Abstract
In an open prospective study, we analyzed the effect of cyclophosphamide (300 mg/m2 body surface daily for 4 days) combined with prednisone (40 mg/m2 body surface daily for 5 days) at 4-week intervals during 6 months in 16 patients with polyneuropathy associated with monoclonal gammopathy of undetermined significance (MGUS). Eleven patients had an IgM-MGUS and five an IgG-MGUS. During a follow-up period of 3 years, eight patients had improvement and six patients stabilized, based on quantitative neurologic examination, the Rankin disability scale, and electrophysiologic studies. These 14 patients had neuropathy with demyelinating and axonal features. One patient with a purely axonal neuropathy had deterioration despite therapy. One other patient developed severe leukopenia as side effect of cyclophosphamide, necessitating withdrawal of treatment. A difference in response was not present in patients with IgM- or IgG-MGUS, nor in patients with or without autoantibodies against myelin-associated glycoprotein. Nine patients had a bone marrow biopsy before and 1 year after treatment. In eight patients, the monoclonal lymphoid IgM or plasma cell IgG infiltration decreased, while in four the monoclonality disappeared after treatment. In the patient who had neurologic deterioration, repeated bone marrow biopsy showed deposits of amyloid. In conclusion, short-term treatment with intermittent cyclophosphamide and prednisone may have a long-term favorable effect in patients with demyelinating polyneuropathy associated with MGUS.
NEUROLOGY 1996;47: 1227-1233
Polyneuropathies associated with monoclonal gammopathy of undetermined significance (MGUS) are incapacitating disorders that often run a progressive course. [1-5] Evidence that MGUS autoantibody activity causes peripheral neuropathy [6-11] has provided a rational basis for therapeutic intervention directed at lowering the monoclonal (M) protein concentration. [12-25] The optimal treatment regimen in terms of benefit and side effects is unknown Table 1. Treatment with chemotherapy may be preferable as a first choice by reducing the number of monoclonal protein-secreting cells for a long time, resulting in lower levels of the M-protein. Therefore, we performed an open prospective study for 3 years on the effect of treatment with cyclophosphamide and prednisone in 16 patients with polyneuropathy associated with MGUS. Based on experience with nonHodgkin lymphoma and multiple myeloma, patients were treated with six intermittent monthly courses of cyclophosphamide and prednisone. [26] As it is not always possible to quantify the M-protein in serum, we examined the relationship between clinical improvement and decrease of plasma cell proliferation by bone marrow biopsy before and 1 year after treatment.
Table 1. Review of studies on treatment of patients with neuropathy associated with MGUS
Patients and methods.
Patients.
From January 1990 to January 1994, 16 patients (9 men, 7 women) with polyneuropathy and MGUS were admitted to the study. Of these patients, 10 patients were drawn from our natural history study, as the progression of the neuropathy and their disability necessitated treatment. [1] The other six patients were referred to us for treatment after diagnosis in other hospitals.
The following criteria had to be met: (1) polyneuropathy and monoclonal gammopathy not associated with amyloidosis, multiple myeloma, osteosclerotic myeloma, Waldenstrom's macroglobulinemia, or lymphoma [1]; (2) progression of neurologic disability; (3) no other cause of neuropathy [27]; (4) no treatment with plasma-exchange, intravenous immunoglobulin (IvIg), or chemotherapy in the preceding 3 years. All patients gave an informed consent. The investigation was approved by the University Hospital Ethical Committee.
Quantification of the neurologic examination.
The patients were examined before and 1/2, 1, 2, and 3 years after onset of treatment by the same neurologist (N.C.N.), who was blinded to the results of previous testing. The neurologic deficit was quantified as follows: (1) Strength (using the MRC grading system): Six muscles of each arm (deltoid, biceps and triceps brachii, finger extensors, finger flexors, and first dorsal interosseus) and six muscles of each leg (iliopsoas, quadriceps femoris, hamstrings, anterior tibial, gastrocnemius, and peroneal) were tested. Summation of test results could lead to a maximum motor sum score of 120. (2) Sensory function: The touch and pin prick sense: normal = 4, distal to wrist/ankle abnormal = 3, distal half forearm/leg abnormal = 2, distal to elbow/knee abnormal = 1, distal to axilla/groin abnormal = 0. Vibration sense: tuning fork perception (128 Hz) on middle finger/hallux = 4, ulnar styloid/medial malleolus = 3, elbow/knee = 2, clavicle/iliac crest = 1, no perception = 0. Joint position sense of middle finger/hallux: normal = 2, diminished = 1, absent = 0. Summation of all sensory modalities could lead to a maximum sensory sum score of 56. (3) For quantification of ataxia we developed and validated a tapping test. [28] This test uses a device consisting of two pushbuttons placed a fixed distance of 35 centimeters apart, connected to an automatic counter. The patient is asked to push the left and right buttons alternately with the dominant limb as fast as possible. Counting starts as soon as the first button is pushed and stops after 15 seconds. (4) Vibration sense was quantified by measuring the vibration perception threshold (VPT) using a Vibrameter type 3 (Somedic AB, Stockholm, Sweden). [1,29] The measurements were performed on the right and left metacarpal two, while the patient was lying down. (5) Disability was scored with the modified Rankin disability scale. [30]
Electrophysiologic studies.
