visual scenes after unilateral temporal- o r frontal-lobe resection. Neuropsychologia 1993;31:1-15.
27. Penfield W, Jasper H. Epilepsy and the functional anatomy of
the human brain. Boston: Little, Brown, 1954.
28. Penfield W, Perot P. The brain's record of auditory and visual
experience. A final summary and discussion. Brain 1963;86:
595-696.
29. Wieser HG. Zur frage der lokalisatorischen bedeutung epilptischer halluzinationen. In: Karbowski K, ed. Halluzinationen
bei epilepsie und ihre differentialdiagnose. Bern: Huber, 1982:
67-92.
30. Schultz G, Melzack R. The Charles Bonnet syndrome: phantom visual images. Perception 1991;20:809-825.
31. Jackson JH. On the anatomical, physiological and pathological investigation of the epilepsies. West Riding Lunatic Asylum Medical Reports 1873;3:315-339.
32. Cogan DG. Visual hallucinations as release phenomena. Albrecht v. Graefes. Arch Klin Exp Ophthalmol 1973;188:139150.

33. Swash M. Visual perseveration in temporal lobe epilepsy.
J Neurol Neurosurg Psychiatry 1979;42:569-571.
34. Kaplan RF, Meadows ME, Verfaellie M, et al. Lateralization
of memory for the visual attributes of objects: evidence from
the posterior cerebral artery amobarbital test. Neurology
1994;44:1069-1073.
35. Ptito A, Zatorre FLJ, Petrides M, Frey S, Alivisatos B, Evans
AC. Localization and lateralization of stereoscopic processing
in the human brain. Neuroreport 1993;4:1155-1158.
36. Njemanze PC, Gomez CR, Hornestein S. Cerebral lateralization and color perception: a transcranial doppler study. Cortex
1992;28:69-75.
37. Harvey M, Milner AD, Roberts RC. Spatial bias in visuallyguided reaching and bisection following right cerebral strokes.
Cortex 1994;30:343-350.
38. Vallar G, Perani D. The anatomy of unilateral neglect after
right-hemisphere stroke lesions. A clinicaVCT-scan correlation
study in man. Neuropsychologia 1986;24:609-622.

Multiple cerebral lesions complicating
therapy with interleukin-2
B. Illowsky Karp, MD; J. C. Yang, MD; M. Khorsand, MD; R. Wood, MD; and T. C. Merigan, MD

Article abstract-We reviewed the records and radiologic studies of eight patients who developed new focal neurologic
abnormalities while receiving interleukin-2 (IL2bbased immunotherapy for malignancy or HIV infection. Initial confusion
and delirium in the patients evolved into coma, ataxia, hemiparesis, seizures, and cortical syndromes including aphasia,
apraxia, and cortical blindness. Imaging studies showed multiple white and gray matter lesions with a predilection for the
occipital poles, centrum semiovale, and cerebellum. After cessation of IL2 treatment, seven patients improved to normal or
near-normal neurologic function paralleled by resolution of the lesions on scans. One patient improved only minimally.
Possible etiologies for the lesions include an IL2-induced cerebral vasculopathy, a direct toxic effect of IL2, or immunologically mediated damage.
NEUROLOGY 1996;47:417-424

Interleukin-2 (IL2) is a lymphokine effective against
advanced metastatic malignancies in some patients,'." which may also act against the human immunodeficiency ~ i r u s .IL2
~ , ~is used alone, in a longacting preparation conjugated to polyethylene glycol
(PEG-ILB), in combination with activated lymphocytes as part of adoptive immunotherapy, or with
other cytokines such as alpha-interferon (alpha-IFN)
or granulocyte-macrophage colony stimulating factor
(GM-CSF).l"
Systemic side effects of IL2 such as fever, chills,
malaise, nausea, vomiting, and diarrhea are common. Fluid extravasation from a capillary leak syndrome may lead to hypotension, multiorgan failure,
and arrhythmias.1.7 The most common neurologic
side effect of IL2 therapy is a reversible encephalopathy characterized by confusion, lethargy, and poor
concentration.R-10

It can progress to delirium with psychotic features
or to coma. Over the past several years, there have
been a few case reports of focal neurologic deficits in
patients receiving IL2. Aphasia, hemiparesis, seizures, visual distortions, and ataxia have occurred as
well as brachial plexopathy and carpal tunnel syndrome.11-17Vecht e t a1.I6 reported a n acute fatal demyelinating leukoencephalopathy in a patient receiving IL2.
As therapy with IL2 has become more widespread,
neurologic illness has become a n important cause of
morbidity and a treatment-limiting complication. We
describe eight patients who developed neurologic deficits and lesions on brain imaging during IL2-based
treatment of malignancy or HIV infection.
Methods. From 1985-1994, over 1,500 patients received
IL2-based therapy for malignancy at our institution, in-

From the National Institute of Neurological Disorders and Stroke (Dr. Illowsky Karp), the National Cancer Institute (Dr. Yang), National Institutes of
Health, Bethesda, MD; the Comprehensive Cancer Center (Dr. Khorsand), University of Wisconsin, Madison, WI; the Center for AIDS Research (Drs. Wood
and Merigan), Stanford University, Stanford, CA.
Received September 20, 1995. Accepted in final form February 1, 1996.
Address correspondence and reprint requests to Dr. Barbara I. Karp, Building 10, Rm. 5N-226, NINDS, NIH, Bethesda, MD 20892.
Copyright 0 1996 by the American Academy of Neurology

417

cluding five who developed neurologic illness with abnormalities on magnetic resonance imaging (MRI) during or
shortly after treatment. No patient had metastases to the
nervous system or a history of neurologic disease or stroke
prior to receiving IL2. The medical records and imaging
studies were reviewed. Three additional patients (cases 4,
7 , and 8) receiving IL2 for HIV infection or malignancy at
other institutions with similar neurologic symptoms,
courses, and imaging studies were also studied.

