вход по аккаунту


Primary central nervous system lymphoma can be histologically diagnosed after previous corticosteroid use A pilot study to determine whether corticosteroids prevent the diagnosis of primary central nervous system lymphoma.

код для вставкиСкачать
Primary Central Nervous
System Lymphoma Can Be
Histologically Diagnosed
after Previous Corticosteroid
Use: A Pilot Study to
Determine Whether
Corticosteroids Prevent the
Diagnosis of Primary
Central Nervous System
Alyx B. Porter, MD,1 Caterina Giannini, MD,2
Timothy Kaufmann, MD,3 Claudia F. Lucchinetti, MD,1
Wenting Wu, PhD,4 Paul A. Decker, MSc,4
John L. D. Atkinson, MD,5 Brian Patrick O’Neill, MD1
The objective is to determine whether corticosteroid administration before biopsy prevents histopathological diagnosis of
primary central nervous system lymphoma (PCNSL). A retrospective review was performed of immunocompetent PCNSL patients from 1985 to 2005. A total of 109 patients
was identified. Sixty-eight (63.6%) patients received corticosteroids before diagnosis. Thirteen patients (of 109; 12%)
had undergone repeat brain biopsy to confirm PCNSL.
These included 8 (of 68) patients who had received corticosteroids (12%), and 5 (of 39) who had not (13%) ( p ⫽ 1.0).
The majority of PCNSL patients who received corticosteroids
before diagnosis did not experience significant radiographic
change or require second biopsy for diagnosis.
Ann Neurol 2008;63:662– 667
Primary central nervous system lymphoma (PCNSL) is
a rare, malignant, non–Hodgkin’s lymphoma confined
to the brain, eyes, spinal cord, and leptomeninges, with
increasing incidence among both immunocompetent
and immunocompromised patients.1
Current clinical practice is to refrain from corticosteroid administration in suspected PCNSL cases for
From the Departments of 1Neurology, 2Pathology, 3Radiology,
Biostatistics, and 5Neurosurgery, Mayo Clinic Rochester, Rochester
Received Sep 5, 2007, and in revised form Jan 14, 2008. Accepted
for publication Jan 25, 2008.
Published online in Wiley InterScience (
DOI: 10.1002/ana.21366
Address correspondence to Dr Porter, 200 First Street SW, Rochester, MN 55901. E-mail: [email protected]
fear of rendering the subsequent biopsy nondiagnostic.2 The goal of this retrospective study of a series of
pathologically documented PCNSL cases was to determine whether the administration of corticosteroids
negatively influenced the pathological diagnosis of PCNSL and delayed patient care.
Subjects and Methods
After Mayo Foundation Institutional Review Board (IRB
#1914-05) approval, we searched the Mayo Lymphoma Registry and the Neuro-Oncology Program database for all PCNSL patients pathologically diagnosed at Mayo Clinic Rochester from January 1985 through December 2005. Patients
who did not consent to research and those who did not meet
criteria for parenchymal PCNSL were human immunodeficiency virus–positive or immunosuppressed after transplant
were excluded from further study (Table 1). In those judged
eligible, we reviewed the clinical and pathological data to
confirm PCNSL and each patient’s clinical course from presentation to diagnosis. Immunohistochemistry for CD3 and
CD20 in the diffuse large B-cell (DLBC) cases and for CD2,
CD3, and CD5 in T-cell cases was always performed. In the
T-cell cases, receptor rearrangement analysis was also performed if the diagnosis was unclear.
Statistical Methods
Age was compared between male and female patients using
the two-sample t test. Fisher’s exact test was used to compare
the rate of repeat biopsy in those patients who did versus did
not receive treatment with corticosteroids. Cumulative survival probabilities were estimated using the Kaplan–Meier
method with time zero being the date of diagnosis.3 The
log-rank test was used to compare groups of patients. p values ⱕ 0.05 were always considered statistically significant.
Of an initial list of 233 patients, 109 eligible immunocompetent PCNSL patients were identified (data
summarized in Fig 1) including 66 men and 43
women (median age, 63 and 70 years, respectively).
