AJR 2000; 175:443-446
© American Roentgen Ray Society
CT-Guided Percutaneous Fibrin Glue Therapy of Cerebrospinal Fluid Leaks in the Spine After Surgery
Mahesh R. Patel1,2,
Paul A. Caruso1,
Naveed Yousuf1 and
Jacob Rachlin3
1
Department of Radiology, Beth Israel Deaconess Medical Center, Harvard Medical
School, Boston, MA 02215.
2
Present address: Santa Clara Valley Medical Center, 3031 Tisch Way, Ste. 02,
Plaza South, San Jose, CA 25128.
3
Department of Surgery, Beth Israel Deaconess Medical Center, Harvard Medical
School, Boston, MA 02215.
Received September 20, 1999;
accepted after revision January 4, 2000.
Address correspondence to M. R. Patel.
Abstract
OBJECTIVE. The objective of this study is to assess CT-guided
percutaneous injection of fibrin glue for the management of cerebrospinal
fluid leaks within the spine.
CONCLUSION. Percutaneous CT-guided placement of fibrin glue can
provide a treatment option for postoperative cerebrospinal fluid leaks,
potentially allowing a major surgical procedure to be avoided. However, the
complication of aseptic meningitis may occasionally result from this
procedure.
Introduction
Cerebrospinal fluid (CSF) leaks involving the spine can occur after surgery
as well as accidental trauma. CSF leaks occur most frequently after back
surgery or, on rare occasions, after a lumbar puncture. Patients with a CSF
leak often present with headaches or a flocculent subcutaneous fluid
collection superficial to the leak with drainage. These leaks represent a
serious problem because of the secondary complication of meningitis. Typical
management consists of surgery with meticulous closure of the dura or closure
using a fascial graft.
Prior studies evaluating CT-guided fibrin glue placement for treatment of
postoperative dural tears have involved only a small number of patients
[1,
2]. We evaluated CT-guided
percutaneous fibrin glue placement to seal postoperative CSF leaks in a larger
group of patients. This therapy has the potential to allow a major surgical
procedure with its associated cost and morbidity to be avoided.
Subjects and Methods
Between June 1994 and February 1999, 23 consecutive patients with a CSF
leak were treated with percutaneous fibrin glue placement. These patients (12
men and 11 women) ranged in age from 26 to 73 years (average age, 49 years),
and they had undergone spinal surgery for either degenerative disk disease
(n = 17) or tumor resection (n = 6). They presented with a
palpable soft-tissue collection, postural headaches, or persistent CSF wound
drainage. Although the first six patients underwent MR imaging within 1 week
before fibrin glue placement to accurately assess the relationship of the dura
to the CSF leak, preoperative MR imaging was not considered essential beyond
this initial experience.
After informed consent was obtained, the patient was placed in the prone
position on a 9800 CT scanner (General Electric Medical Systems, Milwaukee,
WI), and 5-mm contiguous scans (140 kV, 257 mA, 25-cm field of view) were
obtained through the CSF collection. The overlying skin was marked at the site
of suspected leak, and with the patient under local lidocaine anesthesia, an
18- to 20-gauge spinal needle was placed into the CSF fluid collection
adjacent to the location of the suspected dural tear. The maximal amount of
CSF was drained, and repositioning and reaspiration were performed as needed.
The fluid was aspirated slowly to avoid aspiration of a nerve root through the
dural defect. The patient was continuously monitored for associated neurologic
symptoms by the attendant nurse and physician, and the aspiration was titrated
accordingly. The needle was repositioned as needed, usually once or twice, to
maximally drain the fluid. A subcutaneous fluid collection was not drained
unless it was in direct communication with the deeper collection.
The formation of fibrin glue required two components: first,
cryoprecipitate was prepared from the patient's blood
[3] in the absence of
contraindications such as coagulopathy, emergency, or anemia. The preparation
of the autologous cryoprecipitate required at least 3 days. Five hundred
milliliters of whole blood was collected. The resulting plasma was extracted
and frozen at -80°C for 24 hr. The plasma was then thawed gradually over
24 hr at 6°C, at which point a second centrifugation step was performed.
