Primary diffuse multinodular leptomeningeal gliomatosis

Surgical Neurolog

Neoplasm

Primary diffuse multinodular leptomeningeal gliomatosis

Case report and review of the literature

Bertrand Debono, MDa,T, Stephane Derrey, MDa, Charles Rabehenoina, MDa,

Francois Proust, MD, PhDa, Pierre Freger, MD, PhDa, Annie Laquerriere, MD, PhDb

aDepartment of Neurosurgery, and bPathology Laboratory, Rouen University Hospital–Charles Nicolle, 76031 Rouen Cedex, France

Received 10 January 2005; accepted 15 June 2005

Abstract Background: Primary diffuse leptomeningeal gliomatosis is an exceptional neoplasm, and only

www.surgicalneurology-online.com

0090-3019/$ – see fro

doi:10.1016/j.surneu.2

Abbreviations: CN

DNA, deoxyribonucle

hematoxylin-eosin; I

resonance imaging; N

ningeal gliomatosis; W

T Correspondinga

E-mail address: b

30 cases have been reported in the literature. We report a recent case and compare data with

previously published observations.

Methods: A 50-year-old man was admitted to the neurosurgery department for a previous 4-month

history of headache, associated with nonspecific neurological signs. Biologic data and cerebrospinal

fluid examination suggested an inflammatory process. The patient was given an antituberculous

therapy. Magnetic resonance imaging revealed a multinodular enhancement of spinal nerve roots. A

biopsy of sacral rootlets was performed. Histological examination revealed an anaplastic

astrocytoma. Patient’s status worsened, and death occurred 7 months later.

Results: Complete neuraxis postmortem examination revealed no intraparenchymatous glioma and

was conclusive for the diagnosis of primary leptomeningeal gliomatosis (astrocytic, World Health

Organization grade III), with a multinodular pattern in the spinal cord, the brainstem, and the brain

base with diffuse extension into the cerebellar subarachnoid spaces.

Conclusions: Our case illustrates the diagnostic difficulties in making the premortem diagnosis. The

review of the literature indicates that there are no specific clinical or biologic signs. Magnetic

resonance imaging using T1-weighted images with gadolinium enhancement and biopsy material

may be useful diagnostic tools. In most cases, autopsy evaluation alone permits definitive primary

diffuse leptomeningeal gliomatosis diagnosis. Whatever the histological characteristics of

proliferating cells are, the prognosis remains poor. No prognostic factors have been shown to be

correlated with survival time. Unfortunately, no routine treatment has been yet proposed.

D 2006 Elsevier Inc. All rights reserved.

Keywords: Primary diffuse meningeal gliomatosis; Diagnosis; Prognosis; Treatment; Literature review

1. Introduction

Primary diffuse leptomeningeal gliomatosis (PDLG) is a

rare and rapidly fatal disease characterized by diffuse

infiltration of the meninges by greater or lesser differenti-

ated neoplastic glial cells without evidence of CNS

nt matter D 2006 Elsevier Inc. All rights reserved.

005.06.038

S, central nervous system; CSF, cerebrospinal fluid;

ic acid; GFAP, glial fibrillary acidic protein; HE,

RP, intracranial raised pressure; MRI, magnetic

F, neurofilaments; PDLG, primary diffuse leptome-

HO, World Health Organization.

uthor.Tel.: +330232888042; fax:+330232888426.

[email protected] (B. Debono).

intraparenchymatous primary tumor. This condition was

first described by Moore [2,23] in 1954 and must be

distinguished from secondary meningeal gliomatosis, be-

cause of primary CNS tumoral extension [32]. PDLG has

been described as arising most frequently along the spinal

cord, particularly in patients with CNS congenital dysra-

phic malformations [1,13,34]. Clinical presentation

includes a variety of neurological symptoms, where raised

intracranial pressure is the most constant sign [1,10]. Due

to the lack of specificity and of the variability of clinical

presentation, clinical diagnosis is often impossible, and a

biopsy is required to confirm the diagnosis. PDLG is

associated with poor survival, and to date, its histogenesis

remains controversial.

y 65 (2006) 273–282

Fig. 1. MRI T1-weighted and T2-weighted images of spinal multinodular

PDLG. A: Bilateral enhancing nodules at the thoracolumbar cervical level.

B: Similar masses at the lumbosacral level.

B. Debono et al. / Surgical Neurology 65 (2006) 273–282274

We report a patient with PDLG, confirmed by macro-

scopic and histological postmortem evaluations. Patient

initially presented with symptoms of an inflammatory

condition affecting the meninges, with uncommon MRI

findings characterized by multinodular involvement of the

spinal cord.

2. Case report

2.1. Clinical data

A previously healthy 50-year-old man, with no personal

or familial medical history, was referred to a neurologist

with a 4-month history of headache, dizziness, and right

Fig. 2. Macroscopic autopsy findings. A: Bilateral nodular thickening of the ner

masses attached to the dura. C: Macroscopic pattern of the anterior brainstem;

2 vertebral arteries. D: Coronal section of the brain passing through the optic chi

optic chiasm by the leptomeningeal proliferation (arrowhead).

hypoacusia, followed by a progressive deterioration of

clinical status, including asthenia, anorexia, and weight loss.

Neurological examination revealed bilateral papilledema,

neck stiffness, left hemiparesis of the face with a left

Babinski sign, right hypoacusia, nystagmus and abduction

weakness of the right eye, and a slight facial asymmetry.