Cerebrospinal fluid cytology was tested in 38 patients and was positive in 5 (13.2%). Three of the five
patients with positive cerebrospinal fluid cytology had
been treated with steroids before biopsy. None of the
five patients with positive cerebrospinal fluid cytology
required a biopsy for diagnosis.
Clinical data regarding corticosteroid administration
before biopsy were available in 107 of 109 patients (in 2
patients, this information was unable to be obtained).
Of these 107 patients, 68 (63.6%) received corticosteroids and 39 (36.5%) did not. Steroid dose and duration were not standardized. Data were present to calculate total prednisone equivalent dose in 61 of the 68
patients. The median dose administered in this cohort
was 362.5mg (range, 25– 6,325mg). Number of days off
of steroids before biopsy was measured in 58 of 61 patients where specific data were available regarding steroid
dose and administration. Of these 58 patients, 42
© 2008 American Neurological Association
Published by Wiley-Liss, Inc., through Wiley Subscription Services
Table 1. Patients Excluded from Retrospective Chart Review of Cases of Primary Central Nervous System
Lymphoma from 1985–2005
Exclusion Criteria
Did not consent to research
Presented as skull-based tumors
Presentation of lymphoma outside of brain and
spinal cord
24 (1 dural, 13 epidural, 1 meningeal, 9 peripheral nerve)
Systemic lymphoma with CNS involvement
Intravascular lymphoma
Active acute lymphocytic leukemia
Active chronic lymphocytic leukemia
Active multiple myeloma
Anaplastic T-cell lymphoma
Younger than 18 years
Testicular lymphoma
Metastatic squamous cell with radiation
Thought to be lymphoma: biopsy repeated and was
negative for lymphoma
3 (1 HSV encephalitis, 1 reactive process, 1 angiogenic
myeloid metaplasia)
Diagnosed at autopsy
Transplant patients diagnosed with PCNSL
9 (1 lung, 1 kidney/pancreas, 1 liver)
Immunosuppressed because of HIV/AIDS
Worked up in outside institution
Diagnosed before 1985
Total number of cases excluded
CNS ⫽ central nervous system; MALT ⫽ mucosa-associated lymphoid tumors; HSV ⫽ herpes simplex virus; PCNSL ⫽ primary
central nervous system lymphoma; HIV/AIDS ⫽ human immunodeficiency virus/acquired immune deficiency syndrome.
Fig 1. Flow diagram summarizing all patients studied within the cohort.
Porter et al: PCNSL Diagnosed After Steroids
Table 2. Summary of 13 Patients Who Required Repeat Biopsies for Final Diagnosis
Date of
Result of First Biopsy
11/14/85: results consistent
with viral infection
Dexamethasone 4mg PO every
4 hours from 2/25/86
through 3/7/86. The
original biopsy was
performed at an outside
institution, and the data
were not available regarding
corticosteroid administration
before the procedure.
CT 3/4/86: enhancing lesion
with mass effect in the
right superior temporal
lobe and left thalamus
Large B-cell
12/16/85: mild diffuse
gliosis of deep white
matter with occasional
lymphocytic perivascular
Dexamethasone 2mg PO four
times a day for 1 week
before nondiagnostic biopsy
and then was not taking
steroids for 3 months
before second biopsy
MRI 12/16/85:
multiloculated enhancing
lesion in the left
temporoparietal areas with
mass effect CT 8/7/86:
enhancing mass deep
right parietal, decreased
attenuation in the left
hemispheric lesion
Large B-cell
3/10/87: perivascular
lymphocytic infiltrate
with mild atypia
suspicious for lymphoma
Remained off of all
CT 3/4/87: enhancing
periventricular confluent
areas of abnormality
CT 3/24/87: unchanged
from prior
Large B-cell
2/10/88: inflammatory
lesion, no evidence of
Remained off of all
MRI 7/30/92: enhancing
mass in the right
temporoparietal lobe
CT 9/10/92: lesion
unchanged from prior
11/15/93 nondiagnostic,
inflammatory vs
Dexamethasone 4mg PO every
6 hours 1/9/94-1/13/94.
