вход по аккаунту


Chronic relapsing (Dysimmune) polyneuropathy Pathogenesis and treatment.

код для вставкиСкачать
Chronic Relapsing (Dysimmune)
Polyneuropathy: Pathogenesis and Treatment
Marinos C . Dalakas, MD, and W. King Engel, M D
Chronic relapsing polyneuropathy is a distinct dysschwannianldemyelinating polyneuropathy characterized by
usually slow onset, progressive or relapsing-remitting course, elevated cerebrospinal fluid (CSF) protein, marked
slowing of nerve conduction velocity, segmental demyelination demonstrable in sural nerve biopsies, and absence
of systemic illness or abnormal serum immunoglobulins. The cause of the disorder and the mechanisms underlying
its chronicity and relapsing-remitting course are not clear. Immunoglobulin deposition observed in sural nerve
biopsies and abnormal immunoglobulin patterns in the CSF in some cases suggest a dysimmune pathogenesis; thus
the term chronic relapsing (dysimrnune) polyneuropathy (CRDP) is preferred.
The disease is a treatable form of idiopathic polyneuropathy. In our series of 25 patients with CRDP, treatment
with high-single-dose daily prednisone, slowly tapered to an alternate-day program, has been very successful in the
majority. A low (10 to 20 mg) alternate-day single-dose program, maintained indefinitely, seems to be required to
prevent future recurrences. Evidence is provided that other immunosuppressants (azathioprine, cyclophosphamide,
poly-ICLC) and possibly plasmapheresis, alone or in conjunction with corticosteroids, may have a beneficial role in
controlling difficult cases of chronic relapsing polyneuropathy.
Dalakas MC, Engel WK: C h r o n i c relapsing (dysimmune) polyneuropathy: pathogenesis and treatment.
A n n N e u r o l 9(suppl):134-145, 1981
Acute dysimmune polyneuropathy (ADP)-the
Guillain-Bar& syndrome-is
an acute dysschwanniaddemyelinating polyneuropathy that typically progresses to maximum weakness within 3 to 4 weeks
of onset, remits spontaneously, and does not recur.
The small number of patients who d o not recover
completely continue to have a stable, unchanged
neurological deficit. Although the underlying pathogenesis remains uncertain, it is considered dysimmune. The peripheral Schwann cells (with their
myelin sheaths) are attacked, probably by immunologically activated cells or their products [ 4 , 121. The
disease often occurs within 2 to 4 weeks of an infection, especially viral [4].
There is another type of idiopathic dysschwanniaddemyelinating polyneuropathy with similar
clinical, electrophysiological, and pathological features but with slow onset and a chronic or relapsing-remitting course. The terms that have been used
for this disorder include: chronic relapsing polyneuritis [39], recurrent polyneuropathies [7], recurrent and chronic relapsing Guillain-Bark polyrelapsing corticosteroid-dependent
neuritis [@I,
polyneuritis [28], steroid-responsive
polyneuropathy [ 151, relapsing hypertrophic neuritis [ 161, chronic inflammatory polyradiculoneurop-
athy [ 171, subacute demyelinating polyneuropathy
responding to corticosteroid treatment [ 321, and
Guillain-Bar& syndrome with slow, progressive
onset and persistent elevation of cerebrospinal fluid
(CSF) protein [26]. We prefer the term chronic relapsing (dyssimmune) poEyneuropathy (CRDP).
Although it has previously been reported that patients with CRDP may respond t o corticosteroid
medication, no agreement exists as to how best to
administer it. In this review we describe the clinical
and laboratory findings in 25 patients with CRDP and
their response to treatment. O u r results suggest that
a carefully planned corticosteroid regimen may be
critical in the long-term management of such cases.
We also discuss advances in understanding the
pathogenesis of the disorder and its relationship to
ADP as well as to experimental allergic neuritis
From the Neurornuscular Diseases Section, Medical Neurology
Branch, National Institute of Neurological and Communicative
Disorders and Stroke. National Institutes of Health, Bethesda,
M D 20205.
Presented at the Conference o n Guillain-Barre Syndrome sponsored by the Kroc Foundation, Santa Ynez Valley, CA, Sept
22-26. 1980.
Criteria and Classification
For years, our diagnostic criteria [13, 14, 22, 441similar to those used by Prineas and McLeod [391have been:
1. Presence of a sensorimotor polyneuropathy of
slow onset (usually over several months) with
Address reprint requests to Dr Dalakas, Neurology Service (127),
VA Medical Center, East Orange, NJ 07019.
depth of disability usually reached after a 6- to
12-month period
A course that is steadily progressive and/or
Marked slowing of nerve conduction velocity
Elevated CSF protein
Absence of evidence of systemic disease, including diabetes, or of a history of exposure to toxins
and heavy metals; in addition, serum immunoglobulin electrophoresis must be normal, and
there must be no evidence of amyloidosis as determined by crystal violet or congo red stain in
muscle and nerve biopsies
Segmental demyelination in surd nerve biopsies,
as has been noted by others [39], although teased
nerve fiber preparation was not routinely performed in our cases
We further recognized that with regard to relapses,
CRDP cases can be subdivided into autorelapsing,
pharmacorelapsing, or both. In autorelapsing CRDP
we included patients who had experienced at least
two episodes of spontaneous remissions and exacerbations before being treated with corticosteroids or
other immunosuppressants. (These cases may become less common because early commencement of
therapy, as practiced by many neurologists, may
forestall later recurrences and alter the natural tempo
of the disease.) In pharmacorelapsing CRDP we included patients whose relapses appeared to be directly related to pharmacotherapy, either from tooearly withdrawal (“withdrawal polyneuropathy” [7])
or from suboptimal pharmacological dosage. Patients
without spontaneous relapses but with a stepwise
progressive course predominantly developed pharmacorelapses after attaining a pharmacologically in-
duced remission followed by reduction of the drug.
Some relapses, however, occurred while the patients
were being maintained on an unchanged, previously
effective dosage, and thus might have been autorelapses due to a subacute exacerbation of the underlying illness.
Clinical Picture and Laboratory Studies
The most pertinent clinical and laboratory features in
9 representative patients among the 2 5 with CRDP
are summarized in Table 1. In the whole series, persons of all ages-including children-were affected,
and both sexes appeared to be equally susceptible.
