Compartmentalization: An Open Technique for Frontal Sinus Repair and Preservation—Description of Technique and Early Case Series

Abstract

Background  Cranialization or obliteration is widely accepted intervention for traumatic or intentional breach of the frontal sinus. These techniques, however, result in the loss of frontal sinus function and have a persistent risk of cerebrospinal fluid (CSF) leak and mucocele. Compartmentalization is an open technique for repair of the frontal sinus using allograft onlay and a vascularized periosteal flap that allows for preservation of frontal sinus function.

Objective  The main objective of this article is to describe the technique for compartmentalization of the frontal sinus and demonstrate its efficacy and complication rate with an early patient series.

Methods  Our technique includes the following key components: harvesting of a pedicled periosteal flap, frontal sinus repair through a bifrontal craniotomy with minimal mucosa removal, ensuring the patency of the nasal frontal outflow tract, and separation of the brain from the frontal sinus with a dual layer of periosteum and allograft.

All cases of frontal sinus repair using the compartmentalization technique at our institution were reviewed. Charts were reviewed for CSF leak, mucocele, and other complications.

Results  Twenty-three patients underwent the described frontal sinus repair technique 17 for tumor and 6 for trauma. There were no CSF leaks and no mucoceles. One patient experienced postoperative anemia and a “parameningeal reaction” that were managed with a short course of antibiotics.

Conclusions  Compartmentalization, due to its sinus preservation and low complication rate, represents a meaningful step forward in neurosurgical technique for open frontal sinus repair. However, long-term outcomes are necessary to fully evaluate risk of mucocele.

Introduction

Frontal sinus repair remains a complex neurosurgical challenge even after a century of developing the technique. The opening of the frontal sinus, either traumatic or intentional, creates a contaminated wound that must be addressed along with any associated pathology. The repair must also be designed to prevent acute and chronic complications such as cerebrospinal fluid (CSF) leak, infection, and mucocele. Finally, the location of the frontal sinus will result in significant lifelong morbidity if a cosmetic repair is not achieved.

There have been several different methods over the years to treat injury to the frontal sinus. Reidel described one of the earliest techniques, ablation. The technique involved aggressively removing anterior wall, mucosa, supraorbital rims, and proximal nasal bones ^{1 2 3 4 5} ; some reports additionally describe removal of the posterior wall. ^{6 7} The skin is then allowed to heal to the posterior sinus wall or dura if the posterior wall was disrupted secondary to pathology or intervention. This left a severe cosmetic defect that required a difficult subsequent operation to repair. Cranialization came about through a combination of efforts. First, Kline and Nadell published the first evidence that damaged bone fragments of the frontal sinus could be cleaned and replaced in a single operation. ⁸ Donald and Bernstein then formally described cranialization in which the posterior wall is removed, the anterior wall is stripped of mucosa, repaired for cosmesis, and the nasofrontal duct is blocked to prevent communication between the intracranial and extracranial compartments. ^{6 9} Cranialization has been the workhorse of frontal sinus repair for the past few decades; however, recently sinus preservation techniques have emerged. A single neurosurgical technique for frontal sinus preservation is described only for mildly displaced fractures and continues to be complicated by CSF leak. ⁷

Described in this article is the compartmentalization method for the preservation of the frontal sinus after traumatic injury or intentional entry. The method focuses on preventing communication between the intracranial and extracranial compartments while allowing for the continued function of the frontal sinus. By ensuring the separation of the two compartments with allograft and pedicled periosteum, the dural repair need only be up to the standard of typical supratentorial craniotomy. Thus, this technique reduces the technical difficulty of frontal sinus repair and satisfies the requirements of frontal sinus repair: management of acute pathology, control of potential complications, and a cosmetic repair with a single surgery.

Methods

After institutional review board approval, charts were retrospectively reviewed from 2013 to 2018 for patients encountered by the senior author (RC) with frontal sinus fracture or planned bifrontal craniotomy. All patients who underwent the compartmentalization procedure were included in the study. Patients excluded from undergoing the compartmentalization procedure include redo craniotomies, fractures extending beyond the planum sphenoidale, and unintentional entry into the frontal sinus where a pedicled periosteal flap could not be obtained.