Otherwise, physical examination was normal, with no sign

of infection.

MRI on T1-weighted images of the brain and spinal cord

showed contrast-enhancing multiple nodules at the base of

the brain, at the level of the anterior and lateral brainstem

corresponding to the origin of cranial nerves; as well as

intradural extramedullary nodules at the level of anterior

cervical spinal cord (Fig. 1A), lumbar nerve roots, and

cauda equina were observed (Fig. 1B). No intra-axial lesion

was seen. Repeated CSF examinations constantly showed

exclusive lymphocytosis (6-66/mm3), associated with ele-

vated protein content (1.6-2.2 g/L) and hypoglycorrhachia

(2.2-2.6 mmol/L). CSF cytology for either lymphomatous or

carcinomatous malignant cells was negative, as well as

aerobic and anaerobic cultures, cryptococcal antigens, and

fungal cultures. Blood tumor markers were normal, with the

exception of a slight increase in malignancy CA125 marker.

Angiotensin-converting enzyme levels were normal in CSF

and blood, excluding a possible diagnosis of neurosarcoi-

dosis. An extensive screening for an autoimmune disease

(Wegner disease, systemic lupus erythematosus, and Gou-

gerot-Sjfgren syndrome) was negative, as well as for unusual

infectious processes (listeriosis, Whipple disease, Lyme

disease, brucellosis, syphilis, histoplasmosis, toxoplasmosis,

cysticercosis, candidosis, cryptococcosis, Epstein-Barr virus

ve roots. B: Presence of several whitish and relatively well-circumscribed

arrowhead indicates the presence of a bulky nodule located between the

asm, showing no intraparenchymatous lesion, with bilateral encasing of the

Fig. 3. Representative sections of histological postmortem findings. A:

Atypical astrocytic proliferating cells (HE, magnification �400). B: Areas

of pilocytic differentiation, with the presence of Rosenthal fibers

(arrowhead) (HE, magnification �250). C: Diffuse spreading of PDLG,

restricted to the cerebellar meninges (HE, magnification �100). D: Clear-

cut demarcation of PDLG from the spinal parenchyma (HE, magnification

�100). E: Strongly GFAP-immunoreactive tumor cells, surrounding

glomeruloid vessels (arrowhead) (magnification �250). F: NF-immuno-

chemistry displaying the absence of invasion of spinal nerve roots by glial

proliferating cells (magnification �100).

B. Debono et al. / Surgical Neurology 65 (2006) 273–282 275

disease). Tuberculous meningitis was therefore suspected,

and the patient was started on triple antituberculous therapy

combined to corticosteroids. One month later, the patient

Fig. 4. Flow cytometric analysis performed on surgical biopsy. Representative D

hyperdiploid peak (DNA index, 1.74).

status worsened with progressive physical deterioration,

consisting of permanent severe headache and lumbar pain,

right pyramidal signs, persistent left facial palsy, and

intensification of dizziness with gait difficulties and falls.

A biopsy of a nodular lesion infiltrating a rootlet at the L2

level was performed 1 month after admission. Histological

examination of biopsy material was conclusive for an

anaplastic astrocytic glioma, WHO grade III. Despite high

doses of corticosteroid therapy, the patient developed

bilateral pyramidal signs and sixth cranial nerve palsy,

complete bilateral deafness and blindness, and intracranial

raised pressure (IRP). The patient became increasingly

cachectic and confused, and died 7 months later.

2.2. Autopsy findings

An autopsy restricted to the neuraxis was carried out. On

macroscopic examination, the scalp and the skull were

normal. The brain weighed 1440 g. The dural venous

sinuses were normal, but the leptomeninges were opaque

and showed focal areas of whitish thickening, affecting the

base of the brain. Multiple well-circumscribed nodules, up

to 5 mm in diameter, or irregular thickening of the meninges

extended along the spinal cord, located on nerve roots

(Fig. 2A), with invasion of the dura (Fig. 2B). In the pons,

periventricular hemorrhages were observed. Similar masses

or meningeal thickening also affected the anterior brainstem

(Fig. 2C), the optic chiasm, and cranial nerves (Fig. 2D). On

serial coronal sections, no definite intraparenchymatous

tumor was observed.

Histological examination was performed on multiple

samples, taken from the meninges, spinal cord, brainstem

NA histogram, showing an hypodiploid peak (DNA index, 0.90) and an

Table 1

Characteristics of the reported cases of PDLG

No. Case Age, y

(sex)

Meningeal

location

Symptoms Follow-up

(mo)

Clinical

diagnosis

Neurosurgery Postoperative

diagnosis

Treatment Necropsy Histological type

1 Korein et al

(1957) [21]

16 (M) Cb, Cbl Behavioral

changes, S,

visual disturbance

18 Serous

meningitis

Biopsy Negative Lumbar Puncture Yes Anaplastic

oligodendroglioma

2 Sumi and

Leffman

(1968) [36]

61 (M) Cb IRP, confusion,

hemiplegia,

visual loss

6 Chronic

meningitis

NP – Antibiotics Yes Low-grade glioma

3 Ho et al

(1981) [16]

55 (M) Cb, Cbl,

BrS, SpC

S, IRP, multiple

CrN palsy

3 Tuberculous

meningitis

NP – Antituberculous drugs Yes Low-grade

astrocytoma

4 Kitahara et al

(1985) [19]

15 (F) Cb, Cbl IRP, tetraparesis,

multiple CrN palsy

42 Meningeal

tumor of

unknown

origin

NP – Intrathecal

chemotherapy,

radiotherapy (60 Gy)

