1

Introduction

Spontaneous temporal bone cerebrospinal fluid (CSF) leak is a rare but challenging diagnostic and management problem. It often presents with nonspecific symptoms of aural fullness, hearing loss, tinnitus, imbalance, and headaches . Examination of these patients may reveal middle ear effusion, otorrhea, rhinorrhea, and pulsatile movement of the tympanic membrane . The cause of this condition is not fully understood but has been shown to correlate with erosion of the posterior or middle cranial fossa (MCF) by abnormal arachnoid granulations , encephaloceles , and increased intracranial pressure . This entity is distinct from temporal bone leaks of congenital origin that often present in pediatric population secondary to enlargement of preexisting bony pathways such as the cochlear aqueduct, petromastoid canal, Hyrtl fissure, or facial canal . Diagnosis and preoperative identification of defects in the adult onset disease can be challenging, and benefits of laboratory and radiologic ancillary tests have been disputed . Generally, a combination of fluid analysis in cases of frank CSF leak, computed tomography (CT), and magnetic resonance imaging (MRI) has been advocated for diagnosis.

Various approaches for repair have been proposed. Many authors advocate an MCF approach as an initial approach for all defects. However, MCF carries the significantly greater morbidity associated with craniotomy and temporal lobe retraction. The choice of approach carries important implications with respect to surgical outcome and postoperative morbidity.

In the current article, we review our experience with routine use of preoperative MRI and CT imaging for diagnosis and management of defects with a focus on initial transmastoid (TM) approach to repair of posterior cranial fossa and tegmen mastoideum defects. We will summarize our outcomes with this approach and make new recommendations for diagnosis and management of this challenging condition.

2

Materials and methods

Clinical data and imaging results for 18 ears (15 patients) presenting with spontaneous CSF leakage originating in the temporal bone were reviewed after obtaining University of California Irvine’s Institutional Review Board approval. Average follow-up period was 13.5 months. The main outcome measure was presence of persistent CSF leak postoperatively. These cases represent the senior author’s (HRD) experience with this condition over 8 years of practice (2002–2010).

A standard postauricular mastoidectomy is performed. The goal of the mastoidectomy is to expose the dural defect but not create a significant opening of the mastoid antrum. All the facial recess, mastoid tip, and retrofacial air cells are obliterated with bone wax. In our acoustic neuroma experience, we have found that bone wax alone is not sufficient for long-term control of CSF leakage. We reenforce the obliteration with a large temporalis fascia to cover all the air cells and the mastoid antrum. The fascia is then covered with fibrin glue.

The tegmen defect is addressed by first separating the dura from the surrounding bone using an angled hook or joint knife to define the plain between the bone and the dura. Further undermining of the dura is performed using a longer blunt hook. Care must be taken because the bone of the area surrounding a spontaneous defect is generally thin and because the bone may further fracture, creating a larger defect. In addition, the dura tends to be thinned in those areas and can be adherent to the bone. Sharp dissection may be required for separating the dura from the bone. Next, a piece of fascia is placed in an underlay fashion between the dura and the middle (or posterior) fossa floor and overlapping the surrounding bone by approximately 4 to 5 mm around. The fascia is sealed into place with fibrin glue. For larger tegmen mastoideum defects (≥1 cm), we have occasionally used a piece of Duragen (Integra LifeSciences, Plainsboro, NJ) in addition to the fascia to provide better support. A second fascia layer is used as an onlay graft on the mastoid side, and the mastoid is then packed tightly with abdominal fat. A thin titanium mesh is used for closure over the mastoid using 3 or 4 screws. The titanium mesh is then compressed in the middle to create better compression of the fat in the mastoid. No postoperative lumbar drain is placed. The patient receives intraoperative antibiotics and mannitol (1 g/kg). A postoperative CT scan is obtained 4 hours later to rule out a subdural or epidural hemorrhage caused by the dissection. After the operation, the patient is observed for 23 hours before discharge. In the hospital, the patient is started on acetazolamide 125 mg BID (escalating to 500 mg BID after several weeks, titrated based on side effects) for a total of 6 weeks. Outpatient neuro-ophthalmology consultation was obtained postoperatively for all patients to evaluate for development of benign intracranial hypertension (BIH) on every 6-month basis.

2

Materials and methods

Clinical data and imaging results for 18 ears (15 patients) presenting with spontaneous CSF leakage originating in the temporal bone were reviewed after obtaining University of California Irvine’s Institutional Review Board approval. Average follow-up period was 13.5 months. The main outcome measure was presence of persistent CSF leak postoperatively. These cases represent the senior author’s (HRD) experience with this condition over 8 years of practice (2002–2010).

A standard postauricular mastoidectomy is performed. The goal of the mastoidectomy is to expose the dural defect but not create a significant opening of the mastoid antrum. All the facial recess, mastoid tip, and retrofacial air cells are obliterated with bone wax. In our acoustic neuroma experience, we have found that bone wax alone is not sufficient for long-term control of CSF leakage. We reenforce the obliteration with a large temporalis fascia to cover all the air cells and the mastoid antrum. The fascia is then covered with fibrin glue.

