Brain (2002), 125, 996±1004
Occurrence and characterization of peripheral
nerve involvement in neuro®bromatosis type 2
A. D. Sperfeld,1 C. Hein,2 J. M. SchroÈder,3 A. C. Ludolph1 and C. O. Hanemann1
1Department
of Neurology and 2Department of
Neurosurgery, University of Ulm and 3Institut fuÈr
Neuropathologie, UniversitaÈtsklinikum, RheinischWestfaÈlische Technische Hochschule Aachen, Germany
Summary
Neuro®bromatosis type 2 (NF2) is a rare autosomal
dominant disorder characterized by the occurrence of
bilateral vestibular schwannomas, various brain and
spinal tumours as well as peripheral nerve tumours,
cutaneous tumours and juvenile posterior lenticular
opacity. NF2 is caused by mutations in both alleles of a
tumour suppressor gene coding for a protein called
schwannomin or merlin. It is suggested that the
development of NF2 tumours is caused by complete
inactivation of the merlin/schwannomin gene. Interestingly, in a NF2 mouse model, peripheral nerve
pathology was more frequently described than schwannomas. However, review of the literature shows that
patients suffering from NF2 seldom have unexplained
clinical features of peripheral nerve lesion unrelated to
tumour masses. Single case reports describe sural nerve
biopsies, which histologically show onion-bulb-like
formations, seemingly originating from Schwann cells.
Correspondence to: C. O. Hanemann, Department of
Neurology, Zentrum fuÈr Klinische Forschung, University of
Ulm, Helmholtzstrasse 8/1, 89081 Ulm, Germany
E-mail: [email protected]
We have conducted a systematic investigation to determine the occurrence and aetiology of peripheral nerve
involvement in NF2 patients. We investigated 15
patients with de®nite NF2 and in 10 of these found
electrophysiological evidence of neuropathy. In this
study we present the classi®cation of neuropathy, correlation to clinical ®ndings, and histological ®ndings of
a sural nerve biopsy. We conclude that peripheral
neuropathy, mostly of axonal type, is a common clinical
®nding in NF2. We hypothesize that the aetiology of
this frequent peripheral neuropathy syndrome in NF2 is
caused by compression effects of multiple tumourlets,
originating along the length of the peripheral nerves on
adjacent nerve ®bres, by local in¯uences of the endoneurial pathological cells on adjacent nerve ®bres and/
or the inability of these cells to properly adhere to, or
ensheath, the axon.
Keywords: neuro®bromatosis type 2; peripheral neuropathy; preneoplasia
Abbreviations: CMAP = compound motor action potential; DML = distal motor latency; NCV = nerve conduction
velocity; NF2 = neuro®bromatosis type 2; SNAP = sensory nerve action potential
Introduction
Neuro®bromatosis type 2 (NF2) is a rare autosomal dominant
disorder with almost full penetrance and a high rate of
sporadic occurrence (Evans et al., 1992c; Evans et al., 2000).
Stringent diagnostic criteria differentiate NF2 from neuro®bromatosis type 1, formerly known as von Recklinghausen's
disease or peripheral neuro®bromatosis: the occurrence of
bilateral vestibular schwannomas and additional manifestations include various CNS tumours including meningiomas,
schwannomas of other cranial nerves, astrocytomas, ependymomas, as well as juvenile posterior lenticular opacity and, to
a lesser extent, peripheral nerve tumours, cutaneous tumours
and rare cafeÂ-au-lait spots (Kanter et al., 1980; Martuza et al.,
1988; Mautner et al., 1995; Gutmann et al., 1997). NF2 is
caused by mutations in a tumour suppressor gene coding for a
ã Guarantors of Brain 2002
protein referred to as schwannomin or merlin (Rouleau
et al., 1993; Trofatter et al., 1993). Analysis of NF2
mutations has revealed some general genotype/phenotype
correlations, and severe disease has been associated with
mutations that produce a premature termination of translation,
while milder disease forms have been related to missense
mutations (Evans et al., 1998; Gutmann et al., 1998; Evans
et al., 2000; Giovannini et al., 2000). Additionally, somatic
mutations in the second wild-type allele (second hit) of the
NF2 gene are found in both sporadic and familial schwannomas and meningiomas (Gutmann, 1997; MacCollin and
Gusella, 1998). It is suggested that the development of these
tumours is caused by complete inactivation of the merlin/
schwannomin gene, one mutation occuring constitutively in
Peripheral nerve in neuro®bromatosis type 2
almost all cells and the second hit in tumour cells, e.g.
Schwann cells.
Interestingly, polyneuropathy is reported to occur rather
infrequently and has been described in only a few case studies
of NF2. In a large clinical study, peripheral nerve lesions
unrelated to tumour masses were observed in 6% of patients
suffering from NF2 (Evans et al., 1992a). The few case
reports include patients suffering from de®nite NF2 and focal
amyotrophy, distal symmetric sensorimotor neuropathy, or
mononeuritis multiplex (Bosch et al., 1981; Kilpatrick et al.,
1992; Overweg-Plandsoen et al., 1996; Iwata et al., 1998).
Sural nerve biopsies in selected cases showed the histological
appearance of so-called `coreless onion bulb' formations
(Onishi et al., 1972; Thomas et al., 1990; Iwata et al., 1998).
