Endoscopic Transseptal Approach to Frontal Sinus Disease

Abstract

This paper describes an endoscopic transseptal approach to identify and access the frontal sinus and reviews the clinical cases. Between May 2004 and July 2010, endoscopic modified Lothrop procedure (EMLP) with transseptal approach was performed on sixteen patients. The indications for EMLP were complicated frontal sinusitis or cyst, revision surgery for failed frontal sinusotomy or Lynch procedure, or trauma cases. The first step of this procedure was to open a window in the bilateral anterior portion of the middle turbinates and nasal septum. The nasal septum, which could be observed through the window, should be the landmark of the midline during the surgery. A drill bur was raised up just behind the nasal bone along the midline of the nose. After the bilateral frontal sinuses and their posterior walls were confirmed, the interfrontal septum was removed superiorly. We reviewed the clinical records of patients who underwent the EMLP with transseptal approach. We have managed sixteen patients in this fashion. Neither intracranial nor orbital complications were encountered during or after surgery. Endoscopic transseptal frontal sinus surgery is simple to perform, and does not cause severe complications.

Introduction

Endoscopic sinus surgery (ESS) has become a frequently performed procedure for patients with chronic rhinosinusitis. However, ESS for the frontal sinus remains problematic, because of the complex anatomy and large anatomical variation of the frontal recess [1, 2]. More extensive procedures such as the endoscopic modified Lothrop procedure (EMLP) or Draf III procedure is technically demanding but still has a difficulty of surgical access associated with greater risk [3–5]. Thus, a simple technique for safe and effective frontal sinus surgery is critical. In this paper, we describe a technique of EMLP to identify and access the frontal sinus via an endoscopic transseptal approach.

Patients and Methods

Between May 2004 and July 2010, EMLP with transseptal approach was performed on sixteen patients. The indications for EMLP were complicated frontal sinusitis or cyst, revision surgery for failed frontal sinusotomy or Lynch procedure, or trauma cases.

All procedures are performed with the patient under general anesthesia using standard ESS equipment including an XPS drill system (Medtronic Xomed, Jacksonville, FL, USA). The uncinate process is removed. An attempt is made to clear the frontal recess and identify the frontal ostium in the both sides of the nose. Then transseptal frontal sinusotomy is mainly performed under a 4-mm 70° endoscope.

The first step of this procedure is to open a window in the bilateral anterior portion of the middle turbinates and nasal septum using a drill bur immediately caudal to the frontal recess and posterior to the nasal bone: the removal process is begun on one side of the middle turbinates (Figs. 1a, 2a), then proceeded towards the nasal septum and finally the opposite side of the middle turbinate around the most anterior point behind the nasal bone (Figs. 1b, 2b). The main author generally begins the removal process on the right site of the middle turbinate (Figs. 1a, 2a). If the recess or the ostium is not clearly identified, fenestration is first made on the anterosuperior part of the middle turbinate.

Fig. 1.

Intraoperative views with a 70° endoscope. Angles of the endoscopic view are indicated by white arrows in the left row. Illustrations of intraoperative views in a representative case are indicated in the right row. The upper side of the intraoperative views is the anterior side of the nose, the lower side is the posterior and the deeper side is the rostral. a The anterosuperior part of the right middle turbinate (RMT) is removed by a drill burr. The obstructed right frontal sinus ostium (FSO) is then seen. b In succession the nasal septum (NS) and the Opposite side of the middle turbinate are removed and the window is widened out inferiorly. Through the window, the midline of the nose is indicated by the remaining part of NS between both the RMT and left middle turbinates (LFT). The drill is advanced along NS just behind the nasal bone (NB) to the frontal sinus floors. c The fenestration is extent up towards the interfrontal septum (IFS), the frontal sinus floors are penetrated and frontal sinuses are unified. D drill burr, FSO right frontal sinus ostium, IFS interfrontal septum, LFS left frontal sinus, LMT left middle turbinate, RMT right middle turbinate, NB nasal bone, NS nasal septum, RFS right frontal sinus

Fig. 2.

