Endoscopic Management of Facial Fractures

The basic tenets of facial fracture repair include fracture exposure, reduction, and fixation. Traditionally, these repairs have been performed via an open surgical approach. In fact, in recent decades, increasingly wider exposures have been used to ensure accurate =bony repositioning. However, recent experiences have suggested that the use of endoscopes, as in other minimally invasive procedures, may allow repair of facial fractures through smaller incisions with less-extensive exposure.

Because of the lower morbidity rates associated with smaller incisions and exposures, minimally invasive surgical techniques have been rapidly accepted. Endoscopic sinus surgery was described in the 1970s and became the standard of care in the 1980s. Since then, the indications for endoscopic head and neck surgery have continued to expand, with applications in otology (middle ear endoscopy), skull base surgery (cerebrospinal fluid [CSF] leak management, optic nerve decompression, tumor resection), neck surgery (thyroid and parathyroid), and ophthalmologic surgery (dacryocystorhinostomy).

In facial plastic surgery, endoscopy was initially used for forehead rejuvenation. ^{[1, 2]} Now, endoscopic browlifts have, for the most part, replaced traditional open approaches, and many surgeons now use endoscopic techniques for midface rejuvenation. Early applications for endoscopic treatment of facial trauma include subcondylar fractures of the mandible, ^{[3, 4, 5, 6, 7]} orbital blow-out fractures (OBFs), ^{[8, 9, 10, 11, 12, 13]} frontal sinus fractures, ^{[14, 15, 16, 17]} and zygomatic fractures. ^{[18, 19]}

An illustration detailing the incisions for endoscopic repair of anterior table frontal sinus fractures can be seen below.

General requirements for endoscopic surgery include the following:

• The ability to surgically obtain and maintain an optical cavity

• The ability to insert a fiberoptic endoscope

• The ability to maintain adequate hemostasis

• The ability to apply instrumentation

Advantages of endoscopic repair include the following:

• More accurate fracture visualization

• Small external incisions

• Reduced soft tissue dissection

• The potential for visualization around corners

• The possibility of reduced duration of hospital stays

• Improved teaching opportunities (since the procedure can be visualized on a television monitor)

Disadvantages of endoscopic repair include the following:

• A current lack of dedicated instrumentation

• A moderate learning curve for the techniques

• A narrow field of view

• A limited ability for bimanual instrumentation without an assistant

Indications for endoscopic repair are generally related to fracture location, size, degree of comminution, and the surgeon's abilities. Some of the techniques described in this article are still under development, and surgeons contemplating the use of these techniques must determine whether institutional review board approval is necessary.

Anatomy and indications

Most frontal sinus fractures involve injury to some combination of the anterior table, posterior table, and frontal recess. Treatment of these injuries is complex because of the associated risks of brain injury, spinal fluid leaks, and mucocele formation. However, most isolated anterior table fractures are primarily an aesthetic problem. They do not involve the posterior table and are therefore felt to carry a low risk of long-term complications.

Traditional open reduction of isolated anterior table fractures requires a coronal incision for adequate exposure and fracture repair. Success rates are very high, but the procedure produces surgical stigmata, including a large scar, possible alopecia, paresthesias, and even facial nerve injury (temporal branch).

The endoscopic approach allows the surgeon to access isolated anterior table fractures through 2 scalp incisions, which is analogous to an endoscopic browlift approach. However, the repair is performed in a delayed fashion with porous polyethylene sheeting (Medpor, Porex Surgical Inc, Newnan, Ga) or one of the many available cements (generally some form of hydroxyapatite [HA]cement) to camouflage the bony defect. The delayed approach has several advantages over an acute repair. First, it alleviates the need to manipulate unstable bone fragments endoscopically. Second, the repair can be performed in a delayed fashion once the injury has been determined to be severe enough to result in an aesthetic deformity.

In general, frontal sinus fractures are repaired 1-10 days after the injury. This time window often necessitates a decision for repair prior to the complete resolution of forehead swelling. However, with the camouflage technique, a delay of 3-4 months to allow complete resolution of soft tissue edema is not detrimental. In fact, some patients avoid the need for a surgical procedure (see the image below and the video presentation below). The endoscopic technique is applicable for isolated anterior table frontal sinus fractures that do not displace the superior orbital rim.