Electrophysiologic studies were performed before and 1 year after onset of treatment by the same neurophysiologist (H.F.), who was blinded to the results of previous testing. Nerve conduction was investigated by means of surface electrodes using standardized techniques. [31] The median and ulnar nerve (motor, sensory, and F waves) in one arm, the tibial nerve (motor) and the sural nerve in one leg, and the H-reflex of the soleus in the opposite leg were investigated. In all patients concentric needle examination was performed in the anterior tibial muscle and the first dorsal interosseus muscle of the hand. Skin temperature was maintained at 36 degrees C with an infrared heatlamp.
For each patient the following motor conduction variables were measured: distal motor latency (DML); conduction velocity in the lower arm or leg (CV); amplitude, area and duration of the negative part of the compound muscle action potential (CMAP); percentage of amplitude and area reduction on proximal versus distal stimulation (ulnar nerve); percentage of duration increase on proximal versus distal stimulation (ulnar nerve); summated CMAP of ulnar and tibial nerve (Sigma CMAP). [32] The terminal latency index (TLI), defined as distal conduction distance in millimeters/CV in lower arm or leg in meters per second/DML in milliseconds, was calculated to compare the conduction in the distal segment of the nerve with that in the proximal segment. [33,34] The following sensory conduction variables were measured: CV, amplitude of the negative part of the sensory nerve action potential (SNAP); summated SNAP of ulnar and sural nerve (Sigma SNAP). [32]
Axonal degeneration was considered to be present when there was (1) spontaneous muscle fiber activity such as fibrillation potentials, positive sharp waves, or complex repetitive discharges on concentric needle examination, (2) reduction of the amplitude of the CMAP or of the SNAP on distal stimulation. Demyelination was considered to be present when (1) motor or sensory CV was decreased below 70% of the lower limit of normal in our laboratory, [35] (2) DML was increased above 150% of the upper limit of normal in our laboratory, [35] (3) there was evidence of conduction block according to Lange et al., [36] (4) temporal dispersion was abnormal in ulnar nerve motor fibers, [36] (5) the minimal F wave latency minus the M-response latency of 20 F waves was increased above 150% of the upper limit of normal or when F waves were not obtainable. The number of nerves with demyelinating features was compared with the number of nerves with low CMAP amplitude on distal stimulation in order to assess the extent of demyelination with respect to the extent of axonal degeneration.
Laboratory analysis.
Electro- and immunoelectrophoresis was performed before and 1/2, 1, 2, and 3 years after treatment. Electrophoresis of serum was performed on cellulose acetate membrane and the monoclonal component was quantified with densitometry. Immunoelectrophoresis and immunofixation was performed by utilizing monospecific antisera to IgM, IgG, IgA, kappa (kappa), and lambda (lambda). Serum antibodies to MAG were detected by Western blotting as described previously. [37] Serum antibody titers to GM1 ganglioside, sulfatides, and chondroitin sulfate C were measured by ELISA. [38-40]
In all patients bone marrow biopsy and aspiration was performed before treatment and fulfilled our criteria for MGUS. [1] In nine patients repeated bone marrow biopsy was performed 1 year after treatment. Bone marrow biopsies and aspirates from the crista iliaca posterior were obtained with a Yamshidi needle. Bone marrow biopsy specimens were embedded in paraffin and stained with hematoxylin-eosin (HE) and Congo red. Smears from aspirates were stained with May-Grunwald-Giemsa. For immune-phenotyping, cytocentrifuged mononuclear bone marrow cells, separated on Ficoll-Hypaque, were stained with polyclonal rabbit anti-human IgG, IgA, IgM, kappa, and lambda antibodies (DAKO, Santa Barbara, CA).