The observation that many patients with NF2 present with
are¯exia, which cannot be completely explained by tumour
burden, suggests that subclinical neuropathy is underrecognized and that Schwann cells outside schwannomas
might be involved in NF2 (MacCollin, 1999). In this context,
the comparatively frequent occurrence of skin tumours as
opposed to the infrequent occurrence of peripheral neuropathy is surprising (Evans et al., 1992a; Parry et al., 1994).
We therefore conducted a systematic investigation to determine the prevalence of peripheral neuropathy in NF2. In this
study we investigated 15 patients suffering from de®nite NF2
and estimated the frequency and type of neuropathy in
correlation to clinical, MRI and laboratory ®ndings. In
addition, the ®ne structural pathological hallmarks of a
sural nerve biopsy are described.
Methods
Patients
Patients were continuously recruited from our neuro®bromatosis outpatient clinic. All patients ful®l the criteria for
de®nite NF2 (Gutmann et al., 1997). Patients gave informed
consent to participate in the study which was approved by the
local Ethics Committee. Clinical and detailed neurological
examination, including cutaneous inspection and strength
assessment using the Medical Research Council Scale, was
carried out. Touch sensation and vibration were tested with a
mono®lament and a tuning fork, respectively.
Descriptions are focused on medical history and current
symptoms related to NF2, with a special emphasis on skin
manifestations and symptoms related to a probable peripheral
neuropathy. In case of clinical evidence for a peripheral
neuropathy, the neuropathy was graded as: (i) mild (absent
ankle jerks, slight symmetrical hypaesthesia, impaired
pallaesthesia); (ii) moderate (signs as in mild neuropathy
plus atrophy of distal muscles or deformities of the foot); or
(iii) severe (signs as in moderate neuropathy plus trophic
disturbances, analgesia, loss of sensory modalities).
Additional signs and aspects in the medical history of the
patients are also mentioned where relevant. MRI of the
997
thoracal spinal cord and actual status of vestibular schwannomas are not described.
Neurophysiology
Nerve conduction velocity (NCV) studies and EMG methods
followed standard techniques using the electromyograph
Multilinerâ Evolution 1.63 (Jaeger/Toennies, Hoechberg)
(Kimura, 1989). Sensory nerve conduction studies were
performed antidromically for the sural nerve and orthodromically for the median nerve, at least on one side. Motor
nerve conduction studies were also done at least on one side
on the median and the tibial nerves. Semiquantitative EMG
was performed with concentric needle electrodes on the
tibialis anterior muscle and, in cases in which a neurogenic
pattern was found, an additional EMG of the gluteus medius
muscle was performed.
Laboratory analysis
In all patients with electrophysiological evidence for neuropathy we attempted to obtain the following laboratory data:
level of urea, creatinine, C-reactive protein, homocysteine,
serum folate, methylmalonic acid, glyco-haemoglobin or
fasting morning blood glucose levels, levels of free
triiodothyronine, free thyroxine and thyrotropin, quantitative
estimation of immunoglobulin subclasses G, A and M, serum
electrophoresis, anti-Borrelia burgdorferi antibodies, ANA
(antinuclear antibodies), HIV-infection diagnosis and
Treponema pallidum microhaemagglutination assay. In case
of an abnormal test result, further investigations were
undertaken, e.g. estimation of ENA (epithelial cell-derived
neutrophil-activating protein).
Neuroimaging
For brain imaging, sagittal, coronal, and transversal T1weighted and T2 sequences were examined. For spine
imaging, sagittal T2-weighted sequences, as well as sagittal
and transverse T1-weighted sequences were obtained. T1weighted transverse and sagittal images were also obtained
with intravenous administration of gadopentetate dimeglumine. All MRI images were assessed by independent
radiologists. Tumours were assessed for their number,
location, morphology and signal characteristics on the T1and T2-weighted images, and for signal behaviour before and
after administration of contrast medium. In Case 7, only CT
was performed because of a brainstem implant.
