Article Text

Concise report
Takayasu arteritis is characterised by disturbances of B cell homeostasis and responds to B cell depletion therapy with rituximab
  1. Bimba F Hoyer1,2,
  2. Imtiaz M Mumtaz1,2,
  3. Konstanze Loddenkemper1,
  4. Anne Bruns1,
  5. Claudia Sengler3,
  6. Kay-Geert Hermann4,
  7. Sofiane Maza5,
  8. Rolf Keitzer3,
  9. Gerd-Rüdiger Burmester1,
  10. Frank Buttgereit1,
  11. Andreas Radbruch2,
  12. Falk Hiepe1,2
  1. 1Department of Medicine (Clinic for Rheumatology and Clinical Immunology), Charité-Universitätsmedizin Berlin, Berlin, Germany
  2. 2German Rheumatism Research Center, Berlin-Leibniz Institute, Germany
  3. 3Department of Pediatric Rheumatology, Charité-Universitätsmedizin Berlin, Berlin, Germany
  4. 4Department of Radiology, Charité-Universitätsmedizin Berlin, Berlin, Germany
  5. 5Department of Nuclear Medicine, Charité-Universitätsmedizin Berlin, Berlin, Germany
  1. Correspondence to Falk Hiepe, Department of Medicine, Clinic for Rheumatology and Clinical Immunology, Charité–Universitätsmedizin Berlin and German Rheumatism Research Center Berlin-Leibniz Institute, Charitéplatz 1, D-10117 Berlin, Germany; falk.hiepe{at}


Introduction Takayasu arteritis (TA) is a large vessel vasculitis involving the aorta and its major branches. T cell-mediated autoimmunity is thought to play a major role in its pathogenesis, while the role of B cells is still unclear.

Methods B cell subsets in the peripheral blood of 17 patients with TA were analysed and compared with nine patients with active systemic lupus erythematosus (SLE) and nine healthy controls by flow cytometry. Based on these findings, three patients with active refractory TA were treated with B cell depletion therapy (BCDT) using monoclonal anti-CD20 antibodies (rituximab).

Results The absolute number and frequency of peripheral blood CD19+/CD20/CD27high antibody-secreting cells in patients with active TA was significantly higher than in healthy donors. As in active SLE, the majority of these cells are newly generated plasmablasts which significantly correlated with TA activity. Three patients with active refractory TA and expansion of plasmablasts were successfully treated with BCDT, which resulted in remission.

Conclusion Disturbances of B cell homeostasis may be critical in TA. Circulating plasmablasts could be a useful biomarker of disease activity and a tool for selecting appropriate candidates for BCDT. B cells and plasmablasts/plasma cells may therefore represent novel targets for effective therapies for TA.

Statistics from


Takayasu arteritis (TA) is a rare form of chronic large vessel vasculitis involving the aorta and its major branches. In contrast to giant cell arteritis, TA most commonly affects young women and has a higher incidence in Asia and Eastern Europe. Treatment consists of glucocorticoids, which may be supplemented by other immunosuppressive agents and/or tumour necrosis factor (TNF) blockade in refractory patients.1 2 The aetiology of TA is unknown but highly suggestive of T cell-mediated autoimmunity.3 The presence of antiendothelial antibodies and B cell infiltrates in inflamed vessels indicates a pathogenic role of B cells.4 5

In this paper we describe for the first time B cell disturbances leading to increased numbers of newly generated plasmablasts in the peripheral blood and correlating with disease activity. In three patients with highly active refractory TA with plasmablast expansion, B cell depletion therapy (BCDT) resulted in clinical remission. These findings highlight the need to reconsider the role of B cells in TA.



Seventeen women with TA fulfilling the American College of Rheumatology (ACR) criteria6 were analysed. The median age at disease onset was 25 years (range 12–40), median disease duration was 90 months (range 20–266) and median age at analysis was 39 years (range 15–63). Disease activity was evaluated based on National Institutes of Health (NIH) criteria7 and 18F-fluorodeoxy-glucose positron emission tomography. Involved areas were classified according to the revised classification from 1997.8 Detailed characteristics of the patients are listed in table S1 in the online supplement. Nine women with systemic lupus erythematosus (SLE) fulfilling the ACR classification criteria were also analysed. They had a median age of 32 years (21–69) at the time of analysis and high SLE Disease Activity Index scores of disease activity (median 10, range 8–27). The healthy donors (HD) comprised eight women and one man of median age 35.5 years (range 26–57).

