вход по аккаунту


Clinical studies of intrathecal autologous lymphocyte infusions in patients with malignant glioma A toxicity study.

код для вставкиСкачать
Clinical Studies of Intrathecal Autologous
Lymphocyte Infusions in Patients with
Malignant Glioma: A Toxicity Study
Edward A. Neuwelt, MD, Kemp Clark, MD, Joel B. Kirkpatrick, MD, and Howard Toben, PhD
The feasibility and toxicity of intrathecal lymphoid cell infusions in patients with glioma were examined in this
study. Blood rich in lymphoid cells was obtained using the Haemonetics Model 30 cell separator; the lymphoid cells
extracted were further purified on FicoU-Hypaque gradients. Four patients received a total of eighteen autologous
lymphoid cell infusions, with between 1 x lo6 and 5 x loy lymphoid cells being infused on each occasion. No
toxicity was observed, but the CSF glucose declined in 2 patients. I n 1 patient examined at autopsy the lymphoid
cells appeared to have gained access to the tumor bed as well as to the rest of the subarachnoid space.
Neuwelt EA, Clark K, Kirkpatrick JB, et al: Clinical studies of intrathecal autologous lymphocyte infusions in
patients with malignant glioma: a toxicity study. Ann Neurol 4:307-312, 1978
The blood-brain barrier is important in malignant
gliomas as it may partition the central nervous system
from the body's immune surveillance system. In fact,
the presence of the blood-brain barrier and absence
of CNS lymphatics are the main reasons that the central nervous system often is designated an immunologically privileged site [7, 81. Yet patients with
gliomas have been shown to have circulating tumorspecific killer lymphoid cells [2, 3 , 5, 6, 121. As reported by Ridley and Cavanagh [ 101, lymphoid cells,
although present in 50% of malignant gliomas, are
rarely able to penetrate the tumor parenchyma beyond the perivascular Virchow-Robin spaces. Takakura et a1 [ 111 demonstrated enhanced survival in a
small series of glioma patients when autogenous
white blood cells were placed in the tumor cavity;
Young e t a1 1131 have obtained similar results. If
these reports are accurate, patients with malignant
gliomas may derive benefit from direct exposure of
the tumor bed to white cells.
The particular type of white cell that should have
efficacy, if any does, is the lymphoid cell, which is not
only responsible for immunological memory, but is
also the effector of cellular immunity. In addition,
because of its paucity of cytoplasmic organelles compared with other white cells, a lymphoid cell may
have less toxicity when infused into the subarachnoid
space or tumor bed.
In the present communication, a study of the toxicity that results when autologous lymphoid cells
are injected directly into the subarachnoid space of
patients with glioma is described. This followed animal experimentation indicating a lack of toxicity [9].
These studies were carried out in 4 patients with
malignant astrocytoma who had undergone standard
surgical extirpation and postoperative irradiation of
the whole head.
Materials and Methods
Informed consent was obtained from the 4 patients as well
as from their immediate family members. All patients had
histologically proved malignant astrocytoma, i.e., astrocytoma grade 3 or 4. Patients were considered for admission to this study only after they had undergone standard surgical extirpation and postoperative whole-head
biplane megavoltage cranial irradiation. None of the patients had received steroids while on the protocol. All were
alert and able to communicate to varying degrees. The general physical condition of the 4 patients was reasonably
good, and none had any serious infections before or during
the study period. This research was carried out with the
approval of the Human Research Committees of the University of Texas Southwestern Medical School and the Dallas Veterans Hospital.
Lymphocytes were obtained by placing the patients on
the Haemonetics Model 30 blood processor* as described
by Aisner et a1 [I]. The lymphoid cell-rich blood obtained
from the cell separator was purified on Ficoll-Hypaque
gradients [9].The purified lymphoid cells were then resuspended in either Elliot's B solution or 0.85% saline solu"Haemonetics Corp, Natick, MA.
From the Division of Neurosurgery, the Division of Neuropathology, and the Department of Cell Biology, University of Texas
Health Science Center, Dallas, TX.
