Outcomes after Mini-Craniotomy Middle Fossa Approach Combined with Mastoidectomy for Lateral Skull Base Defects

Abstract

Purpose:

Controversy exists regarding the ideal approach for repair of lateral skull base defects. Our goal is to report the outcomes following middle cranial fossa (MCF) mini-craniotomy combined with mastoidectomy for patients with superior semicircular canal dehiscence (SSCD), spontaneous cerebrospinal fluid (CSF) leak, and cholesteatoma.

Materials and Methods:

A retrospective database from chart review was formed consisting of 97 patients who met surgical criteria: SSCD, spontaneous CSF leak, and cholesteatoma. Mini-craniotomy MCF approach (<4 × 2 cm in size) combined with mastoidectomy was performed. All patients were admitted directly to the ICU postoperatively. Multiple factors were assessed, including need for revision surgery, duration of surgery, length of post-operative stay, and hospital readmission.

Results:

Average surgery time was 110 minutes with no intraoperative complications. The average length of hospitalization was 2 days with an average ICU stay of 1 day. There were no neurologic complications; however, there were 3 inpatient complications (3%) which included 1 patient (1%) that had wound breakdown and 2 patients (2%) that had severe post-operative vertigo. A total of 8 patients (8%) required revision surgery and these were primarily for SSCD. The 30-day readmission rate was 3%.

Conclusion:

In the current series, all patients that underwent mini-craniotomy MCF surgery combined with mastoidectomy had minimal complications, short surgical time, limited hospital stay, low revision surgery rate and few hospital readmissions. This combined approach offers superior visualization of lateral skull base defects without the morbidity and risk typically associated with traditional, extensive MCF surgery.

1. Introduction

The tegmen tympani and tegmen mastoideum form the roof of the middle ear/mastoid and floor of the middle cranial fossa (MCF). Defects of the MCF floor can lead to multiple neurotologic pathologies including but not limited to cerebrospinal (CSF) leak, meningoencephaloceles, and superior semicircular canal dehiscence (SSCD). Multiple surgical methods have been described for repair of lateral skull base defects including transmastoid, MCF craniotomy, combined transmastoid and MCF approaches, and MCF with keyhole craniotomy or endoscopic assisted approach [1–10]. Significant debate remains however, regarding the ideal surgical approach based on extent of disease, patient specific factors, and surgeon preference.

The classic MCF approach allows for a top-down repair of a defect; however, it is criticized for extent of temporal lobe retraction, potential need for lumbar drainage, increased morbidity, and lengthened hospital stay [11–15]. The size of a traditional MCF craniotomy is approximately 5 × 5 cm [16]. In contrast, the transmastoid approach allows for a shorter hospital stay without the need for temporal lobe retraction and associated morbidity [6, 17, 18]. Its efficacy, however, is limited by anterior exposure due to the ossicles, limited benefit with larger skull base defects, and issues attendant with a sclerotic mastoid [5, 18, 19]. Furthermore, it is often advocated that the transmastoid approach be reserved for older, infirmed patients due to the morbidity associated with extensive temporal lobe retraction in this cohort, as well as for those with small tegmen defects, while traditional MCF craniotomy is generally utilized for larger tegmen defects with significant CSF leak [20–22]. Most reports describe using either transmastoid or MCF approaches alone and rarely in combination [1, 2, 4, 7, 9, 10, 21].

Over the past 10 years, the authors have performed a combined approach characterized by a mastoidectomy in combination with MCF mini-craniotomy (<4 × 2 cm in size) for repair of most lateral skull base defects. The aim of this study was to review the senior author’s experience, describe the surgical technique, and examine both short- and long-term postoperative outcomes using a defined management pathway. The purpose of this report was to determine whether the combined MCF mini-craniotomy and mastoid approach maximizes the advantages of both approaches while minimizing the morbidity and risk associated with a traditional, more extensive MCF approach.

2. Materials and Methods

With Institutional Review Board approval, the database of patients who underwent a MCF mini-craniotomy combined with mastoidectomy for the diagnoses of spontaneous CSF leak, SSCD, and cholesteatoma was reviewed. The medical records of 97 consecutive patients treated from 2011 to 2018 were retrospectively reviewed. Despite the various indications for surgery, all patients underwent a standard surgical approach and followed a consistent post-operative clinical pathway.