Case reports. Patient 1. A 53-year-old man with renal
cell carcinoma underwent nephrectomy and radiotherapy
before referral for IL2. His immunotherapy included
alpha-IFN and IL2. Five days after the start of treatment,
the patient became progressively confused and lethargic
and lapsed into coma. He had intermittent left focal seizures, which stopped spontaneously within 24 hours. Peritoneal dialysis was begun on the 10th hospital day for
renal failure. Two days after beginning dialysis, the patient was alert but disoriented with hallucinations. The
symptoms improved over the next week. On the 22nd hospital day, left focal seizures recurred and were followed by
a generalized seizure. He was started on phenytoin.
Following the seizure, he had a left spastic hemiparesis.
Cerebrospinal fluid (CSF) was normal except for a slight
elevation in protein to 56 mg/dL. An electroencephalogram
(EEG) showed diffuse background slowing with right temporal lobe periodic epileptiform discharges. A CT of the
head showed nonenhancing white matter lucencies extending from occipital areas forward into the centrum semiovale. White matter hyperintensities, which were especially prominent occipitally, were seen on MRI.
He improved rapidly and was fully awake 48 hours
after the seizure. A repeat MRI 3 weeks later was unchanged, and an EEG was improved. On discharge from
the hospital on the 38th hospital day, the patient had
persistent reflex asymmetry but an otherwise normal neurologic examination. MRI 1 month after discharge showed
resolution of the white matter abnormalities.
Patient 2. A 41-year-old man with renal cell carcinoma
started IL2 2 months after a radical nephrectomy. He received seven doses of IL2, then three weekly doses of PEGIL2. During the second week of PEG-IL2 treatment, he
developed right proximal arm weakness, sensory loss, and
shoulder pain with winging of the scapula. Deep tendon
reflexes were normal. MRI of the cervical spine showed
bony degeneration and bulging disks a t C-5 through C-7.
MRI of the head showed a few small areas of high signal in
the cerebral white matter.
A second cycle of therapy began 1 month later. He received five doses of IL2 followed by PEG-IL2 for 3 weeks.
Two days after the last PEG-IL2 dose, he became delirious
and was treated with haloperidol. On examination, the
patient was lethargic but able to state his name and open
his eyes on command. There was no visual fixation or
response to visual threat. The pupils were small and reactive to light. A right gaze preference was present. Both legs
moved spontaneously, but the right arm was immobile and
hyperreflexic. The left arm withdrew to noxious stimuli
only. Plantar responses were extensor. The patient had a
generalized tonic-clonic seizure, which was treated with
phenytoin.
At the time of neurologic deterioration, the white blood
418

NEUROLOGY 47 August 1996

cell (WBC) count was 23,000/mm3 with 46% eosinophils.
Coagulation studies, serum sodium, magnesium, calcium,
and liver function tests were normal. There was mild renal
insufficiency. CK was markedly elevated at 1,024 U/L. A
nonenhanced CT of the head showed no lesions. MRI
showed new cortical, subcortical, and cerebellar lesions,
including bilateral occipital lesions that did not enhance
with gadolinium (figure 1A). The CSF had 3 white blood
cells and 44 red blood cells/mm". CSF protein, glucose, and
cultures were normal. An EEG done 4 days after the seizure showed diffuse slowing without epileptiform activity.
The patient remained intubated and pharmacologically
paralyzed for 5 days after the seizure. When paralysis and
sedation were discontinued, he gradually awoke. He was
able to read and identify objects and colors, but complained
of difficulty locating objects visually. The right arm remained weak and atrophic. MRI of the head 1 week later
showed persistence of the brain lesions, which now enhanced with gadolinium (figure 1B).
Two months later, his mental status and vision were
completely normal. The right arm weakness had improved.
An MRI 5 months after admission showed only those small
white matter lesions that had been present prior to the
second cycle of IL2.

Patient 3. A 63-year-old woman with renal cell carcinoma of the sternum tolerated an initial course of IL2
followed after 1 week by PEG-IL2. A second, higher dose of
PEG-IL2 was given 1 week later. Forty-eight hours after
the second PEG-IL2 administration, she became confused,
agitated, incontinent, and progressively somnolent. Her
speech was dysarthric. Strength, tone, and deep tendon
reflexes were decreased in the left arm and leg. Snout,
root, and bilateral grasp reflexes were present. The left toe
was upgoing. CSF analysis was normal. An MRI of the
head showed possible bilateral occipital hyperintensities
that had not been present on a baseline scan before admission.
The patient deteriorated. By the next day she was obtunded and unable to follow any commands and had intermittent decorticate posturing of her right arm. A repeat
MRI showed multifocal white matter and cortical hyperintensities including definite bilateral occipital lesions (figure 2A). An EEG showed generalized slowing. Because the
first two patients had seizures in a similar clinical setting,
this patient was given phenytoin prophylactically.
At the time of neurologic deterioration, she had mild
respiratory insufficiency, hypotension, and oliguria. Her
WBC count was 29,00O/mm" with 50% eosinophils. Her
serum CK rose to 694. The prothrombin time was slightly
prolonged a t 18.6; other coagulation studies were normal.
The patient continued to worsen. By the fifth day after
the last IL2 infusion, she could open her eyes spontaneously but did not focus visually, follow objects, or respond
consistently to visual threat. Her pupils were equal and
reactive. She appeared unaware of her surroundings and
did not follow commands. There was no spontaneous movements of her arms or legs, and the left side did not withdraw to pain.
Over the next several weeks, the left spastic hemiparesis persisted. Noxious stimuli to the right side of the body
elicited flexor posturing. Myoclonic jerks were intermittently present in the right arm. A CT without contrast