Yes Low-grade

astrocytoma

5 Bailey and

Robitaille

(1985) [3]

53 (M) Cbl, SpC IRP, depression,

multiple CrN palsy

15 Tuberculous

meningitis

Occipital biopsy Negative Antituberculous and

steroids and antifungal

medication

Yes Possible

DNET with

oligodendroglial

differentiation

6 Whelan et al

(1987) [39]

42 (M) Cbl, BrS IRP, confusion, b3 ND NP – Shunt Yes Fibrillary

astrocytoma

7 Janisch et al

(1991) [17]

22 (F) Cb, SpC Hemiparesis,

multiple CrN

palsy

6 Inflammatory

arachnopathy

Biopsy Negative Steroids Yes Low-grade

astrocytoma

8 Dietrich et al

(1996) [10]

63 (F) Diffuse:

CrB, BrS;

multinodular:

Cb

IRP, gait

disturbance,

meningeal signs,

coma

3 Tuberculous

meningitis

or vasculitis

Temporal lobe

biopsy normal

Biopsy normal Antituberculous

drugs, steroids,

cyclophosphamide

Yes Astrocytoma IV

9 Leproux et al

(1993) [22]

13 (F) CrB, SpC IRP, meningeal

signs, dorsal

pain, epilepsy

ND ND Vermian biopsy Astrocytic

proliferation

Craniospinal

radiotherapy,

chemotherapy

Yes Intermediate

malignancy

astrocytoma

10 Fayet et al

(1994) [11]

53 (F) Cbl, SpC IRP 6 Meningitis NP – Antituberculous drugs Yes Astrocytoma

11 Carpentier et al

(1994) [5]

44 (F) Diffuse IRP, S, multiple

CrN palsy

5 Tuberculous

meningitis

Frontal

meningeal and

cortical biopsy

Arachnoiditis Antituberculous drugs,

steroids, methotrexate

Yes Anaplastic

oligoastrocytoma III

12 Radhakrishnan

et al (1994) [31]

12 (F) Diffuse IRP, S b3 Tuberculous

meningitis

NP NP Antituberculous drugs Yes Astrocytoma III

13 Pingi et al

(1995) [29]

19 (M) Diffuse IRP 3 ND NP NP ND Yes Anaplastic

astrocytoma

14 Giordana et al

(1995) [14]

49 (F) Diffuse IRP, S, multiple

CrN palsy

b3 ND T9 laminectomy Glioblastoma Chemotherapy Yes Glioblastoma

B.Debonoet

al./Surgica

lNeurology65(2006)273–282

276

15 Chen et al

(1995) [6]

17 (F) Diffuse IRP, transient

visual loss,

papilledema

26 ND Frontal

biopsy

Diffuse

leptomeningeal

gliomatosis,

oligodendroglial-type,

grade II

Combined

chemotherapy,

autologous bone

marrow

transplantation

Yes Oligodendroglioma II

16 Rogers et al

(1995) [33]

21 (M) Diffuse IRP, diplopia 48 ND Biopsy of

cerebellar

meninges

Oligodendroglioma II Radiotherapy,

chemotherapy

Yes Oligodendroglioma III

17 Park et al

(1996) [27]

21 (F) BrS, SpC IRP, diplopia 5 ND Cervical

laminectomy

Meningeal nodule

attached to

the dorsal

nerve roots:

low-grade

neoplasm

Radiotherapy,

chemotherapy,

steroids

Yes Astrocytoma II

18 Olivera-Leal

et al

(1997) [26]

24 (F) Cbl, SpC IRP,

lumbosacral

pain,

tetraparesis

4 Vasculitis or

demyelinating

disorder

NP PDLG Steroids Yes Low-grade

astrocytoma

19 Beauchesne

et al

(1998) [4]

17 (M) Cb, SpC IRP,

meningeal

signs, coma

24 PDLG Frontal

meningeal

biopsy

Biopsy 1,

negative;

biopsy 2,

astrocytoma II

Steroids,

chemotherapy,

radiotherapy

No –

20 Paulino et al

(1999) [28]

9 (M) Diffuse IRP, multiple

CrN palsy

36

(Alive)

Pseudotumor

cerebri

Frontal

meningeal

and cortical

biopsy

Malignant

leptomeningeal

gliomatosis

Steroids,

chemotherapy,

radiotherapy

(brain, 45 Gy;

and

spine, 40 Gy)

No No

21 Kastenbauer

et al

(2001) [18]

25 (M) Diffuse: SpC;

multinodular:

BrS, Cbl

IRP 5 CNS

infection then

pseudotumor

cerebri

Cerebellar

stereotactic

biopsy

Leptomeningeal

anaplastic

astrocytoma III

Antiviral and

antituberculous

drugs

Yes Leptomeningeal

anaplastic

astrocytoma III

22 Rees et al 1

(2001) [32]

34 (F) – IRP, diplopia 1.5 Tuberculous

meningitis

Lumbar

meningeal

biopsy

PDLG Steroids and

antituberculous

drugs

No –

23 Rees, et al 2

(2001) [32]

28 (M) – IRP 5 Tuberculous

meningitis

No No Steroids and

antituberculous

drugs

No –

24 Rees et al 3

(2001) [32]

51 (M) Diffuse IRP, back pain 2 Tuberculous

meningitis

then neoplastic

meningitis

No No Antituberculous drugs

and steroids then

chemotherapy

(intrathecal)

