- Email: [email protected]
MR imaging of cerebrospinal fluid rhinorrhea following the suboccipital approach to the cerebellopontine angle and the internal auditory canal: Report of two cases
ELSEVIER
MR IMAGING
OF CEREBROSPINAL FLUID RHINORRHEA FOLLOWING THE SUBOCCIPITAL APPROACH TO THE CEREBELLOPONTINE ANGLE AND THE INTERNAL AUDITORY CANAL: REPORT OF Two CASES Masanori Kabuto, M.D., Toshihiko Kubota, M.D., Hidenori Kobayashi, M.D., Yuji Handa, M.D., Akira Tuchida, M.D., and Hiroaki Takeuchi, M.D. ~epu~ent
of ~eurosu~e~,
Fukui medicos echoes, F~kui, Jupu~
Kabuto M, Kubota T, Kobayashi H, Handa Y, Tuchida A, Takeuchi H. MR imaging of cerebrospinaffluid rhinorrhea following the suboccipital approach to the cerebellopontine angle and the internal auditory canal: report of two cases.Surg Meuroll~,~~3~-~. BACKGROUND
Cerebrospinal fluid (CSF) otorhinorrhea is one of the most common postoperative complications following the suboccipital approach to the cerebellopontine angle and the internal auditory canal. Accurate preoperative detection of the site of CSF leakage is im~rtant because inaccuracy may require a more extensive exploratory surgical procedure in the repair operation. There are few reports on evaluation of magnetic resonance (MR) imaging in diagnosing CSF leakage. CASE DESCRIPTION MR imaging of two cases of postoperative
CSF rhinorrhea is reported. A 53-yearold woman and a 29-yearold man underwent suboccipital operations for microvascular decompression of the facial nerve in hemifacial spasm and for removal of an acoustic schwannoma, respectively. In both cases, MR findings were useful in preoperatively delineating the site of the CSF leakage, which was confirmed during the repair operation. CONCLUSION
MR imaging may be useful in identifying the site of CSF leakage following the suboccipital approach to the cerebellopontine angle. KEY WORDS
Acoustic sc~~nno~~, cere~~sp~nul fluid r~ino~~ea, magi net&z resonance, su~~cipita~ approach, tempoml bone air cells.
fluid (CSF) otorhinorrhea is one of the most common postoperative complications following the suboc~ipit~ approach to the erebrospinaf
c
Address reprint requests to: Masanori Kabuto, M.D., Department Neurosurgery, Fukui Medical School, Matsuoka, Fukui 910-11, Japan. Received February 8, 1995; accepted May 25, 1995. 009~3019/96/$15.00 SSDI ~9~3019(9~002~
of
cerebellopontine angle and the internal auditory canal for acoustic schwannomas and other lesions [6]. It is generally believed that there are two major sites of CSFleakage following the operation. One is the mastoid air cells lateral to the sigmoid sinus, which may be opened during a suboccipital craniectomy [1,3,6,8]. The other site is the posteromedial air cell tract [1,3,4] extending posterior to the bony labyrinth and into the posterior wall of the internal auditory canal which may be opened when drilling the posterior wall of the internal auditory canal for removal of an intracanalicular part of an acoustic schwannoma [3,8]. However, when postoperative CSFotorhinorrhea occurs after the operations described above, it is not easy to accurately identify
the location
of the CSF leakage preopera-
tively. Magnetic resonance (MR) imaging may be useful for determining
the CSF leakage
route
be-
cause T,-weighted MR images can demonstrate fluids such as CSF in the temporal bone air cells. However, to our knowledge, there are few reports on MR findings of CSF leakage [9]. This report describes two types of CSF rhinorrhea following the suboccipital approach to the cerebellopontine angle, in which MR imaging clearly verified the route of the CSF rhinorrhea.