The original biopsy was
performed at an outside
institution (reviewed by
MCR pathologists), and the
data were not available
regarding corticosteroid
administration before the
MRI 1/11/94: multiple
enhancing lesions within
the cerebral hemispheres
largest in the left parietal/
occipital lobe
Large B-cell
5/2/96: changes suggestive
of demyelinating disease
No corticosteroids were given
before first biopsy. Patient
was given solumedrol 1gm
IV daily for 5 days between
biopsies; however, the
patient was not taking
corticosteroids at the time
of the second biopsy
MRI 5/2/96: confluent areas
of patchy T2 signal
involving the white
matter of both
hemispheres and the
corpus callosum
MRI 7/26/96: multiple
areas of increased T2
signal with the largest
being in the medial right
temporal lobe
Large B-cell
1996: nonspecific
lymphocytic process
January 1999: changes
suspicious for lymphoma
Dexamethasone/mannitol (dose
unknown) at outside
institution and 54cGy
MRI 3/9/99: 2.5 cm
enhancing right frontal
lesion with mass effect
3/10/99: unchanged
Large B-cell
2/21/02: perivascular
4/22/02: gliosis
Dexamethasone 3mg IV four
times daily
8/5/02-8/24/02: off steroids
4 days before biopsy
MRI 8/2/02: bihemispheric
patchy T2 signal
MRI 8/25/02: progression
of lesions with increased
contrast enhancement
Large B-cell
2/24/03: mild gliosis
2/28/03: negative for
Prior 1,250cGy of radiation
and steroids given for
presumed metastatic lesion
in February 2003 (outside
hospital: dose and duration
of steroids unknown)
MRI 4/11/06: multiple
enhancing lesions in the
corpus callosum, right
temporal/parietal lobe,
and left frontal lobe
MRI 4/18/03: unchanged
Large B-cell
2/27/03: inflammatory
cells, negative for
Remained off of all
MRI 8/28/03: left cerebellar
enhancing mass
MRI 9/18/03: unchanged
Large B-cell
7/18/05: resembles
demyelination, results
not typical, cannot
exclude lymphoma
Dexamethasone 4mg PO every
6 hours
7/15: biopsy
MRI 7/18/05: enhancing
lesion left thalamus/
internal capsule extending
into the corpus callosum
MRI 1/25/06: 1.1cm
enhancing lesion right
posterior parietal lobe
MRI 2/1/06: unchanged
Large B-cell
Annals of Neurology
Vol 63
No 5
May 2008
Result of Final
Table 2. (Continued)
Date of
Result of First Biopsy
Result of Final
12/6/04: inflammation/
Remained off of all
MRI 3/15/05: large right
frontal enhancing lesion
MRI 3/17/05: unchanged
Large B-cell
4/25/05: inflammation, no
evidence of lymphoma
Remained off of all
MRI 3/31/05: 3cm
enhancing mass right
MRI 4/25/05: mass
slightly smaller than prior,
still enhancing
MRI 5/12/05: mass 4 cm
and enhancing
Large B-cell
PO ⫽ orally; CT ⫽ computed tomography; NHL ⫽ non-–Hodgkin’s lymphoma; MRI ⫽ magnetic resonance imaging; MCR ⫽ Mayo
Clinic Rochester; IV ⫽ intravenously.
(72%), were taking steroids at the time of biopsy with
an average dose of 675.0mg. Steroids were discontinued
1 to 10 days before biopsy in 14 (24%) patients with an
average dose of 1,277.1mg. Of the 2 patients remaining,
one had been off of steroids 28 days and the other 180
days before biopsy, and received 6,250 and 337.5mg total prednisone equivalents, respectively.
All 109 patients had contrast-enhanced preoperative
cranial computed tomography (n ⫽ 39) and/or magnetic resonance imaging (n ⫽ 70). Imaging modality
did not appear to have significance in leading toward
diagnosis. The only trend appreciated would be increased use of magnetic resonance imaging after 1989.
Seventy-four patients had solitary appearing tumors,
and 35 had multifocal disease.