A

13

Figure 1. (A) T,-weighted MRI of patient 2 showing multiple cortical lesions. The lesions did not enhance with gadolinium. (B) TI-weighted MRI with gadolinium obtained 1 week after that i n A showing new enhancement of lesions.
showed multiple well-circumscribed lucencies throughout
the cerebral hemispheres (figure 2B).
There was only slight improvement over the second
month of hospitalization. Although she became responsive
to her surroundings and was able to speak, she had irritability and marked deficits in attention and concentration.
She was able to recognize her family and some objects. All
limbs were spastic and hyperreflexic, and she had no
movement of her right hand. Bilateral grasps persisted.
She was transferred to a rehabilitation facility 7 weeks
after receiving her final dose of IL2.

Patient 4. A 43-year-old HIV-seropositive man on azidothymidine began weekly doses of PEG-IL2. Approximately
60 hours after the sixth infusion, he became confused and
unsteady. The following day he complained of mild bifrontal headache and was unable to feed or clothe himself
properly. On examination, he was alert and oriented to
person and place, but not date. His answers were slow, and
he was unable to follow multi-step commands or do even
simple arithmetic. Left-right confusion and perseveration
were present. He could neither write clearly nor copy figures.
Cranial nerve examination was normal as was sensation to pinprick, light touch, vibration, and position. He

had agraphesthesia and astereognosis. Strength was normal. Rapid alternating movements were performed poorly,
and he was unable to walk with a tandem gait. Babinski
responses were present bilaterally.
Laboratory studies included a WBC count of 18,700/
mm3 with 41% eosinophils. CD4 count was 450 cells/kL.
LDH was elevated a t 2,286 IU/L, and AST a t 139 IU/L.
Renal function was normal. The CSF contained nine RBCs/
mm3 and five WBCs/mm". The CSF protein was 57 mg/dL.
Two CTs were unrevealing. MRI 3 days after admission
showed diffuse, nonenhancing white and gray matter lesions. An EEG showed left temporal slowing with bursts of
frontal intermittent rhythmic delta activity. The patient
improved rapidly and was neurologically normal at discharge on the fifth hospital day.

Patient 5. A 28-year-old woman with metastatic malignant melanoma received IL2 after failing to respond to
conventional chemotherapy. After her seventh dose of IL2,
she developed renal and respiratory insufficiency, hypotension, and peripheral edema. She required a brief period of
respirator support. After extubation, the patient was confused. Although awake and alert, she was oriented to person only. She had little spontaneous speech, mild wordfinding difficulty, poor repetition of words, and
August 1996 NEUROLOGY 47 419

H

A

Figure 2. (A) T,-weighted MRI of patient 3 showing multiple cortical and subcortical lesions. (B) CT obtained 3 weeks
after A. The lesions appear as multiple cortical and subcortical lucencies.
perseveration. The deep tendon reflexes were brisk, and
plantar responses were extensor. A head CT was normal.
However, MRI showed a hyperintense lesion in the left
parietal lobe. The CSF was normal. The patient's symptoms cleared completely over 5 days. The lesion was
smaller but still visible on scans performed up to 10
months later.

Patient 6. A 56-year-old man with metastatic renal cell
carcinoma was referred for immunotherapy after a radical
nephrectomy. He completed the first course of eight doses
of IL2 without difficulty. One week later, he was lethargic
and had recurrent episodes of vertigo lasting about 15
minutes. He complained of unsteadiness while walking
and a loss of grip strength in the left hand. The symptoms
cleared for 2 days, then recurred along with blurred vision.
A left homonymous hemianopia and bilateral limb ataxia
were found on examination.
At the time the neurologic symptoms developed, renal
function was normal, but the alkaline phosphatase and
AST were slightly increased. The peripheral WBC count
was 13,000/mm3with 2% eosinophils. The CSF contained 3
WBC/mm" and was otherwise normal. A head CT was
normal, but an MRI showed multiple hyperintense lesions
on the T,-weighted scan. Lesions involved both gray and
white matter and were found in the occipital cortex, left
thalamus, and cerebellar hemispheres. They did not enhance.
The patient improved over 8 days. MRI 1 week after the
initial scan showed the earlier lesions, but additionally
found that the right occipital lesion enhanced with gadolinium. All lesions were smaller on a scan obtained 6
weeks later.
420