Inclusion criteria were all patients who had a spontaneous CSF leak, temporal lobe herniation, SSCD, or cholesteatoma, where a combined mastoidectomy with MCF mini-craniotomy was performed. Exclusion criteria were patients who underwent solely a MCF approach or mastoidectomy approach for lateral skull base defect repair. Patients who had an iatrogenic CSF leak from a prior lateral skull base surgery (translabyrinthine or retrosigmoid craniotomy) were also excluded from this study.

2.1. Surgical Approach

All patients are initially positioned supine with head turned without fixation or use of stereotactic guidance. A single postauricular incision is made, extending superiorly in the temporal area and then anteromedially to the zygomatic root to incorporate exposure for the MCF mini-craniotomy and mastoidectomy. The soft-tissue exposure involves elevation of an anteriorly-based mastoid periosteal flap to allow for complete mastoid exposure (Figure 1). The temporalis is also elevated from posterior to anterior leaving the muscle pedicled at the level of the zygomatic root, preserving the neurovascular supply and exposing the squamosal portion of the temporal bone. Mastoidectomy is performed initially for identification of landmarks including the ossicular chain and lateral and superior semicircular canals. This initial exposure allows for visualization of any tegmen defect from the sinodural angle up to the anterior epitympanum. If surgery for SSCD is being performed, the superior semicircular canal is also identified in the mastoid and a small fenestration in the tegmen is created immediately lateral to the superior semicircular canal to allow for ease of identification from above (Figure 2). For cholesteatoma cases, the mastoid and middle ear disease is completely removed until middle fossa extension is identified through the mastoidectomy. In all cases involving this approach, a cottonoid is placed through the known or intentionally created floor defect to assist with anatomic exposure from the MCF approach. Figure 3 describes the approach when used for repair of temporal lobe meningoencephalocele.

Figure 1:

General surgical steps for combined transmastoid and minimally invasive MCF craniotomy. A – An incision is outlined approximately 0.5 cm behind the postauricular sulcus and extended superiorly and then anteroinferiorly allowing for adequate exposure for the subtemporal craniotomy. B – A mastoidectomy is performed and tegmen defect is identified. A cottonoid is placed in the mastoid opening to serve as identification landmark for the MCF approach. A minimally invasive MCF craniotomy (<2 by 4 cm in size) is made and craniotomy flap is preserved for repair. C – Tegmen defect is identified from the previously placed cotton (white arrow) and repaired. D – Position of autologous bone graft is reinforced with hydroxyapatite material (white arrow) and mastoid is able to be examined to ensure no extravasation into the middle ear prior to wound closure.

Figure 2:

Repair of SSCD. A – From a transmastoidectomy approach, the superior semicircular canal is identified and middle fossa dura exposed by fenestrating the tegmen immediately lateral (white arrow). B – From a minimally invasive MCF craniotomy approach, dura is elevated medially to the area of the tegmen defect and dissection is then carried in the plane overlying the arcuate eminence where the dehiscence (white arrow) may then be identified. C – Temporalis fascia (white arrow) is harvested and used to plug and resurface the area of dehiscence. D – Previously elevated craniotomy flap is used as an autologous graft (white arrow) to reinforce repair of the MCF floor.

Figure 3:

Repair of meningoencephalocele. A – Mastoidectomy (black arrow) is initially performed followed by a mini-craniotomy. The meningoencephalocele is then reduced back into the middle fossa from below. Craniotomy flap (white arrow) is used as autologous graft for large broad-based defect of middle fossa from above. B – Middle cranial fossa repair is reinforced with hydroxyapatite material (white arrow), which is also utilized to repair the craniotomy defect. Mastoid is examined to ensure that there is no extravasation of material toward the ossicles and middle ear.

An MCF mini-craniotomy is now performed using an 11 mm perforator and craniotome. The previously identified middle fossa floor and tegmen defect aid in optimal placement of the perforator openings. The first burr hole is placed anterior to the defect, typically just above the zygomatic root and the second is placed posterior above the sinodural angle. A craniotome is then used to create a limited craniotomy bone flap between the two burr holes. The bone flap is preserved to be used for the subsequent reconstruction of the lateral skull base defect. The size of the craniotomy is typically no more than 4 cm anterior to posterior and 2 cm inferior to superior.

