Journal of Neuroscience Methods 119 (2002) 45 /50

www.elsevier.com/locate/jneumeth

A microdialysis method for the recovery of IL-1b, IL-6 and nerve

growth factor from human brain in vivo
Craig D. Winter a,*, Fausto Iannotti a,b, Ashley K. Pringle b, Christos Trikkas c,
Geraldine F. Clough c, Martin K. Church c
a
Neurosurgical Department, Wessex Neurological Centre, Southampton General Hospital, Southampton SO16 6YD, UK
b
Clinical Neurosciences, School of Medicine, University of Southampton, Biomedical Sciences Building, Southampton SO16 7PX, UK
c
Division of Infection, Inflammation and Repair, School of Medicine, University of Southampton, Southampton General Hospital,
Southampton SO16 6YD, UK

Received 24 April 2002; received in revised form 4 June 2002; accepted 5 June 2002

Abstract

Intracerebral microdialysis is used extensively as a research tool in the investigation of the neurochemical and metabolic changes
that occur following acute brain injury. Microdialysis has enabled elucidation of intra-cerebral levels of substances such as lactate,
pyruvate and glycerol but, as yet, has not been used effectively to recover macromolecules from the human brain. Traumatic brain
injury is known to result in the generation of cytokines and neurotrophins into extracellular fluid compartment of the brain, with
effects on neuronal damage and repair. We have developed a technique of in vivo sampling of the interstitial fluid of the brain of
patients with severe head injuries which has allowed the measurement of IL-1b, IL-6 and nerve growth factor. This report confirms
the safety and effectiveness of this modified microdialysis method in the clinical setting of a neurological intensive care unit. The
technique provides a timely addition to the armamentarium of the clinical scientist and will potentially lead to a greater
understanding of neuroinflammation following acute traumatic brain injury. # 2002 Elsevier Science B.V. All rights reserved.

Keywords: Microdialysis; Human; Traumatic brain injury (TBI); Extracellular fluid (ECF); Interleukin-1b; Interleukin-6; Nerve growth factor

1. Introduction The subsequent secondary damage results from a series

Trauma is the leading cause of death in people under
45 years of age worldwide and within this group, up to
half of the fatalities are caused by head injury (Gennar-
elli, 1993; Jennett, 1998). In order to make progress in
the management of head injuries, there must be a greater
understanding of the molecular, biochemical and patho-
physiological processes that occur in the brain following
the injury.
The primary injury, resulting directly from the
trauma, causes mechanical damage to parenchymal
structures, such as neurons, axons and blood vessels.
* Corresponding author. Tel.:
/44-23-8079-6149; fax:
/44-23-
8070-4183
E-mail address: mkc@soton.ac.uk (C.D. Winter).
0165-0270/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 5 - 0 2 7 0 ( 0 2 ) 0 0 1 5 3 - X
of biochemical and metabolic processes triggered by the
trauma-induced neurochemical disturbances. One of the
mechanisms which affects neuronal survival after trau-
matic brain injury (TBI) is the neuroinflammatory
cascade (Rothwell and Hopkins, 1995; Arvin et al.,
1996; Morganti-Kossman et al., 1997). This cascade
involves the synthesis and secretion of cytokines and
neurotrophins which may cause, directly or indirectly,
either tissue damage or repair.
To date, studies which have attempted to investigate
the relationship between traumatic brain injury and
cytokine levels have been limited to measurements
within the cerebrospinal fluid and serum of patients
(Goodman et al., 1990; Ross et al., 1994; Shohami et al.,
1994; Morganti-Kossman et al., 1997; Csuka et al.,
1999; Hensler et al., 2000). The variable dysfunction of
46 C.D. Winter et al. / Journal of Neuroscience Methods 119 (2002) 45 /50

the blood-brain barrier, which serves as a functional

barrier between the brain and blood compartments, that
can occur after traumatic brain injury leads to inherent
flaws in any correlations made between levels of
cytokines in the brain and the serum (Betz and
Crockard, 1992). Although the cytokine levels in the
cerebrospinal fluid may be similar to those in the
extracellular fluid of the brain, the insertion of a
ventricular access device can be problematic in patients
with diffuse brain swelling secondary to traumatic brain
injury.
An attractive alternative would be to recover cyto-
kines directly from the brain using microdialysis. How-
ever, limitations in the design of microdialysis probes
has meant that research has concentrated on measuring
the generation of biochemical indicators of inflamma-
tion with a small molecular mass, such as lactate and
pyruvate (Hillered and Persson, 1999).
We now report the development of a technique for the
recovering macromolecules from the interstitial fluid of
the brain of patients with severe head injuries. The
construction of a concentric microdialysis probe using a
plasmaphoresis membrane with a molecular mass cut off
of 3000 kDa has allowed the recovery of IL-1, IL-6 and
nerve growth factor (NGF) and propteins in sufficient
concentrations for quantitative analysis.