The sagittal plane of surgical scheme and the preoperative computed tomographic axial images of the case. Each column corresponds to that of Fig. 1. The surgical removal regions are indicated by dotted lines. AP the anteroposterior diameter between the anterior skull base and nasal bones or frontal bones, FS frontal sinus, MT middle turbinate

The window can be extended down inferiorly by removing the part of the nasal septum and the bilateral middle turbinates for easy access and for the wide view of surgical field (Figs. 1b, c, 2b, c). Through this fenestration, the operator identifies the midline of the nose indicated by the remaining part of the nasal septum clearly viewed between both the sides of middle turbinates (Figs. 1b, 2b). The window is extent up superiorly along the midline just behind the nasal bone keeping its widths less than the limitation that is equal to the anteroposterior diameter between the anterior skull base and nasal bones measured in the preoperative CT image (Fig. 2c). The surgical instruments must skirt the backside of the nasal bone (Figs. 1b, 2c), and not swerve from this safety route, otherwise they shall intrude into the orbit or intracranium. The route guides the instruments towards the direction of frontal sinus floors.

Operating procedures are consistently performed via the middle nasal meatus, and the nasal septum is usually seen in the center of the view point. During the first part of the procedure, the main author handles the endoscope and the drill bur or forceps via the right side of the middle nasal meatus (Fig. 1a). After creating the window, the main author generally advances the endoscope through the right side of the middle nasal meatus and drill bur or forceps through the left side (Fig. 1b). Through the procedure, the endoscope is kept on the right side, except the occasion of clearing the left side of the frontal recess before frontal sinusotomy, so that the operator’s view point is mostly stationary to the end of the surgery (Fig. 1).

The procedures are mainly performed under 70° endoscope, so that the surgeon advances the instruments towards the upper and deeper part of the virtual field on the operative monitor (Fig. 3), that corresponds to the anterior and rostral part of the real nostril (Fig. 1). After the anterosuperior part of the middle turbinate disappears from the operative view, the operator can identify the floor of frontal sinuses and view a link of the nasal septum to the interfrontal septum. After the bilateral frontal sinuses and their posterior walls are confirmed, the interfrontal septum is removed superiorly and bilateral frontal sinuses are unified (Figs. 1c, 2c, 3). Subsequently the largest possible neo-ostium of frontal sinuses can be created, to avoid scarring of the frontal ostium. The resultant frontal ostium is stented with a silicone tube or sheet for several weeks and then removed.

Fig. 3.

Intraoperative photograph with a 70° endoscope that is a source of Fig. 1c illustration. The frontal sinus floors are penetrated and frontal sinuses are unified

We reviewed clinical records of patients who underwent EMLP with transseptal approach and examined operative time, bleeding volume complications during and after surgery and a state of newly created naso-frontal duct of cases. Naso-frontal duct in which bilateral frontal sinuses were well observed from nostril with a 3 mm fiberscope at observation was classified as patent, and that which the fiberscope did not pass through was classified as narrow, and that which was not identified was classified as closed.

Results

The sinus pathology was mucocele in 11 patients and sinusitis in five patients. Sixteen surgeries were performed on sixteen patients. The symptoms encountered, in order of frequency, included pain (44 %, n = 7), lid swelling (44 %, n = 7), sinucutaneous fistula (13 %, n = 2), nasal obstruction (6 %, n = 1). The number of surgical operations including EMLP was 9 in 1 patient, 4 in 2 patient, 3 in 2 patients, 2 in 6 patients, 1 in 5 patients. The average operative time was 147 min (85–328 min). The average intraoperative bleeding volume was 59 ml (50–150 ml). Bleeding volume during or after surgery was not prominently more than that in other classical endoscopic sinus surgery.

Recently we launched the navigation system, which was used in recent three cases. Neither intracranial nor orbital complications were encountered during or after surgery in any cases. One patient complained of subjective olfactory disturbance after surgery, but the symptom disappeared in a few months after the stent being removed.

The average hospital stay was 11 days (6–21 days). One case of sinusitis with sinucutaneous fistula stayed in the hospital for 21 days, because the severe inflammation in the sinus should be washed with the repeated saline irrigation through the fistula.

Preoperative antibiotics or steroids were not used in any case except one case with sinucutaneous fistula including severe inflammation in the sinus. Generally, postoperatively patients underwent macrolide treatment and saline douching for a few months. The silicone stent was put into the naso-frontal duct in most cases, but it was removed in a few months in the all cases except one case of the sinus cutaneous fistula. The average follow-up period was 18 months (5–60 months).

At time of observation, naso-frontal duct was patent in 13 patients, narrow in two patients, and there were no closed cases. However the duct was still stented in one patient at observation. The fiberscope did not pass through the naso-frontal duct in the two narrow-duct patients, but the CT images of these cases showed the well ventilated frontal sinuses. The stented one patient had complicated frontal sinus fractures with sinucutaneous fistula due to trauma and experienced four times surgery. In this case, the stent was put for more than 1 year and this shall be left as long as possible. All patients were asymptomatic at observation.