Sural nerve biopsy
A sural nerve specimen 1.6 cm in length was used for paraf®n
embedding and H&E (haematoxylin±eosin) staining of
sections, and for epoxy resin embedding. Semi-thin sections
of these plastic embedded blocks were stained with
paraphenylenediamine and toluidine blue. Ultrathin sections
998
A. D. Sperfeld et al.
Table 1 Summary of clinical data, NF2-related manifestations, present clinical neuropathic syndrome and
electrophysiological classi®cation of neuropathy in our NF2 patients
Cranial nerve
Case Age
Age of Family NF2
(years) onseta history severity schwannomas
(years)
1
2
37
28
17
14
Neg
Neg
Severe
Severe
3
37
26
Neg
Severe
4
5
56
28
30
12
Pose
Posg
Severe
Severe
6
27
19
Neg
Severe
7
39
14
Neg
Severe
8
32
17
Neg
Severe
9
31
18
Neg
Severe
10
11
12
48
37
49
44
20
35
Neg
Neg
Neg
Severe
Severe
Severe
13
30
26
Neg
Severe
14
15
57
33
34
15
Neg
Neg
Mild
Severe
Bilateral
Bilateral
NV
Bilateral
NV
Bilateral
Bilateral
NV
Bilateral
Menin- Spinal schwannomas
giomas
Cutaneous
Peripheral Electrophysioschwannomas neuropathy logical
classi®cation
of neuropathy
N VIII +
N VIII, +++
C2/C3; whole lumbar areab +++
+++
C2/C5c,b; cauda equina
Moderate
Severe
Ax
Ax
N VIII, ±
C7, L1, L2, cauda equinad
±
Moderate
Demy
N VIII ±
N VIII, +++
+
+++
Mild
Mild
Ax
Ax
+
Mild
Ax-demy
+++
C4, L3d; C6, C7f
C2/C5, C2-Th2b, L5, S1d;
L2, L3f; cauda equina
Conus medullaris, cauda
equina, L3/L4c
ND
+++
Mild
Ax
+
C3/C4c; C3f
+++
±
Ax
+++
C3/C4, C6/C7c; conus
+
medullaris, cauda equinad,c,f
±
Ax-demy
+++
+
+++
C1/C2c,b
C6, C7d,c; conus medullaris
Hole cervical areaf; cauda
equina
C7, Th1f; conus medullaris
±
±
±
±
±
±
Ax
±
±
±
±
±
None
C4f; L4/S1c
±
±
±
±
±
±
N VIII +++
Unilateral N IX
Bilateral N VIII,
NV
Bilateral N VIII,
NV
Bilateral N VIII;
Unilateral N V,
N IV
Bilateral N VIII
Bilateral N VIII
Bilateral N VIII
Bilateral N VIII; +++
Unilateral N V
Bilateral N VIII +++
Bilateral N VIII ±
aAge
of onset of NF2 related symptoms; bintramedullar; cextramedullar; dintraforaminal; epatient's son reported symptoms related to NF2;
was inherited from patient's father.
ND = not done; ± = absent; + = one to ®ve tumours; +++ = six or more tumours; Neg = negative; Pos = positive; N V = trigeminal nerve;
N IV = trochlear nerve; N VIII = vestibular nerve; N IX = glossopharyngeal nerve; L = lumbar nerve root; C = cervical nerve root; Th =
thoracic nerve root; S = sacral nerve root; Ax = axonal type of neuropathy; Demy = demyelinating type of neuropathy; Ax-demy = mixed
type of axonal±demyelinating neuropathy.
Classi®cation of neuropathic syndromes: mild = absent ankle jerks, slight symmetrical hypaesthesia; moderate = `mild' plus atrophy of
distal muscles or deformity of foot; severe = `moderate' plus trophic disturbances, severe hypaesthesia and severe hypalgesia or analgesia.
fextraspinal; gdisease
were contrast enhanced with lead citrate and uranyl acetate
for electron microscopy using routine methods.
Results
We have conducted a systematic investigation to determine
the occurrence and aetiology of peripheral nerve involvement
in NF2 patients. We investigated 15 patients with de®nite
NF2 and in 10 patients found evidence of neuropathy.
Clinical characteristics
Fifteen patients (seven females, eight males) were investigated. The mean age at evaluation was 37.9 years (range 27±
57 years). Clinical information relevant in respect to
peripheral neuropathy is given below. Other NF2 features,
including MRI ®ndings and patients with neither clinical nor
electrophysiological evidence of polyneuropathy, are summarized in Table 1.
Case 1
From age 34 years, this patient had observed exerciseinduced muscle cramps with distal predominance in the legs,
progredient balance and non-speci®c sensory disturbances. At
that time, actual neurological examination revealed signs of
mixed sensory and motor neuropathy. This patient carried a
non-sense mutation (C784T).
Case 2
From her early teens this 28-year-old female had bilateral
cramps in the calves, impaired balance, mild paresis of the
peroneal muscles, and dysaesthesia and hypaesthesia of the
lower extremities with distal predominance. A `polyneuropathy' was diagnosed. Clinical and neurological investigation
showed absent tendon re¯exes, plegia of the anterior tibialis
and peroneal muscles, severe paresis of the calf muscles and
of the bilateral intrinsic hand muscles. In addition, distal
analgesia and severe hypaesthesia of the lower extremities
Peripheral nerve in neuro®bromatosis type 2
999
Table 2 Electrophysiological investigation of all patients
Case
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Tibial nerve
Median nerve
Sural nerve
Median nerve
DML
[ms]
(<5.1)
CMAP
[mV]
(>5)
NCV
[ms]
(>40.6)
f-wave
latency
[ms]
DML
[ms]
(<4.2)
CMAP
[mV]
(>5)
NCV
[ms]
(>50)
f-wave
latency
[ms]
SNAP
[mV]
(>6)
NCV
[ms]
(>40.6)
SNAP
[mV]
(>6.6)
NCV
[ms]
(>46.9)
3.8
0
5
5.7
5.1
5.5
4
3.8
5.4
4.5
3.9
3
4.7
3.3
3.7
11.2
0
11.9
1.7
15
3.6
3
8.8
10
18
14.5
25
22.1
22
12.7
40
0
37.8
51.2
45
35
38.8
43.1
59
51
52
50.4
48
52
45
58.4 [<61]
0 [<58]
56.6 [<61]
NR [<61]
52 [<58]
54.6 [<61]
61 [<58]
57 [<61]
54 [<58]
48 [<58]
60.8 [<61]
48 [<58]
47 [<61]
48 [<58]
45.2 [<61]
3.1
6.4
3.9
2.9
4.4
4
ND
3.2
4
4.1
3.8
3.4
2.9
3.1
3
17.5
3.9
22.4
18.1
15
6
ND
14.4
17
5.2
18.4
11.7
10
17
6.4
63
49
61.4
54.9
53
56
ND
66
51
44
59
60
60
57
67.8
29 [<31.5]
NR [<30.5]
29.4 [<31.5]
NR [<31.5]
27 [<30.5]
29.4 [<31.5]
ND [<30.5]
26.4 [<31.5]
32 [<30.5]
NR [<30.5]
65.7 [<31.5]
27 [<30.5]
26 [<31.5]
24 [<30.5]
25 [<31.5]
3.1
0
8.5
3
4
0
3.5
4.4
0
3
11.5
21
6.4
16
12.9
58
0
42
51.5
51
0
49.7
46.4
0
59
46
44
52
60
48
ND
0
0
7.6
ND
12
ND
11.9
15
5.6
7.4
27
9.9
28
13.9
ND
0
0
49
ND
42
ND
45
43
50
42
51
55
54
49
EMG
Neur
Neur
Neur
Neur
Neur
Norm
ND
Norm
Norm
ND
Norm
Norm
Norm
Norm
ND
CMAP = distal compound motor action potential; SNAP = sensory nerve action potential; NCV = nerve conduction velocity; DML=distal
motor latency; 0 = no motor or sensory response; NR = non-reproducible; ND = not done; Neur = neurogenic pattern; Norm = normal.