Sural nerve biopsy.
Before treatment a sural nerve biopsy was performed in all patients. Five patients had findings consistent with demyelination, two with axonal degeneration, and nine with both. In six patients deposition of IgM was demonstrated. Mononuclear cell infiltrates were not present. Congo red staining showed no amyloidosis.
Treatment protocol.
Patients were treated with cyclophosphamide (300 mg/m2 body surface per day orally for four consecutive days once a month, for a period of 6 months) and prednisone (40 mg/m2 body surface per day orally for 5 consecutive days once a month for a period of 6 months). Drug dosages were according to the original CVD (cyclophosphamide, vincristine, dexamethasone) scheme. [26] Vincristine was omitted because of its neurotoxic side effect.
Improvement criteria.
Improvement of the clinical course was defined as an increase in the score of the tapping test of more than 20% of the score before treatment or improvement of 1 or more points on the Rankin scale. [1,28] Stabilization was defined as an increase or decrease in the score of the tapping test of no more than 20% of the score before treatment or no change on the Rankin scale. Progression of the disease was defined as a decrease in the score of the tapping test of more than 20% of the score before treatment or deterioration of 1 or more points on the Rankin scale.
Side effects of treatment were assessed using WHO criteria. [41]
Statistical analysis.
For the statistical analysis the results of sum scores and tapping tests can be considered as an ordinal scale, and presented in percentiles. The Wilcoxon matched pairs tests were used to compare values before and after treatment. Values of p < 0.05 were considered significant.
Results.
The mean age at onset of symptoms of the neuropathy was 57 years (SD 10 years). The mean age at the beginning of the treatment was 60 years (SD 11 years). The mean follow-up period was 3 years (range, 2-4 years). On clinical examination, all patients had a symmetric sensorimotor polyneuropathy. The sensory symptoms and signs were more pronounced than the motor abnormalities, and the legs were more involved than the arms.
Of the 16 patients who entered the trial, one patient with IgG-MGUS developed severe leukopenia (WHO grade 4) after one cycle with cyclophosphamide and prednisone, necessitating withdrawal of treatment (patient 16, Table 2). After cessation of the cyclophosphamide, the leukopenia recovered within 2 weeks. All 16 patients were included in the analysis.
Table 2. Clinical, hematologic and electrophysiologic features of patients with polyneuropathy associated with MGUS before and 1 year after treatment
Neurologic evaluation 1 year after treatment.
After treatment, eight patients improved and six patients stabilized according to our criteria. Of these 14 patients 10 had IgM-MGUS and four IgG-MGUS. Two patients (numbers 10 and 16) had deterioration.
The motor and sensory sum scores improved significantly after treatment (median motor sumscore before treatment was 110 and after 112, Wilcoxon matched pairs test p = 0.04; median sensory sumscore before treatment was 37 and after 40; Wilcoxon matched pairs test p = 0.03). The scores of the tapping tests of the arm and leg after treatment were significantly better than before treatment (median tapping test score arm before treatment was 30 and after 32, Wilcoxon matched pairs test p = 0.03; median tapping test score leg before treatment was 37 and after 41 Wilcoxon matched pairs test p = 0.008). The vibration perception threshold also showed improvement, but this value did not reach the level of significance (Wilcoxon matched pairs test, p = 0.08). The Rankin score showed significant improvement (Wilcoxon matched pairs test, p = 0.04).
The need of devices was greater before treatment than after treatment. Before treatment three patients had no devices, three needed only a walking cane, eight needed a walking cane and ankle braces, and two patients were wheelchair-bound. After treatment, five patients had no devices, three needed a walking cane only, three needed a walking cane and ankle braces, four needed only ankle braces, and one patient was wheelchair-bound. Two patients (numbers 7 and 12) were severely disabled by sensory ataxia before treatment, but could walk normally afterward. The major improvement started after 3 months of therapy in almost all patients who had a positive response to treatment.
The patient (number 10) in whom the polyneuropathy deteriorated despite therapy complained of weight loss, light-headedness, and syncope. On clinical examination, we saw a sick, frail man with severe orthostatic hypotension. He showed unsteadiness of gait and loss of equilibrium, leading to wheelchair dependency. A rectum biopsy was negative but a repeated bone marrow biopsy showed amyloid deposition. The clinical course of this patient further deteriorated and he died 10 months later.