Typically the peak of disability was reached slowly,
over 6 to 12 months (Figs 1-3), with a plateau of
disability that lasted from several weeks to several
months in untreated patients. Subsequent remissions
were usually incomplete, especially if they had occurred before commencement of pharmacotherapy
(Figs 1, 2). Spontaneous remissions lasted from
months (frequently) to years (rarely) (Fig 3). Second
and later relapses were usually of subacute onset and
slow progression (Figs 2, 3) and tended to be more
severe than the initial episode (Fig 1).
The most common initial symptoms were weakness and paresthesias distally in the limbs. Cranial
nerve involvement was not as common as in ADP,
being present in only 4 of our 2 5 patients (2 with
facial weakness and another 2 with oculomotor disturbances). One patient had pseudopapilledema.
Tendon reflexes were invariably absent and remained
absent or markedly depressed even during the recovery phases. Two patients (see Table 1 and Fig 2)
required assisted ventilation during relapses. Sensory
loss was present in all cases and appeared to affect all
modalities. No patients showed autonomic or central
Table I. Data from 9 Patients with Chronic Relapsing Polyneuropathy
Age (yr)
and Sex
32, M
64, M
44, F
40, M
20, F
32, F
42, F
18, F
53, F
No. of
aNumbers in parentheses represent the relapse, after the first, at which tremor appeared.
nerve conduction velocity;
+ = mild; ++ moderate; ++ + = severe; ++++ = total quadriplegia.
Dalakas and Engel: Relapsing Polyneuropathy
nervous system dysfunction, although both have
been noted previously in CRDP [17].
An interesting phenomenon, present in 6 patients,
was an action and postural tremor prominent in the
distal muscles of the upper extremities. In 5 of the 6
patients it appeared during the second and third relapse and in 1 at the onset of the disease. This tremor,
unlike rhythmical essential tremor, was irregular,
with variable amplitude between oscillations, and was
unlike the pseudoathetotic movements associated
with severe proprioceptive loss. It was also unrelated
to the severity or distribution of muscle weakness
(strength had returned to normal in 2 of our patients
with tremor), to proprioception, which was usually
normal or only slightly impaired, or to corticosteroid
therapy, which in high doses is known to cause a mild
but rhythmical action tremor. The tremor lessened
during treatment with high doses of corticosteroids
and disappeared as the neuropathy improved. It
reappeared, however, after too-early withdrawal or
rapid tapering of corticocosteroids. Propranolol, used
in 1 patient, was ineffective. The mechanism of the
tremor, which most likely reflects disease activity, is
unknown. It has been suggested [3j that such a
tremor may be related to selective disturbance of
large sensory fibers responsible for proprioceptive
input from the muscles. Because the tremor appears
after the second or third relapse, the extent and degree of nerve involvement or the type and patterns of
the preceding nerve remyelinations and muscle reinnervations may determine its manifestation.
In none of our 25 patients was there evidence that
the initial or subsequent attacks of neuropathy were
precipitated by nonspecific illnesses. This is in
agreement with several previous reports (7, 32, 38,
491 but contrasts with others [37, 39, 461. Onset
during pregnancy has been reported by some investigators, and this was noted in 1 patient of our series,
the neuropathy appearing during the late months of
pregnancy, peaking a few months after delivery, and
subsequently pursuing a relapsing and remitting
The precise relationship of CRDP to ADP is uncertain. In none of our 25 CRDP patients was the
initial attack typical of ADP with regard to its temporal course. This is similar to the experience of
others [17, 32, 38, 39, 491.
The hemograms, sedimentation rate, serum immunoglobulins, and serum protein electrophoresis
and immunoelectrophoresis were normal in all cases.
Antinuclear antibody titers and lupus erythematosus
preparations were negative. The CSF total protein
was invariably elevated, and levels above 100 mg/dl
were frequently seen (Table 1). The highest levels
were measured during exacerbations, whereas during
remissions the total protein was decreased and occa-
sionally was close to normal. O n agarose gel electrophoresis a single band was frequently seen in the
IgG region regardless of disease activity or the
amount of total protein [ 131. Another marker of activity was CSF IgM, which was disproportionately
elevated duringrelapses in 3 patients with CRDP who
subsequently responded to corticosteroids [ 8 ] . Cells
were not increased.
One of the hallmarks of the disease was the
marked slowing, usually in the vicinity of 10 to 28
m/sec, of motor nerve conduction velocity (NCV).
Sensory potentials were absent (see Table 1). The
latency of the elicited muscle action potential was
prolonged. Although the motor NCV may improve
during clinical remission, its degree is disproportionately much less than the clinical improvement, and in
none of our 25 patients did it reach normal levels.
Nerve biopsies showed no mononuclear cells
infiltrating the nerves nor evidence of vasculitis on
light microscopy, in contrast to reports by others [ 141
describing lymphoid inflammatory infiltrates. O n
fresh-frozen sections stained with modified trichrome and sudan black [151 there was a relative
paucity of large myelin sheaths, but we did not d o
teased nerve fiber preparations to determine whether
this was caused by segmental demyelination, as described by others [17, 391, or was associated with loss
of axons. Endoneurial connective tissue was increased. Our immunocytochemical studies are discussed later.
Since 1958, when Austin [7j reported a remarkable case of recurrent polyneuropathy responding
dramatically to corticosteroids, several reports have
appeared attesting to the beneficial effects of corticosteroids in CRDP [ 7 , 15, 17, 18,22, 32, 391. There is
little agreement, however, concerning the amount
and type of steroids that should be used, the anticipation of a response, or the duration of therapy.
In recent years we have treated 25 patients with
corticosteroids alone or in conjunction with azathioprine, and the majority have improved. Although a placebo-controlled study was not performed, the patients themselves repeatedly served as
their own controls. Our conclusions concerning the
most promising therapeutic regimens follow.