CASEREPORTS CASE
1
A 53-year-old woman was admitted to our hospital on April 19, 1993 with a l-year history of left progressive facial spasm. A left vertebral angiogram 0 1996 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010
MR Imaging of CSF Rhinorrhea
Surg Nemo1 1996;45:336 -40
337
Axial T,-weighted (2352/80) MR image (A) performed postoperatively demonstrating high-intensity signals corresponding to CSF in the mastoid air cells (asterisk), the middle ear cavity (star), the eustachian
tube (small arrow), and the nasal cavity (double asterisks). The CSF leakage point (large arrow) was identified as an inferomedial part of the mastoid air
cells. Axial plain CT scan with bony windows (B) showing the extent of the craniectomy and the opened air cells. Axial T, weighted MR image (c) performed after repair operation demonstrating no further CSF leakage.
showed dilatation and elongation of the anterior inferior cerebellar artery and elevation of the posterior inferior cerebellar artery. A computed tomographic (CT) scan (GE-CT 9800, GE Medical Systems, Milwaukee, WI) and MR imaging (1.5-T Signa imaging system; GE Medical Systems) showed no tumor and no vascular malformation at the cerebeilopontine angle. On April 27, 1993, a left suboccipital craniectomy was performed for microvascular decompression. The exposed mastoid air cells were sealed with bone wax. The patient noticed a continuous watery discharge into the n~oph~x a few days after the operation, although the facial spasm stopped postoperatively. Continuous spinal drainage of CSF for 10 days failed to stop the CSF leakage. T,-weighted MR images demonstrated high-intensity signals corresponding to CSF in the mastoid air cells, the middle ear cavity, the eustachian tube, and the nasal cavity (Figure 1 A). The CSF leakage point was also identified clearly as an inferomedial part of the mastoid air cells (Figure
1 A) where a plain CT scan with bony windows confirmed the opened air cells (Figure 1 B). On May 20, 1993, a second operation for repairing CSF leakage was performed. Inspection during the operation confirmed the MR findings of CSF leakage described above. The exposed mastoid air cells where sealing bone wax had already been lost were found. This area was packed with muscle fragments and then covered with fibrin glue. It was also found that a part of the dura mater had been loosely sutured and CSF leaked from it. This part of the dura was tightly resutured and covered with fibrin glue. After the second operation, the patient had no further CSF rhinorrhea and her postoperative course was uneventful. MR imaging performed three weeks after the second
operation
showed
no further
CSF leak-
age (Figure 1 C). CASE 2 A 29-year-old man was admitted to our hospital on November 5, 1990, with a half-year history of left
338
Surg Neurol 1996;45:336-40
Kabuto et al
Axial gas CT cisternography an intracanalicular acoustic H eighth nerve. This CT also shows
air cells in the posterior meatal wall were also sealed with bone wax. A histological examination of the surgically obtained specimens revealed an acoustic schwannoma. On the sixth postoperative day, the patient noticed an intermittent watery discharge into the nasopharynx and a developing conductive hearing loss. It was suspected that the CSF was coming from the middle ear via the eustachian tube. continuous spinal drainage of CSFfor 2 weeks failed to stop the CSF leakage. T,-weighted MR images demonstrated high-intensity signals corresponding to CSFin the posterior wall of the internal auditory canal, the mastoid air cells, the middle ear cavity, the eustachian tube, and the nasal cavity (Figures 3 A and B). The CSF leakage point was preoperatively identified as the posterior wall of the internal auditory canal by MR imaging. A residual tumor was also seen. On December 14, 1990, a second operation was performed to remove the residual tumor and to repair CSF leakage. Rxposed air cells from which bone wax had already floated away were observed in the posterior meatal wall. After removing the residual tumor, the bone defect was packed with muscle fragments and then sealed with fibrin glue. After the second operation, the patient had no further CSF rhinorrhea and his postoperative course was uneventful. MR imaging performed 6 weeks after the second operation showed no further CSF leakage (Figure 3 C).
