close

Вход

Забыли?

вход по аккаунту

?

Antibodies to cerebellar Purkinje cells in patients with paraneoplastic cerebellar degeneration and ovarian carcinoma.

код для вставкиСкачать
ORIGINAL ARTICLES
Antibodies to Cerebellar Purkmje Cells
in Patients with Paraneoplastic Cerebellar
Degeneration and Ovarian Carcinoma
John E. Greenlee, MD, and H. Robert Brashear, MD
Sera from 2 patients with ovarian carcinoma and paraneoplastic cerebellar degeneration confirmed postmortem were
reacted with frozen sections of human cerebellum and stained using indirect immunofluorescence methods. Both sera
produced bright cytoplasmic staining of Purkinje cells and of neurons within deep cerebellar nuclei. Titration of these
sera to end point revealed staining at final dilutions of 1:640 and 1:2,560, respectively. Neither of these sera reacted
with sections of human cerebrum, basal ganglia, spinal cord, peripheral nerve, lung, liver, kidney, or ovary. Staining of
Purkinje cells was not obtained with sera from 34 normal, healthy controls, 5 patients with oat cell carcinoma of the
lung, 6 patients with inflammatory central nervous system disorders, or 12 of 14 neurologically normal patients with
ovarian carcinoma. Sera from 2 neurologically normal patients with ovarian carcinoma, however, produced staining of
Purkinje cells and deep nuclei similar to that obtained with sera from patients with paraneoplastic cerebellar degeneration. The present study documents the presence of antibodies to Purkinje cells in patients with ovarian carcinoma and
cerebellar degeneration and demonstrates that development of these antibodies may antedate the onset of clinically
evident cerebellar degeneration.
Greenlee JE, Brashear HR: Antibodies to cerebellar Purkinje cells in patients with paraneoplastic cerebellar
degeneration and ovarian carcinoma. Ann Neurol 14:609-613, 1983
Paraneoplastic cerebellar degeneration is an uncommon complication of human malignant disease. The
disorder is characterized clinically by progressive
ataxia, vertigo, dysarthria, and nystagmus [1-41. Pathological examination of affected brains reveals cerebellar
atrophy with destruction of Purkinje cells, variable loss
of granule cells, and, in some cases, loss of basket cells
{4]. Paraneoplastic cerebellar degeneration has been
reported most frequently with carcinoma of the ovary
and lung. Additional cases have been reported with
Hodgkin’s disease, non-Hodgkin’s lymphomas, and
carcinoma of the breast, uterus, stomach, colon, and
larynx [41. The disorder may occur as an isolated complication of malignancy or may be accompanied by
other paraneoplastic syndromes, including encephalomyelitis, peripheral neuropathy, and Eaton-Lambert
syndrome f41. Onset of symptoms may occur prior to
the clinical appearance of the malignancy, during its
course, o r after its apparently successful cure [l-41.
The cause of paraneoplastic cerebellar degeneration
is unknown. An infectious or inflammatory basis for
the disorder is suggested by the presence of elevated
cerebrospinal fluid immunoglobulin levels in many patients and by the frequent presence in pathological material of lymphocytic infiltrates within meninges, cere-
bellum, and, at times, other parts of the nervous system
[41.Isolation of virus or other infectious agents has
not been reported, however. The only clue to the
pathogenesis of paraneoplastic cerebellar degeneration
has been provided by studies of 2 individual patients.
Trotter and colleagues [7] in 1776 reported the presence of anti-Purkinje cell antibodies in serum from a
patient with Hodgkin’s disease and presumed paraneoplastic cerebellar degeneration. Pathological confirmation of the diagnosis of paraneoplastic cerebellar
degeneration was not obtained in this case, however,
and these observations have not been confirmed in
other patients. More recently Steven and co-workers
(61 studied the cerebrospinal fluid of a patient with
ovarian carcinoma and cerebellar degeneration confirmed postmortem. These investigators, using a solidphase radioimmunoassay, detected higher levels of
antibodies to sonicated neonatal mouse cerebellar
cells in the patient’s cerebrospinal fluid than in the cerebrospinal fluid of 20 patients with a variety of other
neurological diseases. Their observations, although
suggestive of the presence of antibodies to cerebellar
Purkinje cells, were not definitive: there was considerable nonspecific background activity, the reactivity of
the cerebrospinal fluid was not specific for cerebellar
From the Department of Neurology, University of Virginia Medical
Center, Charlottesville. VA 22908.
