|Home | About | Journals | Submit | Contact Us | Français|
Cerebrospinal fluid (CSF) rhinorrhea is a common condition managed by most otolaryngologists with the help of nasal endoscopy (sinoscopy). In the last 2 years, we have used a neuroendoscope with a working sheath to treat nine patients with CSF rhinorrhea. One patient developed a recurrence 1 month after treatment but then responded to conservative treatment. We conclude that the treatment of CSF rhinorrhea by a neuroendoscope with a working sheath is safe, effective, and easy and obviates the need for a separate sinoscope.
Cerebrospinal fluid (CSF) rhinorrhea has been a major treatment challenge for otolaryngologists and skull–base surgeons.1 Traumatic skull–base fractures and iatrogenic injuries are the main causes of CSF rhinorrhea,1 but the latter are rare compared with the former.2 These fistulas must be repaired to avoid imminent life–threatening complications like ascending meningitis and pneumocephalus.1
During the last 25 years, treatment of CSF rhinorrhea has evolved from intracranial approaches3, 4, 5 to extracranial approaches.6, 7, 8 Extracranial approaches are equally successful and associated with significantly fewer complications rates when compared to intracranial approaches.9 Since 1981 when Wigand used endoscopic treatment for the first time to treat CSF rhinorrhea,10 the technique has gained increasing attention. The advantages of endoscopic treatment such as excellent visualization, precise graft placement, and shortened operating time have popularized it worldwide.11, 12, 13 We present our initial experience using a neuroendoscope with a working sheath to treat nine patients with CSF rhinorrhea.
Between March 1998 and November 2001, nine patients (five females and four males; mean age, 21.6 years; range, 2.5 to 36 years) were referred to our department with a possible clinical diagnosis of CSF rhinorrhea. Three patients had spontaneous and six had post–traumatic rhinorrhea. The duration of symptoms ranged from 5 months (in the case of post–traumatic rhinorrhea) to 8 years (in the case of spontaneous rhinorrhea). Three patients had a history of meningitis at some stage of the disorder. All patients had failed conservative treatment.
All patients underwent a thorough clinical examination, and the glucose concentration of the nasal discharge (CSF) was analyzed. Six patients underwent computed tomography (CT), seven underwent magnetic resonance imaging (MRI), and one underwent CT–cisternography. Four patients underwent both CT and MRI. One patient underwent both MRI and CT–cisternography. Only MRI localized sites of leakage. T2–weighted MRI showed an arachnoid pouch prolapsing through the basal defect in two patients and hyperintense CSF leakage into the sinus in four patients or into the nasal cavity in one (Fig. 1). CT and CT–cisternography showed fractured sites in patients with post–traumatic rhinorrhea but were inconclusive regarding the exact location of the site of CSF leakage. In two patients with post–traumatic rhinorrhea, the leakage sites were primarily defined by endoscopy. The leakage sites were at the anterior ethmoid in five patients, the posterior ethmoid in three, and the frontoethmoid in one.
Patients were administered systemic antibiotics. General anesthesia was induced with endotracheal intubation. The head was slightly extended and turned toward the right side (the side of the operating surgeon). The face and nasal cavity were cleaned with soap and Betadine solution. A Gaab universal endoscope (Karl–Storz, Tuttlingen, Germany) was used (working sheath outer diameter, 6.5 mm; 0–degree telescope, 2.7 mm; working channels, 1 and 2.7 mm). A TV monitor and camera were attached to the endoscope for visual control and teaching purposes.
Before the working sheath was introduced into the nasal passage, adrenaline in saline (1:100,000)–soaked cottonoids were left inside 3 to 5 minutes for hemostasis. The working sheath and telescope were introduced under direct visualization and were fixed with the Endoscope Holder (Aesculap, Tuttlingen, Germany). Injury to the mucosa was avoided.
The fistula was localized by diagnostic endoscopy. Leakage sites were identified as a pulsating, glistening white arachnoid pouch in three patients (Fig. 2A) or as CSF leaking through a dural rent in six patients, confirming the findings on MRI. A Valsalva maneuver was performed to confirm the leak through the defect in cases of uncertainty. Fluorescein dye was not used to localize the fistula.
The position of the working sheath changed slightly, as needed, depending on the leakage site. The superior turbinate was partially resected to improve visualization and intraoperative maneuverability. The margin of the defect was defined and made raw by removing any granulation tissue or bone chips (Fig. 2B). Hemostasis was achieved by applying unipolar coagulation. Intermittent saline irrigation through a fine catheter in the working channel was used to clear the surgical field and telescope lens. An appropriately sized fascia lata graft (slightly larger than the defect) was created.
After the telescope and other instruments were removed from the working sheath, the graft was inserted. The telescope was reintroduced to guide the graft to the tip of the working sheath. Held by forceps, the graft was insinuated into the defect a few millimeters, to plug it. The holding forceps were withdrawn gradually by slightly rotating them, and the graft was left in place (Fig. 2C). This was supported by Gelfoam sponge. The working sheath was removed followed by the posterior nasal packing.