For more information about the relevant anatomy, see Facial Bone Anatomy, Facial Nerve Anatomy, and Eye Globe Anatomy.

Surgical technique

Preoperative photographs and CT scans should be obtained to document the injury. Informed consent is obtained, including disclosure of the risks of bleeding, infection, paresthesia, alopecia, and poor aesthetic result and the possible need for open approach if an endoscopic repair cannot be performed.

Endoscopic brow-lifting techniques have previously been well described; ^[1] however, several points merit repeating. After injection of local anesthetic, a 3- to 5-cm parasagittal working incision is made above the fracture, and a 1- to 3-cm incision is made behind the hairline (see the image below). Incisions farther back in the hairline complicate the use of instruments around the intrinsic forehead curvature. The incision length varies depending on the size of the implant to be inserted. Take care to avoid trauma to the hair follicles.

A second 1- to 2-cm endoscope incision is then made 4-6 cm medially to the working incision. An endoscopic periosteal elevator is placed through the working incision, and a dissection is performed in a subperiosteal plane. The endoscope is generally not necessary when dissecting cephalad to the fracture. Take care to avoid tearing the periosteum because this further complicates insertion and manipulation of the endoscope.

A 4-mm 30° endoscope (with rigid EndoSheath and camera) is then inserted through the smaller incision, and the optical cavity is visualized (see the image below). Dissection over the fracture is performed under direct vision to the level of the orbital rims. Use caution to avoid injury to the supratrochlear and supraorbital neurovascular pedicles. The elevation is generally easy because the fracture has previously healed.

Once the entire fracture is exposed, several options are available to the surgeon. If using a sheet of porous polyethylene (0.85 mm thickness), it is trimmed to a size somewhat larger than the defect. The superior edge of the implant is marked with a pen to maintain the orientation endoscopically. The implant is inserted through the working incision and manipulated over the defect; the size is checked. It is then removed and trimmed to a diameter of 5 mm larger than the defect. Several attempts may be necessary to obtain the correct size.

A 25-gauge needle is then passed through the skin over the fracture site and endoscopically visualized to determine the best site for percutaneous screw placement. The ideal site allows placement of 2 screws at opposite edges of the implant without a second incision. A #11 blade is used to make a 2-mm through-and-through stab incision. A 1.7-mm self-drilling screw (4-7 mm in length) is passed through the edge of the implant on either side and into the frontal bone. The self-drilling screw must be placed at least 1 mm away from the implant edge to avoid implant tear. The scalp incisions are then closed in layers, and a head dressing is applied for 48 hours. No drains are used.

An alternative technique is to inject a bone cement (HA cement) to recontour the deficient area. This is injected under endoscopic guidance via one of the incisions, and the contouring can be accomplished by digitally smoothing the surface of the forehead over the skin while monitoring the effect on the cement through the endoscope.

Complications

Potential complications associated with the endoscopic approach, including bleeding, infection, and poor aesthetic result, are very similar to those of traditional open approaches. The development of alopecia at the incision sights is a small risk. Meticulous surgical technique and avoidance of electrocautery reduces this risk. Facial nerve paralysis is possible but highly unlikely because the entire dissection is performed in a subperiosteal plane. Whenever alloplastic implants are used, implant infection or extrusion is a risk. However, porous polyethylene has been used extensively in the maxillofacial skeleton with good clinical results. Also, while a potential for delayed fragmentation of HA cements exists, this has thus far not been reported for this type of reconstruction.

Orbital blow-out fractures (OBFs) most commonly involve the orbital floor or medial orbital wall. They result from traumatic force applied to the globe or bony orbit. Disruption of the bony orbit can cause prolapse and strangulation of the orbital contents, with diplopia, enophthalmos, and even visual loss. Early attempts at OBF repair involved a transmaxillary Caldwell-Luc approach. The orbital contents were reduced from below by packing the sinus with gauze. Although this technique enjoyed some success, poor visualization likely resulted in an increased risk of orbital injury.

In 1972, Walter described a transmaxillary technique for the treatment of OBFs with direct headlight visualization. The reduction, however, was performed blindly with a finger used for tactile sensation. Transconjunctival and subciliary incisions are the current standard of care for treatment of OBFs. These approaches allow direct visualization of the defect and reconstruction of the premorbid bony architecture. Unfortunately, lower-eyelid incisions have known complication rates of 1.2-42%.