Corticosteroids are the drugs of choice. We start
treatment with high-single-dose (HSD) prednisone,
100 mg daily, or 1.5 mg per kilogram of body weight
in children. After an initial 3 to 4 weeks of therapy,
this dose is tapered over 10 weeks to a 100 mg
single-dose, alternate-day (SDAD) schedule by
gradually reducing an alternate “off-day’’ dose by 10
136 Annals of Neurology Supplement to Volume 9, 1981
Table 2. Conservative Method of Reducing High-Single-Dose
Daily Prednisone” t o Low-Dose, Alternate-Day (Maintenance)
Levels in the Treatment of CRDP
Prednisone (me)
High Day
Taper Method
By 10 mg every week
By 5 mg every week
3-4 wk
2 wk
1-3 mo
12-16 wk
36-48 wk
12 mo
20- 10
12-16 m o
By 5 mg every 3-4
By 2.5 mg every 3-4
By 2.5 mg every
By 2.5 mg every 3-4
mo (or by 1 mg
every 1-2 mo)
“Reduction can be slower or faster if dictated by disease status or
drug side effects.
bIf no benefit by this time, patient is probably not responsive to
prednisone; tapertng may then be accelerated to 0.
mg per week (or faster if necessitated by side effects,
though this carries a greater risk of breakthrough of
disease). The tapering program is shown in detail in
Table 2. Prednisone in the high SDAD program
is very slowly reduced, by 5 mg decrements every
3 to 4 weeks, assuming that the drug has started to
have a beneficial effect with only minimal side effects. The SDAD program minimizes side effects
(cushingoid appearance, diabetes, obesity, high blood
pressure, osteoporosis, avascular necrosis) while
controlling the underlying disease. Below a 50 mg
SDAD dosage, the prednisone is reduced by 2.5 mg
every month (Table 2 ) . Patients are maintained
indefinitely on a 10 to 20 mg SDAD dose to avoid
future relapses.
Obligatory with every prednisone treatment is a
collateral program designed to prevent or diminish
side effects. This consists of an antacid (Maalox, Amphojel, Riopan) between meals and at bedtime; a 2
gm sodium, low-carbohydrate, high-protein diet; and
an 80 to 120 mEq potassium supplement, such as
potassium chloride, in liquid form daily in divided
doses (this dosage is decreased as the prednisone is
When, in retrospect, the dose of SDAD prednisone is decreased below a critical level, a relapse may
appear in the next 3 to 12 weeks. It is difficult to predict such relapses, but our experience has been that
the slower the rate of dosage reduction and the
smaller the decrements, the less the chances for relapse. Should such a relapse occur, the disease cannot
be controlled by return to the previous dosage level
but requires an increase to a higher dosage, often to
100 mg SDAD or, in severe cases, to a 100 mg daily
dose. As the disease is brought under control, one
must very gradually decrease the dose again, but this
time more slowly as the “optimal” prednisone level is
approached. (Optimal prednisope level for a given
patient at a given time is considered to be the
minimum dose that will prevent relapse or, empirically, the dose slightly above that at which the previous pharmacorelapse occurred.) A steroid-induced
improvement is recognized by: (1) reversal of the relapsing phase, ( 2 ) alteration in the previous tempo of
worsening of the illness, ( 3 ) arrest of progression,
and (4)prevention of future relapses.
The following general comments relate to our corticosteroid therapy program for CRDP.
1. The most sensitive indicator of improvement
is the clinical picture (strength and sensory
findings). Tendon reflexes usually do not reappear, and NCV may improve only slightly. Total
CSF protein is usually reduced toward or to normal levels, but the abnormal CSF band remains
r 141.
2. There may be a lag period, usually 1 to 4 weeks
(but occasionally up to 3 to 5 months), from the
commencement of therapy until the appearance
of the first signs of improvement. Occasional patients show rapid improvement, within days.
3. A low (10 to 20 mg)’ SDAD maintenance dose is
continued indefintitely, even if the neuropathy is
in complete remission.
4. In patients with longlasting relapse or long, progressive illness, it may be more fruitful to begin
therapy with a combination of HS D daily prednisone and another immunosuppressant, preferably
This program has been successful in controlling the
disease in most of our patients with CRDP. Prednisone has been rather well tolerated. In our various
patients with neuromuscular disease who have been
treated with long-term high SDAD prednisone, osteoporosis (especially in females) and cataracts have
been the most persistent of the well-known side effects. In occasional patients, avascular necrosis of the
femoral or humoral head can develop. Hyperglycemia and hypertension have occurred, apparently in
predisposed individuals. The corticosteroid side effects are fewer on an alternate-day program and as
the dose is decreased. However, because of them, if a
patient is taking both prednisone and azathioprine
(or another cytotoxic agent), we prefer to remove the
prednisone gradually before the azathioprine.
Dalakas and Engel: Relapsing Polyneuropathy
Other Immunosuppressants
For patients who continue to worsen or d o n o t respond to corticosteroids, short- or long-term trials
with other immunosuppressants should be considered. In 1966 Palmer [33] described a case of
CRDP treated with azathioprine after 10 weeks of
progression. Improvement occurred within a week,
but attempted reduction to zero dosage caused a relapse. The patient improved again with reinstitution
of azathioprine. Since then, several authors have successfully used immunosuppressive agents in the
treatment of CRDP: azathioprine in 7 patients [9,
511, cyclophosphamide in 7 [39, 411, and nitrogen
mustard in 1 [39].Other authors, however, in limited
studies, have reported no benefit in 3 patients given
azathioprine and 6-mercaptopurine [ 10, 24, 54).
From among the group of immunosuppressants
whose common denominator is their antilymphocytic
effect, our drug of choice is azathioprine at a 3 mg/kg
single daily dose. It usually takes 10 to 30 days for
the laboratory effect (leukopenia) to become manifest and 1 to 12 weeks for the clinical effect to occur.
One sometimes sees clinical benefit without reduction of white blood cell count, but in the absence of
clinical response one can, in carefully monitored and
reliable patients, gradually increase the dosage to titrate reduction of the polymorphonuclear cell count
(making sure it does not drop below 1,500 to
2,000/mm3 and that platelets and hemoglobulin d o
not fall too low). It is essential to see these patients at
least once weekly at the beginning-and, when their
condition has stabilized, every 3 to 4 weeks-to assess the benefits and side effects of pharmacotherapy,
to reassure those who have become discouraged from
the long-term illness, and to check their blood for
possible bone marrow suppression o r liver damage.
We have tried azathioprine in 4 patients with CRDP
who became or were unresponsive to prednisone.
The drug was beneficial in 3 and had no effect in 1.
The 3 patients have improved to almost 90 to 95% of
normal and have not experienced any relapse in the
last 2 years while azathioprine has been maintained
and prednisone tapered to low doses. In addition to
inducing clinical remissions (see, for example, the
patients shown in Figs 2 and 3), it appears that azathioprine also stabilizes an unstable corticosteroidresponsive and -dependent patient, allowing prednisone to be safely tapered to lower levels at a faster
Cyclophosphamide, an alkylating agent, can be
used at doses of 2 mglkg daily. We have tried it in
only 1 patient, in whom it appeared to be beneficial.
We have also found cyclophosphamide beneficial in
other dysimmune diseases when azathioprine was an
apparent failure, or perhaps was not pursued to effective dose. Cyclophosphamide and azathioprine
138 Annals of Neurology
have somewhat different short- and long-term
toxicities, of which the physician must be aware.