B cell depletion therapy

Three patients were treated with rituximab according to the protocol established for rheumatoid arthritis. Ongoing immunosuppressive therapy was continued in each case (see data in online supplement).

Analysis of circulating B cells by flow cytometry

B cells were isolated from heparin whole blood by lysis using red blood cell lysis buffer (Becton Dickinson, Mountain View, California, USA according to the manufacturer's instructions. Staining of remaining cells in phosphate-buffered saline/EDTA was performed as previously described.9 Cells were analysed for expression of CD20 (PerCP, Clone L27, Becton Dickinson), CD19 (PE, Clone HD37, Dako Cytomation),CD27 (Cy5, Clone 2E4, DRFZ, Berlin, Germany) and HLA-DR (FITC, Clone TU36, Becton Dickinson) as measured by flow cytometry (FACS Calibur, Becton Dickinson) using FlowJo software (TreeStar, Ashland, Oregon, USA). Lymphocytes were gated by forward and side scatter. B cells were defined as CD19-positive lymphocytes. Antibody-secreting cells were classified as CD19+/CD20–/CD27high. Differentiation between plasmablasts and plasma cells was based on their expression of HLA-DR (newly generated plasmablasts express high levels of HLA-DR whereas plasma cells do not).10 Memory and naive B cells were defined based on their expression of CD27 and positivity for CD19 and CD20. Calculations were based on absolute lymphocyte numbers from routine blood cell counts from the same day.

Statistical analysis

Unless otherwise indicated, the Mann–Whitney test and Pearson correlation were used to detect significant differences using Prism5 (GraphPad, San Diego, California, USA). p Values ≤0.05 were considered significant.


Increased number of circulating plasmablasts in active TA

Analysis of B lymphocytes showed an increased number and frequency of CD19+/CD20/CD27high antibody-secreting cells in patients with TA (figure 1A, table 1). About 80% of these cells expressed high levels of HLA-DR, which is characteristic of newly generated plasmablasts. This strongly resembles the findings in SLE, where plasmablast expansion serves as a marker of disease activity.9 We therefore investigated whether plasmablast numbers also correlate with disease activity in TA. Most patients with active TA (≥2 NIH criteria) indeed had higher numbers of plasmablasts than those with inactive disease (figure 1B). TA activity correlated with the number of plasmablasts (figure 1C) and, as expected, with erythrocyte sedimentation rate (ESR). There was also a significant correlation between C-reactive protein and ESR.

Figure 1

Disturbances of B cell homeostasis in patients with Takayasu arteritis (TA). (A) Cytometry analysis of peripheral blood mononuclear cells (PBMC) in a patient with active TA compared with those in a patient with active systemic lupus erythematosus (SLE) and a healthy donor. Boxes on the left show CD27 expression on CD19+ B cells. CD20 and CD27 expression is shown on CD19+ lymphocytes (middle boxes). The red circles in the middle boxes emphasise the CD19+/CD20/CD27high cells. Boxes on the right represent a contour plot of lymphocytes stained with CD27 versus HLA-DR. The red plots depict CD19+/CD20/CD27high/HLA-DR+ cells representing newly generated plasmablasts. (B) Patients with active TA and SLE had an increased number of circulating plasmablasts. The numbers of plasmablasts in patients with active TA (National Institutes of Health (NIH) score ≥2) were significantly higher than those in healthy donors and patients with inactive TA. Horizontal bars represent median values. (C) Significant correlations were found between TA activity and the absolute number of plasmablasts (Pearson r=0.5188, p=0.0329).

Table 1

Absolute numbers of B lymphocyte subpopulations in patients with TA compared with HD and patients with SLE

Unlike active SLE which is characterised by lymphocytopenia,11 total lymphocyte and B cell counts in patients with TA were not significantly lower than those in HD. However, B cell subsets were considerably altered in active TA. The frequency of memory B cells was significantly higher in active TA (49%) than in HD (29.02%), whereas the number and frequency of naive B cells was lower (49.5% vs 68.9%, respectively). Patients with inactive TA had normal naive B cell counts, the only discrepancy being that the memory B cell pool was diminished (table 1).

B cell depletion as a therapeutic option in refractory TA

Taking into account these disturbances of B cell homeostasis with expansion of newly generated plasmablasts, BCDT using rituximab emerged as a new therapeutic option in patients with TA refractory to all previous treatments. Three patients (nos 11, 12 and 13, figure 2 and online supplementary figure S1) with refractory TA and expansion of circulating plasmablasts have so far been successfully treated with BCDT (see online supplement for case reports).

