J Neurol Neurosurg Psychiatry 79:1013-1015 doi:10.1136/jnnp.2007.133520
  • Research paper

Bone-marrow transplantation fails to halt intrathecal lymphocyte activation in multiple sclerosis

  1. T Mondria1,
  2. C H J Lamers2,
  3. P A W te Boekhorst3,
  4. J W Gratama2,
  5. R Q Hintzen1
  1. 1
    Department of Neurology, Erasmus MC, Rotterdam, The Netherlands
  2. 2
    Department of Medical Oncology, Erasmus MC, Rotterdam, The Netherlands
  3. 3
    Department of Haematology, Erasmus MC, Rotterdam, The Netherlands
  1. Dr R Q Hintzen, Department of Neurology, Erasmus University Medical School, Erasmus MC, PO Box 2040, 3000 CA Rotterdam, The Netherlands; r.hintzen{at}
  • Received 30 August 2007
  • Revised 30 November 2007
  • Accepted 30 December 2007
  • Published Online First 25 January 2008


Background: Given the presumed key role for autoreactive lymphocytes in multiple sclerosis (MS), treatment strategies have been developed to ablate lymphocyte activity. Intrathecal lymphocyte activation can be measured by CSF-soluble(s)CD27.

Objective: To determine the effect of maximum whole-body immune ablation on two different markers that detect lymphocyte activation in CSF—oligoclonal IgG bands and levels of CSF-sCD27.

Design, setting and patients: The study quantified sCD27 levels and assessed the presence of oligoclonal IgG bands in CSF samples of secondary progressive patients with MS treated by autologous bone-marrow transplantation. In eight individuals, CSF was taken before and 6–9 months after conditioning. CSF-sCD27 levels were compared with other MS and non-inflammatory neurological disease controls. Regarding the effect of stem-cell transplantation on CSF oligoclonal bands, the study analysed pooled data of this and four other international studies on stem-cell transplantation in MS.

Results: CSF-sCD27 was significantly lower after the extremely immunoablative protocol. However, levels remained elevated compared with non-inflammatory controls and stayed within the range observed in other MS controls. The joint analysis of CSF oligoclonal bands demonstrated persistence of this immune abnormality in 88% of the reported cases (n = 34).

Conclusions: The persistence of CSF lymphocyte activation markers sCD27 and intrathecal oligoclonal IgG bands after maximum immunoablative treatment indicates that complete eradication of activated lymphocytes from the CNS has not been established. This is paralleled by disease progression observed in several studies on the effect of stem-cell transplantation in MS.

Because of a presumed central role for autoreactive lymphocytes in multiple sclerosis (MS),1 strategies have been developed to treat this disease via lymphocyte ablation.2 3 Treatment results have been equivocal at best,48 partly because of selecting patients that have already reached the slowly progressive neurodegenerative stage of the disease and possibly because of limited possibilities to establish complete lymphocyte deletion.3 9 Even the strongest treatment protocols result in a mere log 3–4 reduction of lymphocytes inside a human body, which translates into an absolute reduction of a physiological presence of 1012 to a residual amount of 108 or 109 lymphocytes.10 Among these surviving cells, disease-mediating autoreactive clones could still survive, especially in niches such as intrathecal sites. Nevertheless, immunoablative strategies may specifically influence certain lymphocyte subsets and may select and foster the outgrowth of regulatory T cells that appear to be less susceptible to immunosuppression, which could still result in an effective reduction of autoimmune activation in MS.3

We recently treated 14 patients with MS with a unique protocol for maximum immunoablation.9 Here, we test the effectiveness of this treatment in reducing both the lymphocyte activation marker sCD27,1115 and the presence of oligoclonal IgG bands in the intrathecal compartment.