2. Methods

2.1. Microdialysis apparatus

2.1.1. Microdialysis probe

The microdialysis probe, constructed especially for us
by Metalant (Stockholm, Sweden), comprised a mem-
brane, shaft and two fine-bore ports connected to the
shaft by agarose gel (Fig. 1A). The microdialysis probe
was constructed using an Asahi-Plasmaseparator poly-
ethylene polymer tubular membrane of internal dia-
meter 330 mm and wall thickness 50 mm (Ashai
Plasmaseparator, Diamed Medizitechnik, Cologne, Ger-
many), with a molecular mass cut-off of 3000 kDa and
an effective dialysis length of 12/15 mm. The shaft was
10 cm long and the inflow and outflow tubing were both
15 cm in length. The shaft was concentric in design with Fig. 1. (A) The microdialysis catheter with attached intracerebral
an inner core, for the infusate, and an outer core for housing device/bolt (left of picture). The last 1.5 cm of the catheter is
extraction of the fluid back to the collection point. All formed by the microdialysis membrane. (B) The single stack MAB 20
probes were sterilised before use using ethylene oxide. peristaltic pump attached to the microdialysis probe with the inter-
posed Tygon and Sandopren tubing. (C) A right frontal intrapar-
enchymal microdialysis probe in situ.
2.1.2. Probe introducer
The probe was introduced into the brain, via a twist-
drill hole in the skull (3 mm), using a 25 Gauge Spinal
needle (Yale, Becton Dickinson UK, Oxford, UK), cut
to fit into a 7.5 cm length of Epidural Spinal Catheter
(Portex, Hythe, Kent, UK). Once the two had been withdrawn, leaving the spinal drain to act as the hollow
passed together into the brain, the spinal needle was port for the passage of the probe itself.
 C.D. Winter et al. / Journal of Neuroscience Methods 119 (2002) 45 /50
2.1.3. Intracerebral bolt
To secure the microdialysis probe to the skull, a
stainless steel housing or bolt from a Microsensor Skull
Bolt Kit (Codman, Bracknell, Berkshire, UK) was
screwed into the skull (Fig. 1).
2.1.4. Peristaltic pump
A MAB 20 peristaltic pump (Metalant, Stockholm,
Sweden) was fitted with 0.254 mm internal diameter
Sandopren tubing (Metalant, Stockholm, Sweden) to
provide a constant flow rate of 1 ml/min both into and
out from the microdialysis probe (Fig. 1). This closed
type of circulation was used to prevent either a build up
of fluid in, or an overall loss of fluid from, the brain. All
other tubing was 0.38 mm internal diameter Tygon
tubing (Cole-Palmer Instrument Company, Vernon
Hills, IL). The infusate was sterile normal saline (0.9%
NaCl, Maco Pharma, London, UK).
2.2. Dialysate analysis
Commercial Enzyme Linked ImmunoSorbent Assay
(ELISA) kits were used to assay dialysate samples for
IL-1b and IL-6 (Cytoscreen ELISAs, Biosource Inter-
national, Camarillo, CA), NGF (Promega Corporation,
Madison, WI) and albumin (Universal Biologicals,
Stroud, Glos, UK). Total protein was measured in
both the in vitro and in vivo studies using Coomassie
blue (Coomassie blue-250 reagent, Pierce, Rockford,
IL).
2.3. Experimental protocols
2.3.1. In vitro studies
To investigate their suitability to recover macromole-
cules, three microdialysis probes were immersed in a
bath containing a mixture of known concentrations of
IL-1b, IL-6, NGF and human serum albumin in a 0.9%
NaCl (saline) solution. Prior to immersion of the probe
in the cytokine mixture, the whole system was primed
and the membrane hydrated by perfusion with saline at
a flow rate of 1 ml/min for 1 h. The probe was then
transferred to the cytokine mixture and the first 30 min