The liquid precipitate was harvested again and was frozen for 24 hr. This
process yielded 20-25 mL of cryoprecipitate, which must be thawed 30-45 min
before use. Once thawed, the cryoprecipitate should be used within 4 hr. The
cryoprecipitate served as a concentrated source of fibrinogen. In conjunction
with thrombin and factor XIII, it polymerized into fibrin
[3]. The second component of
fibrin glue was a thrombin and calcium chloride mixture made by mixing 20,000
U of lyophilized thrombin with 10 mL of a 10% calcium chloride solution. This
combined solution had a concentration of 2000 U/mL of thrombin.
Before the procedure, approximately 0.25 mL of iopamidol was added to 3 mL
of the thrombin-calcium chloride solution to allow improved visualization of
the fibrin plug. This mixture and the cryoprecipitate were placed in separate
3-mL syringe assemblies and then connected via a three-way stopcock to an 18-
to 20-gauge spinal needle (Fig.
1). The spinal needle was first placed into the suspected site of
CSF leak adjacent to the suspected dural rent. Equal volumes of the
cryoprecipitate and thrombin-calcium chloride-iopamidol solution were injected
simultaneously. During the injection, the patient was monitored for any
adverse symptoms including pain and numbness or tingling.
View larger version (101K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1. —Preparation of two-component fibrin glue assembly shows attachment
via two 3-mL syringes to three-way stopcock that is then attached to an 18- to
20-gauge spinal needle. No connecting tubing is used to reduce dead
volume.
|
|
The aggregation of the two solutions resulted in formation of the fibrin
plug in vivo at the site of the suspected dural tear. A new syringe assembly
was used for repeated deployment of fibrin glue because aggregation of fibrin
glue within the stopcock and the syringe may occur as a result of reflux. The
total volume instilled ranged from 4 to 24 mL. The aggregation of the fibrin
glue mixture is illustrated in Figure
2. The procedure was terminated when one of the following
conditions occurred: adequate coverage by the fibrin plug adjacent to the site
of suspected dural leak was achieved; appreciable mass effect higher than or
equal to that seen on imaging studies before the intervention was seen; or the
patient complained of symptoms of radiculopathy or back pain.
View larger version (145K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2. —In vitro aggregation of fibrin glue within superior portion of tube
5 min after admixture. Note gravity-dependent air—fluid level in
container adjacent to tube of fibrin glue (arrows).
|
|
The position of the fibrin plug was documented on postprocedure CT scans.
The success of the procedure was determined by chart review, direct patient
interview, and physical examination. In several patients, follow-up MR images
were obtained.
Results
In 15 (65%) of the 23 patients with a postoperative CSF leak, symptoms
either decreased or resolved markedly after the procedure. In five patients
(22%), the complication of aseptic meningitis occurred, which we have found to
require only analgesic therapy and bed rest.
Of the 23 patients, 20 had undergone surgery in the thoracolumbar spine;
one in the sacrum; one in the cervical spine; and one in the skull base.
Seventeen patients had undergone surgery for degenerative disk disease,
whereas six had undergone surgical resection of tumors. The time between
completion of surgery and placement of the fibrin plug ranged from 3 days to 2
years 7 months. Postoperative symptoms of patients presenting for fibrin patch
included wound leak, soft-tissue collection, and headache. Follow-up from the
time of patch placement for the patients who did not require subsequent
definitive repair ranged from 6 months to 3 years 6 months. One of the
successful lumbar spine fibrin plug procedures is illustrated in Figure
3A,3B,3C.
An image of a patient with a complication of aseptic meningitis from
intrathecal placement of fibrin glue is shown in
Figure 4. A skull base leak
that was successfully treated with fibrin glue is shown in Figure
5A,5B,5C.
View larger version (131K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. —39-year-old man with lumbar cerebrospinal fluid (CSF) leak detected
4 months after L5—S1 disk excision who later underwent successful fibrin
glue patch therapy. Axial CT scan shows large posterior CSF collection.
|
|
View larger version (125K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. —39-year-old man with lumbar cerebrospinal fluid (CSF) leak detected
4 months after L5—S1 disk excision who later underwent successful fibrin
glue patch therapy. Axial CT scan after fibrin glue therapy shows placement of
plug over suspected site of CSF leak. Note different attenuation of fibrin
glue components due to different dilutions of contrast material: fibrin plug
(arrows) appears mildly hyperdense to CSF.
|
|
View larger version (152K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C. —39-year-old man with lumbar cerebrospinal fluid (CSF) leak detected
4 months after L5—S1 disk excision who later underwent successful fibrin
glue patch therapy. Sagittal reconstruction shows fibrin glue covering
laminectomy defect. Note extension of fibrin glue (arrows)
subdurally. Subdural extension of fibrin glue is common after this
procedure.
|
|
View larger version (99K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4. —63-year-old woman with lumbar cerebrospinal fluid (CSF) leak who
later underwent fibrin glue patch therapy with intrathecal extension of fibrin
glue, resulting in aseptic meningitis. Axial CT image shows fibrin glue within
thecal sac (arrows). Bony fusion material is seen posterior to spine.