Common complications range from transient scleral show to severe lid malposition. A second limitation is the inability to easily visualize the posterior bony shelf via transconjunctival or subciliary incisions. The angle attack is oblique, and prolapsing orbital fat usually obstructs the surgeon's view. The endoscopic repair of OBF involves a sublabial (Caldwell-Luc) incision and exposure of the orbital floor defect from below. The eyelid anatomy is not violated, and the risk of postsurgical eyelid complications is eliminated. The angle of attack is much more favorable for the visualization of the posterior bone shelf. The orbital floor defect is then reconstructed in a similar fashion to that of the open approach; however, the endoscope improves visualization.

Preoperative evaluation

All patients must undergo a complete preoperative head and neck examination that documents extraocular muscle function and visual acuity. Preoperative ophthalmologic examination is mandatory. An axial and coronal CT scan should be obtained to delineate the bony defect. Indications for surgical repair of OBFs remain controversial. Most authors agree that extraocular muscle entrapment, preoperative enophthalmos, or significant orbital floor disruption (>50% of the surface area) are indications for surgical repair. Patients with uncomplicated trap-door and medial blow-out fractures, as seen in the images below, are candidates for endoscopic repair.

The only absolute indication for the endoscopic technique is extraocular muscle entrapment with hyphema. Traditional teaching recommends observation for 3 days to reduce the risk of rebleeding into the anterior chamber and possible blindness. This likely increases the chance of extraocular muscle strangulation and long-term diplopia. With the endoscopic approach, the extraocular muscle can be released without the need for globe retraction. Currently, the authors consider fractures that extend lateral to the infraorbital nerve, complex orbital fractures, globe injury, visual deficits, or a lone seeing eye to be contraindications for endoscopic repair.

Surgical technique

The patient is prepared and draped in the usual fashion. The surgeon is positioned on the patient's right side; the assistant is on the left side, holding the endoscope. The television monitor is at the head of the bed, as seen in the image below. Forced duction tests are performed with general anesthesia, and the results are documented. A 4-cm sublabial (Caldwell-Luc) incision is performed to expose the maxillary face in a subperiosteal plane. Care is taken to avoid injury to the infraorbital nerve (V2).

A small antrostomy is made in the central portion of the maxillary face with a 4-mm osteotome. A 3-mm Kerrison is used to enlarge the antrostomy to 1 X 2 cm. Alternatively, an oscillating saw can be used. Once the antrostomy is completed, a small notch is placed in the inferior central portion of the antrostomy, just above the dental roots. This notch stabilizes the endoscope and provides tactile feedback to the assistant surgeon while he or she looks at the monitor.

A Greenberg retractor is used to maintain exposure of the maxillary antrostomy. A 30° endoscope with an irrigation sheath is then placed through the antrostomy to visualize the orbital floor fracture. A pulse test should then be performed to assess the fracture pattern, size, and location. A pulse test involves the application of gentle external pressure on the globe while the orbital floor is visualized from below. Fractures are then confirmed as either trap door or medial blowout.

Trap-door fractures can usually be repaired without the use of an implant. The maxillary sinus roof (ie, orbital floor) mucosa is dissected free at the lateral free margin of the bony defect. More extensive mucosal dissection should be avoided because it may destabilize the depressed bone fragment. The orbital contents should then be reduced with a malleable retractor or blunt elevator, and the depressed bone fragment can be hinged back into its premorbid position. Care must be taken to avoid entrapment of the periorbita. Interfragmentary resistance between the hinged bone fragment and the stable bone at the fracture periphery maintains the reduction.

Blow-out fractures of the medial orbital floor require an implant to maintain the reduction. In these cases, the mucosa is dissected from the orbital floor at the periphery of the fracture, protecting the maxillary sinus ostium. The entire sinus is not demucosalized. All depressed bone fragments must be gently teased free and removed. This point cannot be overemphasized because retained bone fragments can be pushed into the bony orbit when the orbital contents are being reduced. All margins of the bony defect are then visualized, including the posterior shelf. This can be extremely difficult with an open approach but is easier with endoscopic assistance.