sample of dialysate discarded. Dialysate was then
collected on ice over the next 2.5 h and then stored at
/808 for assay. Samples of the bath fluid were taken
before and after the dialysate collection and stored
similarly. Data are expressed as percentage recovery
rates calculated for each cytokine with each membrane.
2.3.2. In vivo studies
The study was granted ethical approval by the
Southampton and South West Local Regional Ethics
Committee (LREC number 168/00) and conducted
within the terms of the Declaration of Helsinki. No
patient was entered into the study protocol without
47
prior signed assent given by the appropriate relative or
next-of-kin. Subjects under the age of 18 were excluded.
The care or management of the patients was unaffected
by involvement in the trial.
The three patients involved in this study were
admitted to the Neurological Intensive Care Unit
(NICU) at the Wessex Neurological Centre, South-
ampton General Hospital with severe traumatic brain
injury (TBI). All had been intubated and ventilated
before transfer from a peripheral hospital and were
either taken directly to theatre for an emergency
operation and then to the NICU or transferred straight
to the NICU.
Prior to insertion into the brain, each microdialysis
probe was primed with sterile saline in an aseptic
manner. In order to maintain sterility, the full length
of tubing was covered with a sterile 80 cm Cathgard
Catheter Contamination Shield (Arrow, Erding, Ger-
many). The peristaltic pump was started at a flow rate of
1 ml/min and left for 1 h to equilibrate. The intracerebral
bolt and probe were inserted into the right frontal region
in patients 1 and 2. As patient 3 had a right sided acute
subdural haematoma and an underlying frontal intra-
cerebral haematoma, the probe was placed into the left
frontal region.
Before inserting the probe, the frontal scalp was
cleaned with antiseptic Videne solution (Adams Health-
care, Leeds, UK) to create a sterile surgical field (Fig. 1).
Lidocaine (1% with adrenaline) was injected into the
intended incision area and a stab incision of approxi-
mately 4 mm was made. Using the Codman hand-held
twist drill, a screw-hole was made in the skull bone. The
surface marking for this point was 1.5 /2.0 cm from the
midline, just behind the hairline. A hole was made in the
dura with either the custom built Codman instrument or
a spinal needle. The intracerebral bolt was then screwed
into the bone followed by the introducer, which was
secured within the bolt, after discarding the spinal
needle. Once the system was primed and the membrane
saturated the probe was passed down the lumen of the
spinal drain to a depth of approximately 2.0 cm into the
brain parenchyma and further secured with sterile
Leukostrips (Beiersdorf AG, Hamburg, Germany). A
container at the head of the bed supported the pump
and an insulated receptacle containing ice which allowed
the collection of the dialysate in the Eppendorf tubes to
occur on-ice.
The microdialysis probe was left in situ for 6 days or a
shorter period if the patient was extubated before then.
For practical clinical reasons, it was not possible to
perform dialysis continuously. Consequently, dialysis
was performed for periods of 3 h, with intervals of 5/9 h
in between. At the beginning of each collection period,
the probes were purged for fluid 30 min before collec-
tion, over the following 2.5 h, of the 150 ml dialysate
sample for cytokine and protein assay.
48 C.D. Winter et al. / Journal of Neuroscience Methods 119 (2002) 45 /50