Normal values are listed in parentheses below DML, CMAP, SNAP and NLG. Normal values were established in our laboratory
according to published age-matched, temperature and height corrected values (Ludin, 1993).
In Case 1, axonal lesion was found because of a reduced SNAP of the sural nerve and neurogenic changes including spontaneous activity
in EMG of the anterior tibialis muscle. The EMG of the gluteus medius muscle showed no abnormalities. In Case 2, no tibial nerve
CMAP, no SNAP of the sural and median nerve were detected. EMG revealed neurogenic pattern and axonal polyneuropathy was
diagnosed. Case 3 most likely suffered from demyelinating neuropathy. DML tibial nerve was in upper limit. NCV of the sural nerve was
reduced and SNAP of the median nerve absent. EMG showed mild chronic neurogenic pattern. Case 4 suffered from axonal neuropathy
because of prolonged DML, distinct reduced CMAP of the tibial nerve, reduced SNAP of the sural nerve and neurogenic pattern in EMG.
Case 5 had an axonal neuropathy because of prolonged DML of the tibial nerve and of the median nerve, reduced SNAP of the sural
nerve and neurogenic pattern in EMG. In Case 6, absent sural nerve SNAP, prolonged DML, reduced CMAP and reduced NCV of the
tibial nerve, and reduced sensible NCV of the median nerve were documented leading to mixed axonal±demyelinating neuropathy with
axonal component of demyelination. Case 8 showed reduced SNAP of the sural nerve. In Case 9 prolonged DML of the tibial nerve,
absent sural nerve potential, prolonged f-wave latency and reduced sensible NCV were documented. A mixed axonal±demyelinating
neuropathy was supposed. Case 10 showed reduced SNAP of the sural nerve and median nerve, CMAP at lower limit of the median nerve
and discrete reduced NCV of the median nerve. EMG was not done because of patients declining. An axonal lesion was documented. In
Case 11, prolonged f-wave latencies of the median and tibial nerve were documented.
and, following cutaneous innervation, areas of peripheral
nerves in upper extremities were observed. MRI of the lumbar
spinal cord showed multiple tumour masses of various size
involving multiple nerve roots of the cauda equina. In
addition, MRI of the lumbar plexus revealed enlarged
tumours along both sciatic nerves down to the sciatic
foramen.
Case 3
The parents of this patient noted bilateral foot deformity (pes
cavus) since childhood. On neurological examination he
showed a slightly unsteady gait, normal sensory modalities
(except for a right-accentuated impaired vibration sense of
hand and foot), and mild weakness and wasting of intrinsic
foot muscles, anterior tibialis and peroneal muscles. The deep
tendon re¯exes were sluggish.
Case 4
Clinical investigation showed absent ankle jerks, discrete
bilateral paresis of anterior tibialis and peroneal muscles, and
reduced vibration sense at lower extremities in this patient.
Case 5
This 22-year-old female had a 10-year history of sudden onset
progressive left thigh weakness and wasting. Occasional
fasciculations were evident. Later in life weakness of the
lower extremities progressed. MRI of the lumbar spine
showed an intradural schwannoma affecting the left-sided
fourth and ®fth nerve root and a left-sided extraforaminal
schwannoma at the second lumbar nerve root. After extirpation of tumour, clinical symptoms remained unchanged.
Neurological examination showed profound wasting of the
left gluteus medius, quadriceps and iliopsoas muscle, an
1000
A. D. Sperfeld et al.
Peripheral nerve in neuro®bromatosis type 2
absent left knee jerk, bilateral reduced vibration sense of the
lower extremities, and stand and gait ataxia.
Case 6
On neurological evaluation, gait disturbance, mild reduced
vibration sense and absent ankle jerks were found.
1001
symmetric polyneuropathy. Table 2 shows the classi®cation
of neuropathies. Ten patients (66.7%) revealed electrophysiological evidence of neuropathy. Electrophysiological
®ndings were often moderate but unambiguous. There were
seven patients with axonal neuropathy, one patient (Case 3)
with a demyelinating neuropathy and two patients (Cases 6
and 9) suffered from mixed axonal-demyelinating neuropathy. In Patients 12, 13, 14 and 15, no abnormalities were
found.