In a representative case, the patient was a 45-year-old man who presented with bilateral frontal cysts. Because the bony septa existed above the frontal ostium on CT findings (Figs. 2c, 4a), it was supposed that the surgical identification and access of the frontal sinuses should be difficult with standard ESS technique. In this case, the anteroposterior diameter between the anterior skull base and nasal bones was 1.2 cm in the CT findings (Fig. 2c). Via the transseptal approach, approximately 1 cm-widths fenestration was extent up towards the bilateral frontal sinus floors, the septa were penetrated, and the sinuses were unified (Figs. 1, 3). In this case the navigation system was not used, because we had not it at that time. At 3 months, frontal sinuses remained well ventilated on CT findings (Fig. 4b) and the patient was asymptomatic with 20-month follow-up.

Fig. 4.

Preoperative (a) and postoperative (b) computed tomographic coronal images of the case. a Bony septa obstructed ventilation of frontal sinuses. b Unified frontal sinuses were well ventilated and the superior part of the nasal septum had been removed

Discussion

Chronic frontal sinusitis is one of the most challenging problems faced by sinus surgeons [1, 2]. Extended endoscopic techniques such as EMLP (Draf III) is technically demanding [3–5], but may still involve a difficulty of intraoperative identification of the frontal sinus. In revision cases, the loss of normal anatomic landmarks, scarring, and the occurrence of reactive bone sclerosis make surgery more difficult [6]. In such cases, nasal septum can be uses as a landmark to access to the frontal sinus.

Our EMLP with transseptal approach was a modification of the original EMLP or Draf III [3, 5]. In the original EMLP, initially the bilateral frontal sinuses are opened and continuously interfrontal septum and nasal septum are removed. In our EMLP with transseptal approach, the removal of the nasal septum is the initial step of the surgery, and the opening and the unification of the bilateral frontal sinuses are the final step.

Transseptal frontal sinusotomy has been reported by some authors. Wormald et al. [4] removed a 2 × 2-cm part of the nasal septum at the first step of surgery. McLaughlin et al. [6] created an iatrogenic septal perforation just below the frontal sinus with an approximate diameter of 1.5–2 cm after a classical septoplasty approach. In the previous report [4], frontal sinuses are seeked alternately via the middle nasal meatus or the nasal cavity. In our methods, operating procedures are consistently performed via the middle nasal meatus and the nasal septum is usually seen in the center of the view, so that surgeon’s view point is mostly stationary to the end of the surgery. Besides surgical procedures in the approximately 1 cm behind the nasal bone along the nasal septum are safe, because this area is distant from orbital contents and the anteroposterior diameter between the anterior skull base and nasal bones generally should be approximately 1 cm.

The transseptal approach should be recommended for difficult cases with the risk frontal sinus. We, however, routinely perform this procedure, because the approach may be simple in comparison with other classical EMLP [3]. The fact that prominent complications were not observed during or after surgery was the ground to perform this procedure. In our cases, one patient complained of temporal olfactory disturbance after surgery, but the symptom disappeared in a few months after the stent being removed. This was may be because the stent changed the air current or caused crusting around the olfactory cleft.

Frontal sinus mini-trephination helps the identification of the frontal ostium in combination with frontal recess instrumentation [7]. Trephination itself, however, is not without its own risks, such as intracranial penetration and orbitary injury [8] and can leave unsightly facial scars. In recent years, navigation surgery has been introduced to improve intraoperative localization and we also launched the system in recent series of our surgical cases. There are, however, several potential sources of geometric error, including the matching and registration system employed, anatomic drift and variations in the facial skeleton and skin laxity [9]. Thus frontal sinus mini-trephination or navigation has enhanced surgical techniques, but are not a replacement for a sound knowledge of the regional anatomy and surgical experience [1].

A demerit of this procedure, including classical EMLP approach, is to handle angled endoscope during surgery. The surgeon loses sight of tip of instruments, if he or she has little practice in angled endoscopes. During EMLP, some authors use 30- or 45-degree endoscope [3, 4], and the main author and others use 70° endoscope [10, 11] for direct visualization of the frontal sinus and the removal of the frontal sinus floors. Thus it is desirable that physicians managing frontal diseases should practice handling angled endoscopes.

Conclusions

Endoscopic sinus surgery for the frontal sinus is problematic, because of the complex anatomy and large anatomical variation of the frontal recess. We describe an endoscopic transseptal approach to identify and access the frontal sinus. Endoscopic transseptal frontal sinus surgery is simple to perform, and does not cause severe complications.