A Medpor implant (Porex Surgical, Inc, Newnan, Ga) sized 38 X 50 mm (0.85 mm thick) is then trimmed to a diameter of 1-2 mm larger than the defect, as seen in the image below. A malleable retractor is then inserted to reduce the periorbita back into the orbital cavity. With endoscopic visualization, the implant is inserted through the maxillary antrostomy and positioned just below the orbital defect, and the malleable retractor is removed.

Pressure is applied to the posterior border of the implant until the implant sits above the posterior bony shelf. The instruments are then moved anteriorly, and pressure is applied just behind the orbital rim until the implant moves into the orbital cavity, as seen in the image below. Pressure from the orbital contents pushes the implant inferiorly onto the stable bony shelves and maintains the implant position. Care should be taken to avoid undue pressure or trauma to the infraorbital nerve where it passes lateral to the implant.

The stability of all repairs must be checked with a postreduction pulse test. Forced duction testing is then compared bilaterally to rule out any entrapment. If periorbital entrapment is a concern, the implant is removed and repositioned. If the forced duction test result is normal, the sinus is irrigated out and the incision is closed with a resorbable suture. No dressing is applied. Postoperatively, the patient should receive an ophthalmologic examination and undergo a postreduction CT scan to evaluate the implant placement. As with any orbital reconstruction, if the CT scan shows incomplete reduction of the orbital contents or improper implant placement, the patient should be returned to the operating room to revise the procedure.

In a systematic review of the literature on this topic, Cheung et al recently concluded that the endoscopic approach appears to be safe and effective. They were able to include 9 studies totaling 172 patients. They found resolution of diplopia in 86%, and enophthalmos resolved in 95%. Only 2 patients (1%) needed revision surgery. ^[20]

Complications

The intraorbital complications of endoscopic repair are similar to those seen with an open approach. They include bleeding, infection, extraocular muscle injury with persistent diplopia, globe injury, and poor aesthetic result. When using the endoscopic approach, the surgeon must be confident that all bone fragments are removed prior to reduction of the orbital contents. Before the authors started aggressive bone removal, one patient had a malpositioned implant with rotation of a bone fragment into the orbit. The patient had no symptoms related to the bone fragment, but she was returned to the operating room; the bone fragment was removed, and the implant was repositioned.

Unlike open approaches, postoperative ectropion or entropion is not a risk. However, the surgical approach places the infraorbital nerve at risk, and care should be taken to avoid excessive retraction or trauma to this structure. Most patients notice numbness in the V2 distribution. This is an expected outcome and generally resolves within 2-3 months.

Medial Orbital Wall Blowout Fractures

Blowout fractures of the medial orbital wall represent a separate subset of orbital blowout fractures, and recent reports include a variety of techniques for endoscopic management of these fractures. The procedure involves a transnasal, transethmoid approach to the medial orbital wall similar to that used for endoscopic decompression of a periorbital abscess. However, the main difference is the presence of the orbital contents herniated into the ethmoid sinus, and failure to recognize this can put the orbital contents at risk. The herniated orbital contents should be skeletonized by completing the ethmoidectomy around these tissues. This then allows the surgeon to determine the degree of disruption and the best approach for repair. If the herniated medial orbital wall is 'eggshelled' but continuous, then gentle pressure may result in adequate repositioning. Once repositioning has been accomplished, a temporary stent may suffice to hold the fractured fragments in place for healing.

The most frequently described technique involves rolling a sheet of thick silicon sheeting and placing it into the ethmoid as a stent. ^[21] In its attempt to unroll, it seems to provide just the right amount of pressure to support the medial orbital wall for healing. This stent can be removed in the outpatient setting within 2-6 weeks, depending upon the surgeon's preference. Some authors have described debridement of fractured fragments and placement of implants, such as septal cartilage or porous polyethylene as described for repair of the orbital floor above. ^[22] The proper choice of approach and repair of course depends upon the surgeon's judgment, preferences, and experience.

Subcondylar fractures of the mandible have often been treated with so-called closed reduction, and the indications for open reduction have been limited. ^[23] For the most part, this approach has been used because of the perceived high incidence of complications of open reduction, particularly the highly feared and devastating facial nerve injury, as well as the belief that closed reduction provides satisfactory outcomes. However, recent evidence suggests that closed management also carries complications. These include facial foreshortening on the side of the fracture, increased incidence of deviation upon opening, and pain. ^{[24, 25]}

The so-called closed reduction is, in fact, not closed reduction but closed management. Maxillomandibular fixation (MMF) may reduce the occlusion but does not reduce the fracture. In fact, Ellis suggested in 2000 that rigid MMF should be completely abandoned if a nonsurgical approach is selected. ^[26] He argues that MMF delays the true closed management techniques of functional rehabilitation and physiotherapy, which cannot be initiated until the MMF is removed.