Further studies are needed to determine which drug
is better for long-term use in CRDP. At present, our
drug of choice either after or in combination with
prednisone is azathioprine for maintenance therapy
in CRDP patients. We have used it without major
complications in the treatment of polymyositis/
dermatomyositis and myasthenia gravis lasting several years.
Another promising new antidysimmune-antilymphocytic drug is poly-ICLC, a synthetic polynucleotide, which has been successful in 2 of 4 treated
CRDP patients, 1 of whom showed remarkable improvement starting with the first dose [21]. This
patient has had sustained improvement for 295
years but still requires an infusion of poly-ICLC
every 3 to 6 weeks. The drug produces a flu-like syndrome in the first 24 hours and causes a dramatic
lymphopenic effect (80 to 95% reduction in total
lymphocytes compared to the baseline) lasting about
2 days, with recovery by about 4 to 5 days. Polymorphonuclear cells increase two- to threefold acutely
and transiently. (The serum interferon rises only
modestly, and at 24 hours there is no demonstrable
A recent therapeutic approach in CRDP is plasmapheresis, which has received extensive publicity.
There is no doubt that this procedure has been
beneficial in reversing the neurologic deficit in several, but certainly not all patients with CRDP [ 11, 23,
27, 451; improvement, which coincided with the
onset of plasmapheresis, was rapid, usually noticeable
after the first exchange, and rose in increments after
each session. It is not certain if the procedure has a
longlasting effect and prevents future recurrences;
based on the experience with its use in myasthenia
gravis 1311, a prolonged residual benefit is doubtful.
In fact, 1 patient with CRDP relapsed 2 months after
the last treatment [ 111, but she responded again and
maintained her improvement when plasmapheresis
was done at frequent intervals. Any longlasting superiority to treatment with prednisone and azathioprine has to be determined in future controlled
studies; such a long-term benefit of plasmapheresis is
not clearly evident in myasthenia gravis [ 3 11.
We have treated 3 CRDP patients with this technique, removing 4 liters of blood twice a week for 4
to 5 weeks. Two patients did not improve; the third
(see Fig 3) had minimal benefit that started after the
second exchange and continued for 2 to 3 weeks after
the last treatment. This improvement was transient,
and 4 to 6 weeks after the last exchange she returned
to her pretreatment functional status. If the beneficial
Supplement to Volume 9, 1981
Fig I . (Patient 1 ) Effect of prednisone on percent of estimated
neuromuscularfunction in a patient with CRDP. After the
second relapse a slow spontaneous improvement began; this was
markedly accelerated when prednisone therapy was started
(arrow) and continued t o normal as pvednisone was slowly tapered. In the following years the patient has remained normal,
without recurrence, on low SDAD prednisone therapy.
effect of plasmapheresis in CRDP is fully established
in controlled studies, it may prove to be of value
when: (1) all the described drug combinations, in
adequate doses and duration, have failed; (2) side effects have developed or there are contraindications
to phatmacotherapy; and (3) fast improvement is
necessary in a bedfast, respirator-bound patient until
pharmacotherapy takes effect.
Examples of the Characteristic Course
and Treatment in 3 CRDP Patients
Patient 1
A 32-year-old woman developed slowly progressive weakness and paresthesias, which continued from December,
1973, when they first appeared, to about March, 1975,
when she reached her maximal disability (33% of normal
function); at that point she began to improve spontaneously, reaching a maximal improvement to 5 5 % of normal
in 6 months. The improvement was short-lived and was
followed by another spontaneous relapse, which in 3
months decreased her neuromuscular function from 55%
to 25%. Subsequently, a slow spontaneous improvement
began, elevating her neuromuscular function from 25 % to
about 38% in 3 months. At that point prednisone was
started (arrow, Fig 1) and resulted in dramatic acceleration
of improvement; in 10 days her neuromuscular function
increased from 38% to 75% (sharp rise of curve).
That rapid improvement, which was temporally related
to the commencement of prednisone therapy, contrasted in
tempo and degree to the improvement she experienced
during her first spontaneous remission, which had slowly
reversed her neuromuscular capacity from 33% to 55% in
6 months. Furthermore, the prednisone-induced acceleration and enhancement of improvement continued, and in
the following 12 months, as the drug was tapered to alternate days, the patient became completely normal for the
first time since her disease had begun 3 years earlier. In the
subsequept 3 years up to this writing she has continued to
function normally on low-dose, alternate-day prednisone
(now taking 20 mg every other day) without recurrences.
This clinical and therapeutic pattern is the most common: patients respond to prednisone alone and function
with almost full strength on low maintenance doses.
Patient 2
A 60-year-old man developed a slowly progressive
polyneuropathy that caused him complete quadriplegia and
respiratory failure 10 months after onset. (A 40 mg prednisone trial for 10 days during the worsening period-first
arrow, Figure 2-was considered ineffective and was discontinued.) After 5 months on the respirator, without
therapy, he started to improve and in 8 months reached a
maximum level of 70% of normal. AftFr 4 to 5 months a
second exacerbation began, which again progressed in 7 to
8 months to complete quadriplegia and respiratory failure.
Without drug therapy, he remained totally disabled and on
the respirator for 3 months, when he again began to improve spontaneously, reaching 45% of normal in 3%
At that time (second arrow, Fig 2) commencement of a
40 mg prednisone dose accelerated the pace of improvement and elevated his functional status from 45% to 75%
of normal in just 1 month. A month later, when his already
low prednisone dose was cut in half, he worsened at a
rather rapid rate, and in 2 months his functional status had
declined from 75% to 20% of normal. That continually
worsening phase was arrested as the prednisone was increased to 8 0 mg (third arrow, Fig 2), probably aborting a
third totally disabling relapse. Subsequently a rapid recovery began, reversing his function from 20% to 75% of
normal in 2 months. The rapidity of this recovery was tem-
Dalakas and Engel: Relapsing Polyneuropathy
€ 80-
o 40
2 0 - 1
10 -
porally related to the prednisone and was in contrast to the
previous tempo of his disease and the slow rate of spontaneous remissions. Improvement was maintained as the
prednisone was slowly reduced to 40 mg, but his function
rapidly declined (from 75% to 35%) 2 to 3 weeks after his
physicians completely discontinued the prednisone (fourth
arrow, Fig 2); upon restarting prednisone at 40 mg daily
(fifth arrow), he improved from 35% to 70% in 3 months.
