|Home | About | Journals | Submit | Contact Us | Français|
We propose a surgical approach for select patients that minimizes morbidity while allowing gross total resection of lesions in the anterior portion of the infratemporal fossa. The approach we describe is an extradural approach through a subtemporal craniectomy or craniotomy with the possible addition of a zygomatic osteotomy. Lesions that have a well-defined capsule and a texture that permits manipulation are ideal for this less invasive approach. We retrospectively reviewed six cases from the primary author (C.B.H.) using a temporal craniectomy or craniotomy alone to resect lesions in the infratemporal fossa. All six cases had good clinical outcomes with no unexpected neurological deficits while achieving gross total resections. The only complication included one cerebrospinal fluid leak that was sealed endoscopically. For select lesions, a less morbid surgical approach via an extradural window through a subtemporal craniectomy or small craniotomy may be preferable to transfacial approaches. Adjuvant use of endoscopic techniques may facilitate surgical exposure and resection of large lesions.
The infratemporal fossa is an anatomic space located below the temporal fossa. The boundaries include the greater sphenoid wing, floor of the temporal fossa, pterygoid plates, masseter muscle, mandible, posterior wall of the maxilla, glenoid fossa, tensor veli palatini, styloid process, and mastoid.1,2 The infratemporal fossa hosts several vital neural and vascular structures that have been carefully demarcated in prior cadaveric studies.1,2,3
The infratemporal fossa hosts a variety of uncommon pathological entities. Disease processes include benign lesions such as meningiomas, schwannomas, teratomas, hemangioblastomas, angiofibromas, neurofibromas, and ameloblastomas to more malignant diseases such as adenoid cystic carcinomas, epidermoid carcinomas, adenocarcinomas, chondrosarcomas, plasmacytomas, squamous cell carcinomas, and other various sarcomas.4,5,6,7,8,9
Traditionally, surgical exposures have involved a long facial or neck incision and a large operative window. Such techniques have widely been adopted for resection of malignant lesions and purportedly aggressive surgical resections have translated into prolonged survival.7,8 Early surgical access to the infratemporal fossa was described by Fisch9,13; however, newer surgical techniques have subsequently been devised for varying purposes. Variations to access the infratemporal fossa have included preauricular subtemporal craniotomies, transfacial approaches, transmaxillary venues, osteoplastic osteotomies, mandibular division, and more recent endoscopic approaches.6,10,11
Several combined procedures or variations on previously described techniques have also been published.4,5 The traditional exposures have been associated with significant morbidity and risk.12 We describe another variation that permits access to the anterior compartment of the infratemporal fossa and masticator space while minimizing morbidity.
We retrospectively reviewed the primary author's surgical experience treating infratemporal fossa lesions over the past 13 years. The cases were then narrowed to identify patients who had infratemporal lesions resected via an extradural approach using a temporal craniectomy or craniotomy alone with or without the addition of a zygomatic osteotomy. These patients were identified and their office charts were reviewed for demographic data, outcome measures, pathology reports, clinical results as well as preoperative and postoperative imaging. Institutional review board approval was obtained from our institution's research committee.
Patients are positioned supine or slightly rotated on the surgical bed and placed in three-point Mayfield pin fixation. The head is rotated to expose the desired temporal side, and the vertex of the head is slightly angled toward the floor to minimize temporal lobe retraction. A small area of hair is shaved to expose the planned incision. A linear incision is marked ~1 cm anterior to the tragus extending from below the root of the zygoma to just below the superior temporal line (Fig. 1). Caution is used to preserve the superficial temporal artery. Alternatively, a curvilinear incision standardly used for pterional craniotomies can be adopted if a larger craniotomy with a zygomatic arch osteotomy is planned. After preoperative assessment, we incorporate a zygomatic arch osteotomy to augment our surgical window for very large infratemporal lesions or to minimize temporal lobe retraction for lesions extending into Meckel's cave or the cavernous sinus. The surgical area is prepped in a standard, sterile fashion, and intravenous antibiotics are administered.
After the incision is made, the temporalis muscle is divided in a linear fashion in the plane of the muscle fibers. The root of the zygoma is exposed as well as the inferior aspect of the temporal skull (Fig. 2). A drill is used to perform a low temporal craniectomy or craniotomy. The medial superior aspect of the root of the zygoma is removed as well to provide an additional few millimeters of exposure (Fig. 3). Approximately 1 to 2 cm of the inferior temporal skull is removed, exposing the dura. A zygomatic osteotomy may be performed to increase the exposure rostrally or to allow greater retraction of the temporalis muscle inferiorly if needed. The temporal lobe can then be gently retracted with the dura mater as a protective layer. We do not routinely place a lumbar drain for cerebrospinal fluid (CSF) drainage, but this technique could be used if preferred.