Address reprint requests to Dr Neuwelt, Division of Neurosurgery, University of Texas Health Science Center, 5323 Harry
Hines Blvd, Dallas, TX 7 5 2 3 5 .
Accepted for publication Mar 28, 1978.
0364-5 134/78/0004-04O3$O1.25 @ 1978 by Edward A. Neuwelt 307
tion. Cell counts were determined on a ZB-I Coulter
counter.* T h e cellular composition of the purified lymphocyte suspension was evaluated by examining Wrightstained smears. Viability was determined by dye exclusion,
with 0.05 ml of 0.4%. trypan blue added to an aliquot of the
purified lymphoid cell suspension and examination of the
cells under a light microscope. All lymphoid cell preparations were cultured for bacterial and fungal contaminants.
In 1 patient (No. 3) the purified autologous lymphoid cell
preparation grew Stapbylococcmepidermidis in the thioglycolate broth but not on the agar plates. This single positive
culture was thought to have been produced by a contaminant rather than by a breach of sterile technique.
T h e purified autologous lymphoid cells were infused into
the patient's cerebrospinal fluid by lumbar puncture within
six hours of procurement. The total volume infused was in
the range of 5 ml, and it was administered by barbotage.
Initially, 1 x logviable lymphoid cells were injected. Generally the infusion was repeated every two weeks, with 1 x
lo7, 1 x loR,1 x lo9, and finally 5 x 10' viable lymphoid
cells being given sequentially. To obtain 5 x lo9 purified
autologous lymphoid cells required ten runs on the
Haemonetics Model 30 cell separator, meaning that the
patient was o n the machine for three hours. Patients usually
became fatigued, so no attempt was made to keep them o n
the cell separator longer to obtain larger cell yields.
Three of the 4 patients died prior to the preparation of
this report and underwent a complete autopsy, including
removal and examination of the brain and spinal cord in
their entirety.
times, considerable numbers of platelets in the purified lymphocyte preparation. Despite multiple infusions, systemic hemoglobin remained stable in all patients.
Patient 1 (SCL)
A 56-year-old man presented with a mass in the pineal
region. Surgery included subtotal resection of a glioblastoma in that area. O n e month later the patient began
whole-head irradiation, receiving 4,500 rads. Subsequently
he functioned independently. No steroid medication was
required. Prior to the first intrathecal autologous lymphoid
cell infusion, a CAT scan showed evidence of residual
tumor in both occipital lobes, at which time his only abnormal neurological finding was difficulty with upward gate
and convergence.
Cerebrospinal fluid examinations showed that the initial
lymphoid cell infusion had produced little change in CSF
cell counts. However, each successive increase in the
number of cells infused evoked a transient rise in the total
CSF cell count. Within a few days the cell count returned tp
its preinfusion value. With infusion of 1 x loy and subsequently 5 X loy lymphoid cells, the total CSF white
count rose to as high as 1,000 to 1,500 white cells per cubic
millimeter, 90% being mononuclear cells (Fig 1 ). These
cells were 95 to 1OOF viable by dye exclusion. T h e single
exception to this pattern was the final infusion, when the
cell count remained at 161 mononuclear cells per cubic
millimeter two weeks following infusion.
Despite these markedly increased CSF white cell coynts,
CSF glucose content did not change (Fig 2 ) . (The patient
was diabetic.) There was no alteration in total CSF protein
following the infusion (Fig 3 ) . Total CSF protein gradually
increased over the eight weeks of study, consistent with a
recurrence of tumor. T h e patient's condition gradually deteriorated, and tumor recurred by CAT scan. H e died two
months after entry into the protocol, twelve days following
his last intrathecal lymphoid cell infusion, and nine months
from the time of surgery.
T h e general autopsy revealed lobar pneumonia and acute
pyelonephritis. Pathological examination of the C N S
showed glioblastoma in both occipital lobes, the epi-
Results and Case Studies
Eighteen infusions were carried out in these 4 patients with malignant astrocytoma, with the number
of cells given per infusion ranging from 1 x 106to 5
x lo9 viable lymphoid cells per infusion. As can be
seen from the Table, the viability of the cells infused
was generally in the range of 90 to 95%. The amount
of red cell contamination varied but was in the range
of 1 red cell per 2 to 6 white cells. There were, at
"Coulter Electronics, Hialeah, FL.
