Summary

Both the surgical indications and the surgical techniques appropriate for the frontal sinus have long been a matter of controversy. However, over the last few years, a consensus seems to be emerging that endonasal mucosa-sparing approaches provide the most direct and natural pathway to the frontal sinus. A stepwise approach to the frontal sinus is proposed, with distinct indications for the various techniques.

Introduction

The persisting interest in frontal sinus surgery testifies to our inability, after more than 2 centuries, to come up with a simple, safe, and consistently effective approach. The anatomy of the frontal sinus provides the key to this conundrum: a large bony cavity covered with mucus-producing epithelium that drains in the nose via a narrow, convoluted recess lined by mucosa with a high propensity for scarring. Not unlike attempts to reopen and maintain the patency of other narrow mucosal-covered pathways (the subglottic trachea comes to mind), every approach resembles a battle against the natural tendency of circumferential mucosa to heal by secondary intention and granulation.

Although it would be elegant to present the history of surgical approaches to the frontal sinus as a smooth evolution from more traumatic (external) procedures to less invasive (intranasal) ones, the reality is different. The history of frontal sinus surgery is more like a pendulum, gravitating from external to transnasal to external and then back to transnasal approaches. The first approach to the frontal sinus was via the most direct and less dangerous route, that is, the anterior frontal plate. In 1750, Runge apparently performed the first frontal sinus obliteration procedure,1 although the first published report of frontal sinus surgery appears in The Lancet more than a century later, in 1870.2 In that report, Wells described a combined external and intranasal drainage procedure for a patient with a frontal sinus mucopyocele with intracranial extension. Ogston3 and Luc4 described almost simultaneously during the last decade of the 19th century a combined external trephination and curettage of the nasofrontal duct ( Fig. 22.1 ).

This procedure was predictably associated with high rates of failure related to scarring and subsequent stenosis of the frontal sinus drainage pathway. The next approach, as introduced by Riedel, was to the opposite direction: complete removal of the anterior wall of the frontal sinus combined with stripping of all frontal sinus mucosa and obliteration of the nasofrontal duct from above. It was often successful; however, it was associated with what was described as “an un-sightly forehead deformity that could only be out-weighted by the threat of death.”5 Hence, intranasal approaches were reintroduced in the early 20th century by Schaeffer,6 Halle,7 and Good.8 Inadvertent intracranial entry associated with significant morbidity or even mortality was described; as a result, intranasal approaches gained notoriety and were abandoned. A procedure initially described by Ritter9 and Jansen10 appeared to offer the best of both worlds and was reintroduced and popularized in the United States and United Kingdom (where it was known as the Lynch-Howarth11 procedure) ( Fig. 22.2 ). In the procedure, the frontal sinus was approached via an incision in the intercanthal area; subsequently, the frontal sinus floor and the anterior ethmoids were removed through the lamina papyracea. However, the removal of the medial orbital wall often led to orbital fat prolapse within the frontal duct, scarification, and ultimately failure. Over the next few years, a reluctant consensus emerged: once compromised, the nasofrontal pathway patency cannot be reestablished. This signaled the return to a more radical procedure, namely, the osteoplastic flap procedure. The osteoplastic flap had been described in the late 19th century by Brieger12 as well as Schoenborn and Winkler,13 although concerns about reapproximation of the bony flap and osteomyelitis limited its appeal at the time. Thus, when Goodale and Montgomery reintroduced and popularized this technique,14 it remained the gold standard for most of the remaining 20th century. The advent of three-dimensional (3D) imaging (computed tomography [CT] and magnetic resonance imaging [MRI]) improved our understanding of frontal recess anatomy, while the advent of the operating microscope and endoscope subsequently gave new impetus to transnasal frontal sinus surgery: Draf in 199115 systematized the endonasal (microscopic and endoscopic) approaches to the frontal sinus and described for the first time the type 3 medial drainage procedure (also named endoscopic Lothrop and reintroduced by Gross in 199516) as a valid alternative to frontal sinus obliteration. Although by no means a panacea, there is now significant experience17,18 testifying to its efficacy and safety. It appears that the pendulum has moved, this time irreversibly, toward intranasal approaches.

