Article Text


Rituximab for polyneuropathy with IgM monoclonal gammopathy
  1. J M F Niermeijer1,
  2. M Eurelings1,
  3. H L Lokhorst2,
  4. W-L van der Pol1,
  5. H Franssen1,
  6. J H J Wokke1,
  7. N C Notermans1
  1. 1
    Department of Neurology and Rudolf Magnus Institute of Neurosciences, University Medical Center, Utrecht, The Netherlands
  2. 2
    Department of Haematology, University Medical Center, Utrecht, The Netherlands
  1. Correspondence to Dr J M F Niermeijer, Department of Neurology and Rudolf Magnus Institute of Neurosciences, University Medical Center Utrecht, Room C03.236, PO Box 85500, 3508 GA Utrecht, The Netherlands; j.m.f.niermeijer{at}


Background: Polyneuropathy with IgM monoclonal gammopathy can be a disabling disorder necessitating treatment.

Methods: In a prospective open label trial, 17 patients with disabling IgM MGUS polyneuropathy were treated with rituximab, a chimeric anti-CD-20 monoclonal antibody.

Results: Rituximab induced an improvement of ⩾1 point on the Overall Disability Sum Score in 2/17 patients, an improvement of ⩾5% of the distal MRC sum score in 4/17 and the sensory sum score in 9/17 patients. Bone marrow investigations showed CD 20 B cell depletion in all patients. There were no serious adverse events. Compared with treatment with intermittent cyclophosphamide with prednisone or treatment with fludarabine, it shows a comparable response percentages but fewer side effects. The presence of anti-MAG and a disease duration shorter than 10 years may predict treatment response.

Conclusion: Rituximab is a candidate for treatment of IgM MGUS polyneuropathy and should be further investigated in a double-blind randomised trial.

Statistics from

Polyneuropathy associated with IgM monoclonal gammopathy (IgM MGUS) is a disabling immune-mediated neuropathy. Evidence-based effective treatment strategies are still lacking.1

Rituximab is a chimeric monoclonal antibody directed at the CD 20 surface antigen found on most normal and neoplastic B cells. It is efficacious for the treatment of some haematological malignancies2 and inflammatory diseases3 including a number of neurological disorders such as multiple sclerosis.4 Open label trials and preliminary results of a placebo-controlled trial suggest that it may be of value for the treatment of IgM MGUS polyneuropathy.5 6 7 8

In this prospective open label trial, we studied if rituximab safely induces improvement of disability scores, muscle strength and sensory function, and depletion of CD20-positive B cells in bone marrow in IgM MGUS polyneuropathy.


A prospective open label trial was performed in a tertiary referral centre for polyneuropathy.

The study was approved by the local medical ethics committee, and written informed consent was given by all participants.

Seventeen patients (three female, six anti-MAG antibodies) with disabling or progressive symmetric sensorimotor polyneuropathy associated with IgM monoclonal gammopathy were included. Progression was defined as ⩾1 point deterioration on the Modified Rankin Scale (MRS) or Overall Disability Sum Score (ODSS),9 or ⩾5% deterioration of the distal MRC sum score or sensory sum score in 6 months. Disabling was defined as ⩾2 points on the MRS.

The median age of onset was 55 years (interquartile range (IQR) 50 to 60 years) and median disease duration 7 years (IQR 4 to 12 years).

Five patients had been previously treated for their polyneuropathy: patients 3, 10, 13, 16 and 17 with intermittent cyclophosphamide and prednisone, and patients 16 and 17 with fludarabine. As a previous treatment had resulted in stabilisation or improvement, and was given more than 4 years previous to rituximab treatment, they were considered as candidates for this trial.

Treatment consisted of intravenous rituximab 375 mg/m2 once weekly for 4 weeks, and the median follow-up was 12 months (IQR 12 to 26 m).

Nerve conduction studies were performed before and 9 months after treatment as described elsewhere.10 Polyneuropathies were classified as demyelinating if features of demyelination were found in at least two nerve segments and as axonal if otherwise.11

Bone marrow biopsies and aspirates were stained with May–Grunwald–Giemsa, and with haematoxylin–eosin and Congo Red for the detection of amyloid deposits. Immunophenotyping with CD138 (plasma-cell-specific), CD19 and CD20 (B-cell-specific) and IgM kappa and lambda surface antibodies was performed. The monoclonal B cell population was identified by a predominance of kappa or lambda with a ratio >3:1. The Kyle definition of MGUS was used.12

Outcome measures

The primary outcome measure was ⩾1 point improvement of the ODSS. The secondary outcome measures were: (1) ⩾1 point improvement of the Modified Rankin Scale (MRS, (2) ⩾5% improvement of the distal MRC sum score (a composite score of eight distal muscle groups in arms and legs calculated into a percentage of the maximum score of 80 points) or (3) ⩾5% improvement of the sensory sum score (a composite score of touch, pinprick, vibration and position sense, calculated into a percentage of the maximum score of 56 points),10 (4) disappearance of CD 20 positive B cells in bone marrow biopsy or ⩾50% decrease in M protein concentration (haematological response), (5) ⩾10% improvement of conduction velocity (NCV), (6) adverse events.

Statistical analysis

Outcome measures of first and last visit were compared by the Wilcoxon matched pairs test. Pooled data analysis for characteristics associated with treatment response was performed using the Mann–Whitney U test and χ2 test. p<0.05 was considered statistically significant.


There were no adverse events, except for one patient with transient facial erythaema.

Primary outcome

The ODSS improved in 2/17, remained unchanged in 14/17 and deteriorated in 1/17 patients. The median ODSS changed from 4 (IQR 2.5 to 5) to 3 (IQR 2 to 5.5) (NS) (table 1).