NEUROLOGY 47 August 1996

Patient 7. A 48-year-old man with metastatic renal cell
carcinoma became confused and lethargic after two infusions of IL2 and GM-CSF. His neurologic examination was
otherwise normal, as was MRI. Further immunotherapy
was stopped because his white blood cell count had risen to
95,70O/mrn:' with 55% eosinophils. Hydrocortisone and hydroxyurea were given to decrease the white blood cell
count. At the time neurologic symptoms began, his renal
function was normal; LDH and alkaline phosphatase were
elevated to twice normal levels.
One day later, the patient deteriorated. He complained
of dizziness and a headache and was disoriented with cognitive difficulties, hallucinations, expressive aphasia, and
right-hand weakness. An EEG showed generalized and
right frontal slowing. The CSF was normal. A second MRI
showed nonenhancing, white matter hyperintensities in
both frontal lobes, occipital poles, right parietal lobe, and
right parietotemporal junction. MR angiography was normal. A third MRI 1 day later showed the previous white
matter lesions and new signal abnormalities in the gray
matter of the medial frontal lobes bilaterally. The patient
gradually improved. By 6 weeks after admission he noticed
only a slight impairment of fine motor movements when
playing piano.
Patient 8. A 52-year-old man who underwent nephrectomy for renal cell carcinoma in 1986 had a recurrence of'
disease in 1994. He received IL2 and GM-CSF for 4 days.
The second week of therapy was interrupted by severe
diarrhea. The doses of both drugs were halved when treatment was restarted, and that of GM-CSF was reduced
further when the absolute eosinophil count reached
20,000. During the third week of treatment, the patient

Table Patient demographics and drug doses

--__

Patient

SedAge

Diagnosis

DruglDose'''

Time of Symptom Onset

1

MI53

RCC

H-L IL2 4.5 mU/m2 X 11 doses
alpha-IFN 3 mU X 11 doses

22 days after start of treatment

2

M/4 1

RCC

First course:

3 days after start of second PEG-IL2 course

IL2: 720,000 I U k g X 7 doses
PEG-IL2: 3 mU/m2 X 1 dose
PEG-IL2: 6 mU/m' X 3 doses
Second course:
IL2: 720,000 I U k g X 7 doses
PEG-IL2: 3 mU/mz X 1 dose
PEG-IL2: 6 mU/m2 X 1 dose
3

F/63

RCC

IL2: 720,000 I U k g X 7 doses

3 days after second PEG-IL2 dose

PEG-IL2: 3 mU/m2 X 1 dose
PEG-IL2: 6 mU/m2 X 1 dose
4

M/43

HIV

PEG-IL2: 3 mU/m" X 6 doses

3 days after sixth PEG-IL2 dose

5

F/28

Melanoma

IL2: 720,000 I U k g X 7 doses

6 days after seventh IL2 dose

6

M/56

RCC

IL2: 720,000 I U k g X 8 doses

8 days after first course IL2 completed

7

M/48

RCC

IL2: 1.5 mU/mz X 8 doses

6 days after eighth IL2 dose

GM-CSF: 5 Fg/mz
8

M/53

RCC

IL2: 1.5 mU/m2 X 6 doses

2 days into second week of treatment

GM-CSF: 5 pgln?

* H-L IL2: Hoffman-LaRoche; other patients received the ChirodCetus preparation
RCC = renal cell carcinoma.

became confused and restless. He was able to follow only
simple commands and had poor attention with perseveration. A right homonymous hemianopia was present as well
as right-sided neglect and ocular apraxia. There was right
facial and arm weakness, hyperreflexia, and bilateral
Babinski signs.
At the time of neurologic deterioration, the white blood
cell count was 41,000/mm3 with 50% eosinophils. The
platelet count was 505,0OO/t~L,and serum viscosity was
increased to 4.3 (normal, 1.4 to 1.8).Serum protein electrophoresis was normal. Fibrinogen level was normal and
fibrin split products were absent. CSF was normal. CT and
MRI scans showed nonenhancing lesions in the left cerebellum and left frontal deep white matter. Repeat MRI
several days later showed new nonenhancing lesions in
both gray and white matter involving the thalamus, cerebellum, and occipital lobes. A third MRI the next day
showed gadolinium enhancement in the left cerebellar, occipital, and posterior frontal lesions.
The patient gradually improved. Six weeks later, examination showed slight right arm weakness but was otherwise normal.

Results. The patients included six men and two women
ranging in age from 28 to 63 years (mean age, 48 years).
Six patients received IL2 for renal cell carcinoma, one for
melanoma, and one for HIV infection (table). IL2 was the
sole immunotherapeutic agent in five patients, three of
whom received the long-acting preparation, PEG-IL2. One
patient received alpha-IFN and two patients received GMCSF in addition to IL2.