Serial preoperative imaging was performed in 87 patients. There was no significant radiographic change in
81 of 87 (93%) patients. Of these 81 patients, 49
(60%) patients received corticosteroids. Total prednisone equivalent dose and days off of steroids before
biopsy was available for 43 of these 49 patients. Thirtyone of 43 (72%) patients were taking steroids at time
of repeat imaging after steroid administration of
651.5mg total prednisone equivalents. Steroids were
discontinued between 1 and 10 days in 11 patients
with the average total prednisone equivalent dose of
993.6mg. The remaining 2 patients had been off of
steroids 28 (6,250mg) and 180 (337.5mg) days, respectively, before repeat magnetic resonance imaging.
Of the original 87 patients who had serial preoperative imaging, 6 (6%) patients who received corticosteroids had a substantial difference in their serial preoperative imaging. Two patients had tumor regression
but retained contrast enhancement on corticosteroids,
whereas three patients had tumor progression on steroids. One patient had a “vanishing tumor” that was
initially present as a contrast-enhancing lesion; when
imaging was performed for stereotactic planning, the
lesion had resolved and the biopsy was aborted. This
patient had been treated with dexamethasone 4mg
three times daily. Nearly 15 months later, the lesion
recurred, and this patient was diagnosed with PCNSL.
Thirteen patients (of 109; 12%) had undergone two
biopsies, the first one being not diagnostic of lymphoma (Table 2). Among these 13 patients, 8 (of 68)
patients had received corticosteroids (12%) and 5 (of
39) patients had not (13%). There was no statistical
significance between administration of corticosteroids
and repeat biopsy for diagnosis of PCNSL ( p ⫽ 1.0).
The median time between biopsies was 113 (range,
14 –1,674) days. The median age of patients who had
undergone two biopsies was 59.4 (range, 26 –77) years.
There were 7 men and 6 women, and the prevailing
final diagnosis was DLBC non–Hodgkin’s lymphoma
in 12 patients with only one diagnosis of small lymphocytic lymphoma.
The initial pathological diagnoses from the first biopsy varied, ranging from inflammatory changes to gliosis. After the first biopsy, patients were treated with a
variety of therapies (eight patients) including no treatment at all (five patients). The only predictor of diagnostic biopsy in this group was the presence of an enhancing lesion for target. All but one patient had an
enhancing lesion on neuroimaging before repeat biopsy. Age, sex, tumor location, and solitary versus multifocal disease were all considered but had no predictive
value on repeat biopsy rate.
We began our study with pathologically confirmed
PCNSL defined with strict criteria. Our cohort reflected the typical PCNSL demographic, with onset in
the sixth decade of life, male predominance, and a prevailing DLBC pathology. Approximately 10% of PCNSL cases are of non-DLBC histology, and our cohort
was representative of the rarity of T-cell4 and lowgrade forms.5 The majority of the patients (63.6%) received corticosteroids in the interval leading up to their
diagnosis, most commonly, at the referring institution.
We did not find a statistically significant difference in
nondiagnostic biopsy rates between those treated with
Porter et al: PCNSL Diagnosed After Steroids
corticosteroids before biopsy. The observed difference
in repeat biopsy rate (2%) between those patients
treated with corticosteroids and those not treated was
similarly small and statistically insignificant.
The relative impact of steroids on the diagnosis of
PCNSL is somewhat controversial. In contrast with the
commonly held viewpoint that steroids may lead to
rapid PCNSL regression, and thereby impede the diagnostic yield of biopsy, our report, as well several others,7,8 challenge this premise. Corticosteroids can be
cytotoxic to B lymphocytes6 and have been used as an
adjunct treatment in PCNSL leading to lesion regression. Figure 2 illustrates an example of a patient in our
cohort who was given corticosteroids before biopsy.
Within the field of view, we see the cytotoxic changes
attributable to corticosteroids together with the diagnostic features for PCNSL.
We found the majority of patients who received steroids before diagnosis were also taking steroids at the
Fig 2. The figure illustrates changes attributable to steroid
treatment. This patient is a 55 year old gentleman who presented with a subacute left hemiparesis. He was found to have
an enhancing lesion in the right internal capsule tracking
down into the midbrain and the right middle cerebellar peduncle prior to treatment. After treatment he had increased
lesion burden with multiple enhancing masses seen. He was
treated with IV methylprednisolone 1 gm daily ⫻ 5 days and
then received 4 cycles of plasma exchange. He was subsequently
biopsied 28 days later and found to have diffuse large B cell
lymphoma. He succumbed to his illness on 3 days post biopsy.