Trauma patients were selected for frontal sinus repair based on the following protocol: linear fractures involving frontal sinus and cribriform plate are initially managed conservatively. If a CSF leak is noted, a lumbar drain is attempted. If conservative methods fail, then compartmentalization is undertaken. Severely displaced or comminuted fractures of the frontal sinus and cribriform plate with CSF leak or pneumocephalus are directly treated with compartmentalization. Of note, patients were not excluded due to the degree of displacement of the fracture. Nontrauma patients underwent compartmentalization if the planned bone flap to address the underlying tumor involved the frontal sinus and/or the cribriform plate.

All patients underwent the same procedure:

1. A bicoronal skin flap is made with care to not disturb the periosteum. The galea is fully dissected from the periosteum anteriorly and maximally dissected posteriorly as to maximize the harvest of periosteum. Electrocautery is then used to section the periosteum posteriorly and bilaterally. The periosteum is then dissected from the bone using a periosteal elevator. The flap remains pedicled anteriorly and is protected by a moist sponge until later use. Any defects to the periosteum are repaired with 4–0 braided nylon suture ( Fig. 1A ).

2. A standard bifrontal craniotomy with exposure of the frontal sinuses is performed. Any frontal sinus mucosa present in the bone flap is carefully removed and drilled with a burr drill bit.

3. The anterior and posterior tables of the frontal sinuses are reapproximated and plated if necessary. Only loose or severely damaged mucosa is removed. The frontal sinus outflow tract is inspected for patency. Complete removal of the mucosa is not performed, and the outflow of the frontal sinus is not plugged.

4. The dura is dissected from the anterior skull base and inspected to the level of the planum sphenoidale if fractures are suspected in the anterior skull base. The olfactory nerves are coagulated and cut if necessary.

5. Any significant dural defects from the primary tumor or trauma are approximated with 4–0 braided nylon suture. A watertight seal is not necessary.

6. The dura is then covered with a single large dural allograft (Lyoplant, Aesculap, Center Valley, Pennsylvania, United States). The graft is shaped to reach lengthwise from the point of inferior dissection in step 4 to the superior posterior bony opening. The lateral margins are determined by the width of the anterior fossa skull base and bone flap ( Fig. 1B ).

7. The pedicled periosteal flap is placed over the frontal sinus and then layered between the allograft and the skull base to the depth of the posterior dissection in step 4 ( Fig. 1B-C ).

8. Bone flap is replaced with care to closely approximate the anterior margin without restricting circulation to the periosteal flap.

Fig. 1.

Demonstrating a typical pericranium harvest ( A ), the final placement of the allograft covering the complete area of the bone flap ( B ), and a three-dimensional model demonstrating the ideal coverage of the allograft (blue/light gray) and periosteal flap (red/dark gray) ( C ). Note: in part ( B ) the pericranium has not yet been placed to cover the frontal sinus.

Nasoseptal Flap Modification

1. In cases of combined craniotomy and transnasal procedures where significant removal of anterior fossa floor is undertaken for adequate tumor removal, a nasal septal flap or nasal packing is used at the discretion of the surgeon(s). It is imperative that any flap or packing does not cover the nasal frontal outflow tract as this will lead to mucocele.

2. If the nasal frontal outflow tract must be closed, then compartmentalization must be aborted and cranialization is recommended.

The postoperative protocol for the assessment of repair and frontal sinus function is as follows: Post procedure computed tomography (CT) or MRI is obtained during the hospital stay. Post discharge all trauma patients undergo CT scan at 2 to 3 months and then as needed. Tumor patients undergo CT or MRI as indicated at 2 to 3 months and then every 6 to 12 months depending on the pathology.

Results

During the time period, the senior author performed 23 compartmentalization procedures; six trauma and seventeen tumor-related procedures. The average age was 50.7 (range: 16.7–83.9) and the male to female ratio was 15:8 ( Table 1 ). Average follow-up was 37.5 months (15.4–71.2 months). Specific diagnoses for trauma and tumor are detailed in Tables 2 and 3 .