Classification of Frontal Sinus Approaches/Nomenclature

The first comprehensive classification of endoscopic frontal sinus approaches is credited to Draf, who described in 1991 “the Fulda concept”15 ( Table 22.1 ). In that article, he presented a graduated endonasal approach to the frontal sinus. Type I is a complete ethmoidectomy, with removal of the bulla and suprabullar cells, without violating the mucosa of the frontal recess; type IIa ( Fig. 22.3a ) is the enlargement of the frontal outflow tract with the removal of the agger nasi (variably described as frontal sinusotomy or the “uncapping the egg” technique); type IIb (unilateral frontal sinus drill-out”; Fig. 22.3b ) consists of further enlargement of the frontal sinus ostium by drilling of the floor of the frontal sinus from the lamina papyracea to the septum; and type III ( Fig. 22.3c ) (modified endoscopic Lothrop,16 bilateral frontal sinus drill-out,19 or median drainage procedure20) is the complete drill-out of the floor of the frontal sinus, frontal beak, intersinus septum, and an adjacent part of the nasal septum. Modifications of these techniques include transseptal frontal sinusotomy21 (nasal septum translocation facilitating frontal sinus entry) and the frontal sinus rescue procedure22 ( Fig. 22.4 ), essentially a frontal sinusotomy with complete excision of the bony middle turbinate remnant and coverage of the denuded bone with a mucoperiosteal flap. A similar technique is the axillary flap introduced by Wormald,23 consisting of creating a mucosal flap from the axilla of the middle turbinate, removing its bone, and covering it with the harvested mucosa. May and Schaitkin described in 199524 a classification similar to that described by Draf, with nasofrontal approach type 1 corresponding to the Draf I procedure, type 2 to Draf IIa, type 3 to Draf IIb, and type 4 to Draf III.

Indications

Specific indications for the different types of endoscopic approaches to the frontal sinus ( Table 22.2 ) are a subject of controversy at least as intense as surgical technique. Although in certain cases, surgery is clearly indicated, in the case of a (typical) patient with clinical and endoscopic evidence of chronic rhinosinusitis (CRS) and opacified frontal sinuses on the CT, there are different schools of thought. We advocate surgical conservatism in primary cases, as we believe that, once the inflammation of the ostiomeatal complex and the maxillary and ethmoid sinuses resolves, frontal sinuses will follow.25 However, there are some absolute indications for surgery in the frontal sinus, in the absence of chronic frontal rhinosinusitis.

Indications for Surgery on the Frontal Sinus Excluding CRS

1. 
   

   Complications of frontal acute rhinosinusitis (ARS) or CRS

   
   
   
   
   1. 
      

      Endocranial

      
      
   2. 
      

      Potts puffy tumor

      
      
   3. 
      

      Frontocutaneous fistula

   
   
2. 
   

   Frontal sinus mucocele

   
   
3. 
   

   Benign tumors of the frontal sinus

   
   
   
   
   1. 
      

      Osteoma (symptomatic or large/growing)

      
      
   2. 
      

      Inverted papilloma

      
      
   3. 
      

      Fibrous dysplasias (symptomatic or growing)

      

      

      

   
   
4. 
   

   Scarred or permanently obstructed frontal recess (orbital prolapse after decompression surgery, post-trauma, post–skull base reconstruction) with symptoms (CRS or recurrent allergic rhinitis)

   
   
5. 
   

   Fracture of the anterior frontal plate (for assessment)

   
   
6. 
   

   Access to the anterior skull base

   
   
   
   
   1. 
      

      Anterior skull base tumors: transcribriform approach (Draf III)

      
      
   2. 
      

      Cerebrospinal fluid (CSF) leak repair of anterior skull base defects

The complicated anatomy of the frontal recess and its proximity to critical structures render frontal sinus surgery challenging and potentially dangerous. However, they do not render its pathology or its mucosa qualitatively different from that of other paranasal sinuses. CRS does not run a more indolent course in the frontal sinus. Hence, in the same way that a chronically obstructed and nonfunctional ostiomeatal complex is managed surgically, a chronically obstructed, scarred frontal recess refractive to adequate medical treatment should be surgically drained. However, there are specific anatomical considerations pertaining to the frontal sinus. It has been shown through the groundbreaking work of Draf/Messeklinger that the frontal sinus drains through a complex pathway that includes the anterior ethmoids. The first step in establishing the patency of the frontal recess should be an anterior ethmoidectomy (a Draf I15 or nasofrontal type 1 procedure).24 If that does not resolve inflammation, one should proceed to surgically draining the frontal sinus.