Characteristics of Pur$ed Lymphoid Cell Preparations Used for lntratheral Infuxions
Patient 1
Patient 2
Patient 3
Patient 4
No. of infusions
1 x 106 to
5 x 109
88 (82-90)
1 x 106 to
1 x 107
81 (69-93)
Range of lymphoid cells given
per infusion
Mean %, of mononuclear purified
white cells (range)
Mean viability of infused cells
Median ratio of red cells to
white cells in the cell
suspension (range)
Mean platelet count of purified
lymphoid cell preparation
1 x 106 to
5 x 109
89 (76-91)
1 x 106 to
5 x 109
92 (80-100)
92 (90-95)
97 (95-100)
94 (90-96)
87 (70-98)
1:6 (0-2:5)
1 7 ( 1 17-112)
1:l (114-211)
1 1 3(1120-1:2)
(175,000 to
(23,000 to
Annals of Neurology
Vol 4
No 4 October 1978
(Infusion V )
F i g 1 . Serial cerebrospinalfluid white cell countsfollowing intratberal injusions of large numbers of lymphoid cells in 3 patients with glionias (Nos. l , 3 . a n d 4 ) . The number of lymphoid
cells infused varied from I x IV't o 5 x I()!) viuble cells.
F i g 2. Serial rerebrospinaljluid g1zlro.w lecels following intrathecal infusions oJ1 x loyt o 5 x 10" i,iablelymphoid ceils
i n 3 patients with gliomas (Nos. I , 3 , and4j.
thalamus, and the thalamus, with seeding of the tumor
into the spinal subarachnoid space. No indication of
inflammatory reaction or demyclination was seen within
the CNS. There was no evidence of obstructive hydrocephalus, and the arachnoid granulations were normal.
There was little communication between the tumor and the
subarachnoid space.
Patient 2 (JG)
A 56-year-old man presented with a right parietal mass.
Subtotal resection of a malignant astrocytoma was carried
out. Postoperatively the patient received 6,000 rads o f
whole-head irradiation, after which steroid medications
were discontinued. He was started o n the protocol.
Examination of the CSF after infusions of 1 X LO" and 1
x lo7 lymphoid cells revealed a mild rise in the cell count.
With the second infusion, 22 white cells were seen in the
spinal fluid 24 hours after infusion, but only 405%of these
were mononuclear. T h e CSF glucose and protein did not
change greatly following either infusion. T h e systemic
white cell count remained stable. T h e patient continued to
deteriorate from tumor recurrence and died eleven days
following his second intrathecal lymphoid cell infusion,
which was four months from the time of surgery.
General autopsy revealed acute bronchopneumonia and
multiple organizing and recanalizing thrombotic pulmonary
emboli. The CNS contained recurrent tumor in the right
;, ioo
F i g 3 . Serial rerebrospinaljuid total protein feziels following intrathecal infusion of 1 x 1 O9 to 5 x 1 O9 ziadle lymphoid cells
i n 3 patients with gliomas (Nos. I . 3 , and4j.
Neuwelt et al: Lymphocyte Infusions in Glioma Patients
herniation for at least 2.5 cm across the midline. The tumor
extended down to the level of the midpons. There was
edema of the centrum semiovale but no gross evidence of
demyelination. Microscopic evaluation revealed a marked
perivascular lymphocytic infiltrate throughout the C N S but
particularly in the area of the tumor. T h e degree of perivascular lymphocytic infiltration in the tumor bed was greater
than is normally seen following surgery and irradiation (Fig
5A). These lymphocytic infiltrates not only were perivascular but also seemed to extend into the infiltrating margins
of the tumor and into tumor that had seeded the subarachnoid cisterns (Fig 5B). Away from the area of the
tumor there was no lymphocytic infiltration in the C N S
parenchyma, but there were lymphocytes in the subarachnoid space and in the perivascular spaces (Fig SC, D).
With the exception of one small perivascular area adjacent
to the tumor, there was no evidence of demyelination anywhere in the CNS. The choroid plexus was normal and
without sign of inflammation (Fig 5E).