Figure 2

Effect of B cell depletion therapy on vessel imaging. (A) Patient #11: MR angiogram from 2006 showing typical changes such as stenosis and aneurysm of large vessels including the aorta and its main branches (left, white arrow). 18F-fluorodeoxy-glucose positron emission tomography performed before the start of rituximab therapy in 2005 showing enhanced glucose uptake of the subclavian and femoral arteries and diffuse enhancement of the aorta (middle, arrows). In contrast, no such enhancement was seen on images taken after the second cycle of rituximab (right).


This study shows for the first time that B cell disturbances can be found in the peripheral blood of patients with TA and that BCDT may be beneficial in refractory patients. Patients with active TA had markedly increased levels of CD19+/CD20/CD27high/HLA-DR+ B cells, corresponding to newly generated plasmablasts. As previously described,9 11 these disturbances are comparable to those found in patients with active SLE whom we used as controls. Patients with inactive TA and SLE do not exhibit this plasmablast expansion. Whereas SLE is associated with pronounced B cell lymphocytopenia,11 there was no significant decrease in B cell numbers in TA.

The observed changes in the other B cell subpopulations from patients with active TA were interpreted as the result of plasmablast expansion; this reflects B cell hyperactivity, which can be induced by cytokines. Elevated interleukin 6 (IL-6) and B cell-activating factor (BAFF) levels, which have been described in active TA,12 13 may stimulate B cells to differentiate into plasmablasts. Clinical trials have shown that inhibition of BAFF levels in active SLE leads to a striking reduction of these expanded plasmablasts, which is accompanied by a clinical response.14

The observed correlation between the plasmablast numbers and active TA, as well as the beneficial effects of BCDT, suggest a potential of plasmablasts as biomarkers of disease activity and for interventions targeting B cells. BCDT interrupts the generation of plasmablasts as plasma cell precursors, which are short-lived. Interestingly, relapse was accompanied by a renewed increase in plasmablast numbers in the two patients successfully treated with a second cycle of rituximab.

These findings suggest that B cells could play a critical role in the pathogenesis of TA. Interestingly, immunohistochemical analyses of aortic wall samples from patients with TA showed that, besides T cells, B cells were among the most prominent cells in the inflamed adventitia.5 Moreover, autoantibodies to endothelial cells (antiendothelial cell antibody (AECA)) and hypergammaglobulinaemia have also been detected in patients with active TA.4 15 A recent publication shows that circulating B cells isolated from the blood of patients with TA secrete antiendothelial antibodies in vitro, correlating with disease activity.16 In our patients investigated with the only commercially available test for AECA (Euroimmun, Lübeck, Germany), antibodies did not correlate with plasmablast counts nor disease activity. Further characterisation of the expanded plasmablast population may contribute to the identification of pathogenic autoantibodies.

To the best of our knowledge, this is the first extensive report documenting the successful treatment of TA with rituximab. Remarkably, these patients had previously developed resistance to agents described as effective options for the treatment of TA, including TNF blockers. Apparently, two previous cases were treated successfully with rituximab,17 but detailed information about the patients is lacking. One other case report describes the efficacy of rituximab as an adjunct in the treatment of polymyalgia rheumatica/giant cell arteritis.18 Interestingly, a successful case of the inhibition of IL-6 (which activates B cells) by the monoclonal anti-IL-6 receptor antibody tocilizumab in a patient with TA has already been reported.12

The effect of BCDT might not be limited to B cells and their differentiation to autoantibody-secreting plasmablasts and plasma cells. It has become increasingly evident that B cells influence T cell activation and the production of proinflammatory cytokines.19

Our findings indicate that the pathogenic role of the B cell compartment in TA, which was previously thought to have a mainly T cell-mediated pathogenesis, must be reconsidered. The number and frequency of plasmablasts in patients with TA might be useful biomarkers for disease activity and for identifying appropriate candidates for BCDT. BCDT seems to be a useful option for refractory TA, and its potential should be evaluated in controlled trials. Other factors influencing B cell differentiation and plasma cells may also emerge as novel therapeutic targets.20


The authors thank Ann Carolin Longardt, Capucine Daridon, Thomas Dörner and Henrik Mei for excellent technical assistance, critical reading of the manuscript and fruitful discussions.


Supplementary materials


  • Funding This work was supported by the Deutsche Forschungsgemeinschaft (SFB650) and Roche Pharma AG provided financial support for writing the manuscript.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval This study was conducted with the approval of the ethics committee of Charité-Universitätsmedizin Berlin.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.