Table 1. Demographics and clinical characteristics for patient series.

Patient demographics
Age (range)	50.7(16.7–83.9)
Gender: female (%)	34.8%
Clinical characteristics
Trauma	6/23
Tumor	17/23
Length of follow-up (m)	37.5 (15.4–71.2)

Table 2. Clinical details for trauma patients who underwent compartmentalization.

Trauma
Case	Age	Mechanism	Fracture location			Displacement (cm) a	CSF diversion	Modality b
			AT	PT	CP
1	37	GSW	C	C	D	0.5	Lumbar drain	CT
2	17	MVC	L	L	D	0.5	No	CT
3	19	GSW	N	D	C	0.6	No	CT
4	61	GSW	C	C	C	N/A	No	CT
5	35	MVC	D	D	L	1.5	No	CT
6	50	GSW	L	C	C	N/A	Shunt	CT

Abbreviations: AT, anterior table of frontal sinus; C, comminuted; CP, cribriform plate; D, displaced; L, linear; N, no fracture; PT, posterior table of the frontal sinus.

^aMaximal displacement if applicable.

^bModality used to assess frontal sinus outflow tract patency.

Table 3. Clinical details for tumor patients who underwent compartmentalization.

Tumor
Case	Age	Diagnosis	Size (cm) a	Location b	Ventral adjunct	Modality c
7	46	Squamous cell carcinoma	6.2	Centered on cribriform plate	NSF	MRI
8	43	Esthesioneuroblastoma	4.1	Centered on cribriform plate	NSF	MRI
9	71	Poorly differentiated carcinoma	5.5	Invasive through cribriform plate	NSF	MRI
10	50	Poorly differentiated carcinoma	5.5	Centered on cribriform plate	NSF	MRI
11	62	Squamous cell carcinoma	8.9	Invasive along cribriform plate	NSF	CT
12	31	Esthesioneuroblastoma	3.0	Centered along cribriform plate	NSF	CT
13	68	Melanoma	3.5	Invasive through cribriform plate	NSF	MRI
14	38	Fibrodysplasia	5.0	Involving cribriform plate	NSF	CT
15	60	Sinonasal carcinoma	4.1	Invasive along cribriform plate	NSF	MRI
16	51	Olfactory neuroblastoma	5.0	Centered on Cribriform plate	NSF	MRI
17	47	Meningioma	3.0	Based along cribriform plate	None	MRI
18	62	Atypical meningioma	7.0	Frontal bone (CN I preserved)	None	MRI
19	33	CN I schwannoma	2.8	CN I directly involved	None	MRI
20	80	Poorly differentiated carcinoma	5.6	Invasive through cribriform plate	Fat	CT
21	64	Olfactory neuroblastoma	5.8	Invasive along cribriform plate	None	MRI
22	57	Metastasis	5.7	Based along cribriform plate	None	MRI
23	84	Meningioma	6.5	Based along cribriform plate	None	MRI

Abbreviations: CN, cranial nerve; CT, computed tomography; MRI, magnetic resonance imaging; NSF, nasoseptal flap.

^aTumor size in largest dimension.

^bLocation in relation to cribriform plate or olfactory nerve involvement.

^cModality used to assess frontal sinus outflow tract patency.

The trauma population was predominantly gunshot wounds followed by blunt trauma secondary to motor vehicle collision. Three trauma patients had displacement of the frontal sinus fracture ranging from 5.47 to 14.95 mm. One patient, case 2, had a nondisplaced frontal sinus and 5 mm displaced cribriform plate fracture from blunt injury that was initially managed conservatively; however, the patient presented to clinic after discharge with pneumocephalus and rhinorrhea. The two remaining traumas had severely comminuted fractures that would not allow for the measurement of displacement. All fractures had CSF leak and involvement of the cribriform plate requiring olfactory nerve dissection as in step 4. All trauma patients underwent the compartmentalization procedure alone. All patients underwent postoperative CT scans to assess frontal sinus structure and function ( Fig. 2 ). All but one patient met the standard clinical and imaging follow-up of at least 2 months.