Neurologic symptoms did not begin immediately with
the start of IL2 treatment. Five patients became symptomatic 3 to 8 days after completing a course of therapy, the
other three during a course of treatment.
In each patient, the first sign of deterioration was a
change in mental status with confusion, disorientation,
and agitation. Three patients had hallucinations. Two patients became obtunded and one comatose. Two patients
had seizures. Speech difficulties, including aphasia, dysarthria, and mutism, were present in four patients. Two
patients had cortical blindness, and two had a homonymous hemianopia. Corticospinal tract involvement was
manifest in seven of the eight patients who had weakness,
spasticity, hyperreflexia, or Babinski signs. Three patients
had perseveration, and one had primitive reflexes, including snout, bilateral grasp, and rooting reflexes. Four patients had limb or gait ataxia. Two patients had hemisensory deficits. One patient had left-right confusion,
agraphesthesia, and astereognosis.
All patients had abnormal brain imaging. Multiple lesions were present in both gray and white matter in cortical and subcortical territories except for one patient who
had an isolated left cortical parietal lesion. The lesions
were not apparent or hypodense on CT (see figure 2B). On
MRI, the lesions were either not visible or hypointense on
TI-weighted images, and hyperintense on T,-weighted images. In one patient, the lesion enhanced with gadolinium
on the first MRI scan obtained. In three other patients,
lesions that did not enhance on the initial MRI scans did
show enhancement on scans obtained 5 to 7 days later (see
August 1996 NEUROLOGY 47 421

figure 1).MR angiography, performed in only one patient,
was normal.
Seventy-five percent of the patients had occipital lobe
involvement, which was often bilateral. Half had parietal
lesions and half frontal lesions. The cerebellum was involved in 33% of patients and the thalamus in 25%. No
patient had brainstem lesions.
EEGs, recorded in five patients, showed generalized
slowing. Additional focal slowing was noted in two patients. One patient had periodic localizing epileptiform discharges, and one patient had frontal intermittent rhythmic
delta activity. CSF analysis was usually normal. Two patients had a slight increase in CSF protein. Three patients
had a mild lymphocytic pleocytosis.
At the time of neurologic compromise, one patient had
renal failure and several others had mild renal or hepatic
dysfunction. Two patients had hypotension, which improved with intravenous fluid infusion. Routine coagulation studies were usually normal. The serum viscosity was
elevated in the single patient in whom it was measured.
All patients except one had an IL2-induced eosinophilia.
Eosinophilia was mild in one patient whose absolute eosinophil count was 700/mm", but was marked in the others
whose absolute eosinophil counts ranged from 1,372/mm3
to 57,OOO/mm:'.
No patient received additional IL2 after the development of neurologic symptoms. While patients did not receive specific treatment for their neurologic symptoms, all
received supportive care. Several required medication for
agitation o r seizures. Seven patients had complete or
nearly complete recovery of neurologic function. Improvement usually began within a week of the onset of neurologic symptoms. One patient, however, had only minimal
recovery. This patient remained severely encephalopathic
with a spastic quadriparesis.

Discussion. The patients presented here developed neurologic illness during or shortly after treatment with IL2 with corresponding cortical and subcortical lesions on imaging studies. The initial
behavioral changes were often similar to the more
common IL2-induced encephalopathy, but focal signs
rapidly evolved reflecting the areas of the brain affected.
There was no evidence to suggest a n infectious or
malignant cause for the neurologic illness. Patients
with nervous system metastases were excluded from
the IL2 trials, and CSF cytology and cultures were
negative. Metabolic disturbances, including hyponatremia, hepatic failure, renal insufficiency, or hypoxia, were present in some of our patients, but they
are not likely to have caused the neurologic syndrome as they were often mild and transient. One
patient had renal failure severe enough to require
dialysis, but the others had only oliguria or a slight
increase in BUN and creatinine. The disturbance in
three patients was limited to transaminase elevation. No single abnormality was uniformly present in
all patients, and none is known to cause focal symptoms with multiple cortical and subcortical lesions.
Two of the patients were co-treated with GM-CSF
and one with alpha-IFN. The other five patients re422

NEUROLOGY 47 August 1896

ceived IL2 alone. Three of the patients received IL2
conjugated to polyethylene glycol (PEG), which decreases the clearance rate 15-f0ld.'~The neurologic
illness, however, cannot be due to GM-CSF, alphaIFN, or the PEG moiety, since nonconjugated IL2
alone caused a similar picture and GM-CSF has no
known neurologic complications. This neurologic toxicity occurred in 2 of 92 patients treated with PEGIL2 a t the National Cancer Institute. While PEG-ILS
may prolong exposure to high levels of IL2, in the
absence of a prospective study it is not clear that
neurologic complications are more frequent in patients receiving PEG-IL2 than in those receiving unconjugated IL2.
Patients with hypereosinophilia may develop peripheral neuropathy, cerebral infarction, or dementia
with white matter lesions on MRI.19-22
While we cannot dismiss the possibility that hypereosinophilia
contributed to the development of neurologic symptoms, marked eosinophilia was not present in all of
our patients, and there are no previous reports of
this constellation of signs and reversible lesions in
patients with idiopathic hypereosinophilia.
There are a few reports of similar neurologic illness associated with IL2. Aphasia, hemiparesis, seizures, visual distortions, and ataxia occurred in a
few patients receiving IL2I2 or PEG-IL2" for metastatic renal cell carcinoma reported by Philip et al.
and Bukowski et al. Meyers and Yung2" described a
melanoma patient who developed a cerebellar syndrome and dementia 3 months after completing
treatment with IL2 who had multiple white matter
lesions on a later MRI scan.23The melanoma patient
of Somers e t al.I5 developed a spastic quadriparesis
following IL2 with subcortical and cortical hyperintensities on MRI. The neurologic symptoms and the
lesions gradually resolved. Bernard et al. described
two IL2 patients with brief episodes of focal neurologic deficits. Brain imaging was normal in both.
Vecht et al.I6 reported a melanoma patient who developed a rapidly fatal encephalopathy with blindness, ataxia, and bilateral Babinski signs during a n
initial course of IL2. A CT was normal; MRI was not
reported. At autopsy, the gray matter was normal
but the white matter contained multiple small foci of
perivascular demyelination. Because of the pathologic resemblance to acute perivenous encephalomyelitis, they postulated that the lesions were autoimmunologically mediated.
IL2 penetrates the blood-brain barrier. The CSF
IL2 concentration averages 50% of serum leveL2-'
Specific cerebral IL2 binding sites are present in rat
brain with highest density in the hippocampus. Similar receptors are likely present in human
IL2 can alter adrenergic, dopaminergic, and cholinergic neurotransmission.'La.L7.30
'
Direct neuromodulatory effects mediated by endogenous IL2 receptors
may underlie the acute delirium frequently associated with IL2 the rap^,^^;^^ but are unlikely to have
caused structural lesions.
IL2 can both promote the proliferation and matu-