tinnitus and hearing disturbance. On neurological lamination, he had only a moderate sensorineur~ hearing disturbance on the left side, but this hearing was preserved to a certain extent. The CT scan showed an enlargement of the left internal auditory canal but failed to show a tumor. Air cells in the posterior wall of the left internal auditory canal as well as well-developed mastoid air cells were disclosed on the CT scan (Figure 2 A). A gas CT cisternography [2] was performed. This cisternography clearly delineated an intracanalicular acoustic tumor, facial nerve, and 8th nerve (Figure 2 A). An enhanced T,-weighted MR image demonstrated an intracanalicular acoustic tumor (Figure 2 B). On November 16, 1990, a left suboccipital craniectomy was performed and exposed mastoid air cells were sealed with bone wax. The posterior wall of the internal auditory canal was removed by drilling, and the intracanaIicular tumor was exposed and then removed. After removing the tumor, exposed
The incidence of CSF otorhinorrhea following the suboccipital approach to the cerebellopontine angle and the internal auditory canal is thought to vary from 4% to 18% [3,6]. The major sites of postoperative CSFleakage are the group of mastoid air cells lateral to the sigmoid sinus, which are always opened at the lateral margin of the suboccipital craniectomy unless the degree of pneumatization is small [ 1,3,6,8]. These air cells are interconnected and all of them eventually drain into the middle ear cavity connecting with the nasopharynx through the eustachian tube. Case 1 is one of this type. In this case, T,-weighted MR images clearly demonstrated this CSFrhinorrhea route. Another possible site is the posterior wall of the internal auditory canal, of which pneumatization is reported to be observed in about 20% of patients [ 1,3,4,83. This posteromedial air cell tract connects with the mastoid air cells. Case 2 is one of this type. Also in this case, MR imaging accurately depicted the route of the CSF rhinorrhea. Accurate preoperative detection of the site of CSF leakage is very important because inaccuracy may require a more extensive
(A) clearly delineating tumor, facial nerve, and welldeveloped mastoid air cells and pneumatization of the posteromedial air cell tract in the posterior meatal wall (arrow). AxiaI gadolinium-DTPA-enhanced T,-weighted (317/20) MR image (B) showing a homogenously enhanced tumor in the left internal auditory canal.
MR Imaging of CSF Rhinorrhea
Surg Neurol 339 199~45:336~0 rhinorrhea following the suboccipital approach to the cerebellopontine angle and the internal auditory canal. Although MR imaging (especially increased T2weighted signal) is very useful for detection of fluid within the temporal bone air cells and middle ear cavity, it is not specific for the presence of CSF within these cavities. CSF may be distinguished from highly proteinaceous fluid secondary to inflammatory reaction or bleeding because proteinaceous fluid has a higher signal intensity than simple fluid in TV-weighted images 151. However, the signal from simple fluid due to serous otomastoiditis may be identical to the CSFsignal. Therefore, in the presence of preexisting otomastoiditis, it may be impossible to detect CSF by relying only on the signal intensity of MR imaging within the mastoid air cells and middle ear cavity. In patients with clinical symptoms suggesting postoperative CSF rhinorrhea, however, if the continuity of high-intensity signals like CSF from subarachnoidal space to the eustachian tube through the mastoid air cells can be determined from the T,-weighted images as demonstrated in our cases, MR imaging can strongly suggest the presence of CSF within the mastoid air cells as well as suggesting the site of leakage. Positive
‘[‘,-weighted (25~1/80) MR images (A,B) qintensityAxial performed postoperatively demonstrating highsignals corresponding to CSF in the posterior wall of the left internal auditory canal (large arrow), the posteromedial air cell tract (double arrows), the mastoid air cells (asterisk), the middle ear cavity (star), and the eustachian tube (small arrow). The CSF leakage point (large arrow) was identified as the posterior meatal wall. Axial T,-weighted MR image (C) performed after repair operation demonstrating no further CSF leakage.
exploratory surgical procedure in the repair operation, especially if both the mastoid air ceils and the posteromedial air cell tract are opened as in case 2. In both cases here, MR imaging was considered a noninvasive and simple method enabling direct and accurate identification of the route of the CSF oto-
contrast
CT cisternography
remains
the
gold standard for any CSF rhinorrhea or otorhinorrhea. In many cases, where the bone defect is relatively large, however, the combination of MR imaging and CT scan with bony windows may be adequate to determine the location of the CSF leakage in the temporal bone, eliminating the need for CT cisternography. To our knowledge, there is only one report evaluating MR imaging in diagnosing CSF fistulas at the anterior skull base due to a trauma and an encephalomeningocele [ 91. Our report may be of value with respect to MR imaging in diagnosing CSF leakage and-detesting the route. In our cases, opened air cells were first sealed with bone wax. Postoperative CSF rhinorrhea occurred, however, and in the repair operation it was found that a part of the sealing bone wax had been lost, possibly due to being floated away by CSF. Sealing only with bone wax is thought to be inadequate for preventing CSFleakage [ 81.Fibrin glue has been applied as an adjunctive sealant for dural and bone defects, resulting in reduction of the occurrence of postoperative CSF leakage [7,8]. In our cases, fibrin glue was used as a sealant after packing the exposed air cells with muscle fragments in the repair operation, resulting in complete sealing. These findings indicate that fibrin glue should be used as an adjunctive sealant after packing with
340
Kabuto et al
Surg Nemo1 1996:45:336-40
various substances such as muscle fragments or compressed bone dust [8] at the time of exposure of the air cells. In conclusion, it is considered that if the postoperative CSF otorhinorrhea was suspected, MR imaging (especially T,-weighted images) may be very helpful in diagnosing CSF leakage and detecting the route in the temporal bone, and that fibrin glue may be useful as an adjunctive sealant for preventing CSF otorhinorrhea following the suboccipital approach to the cerebellopontine angle and the internal auditory canal. REFERENCES 1. Allam AF. Pneumatization of the temporal bone. Ann Otol 1969;78:49-64. 2. Clark WC, Acker JD, Robertson .JH, Gardner G, Dusseau JJ, More&W. Neurorad~olog~cal detection of small and intracanalicular acoustic tumors: An emphasis on COZ contrast-enhanced computed tomographic cisternography. Neurosurgery 1982;11:733-8.