The floor of the temporal fossa is then removed with a drill extending medially until the foramen ovale is encountered. Any exposed air cells should be carefully sealed with bone wax. The surgical exposure can be extended extradurally into Meckel's cave or the cavernous sinus as needed. The operative corridor is illustrated through the preoperative coronal magnetic resonance imaging by arrows in Fig. Fig.4.4. Gentle retraction of the inferior lateral temporal dura provides about 2 cm of vertical working space. The anterior to posterior working space is 3 to 4 cm (Fig. 5). Most benign tumors have a plane that is advantageous to the surgeon. The tumor is debulked internally, and then the tumor edges can be retracted into view and the mass further debulked to allow greater manipulation and resection of the lesion. Electrophysiological monitoring may help identify the motor fascicles of the third division of the trigeminal nerve. Figure Figure55 depicts the operative view prior to tumor resection, whereas Fig. Fig.66 highlights the tumor resection cavity in the infratemporal fossa. Figure Figure77 is a cadaveric dissection illustrating the normal anatomy identified within the infratemporal fossa.
We occasionally place a piece of thrombin-soaked Gelfoam (Pharmacia & Upjohn, Kalamazoo, MI) into the resection cavity, but we do not typically seal the skull base defect. Alternatively, other adjuvant sealant tools could be adopted such as Hydroset (Stryker-Leibinger, Freiburg, Germany), dural repair substitutes, fat grafts, or liquid dural sealants (i.e., Duraseal [Baxter Healthcare, Round Lake, IL] or Tissal [Baxter Healthcare]).
We have performed 17 operations for infratemporal lesions over the past 13 years. A total of six cases were managed solely using an extradural approach via a small temporal craniectomy or craniotomy. The operative approach was selected based on the radiographic appearance of a well-defined lesion, the extent of tumor invasion, and characteristics of benign tumor pathology. The most common presenting symptom was a headache in four patients, followed equally by diplopia, trigeminal nerve numbness, hearing loss, and a facial mass, each in two patients. The mean age of the patients was 26.3 years; however, excluding the 7-day-old neonate, the readjusted mean was 31.4 years of age. The lesions ranged from 2.5 to 6.5 cm in maximal diameter.
The 7-day-old patient was hospitalized for 34 days due to reasons unrelated to the surgery. Of the remaining patients, the mean hospitalization was 5.4 days, although there was a trend toward a shorter length of stay for recent surgeries as compared with patients early in the series. The recent hospitalizations ranged from 2 to 3 days postoperatively, whereas patients earlier in the series ranged 9 to 10 days. The mean estimated blood loss was 405 mL, although it ranged from 85 to 1115 mL.
Postoperative imaging showed gross total resection in all cases, but three patients had reoperations for recurrent lesions. One of these lesions was a recurrent hemangiopericytoma, which recurred for the fourth occasion 29 months after the initial surgery and was retreated through the same approach. The other recurrence was a mucocele that redeveloped 26 months later and was treated through a transnasal endoscopic venue. The last recurrence was a neonate who was felt to have had some residual mature teratoma postoperatively that regrew and was reoperated upon 3 months later.
On final pathology, the remaining lesions included two trigeminal schwannomas and one malignant peripheral nerve sheath tumor (MPNST). Two patients developed new sensory trigeminal deficits after branches were sectioned to remove tumor. However, no unplanned neurological sequelae developed with any of the six patients. None of the patients were noted to experience any significant facial edema. We encountered no complications of unplanned neurological deficits, trismus, or wound infections; and we only experienced one case of CSF leakage. One patient had a MPNST resected that invaded the eustachian tube and developed postoperative rhinorrhea. The orifice was closed endoscopically and no further leakage was observed. Only one patient died 4 years after surgery from progression of a hemangiopericytoma, but did not suffer any surgical complications perioperatively. The mean follow-up was 54.2 months, but extended from 3 to 146 months. The summary of the case series are included in Table Table11.
Over the last few decades, access to previously challenging lesions has been significantly facilitated by the advent of skull base approaches. With progressive technological advances and improved surgical windows, more recent trends have been toward smaller exposures and less invasive procedures. Proponents of less invasive approaches advocate that they are less morbid with equally efficacious results. We propose that an extradural approach through a small temporal craniectomy or craniotomy is a viable alternative for select infratemporal fossa tumors located anteriorly. This approach reduces morbidity and limits the size of the incision while preserving equivalent results.