Case 7
Detailed neurological examination showed mild stand and
gait ataxia, and reduced ankle jerks. Other deep tendon
re¯exes, muscle strength and sensory function were normal
except for hypaesthesia of all right ®ngertips.
Cases 8±15
In Cases 8±15 no clinical evidences for a neuropathy were
found.
In summary, clinical investigation showed a motor and/or
sensory distal peripheral neuropathy in seven of 15 patients
(46.7%). One patient clinically presented with mononeuropathy multiplex. Following grading there were four patients
with a mild, two patients with a moderate and one patient with
a severe neuropathic syndrome. According to previously
published criteria, severity of grading in NF2 is based on
tumour burden and age of onset (Evans et al., 1992c; Parry
et al., 1994). Following these criteria, all patients except one
had a severe form of NF2, although few had late onset (which
is one criterion) or a mild form. One patient with mild NF2
showed no evidence for neuropathy. One patient (Case 3)
showed bilateral foot deformities from early childhood. Two
patients (Cases 2 and 3) suffered from neuropathic syndrome
years before they experienced common symptoms due to NF2
(Table 1).
Electrophysiological ®ndings
NCV studies were performed in all patients with neuropathy.
Table 2 summarizes the results of motor and sensory nerve
conduction studies. Electrophysiologically, all patients had a
Laboratory ®ndings
In all patients with electrophysiological criteria for neuropathy the laboratory investigations as listed above were
investigated. In none of the patients was evidence found for
other causes for neuropathy. Case 3 declined duplication
testing for hereditary motor sensory neuropathy type 1
(HMSN1A gene). In Case 2, laboratory investigations
revealed slightly elevated levels of ANA, but ENA levels
were in the normal range. Further investigations revealed no
evidence of in¯ammatory or immune-mediated diseases. In
this case, a sural nerve biopsy was performed. Five of our
patients underwent mutation analysis. One patient (Case 1)
carried a nonsense mutation (C784T). In one additional
patient, we found a loss of heterozygosity (LOH) in a
vestibular schwannoma.
Nerve biopsy ®ndings
The sural nerve of Case 2 comprised 10 nerve fascicles
surrounded by a perineurium, which was focally increased in
thickness. Some fascicles were moderately distended by
endoneurial oedema. Very few myelinated and unmyelinated
nerve ®bres were preserved. Rare clusters of some regenerated nerve ®bres were occasionally seen. Numerous bands of
BuÈngner indicated preceding loss of myelinated and unmyelinated axons (Fig. 1A and B). In addition, there were
either isolated pathological Schwann cells with sometimes
abnormally large nuclei (Fig. 1E and F), or groups of
abnormal Schwann cells arranged in small or larger complexes of multiple, ¯at, closely attached, interdigitating cell
Fig. 1 (A±F) Light and electron micrographs of the sural nerve biopsy of Case 4. (A) Semi-thin cross sections of two segments of nerve
fascicles showing severe loss of myelinated nerve ®bres with only two remaining ®bres (arrowheads), numerous bands of BuÈngner at the
site of degenerated nerve ®bres, and circumscribed complexes of abnormal cells (arrows). The oedematous area in (B) comprises
vacuolated cells and one unequivocal tumour cell with an abnormally large nucleus (arrowhead) adjacent to an inconspicuous capillary.
Toluidine blue staining. Magni®cation: 3360. (C) Borderline (arrowheads) between pathological Schwann cells on the right and normal
Schwann cells with (arrows) or without unmyelinated axons on the left. Collagen ®bres are surrounding the normal Schwann cells
individually whereas the pathological cells are coherent and arranged in a large complex of multiple, ¯at, interdigitating cells and cell
processes that are surrounded by a basal lamina. Clusters of collagen ®brils are located between the abnormal cells. Magni®cation:
31940. (D) Similar complex of abnormal Schwann cells as shown in (A), but focally separated from adjacent, pre-existing unmyelinated
nerve ®bres (arrows) and bands of BuÈngner by several layers of basal laminae (arrowheads). Many spaces between the abnormal cells are
®lled with electron optically lucent or ¯occular mucoid material containing only rare collagen ®laments. Magni®cation: 34600. (E and F)
Isolated tumour cells cut at the level of the nucleus with multiple indentations and ¯at processes at the periphery. Some of the cytoplasmic
vacuoles are covered by a basal lamina (arrowheads), indicating their connection with the extracellular space. No axons are enclosed by
the tumour cells. The severely enlarged (giant) nucleus in (F) shows several cytoplasmic invaginations with dilated components of the
endoplasmic reticulum. (E) Magni®cation: 36700. (F) Magni®cation: 39100.
1002
A. D. Sperfeld et al.
processes, surrounded by a basal lamina and separated by
electron optically empty extracellular spaces, some of which
contained collagen ®laments (Fig. 1C and D). No axons were
detected within these tumour cell complexes, which clearly
differed in appearance from bands of BuÈngner and onion bulb
formations. Some isolated Schwann cells with abnormally
long or semicircular processes surrounding empty spaces
with ¯occular material resembled pathological ones.