0
UR KNOWLEDGE EXPERIENCE
OF
3. Gordon DS, Kerr AG. Cerebrospinal fluid rhinorrhea following surgery for acoustic neurinoma. Report of two cases. J Neurosurg 1986;64:676-8. 4. Lang J, Kerr AG. Pneumatization of posteromedial aircell tract. Clin Otolaryngol 1989;14:425-7. 5. Mitchell DG, Burk DL Jr, Vinitski S, Rifkin MD. The biopbysical basis of tissue contrast in extracraniaf MR imaging. Am J Roentgenol 1987;149:831-7. 6. Montgomery WW. &ommon complications following removal of vestibular schwannoma. Adv Otorhinolaryngol 1983;31:228-39. 7. Shaffrey CI, Spotnitz WD, Shaffrey ME, Jane JA. Neurological application of fibrin glue: Augmentation of dural closure in 134 patients. Neurosurgery 1990;26:20710. 8. Symon L, Peli MF. Cerebrospinal fluid rhinorrhea following acoustic neurinoma surgery. Technical note. J Neurosurg 1991;74:152-3. 9. Wakhloo AK, van Velthoven V, Schumacher M, Krauss JK. Evaluation of MR imaging, digital subtraction cisternography, and CT cisternography in diagnosing CSF fistula. Acta Neurochir (Wien) 1991;lll: 119-27.
IS TliE AMASSED INNUMERABLE RALPH
WALDO
THOUGI-IT MINDS. EMERSON
AND (I 803-82)
MR IMAGING
OF CEREBROSPINAL FLUID RHINORRHEA FOLLOWING THE SUBOCCIPITAL APPROACH TO THE CEREBELLOPONTINE ANGLE AND THE INTERNAL AUDITORY CANAL: REPORT OF Two CASES Masanori Kabuto, M.D., Toshihiko Kubota, M.D., Hidenori Kobayashi, M.D., Yuji Handa, M.D., Akira Tuchida, M.D., and Hiroaki Takeuchi, M.D. ~epu~ent
of ~eurosu~e~,
Fukui medicos echoes, F~kui, Jupu~
Kabuto M, Kubota T, Kobayashi H, Handa Y, Tuchida A, Takeuchi H. MR imaging of cerebrospinaffluid rhinorrhea following the suboccipital approach to the cerebellopontine angle and the internal auditory canal: report of two cases.Surg Meuroll~,~~3~-~. BACKGROUND
Cerebrospinal fluid (CSF) otorhinorrhea is one of the most common postoperative complications following the suboccipital approach to the cerebellopontine angle and the internal auditory canal. Accurate preoperative detection of the site of CSF leakage is im~rtant because inaccuracy may require a more extensive exploratory surgical procedure in the repair operation. There are few reports on evaluation of magnetic resonance (MR) imaging in diagnosing CSF leakage. CASE DESCRIPTION MR imaging of two cases of postoperative
CSF rhinorrhea is reported. A 53-yearold woman and a 29-yearold man underwent suboccipital operations for microvascular decompression of the facial nerve in hemifacial spasm and for removal of an acoustic schwannoma, respectively. In both cases, MR findings were useful in preoperatively delineating the site of the CSF leakage, which was confirmed during the repair operation. CONCLUSION
MR imaging may be useful in identifying the site of CSF leakage following the suboccipital approach to the cerebellopontine angle. KEY WORDS
Acoustic sc~~nno~~, cere~~sp~nul fluid r~ino~~ea, magi net&z resonance, su~~cipita~ approach, tempoml bone air cells.