The advent of minimally invasive techniques for mandibular fracture repair has further pushed surgeons in the direction of open reduction. Although still too early to draw conclusions, the endoscopic technique appears to minimize (if not eliminate) the risk of permanent facial nerve injury. ^{[27, 28, 29]} If satisfactory reduction and fixation of these fractures can be accomplished while reducing patient risk, the risk-benefit equation is likely to shift in favor of open reduction. Early data appear to favor this approach.

Preoperative evaluation

In addition to the standard general health assessment, all patients must undergo an adequate radiological evaluation. The minimum requirements remain controversial. Some authors advocate that a 3-dimensional coronal CT scan be obtained in all cases to assess the precise position of the condylar segment relative to both the glenoid fossa and the distal mandibular fragment. Although most surgeons obtain an orthopantomogram, it alone would not be enough to clearly demonstrate the position of the proximal fragment. Therefore, at a minimum, a complete mandible series should be obtained in addition to the orthopantomogram. When available, a coronal CT scan is recommended. The clinical evaluation should include a thorough assessment of all facial injuries. The occlusion, status of the dentition, and dental history should also be carefully evaluated. Obviously, the presence of other mandible fractures makes the repair more difficult.

Surgical technique

Arch bars are recommended in order to maintain the occlusal relationship. In addition, they may be used postoperatively for elastic training of the muscles. (Of course, once rigid fixation has been placed, MMF cannot be used to alter the bony relationship.) A vasoconstrictor is injected at the incision site. The incision is made along the anterior border of the ramus. Elevation is carried out over the lateral border of the mandible between the masseter and the bone. Keeping the periosteum as intact as possible to maintain the integrity of the optical cavity is important; this minimizes bleeding and muscular prolapse, which may obscure the operative field. The elevation begins under direct vision. When the mandible and the muscle are adequately spaced, the endoscope can be introduced. Generally, a 30° 4-mm endoscope is used through a brow lift sheath. Irrigation may be introduced through the brow lift sheath, if desired. Elevation is completed from the angle of the mandible inferiorlyto thefracturesitesuperiorly. Once the fracture is identified, care must be taken to find the lateral surface of the proximal segment. Otherwise, elevating medially to the proximal fragment is possible, which is likely to lead to bleeding.

Once the fracture is identified, attention is turned to reduction. One of the key steps in reducing these fractures is inferior traction on the angle of the mandible. This allows laterally displaced proximal fragments to be pushed medially into proper position. For medially displaced fractures, distracting the distal fragment far enough inferiorly to allow the proximal fragment to be pulled into reduction is essential; if that is too difficult at first, pulling it into a laterally displaced position is recommended. Inferior traction on the angle can be applied in several ways. The first and most obvious is to push down on the ipsilateral mandibular molar teeth. If that fails, the mandibular angle can be exposed externally through a small incision. A wire may be placed through a hole drilled in the angle, and inferior traction applied. Alternatively, a screw may be placed into the angle, and a screw-holding device may be used to apply inferior traction.

The proximal fragment may be manipulated with instruments introduced transorally through the incision. Instruments may also be placed through one or more transbuccal stabs through the cheek. A transbuccal trocar placed through the cheek for drilling and screw placement may be used to push the proximal fragment into reduction. Another option is to place a threaded fragment manipulator through a second transbuccal stab. This device has a self-drilling threaded end on a solid shaft that attaches to a handle. This can be screwed into the proximal fragment and can then be used to manipulate the fragment into proper reduction.Note that this must be performed with care, since the use of excessive force may break the device, leaving the threaded portion in the bone. Also note that the thread on this device in the Synthes set is designed to match the 2-mm screw thread, which allows it to be placed through a plate hole initially; later, after it is removed, a screw (usually an emergencyscrewwithanoversized thread) can be placed into the hole.