Dense radiopaque object within left paraspinal soft tissues represents an
electrical stimulator device. Because of large size of defect, surgical repair
was subsequently performed.
|
|
View larger version (108K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A. —29-year-old man with suboccipital cerebrospinal fluid leak after
hemangioblastoma resection who later underwent successful fibrin glue patch
therapy. Axial CT image shows collection posterior to C2 level.
|
|
View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B. —29-year-old man with suboccipital cerebrospinal fluid leak after
hemangioblastoma resection who later underwent successful fibrin glue patch
therapy. Axial CT image shows spinal needle placed into collection with
introduction of fibrin glue.
|
|
View larger version (71K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C. —29-year-old man with suboccipital cerebrospinal fluid leak after
hemangioblastoma resection who later underwent successful fibrin glue patch
therapy. Sagittal reconstruction shows fibrin plug covers site of craniotomy
defect.
|
|
Discussion
A simple fibrin plug procedure for the management of postoperative CSF
leaks may help patients avoid a major surgical procedure with its associated
costs and morbidity. We have shown its utility in sealing leaks in the
lumbosacral spine and in the skull base; this procedure is most frequently
used in the setting of a postoperative CSF leak. Our current results, a
success rate of 65%, as opposed to the previously reported success rate of 50%
[2] suggest that with
experience and appropriate patient selection, this technique provides a good
alternative to surgical therapy.
The surgical management of postoperative CSF leaks entails providing an
adequate seal that is able to withstand CSF pressure during the healing
period. Prompt surgery has been advocated to prevent the complication of
meningitis, CSF fistulas, and pseudocyst formation with potential resultant
neural compression. Surgical methods include meticulous primary closure of the
defect and potential interposition of fascia, muscle, or fat graft over the
defect.
More recently, direct placement of fibrin glue over the defect has been
performed, with or without associated graft placement. Placement of a fibrin
plug has been postulated to create adhesion at the site of a dural tear, and
this adhesion may also promote healing
[3]. The maximal bonding effect
is reached within 30-90 min, with 70% of the bonding effect occurring within 2
min of the mixing of the two adhesive components. Subsequent granulation
tissue and fibrosis are believed to result in a more definitive repair
[4]. It has further been
proposed that the fibrin adhesive is completely absorbed so that the process
of wound healing remains undisturbed and that wound healing is promoted by the
immediate stimulation of fibroblasts
[3].
A risk of the procedure is that a small amount of the fibrin glue may be
introduced intrathecally and may result in aseptic meningitis. In our
experience, aseptic meningitis is a self-limiting complication that may occur
in the setting of a CSF leak with or without a fibrin glue patch; this
complication resolves in 2-3 days and can be treated with conservative
therapy. Bed rest and analgesics are considered appropriate. We believe that
antibiotic therapy is not indicated. In rare cases, a large dural tear may be
a relative contraindication to this technique because it may result in free
introduction of fibrin glue into the subarachnoid space. This contraindication
can be determined by injecting myelographic contrast material into the
collection to see whether free communication with the thecal sac is present.
Although the fibrin glue may serve as a nidus for infection, this finding was
not observed in our patients. Also in follow-up to date, neither arachnoiditis
nor fibrous adhesions have occurred.
Originally, we performed MR imaging before the procedure to locate the
potential site of the CSF leak and to optimally delineate the thecal sac from
the extradural CSF collection. This additional imaging may also prevent the
inadvertent introduction of fibrin glue into the subarachnoid space. Although
CT myelography may also define the dural tear, we did not think it was
necessary before performing the fibrin glue procedure. Having gained more
experience with the technique, we no longer routinely perform imaging besides
CT at the time of fibrin glue placement.