Neurologic evaluation 3 years after treatment.
During the complete follow-up period improvement or stabilization was retained in all patients in whom the polyneuropathy improved or stabilized.
Type of gammopathy.
No differences in responding to treatment were found on quantitative neurologic examination between patients with IgM and IgG-MGUS. In the group of patients with IgM gammopathy there were no significant differences between five MAG-positive and six MAG-negative patients regarding quantitative neurologic examination.
Electrophysiological studies.
Based on motor conduction studies, 13 of the 16 patients had predominantly demyelinating or only demyelinating features (see Table 2). Eight of these 13 patients had spontaneous muscle fiber activity, indicating coexistent axonal degeneration. The other three patients had predominantly axonal or mixed axonal-demyelinating features on motor nerve conduction study; two of these three patients had spontaneous muscle fiber activity. The mean TLI was 0.49 in nerves without electrophysiologic signs of demyelination (DML, CV), and 0.47 in nerves with signs of demyelination. If values of DML fulfilled the criteria for demyelination and values of CV did not, mean TLI was 0.28. On the other hand, if values of CV fulfilled the criteria for demyelination and values of DML did not, mean TLI was 0.57. In four patients (numbers 1, 2, 3, and 5) low TLI values (<or=to0.25) were present; all four patients had anti-MAG antibodies.
After treatment ulnar nerve conduction variables (DML, motor CV, CMAP amplitude, percentage of CMAP amplitude reduction, sensory CV, SNAP amplitude, Sigma CMAP, and Sigma SNAP after treatment) were significantly better than before treatment Table 3. In one patient the sural SNAP, which was not obtainable before treatment, could be obtained after treatment. In the patient with conduction block before treatment there was a decrease of the percentage of amplitude reduction from 86% to 60% and a decrease of the area reduction from 85% to 50% after treatment.
Table 3. Median values (50th percentile) of electrophysiologic studies before and 1 year after treatment
Laboratory analysis.
In none of the patients did the M-protein disappear after treatment. In 10 patients the M-protein was present but could not be measured quantitatively. In the patients with a measurable M-protein four patients showed a decrease of M-protein after therapy. Only in one patient (patient 9) was the level of M-protein not related to the response to treatment.
In nine patients bone marrow biopsy and aspirate was evaluated before and after treatment. After treatment, in eight patients the monoclonal lymphoid IgM or plasmacel IgG infiltration decreased as measured in the bone marrow biopsy specimens and by immune phenotyping. In four of the nine patients the monoclonality could not be detected. (Absence of monoclonality was defined as a kappa:lambda ratio <or=to 70-30 within the malignant heavy chain.) In one patient amyloidosis was diagnosed. (see Table 2.)
Side effects of treatment.
Three patients (19%) had severe loss of hair (WHO grade 3) during the treatment cycles; 2 months after the last treatment hair growth was restored. Three days after the beginning of treatment all patients had nausea (WHO grade 1) for 2 days.
Discussion.
This prospective study shows that treatment with intermittent cyclophosphamide and prednisone during 6 months can prevent worsening of neuropathy associated with MGUS (38%) or lead to amelioration (50%), as measured by quantitative neurologic examination and electrophysiologic studies. In addition, plasma cell infiltration of bone marrow decreased. The improvement of the neuropathy was evident not only with the quantified neurologic examination, but also on the Rankin disability scale. Two patients who were wheelchair-bound before treatment were able to walk after treatment. During the follow-up period of 3 years, the improvement persisted and the polyneuropathy stabilized. Although this study is an open trial with regard to the evaluation of the clinical data, the improvement in the electrophysiologic and hematologic data cannot be attributed to observer bias, because they were blinded to the results of previous testing.