Fig. 2.

Preoperative ( A and B ) and postoperative ( C and D ) computed tomography scan images from a compartmentalization repair. Note the preservation of the nasofrontal outflow tract and resolution of the fluid air level in the frontal sinus at the time of injury ( A , C ).

One lumbar drain was placed with the initial trauma patient in this series for CSF diversion during the early healing process. The benefit of the additional drainage was questionable and was not used in subsequent compartmentalization repairs. An additional trauma patient had a pre-existing shunt for a diagnosis of pseudotumor.

Tumor patients had a variety of diagnosis detailed in Table 3 . All but one tumor, case 18, had tumor invading the cribriform plate or direct involvement of the olfactory nerve. For these cases, the olfactory nerve was dissected as in step 4. Case 18 was an invasive meningioma requiring bony removal in the context of large frontal sinus. For this case, the posterior–inferior limit of the layered closure was anterior to the olfactory nerves and the olfactory nerve was preserved. For cases with significant nasal components that required transnasal endoscopic removal of tumor, additional ventral skull base repair with a vascularized nasoseptal flap or fat packing was used ( Table 3 ). No CSF diversion was utilized for any of the tumor patients. All patients underwent postoperative CT or MRI following surgery. Four patients did not reach the standard follow-up timing for delayed imaging of 6 months.

The patient series had two major acute complications in the same patient: anemia and suspicion for meningitis. The patient had undergone resection of sinonasal carcinoma requiring removal of the anterior fossa floor and use of nasoseptal flap. Post-op day 1, the patient was noted to have anemia (Hgb 5.7) that was resolved with two units of packed red blood cells. Patient was also noted to have concern for acute or chronic sinusitis on imaging and given a short course of Augmentin prior to discharge. On post-op day 19, the patient was readmitted for concerns of meningitis. Extensive workup including blood and CSF cultures did not result in an infectious agent. The patient was treated with 14-day course of intravenous antibiotics and then continued on oral antibiotics for sinusitis. Ultimate diagnosis by the consulted infectious disease physicians was “unlikely meningitis” in favor of a “parameningeal reaction.”

All chronic complications related to the procedure were mild including five patients (21.7%) with headache or incisional pain and one patient (4.3%) with maxillary sinusitis. Three of the headaches or incisional pain had complete resolution at 1, 2, and 4 months. The two remaining patients had either mild chronic headache controlled on over the counter medications or returned to regular headache baseline associated with their preexisting pseudotumor. The chronic maxillary sinusitis was diagnosed via clinical symptoms and endoscopic evaluation. Treatment was initially attempted with multiple rounds of oral antibiotics. At 5 years post-op, the patient continues to require intermittent treatment with gentamycin nasal flushes ( Table 4 ).

Table 4. Denoting complications and outcome of complications.

Complications
Case	Age	Dx	Acute complications	Delayed complications	Complication outcome
1	37	Trauma		Headache	Chronic mild headaches
3	19	Trauma		Headache	Resolved at 4 months
6	50	Trauma		Headache	Return to pseudotumor baseline at 7 months
8	43	Tumor		Sinusitis; incisional pain	Chronic sinusitis managed on gentamycin and saline flushes; incisional pain resolved at 2 months
10	50	Tumor	Parameningeal reaction; anemia		Resolved with antibiotics at 1 month; anemia resolved with 2 units of pRBC
17	47	Tumor		Incisional pain	Resolved at 1 month

Abbreviations: Dx, diagnosis category; pRBC, packed red blood cells.