Yes Anaplastic

astrocytoma III

(continued on next page)

B.Debonoet

al./Surgica


277

Table 1 (continued)

No. Case Age, y

(sex)

Meningeal

location

Symptoms Follow-up

(mo)

Clinical

diagnosis

Neurosurgery Postoperative

diagnosis

Treatment Necropsy Histological type

25 Baborie et al

(2001) [1]

71 (M) Multinodular

thickening of

meninges

IRP, back

pain,

diplopia,

visual

impairment

4 Tuberculous

meningitis

No No Steroids and

antituberculous

drugs

Yes Anaplastic

astrocytoma III

26 Tsui et al

(2001) [37]

23 (F) Diffuse: CrB IRP, S b3 Possible

encephalitis

then

tuberculous

meningitis

No No Steroids and

antituberculous

drugs

Yes Astrocytoma III

27 Corsten et al

(2001) [8]

44 (M) CrB, SpC,

CaEq

IRP, S,

multiple

CrN palsy

NP Tuberculous

meningitis

Thoracolumbar

biopsy

Glioblastoma

multiforme

Steroids and

antituberculous

drugs then

radiotherapy

(brain, 50 Gy;

and

spine, 40 Gy)

No –

28 Havlik et al

(2001) [15]

28 (F) Nodular: Cb,

Cbl, BrS,

CrN, SpC,

CaEq

IRP b3 – – – – Yes Astrocytoma II

29 Our case

(2004)

50 (M) Nodular: Cb,

Cbl, BrS,

CrN, SpC,

CaEq

IRP, S,

multiple

CrN

palsy

7 Tuberculous

meningitis

Biopsy CaEq Leptomeningeal

anaplastic

astrocytoma III

Steroids and

antituberculous

drugs

Yes Astrocytoma III

Cb indicates cerebral; Cbl, cerebellar; BrS, brainstem; SpC, spinal cord; CrN, cranial nerve; CrB, cranial base; CaEq, cauda equina; Gy, Grays; S, seizure; DNET, dysembryoplastic neuroepithelial tumor;

ND, no data available; NP, not performed.

B.Debonoet

al./Surgica


278


and cerebellum, and from the cerebral hemispheres. Seven-

micrometer sections were stained using HE. In all structures

assessed, a diffuse glial proliferation infiltrating themeninges,

with some protrusion into Virchow-Robin spaces in a nodular

fashion. Tumor cells were pleomorphic, composed of an

admixture of ovoid and spindle-shaped cells exhibiting

irregular and vesicular or multilobulated nuclei, with a

relatively abundant faintly eosinophilic cytoplasm (Fig. 3A).

The presence of some elongated bipolar-shaped elements

suggested a piloid differentiation, as well as the presence of

Rosenthal fibers (Fig. 3B). Tumor cells were arranged in

short fascicles, separated by a prominent vascular network,

often displaying a glomeruloid pattern. Scattered giant multi-

nucleated cells and abnormal mitoses were observed, but

necrosis and endotheliocapillar proliferation were absent.

Neither tumoral proliferation nor parenchymal destruction

within the brain, cerebellum (Fig. 3C), or spinal cord

(Fig. 3D) was observed. However, tumor cells encased optic

chiasm and cranial nerves, particularly the trigeminal nerve.

Although brain convexity was spared, the corticospinal tracts

were compressed by the leptomeningeal proliferation and

exhibited degenerative changes. There was no associated

lesion, that is, meningiomatosis or schwannomatosis, which

could suggest the existence of a phakomatosis.

Immunohistochemical studies were performed using anti-

bodies directed against glial fibrillary acidic protein (GFAP;

diluted 1:100), vimentin (1:300), S100 protein (PS100;

1:500), smooth muscle actin (1:200), desmin (1:100), neuro-

filaments (NF; 1:200), p53 protein (1:100), the epithelial

membrane antigen (1:120), the melanocytic marker HMB45

(1:200), and the proliferative antigen Ki-67 (1:100; Dako-

patts, Trappes, France).

Tumor cells were strongly immunoreactive for GFAP

(Fig. 3E) and vimentin, and weakly for PS100. Smooth

muscle actin underlined the glomeruloid hypervasculariza-

tion. Epithelial membrane antigen, as well as desmin and

HMB45, was negative, which permitted to rule out a

possible meningeal sarcomatosis, carcinomatosis, or mela-

nomatosis. In the spinal cord, NF was positive in nerve roots

and pyramidal tracts (Fig. 3F), which revealed numerous

axonal spheroids. Ki-67 immunolabeling varied from 1%

to 5% on biopsy samples but was not available on post-

mortem blocks. For the same reason, p53 immunolabeling

was negative.

A retrospective DNA flow cytometric study, performed on

biopsy material and on 5 postmortem paraffin-embedded

blocks, revealed DNA multiploidy on the biopsy material,

with DNA indices of 0.90, 1, and 1.74 (Fig. 4). Postmortem

block analysis was inconclusive. Proliferative index of the

hyperdiploid population was 9.4% and represented 25% of

the cells analyzed. Cell cultures were carried out, but

karyotypic analysis was not available, because cells failed

to grow. The final neuropathologic diagnosis of primary

diffuse multinodular leptomeningeal gliomatosis, of ana-

plastic astrocytic type, grade III according to the WHO

classification was established.

3. Discussion

As in previous reports, our case illustrates the difficulties

encountered in performing the clinical diagnosis of PDLG.