F i g 4. Serial cerebrospinal~%id cell countsfollowing intrathecal
infisions of I avying nirnzhers oj'lymphoid cells i n a patient with
a gliorna (No. 3).
parietal area but no evidence of inflammation or demyelination. The C N S system was examined, including the spinal cord and overlying leptomeninges. Cultures obtained
before his death revealed no growth from the CSF.
Patient 3 (AVB)
A 52-year-old white man underwent left temporal lobectomy for a malignant astrocytoma. He received 6,000 rads
of whole-head irradiation. Two months later the patient
was entered into the protocol and received six intrathecal
infusions (Fig 4 ) . An aliquot of the purified lymphoid cells
used for the final two infusions was evaluated for percentage of T-cells present. Using the sheep red cell E rosette
method [4],72% and 7 4 % , respectively, of the lymphoid
cells were found to be T-cells. Transient hypoglycorrhachia
followed infusion of 1 x 10' to 5 x 10' lymphoid cells (see
Fig 2). The peripheral white cell count did not change.
During the first eight weeks of the protocol the patient
showed some improvement in speech. A C A T scan after
five intrathecal lymphoid cell infusions over a thirteenweek period demonstrated a diminished shift of the midline, decreased deformity of the lateral ventricles, and less
tumor enhancement in the tumor bed. Subsequently, however, his condition deteriorated, and he died seven months
Autopsy revealed pulmonary edema and hepatic cirrhosis. Ncuropathological examination showed cerebral
310 Annals of Neurology
Vol 4 N o 4 October 1978
4 (NG)
A 58-year-old man underwent subtotal excision of a grade
3 astrocytoma of the left temporal lobe. He had a mixed
expressive and receptive dysphasia and right hemiplegia.
H e received 6,000 rads of whole-head irradiation.
T h e patient was begun o n the protocol six months following tumor resection. His neurological examination remained unchanged throughout the course of the five lymphoid cell infusions. A CAT scan after two infusions
showed some decreased mass effect. Studies of the CSF
revealed that the initial infusions of 1 x lofi and 1 x l o 7
lymphocytes had no appreciable effect on the CSF white
cell count, but following infusions of 1 x 10' and 9.9 x lo8
of viable white cells, the CSF cell count rose to over 1,000
cells per cubic millimeter, of which 75% were mononuclear cells (see Fig 1). T h e CSF total protein showed little
consistent change with the lymphocyte infusions (see Fig
3). With a lymphoid cell infusion in the range of 1 x lo9
cells, the CSF glucose fell (see Fig 2); i t returned to normal
within two to three days.
About three weeks following the final lymphocyte infusion, the patient developed decreased consciousness. A
C A T scan revealed increased tumor enhancement suggestive of recurrence. This was nine months from the time of
No conclusion can be made from this study either in
support of o r against the efficacy of intrathecal autologous lymphocyte infusions in patients with primary CNS tumors. The present study does indicate
that intrathecal infusions of autologous lymphocytes
have n o apparent toxicity in patients with glioma. O n
the basis of t h e cisternograms performed in our patients, it is very likely that few, if any, infused cells
actually reached the tumor bed except in Patient 3.
T h e pathological suggestion of cell-mediated tumor
destruction in that patient is encouraging. On the
other hand, 2 patients had mild tumor regression dur-
Fig 5. Histopathologicalexamination in Patient 3 . (A)Lowpower view shows a cessel within the tumorsurrounded by a
dense infiltrate of lymphocytes and other mononuclear cells. ( B )
Lymphocyte infiltration ofglioma that has seeded into thepontine cistern. Uninvoluedpons can beseen at the top of thephotomicrograph. (C) Subarachnoid space at autopsy, four weeks following the last lymphoid cell infusion. Note the persistent
marked lymphocytic infiltration of the subarachnoidspace (top,.
( D ) Perivascular Lymphocyte infiltration at autopsy in part of
brain uninvolved with tumor. There was no evidence of periziascukzr demyelination in association with these lymphocytic
infiltrates. (E) Normal cboroidplexus at autopsy. (AllHOE;
A , B, C , and E x 40, D x 100, all before 3 5 p,reduction.)