3. Results

3.1. In vitro experiments

For the estimation of dialysis efficiency in vitro, the

bath concentrations of IL-1b, IL-6 and NGF and
albumin at the beginning of the experiment were 45
pg/ml, 43 pg/ml, 962 pg/ml and 6.07 mg/ml, respectively.
These concentrations did not change throughout the
experiment. The calculated efficiencies of dialysis were
(mean9/S.E.M.) for IL-1b, 27.89/4.1%; for IL-6, 45.29/
8.4%; for NGF, 22.29/7.9%; and for protein, 17.59/
1.5% (all n /3).

3.2. In vivo experiments

Patient 1 had a large left frontal intracerebral

haematoma and bilateral subfrontal contusions. The
probe was inserted approximately 30 h after the head
injury. Three 2.5 h dialysate aliquots of brain extra-
cellular fluid were collected with interim rest periods of
between 8 and 9 h when the pump was switched off. The
patient showed signs of improvement and the probe was
removed immediately prior to extubation on day 3.
Patient 2 sustained a left parietal compound depressed
skull fracture with underlying contusion and extradural
haematoma. A right frontal microdialysis probe was
inserted approximately 24 h after the injury and six 2.5 h
dialysate samples were collected over the next 3 days
with interim rest periods of 6/8 h. The patient
improved and was extubated on day 4.
Patient 3 had a large right-sided acute subdural
haematoma and a right frontal intracerebral haema-
toma. The microdialysis probe was inserted into the left Fig. 2. Levels of IL-1b (open bars), IL-6 (grey bars) and NGF (black
frontal lobe approximately 10 h after the trauma and bars) recovered by microdialysis the interstitial fluid of the brain of
eight 2.5 h dialysate samples were collected over the next patients with severe head injury. Dialysate collections were made for
periods of 3 h with intervals of 5 /9 h in between. The arrows indicate
5 days with interim rest periods of 6 /8 h. The patient
probe insertion and removal.
deteriorated and died on day 5 after the injury.
No complications of infection or bleeding were
sufficient fluid was available. The levels of both of these
encountered with any of the patients.
remained relatively constant throughout the study
periods for all patients. The mean levels (9/S.E.M.) of
3.2.1. Cytokine recovery
total protein recovered by dialysis for patients 1, 2 and 3
The recoveries of IL-1b, IL-6 and NGF from the
were 1649/33 (n /3), 709/21 (n /4) and 949/12 mg/ml
brain are shown in Fig. 2. In general, the levels of IL-1b
(n /6) respectively. The corresponding albumin levels
were some 15 times lower than those of IL-6 and NGF,
were 1309/41, 449/15 and 659/15 mg/ml.
maximum levels being approximately 66, 1000 and 1100
pg/ml, respectively. Peak levels of all cytrokines were
seen within 36 h of the initial trauma. Interestingly, the
two patients who survived had higher peak IL-6 levels 4. Discussion
and lower peak NGF levels than the patient who died.
The IL-1 levels in the dialysates did not reveal any We report an effective method for the recovery of
consistent patterns between the patients. high molecular weight molecules, such as cytokines and
neurotrophin, from the human extra cellular fluid of the
3.2.2. Protein recovery brain in vivo using an intraparenchymal microdialysis
In addition to cytokines, dialysate fluid was also used probe with a membrane of 3000 kDa molecular mass cut
for the assay of total protein and albumin whenever off. We have modified a housing device currently used
 C.D. Winter et al. / Journal of Neuroscience Methods 119 (2002) 45 /50
for intracerebral invasive monitoring to secure safely the
microdialysis probe in the frontal lobe of the brain of
severely head injured patients.
At present, the only microdialysis probes available
commercially for clinical use have molecular mass cut
off of 20 and 100 kDa (Anderson et al., 1995).
Theoretically, the probe with the larger molecular
mass cut off may appear to be satisfactory for the
dialysis of IL-1b, IL-6, NGF and albumin, whose
molecular masses are 17, 26, 13.6 and 66 kDa, respec-
tively. However, practically, the number of pores within
a fibre capable of allowing the passage of such large
molecules is small, resulting in an efficiency of dialysis of
only B/1% (Anderson et al., 1995). Consequently, we
have developed a concentric microdialysis probe using a
plasmaphoresis membrane with a molecular mass cut-
off of 3000 kDa. Used in vitro as a linear probe of 1.5