However, usually the abnormal cells showed a tendency
towards coherence and compact growth. The shape of the
pathological nuclei varied considerably. Some were indented
by cytoplasmic inclusions with dilated components of the
endoplasmic reticulum (Fig. 1F).
Comparison of the clinical, imaging and
electrophysiological data
The results are summarized in Table 1. There were three
patients with electrophysiological evidence of neuropathy
with no clinical abnormalities related to peripheral nerve
lesions and four patients with a mild neuropathic syndrome.
Only three patients showed a moderate or severe, and
therefore striking, neuropathic syndrome.
Overall, 10 patients showed cutaneous abnormalities
related to NF2, but only eight patients showed skin tumours
(subcutaneous schwannoma and `NF2 plaques'). Comparing
the presence of cutaneous abnormalities related to NF2 and
evidence for neuropathy, all eight patients with skin tumours
showed electrophysiological evidence for neuropathy. In
contrast, all but two patients with neuropathy showed NF2related skin tumours. The patient with demyelinating
neuropathy showed no cutaneous abnormalities (Case 3).
This patient, according to clinical and electrophysiological
criteria, may well suffer from demyelinating hereditary motor
sensory neuropathy. The second patient with neuropathy but
without cutaneous abnormalities is Case 10. In this patient, by
comparison, ®rst symptoms related to NF2 became evident
relatively late in life and until today only bilateral vestibular
schwannomas, various meningiomas and one intramedullar
tumour in the cervical spinal cord have been detected. In all
patients with multiple (more than six) skin tumours, axonal
neuropathy was documented. In both patients suffering from
mixed axonal-demyelinating neuropathy, less than six
cutaneous abnormalities related to NF2 were found. In
summary, this indicates that there is a possible relationship
between the presence of skin tumours and peripheral
neuropathy.
Comparing peripheral nerve lesions to MRI of the spinal
cord in Case 2, tumour masses on proximal nerve sections and
roots can be responsible for sensory and motor peripheral
symptoms. In Case 2, large tumour masses on sciatic nerve
sections in the pelvic region could explain reduced sensory
nerve action potential (SNAP) of sural nerves. However, on
upper extremities, which also had evidence of polyneuropathy on examination, comparable MRI images were not
found. Additionally, the neuropathic syndrome was the ®rst
clinical abnormality in this patient. Furthermore, a sural nerve
biopsy was performed in this patient. In all other cases with
documented intra- and/or extraforaminal schwannoma, these
results do not explain the clinical ®ndings. Only in Case 11
are isolated prolonged f-wave latencies of the median and
tibial nerve due to small schwannomas on the corresponding
proximal nerve roots.
Discussion
We present systematic data on the occurrence of peripheral
neuropathy in NF2. Clinical signs manifesting a peripheral
neuropathy occurred in 46.7% of patients. Electrophysiological examination revealed evidence of neuropathy in
66.7%. This frequency is higher than described so far.
Previously, polyneuropathy was observed in one larger series
in up to 6% of patients suffering from NF2. A possible reason
for this discrepancy might be a patient selection bias. Our
study was done in isolated patients subsequently recruited
from our outpatient clinic. Usually, these patients are looking
for medical care in the more advanced stages of disease. In
accordance with this, all patients with neuropathy had severe
NF2 according to previously published criteria, although a
few had late onset (Evans et al., 1992c; Parry et al., 1994). In
one patient, however, polyneuropathy developed years before
symptoms related to NF2 became evident. Family investigations were not performed. However, in a mouse model
describing conditional biallelic NF2 mutations, subclinical
peripheral nerve involvement including Schwann cell hyperplasia and myelination defects was found at a high frequency
(Giovannini et al., 2000). Clinical neuropathy ®ndings in our
patients varied from mild to severe neuropathic syndromes.
Three patients showed electrophysiological evidence of
neuropathy and no clinical signs of polyneuropathy. A mild
neuropathic syndrome documented in four patients consisted
of absent ankle jerks and slight symmetrical hypaesthesia.
Absent or reduced deep tendon re¯exes are known to be
caused by affected extra- or intraspinal nerve roots in NF2
patients. Hypaesthesia can also result from intraspinal tumour
masses (Evans et al., 1992a). In summary, seven of 10
patients with neuropathy showed no or mild clinical signs
related to neuropathy, leading to the assumption that the
lesion of the peripheral nerves might be subclinical or masked
by signs related to damage of the CNS and nerve roots,
respectively. On the other hand, two patients suffered from a
moderate neuropathic syndrome with absent ankle jerks,
symmetrical hypaesthesia and atrophy of distal muscles. In
one female we observed a severe neuropathy.
There have been few case reports describing patients with
initial polyneuropathy, where later in life other NF2-related
symptoms developed. One young boy suffered from muscle
wasting and progressive weakness since childhood. After the
diagnosis of a hereditary motor and sensory neuropathy,
various subcutaneous tumours developed. Electrophysio-
Peripheral nerve in neuro®bromatosis type 2
logical ®ndings were consistent with an axonal type of
neuropathy (Overweg-Plandsoen et al., 1996).