Received Apr 4 , 1983. Accepted for publication May 15, 1983.
Address reprint requests to Dr Greenlee, Box 394, University of
Virginia Medical Center, Charlorresville, VA 22908.
609
cells, and the levels of antibody were lower than those
previously measured in sera of rabbits immunized with
cerebellar preparations. In the present report we document the presence of circulating antibodies to cerebellar Purkinje cells in 2 patients with ovarian carcinoma
and paraneoplastic cerebellar degeneration demonstrated postmortem and also in 2 of 14 neurologically
normal patients with biopsy-proved ovarian carcinoma.
Case Reports
Patient 1
A 50-year-old woman was in good health until October
1973, when she developed severe vertigo, nausea, and vomiting. Evaluation at another hospital showed bilateral horizontal and rotatory nystagmus, dysarthria, cerebellar ataxia, and
photophobia. Laboratory findings were normal except for a
cerebrospinal fluid pleocytosis of 140 cells per cubic millimeter (97%. polymorphonuclear leukocytes), with normal glucose and protein levels. The patient’s ataxia and dysarthria
worsened steadily, and she was transferred to the University
of Virginia Medical Center on November 15, 1973.
O n admission, the patient showed intact mental function
but was unable to raise her head, sit, or stand. She exhibited
bilateral horizontal nystagmus on lateral gaze, vertical nystagmus, severe dysarthria, generalized hypotonia, and truncal
and appendicular ataxia. Sensory examination showed no abnormalities. Reflexes were depressed, with absent triceps and
brachioradialis reflexes. Babinski’s sign and ocher pathological reflexes were absent. Results of hematological studies,
blood chemistries, and collagen vascular screening were normal except for demonstration of a positive rheumatoid factor.
Cerebrospinal fluid contained 16 lymphocytes per cubic millimeter, with normal glucose and protein levels. An electroencephalogram showed bursts of generalized theta and delta
activity. A pneumoencephalogram demonstrated cerebellar
atrophy. Bacterial, fungal, and viral cultures were negative.
Efforts to find an associated neoplasm were initially unsuccessful. Later in the patient’s hospitalization, an enlarged,
cystic right ovary was detected, but deterioration in the patient’s condition precluded exploratory laparotomy. The patient was transferred to a hospital closer to her home and died
there on May 24, 1974.
The patient was found postmortem to have adenocarcinoma of the right ovary (Fig 1) with extension to the
peritoneum and metastatic involvement of spleen. The cerebellum was diffusely atrophic, with extensive loss of Purkinje
cells (Fig 21, reactive gliosis of Bergmann cells, and thinning
of the granule cell layer. The remainder of the patient’s brain
and spinal cord were histologically normal and without evidence of metastatic disease or hypoxic injury.
Patient 2
A 58-year-old right-handed woman was in good health until
July 1975, when she developed dizziness, nausea, and vomiting while playing golf. O n the following day she noticed
diplopia, which was most severe on right lateral gaze. Evaluation at another institution revealed nystagmus but no other
neurological abnormalities; electroencephalography and
radionuclide brain scan findings were normal. Over the next
610 Annals of Neurology
Vol 14
Fig 1 . Section of oziavy from patient 1 showing poorly d‘ierentiated adenocarcinoma.(H&E; x 200 before 20% reduction.)
~
Fig 2. Section of cerebellumfrom patient I . Purkinje cells are absent (arrows), and there is thinning ofthe granule celL layer.
(H6.E; x 70 before 20% reduction.)
several weeks, the patient developed slurred speech and
progressive unsteadiness of gait. She tended to veer to the
left when she walked. She was admitted to Northwestern
Memorial Hospital on August 26, 1975.