Neuwelt et al: Lymphocyte Infusions in Glioma Patients 311
ing lymphocyte infusions, but this may have resulted
from the previous radiotherapy rather than from any
effect of the infused lymphocytes. Both these patients showed hypoglycorrhachia following infusions
of autologous lymphoid cells. In previous studies of
infusion of human lymphoid cells into the subarachnoid space in rabbits, in even greater numbers
than were used here, no hypoglycorrhachia was observed [C,]. Hypoglycorrhachia may be a reflection of
lymphocytes undergoing blastogenesis and subsequently exerting a tumoricidal effect. Despite multiple lymphoid cell infusions, none of the patients
developed inflammation of the choroid plexus. In
previous animal studies, choroid plexitis was observed following repeated xenogenic lymphoid cell
infusions, but not after multiple syngeneic infusions.
Finally, it should be mentioned that intrathecal
autologous lymphoid cell infusions into a tumor bed
exposed to the subarachnoid space may be better
than direct intratumoral infusion for two reasons.
First, the infused lymphocytes, after exposure to the
tumor bed and subsequent return to the systemic
circulation, may further activate the afferent limb of
the immune response. Thus, the proportion of sensitized lymphoid cells may progressively increase
after each infusion. Second, there is less risk of
creating a localized mass effect. O n the other hand, if
there is not good communication between the tumor
bed and the subarachnoid o r ventricular CSF, then
direct intratumoral infusion may be better.
1. Aisner J, Schiffer CA, Wolff JH, et al: A standardized tech-
The technical assistance of Sybille Galosy as well as the assistance
of the Wadley Blood Bank and its staff were greatly appreciated.
312 Annals of Neurology
Vol 4
N o 4 October 1978
nique for efficient platelet and leukocyte collection using the
Model 30 blood processor. Transfusion 16437-445, 1076
Brooks WH, Netsky MG, Normansell DE, et al: Depressed
cell-mediated immunity in patients with primary intracranial
tumors. Characterization of a humoral immunosuppressive
factor. J Exp Med 136:1631-1647, 1972
Eggers AE: Auto-radiographic and fluorescence antibody
studies of the human host immune response to gliomas. Neurology (Minneap) 22~246-250, 1972
EvansJ, Smith MA, Steel CM: Rosetting test. Lancet 1:96-97,
Jelsma R, Buch PC: Glioblastoma multiforme: its treatment
and some factors in patients with tumors of the central nervous system. Int J Cancer 12:194-205, 1973
Levy NL, Mahaley MS Jr, Day ED: In v i m demonstration of
cell-mediated immunity to human brain tumors. Cancer Res
32:477-482, 1972
Medawar PN: Immunity to homologous grafted skin 111: Fate
of skin homografts transplanted to the brain, to subcutaneous
tissue, and to the anterior chamber of the eye. Br J Exp Pathol
29:58-69, 1948
Neuwelt E, Clark K: Clinical Aspects of Ncuroimmunology.
Baltimore, Williams & Wilkins, 1978, in press
Neuwelt E, Doherty D: Toxicity, kinetics and clinical potential of subarachnoid lymphocyte infusions. J Neurosurg
47:205-217, 1977
Ridley A, Cavanagh B: Lymphocytic infiltration in gliomas:
evidence of possible host resistance. Brain 94: 1 17- 124, 197 1
Takakura K, Mili Y,Kubo 0,et al: Adjuvant immunotherapy
for malignant brain tumors. Jpn J Clin Oncol 2:109-120,
Wahlstrom T, Saksela E, Troupp H: Cell-bound antiglial immunity in patients with malignant tumors of the brain. Cell
Immunol6161-170, 1973
Young HC, Kaplan A, Regelson W, et al: Immunotherapy
with autologous white cell infusions (lymphocytes) in the
treatment of recurrent glioblastoma-a preliminary report.
Cancer 40:1037-1044, 1977
Без категории
Размер файла
635 Кб
patients, toxicity, clinical, stud, malignant, glioma, autologous, intrathecal, studies, lymphocytes, infusion
Пожаловаться на содержимое документа