We also recommend that autologous cryoprecipitate be used to avoid the
risks of blood-born pathogens, including hepatitis C. The blood needs to be
harvested several days before the procedure to allow preparation of the
cryoprecipitate as described. After the procedure, we generally recommend a
period of overnight absolute bed rest.
Other alternative procedures to surgical closure have been proposed as
therapy for CSF leaks, including epidural blood patches
[5,6,7]
and persistent lumbosacral drainage
[8]. These alternative
therapies can also have complications. Epidural blood patches can result in
persistent headache, neurologic deterioration, and subdural hematoma
[9,
10]. The intrathecal
administration of RBCs can cause severe headaches. According to one large
series, the success rate of lumbar subarachnoid catheterization and drainage
was 94%; however, complications can occur such as infections (5%), including
meningitis (2%); overdrainage with temporary neurologic decline (3%);
occlusion of the drain requiring replacement (9%); and transient nerve root
irritation (14%) [11].
Persistent CSF leak, coma, and injury to the soft tissues as well as to the
nerve roots have also been reported
[12]. The only complication of
the fibrin patch has been self-limiting aseptic meningitis. Furthermore, the
lumbar catheter technique requires a 14- to 16-gauge tear in the dura above
the site of the leak, and this additional rent may not heal spontaneously.
Also, the subsequent continuous bed rest for at least 3 days necessitates a
prolonged hospital stay and results in a predisposition to infection.
In summary, CT-guided fibrin glue therapy has significant potential to
lower the morbidity and cost of treating patients with postoperative CSF
fistulas. Given our experience to date, we advocate percutaneous CT-guided
fibrin glue therapy as an initial treatment for a CSF leak.
References
-
Fraioli B, Pastore FS, Floris R, et al. Computed tomography-guided
transsphenoidal closure of post-surgical cerebrospinal fluid fistula: a
transmucosal needle technique. Surg Neurol
1997;48:409
-413[Medline]
-
Patel MR, Louie W, Rachlin J. Postoperative cerebrospinal fluid
leaks of the lumbosacral spine: management with percutaneous fibrin glue.
AJNR
1996;17:495
-500[Abstract]
-
Sidentop KH, Harris DM, Ham K, Sanchez B. Extended experimental and
preliminary surgical findings with autologous fibrin tissue adhesive made from
the patient's own blood. Laryngoscope
1986;96:1062
-1064[Medline]
-
Bundschuh CV. Imaging of the postoperative lumbar spine.
Neuroimaging Clin N Am
1993;3:499
-516
-
Chauchan C, Frances GA, Kemeny AA. The avoidance of surgery in the
treatment of subarachnoid cutaneous fistula by the use of an epidural patch:
technical case report. Neurosurgery
1995;36:612
-613[Medline]
-
Di Giovani AJ, Galbert MW, Wahle WM. Epidural injection of
autologous blood for post-lumbar headache. Anesth
Analg 1972;51:226
-232[Free Full Text]
-
Maycock NF, van Essen J, Pfitzner J. Post-laminectomy cerebrospinal
fluid fistula treated with an epidural blood patch.
Spine
1994;19:2223
-2225[Medline]
-
Findler G, Sahar A, Beller AJ. Continuous lumbar drainage of
cerebrospinal fluid in neurosurgical patients. Surg
Neurol 1977;8:455
-457[Medline]
-
Seeberger MD, Urwyler A. Lumbosacral syndrome after extradural
blood patch. Br J Anaesth
1992;69:414
-416[Abstract/Free Full Text]
-
Woodward WM, Levy DM, Dixon AM. Exacerbation of post-dural puncture
headache after epidural blood patch. Can J Anaesth
1994;41:628
-631[Abstract/Free Full Text]
-
Shapiro SA, Scully T. Closed continuous drainage of cerebrospinal
fluid via a lumbar subarachnoid catheter for treatment or prevention of
cranial/spinal cerebrospinal fluid fistula.
Neurosurgery
1992;30:241
-245[Medline]
-
Francel PC, Persing JA, Cantrell RW, Levine PA, Newman SA.
Neurological deterioration after cerebrospinal fluid drainage. J
Craniofac Surg 1992;3:145
-148[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
D. K. Turnbull and D. B. Shepherd
Post-dural puncture headache: pathogenesis, prevention and treatment
Br. J. Anaesth.,
November 1, 2003;
91(5):
718 - 729.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