Indications for endoscopic frontal sinus approaches

Procedure	Relative indications	Absolute indications	Contraindications
Draf I	1. Primary ESS for frontal sinus CRS
Draf IIa	Secondary ESS for frontal CRS Frontal sinus mucocele Primary ESS for frontal sinus CRS (ASA triad, CF, massive polyposis) Diagnostic for frontal sinus tumors Acute frontal sinusitis with intracranial, orbital, or cutaneous complications
Draf IIb	Limited/unilateral benign tumors of frontal sinus (inverted papilloma, mucocele, etc)	Access for frontal/skull base tumors
Draf III	Revision ESS for frontal CRS (ASA triad, CF, massive polyposis)	Access for frontal/skull base tumors Surgery for frontal sinus mucoceles that cannot be reached via a unilateral approach	Tumors involving the anterior frontal plate/extreme lateral tumors
ASA, acetylsalicylic acid; CF, cystic fibrosis; CRS, chronic rhinosinusitis; ESS, endoscopic sinus surgery; IP, inverted papilloma.

Surgical Anatomy of the Frontal Recess

Traditionally, the uncinate process was considered as the “key” to the frontal sinus, with the three different types of uncinate attachment superiorly (to the orbit, skull base, and middle turbinate) defining the three different types of frontal sinus drainage ( Fig. 22.5a, b ). However, Wormald27 showed that the location of the superior insertion of the uncinate process is often a secondary effect of the pneumatization and morphology of the agger nasi ( Fig. 22.5c ), the almost ubiquitous (92%) anteriormost cell of the ethmoid labyrinth. Most of the agger nasi cell is anterior to the uncinate, but the posterior half of the agger nasi merges with the superior extension of the uncinate process. Drainage pathways are thus defined by the combined morphology of the uncinate process and the agger nasi. In addition to the agger nasi cell, there are several cells that may narrow the recess and were classified by Kuhn.28 These are also discussed in Chapter 7 (see Figs. 7.34 , and 7.35 ) and will only be briefly presented here:

• 
  

  Type 1 cell: Single cell above the agger nasi; posterior wall is free partition within the frontal recess, not attached to skull base ( Fig. 22.6 )

  
  
• 
  

  Type 2 cell: Two or more cells pneumatizing above the agger nasi; posterior wall is free partition within the frontal recess, not attached to skull base ( Fig. 22.7 )

  
  
• 
  

  Type 3 cell: Single large cell pneumatizing above the agger nasi, extending well into the frontal sinus ( Fig. 22.8 )

  
  
• 
  

  Type 4 cell: Isolated cell within the frontal sinus; alternatively, cell occupying more than 50% of the frontal sinus height anteroinferior wall is the anterior wall or floor of the frontal sinus ( Fig. 22.9 )

  
  
• 
  

  Frontobullar cell: Ethmoid cell above the ethmoid bulla; the posterior wall is skull base; anterior wall pneumatizes within the frontal sinus (see Fig. 22.6 )

  

  

  
  
• 
  

  Suprabullar cell: Ethmoid cell above the ethmoid bulla; inferior wall is ethmoid bulla, and anterior wall does not extend within the frontal sinus (see Figs. 22.6 and 22.7 )

  
  
• 
  

  Interfrontal sinus septal cell: Pneumatization of the frontal sinus septum, draining in the frontal recess and associated often with pneumatized crista galli

This classification remains pertinent, although it has been suggested that Kuhn type 4 cells always have a drainage pathway and hence should be redefined as cells that occupy > 50% of the frontal sinus height rather than cells exclusively originating within the sinus.

It is clear from the anatomy that approaching the frontal recess brings one very close to several structures that should not be violated: the anterior skull base, the anterior ethmoid artery (which is usually the posterior limit of the frontal recess and separated from the frontal ostium by about one cell width), and the orbit. The “building-block” approach (a way of mentally visualizing in 3D the frontal recess and the surrounding individual cells), as introduced by Wormald27 ( Fig. 22.12 ), can be a useful teaching tool and works ideally in well-pneumatized sinuses (albeit not always in completely opacified ones). A similar 3D reconstruction or even an endoscopic “run-through” can be produced by DICOM readers such as OsiriX (open source software developed by Pixmeo, Geneva, Switzerland, www.osirix.com) ( Figs. 22.13 , 22.14 , and 22.15 ).