According to Cooper and Kernohan [7], the definite

diagnosis of PDLG is based on the following criteria: no

apparent attachment of extramedullary meningeal tumor to

the neural parenchyma, no evidence of primary neoplasia

within the neuraxis, and the existence of distinct leptome-

ningeal encapsulation around the tumor. Thus, this defini-

tion has always implied a complete postmortem

neuropathologic examination, to exclude a primary paren-

chymal tumor, such as oligodendroglioma, medulloblasto-

ma, and gliomatosis cerebri, where diffuse leptomeningeal

invasion is a well-recognized and frequent complication,

initially described by Russell and Cairns in 1930 and present

in 4% to 30% of the series published [11,12,33]. Neverthe-

less, and particularly after radiation or chemotherapy, a

small intraparenchymatous tumor could have disappeared at

the time of postmortem examination, leading to morpho-

logical diagnostic difficulties [5,6,9,33].

PDLG is a very rare condition, and to our knowledge, only

30 cases have been previously reported. The extensive review

of the literature is summarized in Table 1, which includes

clinical, surgical, follow-up, and diagnostic hallmarks. The

mean age of clinical onset was 35 years, ranging from 9 to

71 years, with no sex predominance (sex ratio male/female,

1:1.06). Our case also illustrates the wide constellation of

symptoms and signs that PDLG may manifest. IRP was the

most frequent sign, observed in 83.8% (26/31) of patients

studied. Cranial nerve palsy was the second most frequent

sign, present in 35.5% (11/31) of cases. Epilepsy was

observed in 29.03% (9/31) of cases. Other signs were less

frequent such as visual disturbance (19.3%), confusion

(19.3%), back pain (16.1%), and tetraparesis (9.7%) [19].

In all cases, clinical presentation depended on the preferential

location of the lesions but appeared to develop in 2 stages.

The prodromal phase consisted of various signs, such as

subacute meningism, headaches, or confusion, suggesting a

psychiatric condition or an encephalomyelitis. A second

phase followed, consisting of nonspecific neurological

impairment, which could mimic numerous neurological

diseases [3,11,28,31].

CSF examination (obtained in 27 cases) had confirmed an

increase of protein concentration in all cases, although

glucose levels were low in 68% of cases. Cytological

examination revealed a low or moderate pleocytosis.

Exclusive lymphocytosis was infrequent (2 cases), and in

only 1 case CSF was considered acellular. Conversely to

secondary meningeal spread from gliomas, carcinomas, or

lymphomas, repeated CSF examinations generally did not

show any malignant cells. In the blood, a biologic inflam-

matory syndrome has been rarely reported [15,36].

In our patient, neuroradiological findings were similar to

those of Kastenbauer et al [18], unusual for the bilateral

nodular pattern observed in the spinal cord. Most frequently,


neuroimaging was nonspecific: ventriculomegaly in 50% of

cases, with diffuse or focal contrast-enhancing leptomenin-

geal thickening. Spinal cord involvement was predominant

(17/29 cases, 58.6%), and lesions were diffuse or involved

only 1 level (cervical, dorsal, or lumbar). These leptome-

ningeal abnormalities can be restricted to the spinal cord or

are also associated with contrast enhancement of the basal

cisterns (our case and 8 other reported cases), the

cerebellum, brainstem, and convexity (4 cases in each

location). Cranial base infiltration is possible (4 cases), as

well as extension to sinuses and dura mater [22]. Ischemic

changes have also been reported and interpreted as

secondary to vascular compression by the neoplastic cells

[39]. In 1 case, computed tomography scan neuroimaging

was considered as normal. In fact, MRI using T1-weighted

images with gadolinium enhancement is required to

visualize the lesions [22,35].

On the basis of the lack of specificity concerning

symptoms and CSF data, PDLG is often misinterpreted as

being of infectious or autoimmune origin [26,32,37].

Among infectious causes, tuberculous meningitis was the

most often suggested diagnosis (10 cases). In 5 cases, the

diagnosis was chronic meningitis. An autoimmune process

was suspected in 3 cases and a cerebral pseudoinflammatory

tumor in 2 cases. It is noteworthy that, in all patients,

iterative investigations for mycobacteria, parasitic agents, or

viruses have always been inconclusive. In approximately

30% of the cases, no clinical diagnosis was established, and

the patients were treated by steroid drugs often combined

with antituberculous treatment. With the exception of the

case described by Rees et al [32], who performed the correct

diagnosis by means of CSF examination and MRI, the

diagnosis of neoplastic proliferation has never been estab-

lished in the other cases, despite radiological findings.

Patients failed to respond to the anti-infectious and anti-

inflammatory treatment, implying the search for malignan-

cy, although CSF examination rarely demonstrated the

presence of neoplastic glial cells [33,38].

A surgical meningeal biopsy, performed in a contrast-

enhancing area, should be considered as an accurate diag-

nostic tool for primary tumoral involvement of the lepto-

meninges. Nevertheless, biopsy material may be normal or

reveal a glioma, as in our case. Surgical biopsy was obtained

in 55% of the reported patients (17 cases): 7 laminectomies

with meningeal biopsy, 3 cerebellar meningeal biopsies, and

7 subtentorial biopsies. Morphological analyses were incon-

clusive in 4 cases and revealed a nonspecific inflammatory

reaction in 1 case. A diagnosis of glioma was concluded in

11 cases without grading in 1 case, low grade (WHO grade II)

in 4 cases, anaplastic (WHO grade III) in 4 cases, and

glioblastoma (WHO grade IV) in 1 case. PDLG, oligoden-

droglioma grade II type, was concluded in 1 case [6].