After informed consent, 14 patients with secondary progressive MS received an autologous bone-marrow transplantation according to the Rotterdam protocol.9 Inclusion criteria were: age 18–50 years, EDSS of 5.0–7.0 with an EDSS of at least 3.0 at 2 years after diagnosis and an EDSS increase of more than 1.0 in the previous 2 years if EDSS was 5.5 or an EDSS increase >1.5 if EDSS was 5.0. Follow-up was performed by clinical evaluation, brain and spinal cord MRI and CSF analysis.9 CSF samples were taken by lumbar puncture at baseline and between 6 and 9 months follow-up. We obtained pre-transplantation samples in 10 of the 14 transplanted patients (lumbar puncture refused by 4 patients) within 3 months before conditioning. A follow-up sample, obtained between 6 and 9 months after conditioning, was available in 8 patients. CSF samples obtained from 12 patients with relapsing remitting multiple sclerosis (MS) and 17 patients with other non-inflammatory neurological disease (OND) (headache and neurodegenerative disease) were used as controls. All samples were obtained during the same time period as the CSF samples taken from the trial patients after informed consent. sCD27 concentrations in CSF were quantified with an enzyme-linked immunosorbent assay kit (CLB, Amsterdam, The Netherlands).13 Presence of CSF oligoclonal bands and IgG indices were tested using standard methodology.16


Mann–Whitney U test was used to compare CSF-sCD27 concentrations between the different patient groups (bone-marrow transplantation (BMT) patients, MS controls and OND controls). Levels of CSF-sCD27 before and after transplantation were compared using paired t test. Spearman’s rank test was used to test the correlation between the sCD27 concentration, clinical scores and other laboratory tests.


CSF-sCD27 levels in different patient groups

Baseline CSF-sCD27 levels in the 10 BMT patients (median 112, range 6–480) were significantly higher (p<0.01) than those of the OND control group (median 4.0, range 0.0–21 U/ml), in accordance with previous findings.13 14 No significant difference was found between the MS control group (median 54.5, range 20–280 U/ml) and the BMT patients that were selected for highly aggressive disease courses (p = 0.49).9

Baseline CSF-sCD27 levels and clinical characteristics

We compared CSF sCD27 levels at baseline with several disease characteristics of the BMT group prior to treatment. No correlations were found between the CSF sCD27 levels at baseline and parameters such as EDSS at baseline, number of relapses and EDSS progression in the year preceding BMT. In addition, there was no correlation between baseline CSF sCD27 levels and the course of disease after BMT, reflected by exacerbation rate and EDSS progression.

CSF-sCD27 levels after bone-marrow transplantation

In the 8 patients for which paired pre- and post-BMT samples were available, CSF sCD27 levels dropped significantly after transplantation, from a pre-BMT median value of 112 (range 6–408) to a post-BMT median of 55 (range 20–138) (p = 0.02), as depicted in figure 1. CSF-sCD27 levels post-BMT were similar to those observed in the MS control group, indicating ongoing inflammatory activity (p = 0.54). CSF-sCD27 levels before and after transplantation did not correlate with CSF IgG levels or CSF monocyte cell counts.

Figure 1 Distribution of CSF-sCD27 concentrations expressed as units per ml (U/ml) in the different groups: non-inflammatory, neurological disease (OND), MS control patients (MS) and the MS transplant group before (Baseline) and after (post-BMT) treatment. A horizontal line represents the median for each group. A significant decrease in sCD27 level was seen after treatment (Baseline vs. post-BMT; p = 0.02), but levels were similar to those observed in the MS control patients (MS vs post-BMT; p = 0.54), indicating ongoing inflammatory activity.

Patients with CSF oligoclonal bands had higher CSF-sCD27 (median 104, range 20–408) than those without (median 50, range 6–136) (not significant, p = 0.07).

Intrathecal oligoclonal bands persist after maximum immunosuppression

In line with these indications of ongoing intrathecal inflammatory activity even after severe immunosuppression, we observed that in 5 out of 6 patients (83%) with oligoclonal bands present in CSF before BMT, these bands persisted in the period post-BMT. In one case without oligoclonal bands at baseline, these bands appeared after transplantation. Signs of ongoing B-lymphocyte activity after immunoablative conditioning were also observed in several other studies.6 8 17 18 In table 1, we demonstrate a group analysis of the presence of oligoclonal bands in CSF before and after conditioning in several other small studies. Although conditioning protocols in the studies differed, it can be concluded that none of the strategies applied thus far succeeded in eliminating intrathecal production of oligoclonal bands.