In many of these conditions, the surgeon navigating the recess can find his or her way by being acutely aware of the location of the lamina papyracea and the cribriform plate, including its anteromedial extension, and utilizing the only safe structure of the recess: the thick, unyielding bone of the frontal beak.

It is important to keep in mind that once the frontal recess is violated, there is no way back—a small hole drilled or punctured through the frontal recess cells is a one-way path to stenosis and iatrogenic frontal sinusitis. The least invasive surgery of the frontal recess should always include complete removal of the agger nasi cell.

Instrumentation

Operating endoscopically in the frontal recess is challenging, not only because of the complex anatomy but also because of ergonomic considerations. The surgeon is working at a vertical angle, with blood trickling from the operating field above and onto the endoscope, potentially obscuring his or her view. Although there is an argument for the use of a 0-degree endoscope (direct field of view and ease of instrumentation),29 its use necessitates the routine removal of the middle turbinate axilla.8 In most of our operations, we use the 30-degree endoscope, which, in our view, represents the ideal compromise between superior visualization and ease of instrumentation. Purpose-designed angled instruments are vital: frontal (J) curettes of various sizes and angles, as well as a 15-degree 5-mm diamond drill, a 40-degree 2-mm drill, or a thicker, 55-degree 3.6-mm angled drill (for thicker bone) or a 70-degree tapered diamond drill (for more delicate work) operating at 12,000 rpm or rough diamond, new-generation, 75,000-rpm drill. These can be much more efficient and provide better handling, although at a cost of reduced angle (up to 20 degrees). We routinely use intraoperative imaging for confirmation, also recently we have been using, in certain cases, a 3D scope (Visionsense Corp, New York, NY). The improved sensation of depth, however, is (for the time being) associated with reduced image resolution.

With this proviso, in more than 1000 frontal sinus procedures, we have not found the use of frontal sinus trephination to add to our approach.

Approaches: Anatomical and Technical Description

Surgical Rationale

In concordance with a significant body of literature,20 we feel that drilling of the frontal recess is associated with high rates of scarring and stenosis; thus, in most cases, one must follow the “all or none” philosophy: the use of a drill in the frontal recess with the resultant unavoidable mucosal damage normally mandates proceeding to a complete Draf III. We feel that what was described as type IIb drainage by Draf or a unilateral frontal sinus drill-out procedure has few indications, with the exception of tumors or the presence of neo-osteogenesis blocking the recess. In patients with persistent frontal sinusitis, we perform a complete clearance of the cells obstructing the frontal recess by removing the agger nasi and any obstructing suprabullar/Kuhn type 2 cells, trying to get the biggest possible exposure of the frontal ostium, while preserving as much of the mucosa as possible. It is our feeling that if the surgeon has to use a drill to remove bone, then he or she is required to go all the way to avoid stenosis and create the biggest possible ostium by performing a median drainage procedure (Draf III).

Draf IIa: Frontal Sinusotomy

We start the approach to the frontal sinus with a complete uncinectomy ( Fig. 22.16 ; see Video 22, Draf IIa ). We use 45-degree up-cutting Blakesley forceps to remove the most superior attachment of the uncinate process, where it fuses with the axilla of the middle turbinate. In most cases, the uncinate process fuses superiorly with the posteromedial wall of the agger nasi to attach to the lamina papyracea ( Fig. 22.17 ). In such cases, a curette should be inserted posteriorly to its attachment to completely remove the agger nasi (see Fig. 22.10 ). Subsequently, the surgeon is faced with the axilla of the middle turbinate and several half-opened cells facing the skull base (see Fig. 22.11 ). A good description of the action performed is “uncapping the egg.”

Note

Conformation of being in the frontal sinus can be obtained by

1. 
   

   Navigation ( Fig. 22.18 )

   
   
2. 
   

   Transillumination: approaching the light source to the recess and dimming the lights usually allows the transillumination of the anterior frontal sinus wall ( Fig. 22.19 ).

   
   
3. 
   

   Removing the curette and by replacing it on the nose assessing the position of its end

   
   
4. 
   

   Palpating the anterior frontal beak and seeing clearly the roof of the orbit

   
   
5. 
   

   Visualizing the anterior ethmoid artery posterior to the ostium

   
   
6. 
   