ration of oligodendrocyte~~l
and damage m ~ e l i n . ~ References
~*~~
Demyelination was the primary pathology in the
1. Rosenberg SA, Lotze MT, Yang JC, e t al. Experience with the
use of high-dose interleukin-2 in the treatment of 652 cancer
case of fatal IL2 encephalopathy reported by Vecht
patients. Ann Surg 1989;210:474-485.
et a1.I6There was, however, no clear evidence of my2. Vlasveld LT, Rankin EM. Recombinant interleukin-2 in canelin breakdown in our patients. The lesions affected
cer: basic and clinical aspects. Cancer Treat Rev 1994;20:275311.
both gray and white matter, and the CSF protein,
3. Oldham RK. Therapy with interleukin-2 and tumor-derived
which is often elevated in demyelinating diseases,
activated lymphocytes. Immunol Ser 1994;61:251-271.
was normal in six of the eight patients.
4. Klimas N, Patarca R, Walling J , et al. Clinical and immunoIL2 causes a systemic capillary leak syndrome.34
logical changes in AIDS patients following adoptive therapy
with activated autologous CD8 T cells and interleukin-2 infuAlthough Elison et al.32reported increased cerebral
sion. AIDS 1994;8:1073-1081.
vascular permeability in cats, others have not found
5 . Kovacs JA, Baseler M, Dewar RJ, e t al. Increases in CD4 T
blood-brain barrier disruption with IL2 i n f u ~ i o n . ~ ~ , ~lymphocytes
~
with intermittent courses of interleukin-2 in patients with human immunodeficiency virus infection. N Engl
The absence of gadolinium enhancement at the time
J Med 1995;332:567-575.
of disease onset in our patients argues against blood6. Bukowski RM, Young J , Goodman G, e t al. Polyethylene glycol
brain barrier dysfunction as a precipitating factor in
conjugated interleukin-2: clinical and immunologic effects in
the illness.
patients with advanced renal cell carcinoma. Invest New
Drugs 1993;11:211-217.
The brain lesions in our patients were not typical
7. Sculier J P , Bron D, Verboven N, Klastersky J . Multiple organ
of arterial strokes. Multiple arterial territories were
failure during interleukin-2 administration and LAK cells inaffected simultaneously and lesions resolved quickly.
fusion. Intensive Care Med 1988;14:666-667.
The subacute evolution of lesions, absence of hemor8. Triozzi PL, Kinney P, Rinehart JJ. Central nervous system
toxicity of biological response modifiers. Ann NY Acad Sci
rhage, and lack of an embolic source make it unlikely
1990;594:347-354.
that the lesions were caused by emboli.
9. Denicoff KD, Rubinow DR, Papa MZ, et al. The neuropsychiIL2 induces IL1 and tumor necrosis factor (TNF)
atric effects of treatment with interleukin-2 and lymphokinewhich affect endothelial cell function and coagulaactivated killer cells. Ann Intern Med 1987;107:293-300.
10. Forman AD. Neurologic complications of cytokine therapy.
t i ~ n . ~ , ~ TNF
~ , " , alone
" ~ can damage axons and myelin
Oncology 1994;8:105-110.
in vitro and can cause a transient aphasia and focal
11. Bernard JT, Ameriso S, Kempf RA, Rosen P, Mitchell MS,
neurologic deficits in patient^.'^^:'^.^^ IL2 may, thereFisher M. Transient focal neurologic deficits complicating
interleukin-2 therapy. Neurology 1990;40:154-155.
fore, cause neurologic dysfunction by inducing other
12. Philip T, Mercatello A, Negrier S, et al. Interleukin-2 with
cytokines.
and without LAK cells in metastatic renal cell carcinoma: the
IL2 can prolong the prothrombin time (PT) and
Lyon first-year experience in 20 patients. Cancer Treat Rev
partial thromboplastin time (PTT) and decrease co1 9 8 9 ; 1 6 ( ~ ~ pA):91-104.
pl
13. Loh FL, Herskovitz S, Berger AR, Swerdlow ML. Brachial
agulation factor^,^^.^' but the PT and PTT were usuplexopathy associated with interleukin-2 therapy. Neurology
ally normal in our patients. More extensive monitor1992;42:462-463.
ing of coagulation at one of our institutions showed
14. Heys SD, Mills KLG, Eremin 0. Bilateral carpal tunnel synan increase in von Willebrand's factor, but no abnordrome associated with interleukin-2 therapy. Postgrad Med J
1992;68:587-588.
malities in protein C, protein S, antithrombin 111,
15. Somers SS, Reynolds JV,Guillou PJ. Multifocal neurotoxicity
fibrinogen, PT, or PTT in patients receiving IL2.
during interleukin-2 therapy for malignant melanoma. Clin
The pattern of illness and distribution of lesions
Oncol (R Coll Radiol) 1992;4:135-136.
16. Vecht CJ, Keohane C, Menon RS, Henzen-Logmans SC, Punt
associated with IL2 are similar to those complicating
CJA, Stoter G. Acute fatal leukoencephalopathy after
cyclosporine therapy and occasionally seen with cheinterleukin-2 therapy. N Engl J Med 1990;323:1146-1147.
motherapy, alpha-IFN, hypertensive encephalopa17. Wood R, Montoya JG, Kundu SK, Schwartz DH, Merigan
thy, and e c l a m p ~ i a .The
~ ~ -etiology
~~
of neurologic disTC. Safety and efficacy of polyethylene glycol-modified
interleukin-2 and zidovudine in human immunodeficiency viease in these circumstances is unknown, but has
rus type 1 infection: a phase I/II study. J Infect Dis 1993;167:
been attributed to direct toxic effects of medications,
519-525.
cortical venous thrombosis, or a local vasculopathy
18. Teppler H, Kaplan G, Smith KA, Montana AL, Meyn P, Cohn
ZA. Prolonged immunostimulatory effect of low-dose polyethwith fluid extravasation. Cytokine activation is also
ylene glycol interleukin 2 in patients with human immunodelikely in all these circumstances.
ficiency virus type 1 infection. J Exp Med 1993;177:483-492.
In summary, we describe eight patients who de19. Weaver DF, Heffernan LP, Purdy RA, Ing VW. Eosinophilveloped cortical and subcortical neurologic signs and
induced neurotoxicity: axonal neuropathy, cerebral infarction,
and dementia. Neurology 1988;38:144-146.
cerebral lesions during treatment with IL2-based im20. Kaplan PW, Waterbury L, Kawas C, Bolla-Wilson K, Durack
munotherapy. The etiology of this illness is not
D. Reversible dementia with idiopathic hypereosinophilic synknown, but the similarity to neurologic complications
drome. Neurology 1989;39:1388-1391.
of cyclosporine, chemotherapy, and other immuno21. Lazzarino LG, Nicolai A, Cavallaro T, Rizzuto N. Neurotoxicity of idiopathic hypereosinophilia on a case with simultatherapy suggests a common pathogenesis.
neous CNS and PNS involvement. Ital J Neurol Sci 1993;14:
The occurrence of new neurologic signs and le259-262.
sions in patients receiving IL2 should not be pre22. Moore PM, Harley JB, Fauci AS. Neurologic dysfunction in
the idiopathic hypereosinophilic syndrome. Ann Intern Med
sumed t o be progression of the underlying disease.
1985;102:109-1 14.
Since the neurologic illness associated with IL2 may
23. Meyers CA, Yung WKA. Delayed neurotoxicity of intraventricbe completely reversible, further IL2 infusion should
ular interleukin-2: a case report. J Neurooncol 1993;15:265be stopped and supportive care provided to maximize
267.
24. Saris SC, Rosenberg SA, Friedman RB, Rubin JT, Barba D,
neurologic recovery.
August 1996 NEUROLOGY 47 423