fluid (CSF) otorhinorrhea is one of the most common postoperative complications following the suboc~ipit~ approach to the erebrospinaf
c
Address reprint requests to: Masanori Kabuto, M.D., Department Neurosurgery, Fukui Medical School, Matsuoka, Fukui 910-11, Japan. Received February 8, 1995; accepted May 25, 1995. 009~3019/96/$15.00 SSDI ~9~3019(9~002~
of
cerebellopontine angle and the internal auditory canal for acoustic schwannomas and other lesions [6]. It is generally believed that there are two major sites of CSFleakage following the operation. One is the mastoid air cells lateral to the sigmoid sinus, which may be opened during a suboccipital craniectomy [1,3,6,8]. The other site is the posteromedial air cell tract [1,3,4] extending posterior to the bony labyrinth and into the posterior wall of the internal auditory canal which may be opened when drilling the posterior wall of the internal auditory canal for removal of an intracanalicular part of an acoustic schwannoma [3,8]. However, when postoperative CSFotorhinorrhea occurs after the operations described above, it is not easy to accurately identify
the location
of the CSF leakage preopera-
tively. Magnetic resonance (MR) imaging may be useful for determining
the CSF leakage
route
be-
cause T,-weighted MR images can demonstrate fluids such as CSF in the temporal bone air cells. However, to our knowledge, there are few reports on MR findings of CSF leakage [9]. This report describes two types of CSF rhinorrhea following the suboccipital approach to the cerebellopontine angle, in which MR imaging clearly verified the route of the CSF rhinorrhea.
CASEREPORTS CASE
1
A 53-year-old woman was admitted to our hospital on April 19, 1993 with a l-year history of left progressive facial spasm. A left vertebral angiogram 0 1996 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010
MR Imaging of CSF Rhinorrhea
Surg Nemo1 1996;45:336 -40
337
Axial T,-weighted (2352/80) MR image (A) performed postoperatively demonstrating high-intensity signals corresponding to CSF in the mastoid air cells (asterisk), the middle ear cavity (star), the eustachian
tube (small arrow), and the nasal cavity (double asterisks). The CSF leakage point (large arrow) was identified as an inferomedial part of the mastoid air
cells. Axial plain CT scan with bony windows (B) showing the extent of the craniectomy and the opened air cells. Axial T, weighted MR image (c) performed after repair operation demonstrating no further CSF leakage.
showed dilatation and elongation of the anterior inferior cerebellar artery and elevation of the posterior inferior cerebellar artery. A computed tomographic (CT) scan (GE-CT 9800, GE Medical Systems, Milwaukee, WI) and MR imaging (1.5-T Signa imaging system; GE Medical Systems) showed no tumor and no vascular malformation at the cerebeilopontine angle. On April 27, 1993, a left suboccipital craniectomy was performed for microvascular decompression. The exposed mastoid air cells were sealed with bone wax. The patient noticed a continuous watery discharge into the n~oph~x a few days after the operation, although the facial spasm stopped postoperatively. Continuous spinal drainage of CSF for 10 days failed to stop the CSF leakage. T,-weighted MR images demonstrated high-intensity signals corresponding to CSF in the mastoid air cells, the middle ear cavity, the eustachian tube, and the nasal cavity (Figure 1 A). The CSF leakage point was also identified clearly as an inferomedial part of the mastoid air cells (Figure
1 A) where a plain CT scan with bony windows confirmed the opened air cells (Figure 1 B). On May 20, 1993, a second operation for repairing CSF leakage was performed. Inspection during the operation confirmed the MR findings of CSF leakage described above. The exposed mastoid air cells where sealing bone wax had already been lost were found. This area was packed with muscle fragments and then covered with fibrin glue. It was also found that a part of the dura mater had been loosely sutured and CSF leaked from it. This part of the dura was tightly resutured and covered with fibrin glue. After the second operation, the patient had no further CSF rhinorrhea and her postoperative course was uneventful. MR imaging performed three weeks after the second
operation
showed
no further
CSF leak-
age (Figure 1 C). CASE 2 A 29-year-old man was admitted to our hospital on November 5, 1990, with a half-year history of left
338
Surg Neurol 1996;45:336-40
Kabuto et al
Axial gas CT cisternography an intracanalicular acoustic H eighth nerve. This CT also shows
air cells in the posterior meatal wall were also sealed with bone wax. A histological examination of the surgically obtained specimens revealed an acoustic schwannoma. On the sixth postoperative day, the patient noticed an intermittent watery discharge into the nasopharynx and a developing conductive hearing loss. It was suspected that the CSF was coming from the middle ear via the eustachian tube. continuous spinal drainage of CSFfor 2 weeks failed to stop the CSF leakage. T,-weighted MR images demonstrated high-intensity signals corresponding to CSFin the posterior wall of the internal auditory canal, the mastoid air cells, the middle ear cavity, the eustachian tube, and the nasal cavity (Figures 3 A and B). The CSF leakage point was preoperatively identified as the posterior wall of the internal auditory canal by MR imaging. A residual tumor was also seen. On December 14, 1990, a second operation was performed to remove the residual tumor and to repair CSF leakage. Rxposed air cells from which bone wax had already floated away were observed in the posterior meatal wall. After removing the residual tumor, the bone defect was packed with muscle fragments and then sealed with fibrin glue. After the second operation, the patient had no further CSF rhinorrhea and his postoperative course was uneventful. MR imaging performed 6 weeks after the second operation showed no further CSF leakage (Figure 3 C).