While he was functioning at this level on a stable prednisone dose (40 mgiday), a severe sinusitis coincided with the
beginning of a slow decline of his neuromuscular function
(from 70% to 45% in 4 months), suggesting that a
nonspecific infection may provoke an exacerbation in
CRDP, as in other dysimmune neuromuscular diseases
(e.g., myasthenia gravis or dermatomyositis/polymyositis).
After his infection was treated, a combination of high
(100 mg) single-dose prednisone plus azathioprine ( 3
mglkg) daily was begun (sixth arrow, Fig 2). The ensuing
response was slower but eventually more rewarding than
previously; he improved during a 10-month period from
45% to 90% of normal, the first time in 5 years that he had
reached 90%. That degree of improvement was in contrast
to a maximum of 70% reached spontaneously and 75% attained with prednisone alone. In the 12 to 14 months prior
to this writing he has continued to function at 90% of normal capacity, and as the prednisone was gradually tapered
(now 60 mg on alternate days), he experienced no relapses.
This case demonstrates the importance of the optimal
steroid dose required for achieving and maintaining improvement as well as the effect of azathioprine in sustaining
Patient 3
At the age of 19 years this woman developed insidious
onset of slowly progressive weakness and paresthesias,
140 Annals of Neurology
F i g 2. (Patient 21 Effect of prednisone and azathioprine on
percent of estimated neuromuscularfunction i n a patient with
CRDP. The onset of a slow spontaneous improvement was
noted after the second relapse; low-dose prednisone (second
arrow) accelerated the pace of recovery, but as the dose was cut
in ha& the patient worsened; upon increasing prednisone t o
higher dosages (third arrow) the rapid worsening stopped and
a swift recovery began. When prednisone was discontinued
(fourth arrow), the patient quickly regressed but improved
again upon reinstitution of therapy (fifth arrow). Commencement of HDSD prednisone in conjunction with azathioprine
(sixth arrow) resulted in a remarkable (90% of normal) and
sustained improvement without recurrences in the following
1'12 years (see details in text).
which continued from March, 1958, when they first appeared, to about September of that year, when she reached
her maximal disability (555% of normal function). After a
3-month plateau of disability she began to improve spontaneously, reaching her normal status in 6 months.
She had remained completely normal for the following
10 or 11 years when in November, 1969, a second exacerbation began, causing in 4 months a 50% decline of her
neuromuscular function. After a 2- to 3-month plateau of
disability she started to improve spontaneously, reaching
her normal function in 4 months. Six years later (April,
1976), a third insidious exacerbation began, which in 6
months decreased her neuromuscular function to 45% of
normal. At that point (first arrow, Fig 3) commencement of
prednisone, 100 mg every other day, stopped the worsening phase and resulted in fast clinical recovery, reversing
her neuromuscular function from 45% to 80% of normal
in just 1 month (sharp rise of curve, Fig 3). T h e rapidity of
this recovery, which was temporally related to the com-
Supplement to Volume 9, 1981
Rednironr q.0.d.
0 90
a 70
F i g 3. (Patient 3 ) Effect of prednisone, plasmapheresis,
poly-ICLC, and azathioprine on percent of estimated
neuromuscularfunction in a patient with CRDP. Prednisone
at a dosage of 100 mg on alternate days (first arrow) resulted
in recovery, faster than the rate of the previous spontaneous remissions. As prednisone was decreased below 47.5 mg every
other day, the patient worsened; elevating prednisone to 100
mg on alternate days bad no effect (daily prednisone was never
tried in this patient). PlasmaphereJis (first set of small arrows) bad a minimal but transient effect, andpoly-ICLC
(second set of small arrows) was ineffective. Azathioprine
resulted in a slow but impressive ( 9 0 % of normal) and sustained improvement, without recurrences, in the following 2
years (see details in text).
mencement of prednisone therapy, was in contrast to the
slower rate of the previous spontaneous remissions. In the
following 2 months a maximal improvement to 95% of
normal was reached and was sustained as predriisone was
gradually tapered to lower dosages. When prednisone was
lowered to 47.5 mg on alternate days the patient started to
worsen, and despite prednisone being increased to 100 mg
on alternate days, her worsening phase continued.
At that time, plasmapheresis (first set of two small arrows, Fig 3) was performed twice a week for 5 weeks while
the dosage of prednisone remained unchanged. Plasmapheresis was beneficial in reversing some of the neurological deficit, elevating her neuromuscular function from
45% to 55% of normal in 2 months; improvement was
noticeable after the second plasma exchange and rose in
increments after each session. This mild beneficial effect
was transient, and a month after the last plasma exchange,
with the prednisone unchanged, she started to worsen. A
trial of poly-ICLC (second set of small arrows, Fig 3) was
ineffective. At that point azathioprine, 3 mg/kg, was begun.
This arrested the continually worsening phase and resulted in a slow but eventually rewarding response. H e r
neuromuscular function improved during a 12-month
period from 40% to 80% of normal and reached 90% of
normal in the next 6 months while the prednisone was
being tapered to lower dosages. For 1% years prior to this
writing she had continued to function at 90% of normal
while azathioprine was kept unchanged and prednisone was
tapered, now to 40 mg every other day.
This case illustrates the mild and short-lived beneficial
effect of plasmapheresis and the effect of azathioprine in
inducing and sustaining improvement.
CRDP is a distinct form of chronic dysschwannianldemyelinating polyneuropathy. Examining its
relationship to an acute form of dysschwannianl
demyelinating polyneuropathy, ADP (Guillain-Barre
syndrome), and its analogies to EAN may be useful in
understanding the possible pathogenetic mechanisms
underlying CRDP.
The morphological features of segmental myelin
destruction in nerves from patients with CRDP are
similar to those seen in nerves from patients with
ADP and animals with EAN [4, 5, 391, indicating
that ADP, EAN, and CRDP could have similar
pathogenetic mechanisms [39]. EAN, a useful model
for ADP, can also take a relapsing-remitting course
and can serve as a model for CRDP. Waksman [50]
observed in one experiment that more than half of
the animals with EAN inoculated with heterologous
sciatic nerve and adjuvant showed spontaneously recurring episodes of disease activity; in the study of
Wisniewski et a1 [53], 2 of 5 monkeys with EAN had a
biphasic illness; and of the 52 animals inoculated with
heterologous nerves by Pollard et a1 [36], 3 developed a relapsing-remitting course. A more reproducible model of chronic or relapsing-remitting EAN
can be produced by rechallenging the animals with
the initial peripheral nerve antigen [36, 471, suggesting that a mechanism of chronicity may be
Dalakas and Engel: Relapsing Polyneuropathy
through reexposure of the previously sensitized host
to the same agent that initially triggered the disease.