O n admission, the patient was alert and showed normal
intellectual function but demonstrated severe truncal and appendicular ataxia, dysarthria, and scanning speech. She had
truncal and head titubation when sitting and was unable to
walk unsupported. There were no cranial nerve abnormalities
except for coarse horizontal nystagmus on lateral gdzc bi1att.rally and limitation of left lateral gaze. Strength was normd,
but the patient was hypotonic, with depressed reflexes. Chaddock’s sign was present bilaterally. Initial hematological
findings were normal, as were results of thyroid function tests
and collagen vascular screening. Cerebrospinal fluid at the
time of admission contained 28 mononuclear cells per cubic
millimeter, normal glucose level, and 3 1 mgidl protein with
4595 gamma globulin. Two weeks later cerebrospinal fluid
No 6 December 1983
Fig 3 . Section of cerebellum from patient 2, showing losf of Purkinje a n d granule cells similar to that observed in patient l .
( H 6 E ; x 24 before 20% reduction.)
contained 8 mononuclear cells with a protein level of 64 mgi
dl, of which 17% was gamma globulin. Results of electroencephalography, computed tomographic scanning, and
four-vessel cerebral angiography were normal. Pneumoencephalography demonstrated a prominent superior
cerebellar cistern. Cultures for bacteria, mycobacteria, fungi,
and viruses were negative. Extensive search for an occult
malignancy was unrevealing. The patient was treated with
adrenocorticotropic hormone without improvement and was
discharged on October 10, 1975, unable to walk, transfer
from a wheelchair, or feed herself.
The patient's neurological condition remained stable on
follow-up examinations in 1976 and early 1977. In June 1977
an abdominal mass was detected by the patient's daughter.
Laparotomy revealed a poorly differentiated ovarian adenocarcinoma with spread throughout the abdominal cavity. The
patient was treated with cyclophosphamide.
The patient was readmitted to Northwestern Memorial
Hospital in October 1977 because of severe dysphagia. Examination showed the patient to be cachectic with preserved
intellect. H e r cerebellar deficit was unchanged, but she had
weakness of her right iliopsoas, hip abductors, and hamstrings. Sensation was diminished over the right second and
third lumbar dermatomes. The patient was found to have an
esophageal stricture secondary to infiltration by neoplasm.
She underwent local radiation therapy and esophageal dilatation but became increasingly cachectic, developed ascites and
bilateral pleural effusions, and died on November 17, 1977.
Postmortem examination revealed moderately well differentiated ovarian adenocarcinoma metastatic to peritoneum,
abdominal and pelvic viscera, right lumbar spinal roots, and
thyroid. The cerebellum was atrophic, with extensive loss of
Purkinje cells (Fig 3) and moderate reduction of the granule
cell layer. The central nervous system was without evidence
of metastatic involvement or hypoxic injury. Numerous microabscesses were present in the kidneys, lungs, and brain.
Materials and Methods
Sera from patients 1 and 2 had been frozen at - 70°C since
1973 and 1975, respectively. Control sera included 34 sera
from normal, healthy adults, of whom 24 were women of
childbearing age; 18 sera from 14 neurologically normal patients with ovarian carcinoma; sera from 5 patients with oat
cell carcinoma of the lung; and sera from 6 patients with a
variety of inflammatory central nervous system disorders, including multiple sclerosis, recurrent meningitis, Epstein-Barr
virus-associated encephalitis, and Toxoplasma gondii encephalitis. Tissues for immunofluorescence staining were obtained postmortem from patients without neoplastic disease.
All sera and tissue samples were obtained with the approval
of the Human Investigation Committee at the University of
Virginia.
Serial twofold dilutions of each serum were prepared in
phosphate-buffered saline (PBS), beginning at a dilution of
1 : 10. Portions of cerebellum, cerebrum, basal ganglia, spinal
cord, peripheral nerve, lung, liver, kidney, and ovary were
frozen in liquid nitrogen-chilled isopentane, sectioned at
6 IJ., and fixed in acetone at -20°C. Sections were reacted
with serial dilutions of each serum at 37°C for 60 minutes,
washed in PBS, overlaid with fluorescein-conjugated rabbit
antihuman IgG globulin (Miles Laboratories) for 60 minutes
at 37"C, washed in PBS and distilled water, air dried,
mounted using a 1 :20 dilution of glycerol in PBS, and examined using a Leitz Ortholux I1 microscope with a 495 nm
excitation filter and a K530 barrier filter. Sections were
coded and read by two different observers before identification of patient source and dilution.