cm length, this membrane has an efficiency of dialysis
for albumin of around 10% (Schmelz et al., 1997, 1999;
Rukwied et al., 2000; Trikkas et al., 2001). Our findings
of an efficiency of dialysis for albumin of 17.5% suggest
a similar performance of the concentric and linear
probes.
Total protein and albumin levels were also measured
in the in vivo studies. That the levels of both remained
relatively constant may be taken as evidence that the
probes did not lose their dialysis efficiency over periods
of up to 5 days. Interestingly, the levels of total protein
of around 100 mg/ml in dialysis fluid from the brain are
approximately 3-fold lower than those found in dialysis
fluid from the skin using the same probes (Trikkas et al.,
2001, 11443/id). However, the finding that approxi-
mately 70% of the total protein was albumin agrees with
similar observations in the skin (Clough et al., unpub-
lished observations).
All severe head injuries can be considered to consist of
elements of focal parenchymal damage (contusions and
intracerebral haematomas) and diffuse damage (diffuse
axonal injury). For ethical reasons, it was only possible
to insert microdialysis probes into the frontal lobes of
the brain, some distance from the focal brain damage. It
is highly unlikely that the probes are recovering
cytokines generated by the area of focal damage but
rather those arising from secondary diffuse-type damage
occurring throughout the brain if cytokines diffuse a
matter of a few millimetres in biological tissues as do
smaller molecules (Petersen et al., 1997; Westerink and
De Vries, 2001). If this is the case, then, although
appearing macroscopically normal, such tissues are
providing biochemical markers that suggest that they
are damaged.
The possibility that trauma due to the implantation of
the microdialysis probes may be a stimulus for cytokine
generation must be considered. If this were so, then it
would be difficult to interpret the microdialysis data in
terms of head injury. However, we are confident for
49
three reasons that implantation trauma is not giving
artefactual information. First, histological studies have
shown that the insertion of a microdialysis catheter into
rat brain or human subcutaneous adipose tissue does
not result in major oedema or bleeding (Hamberger et
al., 1983; Lonnroth and Smith, 1990) although electron
microscopic studies have indicated some local neuronal
damage and axonal degeneration (Clapp-Lilly et al.,
1999). Furthermore, in the subcutaneous adipose tissue,
the concentration of adenosine */a biochemical marker
of tissue damage */has been found to be low immedi-
ately surrounding a microdialysis catheter (Lonnroth
and Smith, 1990). Second, if significant amounts of
cytokines or neurotrophin were generated following
probe insertion, then their peak levels would be likely
to occur 24/48 h afterwards (Woodroofe et al., 1991).
No such peak was seen. Third, the maximum volume of
irreversibly damaged brain parenchyma surrounding a
microdialysis probe approximates to 1.5 mm3 (Benve-
niste and Diemer, 1987; Landolt et al., 1996) whereas
the catchment volume is approximately 100 times larger
(Benveniste and Huttemeier, 1990). Consequently, cyto-
kine and neurotrophin generation secondary to probe
insertion are likely to contribute minimally to the levels
measured in the dialysate.
In conclusion, we have developed a safe and effective
method of cytokine and neurotrophin recovery from the
extracellular fluid of the brain of patients with severe
head injuries. While changes in levels of these macro-
molecules were observed during the period of dialysis in
all three patients, studies in an increased number of
individuals is necessary before any attempt may be made
at interpreting their relevance to outcome. However, we
believe that the development of this technique is a timely
addition to the armamentarium of the clinical scientist
for the investigation of neuroinflammation following
severe head injury.
Acknowledgements
The authors would like to thank Mr I. Tatlow and
Mr I. Chapman for their significant technical support on
the Neurosurgical Intensive Care Unit, Mr B. Barber for
full financial backing, Mrs I. Geary for help with patient
data collection and Mrs J. Mepham for assistance in the
laboratory. CT was supported by an educational grant
from Unilever.
References
Anderson C, Svensson C, Sjogren F, Andersson T, Wardell K. Human
in vivo microdialysis technique can be used to measure cytokines in