All 14 patients in whom the polyneuropathy improved or stabilized had a demyelinating or a mixed polyneuropathy on the basis of nerve conduction studies. In most of these patients the neuropathy was predominantly demyelinating. We may have underestimated the number of patients with demyelination because a low CMAP amplitude on distal stimulation, which was considered to be evidence of axonal degeneration, can also be caused by distal conduction block. The clinical improvement may be attributed either to improvement of myelin function leading to a decrease of the number of fibers with conduction block or alternatively to axonal regeneration. Decrease of fibers affected by conduction block is supported by the decrease in the percentage of ulnar nerve CMAP amplitude reduction. Further evidence of improved myelin function is provided by the decrease in ulnar nerve DML and increase in ulnar nerve CV. The increase of ulnar nerve CMAP and SNAP amplitude is related either to improvement of distal conduction block or to axonal regeneration. Electrophysiologic and histologic examination of the patient who had deterioration despite treatment showed an axonal polyneuropathy. Deterioration in this patient may be attributed to the gradual deposition of amyloid, which was not found at the start of the study. A low TLI indicates that demyelination in the distal segment of the nerve is more pronounced compared with the proximal segment [33,34]; this electrophysiologic pattern is characteristic in patients with anti-MAG antibodies. Kaku et al. [34] suggested that demyelination starts distally and progresses to more proximal nerve segments in the course of the disease. [34] In our study these were low TLI values in four of the five patients with anti-MAG antibodies.
There is a debate about whether the M-proteins cause the neuropathy in all patients with MGUS or whether there is simply a chance association in some patients, since 1 to 3% of the population over 50 years have M-protein in their serum, often with no evidence of any related disease. [42,43] Most people with MGUS and no disease have IgG but not IgM M-proteins. [42] That 11 patients in our study had IgM M-proteins and only five IgG suggests that the M-protein is not likely to be simply an associated problem. In addition, since axonal neuropathies are much more frequent than acquired demyelinating neuropathies, that so many of the neuropathies in our study were associated with demyelination also suggests that the M-protein and neuropathy are not likely to be simply chance associations. [1,3-5,7,25,31,33]
Dyck et al. [21] (see Table 1) showed that polyneuropathy associated with IgM gammopathy responds less well to plasmapheresis treatment than polyneuropathy associated with IgG and IgA-MGUS. We found no difference in response in patients with IgM or IgG, but the number of patients may be too small for definite conclusions. Blume et al. [23] described four patients with polyneuropathy associated with high-titer serum IgM autoantibodies against MAG who had improvement after treatment with a combination of plasmapheresis and cyclophosphamide. About half of our patients with IgM M-protein had anti-MAG antibodies, in agreement with earlier observations. [44] In our study, we found no difference in improvement between patients with and without anti-MAG antibody activity.
In none of our patients did the M-protein disappear or diminish significantly after therapy, but this might be explained by more than half of our patients having the M-protein before treatment but too low to be quantified. Bone marrow evaluation, however, showed a significant reduction in monoclonal B and plasma cells. The reduction of monoclonal B and plasma cells is probably responsible for the neurologic improvement.
Although the side effects of plasmapheresis and IVIg are less severe than the side effects of continuous treatment with chemotherapy, the improvement is limited in duration. [25] In our study, the patients received intermittent, instead of continuous, chemotherapy to lessen the side effects. [41] Intermittent doses of cyclophosphamide and prednisone are generally well tolerated. [41] However, chemotherapeutics have potentially serious side effects. Therefore, careful consideration is warranted before treating patients with MGUS who show significant disability and progression. In our study there were only minor side effects, such as nausea and transient hair loss in three patients. In one patient, treatment was stopped due to severe neutropenia occurring after the first cycle of treatment.
The ideal duration of chemotherapy is unknown. Kelly et al. [16] advise long-term and aggressive treatment to maintain the M-protein levels at, or lower than, 50% of baseline levels for several months. We have chosen six monthly courses of chemotherapy based on experience in non-Hodgkin lymphoma and multiple myeloma. Maximum tumor reduction in these malignancies is achieved in most patients within 6 months, whereas continuation of chemotherapy does not prolong remission duration and may be associated with increased risk for the occurrence of a myelodysplastic syndrome or acute leukemia. [45] As in the study of Blume et al., [23] 6 months of intermittent therapy appeared to be sufficient, as neurologic improvement occurred in most patients early in the course of treatment and persisted during the follow-up of 3 years.
In conclusion, treatment with intermittent cyclophosphamide and prednisone therapy, based on bone marrow biopsy results and neurologic scores, may be favorable in patients with demyelinating polyneuropathy associated with MGUS. The chemotherapy can be discontinued after 6 months without deterioration of neurologic function.
- Copyright 1996 by Advanstar Communications Inc.
REFERENCES
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Use of Whole-Genome Sequencing for Mitochondrial Disease Diagnosis
Dr. Robert Pitceathly and Dr. William Macken