Results
Sera from both patient 1 and patient 2 produced bright
immunofluorescent labeling of cerebellar Purkin je
cells and cells within deep cerebellar nuclei. Specific
staining of these cells could be detected through dilutions of 1: 640 and 1:2,560, respectively. Staining involved cell cytoplasm diffusely and appeared to spare
cell nuclei (Figs 4 and 5). Neither serum reacted with
cells within cerebrum, basal ganglia, spinal cord, liver,
kidney, or ovary, or with peripheral nerve. Labeling of
Purkinje cells or other cells within cerebellum did not
occur with sera from normal, healthy controls, patients
with carcinoma of the lung, or patients with any of the
other neurological disorders listed previously. Sera
from 12 of 14 patients with ovarian carcinoma also
produced no labeling. Sera from two neurologically
normal patients with ovarian carcinoma, however, produced immunofluorescent labeling of cerebellar Purkinje cells and deep nuclei through dilutions of l :640
and 1 : 10,240, respectively. The immunofluorescent
staining produced by these sera was similar to that obtained with sera from patients 1 and 2. These sera did
not stain cells in other areas of the central nervous
system or in extraneural tissues.
Discussion
The present study demonstrates antibodies to cerebellar Purkinje cells and cells within deep cerebellar nuclei
in the sera of 2 patients with paraneoplastic cerebellar
degeneration associated with ovarian carcinoma and in
Greenlee and Brashear: Purkinje Cell Antibodies
611
Fig 4. Frozen section of human cerebellum reucted with serum
from patient 1 and stained by indirect immunoJ7uorescence
methods. Purkinje cells are brigbtly labeled. Fluoresience is predominantly rytophsmic and .\pares cell nuclei. x 3.50 before
20% reduction.)
Fig 5 . Section of humun cerebellum reacted with serum from patient 2. Purkinje r d s show bright, cytoplasmic labeling similar
t o that obtuined u'ith serum from potient I . ( x 350 brfore 20%
reduction.I
2 of 14 neurologically normal patients with ovarian
carcinoma. These sera did not contain antibodies to
other areas of the central nervous system, peripheral
nerve, or extraneural tissues. Antibodies to Purkinje
cells were not found in sera of normal controls or in
sera of patients with oat cell carcinoma of the lung or a
variety of inflammatory central nervous system disorders. The immunofluorescent staining produced by
these sera is similar to that obtained by Trotter (71,
using serum from a patient with Hodgkin's disease and
cerebellar ataxia, and does not appear to be directed
612 Annals of Neurology
Vol 14 No 6
against an antigen common to other cell populations in
neural or extraneural tissues. The finding that
neurologically intact patients with ovarian carcinmoma
may have similar anti-Purkinje cell antibodies makes it
unlikely that these antibodies are produced in response
to antigens released during destruction of Purkinje
cells. These observations lend support to the hypothesis that paraneoplastic cerebellar degeneration is
an immune-mediated disorder.
The presence of circulating antibodies specific
against cerebellar Purkinje cells in these patients raises
questions about the nature and origin of the antigen
against which the antibodies are directed, the role of
humoral immunity in the pathogenesis of paraneoplastic cerebellar degeneration, and the relationship of
this antibody response to the patients' underlying neoplasms. The antigen stained within Purkinje cells appears to be a cytoplasmic protein, but the fluorescent
antibody methods used in the present study did not
allow its more precise identification. Additional studies
employing immunoenzymatic staining, electron microscopy, and immunoblot methods are in progress,
with the aim of characterizing this cerebellar antigen
further. It is unlikely, however, that the antigen that
initiates this antibody response is derived from Purkinje cells, because an identical antibody response can
be detected in the absence of clinically evident cerebellar degeneration. It is more probable that the observed
antibody response is directed against some other intracellular antigen. The failure of these sera to stain a
variety of neural and extraneural tissues suggests that
the source of the antigen is the patients' tumors. Unfbrtunately, frozen tissue or cultured cells were not available from any of these patients. Experiments using immunoperoxidase methods are in progress to determine
whether these patients' sera react with formalin-fixed,
paraffin-embedded sections of their tumors and
whether this reaction can be abolished by absorption of
the sera with cerebellar tissue.
The role of humoral immunity in the pathogenesis of
paraneoplastic cerebellar degeneration is uncertain.