contact reactions. Curr Probl Dermatol 1995;23:121 /30.
50 C.D. Winter et al. / Journal of Neuroscience Methods 119 (2002) 45 /50
Arvin B, Neville LF, Barone FC, Feuerstein GZ. The role of
inflammation and cytokines in brain injury. Neurosci Biobehav
Rev 1996;20:445 /52.
Benveniste H, Diemer NH. Cellular reactions to implantation of a
microdialysis tube in the rat hippocampus. Acta Neuropathol
(Berl) 1987;74:234 /8.
Benveniste H, Huttemeier PC. Microdialysis */theory and application.
Prog Neurobiol 1990;35:195 /215.
Betz A, Crockard A. Brain edema and the blood-brain barrier. In:
Crockard A, Hayward R, Hoff J, editors. Neurosurgery: The
Scientific Basis of Clinical Practice. Cambridge: Blackwell,
1992:353 /73.
Clapp-Lilly KL, Roberts RC, Duffy LK, Irons KP, Hu Y, Drew KL.
An ultrastructural analysis of tissue surrounding a microdialysis
probe. J Neurosci Methods 1999;90:129 /42.
Csuka E, Morganti-Kossmann MC, Lenzlinger PM, Joller H, Trentz
O, Kossmann T. IL-10 levels in cerebrospinal fluid and serum of
patients with severe traumatic brain injury: relationship to IL-6,
TNF-alpha, TGF-beta1 and blood-brain barrier function. J
Neuroimmunol 1999;101:211 /21.
Gennarelli TA. Mechanisms of brain injury. J Emerg Med
1993;11(Suppl 1):5 /11.
Goodman JC, Robertson CS, Grossman RG, Narayan RK. Elevation
of tumor necrosis factor in head injury. J Neuroimmunol
1990;30:213 /7.
Hamberger A, Nystrom B, Berthold CH, Karlsson B, Lehmann A.
Extracellular GABA, glutamate and glutamine in vivo */perfusion-
dialysis of the rabbit hippocampus. In: Hertz L, Kvamme E,
McGeer EG, Schousboe A, editors. Glutamine, Glutamate and
GABA in the Central Nervous System. New York: Alan R Liss,
1983:473 /92.
Hensler T, Sauerland S, Riess P, Hess S, Helling HJ, Andermahr J,
Bouillon B, Neugebauer EA. The effect of additional brain injury
on systemic interleukin (IL)-10 and IL-13 levels in trauma patients.
Inflamm Res 2000;49:524 /8.
Hillered L, Persson L. Neurometabolic monitoring of the acutely
injured human brain. Scand J Clin Invest 1999;229(Suppl.):9 /18.
Jennett B. Epidemiology of head injury. Arch Dis Child 1998;78:403 /
6.
Landolt H, Langemann H, Alessandri B. A concept for the introduc-
tion of cerebral microdialysis in neurointensive care. Acta Neu-
rochir Suppl (Wien) 1996;67:31 /6.
Lonnroth P, Smith U. Microdialysis */a novel technique for clinical
investigations. J Intern Med 1990;227:295 /300.
Morganti-Kossman MC, Lenzlinger PM, Hans V, Stahel P, Csuka E,
Ammann E, Stocker R, Trentz O, Kossmann T. Production of
cytokines following brain injury: beneficial and deleterious for the
damaged tissue. Mol Psychiatry 1997;2:133 /6.
Petersen LJ, Church MK, Skov PS. Histamine is released in the weal
but not the flare following challenge of human skin in vivo: a
microdialysis study. Clin Exp Allergy 1997;27:284 /96.
Ross SA, Halliday MI, Campbell GC, Byrnes DP, Rowlands BJ. The
presence of tumour necrosis factor in CSF and plasma after severe
head injury. Br J Neurosurg 1994;8:419 /25.
Rothwell NJ, Hopkins SJ. Cytokines and the nervous system II:
actions and mechanisms of action. Trends Neurosci 1995;18:130 /
6.
Rukwied R, Lischetzki G, McGlone F, Heyer G, Schmelz M. Mast cell
mediators other than histamine induce pruritus in atopic dermatitis
patients: a dermal microdialysis study [see comments]. Br J
Dermatol 2000;142:1114 /20.
Schmelz M, Luz O, Averbeck B, Bickel A. Plasma extravasation and
neuropeptide release in human skin as measured by intradermal
microdialysis. Neurosci Lett 1997;230:117 /20.
Schmelz M, Zeck S, Raithel M, Rukwied R. Mast cell tryptase in
dermal neurogenic inflammation. Clin Exp Allergy 1999;29:695 /
702.
Shohami E, Novikov M, Bass R, Yamin A, Gallily R. Closed head
injury triggers early production of TNF alpha and IL-6 by brain
tissue. J Cereb Blood Flow Metab 1994;14:615 /9.
Trikkas C, Church MK, Clough GF. Measurement of protein
extravasation in human skin in vivo using microdialysis. J Vasc
Res 2001;38:199.
Westerink BH, De Vries JB. A method to evaluate the diffusion rate of
drugs from a microdialysis probe through brain tissue. J Neurosci
Methods 2001;109:53 /8.
Woodroofe MN, Sarna GS, Wadhwa M, Hayes GM, Loughlin AJ,
Tinker A, Cuzner ML. Detection of interleukin-1 and interleukin-6
in adult rat brain, following mechanical injury, by in vivo
microdialysis: evidence of a role for microglia in cytokine produc-
tion. J Neuroimmunol 1991;33:227 /36.