Panel A and B illustrate separate field of a biopsy diagnostic
of DLBC lymphoma. Panel A demonstrates the presence of
residual large atypical cells with large nuclei and prominent
nucleoli diagnostic of lymphoma (black arrow heads) scattered
in a markedly reactive background. In panel B, no lymphoma
cells remain. A prominent infiltrate composed of single and
clustered macrophages with ample pale “granular” cytoplasm
(black arrows) is present, associated with a number of small,
dark apoptotic bodies (asterisk), likely the remaining of lymphoma cells. A reactive astrocyte with abundant glossy cytoplasm is visible at the center of the field (white arrow).
Annals of Neurology
Vol 63
No 5
May 2008
time of biopsy (72%). This observation is interesting,
given that serial preoperative neuroimaging was not
significantly changed (72%) in this subgroup as well.
This subset of patients retained the original contrastenhancing lesions that raised the initial possibility of
PCNSL within the differential diagnosis and went on
to have diagnostic biopsies.
We identified only 3 DLBC PCNSL patients with
lesion regression, one of which was considered a “vanishing tumor.” In their retrospective study evaluating
the relation of PCNSL and “vanishing tumors,” Bromberg and colleagues7 identified 12 patients, 5 of whom
subsequently received a diagnosis of PCNSL; the remaining had a myriad of diagnoses including demyelinating disease, infarction, sarcoidosis, and renal cell
carcinoma. Thus, steroid-induced lesion regression is
not only relatively uncommon, its occurrence is not
necessarily pathognomonic for PCNSL.
Similar to our results indicating that steroids may
not significantly reduce diagnostic yield of biopsies,
Haldorsen and colleagues,8 in their review of 58 nonAIDS PCNSL cases, found the first biopsy to be nondiagnostic in 32% of patients younger than 65 years
and in 4% of patients older than 65 years, with the
median time from first biopsy to final diagnosis in
their series 100 (range, 9 –290) days. Approximately
22% of these patients were treated with corticosteroids
before their biopsy ( p ⫽ 0.18). Haldorsen and colleagues8 conclude that “the unfavourable effect of steroids on the yield of biopsy. . .was not statistically significant in our small series.”
Several caveats are in order for our results. The experience at our tertiary referral center may be skewed.
We recognize that there is a large population of patients who may have had a nondiagnostic biopsy and
did not return for further workup. The similar nondiagnostic rate in steroid-treated and steroid-naive patients may reflect a stage in PCNSL tumorigenesis. In
posttransplant lymphoproliferative disorders of the central nervous system, there appears to be a transition
stage during which the patient may be saved if immunocompetence is improved.9
In conclusion, we found that the administration of corticosteroids in patients proved to have PCNSL did not
affect the ability to make a final diagnosis in the majority of cases. The use of corticosteroids should be determined by clinical circumstance and necessity rather
than concern of obscuring PCNSL diagnosis. Because
of the retrospective nature of this study, there is surely
some bias in our results, and we may be underestimating the impact of corticosteroids on biopsy results in
cases of PCNSL. Nevertheless, the so-called vanishing
tumor phenomenon, in our experience and as reported
by other groups, is an exceptionally rare situation. As a
corollary, our results suggest that in the context where
contrast enhancement persists after administration of
corticosteroids, a biopsy can be performed with high
diagnostic yield.
1. Olson JE, Janney CA, Rao RD, et al. The continuing increase in
the incidence of primary central nervous system non-Hodgkin
lymphoma: a surveillance, epidemiology, and end results analysis.
Cancer 2002;95:1504 –1510.
2. Gunduz K, Pulido JS, McCannel CA, O’Neill BP. Ocular manifestations and treatment of primary central nervous system lymphomas. Neurosurg Focus 2006;21:E9.
3. Kaplan EL, Meier P. Nonparametric estimation from incomplete
observations. J Am Stat Assoc 1958;53:457– 481.