Analogous to the induction of chronic or relapsingremitting EAN may be the relapses that occurred
in some patients with CRDP upon exposure to a
known exogenous agent. In 1 patient [37], three relapses of polyneuropathy occurred, each after an injection of tetanus toxoid; in 2 other patients [461, two
relapses occurred, each after an injection of influenza
vaccine, suggesting that repeated exposure or reactivation of the specific agent that originally triggered
the neuropathy may also be responsible for the subseqyent relapses or the chronicity of CRDP, at least
in some cases. If this is a mechanism of relapse, one
would expect at least some patients with ADP to relapse each time they are reexposed to the same (or an
antigenically similar) agent that preceded the acute
illness in 70% of them. However, only 2% of ADP
patients relapse, and in none of our patients with
CRDP did the illness start as acute ADP with a
known precipitating agent or event. It may be that
the same immunoregulatory mechanisms that protect
ADP patients from having their illness reactivated
upon reexposure are the ones deficient or inadequate
in the patients who develop the relapsing-remitting
course of CRDP. Such a defect in immune
regulation-e.g., impaired function of suppressor T
lymphocytes or formation of antibodies against
them-has not yet been described in patients with
CRDP, nor has excessive activity of helper T lymphocytes. The fine interplay of a person’s activating
and deactivating immune regulation is probably determined to a large degree by his genetic status; different genetic factors certainly play a role in various
forms of immunoregulation in animals and humans
and may be factors in determining chronicity of
CRDP or the monophasic limitation of ADP. Of relevance are recent reports by two groups of inves-
tigators [2, 481 who suggested that HLA-linked genetic factors may influence susceptibility to CRDP
and that in CRDP (but not in ADP) there is a
disease-susceptibility gene associated with HLAAW30, AW31, DRW3, and DW3 haplotypes. Regardless of the mechanisms that determine a
monophasic or relapsing course, ADP and CRDP
have some clinical and laboratory differences, as
summarized in Table 3.
Evidence that humoral immunopathogenetic mechanisms may be operating in CRDP is suggested by
our immunocytochemical findings on sural nerve
biopsies from 9 patients with CRDP [14]. Examining
frozen sections of nerve biopsies by the immunofluorescence technique, we found linear deposition of IgM in 6 patients and IgG in 3; complement
(C3) was not found. Although the exact ultrastructural location and absolute molecular specificity
of these immunoglobulin deposits require further
study, the deposits appear to be on the exterior portion of the Schwann cell plasmalemma, which is rich
in carbohydrate groups and glycoproteins (such as
PO, which contains abundant galactose and galactosamine). Because glycoproteins, lipid groups, and
polysaccharides, when they serve as antigens, usually
provoke a thymus-independent antibody response
found exclusively in the IgM class (not converting to
IgG), the finding of predominantly IgM deposits in
CRDP nerves would be in accord with the possibility
that the major antigen in CRDP is a superficial plasmalemmal carbohydrate group eliciting a persistent
thymus-independent response of IgM, which binds
to the Schwann cell surface. Harmonious with the
immunocytochemical observations is the finding of
preferentially elevated IgM levels, as measured by
radioimmunoassay, in the CSF of patients with
CRDP during an exacerbation of the illness and its
fall during a corticosteroid-induced remission [81.
Table 3. Differences between CRDP and GBS in Patients Studied
Antecedent infection
Acute; maximum neurological deficit in
2-3 wk
Present in 70%
Sensory abnormalities
May not be prominent
Cranial nerve involvement
HLA antigens
No significant association (Lato et al,
CSF bands
Oligoclonal bands (transient) (Link et
al, 1978, Houff et al, 1980)
Subacute; maximum neurological deficit
in 6-12 mo
Remitting-relapsing or steadily progressive
Large fiber sensory loss invariably present
HLA-AW30 and AW31 (Stewart et al,
1778); HLA-A1, B8, DRW3, and
DW3 (Adams et al, 1979)
Single, “monoclonal” bands (Dalakas et
al, 1980)
Response to corticosteroids
142 Annals of Neurology
Supplement to Volume 9, 1981
Humoral factors were thought to be mediating the
nerve injury in a recently reported model of EAN
[42] (also a possible model for CRDP in a chronic
form); and passive transfer of demyelination has been
accomplished through intraneural injection of EAN
serum [431.
Another finding in our immunofluorescence study
was that the seemingly intact endoneurial blood vessels demonstrated granular or nodular patterns of
deposits containing complement plus IgM or IgG (or
sometimes both) [14]. We are not certain about the
specificity and signification of those deposits, nor if
their components arrived at the blood vessels free or
as already formed complexes, nor why they are there.
If they prove to be true immune deposits, they may
play a role in the immunopathogenetic mechanism of
the neuropathy through impairing the blood-nerve
barrier by inducing vascular permeability changes in
the endoneurial vessels (with or without vasoconstriction), which would greatly enhance access of a
detrimental circulating antibody to Schwann cells. In
ADP, demyelination in the perivascular region is
common [4]; whether it is in CRDP is not clear.
Another dysimmune indicator in CRDP may be
the presence of an abnormal CSF protein pattern on
agarose gel electrophoresis, as we observed in 14 of
1 6 patients [13]. In contrast to the occurrence of
transient oligoclonal IgG bands in patients with
ADP, CRDP patients have a single (“monoclonal”)
IgG band which: (1) remains unchanged with the
course of the disease, regardless of remissions or relapses; (2) is unaffected by corticosteroid therapy;
and ( 3 ) does not correlate with the total CSF protein
or the chronicity and severity of the disease. This
band in CRDP patients may serve as a dysimmune
indicator of chronicity and may reflect an antibody
response to stimulation by a not yet identified persistent antigen.
In one study [l] of 4 patients, it was found that
CRDP lymphocytes were sensitized to both peripheral and central myelin basic protein. A role of such
sensitization, if confirmed in other patients, remains
to be determined.
Regardless of the exact dysimmune mechanism (or
mechanisms) underlying CRDP, the disease forms an
important group of treatable polyneuropathies, and
reasonable functional recovery can be anticipated in
virtually every patient adequately treated. Although
the therapeutic program we effected in our patients
was not placebo controlled or double blind, the
efficacy of pharmacotherapy can be-and
wasclearly demonstrated by carefully observing in each
individual CRDP patient the natural history and
tempo of the disease and the rate of progression o r
degree of improvement before and during adequate
therapy and after interruption of adequate therapy.