Once reduction has been accomplished, 1 or 2 plates are applied to rigidly fix the fracture in reduction. Note that the minimum amount of fixation for subcondylar fractures remains unclear; however, most surgeons who use this technique have used 2.0 zygomatic or mandibular fixation plates. If zygomatic plates are used, the placement of 2 is preferable. A 4- or 5-hole plate is generally introduced through the wound so that 2 holes overlie the proximal fragment. A plate-delivering device has been developed to make this process easier, although its use is not required. If a threaded fragment manipulator is used, it can be placed through the most proximal hole into the most superior portion of the proximal fragment. A screw can then be placed into the second hole on the proximal fragment. The reduction is then ensured, and a screw is placed into the distal fragment. The threaded fragment manipulator is removed, and the remaining screws are placed.Again, if adequate space is available,asecondplateshould be placed. Because fracture reduction has already been achieved, placement of a second plate should be much easier. The wound is then irrigated and closed. Note that although the description of the procedure is relatively straightforward, in practice, a fairly steep learning curve exists. Do not be discouraged if initial cases prove difficult and time consuming. Also, do not be deceived if the initial cases go exceedingly smoothly.

Intraoperative challenges

Various challenges may further complicate this procedure. The most significant is the inability to reduce or fix the fracture (see Bailing out). Because visualization is key to this technique, any bleeding must be stopped before the procedure can continue. Although a vessel that can be cauterized is occasionally encountered, bleeding is generally controlled with pressure. Shredding the masseter muscle usually obstructs the vision because of the inability to adequately retract the shredded muscle. Therefore, the muscle must be handled carefully while the procedure progresses. Occasionally, the muscle is large or firm, which results in a tight optical cavity. Other challenges include difficulty manipulating instruments in the cavity and breakage of instruments. For instance, the threaded fragment manipulator can fracture, which leaves the threaded portion in the bone. This is generally left in place. Screws may drop into thecavity,andretrievalcan be difficult. They are probably best removed with a suction device. Fragmentation of the proximal fragment during manipulation or screw placement is a frustrating problem. When this occurs, reduction becomes more difficult, and, in some cases, not enough solid bone may remain to allow for plate placement.

Bailing out

In some cases, the fracture may prove extremely difficult or impossible to satisfactorily reduce or fixate. If this occurs, the surgeon needs to decide whether to perform a standard open reduction or to resort to a closed management approach. Based on experience thus far, fractures that result in lateral displacement of the proximal fragment seem to have the greatest likelihood of successful repair. As might be predicted, the lower the fracture, the greater the ease and likelihood of successful repair. Therefore, when a medially displaced proximal fragment proves difficult to reduce or fixate or if the fracture is high or comminuted, the surgeon may opt to abort the endoscopic approach. The decision of whether to open via an external approach should depend on the clinician's judgment concerning the absolute need for open reduction. If closed management is considered, reduction of the fracture should be attempted before the procedure is aborted.

Postoperative management

A postoperative radiologic study should be used to confirm the reduction. Some surgeons advocate the use of a coronal CT scan to ensure the proper alignment of the condylar fragment in the glenoid fossa. No special wound care is required other than routine oral hygiene. A soft diet of 4-6 weeks is recommended. Because the bones have been rigidly fixated, rigid MMF should not be necessary. However, a period of elastic training of the occlusion may prove beneficial. Arch bars may be removed at any time, although some surgeons prefer to wait until satisfactory healing is evident.

Complications

As in any mandible fracture, malunion and, thus, malocclusion are always possible. If the CT scan indicates good position of the condyle in the glenoid fossa, physical therapy should be likely to overcome an early malocclusion. Of course, if the malocclusion is due to malalignment of the bone fragments, reoperation should be considered. Although facial nerve injury is certainly possible with this approach, early experience with this technique has been excellent, with no permanent facial nerve paralyses. However, as experience grows, this complication may become more common. Thus far, nonunion has not been an issue. Postoperative swelling should be anticipated, and hematoma in the cheek may also occur.

In 2011, Domanski et al reported 3 cases in which they had difficulty and complications performing the endoscopic approach. Unfortunately, it is unclear what their level of experience was and how many such cases they had attempted. ^[30]

Last year, Arcuri et al analyzed complications in their series of 17 patients. They had 4 complications in their series, for a rate of 33%. They included aborting the technique because of bleeding, a temporary facial nerve palsy, one poor reduction, and one delayed condylar resorption. ^[31]