Oldfield EH. Penetration of recombinant interleukin-2 across
the blood-cerebrospinal fluid barrier. J Neurosurg 1988;69:
29-34.
25. Araujo DM, Lapchak PA, Collier B, Quirion R. Localization of
interleukin-2 immunoreactivity and interleukin-2 receptors in
r a t brain: interaction with the cholinergic system. Brain Res
1989;498:257-266.
26. Luber-Narod J , Rogers J . Immune system associated antigens
expressed by cells of the human central nervous system. Neurosci Lett 1988;94:17-22.
27. DeSarro G, Nistico G. Effects of pertussis toxin, dibutyrylcyclic-AMP, bromo-cyclic-AMP and forskolin on the behavioural and electrocortical power spectrum changes induced by
microinfusion of interleukin-2 into the locus coeruleus. Int
J Neurosci 1991;59:67-79.
28. Zalcman S, Green-Johnson JM, Murray L, e t al. Cytokinespecific central monoamine alterations induced by
interleukin-1, -2, and -6. Brain Res 1994;643:40-49.
29. Hanisch U, Set0 D, Quirion R. Modulation of hippocampal
acetylcholine release: a potent central action of interleukin-2.
J Neurosci 1993;13:3368-3374.
30. Nemni R, Iannaccone S, Quattrini A, et al. Effect of chronic
treatment with recombinant interleukin-2 on the central nervous system of adult and old mice. Brain Res 1992;591:248252.
31. Benveniste EN, Merrill J E . Stimulation of oligodendroglial
proliferation and maturation by interleukin-2. Nature 1986;
3211610-613.
32. Ellison MD, Povlishock JT, Merchant RE. Blood-brain barrier
dysfunction in cats following recombinant interleukin-2 infusion. Cancer Res 1987;47:5765-5770.
33. Eitan S, Schwartz M. A transglutaminase that converts
interleukin-2 into a factor cytotoxic to oligodendrocytes. Science 1993;261:106-108.
34. Cotran RS, Pober JS, Gimbrone MA, et al. Endothelial activation during interleukin 2 immunotherapy: a possible mechanism for the vascular leak syndrome. J Immunol 1988;140:
1883-1888.
35. Kobiler D, Lustig S,Gozes Y, Ben-Nathan D, Akov Y. Sodium
dodecylsulphate induces a breach i n the blood-brain barrier
and enables a West Nile virus variant to penetrate into mouse
brain. Brain Res 1989;496:314-316.
36. Banks WA, Kastin AJ. The interleukins-la, - l b , and -2 do not
acutely disrupt the murine blood-brain barrier. Int J Immunopharmacol 1992;14:629-636.
37. Ellison MD, Merchant RE. Appearance of cytokine-associated
central nervous system myelin damage coincides temporally
with serum tumor necrosis factor induction after recombinant
interleukin-2 infusion i l i rxts. J Neuroimmunol 1991;33:245251.
38. Silverman P, Goldsmith GH, Spitzer TR, Rehmus EH, Berger