Dural elevation is performed from posterior to anterior allowing for safe elevation over the geniculate ganglion, if necessary. A fixed dural retractor is not used in an effort to minimize temporal lobe retraction. In the case of SSCD, care is taken with elevation proximal to the superior canal dehiscence to minimize trauma to the membranous portion of the labyrinth. SSCD defects are repaired by canal plugging with fascia. Temporal lobe herniation is elevated out of the mastoid. Bony defects of the middle fossa floor are then corrected in a multilayered fashion. Any dural defect is repaired with collagen matrix. The previously harvested craniotomy bone flap is then used for repair of the bony defect of the middle fossa floor and in the case of SSCD, to further reinforce the canal plugging. Surgical closure includes the use of hydroxyapatite bone substitute over the remaining portion of the MCF floor, as well as to repair the craniotomy defect. No titanium instrumentation was used for these repairs. The previously placed cottonoid in the mastoid is then removed to allow for visualization of the repair from below to be certain there is no extravasation of hydroxyapatite towards the ossicles and in the case of spontaneous CSF leak, to confirm the repair. The temporalis and periosteal flaps are reapproximated. A compression dressing is applied after closure is completed.

2.2. Postoperative Care Pathway

A standardized clinical care pathway is used for all patients with an admission to the ICU for post-operative monitoring. By the morning of the first post-operative day, the Foley catheter is removed, and the patient is moved to a chair prior to transfer to the floor. All patients are ambulated on postoperative day one. On the second postoperative day, the compression dressing is removed, and the patient is discharged home when demonstrating adequate independent ambulation and having a bowel movement. Neurological assessments are performed initially in the ICU hourly and then every eight hours on the floor.

Medical therapy includes perioperative intravenous antibiotics for 24 hours. Mannitol is perfused immediately prior to the incision. Patients are placed on a two-day steroid taper. For patients under 75 years of age, a scopolamine patch is used for nausea and vertigo control with a combination of other anti-emetics (ie, ondansetron, metoclopramide) as needed. Venous thromboembolism prophylaxis includes use of compression stockings and aggressive mobilization characterized by ambulating all patients by postoperative day one, typically with independent activity by post-operative day two without chemical prophylaxis.

2.3. Data Analysis

Statistical analysis was performed using SAS Version 9.4 statistical software (SAS Institute, Cary, NC). Standard descriptive statistics were used to highlight frequency and percentages for categorical data. The median values with minimums and maximums were provided for continuous data.

3. Results

A total of 97 patients underwent surgical repair for lateral skull base defect with a combined mastoid and minimally invasive MCF craniotomy approach. The patient demographics associated with each surgical indication are summarized in Table 1. The most common indication for surgical repair was SSCD (66.0%) followed by spontaneous CSF leak (28.9%), typically with temporal lobe herniation, and cholesteatoma eroding the middle fossa floor (5.1%). There was a higher percentage of female patients in the SSCD and spontaneous CSF leak cohort. BMI was also elevated for the spontaneous CSF leak cohort with a median score of 38 (Range 19–65).

Table 1:

Baseline Demographics Stratified by Surgical Indication

	Total Cohort (n=97)	SSCD (n=64)	Spontaneous CSF Leak (n=28)	Cholesteatoma (n=5)
Age (years)
Median (Min-Max)	51 (17–79)	51 (17–75)	53 (31–79)	39 (20–60)
Gender
Male	23 (24%)	15 (23%)	6 (21%)	2 (40%)
Female	74 (76%)	49 (77%)	22 (79%)	3 (60%)
Side
Left	50 (52%)	30 (47%)	17 (61%)	3 (60%)
Right	47 (48%)	34 (53%)	11 (39%)	2 (40%)
BMI Score Median (Min-Max)	30 (18–65)	26 (18–53)	38 (19–65)	28 (22–52)

Duration of surgery was consistent regardless of the indication with a median surgery time of 110 minutes (Range 53–234). No intraoperative complications were encountered. Inpatient complications included two cases of prolonged post-operative vertigo after undergoing repair for SSCD. One patient had serosanguinous drainage from their incision which resolved following reapplication of the compression dressing. These three complications resulted in a prolonged hospitalization when compared to the average stay of two days (Figure 4). Two other patients had extended stays due to pre-operative admission, one for antiplatelet therapy and the other for placement of a lumbar drain. The latter was the only case that underwent lumbar drain placement in the entire cohort of 97 patients. There were 8 cases (8.2%) that required revision surgery (Table 2). The majority of these were for SSCD (N=7). The other case requiring revision surgery was for recurrent CSF leak.