Surgical approaches to the infratemporal fossa have ranged from more anterior or lateral exposures. Some of the earlier approaches were described by Fisch. His type A to C approaches necessitate sacrifice of conductive hearing and therefore involve significant morbidity.9,13,14,15,16 Alternatively, his type D approach theoretically preserves hearing although closure of the eustachian tube will lead to secretory otitis obviating the middle ear function.17 Whereas the type A Fisch exposure is ideal for jugular fossa lesions and his type B exposure better suited for surgical access to the petrous apex, the subtemporal craniotomy provides an excellent surgical window for lesions in the anterior aspect of the infratemporal fossa or masticator space. Some authors have also advocated that this approach is limited to lesions less than 2.5 cm in size.5 We feel that the size of the lesion is not as limiting a factor with benign lesions; the tumor texture and surrounding plane in benign lesions permit greater manipulation and resection of larger lesions through a smaller exposure.4
However, when operating through a smaller surgical exposure, anatomic localization is critical to minimize complications. The normal anatomy of the infratemporal fossa has been well documented, and distances between crucial structures have been measured in prior series.1,2,3,18,19 It is also critical to carefully select appropriate cases. The surgical plane surrounding most benign lesions allows dissection of the tumor from critical neurovascular structures and permits safe gross total resection of lesions. The safety of the surgical resection can be augmented by adjuvant use of intraoperative navigation. The surgical exposure can further be maximized through a small window with the aid of endoscopic tools.
Goel described a similar exposure for resection of trigeminal schwannomas using an intradural approach and performing an osteotomy of the zygoma.20 Although the zygomatic osteotomy may help exposure in some cases, we have found that it is unnecessary in most cases. However, if the tumor extends into the cavernous sinus or significantly into Meckel's cave, then a zygomatic arch osteotomy allows inferior displacement of the temporalis muscle. This permits a viewing angle from an inferolateral to superior and medial direction, which minimizes the need for temporal lobe retraction when removing tumor from the cavernous sinus or Meckel's cave.
Endoscopic resection of infratemporal fossa lesions in the published literature has largely been limited to juvenile nasal angiofibromas.6 The operative approach has been through a transnasal venue. However, an anterior endoscopic approach may provide a suboptimal operative window for lateral or large lesions extending inferiorly and limit a surgeon's ability to achieve a complete resection. We have opted to use the endoscope as an adjuvant instrument through a small temporal craniotomy to facilitate our tumor resection.
We propose that a small temporal craniotomy also minimizes the procedural morbidity. Traditional surgical approaches to the infratemporal fossa are associated with significant morbidity. Complications reported include infection, CSF leak, postoperative hematoma, cosmetic disfigurement, flap necrosis, malocclusion, trismus, limiting food to soft solids for weeks, nerve dysfunction, stroke, and ocular problems. Although most series include malignant pathologies, the reported procedural risk ranges from 21 to 70%.4,7,11 More benign lesions are most likely associated with less risk given a lower association with local invasion of tissue and vascular compromise. However, most of the series do include a significant portion of benign pathologies and still have a relatively high associated risk. In our small series, we only encountered one unexpected procedural complication of CSF rhinorrhea. A small temporal craniectomy or craniotomy eliminates the risk of flap necrosis, minimizes the risk of infection and cosmetic disfigurement, and avoids the problems of trismus and orbital involvement. Our mean length of hospital stay was 5 days, which is considerably shorter than the average stay of patients who have undergone more traditional approaches at our institution. Patients can also initiate an oral diet earlier, which may also contribute to a more expeditious recovery.
Most series reporting recurrences contain a mixture of both malignant and benign lesions, but the benign subgroups favor a very low recurrence after gross total resection. We advocate this approach for lesions that have radiographic and clinical features supporting a benign pathophysiological process to increase the likelihood of encountering a well-defined tumor border and a good operative plane. Most series have shown excellent control with low recurrence rates after gross total resection among benign lesions.11
We have noted that on postoperative imaging, intense enhancement of the muscle tissue in the infratemporal fossa is often present. This was particularly prominent in the pterygoid muscles that were compressed by a large infratemporal fossa tumor preoperatively (Fig. 8). This enhancement often obscures interpretation of the postoperative magnetic resonance images. In our experience, the enhancement typically dissipates by 3 months and has not been associated with recurrences (Fig. 9).
We describe a variation on previously described techniques to minimize morbidity when managing infratemporal fossa lesions. The use of a temporal craniotomy to resect infratemporal fossa masses is a feasible alternative associated with a lower risk of complications, but careful selection of appropriate patients is critical. Alternatively, this approach can be used safely for an open biopsy of infratemporal fossa lesions.