As is evident from the 3 cases detailed here and
shown in Figures 1 through 3, one can discern by
careful study over a long period a definite temporal
relationship between commencement of therapy and
induction of remission and between dosage decrease
or discontinuation and the appearance of relapses.
It is not clear exactly how prednisone, azathioprine, plasmapheresis, or poly-ICLC improves the
functional status of patients with CRDP, although in
a general sense they represent antidysimmune therapy. Plasmapheresis must benefit the patient by
removing circulating putative pathogenic immunoglobulins, immune complexes, or both. (We recommend concurrent lymphocytopheresis; in that situation putatively detrimental lymphocytes, B and T, are
also removed.) We and others have been impressed
with the occasional rapidity of a remission induced by
antidysimmune therapy, which can be within days or
faster in occasional patients-much too fast to be
caused by complete remyelination. It appears, particularly with plasmapheresis and poly-ICLC therapy
[21], that what we are reversing is a “minute-tominute” detrimental phenomenon, because patients can
be better after 48 hours and then regress after 1 or a
few weeks, requiring another treatment for reversal.
That clinical change must reflect axonal recovery. We
suggest, and emphasize, that although the pathological changes in nerve and the very slow conduction
velocities reflect Schwann cell/myelin sheath abnormality, there is a major axonal malfunction in CRDP
because, ultimately, it is axonal malfunction that
causes muscle weakness, muscle atrophy, and sensory
loss [ZO]; furthermore, the remarkable clinical improvement that can occur in CRDP typically is accompanied by only a slight increase in nerve conduction velocity, which often remains below 30 m/sec.
The muscle abnormality in CRDP suggests functional
but not anatomical deinnervation, because when there
is much weakness clinically, there is only moderate
atrophy of muscle fibers with very little, if any, muscle fiber type grouping (Engel WK, unpublished observations), the converse of what one sees with
chronic diseases of motor neurons. Accordingly, the
presumed functional defect of axons could be entirely secondary to Schwann cell malfunction, or it
could be a concomitant malfunction of axons related
more directly to the immunopathogenetic mechanisms; correction of either dysfunction could result
in the rapid clinical improvement. Austin [7] suggested that in CRDP there may be increased permeability of the endoneurial capillaries, causing
endoneurial edema and affecting fluid-electrolyte
distribution and other metabolic functions of the
axon-Schwann cell complex and resulting in conduction block; and Prineas [39] found endoneurial
and intramyelinic edema and “watery” Schwann cells
Dalakas and Engel: Relapsing Polyneuropathy 143
in CRDP. It is possible that endoneurial interstitial
edema provoked by immune mechanisms such as
immunoglobulin deposition, as we demonstrated,
although requiring some time to appear and disappear, may need only a few hours to cause a functional
conduction block or relief of functional block by its
removal, affecting axons (and maybe also Schwann
cells). Or, hypothetically, molecules toxic to axonal,
or Schwann cell, function may also increase or decrease in a period of hours to days to account for
rapid changes. Corticosteroids, in addition to their
other antidysimmune actions [29, 30, 353, may affect
such mechanisms locally by their antiinflammatory
and antiedema actions, helping to reverse a temporary blockade more rapidly.
1. Abramsky 0, Webb C, Teitelbaum D, et al: Cell-mediated
immunity for neuronal antigens in idiopathic polyneuritis and
myeloradiculitis. Neurology (Minneap) 25: 1154-1 159, 1975
2. Adams D, Festenstein H , Gibson JB, et al: HLA antigens in
chronic relapsing idiopathic inflammatory polyneuropathy. J
Neurol Neurosurg Psychiatry 42:184-186, 1979
3. Adams RD, Shahani BT, Young R R Tremor in association
with polyneuropathy. Trans Am Neurol Assoc 97:44-48,
4. Arnason B G W Inflammatory polyradiculoneuropathies. In
Dyck PJ, Thomas PK, Lambert EH (eds): Peripheral
Neuropathy. Philadelphia, Saunders, 1975, pp 1110-1 148
5. Asbury AK, Arnason BGW, Adams RD. The inflammatory
lesion in idiopathic polyneuritis. Medicine (Baltimore) 48:
173-215, 1969
6. Ashworth B, Smyth GE: Relapsing motor polyneuropathy.
Acta Neurol Scand 45:342-350. 1969
7. Austin JH: Recurrent polyneuropathies and their corticosteroid treatment with five year observations of a placebocontrolled case treated with corticotrophin, cortisone and
prednisone. Brain 81:157-194, 1958
8. Brooks BR, Cook JD, Trotter JL, et al: Effect of prednisone
therapy on cerebrospinal fluid immunoglobulins and lymphocytes in chronic idiopathic relapsing polyneuropathy and
other neuromuscular diseases (abstract). Ann Neurol 1:5 10,
9. Cendrowski W: Treatment of polyneuropathy with azathioprine and adrenal steroids. Acta Med Pol 18(2):147-156,
10. Colin-Jones D G , Heathfield K W G 6-Mercaptopurine in
polyradiculoneuropathy. Lancet 2:739, 1965
11. Cook JD, Tindall RAS, Walker J, et al: Plasma exchange as a
treatment of acute and chronic idiopathic autoimmune
polyneuropathy: limited success (abstract). Neurology 30:
361, 1980
12. Cook SD, Dowling PC, Murray MR, et al: Circulating demyelinating factors in acute idiopathic polyneuropathy. Arch
Neurol 24:136-144, 1971
13. Dalakas M, Houff S, Engel WK, et al: CSF “monoclonal”
bands in chronic relapsing polyneuropathy. Neurology
30~864-867, 1980
14. Dalakas MC, Engel WK: Immunoglobulin and complement
deposits in peripheral nerves of patients with chronic relapsing polyneuropathy. Arch Neurol 37:637-640, 1980
15. DeVivo DC, Engel WK: Remarkable recovery of a steroidresponsive recurrent polyneuropathy. J Neurol Neurosurg
Psychiatry 33:330-337, 1970
144 Annals of Neurology
16. Dolman CL, Allan BM: Relapsing hypertrophic neuritis. Arch
Neurol 28:351-353, 1973
17. Dyck PJ, Lois AC, Ohta M, et al: Chronic inflammatory
polyradiculoneuropathy. Mayo Clin Proc 50:621-637, 1975
18. Dyck PJ, OBrien PC, Oviatt KF, et al: Suggestive preliminary
evidence from controlled three-month clinical trials that
prednisone improves chronic inflammatory polyradiculoneuropathy (abstract). Ann Neurol4:165, 1978
19. Eiben RM, Gersony WM: Recognition, prognosis and treatment of the Guillain-Bark syndrome (acute infective
polyneuritis). Med Clin North Am 47:1371-1380, 1963
20. Engel W K Introduction to disorders of the motor neuron,
nerves and related abnormalities. In Goldensohn ES, Appel
S H (eds): Scientific Approaches to Clinical Neurology.