424

NEUROLOGY 47 August 1996

NA. Effect of tumor necrosis factor on the human fibrinolytic
system. J Clin Oncol 1990;8:468-475.
39. Sherman ML, Spriggs DR, Arthur KA, Imamura K, Frei E,
Kufe DW. Recombinant human tumor necrosis factor administered as a five-day continuous infusion in cancer patients:
phase I toxicity and effects on lipid metabolism. J Clin Oncol
1988;6:344-350.
40. Siege1 J P , P u n RK. Interleukin-2 toxicity. J Clin Oncol 1991;
91694-704.
41. Oleksowicz L, Strack M, Dutcher JP, e t al. A distinct coagulopathy associated with interleukin-2 therapy. Br J Haematol
1994;88:892-894.
42. Byrd RL, Rohrbaugh TM, Raney RB, Norris DG. Transient
cortical blindness secondary to vincristine therapy in childhood malignancies. Cancer 1981;47:37-40.
43. Rippe DJ, Edwards MK, Schrodt JF, Bognanno JR, DAmour
PG, Boyko OB. Reversible cerebral lesions associated with
tiazofurin usage: MR demonstration. J Comput Assist Tomogr
1988;12:1078-1081.
44. Ghalie R, Fitzsimmons WE, Bennett D, Kaizer H. Cortical
blindness: a rare complication of cyclosporine therapy. Bone
Marrow Transplant 1990;6:147-149.
45. Scheinman SJ, Reinitz ER, Petro G, Schwartz RA, Szmalc FS.
Cyclosporine central neurotoxicity following renal transplantation. Transplantation 1990;49:215-216.
46. Patronas NJ, Argyropoulou M. Intravascular thrombosis a s a
possible cause of transient cortical brain lesions: CT and MRI.
J Comput Assist Tomogr 1992;16:849-855.
47. Liehowitz HA, Hall PE. Cortical blindness as a complication of
eclampsia. Ann Emerg Med 1984;13:365-367.
48. Mama TR, Shah M, Mikus MA. Transient cortical blindness
due to hypertensive encephalopathy: magnetic resonance imaging correlation. J Clin Neuroophthalmol 1993;13:35-37.
49. Merimsky 0, Nisipeanu P, Loewenstein A, Reider-Groswasser
I, Chaitchik S. Interferon-related cortical blindness. Cancer
Chemother Pharmacol 1992;29:329-330.
50. VanGelder T, Geurs P, Kho GS, Dippel DWJ, Vecht CJ, Splinter TAW. Cortical blindness and seizures following cisplatin
treatment: both of epileptic origin? Eur J Cancer 1993;29A:
1497-1498.
51. Highley M, Meller ST, Pinkerton CR. Seizures and cortical
dysfunction following high-dose cisplatin administration in
children. Med Pediatr Oncol 1992;20:143-148.
52. Hauser RA, Lacey DM, Knight MR. Hypertensive encephalopathy: magnetic resonance imaging demonstration of reversible
cortical and white matter lesions. Arch Neurol 1988;45:10781083.
53. Heran F, Defer G, Brugieres P, Brenot F, Gaston A, Degos
J-D. Corticai blicdness during chemotherapy: clinical, CT,
and MR corre1atii;:ia J Comput Assist Tomogr 1990;14:262266.

Multiple cerebral lesions complicating therapy with interleukin-2
B. Illowsk Karp, J. C. Yang, M. Khorsand, et al.
Neurology 1996;47;417-424
DOI 10.1212/WNL.47.2.417
This information is current as of August 1, 1996
Updated Information &
Services

including high resolution figures, can be found at:
http://n.neurology.org/content/47/2/417.full.html

References

This article cites 52 articles, 3 of which you can access for free at:
http://n.neurology.org/content/47/2/417.full.html##ref-list-1

Citations

This article has been cited by 4 HighWire-hosted articles:
http://n.neurology.org/content/47/2/417.full.html##otherarticles

Permissions & Licensing

Information about reproducing this article in parts (figures,tables)
or in its entirety can be found online at:
http://n.neurology.org/misc/about.xhtml#permissions

Reprints

Information about ordering reprints can be found online:
http://n.neurology.org/misc/addir.xhtml#reprintsus

Neurology ® is the official journal of the American Academy of Neurology. Published continuously
since 1951, it is now a weekly with 48 issues per year. Copyright Copyright 1996 by Advanstar
Communications Inc.. All rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.