Based on the definition of PDLG, a postmortem morpho-

logical study of the entire neuraxis should be required to

confirm the diagnosis. An autopsy was obtained from

25 patients (80.6%), and a concordance between the surgical

and necropsic diagnosis was found in 4 cases alone. In the

other cases, either the grade was underestimated, or the

analysis of biopsy material was negative. A correlation

between duration of the disease, treatment, and the final

diagnosis could not be evaluated because of the limited

number of patients studied: 9 low-grade gliomas, 14

anaplastic gliomas, 2 malignant gliomas, and to the multi-

plicity of treatment schedules, which did not permit an

accurate statistical analysis.

Two major anatomoclinical forms of PDLG have been

established: a nodular form and a diffuse form. The nodular

form was initially described by Freeman in 1926, cited by

Bailey and Robitaille [3], then by Dietrich et al [10] as basolitary or focal leptomeningeal gliomatosis, defined by

limited tumor masses in cranial or spinal leptomeninges,Qwhich corresponds to our observation. The diffuse form has

been described by Korein et al [21] as a diffuse extension,

outside the nervous parenchyma, of glial tumor cells over a

wide area of the CNS. More recently, leptomeningeal

gliomatosis was redefined as ba neoplasm, although largely

leptomeningeal, associated with a parenchymal component

small enough to be considered as an ingrowth from the

meningeal lesionQ in the WHO classification of tumors [20].

The nature of proliferating cells is most often of a

astrocytic phenotype (14 cases), confirmed by GFAP

immunohistochemistry. In rare cases, the proliferation

exhibited an apparently exclusive oligodendroglial differen-

tiation (3 cases), or presents with both oligodendroglial and

astrocytic components (3 cases). The tumors range from well

differentiated (9 cases) to malignant lesions classified either

as anaplastic (14 cases) or glioblastoma (2 cases). In the other

reports, the precise morphological classification was

not specified.

PDLG and gliomatous meningitis secondary to gliomas

are clinically difficult to separate, particularly in the absence

of recurrence of a known primary intraparenchymatous lesion

[14]. In a series of 22 patients, Poisson et al [30] demonstrated

that clinical signs were not different from those of PDLG,

associating a combination of meningism, hydrocephalus, and

several neurological signs of cerebral, cranial nerve roots, and

spinal cord dysfunction. Furthermore, CSF data consisted

also of hyperproteinorachia in 94% of the cases and

hypoglycorrhachia in 47%.

Radiological and macroscopical appearance of diffuse or

multinodular forms PDLG can mimic several tumoral or

inflammatory processes related to chronic meningitis and

parasitic and autoimmune processes. Most often, diffuse

tumoral spread in the meninges is due to the extension of

primary intra-axial glial tumors or gliomatosis cerebri.

Nodular forms of PDLG can mimic multiple neuromas, seen

on neuroimaging or macroscopy. However, histological and

immunohistochemical studies usually allow exclusion of

other conditions, such as inflammatory pseudotumors,

infections, leptomeningeal carcinomatosis (particularly

breast and lung primary tumors), lymphomatous meningitis,

meningeal sarcomas, and meningioangiomatosis, as well as


meningeal dissemination of brain or spinal cord primary

tumors, spreading through the subarachnoid spaces, although

clinical presentation may be similar [8,16,32]. Primary

meningeal tumors other than meningioma are extremely rare.

Furthermore, true solitary meningeal gliomas can be difficult

to differentiate from nodular forms of PDLG, but they

generally present as exophytic masses with an intraparen-

chymatous component, extending sometimes massively into

the subarachnoid spaces [25]. Davila et al [9] emphasized that

the multinodular pattern could be related to the disability of

glioma cells to detach from the primary focus, because of a

desmoplastic reaction limiting the dissemination of

tumor cells.

The precise histogenesis of PDLG remains unclear.

Cooper and Kernohan [7] suggested that heterotopic

neuroglial nests could give rise to this pathological

condition. It is worth pointing out that, in the literature,

such anomalies have been observed in only 2 cases [7,27].

The occurrence of glial meningeal heterotopias has been

reported in association with congenital malformations of the

CNS. In a 100-autopsy series, Cooper and Kernohan [7]

found spinal dysraphism in 25% of the patients. The

location of the lesions was bulbar in 53% of the cases,

pontine in 15%, lumbosacral in 20%, and cervical in 10%.

Heterotopic glial nests could derive from overmigrating

cells during antenatal neuronal migration. However, they

have never been reported in association with generalized or

focal cortical dysplasias, suggesting that these abnormalities

could be acquired, resulting from an inappropriate migration

of glioneuronal elements in the subarachnoid spaces,

possibly due to an environmental event. As in the majority

of cases and despite extensive histological sampling, we did

not observe any heterotopic glial cells in the meninges of

our patient.

PDLG poses considerable prognostic and therapeutic

problems. The optimal therapy is not well defined, and

treatment schedules are highly variable among reported

cases. Despite numerous proposed treatment schedules, the

prognosis usually remains poor, with the exception of

2 cases published by Beauchesne et al [4] and by Rogers

et al [33], who reported the longest survival time. The

review of the literature suggests that a longer survival time

is observed when concomitant radiochemotherapy is ap-

plied, and when the latter associates intravenous and

intrathecal route [4]. In some instances, a localized

presentation and a low-grade tumor proliferation could

explain a better prognosis than in diffuse forms [10,17].