The tegmen defect is addressed by first separating the dura from the surrounding bone using an angled hook or joint knife to define the plain between the bone and the dura. Further undermining of the dura is performed using a longer blunt hook. Care must be taken because the bone of the area surrounding a spontaneous defect is generally thin and because the bone may further fracture, creating a larger defect. In addition, the dura tends to be thinned in those areas and can be adherent to the bone. Sharp dissection may be required for separating the dura from the bone. Next, a piece of fascia is placed in an underlay fashion between the dura and the middle (or posterior) fossa floor and overlapping the surrounding bone by approximately 4 to 5 mm around. The fascia is sealed into place with fibrin glue. For larger tegmen mastoideum defects (≥1 cm), we have occasionally used a piece of Duragen (Integra LifeSciences, Plainsboro, NJ) in addition to the fascia to provide better support. A second fascia layer is used as an onlay graft on the mastoid side, and the mastoid is then packed tightly with abdominal fat. A thin titanium mesh is used for closure over the mastoid using 3 or 4 screws. The titanium mesh is then compressed in the middle to create better compression of the fat in the mastoid. No postoperative lumbar drain is placed. The patient receives intraoperative antibiotics and mannitol (1 g/kg). A postoperative CT scan is obtained 4 hours later to rule out a subdural or epidural hemorrhage caused by the dissection. After the operation, the patient is observed for 23 hours before discharge. In the hospital, the patient is started on acetazolamide 125 mg BID (escalating to 500 mg BID after several weeks, titrated based on side effects) for a total of 6 weeks. Outpatient neuro-ophthalmology consultation was obtained postoperatively for all patients to evaluate for development of benign intracranial hypertension (BIH) on every 6-month basis.

3

Results

Fifteen patients underwent surgery for repair of CSF leak in 18 ears (3 patients had bilateral disease). Of the 18 cases with spontaneous CSF leak, 8 (44%) presented initially with eustachian tube dysfunction (ETD) or serous otitis media; 5 (28%), with otorrhea after pressure equalization tube (PET) placement or TM perforation; and 5 (28%), with recurrent meningitis ( Table 1 ). All patients were evaluated with high-resolution CT scans. All but 2 patients also underwent an MRI scan. Diagnostic imaging localized the temporal bone defect to mastoid tegmen in 14 patients. Three of those patients had a combined mastoid tegmen and tegmen tympani defect. There were 3 isolated posterior fossa cases and 1 isolated tegmen tympani defect. Isolated mastoid tegmen defects were successfully repaired via TM approach in 8 of 11 cases and via MCF or combined approach in the other 3 cases. The MCF approach was used in these 3 patients in our early experience because of the preference of the neurosurgeon in the institution who referred the patient. Transmastoid approach alone was also used in repair of 3 posterior fossa defects with postoperative CSF leak in 1 case. Combined defects of tegmen tympani and mastoid tegmen were treated via MCF approach in 2 cases and combined approach in 1 case without evidence of CSF leak on postoperative follow-up. The case of isolated tegmen tympani defect occurred in an 83-year-old patient with multiple comorbidities who was treated with a transcanal eustachian tube plug using fascia and fibrin glue under local anesthesia to avoid morbidity of prolonged operative repair. The sole case of postoperative leak in our series occurred in a case in which no fibrin glue or tissue glue was used according to the patient’s religious belief.

Table 1

Results of an 8-year experience with various presentations and treatment options for spontaneous CSF leakage originating in the temporal bone

No.	Age	Sex	Presenting symptom	Workup	Treatment	Location	Follow-up
1A	53	F	Meningitis	CT, MRI	TM	Mastoid tegmen	12 mo, no leak
1B	55	F	ETD	CT, MRI	MCF extradural	Mastoid tegmen	24 mo, no leak
2	66	F	Otorrhea post PET	CT, MRI	MCF intradural + LD	Mastoid tegmen	48 mo, no leak
3	62	F	Otorrhea post PET	CT	TM	Posterior fossa	2 mo, leaked
4	83	M	Serous otitis media	CT	Transcanal ET plug	TT	6 mo, no leak
5	59	F	Otorrhea	CT, MRI	Combined approach + LD	TT + mastoid tegmen	12 mo, no leak
6	62	F	Serous otitis media	CT, MRI	Combined approach + LD	Mastoid tegmen	6 mo, no leak
7	56	F	Serous otitis media	CT, MRI	MCF extradural + LD	TT + mastoid tegmen	68 mo, no leak
8	58	F	Recurrent meningitis	CT, MRI	MCF extradural + LD	TT + mastoid tegmen	9 mo, no leak
9	57	F	Serous otitis media	CT, MRI	TM	Mastoid tegmen	12 mo, no leak
10	68	F	Otorrhea post PET	CT, MRI	TM	Mastoid tegmen	12 mo, no leak
11	77	M	Serous otitis media	CT, MRI	TM	Posterior fossa	3 mo , no leak
12	57	F	ETD	CT, MRI	TM	Mastoid tegmen	3 mo, no leak
13A	75	F	Serous otitis media	CT, MRI	TM	Mastoid tegmen	2 mo, no leak
13B	75	F	Serous otitis media	CT, MRI	TM	Mastoid tegmen	2 mo, no leak
14	52	F	Pulsatile tinnitus	CT, MRI	TM	Posterior fossa	2 mo, no leak
15A	33	M	Meningitis	CT, MRI	TM	Mastoid tegmen	3 mo , no leak
15B	33	M	Meningitis	CT, MRI	TM	Mastoid tegmen	3 mo , no leak

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Delayed lymph node metastases after elective neck dissection in patients with oral and oropharyngeal cancer and pN0 neck
2. Congenital cholesteatoma: clinical features and growth patterns
3. Comparing capsule exposure using extracapsular dissection with partial superficial parotidectomy for pleomorphic adenoma
4. Cochlear implantation outcomes in patients with far advanced otosclerosis

Stay updated, free articles. Join our Telegram channel

Join