivascular dermatitis consisting of CD3⁺ and CD8⁺ T cells (figure, A), the latter expressing T-cell-restricted intracellular antigen-1, while CD20⁺ B cells were almost absent (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure CD8+ T-cell-driven pathology in alemtuzumab-treated patients with relapsing-remitting multiple sclerosis (RRMS) developing vitiligo

(A) Patient 1: circumscribed, symmetrical white macules and patches typical for vitiligo (A.a). Exemplary skin biopsy from the same patient: melanocyte staining (A.b), panMel antibody cocktail directed to tyrosine, melanA, HMB45. Positive stainings for CD3 (A.c) and CD8 (A.d). (B) Patient 3: typical skin lesions for vitiligo (B.a). Exemplary skin biopsy from the same patient: melanocyte staining (B.b). Positive stainings for CD3 (B.c) and CD8 (B.d). Scale bars represent 100 μM in A and B. (C) Proportions of CD8⁺ T cells and CD8⁺ T-cell subsets (defined by the displayed markers) within peripheral blood mononuclear cells (PBMC) isolated from alemtuzumab-treated patients with RRMS (n = 30 for baseline, n = 29 at 6 months, n = 25 for 12 months, n = 17 for 18 months) and patient 2 (red triangle) were determined by flow cytometry. For the analysis of cytokine production, cells were treated with leukocyte activation cocktail (BD) for 4 hours. The red arrow indicates the timepoint of vitiligo onset. Boxplots show median, 25% and 75% percentile, whiskers represent 5% and 95% percentile. (D) Unique clones, clonality, as well as maximum frequency of distinct clones and top 5 clones in the CD8 T-cell receptor (TCR) repertoire was analyzed by deep sequencing. Prior to analysis, CD8⁺ T cells were magnetic-activated cell sorted from PBMC of healthy controls (ctrl, n = 10), treatment-naive patients with RRMS (MS naive, n = 11), and patient 2 at baseline (vitiligo). RNA of at least 2 × 10⁶ CD8⁺ T cells from patient 2 at baseline was purified and cDNA was sequenced by Adaptive Biotechnologies, resulting in a range from 1.5 to 2.5 × 10⁶ total reads in each analyzed sample. (E) 3D histograms show expansion of single clones in patient 2 in comparison to a representative treatment-naive patient with RRMS (i.e., sex- and age-matched and exhibiting a similar mean number of relapses and mean EDSS progression in the last 2 years as compared to patient 2). The x axis lists all Vβ genes, the z axis the Jβ genes, and the column height indicates the total reads of this specific V/J combination. (F) Clone tracking of the top 10 expanded clones (highest number of total reads) of patient 2 before and 18 months after alemtuzumab treatment. The proportion of clones in the whole TCR repertoire is depicted. HLA = human leukocyte antigen; IFN-γ = interferon-γ; TNF-α = tumor necrosis factor–α.

A second patient, a 34-year-old man, was referred to our clinic with skin lesions suspicious for vitiligo in February 2017 after having received courses of alemtuzumab in August 2015 and 2016. He was diagnosed with MS in 2001, fulfilling McDonald criteria with typical findings on MRI and in the CSF. Several immunomodulatory treatments including IFN-β-1a, natalizumab, fingolimod, and dimethyl fumarate were ineffective. Since alemtuzumab initiation the patient has demonstrated clinical and radiologic stable disease. However, in February 2017, he reported patchy, well-defined depigmentation of the skin progressively affecting the whole integument in September 2017. Dermatologic examination established the diagnosis of vitiligo due to characteristic clinical presentation and disease course; the patient refused a skin biopsy.

A third patient, a 42-year-old woman diagnosed with MS in August 2015, was treated with alemtuzumab in the University Hospital Essen in February 2017 and 2018 due to ongoing disease activity with dimethyl fumarate and fingolimod. Fourteen months after treatment initiation, she reported vitiligo-characteristic depigmentation of the skin starting from the extremities and then affecting the whole body including the face (figure, B). Skin biopsy confirmed the diagnosis, showing interface dermatitis with predominant CD3⁺CD8⁺ lymphocyte infiltration along the basal lamina and missing melanocytes in central parts of the biopsy specimen (figure, B).

The pathophysiologic mechanisms underlying autoimmunity secondary to alemtuzumab remain insufficiently understood. However, currently it is assumed that B cells are the central drivers of these autoimmune conditions due to predominance of antibody-mediated autoimmune disorders.^5,6 In contrast, vitiligo is a T-cell-driven autoimmune disorder, where autoreactive, melanocyte-specific CD8⁺ T cells are recruited to the skin and target melanocytes.⁷

As patient 2 is part of a prospective biobanking cohort, we were able to analyze PBMCs for vitiligo-related immune patterns. Interestingly, he showed high proportions of CD8⁺ T cells with an activated (HLA-DR⁺), memory (CD45RO⁺), and type 1 cytokine (IFN-γ⁺ and TNF-α⁺) phenotype at vitiligo onset compared to a control cohort of alemtuzumab-treated patients with RRMS (figure, C). Analysis of CD8 TCR repertoire in patient 2 at baseline revealed a highly increased clonality and reduced repertoire diversity (figure, D) as compared to healthy controls and treatment-naive patients with RRMS. Furthermore, a predominance of single clones could be observed at baseline in this patient (figure, E), illustrated in a 3D histogram, which provides an overview of the clonal distribution within the TCR repertoire. Of note, alemtuzumab treatment did not substantially affect the proportions of the top 10 clones over time (figure, F).