Nevertheless, such cases are rarely observed, and in the

majority of reported autopsy cases, the proliferation was

diffuse, although radiochemotherapy had been administered.

This suggests that treatment is generally ineffective, and the

aggressive behavior of PDLG, as that of gliomatosis cerebri,

contrasts with an often low histological grading. These data

imply that there is no correlation between the course of the

disease and histological features, as emphasized by Davila

et al [9]. Furthermore, as in glioblastomas and gliomatosis

cerebri, although exceptional, extraneuraxic metastases have

been described, especially bone metastases. These are

usually restricted to one or a few sites and are more likely

osteolytic, rather than osteoblastic [24,29]. The causes of

ineffective treatment remain unknown, but up to date,

prognostic factors have not been extensively studied, such

as vascular endothelial growth factor, platelet-derived

growth factor, and their receptor expression, as well as

proliferative indices, using either immunohistochemistry or

flow cytometry. Similarly, molecular prognostic factors such

as mutated p53 expression, chromosomes 1p 19q deletions,

or hypermethylation of O-6-methylguanine-DNA-methyl-

transferase [20] have not been reported. Lastly, a routine

screening for somatic and/or constitutive mutations on NF1

and NF2 genes has never been performed.

4. Conclusion

Although rare, PDLG should be evoked whenever

patients present with aseptic subacute or chronic meningitis,

in association with nonspecific neurological signs, and

where a precise etiology cannot be ascertained. CSF

examination is a useful diagnostic tool if atypical PS100/

GFAP–positive cells are present, when MRI reveals diffuse

or multinodular meningeal contrast-enhancing thickening or

masses, without CNS parenchyma involvement. Surgical

biopsy is therefore strongly recommended, but most often,

the final diagnosis is established on the basis of postmortem

findings. No correlation has been demonstrated between

location, macroscopic and histological type, tumor grading,

and survival time. The major question that is raised concerns

the search for accurate prognostic factors, which have not

yet been extensively studied. Finally, there is no consensus

regarding treatment schedules because it is too rare an

entity. Nevertheless, combining cerebrospinal radiation with

intravenously and intrathecally administered chemotherapy

appears to improve patient survival.

Acknowledgments

The authors thank Nicole Rappaport for her medical

assistance, Marie Magdeleine Ruchoux in confirming the

diagnosis, Marianne Paresy for her technical assistance,

Siegfried Le Roy for the iconography, and Richard

Medeiros, medical editor of Rouen University Hospital,

for his valuable advice in editing the manuscript.

References

[1] Baborie A, Dunn EM, Bridges LR, et al. Primary diffuse leptome-

ningeal gliomatosis predominantly affecting the spinal cord: case report

and review of the literature. J Neurol Neurosurg Psychiatry 2001;70:

256 -8.

[2] Bae JY, Choi BO, SunWoo IN, et al. Diffuse cerebrospinal gliomatosis

with extensive leptomeningeal spread. Yonsei Med J 2000;41:517 -21.

[3] Bailey P, Robitaille Y. Primary diffuse leptomeningeal gliomatosis.

Can J Neurol Sci 1985;12:278-81.


[4] Beauchesne P, Pialat J, Duthel R, et al. Aggressive treatment with

complete remission in primary diffuse leptomeningeal gliomatosis—a

case report. J Neurooncol 1998;37:161-77.

[5] Carpentier AF, Sanson M, Kujas M, et al. Primary meningeal

gliomatosis. Rev Neurol (Paris) 1994;150:232-5.

[6] Chen R, Macdonald DR, Ramsay DA. Primary diffuse leptomeningeal

oligodendroglioma. Case report. J Neurosurg 1995;83:724-8.

[7] Cooper IS, Kernohan JW. Heterotopic glial nests in the subarachnoid

space; histopathologic characteristics, mode of origin and relation to

meningeal gliomas. J Neuropathol Exp Neurol 1951;10:16 -29.

[8] Corsten LA, Raja AI, Wagner Jr FC. Primary diffuse leptomeningeal

gliomatosis. Br J Neurosurg 2001;15:62 -6.

[9] Davila G, Duyckaerts C, Lazareth JP, et al. Diffuse primary

leptomeningeal gliomatosis. J Neurooncol 1993;15:45-9.

[10] Dietrich PY, Aapro MS, Rieder A, et al. Primary diffuse leptome-

ningeal gliomatosis (PDLG): a neoplastic cause of chronic meningitis.

J Neurooncol 1993;15:275-83.

[11] Fayet G, Bertout-Lemercier C, Gugenheim M, et al. Leptomeningeal

gliomatosis. 2 Clinicopathological cases. Rev Neurol (Paris) 1994;

150:700-3.

[12] Fortuna A, Celli P, Palma L. Oligodendrogliomas of the spinal cord.

Acta Neurochir (Wien) 1980;52:305-29.

[13] Freeman W. Cortical heterotopia in the pontile meninges. Arch Pathol

1926;2:252-4.

[14] Giordana MT, Bradac GB, Pagni CA, et al. Primary diffuse

leptomeningeal gliomatosis with anaplastic features. Acta Neurochir

(Wien) 1995;132:154-9.

[15] Havlik DM, Becher MW, Nolte KB. Sudden death due to primary

diffuse leptomeningeal gliomatosis. J Forensic Sci 2001;46:392-5.