Philadelphia, Lea & Febiger, 1977, pp 1250-1321
2 1. Engel WK, Cuneo RA, Levy HB: Polyinosinic-polycytidylic
acid treatment of neuropathy. Lancet 1:503-504, 1978
22. Engel WK, DeVivo DC, Warmolts JR, et al: High-single-dose
alternate-day prednisone (HSDAD-PRED) in neuromuscular
diseases (abstract). In Kakulas B (ed): Progress in Myology
(International Congress Series No. 2 37). Amsterdam, Excerpta Medica, 1971, p 59
23. Fowler H, Vulp M, Marks G, et al: Recovery from chronic
progressive polyneuropathy after treatment with plasma exchange. Lancet 2: 1193, 1979
24. Gamstrorp I, Aronson S, Lndquist B: Mercaptopurine in
polyradiculoneuropathy. Lancet 1:99-100, 1966
25. Hewer RL, Hilton PJ, Smith AC, et al: Acute polyneuritis
requiring artificial respiration. Q J Med 37:479-491, 1968
2 6. Hinman R, Magee KP: Guillain-Barre syndrome with slow
progressive onset and persistent elevation of spinal fluid protein. Ann Intern Med 67:1007-1012, 1967
27. Levy RL, Newkirk R, Ochoa J: Treating chronic relapsing
Guillain-Barri syndrome by plasma exchange. Lancet 2:
259-260, 1979
2 8. Matthews WB, Howell DA, Hughes RC: Relapsing
corticosteroid-dependent polyneuritis. J Neurol Neurosurg
Psychiatry 33:330-337, 1970
29. McMillan R, Longmire R, Yelenosky R: The effect of corticosteroids on human I g G synthesis. J Immunol 116: 1592-1 595,
30. Melby JC: Clinical pharmacology of systemic corticosteroids.
Annu Rev Pharmacol Toxicol 17:511-527, 1977
3 1. Newsom-Davis J, Wilson SG, Vincent A, et al: Long-term effects of repeated plasma exchange in myasthenia gravis. Lancet 1:464-468, 1979
32. Oh SJ: Subacute demyelinating polyneuropathy responding to
coricosteroid treatment. Arch Neurol 35:509-5 16, 1978
33. Palmer KNY: Polyradiculoneuropathy treated with cytotoxic
drugs. Lancet 1:265, 1966
34. Peterman AF, Daly D D , Dion FR, et al: Infectious neuronitis (Guillain-Barre syndrome) in children. Neurology
9:533-539, 1959
35. Petranyi GY, Benizuz M, Alfoldy P The effect of single large
dose hydrocortisone treatment on I g M or I g G production,
morphological distribution of antibody producing cells and
immunological memory. Immunology 21:151-158, 1971
36. Pollard JD, King RHM, Thomas PK: Recurrent experimental
allergic neuritis. J Neurol Sci 24:365-383, 1975
37. Pollard JD, Selby G : Relapsing neuropathy due to tetanus
toxoid. J Neurol Sci 37:113-125, 1978
38. Prineas JW: Polyneuropathies of undetermined cause. Acta
Neurol Scand 46suppl44, 1970
39. Prineas JW, McLeod JG: Chronic relapsing polyneuritis. J
Neurol Sci 27:427-458, 1976
40. Ravn H: The Landry-Guillain-Barre syndrome. A survey and
a clinical report of 127 cases. Acta Neurol Scand 43:suppl 30,
1-64, 1967
S u p p l e m e n t to V o l u m e 9, 1981
4 1. Rosen AD, Vastola EF: Clinical effects of cyclophosphamide
in Guillain-Barre polyneuritis. J Neurol Sci 30:179-187,
Saida T, Saida K, Dorfman SH, et al: Experimental allergic
neuritis induced by sensitization with dactocerebroside. Science 204:1103-1106, 1979
Saida T, Saida K, Silberberg DH, et al: Transfer of demyelination by intraneural injection af experimental allergic
neuritis serum. Nature 272:639-641. 1978
Schwwzman RJ, Engel WK, Rapaport A: Criteria for the
prednisone treatment of idiopathic polyneuropathy. Neurology (Minneap) 27:361, 1977
Server AC, Lefkowith J, Braine H, McKhann GM: Treatment
of chronic relapsing inflammatory polyradiculoneuropathy by
plasma exchange. Ann Neurol 6:258-261, 1979
Seyal M, Ziegler DY, Couch JR: Recurrent Guillain-Bark
syndrome following influenza vaccine. Neurology 28725726, 1978
Sherwin AC: Chronic allergic neuropathy in the rabbit. Arch
Neurol 15:289-293, 1966
48. Stewart GJ, Pollard JD, McLeod JG, Wolnizer C M HLA
antigens in the Landry-Guillain-Barre syndrome and chronic
relapsing polyneuritis. Ann Neurol4:285-289, 1978
49. Thomas PK, La Sceller RG, Hallpike JF, et al: Recurrent aqd
chronic relapsi& Guillain-Bark polyneuritis. Brain 92:589606, 1969
50. Waksman BH: Experimental immunalogical diseases of the
peripheral nervous system. In Rose AS, Pearson CM (eds):
Mechanis~sof Demyelinqtiop. Neb York, McGraw-Hill, 1963
5 1. W$ker GL ProgresSive polyradiculoneuropathy . Treatment
with arathioprine. Apst N Z J Mea 9184-187, 1979
52. Wiederholt WC, Mulder DW, Lambert EH: The LandryGuillain-Barr6-Strahl syndrome or polyradiculoneurspathy:
historical review, report on 97 patients, and present concepts.
Mayo Clin Proc 39:427-451, 1964
53. Wisniewski H, prostoff SW, Cantor p,et al: Recurrent experimental Jlergic polyganglioradiculoneuritis. Arch Neurol
30:347-359, 1974
54. Yuill G, Swinburn W, Liversedge L Treatment of polyneuropathy with azathioprine. Lancet 2:854, 1970
DaIakas and Engel: Relapsing Polynepropathy 145
Без категории
Размер файла
1 173 Кб
treatment, relapsing, dysimmune, pathogenesis, chronic, polyneuropathy
Пожаловаться на содержимое документа