Table 1 Persistence of cerebrospinal fluid oligoclonal bands after stem-cell transplantation in multiple sclerosis


The clinical outcome of this strong conditioning protocol on selected patients with aggressive MS has been described in detail elsewhere.9 The protocol aimed at rigorous T-cell ablation, including total body irradiation, high-dose cyclophosphamide and administration of ATG, followed by infusion of autologous bone-marrow-derived purified CD34+ cells, which were maximally depleted of T cells. The effects on circulating immune cells were strong, with prolonged suppression of circulating CD4 and CD8 lymphocytes for more than 1 year.19 Disease activity, as measured with contrast MRI, showed no contrast-enhancing lesions post-treatment on sequential MRI scans until 36 months follow-up. Regarding the clinical course of MS, less strong effects were observed; 9 out of 14 patients worsened.9

CD27 is shed by B and T cells at activation, thus providing a tool to monitor immune activation.1114 Raised concentrations of sCD27 have been reported in various immunopathological conditions.12 15 20 21 Soluble CD27 has been applied as a marker to monitor immune activation, as well as to evaluate the therapeutic effects of immunosuppressive and antiviral treatment.21 In this study, we confirmed our earlier findings,12 13 showing significantly higher levels of CSF-sCD27 in MS. We did not measure serum sCD27 levels, as we have previously demonstrated that, in patients with MS, these levels do not differ from controls.12 The main observation of this study is that CSF-sCD27, a robust marker of intrathecal lymphocyte activation, was significantly lowered by our extreme conditioning protocol, which also included irradiation of the neuraxis.9 However, despite this intensified and long-acting immunosuppressive regimen, the levels of CSF sCD27 remained elevated compared with non-inflammatory controls. These findings imply that the treatment did not lead to the complete eradication of activated lymphocytes in the brain. This conclusion is supported by the observation here and elsewhere, that following different protocols of immunosuppressive conditioning, activated B-lymphocytes persist in the brain, as reflected by the persistence of oligoclonal IgG molecules in CSF.6 8 9 17 18 Combined analysis of these studies shows that, following immunosuppressive conditioning, oligoclonal bands in the CSF persisted in 88% of the reported cases.6 8 9 17 18 Thus, we can conclude that it has not been possible to completely eradicate intrathecal immune activation in MS. This is corroborated by the persistence of disease, as reflected by histopathology,22 MRI and clinical course, as observed in the longer follow-up series.7

It appears that autologous stem-cell transplantation results in only partial suppression of inflammation, and that demyelination and axonal damage remain active. This was clearly associated with macrophage/microglial activation in lesions and the normal-appearing white matter. Despite a strong depletion of T cells after AHSCT, Metz et al. still found substantial numbers of CD8+ T cells.22 These cells express and shed CD2723 and may be important effectors of further axonal damage. Thus, enhanced innate immune activation after AHSCT may still be accompanied by some low-grade adaptive immune activation. This is in line with at least some remaining relapse activity that is frequently observed after AHSCT.7 9 17 24

Soluble CD27 can also be released from activated B lymphocytes, which are largely CD27-positive in CSF.25 It is unclear whether persisting B-cell clones after AHSCT are directly involved in the ongoing CNS pathology or whether they are ignorant bystander Ig-producing clones that have escaped the strong immunosuppressive regimen.

Whatever the precise roles of the different innate and adaptive immune arms, serious doubt remains as to whether we will be able to influence these compartmentalised cells via the conditioning protocols chosen. After the initial enthusiasm that it would be possible to completely renew the immune system in CNS autoimmune diseases, including inside the CNS, a more temperate position is warranted.

Possibly, the future lies in an approach with less aggressive therapies, with more emphasis on immune regulation rather than immune eradication.


This study was supported by grants from MS Research Netherlands (RQH), the Netherlands Organisation for Scientific research (ZON-MW, RQH) and Erasmus MC, Rotterdam.


  • Competing interests: None declared.

  • Funding: This study was supported by grants from MS Research Netherlands (RQH), the Netherlands Organisation for Scientific Research (ZON-MW, RQH) and Erasmus MC, Rotterdam.

  • Ethics approval: Ethics approval was obtained.


Visit the full archive of podcasts for JNNP here >>

Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of JNNP.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.