4. Shenkier TN, Blay JY, O’Neill BP, et al. Primary CNS lymphoma of T-cell origin: a descriptive analysis from the international primary CNS lymphoma collaborative group. J Clin Oncol 2005;23:2233–2239.
5. Jahnke K, Korfel A, O’Neill BP, et al. International study on
low-grade primary central nervous system lymphoma. Ann Neurol 2006;59:755–762.
6. Gametchu B. Glucocorticoid receptor-like antigen in lymphoma cell
membranes: correlation to cell lysis. Science 1987;236:456 – 461.
7. Bromberg JEC, Siemers MD, Taphoorn MJB. Is a “vanishing
tumor” always a lymphoma? Neurology 2002;59:762–764.
8. Haldorsen IS, Espeland A, Larson JL, Mella O. Diagnostic delay
in primary central nervous system lymphoma. Acta Oncol 2005;
44:728 –734.
9. Phan TG, O’Neill BP, Kurtin PJ. Post-transplant CNS lymphoma. Neuro Oncol 2000;2:229 –238.
Reversible Optic Neuropathy
with OPA1 Exon
5b Mutation
Karen Cornille, BS,1 Dan Milea, MD, PhD,2
Patrizia Amati-Bonneau, MD,3
Vincent Procaccio, MD, PhD,4 Lydie Zazoun, MD,5
Virginie Guillet, BS,3 Ghizlane El Achouri, BS,1
Cécile Delettre, PhD,1 Naı̈g Gueguen, PhD,3
Dominique Loiseau, PharmD,3 Agnès Muller, PhD,1
Marc Ferré, PhD,3 Arnaud Chevrollier, PhD,3
Douglas C. Wallace, PhD,4
Dominique Bonneau, MD, PhD,3
Christian Hamel, MD, PhD,1 Pascal Reynier, MD, PhD,3
and Guy Lenaers, PhD1
A new c.740G⬎A (R247H) mutation in OPA1 alternate
spliced exon 5b was found in a patient presenting with bilateral optic neuropathy followed by partial, spontaneous visual recovery. R247H fibroblasts from the patient and his
unaffected father presented unusual highly tubular mitochondrial network, significant increased susceptibility to apoptosis, oxidative phosphorylation uncoupling, and altered
OPA1 protein profile, supporting the pathogenicity of this
mutation. These results suggest that the clinical spectrum of
the OPA1-associated optic neuropathies may be larger than
previously described, and that spontaneous recovery may occur in cases harboring an exon 5b mutation.
Ann Neurol 2008;63:667– 671
Nonsyndromic inherited optic neuropathies consist in
the Leber hereditary optic neuropathy (LHON;
OMIM #535000) and the autosomal dominant optic
atrophy (ADOA, OMIM #165500), which similarly
end by a major degeneration of the optic nerve, affecting mainly the papillomacular fibers, resulting in a central scotoma.1 Although both conditions affect mainly
the retinal ganglion cells, their initial clinical presentaFrom the 1Institut National de la Santé et de la Recherche Médicale
U583, Institut des Neurosciences de Montpellier, Université de
Montpellier I et II, Montpellier, France; 2Copenhagen University
Hospital, Glostrup, Denmark; 3Institut National de la Santé et de la
Recherche Médicale U694, Département de Biochimie et Génétique, Centre Hospitalier Universitaire, Angers, France; 4Center for
Molecular and Mitochondrial Medicine and Genetics, University of
California, Irvine, Irvine, CA; and 5Cabinet d’Ophtalmologie, Paris,
Received Sep 7, 2007, and in revised form Jan 4, 2008. Accepted
for publication Feb 4, 2008.
This article includes supplementary materials available via the Internet
Published online Mar 21, 2008, in Wiley InterScience
( DOI: 10.1002/ana.21376
Address correspondence to Dr Lenaers, Institut des Neurosciences
de Montpellier; BP74103, 34091 Montpellier cedex 5, France.
E-mail:[email protected]
Cornille et al: OPA1 Reversible Optic Atrophy
Без категории
Размер файла
216 Кб
prevent, diagnosed, primary, corticosterone, system, central, can, nervous, determiners, stud, whether, pilot, use, previously, diagnosis, histological, lymphoma
Пожаловаться на содержимое документа