tinnitus and hearing disturbance. On neurological lamination, he had only a moderate sensorineur~ hearing disturbance on the left side, but this hearing was preserved to a certain extent. The CT scan showed an enlargement of the left internal auditory canal but failed to show a tumor. Air cells in the posterior wall of the left internal auditory canal as well as well-developed mastoid air cells were disclosed on the CT scan (Figure 2 A). A gas CT cisternography [2] was performed. This cisternography clearly delineated an intracanalicular acoustic tumor, facial nerve, and 8th nerve (Figure 2 A). An enhanced T,-weighted MR image demonstrated an intracanalicular acoustic tumor (Figure 2 B). On November 16, 1990, a left suboccipital craniectomy was performed and exposed mastoid air cells were sealed with bone wax. The posterior wall of the internal auditory canal was removed by drilling, and the intracanaIicular tumor was exposed and then removed. After removing the tumor, exposed
The incidence of CSF otorhinorrhea following the suboccipital approach to the cerebellopontine angle and the internal auditory canal is thought to vary from 4% to 18% [3,6]. The major sites of postoperative CSFleakage are the group of mastoid air cells lateral to the sigmoid sinus, which are always opened at the lateral margin of the suboccipital craniectomy unless the degree of pneumatization is small [ 1,3,6,8]. These air cells are interconnected and all of them eventually drain into the middle ear cavity connecting with the nasopharynx through the eustachian tube. Case 1 is one of this type. In this case, T,-weighted MR images clearly demonstrated this CSFrhinorrhea route. Another possible site is the posterior wall of the internal auditory canal, of which pneumatization is reported to be observed in about 20% of patients [ 1,3,4,83. This posteromedial air cell tract connects with the mastoid air cells. Case 2 is one of this type. Also in this case, MR imaging accurately depicted the route of the CSF rhinorrhea. Accurate preoperative detection of the site of CSF leakage is very important because inaccuracy may require a more extensive
(A) clearly delineating tumor, facial nerve, and welldeveloped mastoid air cells and pneumatization of the posteromedial air cell tract in the posterior meatal wall (arrow). AxiaI gadolinium-DTPA-enhanced T,-weighted (317/20) MR image (B) showing a homogenously enhanced tumor in the left internal auditory canal.
MR Imaging of CSF Rhinorrhea
Surg Neurol 339 199~45:336~0 rhinorrhea following the suboccipital approach to the cerebellopontine angle and the internal auditory canal. Although MR imaging (especially increased T2weighted signal) is very useful for detection of fluid within the temporal bone air cells and middle ear cavity, it is not specific for the presence of CSF within these cavities. CSF may be distinguished from highly proteinaceous fluid secondary to inflammatory reaction or bleeding because proteinaceous fluid has a higher signal intensity than simple fluid in TV-weighted images 151. However, the signal from simple fluid due to serous otomastoiditis may be identical to the CSFsignal. Therefore, in the presence of preexisting otomastoiditis, it may be impossible to detect CSF by relying only on the signal intensity of MR imaging within the mastoid air cells and middle ear cavity. In patients with clinical symptoms suggesting postoperative CSF rhinorrhea, however, if the continuity of high-intensity signals like CSF from subarachnoidal space to the eustachian tube through the mastoid air cells can be determined from the T,-weighted images as demonstrated in our cases, MR imaging can strongly suggest the presence of CSF within the mastoid air cells as well as suggesting the site of leakage. Positive
‘[‘,-weighted (25~1/80) MR images (A,B) qintensityAxial performed postoperatively demonstrating highsignals corresponding to CSF in the posterior wall of the left internal auditory canal (large arrow), the posteromedial air cell tract (double arrows), the mastoid air cells (asterisk), the middle ear cavity (star), and the eustachian tube (small arrow). The CSF leakage point (large arrow) was identified as the posterior meatal wall. Axial T,-weighted MR image (C) performed after repair operation demonstrating no further CSF leakage.