The high levels of circulating anti-Purkin je cell antibodies observed in 2 neurologically normal patients
with ovarian carcinoma indicate that circulating antibody alone may not be sufficient to produce the syndrome; some other factor may be required, such a!;
disruption of the blood-brain barrier, activation of cellmediated immunity, or entry of immunocompeteni:
cells into cerebrospinal fluid and brain. The detection
by Steven and colleagues (61 of anti-Purkinje cell antibodies in the cerebrospinal fluid of a patient with ovarian carcinoma and cerebellar degeneration, although
inconclusive, is of great potential relevance to the
pathogenesis of the disorder. A unique pathoanatomic
study by Schmid and Rtede (31, however, suggests that
December 1983
clinically inapparent destruction of Purkinje cells may
be a characteristic occurrence in patients with malignant disease and particularly in patients with ovarian
carcinoma. These investigators carried out careful morphometric and automatic measuring studies of cerebellar tissue from patients with a variety of neoplasms.
They found that minor loss of Purkinje cells and other
cerebellar cortical neurons was common in patients
with carcinoma of the lung, breast, stomach, colon, kidney, and prostate. In patients with carcinoma of the
ovary, however, the severity of the loss of cerebellar
Purkinje cells and granule cells was similar to that encountered in pathologically confirmed cerebellar degeneration syndromes.
A third question posed by this study concerns the
relationship of the antibodies observed in these patients to the underlying neoplasms. Anti-Purkinje cell
antibodies were detected in 4 patients with ovarian carcinoma and were present in 1 patient 20 months before
her tumor was diagnosed. Three of these 4 patients had
adenocarcinoma; the fourth patient had a poorly differentiated tumor believed to be a mixed mesodermal
sarcoma. These tumor types were not different from
those found in patients without detectable levels of
antibodies. The 2 patients still living, however, in contrast to the other tested patients with ovarian carcinoma, have disease poorly responsive to chemotherapy. We are currently attempting to determine
whether more widespread anti-Purkinje cell antibodies can be detected in patients with ovarian carci-
noma by means of a more sensitive assay system, and
we are studying additional patients with ovarian carcinoma to determine whether high titers of antibody to
Purkinje cells reflect in any way the biological behavior
of the underlying tumor.
The authors thank D r Carolyn Brunner for providing sera from
patient 1, D r Howard Lipton (Northwestern University School of
Medicine) for providing sera and pathological material from patient
2, and Dr Peyton Taylor for his assistance in obtaining sera from
patients with ovarian carcinoma. The technical assistance of Mr Wendell Dodd and Mrs Paula Keeney is gratefully acknowledged.
References
1. Brain WR, Daniel PM, Greenfield JG: Subacute cortical cerebelNeurol
lar degeneration and its relation to carcinoma.
Neurosurg Psychiatry 14:59-75, 1951
2. Brain WR, Wilkinson M: Subacute cerebellar degeneration associated with neoplasms. Brain 88:465-478, 1965
3. Greenfield JG: Subacute spinocerebellar degeneration occurring
in elderly patients. Brain 57:161-176, 1934
4. Henson RA, Urich H: Cortical cerebellar degeneration. In Cancer and the Nervous System: The Neurological Manifestations of
Systemic Malignant Disease. Oxford, Blackwell, 1982, pp 346357
5. Schmid AH, Riede UN: A morphometric study of the cerebellar
cortex from patients with carcinoma: a contribution on quantitative aspects in carcinotoxic cerebellar atrophy. Acta Neuropathol
(Berl) 28:343-352, 1974
6. Steven MM, Carnegie PR, Mackay IP, et al: Cerebellar degeneration with ovarian carcinoma. Postgrad Med J 58:47-5 1, 1982
7 . Trotter JL, Hendin BA, Osterland K Cerebellar degeneration
with Hodgkin disease: an immunological study. Arch Neurol
3 3~660-661 , 1976
Greenlee and Brashear: Purkinje Cell Antibodies
613
Документ
Категория
Без категории
Просмотров
0
Размер файла
1 210 Кб
Теги
carcinoma, patients, antibodies, cerebellar, ovarian, degeneration, purkinje, paraneoplastic, cells
1/--страниц
Пожаловаться на содержимое документа