In our study there were seven patients with an axonal
neuropathy type, two patients with a mixed axonaldemyelinating neuropathy type, and one patient ful®lled
criteria for a demyelinating neuropathy (Ad Hoc Subcommittee of the American Academy of Neurology AIDS
Task Force, 1991). Predominant electrophysical signs of
axonal damage were found in one patient with mixed
polyneuropathy. Cases described in the literature are of
axonal or mixed type neuropathies, and those with detailed
electrophysiological data show a mainly axonal type of
neuropathy (Bosch et al., 1981; Thomas et al., 1990;
Overweg-Plandsoen et al., 1996; Iwata et al., 1998).
Further case reports document symptomatic focal amyotrophy without evidence of root or peripheral nerve tumours
showing axonal damage (Trivedi et al., 2000).
Peripheral nerve lesions in NF2 can result from tumour
masses within the proximal nerve roots, or along the
peripheral nerve (Grazzi et al., 1998). In our series there
were some patients with multinodular tumours in the area of
the conus and cauda equina, in the proximal nerve roots, or in
the extramedullar affecting neighbouring nerve roots. Lesions
described were possibly responsible for local abnormalities,
as in Case 5, in which proximal nerve tumour led to
amyotrophy of the unilateral thigh, or Case 2, with tumour
masses along the sciatic nerve. However, abnormal nerve
conduction velocities of the sensory nerves and reduced
SNAPs and the distribution of neurogenic changes revealed
by electrophysiological investigations clearly suggested that
the lesions were at the peripheral nerve level in our patients.
In clinical studies, evidence of skin tumours occurred
between 32% and 68%, and skin tumours appeared to be of
two or three different types (Kanter et al., 1980; Evans et al.,
1992b; Parry et al., 1994). In our study, skin tumours due to
NF2 were found in 10 of 15 patients (66.7%). One patient
was not included in this calculation due to the presence of
a histologically proven lipoma. Clinical differentiation
between cutaneous tumour types as described in the literature
was not systematically achieved (Evans et al., 1992b). We
differentiated patients with more than six (multiple) and less
than six cutaneous tumours. Interestingly, multiple dermal
schwannoma tumours on the back were often situated in a
paraspinal position corresponding to the region of the short
spinal nerves. Others found that the most common location
was the back, with no further details listed (Parry et al., 1994).
Based on current knowledge of the origin of skin tumours,
observations in mouse models and our present ®ndings, we
show that there is a relation between the presence of skin
tumours and peripheral neuropathy: a peripheral neuropathy
was found in all of our patients with cutaneous schwannoma,
and most patients with neuropathy showed cutaneous lesion
related to NF2. In the patient with a demyelinating type of
neuropathy and one female with an axonal type of neuropathy, no cutaneous abnormalities were found. As described
1003
above, the latter only developed clinical signs related to NF2
later in life.
In conclusion, polyneuropathy in NF2 was much more
common than expected when electrophysiological methods
were applied for examination. We noted an apparent correlation between disease duration and the development of skin
tumours that were associated with peripheral nerve lesions.
In accordance with previous histopathological data from
case reports we found ®bre loss, Schwann cell complexes and
increased collagen. Previously, however, the Schwann cells
complexes were described as somehow irregular `coreless
onion bulbs' (Thomas et al., 1990; Iwata et al., 1998). The
most interesting ®nding in ultrastructural pathology in our
biopsy, however, was the demonstration of `de-differentiated'
Schwann cells, either isolated or in complexes. The multiple
interdigitating cell processes are very well compatible with a
malfunction of mutated merlin, since merlin supposedly acts
as a membrane±cytoskeleton linker. During the preparation
of this manuscript, Gijtenbeek and colleagues have published
a report on a patient with NF2 and similar peripheral nerve
pathology (Gijtenbeek et al., 2001). Based on these ®ndings
we hypothesize that the aetiology of this frequent peripheral
neuropathy syndrome in NF2 is caused by: (i) compression
effects of multiple tumourlets originating along the length of
the peripheral nerves on adjacent nerve ®bres; (ii) unknown
local toxic or metabolic in¯uences of the endoneurial
pathological cells on adjacent nerve ®bres, which requires
further clari®cation; or (iii) the inability of these cells to
adhere to/ensheath the axon.
References
Ad Hoc Subcommittee of the American Academy of Neurology
AIDS Task Force. Research criteria for diagnosis of chronic
in¯ammatory demyelinating polyneuropathy (CIDP). Neurology
1991; 41: 617±8.
Bosch EP, Murphy MJ, Cancilla PA. Peripheral neuro®bromatosis
and peroneal muscular atrophy. Neurology 1981; 31: 1408±14.
Evans DG, Huson SM, Donnai D, Neary W, Blair V, Newton V,
et al. A clinical study of type 2 neuro®bromatosis. Q J Med 1992a;
84: 603±18.
Evans DG, Huson SM, Donnai D, Neary W, Blair V, Newton V,
et al. A genetic study of type 2 neuro®bromatosis in the United
Kingdom. II. Guidelines for genetic counselling. J Med Genet
1992b; 29: 847±52.
Evans DG, Huson SM, Donnai D, Neary W, Blair V, Teare D, et al.
A genetic study of type 2 neuro®bromatosis in the United Kingdom.
I. Prevalence, mutation rate, ®tness, and con®rmation of maternal
transmission effect on severity. J Med Genet 1992c; 29: 841±6.
Evans DG, Trueman L, Wallace A, Collins S, Strachan T.
Genotype/phenotype correlations in type 2 neuro®bromatosis
(NF2): evidence for more severe disease associated with
truncating mutations. J Med Genet 1998; 35: 450±5.