Table 1

Outcome after treatment with rituximab in 17 IgM MGUS polyneuropathy patients

Secondary outcome

The MRS improved in 5/17 patients. The grouped median MRS changed from 2 (2 to 3) to 2 (1.5 to 2.5) (p = 0.025). The distal MRC sum score improved ⩾5% in 4/17 patients, and improved <5% in 11/17. The grouped median distal MRC sum score increased from 90% to 93% after treatment (p = 0.006). The sensory sum score improved ⩾5% in 9/17 and deteriorated in 4/17 patients. The grouped median sensory sum score of the legs improved from 57% to 71% after treatment (p = 0.03). The median IgM concentration decreased from 6.31 g/l (3.67 to 9.85) before to 3.91 g/l (2.87 to 7.42) after treatment (p<0.001) (table 1).

Bone marrow investigations in 17 patients did not show haematological malignancy. After treatment, 10/13 evaluated bone marrow samples showed complete disappearance of CD 20+ B cells, and 3/13 a significant decrease. The median CD20+ B cell percentage decreased from 5% (2 to 6) before to 0% (0 to 0.5%) after treatment (p = 0.002) (table 1). In all cases, disappearance of CD 20+ B cells coincided with a decrease in CD19+ B cells. In one patient (5), the B cell population was further specified before and after rituximab treatment showing elimination of the CD20+sIgM+B cells leaving predominantly pre- and pro-B cells.

Nerve conduction studies showed ⩾10% improvement of NCV in 4/17 patients (1, 5, 12 and 14) in two or more nerves and in 2/17 patients (6, 7) in one nerve (data not shown).

Comparison with data from previous trials performed in the centre

Response to rituximab treatment was compared with response to double-blind cyclophosphamide with prednisone (CP) (n = 25)10 and open label fludarabine (n = 16)13 treatment in other IgM MGUS polyneuropathy patients. All three treatment strategies resulted in an equal percentage of responders: response on the MRS was 31% after fludarabine, 29% after rituximab and 36% after CP. Response on the MRC sum score was 31% after fludarabine, 24% after rituximab and 32% after CP, on the sensory sum score 56% after fludarabine, 53% after rituximab and 20% after CP. Response of NCV was 43% after fludarabine and rituximab, and 60% after CP. Intermittent CP and fludarabine treatment both resulted in a higher frequency (56% after CP and 25% after fludarabine) and severity of side effects compared with rituximab (6%). After analysis of pooled data, specific patient characteristics associated with treatment response could not be identified. Age of onset, disease duration, number of MAG positive patients and number of patients with a demyelinating EMG were equal between responders and non-responders. (data not shown)


Improvement of the ODSS occurred in 2/17 patients with polyneuropathy associated with IgM monoclonal gammopathy after treatment with rituximab. However, in 11/17 patients the MRC sum score improved, and in 10/17 the sensory sum score improved

This may suggest that improvement in strength or sensory function does not coincide with reduced disability, or that the disability scores we used lack sensitivity. Because the trial was unblinded, the observed improvement of exam scores could be caused by a placebo effect; however, spontaneous improvement has not been observed in IgM MGUS polyneuropathy.

The lack of treatment response may be explained by the long disease duration of more than 10 years in some patients (10, 15, 16, 17). This suggests that early treatment, at the moment when axonal degeneration supposedly is not yet a prominent feature, increases the chances of treatment response. This concept is supported by the finding that improvement of nerve conduction velocity occurred primarily in patients with shorter disease duration, suggesting that nerve dysfunction is partly reversible early in the disease course.

Three out of six patients with anti-MAG IgM showed clear improvement, indicating that anti-MAG antibodies may predict treatment response. However, the small numbers preclude sound statistical analysis.

Deterioration of scores in some patients in this study was not treatment-related, as reported by others,14 15 but, in our opinion, reflects the ongoing deterioration of the polyneuropathy itself.

Elimination of CD 20+B cells occurred in all patients and was paralleled by decreased M protein concentrations. Knowledge emerges that rituximab may be effective through a set of different haematological and immunological mechanisms. Next to depletion of CD 20+B cells to lower the level of circulating auto-antibodies, rituximab may also lower CD 20 negative plasma cells that produce auto-antibodies, through suppression of memory-B cells. It may also directly inhibit binding of autoantibodies to the neural antigen by scavenging the circulating anti-neural antibodies, which may explain early treatment effects.3

There is debate about what dosage of rituximab is most effective for polyneuropathy associated with IgM monoclonal gammopathy.7 Considering the level of suppression of the CD 20 and CD 19 B cell clones in the investigated bone marrow biopsies, and the decline in IgM concentrations, the dosage that was used in this study is presumed to be adequate.

Analysis of pooled data of this trial with those from previous trials examining the effects of intermittent cyclophosphamide combined with prednisone10 and fludarabine13 found no difference in treatment effects, nor identification of predictive factors for good clinical treatment response. Importantly, rituximab treatment was safe and caused fewer side effects than fludarabine and cyclophosphamide with prednisone.

In this open label trial, individual patients experienced benefit by reduction in disability scores, and more than half showed improved neurological examination after rituximab treatment that was well tolerated. Presence of anti-MAG and a disease duration shorter than 10 years may predict treatment response. These data suggest that efficacy of rituximab for polyneuropathy associated with IgM monoclonal gammopathy should be further investigated in a double-blind randomised trial.


View Abstract


  • Competing interests None.

  • Ethics approval Ethics approval was provided by the medical ethics committee of the University Medical Centre, Utrecht.

  • Patient consent Obtained.

  • Statistical analysis was performed by JMFN.

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.