Figure 4:

Length of hospital stay is illustrated by surgical indication.

Table 2:

Duration of Surgery, Revision Surgeries, and 30-day Readmissions

	Total Cohort (n=97)	SSCD (n=64)	Spontaneous CSF Leak (n=28)	Cholesteatoma (n=5)
Duration of Surgery Median Minutes (Min-Max)	110 (53–234)	107 (82–180)	124 (53–234)	107 (82–137)
Need for Revision Surgery	8% (8/97)	11% (7/64)	4% (1/28)	0% (0/5)
30 Day Hospital Readmission	3% (3/97)	3% (2/64)	4% (1/28)	0% (0/5)

Only 3% of cases had an unplanned all-cause 30-day readmission (Table 2). One patient developed a recurrent CSF leak on post-operative day 7 requiring revision surgery. Another patient developed right-upper-quadrant pain and was diagnosed with cholelithiasis and a third patient developed a seroma requiring incision and drainage.

4. Discussion

The MCF approach provides superior visualization and access to repair various neurotologic skull base pathologies; however, surgeons often hesitate because it is commonly associated with increased morbidity, longer hospital stays, and need for extensive temporal lobe retraction [23]. Meanwhile, the transmastoid approach provides an inferiorly based view to the MCF and has evolved as a solution to address some of the disadvantages of the traditional MCF craniotomy. Limitations with the transmastoid approach, however, include difficulty with repairing large defects and accessing the anterior portion of the tegmen. There have been limited number of reports utilizing a combined transmastoid and MCF craniotomy and most of these have reviewed small sample sizes [3, 10, 21, 24]. Furthermore, most of these previous studies do not use a MCF mini-craniotomy which minimizes the morbidity associated with a traditional middle fossa craniotomy. In this study, experience with a combined transmastoid and MCF mini-craniotomy exclusively used for repair of lateral skull base defects is reviewed. This report represents one the largest series to date utilizing the combined mini-craniotomy and transmastoid approach. It is evident from this study that this combined approach offers a safe and effective treatment option for repairing lateral skull base defects.

Even with these various approaches, the success rates for closure of lateral skull base defects vary widely in the literature, ranging from 50 to 100%, independent of surgical technique [1, 5, 6, 17, 25–29]. We observed an overall revision rate across all lateral skull base pathologies of 8% (8/97) with SSCD representing the most common pathology (11%; 7/64). This rate is similar to more recent reports including a systematic review [21]. The low revision rate for the combined transmastoid and mini-craniotomy approach is due to several surgical advantages when compared to mastoidectomy alone or traditional extensive MCF craniotomy alone. The combined technique allows for rapid identification of the middle fossa floor with the mastoidectomy. By visualizing the defect through the mastoid, the surgeon can then limit the size of the MCF craniotomy. Once the craniotomy is performed, the defect can be easily identified and the repair can be then be performed from the MCF approach in an extradural overlay fashion, which allows for precise placement of a graft or plug. This again minimizes the need for any significant temporal lobe retraction. Adequacy of repair can then be confirmed via the mastoid approach.

In previous studies reviewing the combined approach, the size of the craniotomy has varied significantly. Most of these studies have used a craniotomy significant larger than 4 × 2 cm window which has known increased morbidity [30, 31]. Morbidity of the MCF craniotomy approach is related to extensive temporal lobe retraction, which can lead to postoperative seizures, focal neurological deficits, aphasia, and venous complications [32, 33]. In the current study, the craniotomy is no larger than 4 × 2 cm, whereas a traditional MCF craniotomy is approximately 5 × 5 cm in size. In assessing the literature, only two recent studies were identified that have reviewed experiences with a combined mini-craniotomy and mastoidectomy [5, 10]. Marchioni et al showed effectiveness of this surgical technique in 22 patients without significant intraoperative or postoperative complications using a mini-craniotomy of 4 × 4 cm in size [5]. Hernandez et al also used this approach for 11 patients and performed a smaller craniotomy, similar to the current study, measuring approximately 2 × 3 cm in size [10]. Both studies highlighted the low revision rate and low perioperative complication rate. The current report confirms similar surgical outcomes with a more robust cohort.