Evans DG, Sainio M, Baser ME. Neuro®bromatosis type 2.
[Review]. J Med Genet 2000; 37: 897±904.
1004
A. D. Sperfeld et al.
Gijtenbeek JMM, Gabreels-Festen AA, Lammens M, Zwarts MJ,
van Engelen BG. Mononeuropathy multiplex as the initial
manifestation of neuro®bromatosis type 2. Neurology 2001; 56:
1766±8.
Giovannini M, Robanus-Maandag E, van der Valk M, NiwaKawakita M, Abramowski V, Goutebroze L, et al. Conditional
biallelic Nf2 mutation in the mouse promotes manifestations of
human neuro®bromatosis type 2. Genes Dev 2000; 14: 1617±30.
Grazzi L, Chiapparini L, Parati EA, Giombini S, D'Amico D, Leone
M, et al. Type II neuro®bromatosis presenting as quadriceps
atrophy. Ital J Neurol Sci 1998; 19: 94±6.
Gutmann DH. Molecular insights into neuro®bromatosis 2.
[Review]. Neurobiol Dis 1997; 3: 247±61.
Gutmann DH, Aylsworth A, Carey JC, Korf B, Marks J, Pyeritz RE,
et al. The diagnostic evaluation and multidisciplinary managment of
neuro®bromatosis 1 and neuro®bromatosis 2. [Review]. JAMA
1997; 278: 51±7.
Gutmann DH, Geist RT, Xu HM, Kim JS, Saporito-Irwin S. Defects
in neuro®bromatosis 2 protein function can arise at multiple levels.
Hum Mol Genet 1998; 7: 335±45.
Iwata A, Kunimoto M, Inoue K. Schwann cell proliferation as the
cause of peripheral neuropathy in neuro®bromatosis-2. J Neurol Sci
1998; 156: 201±4.
Kanter WR, Eldridge R, Fabricant R, Allen JC, Koerber T. Central
neuro®bromatosis with bilateral acoustic neuroma: genetic, clinical
and biochemical distinctions from peripheral neuro®bromatosis.
Neurology 1980; 30: 851±9.
Kilpatrick TJ, Hjorth RJ, Gonzales MF. A case of neuro®bromatosis
2 presenting with a mononeuritis multiplex. J Neurol Neurosurg
Psychiatry 1992; 55: 391±3.
Kimura J. Electrodiagnosis in diseases of nerve and muscle:
principles and practice. 2nd ed. Philadelphia: Davis; 1989.
Ludin HP. Praktische Elektromyographie. Stuttgart: Enke Verlag;
1993.
MacCollin M. Clinical aspects. In: Friedman JM, Gutmann DH,
MacCollin M, Riccardi VM, editors. Neuro®bromatosis. 3rd ed.
Baltimore: Johns Hopkins University Press; 1999. p. 299±326.
MacCollin M, Gusella J. Neuro®bromatosis type 2. In: Vogelstein
B. Neuro®bromatosis type 2. New York: McGraw-Hill; 1998. p.
443±53.
Martuza RL, Eldridge R. Neuro®bromatosis 2 (bilateral acoustic
neuro®bromatosis). [Review]. New Engl J Med 1988; 318: 684±8.
Mautner VF, Tatagiba M, Lindenau M, Funsterer C, Pulst SM,
Baser ME, et al. Spinal tumors in patients with neuro®bromatosis
type 2: MR imaging study of frequency, multiplicity, and variety.
AJR Am J Roentgenol 1995; 165: 951±5.
Onishi A, Nada O. Ultrastructure of the onion bulb-like lamellated
structure observed in the sural nerve in a case of von
Recklinghausen's disease. Acta Neuropathol (Berl) 1972; 20:
258±63.
Overweg-Plandsoen WC, Brouwer-Mladin R, Merel P, de Vries L,
Bijlsma EK. Neuro®bromatosis type 2 in an adolescent boy with
polyneuropathy and a mutation in the NF2 gene. J Neurol 1996;
243: 724±6.
Parry DM, Eldridge R, Kaiser-Kupfer MI, Bouzas EA, Pikus A,
Patronas N. Neuro®bromatosis 2 (NF2): clinical characteristics of
63 affected individuals and clinical evidence for heterogeneity. Am
J Med Genet 1994; 52: 450±61.
Rouleau GA, Merel P, Lutchman M, Sanson M, Zucman J,
Marineau C, et al. Alteration in a new gene encoding a putative
membrane-organizing protein causes neuro®bromatosis type 2.
Nature 1993; 363: 515±21.
Thomas PK, King RH, Chiang TR, Scaravilli F, Sharma AK,
Downie AW. Neuro®bromatous neuropathy. Muscle Nerve 1990;
13: 93±101.
Trivedi R, Byrne J, Huson SM, Donaghy M. Focal amyotrophy in
neuro®bromatosis 2. J Neurol Neurosurg Psychiatry 2000; 69:
257±61.
Trofatter JA, MacCollin MM, Rutter JL, Murrell JR, Duyao MP,
Parry DM, et al. A novel moesin-, ezrin-, radixin-like gene is a
candidate for the neuro®bromatosis 2 tumor suppressor. Cell 1993;
72: 791±800.
Received April 30, 2001. Revised August 6, 2001.
Second revision November 7, 2001.
Accepted December 6, 2001