Discussion

Treatment for a frontal sinus fractures has been a major challenge in healthcare for millennia. The first documentation of a treatment algorithm can be found in the Edwin Smith papyrus written in 1501 BC. Imhotep described two patients who had fractures that included the frontal sinus. At that time, they recognized the importance of the mucosal layer and the potential sequelea of frontal sinus damage including deterioration from infection. Albeit one of the two treatment methods included simple wrapping and repetition of a healing chant. ^{10 11} Contemporary frontal sinus repair requires addressing acute pathology, prevention of short- and long-term complications, and restoration of cosmesis. ^{6 12} Reidel documented the first neurosurgical technique when he described the ablation method for frontal sinus fractures. ^{6 13} Meaningful improvements in neurosurgical technique came by way of Nadel and Klein, who reported a series of single surgery debridement and repair, and Donald et al, with the first description of cranialization in 1982. ^{8 14}

Over the last three decades, the literature regarding frontal sinus fractures has focused on treatment methods based on the type of fracture. ^{5 6 13 15 16 17 18} There are two major series that outline treatment algorithms: Rodriguez and Bell. ^{5 6 19} The Bell algorithm evaluates each fracture based first on displacement, then nasofrontal duct injury, and finally posterior table injury. Rodriguez proffered a simpler algorithm that focused heavily on nasofrontal tract outlet obstruction. Bell and Rodriguez ultimately endorse a conservative approach to frontal sinus fractures, meaning fractures with outflow tracts intact and minimal displacement should be observed and any outer table injury should be reconstructed assuming the outflow was intact.

Despite the focus on minimizing intervention, there are still patients 18.1 to 51.9% who require a major intervention due to displaced posterior table injury, outflow tract obstruction, overt rhinorrhea, or need for emergent intracranial intervention. In these situations, the literature for the best surgical technique is relatively sparse. ^{5 13 17 20 21} The majority of frontal sinus literature is found in maxillofacial, plastics, or ENT journals. Though they provide a detailed analysis of indications for the various management strategies and outcomes from those decisions, the specifics of the intervention that would be required for neurosurgery are limited to naming the technique, that is, cranialization or obliteration. To date the most popular and successful technique used in the neurosurgical repair of frontal sinus fractures is cranialization, though obliteration is also undertaken. ^{6 8 14} The contemporary techniques discussed in various papers still result in a 8.2 to 26.1% complication rate including 3 to 10% rate of CSF leak. ^{5 7 13 17 22} Furthermore, the contemporary procedures all result in the loss of a normally functioning frontal sinus that can lead to long-term consequences such as mucocele. There are also situations in which the frontal sinus is intentionally breached in the course of a surgical approach; for elective surgeries, the standard for the prevention of complication is significantly higher. Thus, there appears a need for continued innovation with regard to repair of frontal sinus injury to work closer to the goal of zero morbidity.

A novel approach to sinus injury repair described by Freeman and Winston advocates for the preservation of the sinus. ⁷ This addresses the next frontier in frontal sinus repair by allowing for continued function of the disrupted anatomy. The benefits of frontal sinus preservation are somewhat theoretical but include restoration of normal anatomy, drainage pathway for mucous, and avoidance of trapping bacteria in the repair. ^{6 12} However, with a patent nasal frontal outflow tract there must be sufficient separation between the frontal sinus and the intracranial space to prevent meningitis, empyema, and CSF leak. This is seen in the Freeman and Winston series wherein 4/27 (14.8%) patients who underwent surgery had complications of CSF leak. Notably, with their technique all CSF leak patients recovered with bed rest and a lumbar drain. In the present series, there were no leaks throughout 23 patients who underwent compartmentalization. There was one patient in the current series who presented with symptoms concerning for meningitis. This patient was treated and recovered while on antibiotics; no infectious agent was isolated. Ultimately, the complication rate of our series was 4.5% including no CSF leaks, which is favorable as compared with the existing literature for cranialization and obliteration.