exploratory surgical procedure in the repair operation, especially if both the mastoid air ceils and the posteromedial air cell tract are opened as in case 2. In both cases here, MR imaging was considered a noninvasive and simple method enabling direct and accurate identification of the route of the CSF oto-
contrast
CT cisternography
remains
the
gold standard for any CSF rhinorrhea or otorhinorrhea. In many cases, where the bone defect is relatively large, however, the combination of MR imaging and CT scan with bony windows may be adequate to determine the location of the CSF leakage in the temporal bone, eliminating the need for CT cisternography. To our knowledge, there is only one report evaluating MR imaging in diagnosing CSF fistulas at the anterior skull base due to a trauma and an encephalomeningocele [ 91. Our report may be of value with respect to MR imaging in diagnosing CSF leakage and-detesting the route. In our cases, opened air cells were first sealed with bone wax. Postoperative CSF rhinorrhea occurred, however, and in the repair operation it was found that a part of the sealing bone wax had been lost, possibly due to being floated away by CSF. Sealing only with bone wax is thought to be inadequate for preventing CSFleakage [ 81.Fibrin glue has been applied as an adjunctive sealant for dural and bone defects, resulting in reduction of the occurrence of postoperative CSF leakage [7,8]. In our cases, fibrin glue was used as a sealant after packing the exposed air cells with muscle fragments in the repair operation, resulting in complete sealing. These findings indicate that fibrin glue should be used as an adjunctive sealant after packing with
340
Kabuto et al
Surg Nemo1 1996:45:336-40
various substances such as muscle fragments or compressed bone dust [8] at the time of exposure of the air cells. In conclusion, it is considered that if the postoperative CSF otorhinorrhea was suspected, MR imaging (especially T,-weighted images) may be very helpful in diagnosing CSF leakage and detecting the route in the temporal bone, and that fibrin glue may be useful as an adjunctive sealant for preventing CSF otorhinorrhea following the suboccipital approach to the cerebellopontine angle and the internal auditory canal. REFERENCES 1. Allam AF. Pneumatization of the temporal bone. Ann Otol 1969;78:49-64. 2. Clark WC, Acker JD, Robertson .JH, Gardner G, Dusseau JJ, More&W. Neurorad~olog~cal detection of small and intracanalicular acoustic tumors: An emphasis on COZ contrast-enhanced computed tomographic cisternography. Neurosurgery 1982;11:733-8.
0
UR KNOWLEDGE EXPERIENCE
OF
3. Gordon DS, Kerr AG. Cerebrospinal fluid rhinorrhea following surgery for acoustic neurinoma. Report of two cases. J Neurosurg 1986;64:676-8. 4. Lang J, Kerr AG. Pneumatization of posteromedial aircell tract. Clin Otolaryngol 1989;14:425-7. 5. Mitchell DG, Burk DL Jr, Vinitski S, Rifkin MD. The biopbysical basis of tissue contrast in extracraniaf MR imaging. Am J Roentgenol 1987;149:831-7. 6. Montgomery WW. &ommon complications following removal of vestibular schwannoma. Adv Otorhinolaryngol 1983;31:228-39. 7. Shaffrey CI, Spotnitz WD, Shaffrey ME, Jane JA. Neurological application of fibrin glue: Augmentation of dural closure in 134 patients. Neurosurgery 1990;26:20710. 8. Symon L, Peli MF. Cerebrospinal fluid rhinorrhea following acoustic neurinoma surgery. Technical note. J Neurosurg 1991;74:152-3. 9. Wakhloo AK, van Velthoven V, Schumacher M, Krauss JK. Evaluation of MR imaging, digital subtraction cisternography, and CT cisternography in diagnosing CSF fistula. Acta Neurochir (Wien) 1991;lll: 119-27.
IS TliE AMASSED INNUMERABLE RALPH
WALDO
THOUGI-IT MINDS. EMERSON
AND (I 803-82)