rim of the FR gave access to the inferior rim of the superior orbital fissure. Our measurements indicated that the
pterygopalatine fossa was exposed on a mean 6 mm width in a coronal plan as far as the vertical plate of the palatine bone,
medially. In all cases, the pterygopalatine ganglion and palatine n. were identified.
The lateropterygoid space(Fig. 6)
The lateropterygoid space was localized posterior to the retroantral space, lateral to the lateral pterygoid plate and under the
greater sphenoid wing. Exposure of this structure via the transmaxillary approach required the use of optic magnification.
This space was filled by the LPM. This fleshy muscle ran transversely in the horizontal plane and its two distinct heads, a
smaller upper infratemporal and a lower pterygoid one, were clearly exposed. The MA was intimately related to the pterygoid
head of the LPM, running lateral to the pterygoid head or between the two bundles. Following the posterior edge of the
pterygoid process required a 12 mm additional progression and dissection of the PVP. This plexus was located between the
fibers of the LPM and, at its upper surface, connected to the greater sphenoid wing via transosseous channels. In the present
study, these multiple venous channels were in the area of the FO. The anterior rim of the FO was reached at a 56 mm mean
depth and its inner portion was not totally exposed, concealed by the lateral pterygoid plate. The mean angle of exposure of
the FO was 20° in the horizontal plane.
Discussion
The ITF is grossly defined as the anatomic space under the floor of the middle cranial fossa and posterior to the maxilla. The
medial wall of the ITF is formed anteriorly by the lateral pterygoid plate and posteriorly by the tensor veli palatini m. The
anterior and medial walls are separated superiorly by the pterygomaxillary fissure through which the ITF communicates with
the pterygopalatine fossa. The lateral limit of the ITF is the ramus of the mandible. The posterior boundary of the IFT is
variously defined. Some authors define it by the prevertebral layer of the cervical fascia and the underlying muscles. This
broad definition includes the structures localized behind the styloid mm., including the internal carotid a., internal jugular v.
and lower cranial nn. [1, 9]. For other authors, these structures belong to the parapharyngeal space [ 8]. With them, we define
the ITF as the area located below the greater sphenoid wing and where the posterior limit is a plane descending vertically
from the sphenosquamosal suture. This is in accordance with the data provided by Robert et al [10 ], who define this posterior
limit by the fascia of the medial pterygoid m. The ITF can be subdivided in a retromaxillozygomatic region, the region of the
pterygoid mm. and the pterygopalatine fossa [10]. It may be involved by various pathologic processes and particularly by
tumors. Tumors originate primarily from Schwann, vascular, epithelial or mesenchymal tissues of the ITF or involve it
secondarily from a neighbouring area. Tumors growing from the parapharyngeal and nasopharyngeal spaces or from the
middle cranial fossa are classified in this latter group. To reach those various lesions located in a deep and complex anatomic
space, cranial surgeons have developed numerous approaches. In such situations, the appropriate choice of approach is
dictated by the origin of the lesion and its exact extension in the IFT. Thus, lesions involving the posterior cranial fossa and
invading the ITF are preferentially removed by lateral approaches [4, 11]. Lesions starting in the middle cranial fossa
(meningiomas, schwannomas, chordomas, etc.) and secondarily extending into the ITF, are preferentially approached
superiorly via a temporal craniotomy. Lesions extending from the parapharyngeal space are exposed after a transmandibular
median splitting [6, 8]. In a recent anatomic study, Hitotsumatsu and Rhoton [5] described various combinations of subtotal
maxillectomies in order to expose the totality of the elements of the infratemporal fossa. Traditionally, approaches of the ITF
can also be classified as extensive for carcinologic surgery, or minimally invasive if the aim is a simple exploration or biopsy.
Nevertheless, recent refinements of imaging techniques have reduced the role of exploratory surgery and, in order to
minimize cosmetic and functional morbidity, surgeons nowadays attempt to perform, if possible, tumor removal via
conservative routes. Using the natural space offered by air cavities, the transmaxillary approach appears to be a safe and
simple procedure. It preserves the functional anatomy of the nose and allows the initial control of the neurovascular
structures, namely the infraorbital n. and the MA. The infraorbital n. course is used as a guide during anteroposterior
progression in the antrum. Careful opening of the posterior wall of the antrum avoids the risk of MA bleeding. This
serpentine artery is enclosed in a sheath of venous network that can be dissected without problems. Mobilization and
sometimes ligation and section of this artery are required to progress deeper into the ITF. Vrionis et al [12] demonstrated that
the tortuous feature of this artery could allow tension-free in situ anastomosis to the supraclinoid part of the carotid a. when
using temporal craniotomy.
Some limitations of this approach can be deduced from the present study. A sublabial incision appeared to be insufficient for
exposure of the anterior surface of the maxilla and puts the infraorbital n. at risk of traction injury. Moreover, it is broadly
admitted that the required paralateronasal skin incision is cosmetically acceptable. As shown by our measurements, this route
is deep and narrow for exposure of structures located posterior to the plane of the pterygoid process. Thus, delineation of the
boundaries of the FO is allowed only by a 20˚ angle of vision at more than 55 mm depth. The reduced exposure of this
foramen is additionally due to its outer masking by the lateral pterygoid plate. Moreover, two major structures reduce the
access to the retropterygoid region. In all cadavers, we have seen that the LPM tightly filled the lateropterygoid space. By the
transmaxillary route, single retraction of this fleshy muscle was not possible and further progress required desinsertion and
resection of its superior bundle. The other important obstacle is the complex anatomy and density of the PVP. A recent