Endoscopic approaches for traumatic or iatrogenic frontal sinus injury have been steadily evolving since the first case series was reported by Woodworth in 2005. ²³ Shortly after a case series describing the reduction in anterior table fractures with a transnasal or trephine-assisted endoscopic approach was reported by Steiger et al. ²⁴ They reported five patients treated over 2 years all of whom noted good cosmetic repair with continued sinus function. Over the next decade, advancements in materials for fracture reduction and CSF leak repair allowed for posterior table repair. At that time, Chabaan and Woodworth et al reported 13 patients with an average defect of 13.5 × 4.5 mm; all with CSF leaks. ²⁵ Of the 13 patients, 1 required revision of the sinusotomy. They reported resolution of all CSF leaks and all patients had patent frontal sinus. Recently, the Woodworth group published a call to change the current dogma on the repair of traumatic frontal sinus injury. ²⁶ They report a new prospective case series of 46 patients who underwent endoscopic frontal sinus repair after trauma with notable success. The majority were posterior table injuries and repairs were done with porcine submucosal graft, nasoseptal flap, cadaveric pericardium, or a combination. CSF diversion was used in 14/46 of the patients. They report no failures with regard to CSF leak, and no need for further repair and two complications: scarring of the outflow tract and periorbital cellulitis. The remarkable success of this group prompts significant consideration to using endoscopic approaches for frontal sinus repair. However, the authors note that the treatment of frontal sinus injury is complicated by the complex nature of injury mechanisms and wide variety of repair options. For example, endoscopic repair is impractical when open approaches are required for traumas with underlying injury or tumor removal. Furthermore, with the wide variety of endoscopic approaches, a given neurosurgeon may not have an ENT available who is adequately trained to perform an endoscopic repair. In these examples, compartmentalization offers an open approach to repair and preserve the frontal sinus.

The compartmentalization technique evolved from the prior strategies of severe frontal sinus injury repair. A pedicled periosteal flap covering the frontal sinus is a key first step adapted from prior descriptions of frontal sinus repair. ^{5 8 12 21 27 28} By retaining the vascularization of the tissue, the blood supply can aid in resisting infection and allow for antibiotics to directly access the site of separation between the frontal sinus and intracranial space. ²⁹ Allograft is then used to reinforce the periosteum creating a dual layer that further differentiates the extracranial and intracranial compartments. Among the benefits of a dual layer separation is the reduced need for watertight dural closure. In the present series, the dura was approximated using 4–0 braided nylon suture; however, a watertight seal was not regularly achieved. The Freeman and Winston sinus preservation technique endorsed rigorous dural closure with free flaps of periosteum, but their surgical patient series showed a 14.8% leak rate; all of whom were treated with lumbar drain alone. Additionally, the technique was only used in minimally displaced fractures, while the present series includes fractures with displacement up to 15mm. The early experience with this technique has thus far shown no CSF leak or rhinorrhea. Further notable is that the technique does not demand any specialized technique, novel equipment, or high technical demand. Thus, any practicing neurosurgeon who encounters severe frontal sinus injuries could potentially employ the compartmentalization technique.

There are limits to the technique. Principally, compartmentalization requires a pedicled periosteal flap. In cases of redo craniotomy where the periosteum is already disrupted or scarred, other techniques such as obliteration or cranialization should be used. Additionally, there is almost certain loss of olfactory sensation if there is the need to dissect to the level of the planum sphenoidale. For fractures that extend beyond the planum sphenoidale, that is, sella turcica or clivus, other approaches should be considered. Compartmentalization is also a more invasive procedure and should be undertaken with appropriate indications as described by Rodriguez et al and Bell et al. ^{5 6} Further, the early case series presented here is limited in patient numbers and complications such as CSF leak are likely to present with sufficient volume. There are also limits to determining frontal sinus function by relying on imaging to assess the patency of the nasofrontal outflow tract. Often, long-term post-op imaging cannot be obtained due to the nature of trauma populations and mortality with severe tumor pathologies. More delayed outcomes are also necessary to truly define the risk of mucocele. Regardless, the series demonstrates the overall safety of the procedure and potential for improvement in patient outcomes.

Conclusion

Compartmentalization represents the latest evolution of open technique in frontal sinus repair. The method allows for continued function of the frontal sinus after repair and early case series shows low complication rates. Further, the intervention only requires basic neurosurgical competency allowing for widespread use of the method. Further study in more patients and comparison trials are needed to verify the efficacy and safety of compartmentalization in comparison to traditional techniques.

Funding Statement

Funding Sources None.