   Checking the curve of the skull base, remembering that the frontal sinus starts at the point of maximum curve of the anterior skull base

Referring back to the CT and mentally reconstructing the anatomy is often useful at this stage. Any Kuhn frontal cells should be removed to gain uninterrupted entry into the frontal sinus. A rule of thumb is that the frontal ostium is usually more medially higher and more anterior than most people would tend to visualize with a 30-degree scope. The orbital wall (as followed up from inferiorly), the anterior ethmoidal artery (lying one cell behind the opening to the frontal recess), and the very thick bone of the frontal beak are all landmarks that are helpful in identifying the recess. The frontobullar and suprabullar cells may also need to be removed using “giraffe” 90-degree angled through-cutting forceps and 40- to 70-degree microdebrider blades to expand the neo-ostium posteriorly. It is important to preserve the mucosa all around if possible and if not, at least > 50% ( Fig. 22.20 ).

Draf IIb

The Draf IIb (or nasofrontal type 3 or unilateral frontal sinus drill-out) procedure consists of, in addition to the removal of agger nasi and Kuhn cells, thinning of the frontal beak and the medial wall of the frontal recess. Although it results initially in the maximum possible opening (from the nasal septum to the lamina papyracea), the mucosal damage produced by the drill is associated with a high risk of stenosis, especially in patients with CRS and chronically inflamed mucosa. This may be countenanced in patients with a wide anteroposterior and lateral nasofrontal diameter; however, in most patients we feel that, once a drill is used in the frontal recess, a Draf III procedure is often unavoidable. Hence, we rarely use Draf IIb for patients with CRS, although we have found it useful for patients with tumors of the frontal sinus like inverted papilloma and sometimes mucoceles that cannot simply be marsupialised (see Video 23, Marsupialization of a Two-lobed Frontal Mucocele Medial of Middle Turbinate, and Video 24, Draf IIb for Frontal Inverted Papilloma ). Indeed, in such patients where the mucosa is not pathologic, the chance of restenosis is less.

Draf III

A complete anterior and posterior ethmoidectomy is performed (see Video 25, Draf III for Large Frontal Sinus Osteoma Removed Endoscopically ), including all ethmoid cells, as their remnants can compromise the operation′s success, both by limiting the frontal sinus outflow pathway and by acting as a nidus of persisting inflammation. A Draf IIa (simple frontal sinusotomy) is performed ideally on both sides ( Fig. 22.21 ), although neo-osteogenesis associated with chronic inflammation may render simple frontal sinusotomy without the use of a drill impossible. Keeping in view the frontal ostium at all times during the procedure is crucial, as it is the most important safety landmark. An incision through the anterosuperior septum is subsequently performed either with a no. 15 blade or with monopolar cutting needle diathermy, starting just anterior to the frontal ostium and advancing 2 cm inferiorly and anteriorly ( Fig. 22.22 ). Its mucosa is then removed with a shaver and the septal cartilage with cutting Blakesley forceps or scissors. The opening in the septum thus created should be wide enough to allow visualization of both frontal ostia but not more posterior than the first olfactory fibers. The axilla of the middle turbinate is removed ( Fig. 22.23 ). With a 40- or 60-degree shaver, the soft tissue and mucosa between the two frontal ostia are then removed, exposing the thick, hard bone of the nasal beak. Using a 40-degree cutting finesse bur, the nasal beak is subsequently drilled in a rolling, smooth, anteromedial direction, keeping in mind that the skull base extends anterior in the midline ( Fig. 22.24 ). The drilling is continued superiorly, removing the intersinus septum as high as possible, and anteriorly, up to the skin, aiming for the largest possible, smooth-edged, common outflow pathway, as defined by the lacrimal bones laterally, the anterior skull base posteriorly, and the nasal skin anteriorly. Alternatively, the frontal beak is removed outside-in with a 15-degree 5-mm diamond drill, as described by Harvey. We aim to preserve a thin (< 1 mm) bone layer at the skin side to prevent skin collapsing in the neo-ostium. The opening proceeds until the drill can be felt under the skin, both anteriorly on the nasal bridge and laterally on the frontal process of the maxilla, removing as much of the beak as possible, aiming for a neo-ostium where the anterior frontal wall can be easily visualized with a 30-degree scope ( Fig. 22.25 ).

We aim to preserve as much mucosa as possible, especially laterally and posteriorly. However, the anterior projection of the anterior skull base in the midline usually requires some very careful drilling after identification of the first olfactory nerve, to produce the maximal anteroposterior diameter ( Fig. 22.26 ).

Stay updated, free articles. Join our Telegram channel

Join