Since this combined approach is performed routinely, an established postoperative clinical care pathway has been developed. All patients were admitted to the ICU for immediate postoperative care. There were no complications in the ICU, which may suggest that patients could probably be managed effectively with initial postoperative admission to the floor with adequate neurological monitoring and immediate ambulation. This could potentially allow for discharge for most patients on the first postoperative day. This would be similar to the transmastoid approach, which is generally considered an outpatient procedure with hospital stays ranging from 0.9–1.6 days [17, 29]. In contrast, studies reviewing outcomes of the traditional MCF craniotomy approach describe an average hospital stay of at least 3 days and often longer with concominant use of lumbar drainage [4, 19, 27].

In the current series, a lumbar drain was not used. This allowed for early patient ambulation and decreased hospital stay. In reviewing multiple studies, it is clear that there is no consensus regarding preoperative use of a lumbar drain. Lumbar drains have been associated with increased number of complications including meningitis, pneumocephalus, uncal herniation, persistent headache, and lumbar radiculopathy [27]. For cases of spontaneous CSF leak, placement of a lumbar drain allows for identification of intracranial hypertension [27]. However, in the setting of an active CSF leak, which is the most typical situation, the sensitivity of the opening pressure in identifying intracranial hypertension decreases [34]. The utility of lumbar drain placement may be in situations in which the site of the CSF leak needs to be identified with fluorescein, although this was not an issue in the current study. Others prefer using a lumbar drain postoperatively to aid in healing of a CSF leak [27]. The current analysis, as well as previous reports, confirm that there is a very low incidence of CSF leak requiring revision surgery. This suggests that routine use of lumbar drain placement in otherwise uncomplicated situations is likely unnecessary [25]. Furthermore, in this series, no patients required ventriculoperitoneal shunt diversion due to persistent leak, which is similar to conclusions drawn by previous studies [2, 35].

This study has several limitations. Due to the retrospective nature of this study, the patients could not be randomized; thus, this study is at risk for selection bias. However, the selection bias is limited by the fact that these lateral skull base defect repairs were done for various pathologies instead of a single indication. In this series, severity of the spontaneous CSF leak could not be confirmed since lumbar punctures were not obtained preoperatively to assess for presence of idiopathic intracranial hypertension. Severity or presence of obstructive sleep apnea was also not assessed in this series. Previous studies have shown that OSA is associated with worse outcome [36]. The current cohort of patients did include a wide range of BMI, and it is suspected that many patients may have had OSA. Another limitation of this study is that there is no comparison group for this technique (ie, traditional MCF craniotomy or mastoidectomy), which might otherwise have been achieved by prospective randomization. As there is a divergence of opinion regarding the ideal treatment for these lateral skull base defects, a multi-institutional review may provide a sufficiently large cohort to allow for univariate and multivariate analyses to compare outcomes.

5. Conclusion

This is one of the largest series to analyze outcomes utilizing a combined transmastoid and mini-craniotomy (< 4 × 2 cm) approach for repair of lateral skull base defects including SSCD, spontaneous CSF leak, and defects secondary to cholesteatoma. Through this surgical technique, one is able to maximize the benefit from improved visualization while minimizing the morbidity typically associated and the traditional extensive MCF craniotomy. As a result of the combined approach and a well-defined postoperative care pathway, patients can expect to have a short hospital stay, low incidence of postoperative complications, and minimal readmission rate.

Highlights.

• Middle cranial fossa (MCF) floor defects can lead to neurotologic pathologies

• Pathologies include canal dehiscence, spontaneous CSF leak, and cholesteatoma

• Mini-craniotomy and mastoidectomy offers superior visualization of MCF defect

• Combined approach results in good surgical outcomes with decreased morbidity

• Mini-craniotomy (<4 × 2 cm size) minimizes the risks of traditional MCF craniotomy

Abbreviations:

MCF

middle cranial fossa

SSCD

superior semicircular canal dehiscence

CSF

cerebrospinal fluid

ICU

intensive care unit