[16] Ho KL, Hoschner JA, Wolfe DE. Primary leptomeningeal gliomatosis:

symptoms suggestive of meningitis. Arch Neurol 1981;38:662 -6.

[17] Janisch W, Engel U, Leonhardt T. Diffuse primary leptomeningeal

gliomatosis. Zentralbl Pathol 1991;137:523 -30.

[18] Kastenbauer S, Danek A, Klein W, et al. Primary diffuse leptome-

ningeal gliomatosis: unusual MRI with non-enhancing nodular lesions

on the cerebellar surface and spinal leptomeningeal enhancement.

J Neurol Neurosurg Psychiatry 2000;69:385-8.

[19] Kitahara M, Katakura R, Wada T, et al. Diffuse form of primary

leptomeningeal gliomatosis. Case report. J Neurosurg 1985;63:283-7.

[20] Kleihues P, Cavenee WK. World Health Organization classification of

tumours. Pathology and genetics of tumours of the nervous system.

Lyon7 IARC Press; 2000.

[21] Korein J, Feigin I, Shapiro MF. Oligodendrogliomatosis with

intracranial hypertension. Neurology 1957;7:589 -94.

[22] Leproux F, Melanson D, Mercier C, et al. Leptomeningeal gliomatosis:

MR findings. J Comput Assist Tomogr 1993;17:317-20.

[23] Moore M. Diffuse cerebrospinal gliomatosis, masked by neurosyphilis.

J Neuropathol Exp Neurol 1954;13:129-43.

[24] Myers T, Egelhoff J, Myers M. Glioblastoma multiforme presenting as

osteoblastic metastatic disease: case report and review of the literature.

AJNR Am J Neuroradiol 1990;11:802-3.

[25] Ng HK, Poon WS. Diffuse leptomeningeal gliomatosis with oligo-

dendroglioma. Pathology 1999;31:59 -63.

[26] Olivera-Leal IR, Gomez-Viera N, Gonzalez-Espinosa L, et al. Primary

diffuse leptomeningeal gliomatosis. Presentation of one case. Rev

Neurol 1997;25:1419-21.

[27] Park JS, van den Noort S, Kim RC, et al. Primary diffuse

leptomeningeal gliomatosis with signs of increased intracranial

pressure and progressive meningeal enhancement on MRI. J Neuro-

imaging 1996;6:250-4.

[28] Paulino AC, Thomas C, Slomiany DJ, et al. Diffuse malignant

leptomeningeal gliomatosis in a child: a case report and review of the

literature. Am J Clin Oncol 1999;22:243 -66.

[29] Pingi A, Trasimeni G, Di Biasi C, et al. Diffuse leptomeningeal

gliomatosis with osteoblastic metastases and no evidence of intraaxial

lesions. AJNR Am J Neuroradiol 1995;16:1018-20.

[30] Poisson M, Stragliotto G, Davila G, et al. Gliomatous meningitis of

hemispheric tumors. Study of 22 cases in adults. Rev Neurol (Paris)

1995;151:177-89.

[31] Radhakrishnan VV, Saraswathy A, Rout D. Diffuse form of primary

leptomeningeal gliomatosis—a case report. Indian J Pathol Microbiol

1994;37(Suppl):S51-3.

[32] Rees JH, Balakas N, Agathonikou A, et al. Primary diffuse leptome-

ningeal gliomatosis simulating tuberculous meningitis. J Neurol

Neurosurg Psychiatry 2001;70:120-2.

[33] Rogers LR, Estes ML, Rosenbloom SA, et al. Primary leptomeningeal

oligodendroglioma: case report. Neurosurgery 1995;36:166 -8.

[34] Russell DS, Rubinstein LJ. Pathology of tumours of the nervous

system. 5th ed. Baltimore7 Williams and Williams; 1977.

[35] Singh M, Corboy JR, Stears JC, et al. Diffuse leptomeningeal

gliomatosis associated with multifocal CNS infarcts. Surg Neurol

1998;50:356 -62.

[36] Sumi S, Leffman H. Primary intracranial leptomeningeal glioma with

persistent hypoglycorrhachia. J Neurol Neurosurg Psychiatry 1968;

31:190-4.

[37] Tsui EY, Loo KT, Mok CK, et al. Primary multifocal leptomeningeal

gliomatosis. Eur J Radiol 2001;37:5 -7.

[38] Wechsler LR, Gross RA, Miller DC. Meningeal gliomatosis with

bnegativeQ CSF cytology: the value of GFAP staining. Neurology

1984;34:1611 -5.

[39] Whelan HT, Sung JH, Mastri AR. Diffuse leptomeningeal

gliomatosis: report of three cases. Clin Neuropathol 1987;6:

164 -8.

Commentary

For most, the finding of diffuse leptomeningeal involve-

ment would signal the probability of it being secondary to a

previously known malignancy or inflammatory process.

Hence, one would expect to be able to image a primary

problem either elsewhere in the CNS or in another

organ system.

Failure to find a bprimaryQ to biopsy obviously signals theneed for surgical intervention to determine the etiology of the

spinal lesions in question. To find a grade III glioma is

certainly unusual andwould not be suspected pre-operatively.

While the surgical indications are not changed by the

finding of leptomeningeal spread in the absence of a primary

lesion or cause, the etiology is decidedly unusual and of

interest to us all.

Robert G. Selker, MD

Pittsburgh, PA 15243, USA

Primary diffuse multinodular leptomeningeal gliomatosis

Documents