Post–operatively patients were confined to bed rest with their heads elevated 30 degrees. Intermittent lumbar drainage of CSF was done twice a day for 3 to 5 days. Nasal packing was removed 48 to 72 hours after surgery. Patients were advised to avoid straining and nose blowing during the immediate postoperative period.
The rhinorrhea resolved completely in eight patients. One case of post–traumatic rhinorrhea recurred a month later but responded to conservative treatment. There were no procedure–related complications. Transient anosmia occurred in two patients but recovered spontaneously within a month.
Most neurosurgeons prefer the intracranial approach.14 Sphenoid sinus fistulas are approached with great difficulty and may be inaccessible through intracranial approaches because of adjacent neural and vascular structures.15 Exposure of the skull base and the necessity of brain retraction during intracranial procedures are associated with a significant risk of anosmia, postoperative intracerebral hemorrhage, and brain edema.16 The failure rate associated with the management of CSF leaks via an intracranial approach has ranged from 20 to 40%.7, 17, 18
In contrast, extracranial approaches have lower morbidity rates, higher success rates, and seldom result in anosmia.6, 7, 11, 12, 16 They provide the best exposure of the sphenoid, parasellar, and posterior ethmoidal regions and offer excellent visualization of fistulas in the posterior wall of the frontal sinus, the cribriform plate, and the fovea ethmoidalis.6, 16, 19, 20, 21 Transnasal endoscopic surgery minimizes intranasal trauma and preserves the bony framework supporting the frontal recess and other critical areas.22
Mostly otolaryngologists use a 4–mm sinoscope to perform transnasal endoscopic treatment of CSF rhinorrhea. The sinoscope, which is not fixed, is usually held in one hand while the other hand guides the instrument. This configuration risks injury to the passage. A system with a working sheath, which is fixed with an Endoscope Holder, eliminates unwanted movement and frees both hands for surgical maneuvering. Working channels in the sheath allow other instrumentation to be inserted without causing injury. The field and lens can also be irrigated when obscured by bleeding or cauterization. Once inserted, the working sheath remains until the procedure is completed. In contrast, a sinoscope must be withdrawn multiple times for cleaning and surgical maneuvers.
Various dyes like methylene blue, phenolsulfonphthalein, indigo carmine, and fluorescein have been used to demonstrate the osculum of the fistula.23, 24 Fluorescein is still in use but is not preferred because it is associated with complications like transverse myelitis and allergic reactions.25 The Valsalva maneuver has been used to detect ambiguous sites of leakage in CSF rhinorrhea. We have also used the Valsalva maneuver, which clearly helped demonstrate the location of the CSF leak.
A pedicled flap–like septal mucoperiosteum or a free graft from temporalis fascia, fascia lata, free muscle, tragal perichondrium, abdominal fat, or even an omental free flap of synthetic dural substitute can be used for the endoscopic repair of CSF fistulas.1, 16, 26, 27, 28 Free grafts are less bulky and are thought to interfere less with postoperative nasal function.9 Theoretically, tenting or folding the pedicled flap could cause the defect to seal inadequately.12
Fibrin glue has been used to secure the graft into position in previous studies.28, 29 In our series, plugging the graft into the defect required no further reinforcement by fibrin glue, thereby reducing the cost of treatment. We used autologous fascia lata graft, which can easily be obtained from thigh through a very small incision, in all our cases.
We conclude that endoscopic treatment of CSF rhinorrhea with a neuroendoscope with a working sheath is relatively inexpensive, effective, safe, and less traumatic.
This article reviewed the endoscopic management of cerebrospinal fluid (CSF) leaks, a technique that has been used since the 1980s. The authors treated nine cases over 3 years and had good results in eight of the nine cases after the original operation.
This technique is well known to otolaryngologists. At many hospitals, it is the first–line treatment offered for a CSF leak from the skull base judged to be reachable with an endoscope. It is reasonable to use the technique, rather than intracranial or transcranial approaches (which can be held as back–ups for endoscopic failure), to deal with the problem. The technique is advantageous because of its ease of access. In fact, it is an extracranial procedure and is performed on an outpatient basis with local anesthesia and intravenous sedation. In experienced hands, the morbidity rate is minimal.
The authors report nine patients who underwent endoscopic repair of a cerebrospinal fluid (CSF) leak. They achieved excellent results using a much less invasive approach than a traditional bifrontal craniotomy. We prefer to use septal or conchal cartilage to fill the defect. We also use temporalis fascia. We have not used fluorescein dye and have been able to visualize CSF leakage without difficulty. Furthermore, the use of flourescein intrathecally has been associated with seizures. We also supplement the repair with Gelfoam and fibrin glue to seal the defect. Finally, we use frameless image guidance during surgery to avoid perforation through the anterior cranial fossa and to achieve the most direct approach. Clearly, this approach has become a very attractive, less invasive option for the treatment of CSF leaks. In most cases, it should be attempted before a craniotomy.