Case report

A 34-year-old woman was referred for surgical correction of severe maxillary transverse deficiency. The patient’s medical history was unremarkable.

SARME was performed under general anesthesia. After local anesthesia was administered with a vasoconstrictor, an incision was made in the area of the future abutment plate over the second premolar root. A rapid palatal expander distractor (KLS Martin, Tuttlingen, Germany) was placed with the plates on the bone. The distractor was slightly activated. The pins penetrated the bone and stabilized the distractor. The distractor was secured using two additionally supplied drill-free screws through the holes of the distractor plates, following the protocol of Mommaerts .

The distractor was activated before standard corticotomies for better adaptation of transpalatal distractor. A small surgical bur was used to make horizontal bone cuts bilaterally, and a corticotomy was performed through the zygomatic buttress from the piriform rim to the maxillopterygoid junction. The medial wall of the maxillary sinus was not included in the cut. Pterygomaxillary separation was performed using a curved osteotome, and the maxilla was split sagittally using a thin, straight osteotome.

The distractor was activated by opening a midline diastema of approximately 3 mm to check if it would expand properly after surgery; it was then reset to the starting position. The appliance was decompressed, and the surgery was completed with 2 mm diastema.

On the day after the surgery, the patient complained of ptosis in the right eyelid, ‘blurry’ vision, and right eye pain. Lesions on the III, IV and VI cranial nerves were observed. The patient was unable to move the right eye. The pupil was dilated and unreactive to light. On the second postoperative day, she complained of paresthesia in the trigeminal nerve area (ophthalmic branch).

A computed tomography (CT) scan showed a linear fracture on the posterior surface of the maxillary sinus that extended to the left minor wing of the sphenoid. Some other fractures were noted on the left sphenoid sinus, and the authors observed displacement of the edge of the anterior clinoid process over the internal carotid artery. Magnetic resonance imaging (MRI) showed an abnormal saccular structure in the right cavernous sinus with flow-void signals and vascular ectasia that compressed and thus congested the lateral cavernous sinus. The findings suggest a saccular aneurysm caused by laceration at the cavernous segment of the internal carotid artery.

An angiogram showed a high-flow direct carotid-cavernous fistula that was fed by the internal carotid artery at the right side, which had a large laceration extending to the upper and medial surfaces. The angiogram showed that the extent of the injury was higher at the internal carotid artery but lower at the carvernous sinus, so an arteriovenous shunt was required ( Fig. 1 ). The shunt was high, and therefore, the blood did not flow through the medial and anterior group of cerebral arteries on the ipsilateral side, which allowed almost all the blood to flow through the venous route. The venous drainage route was antegrade through to the basilar plexus via the ipsilateral cavernous sinus, and no cortical venous reflux was observed. An extensive venous dilatation into the cavernous sinus was observed at the shunt. The patient underwent shunt occlusion via the remodelling technique for transarterial and homogeneous coil packing in the cavernous sinus. The final control during the therapeutic procedure showed shunt occlusion, improved arteriovenous timing and reduced cavernous sinus congestion ( Fig. 2 ).

Angiogram showing carotid cavernous fistula and venous dilation on the right side.

Final control after coil embolization without the shunt (A). Frontal view without subtraction showing the coil packing (B).

The patient did not experience any right orbital pain immediately after the procedure. At the 3-month follow-up, the patient did not have any neurological deficits. An angiogram taken 6 months after the procedure showed that coiling stability without any recanalization at the point of the shunt.

Discussion

SARME has potential repercussions beyond maxillary widening because the expansion of the appliance allows the forces that are generated to be widely distributed in other regions of the craniofacial complex . Rapid maxillary expansion was performed in an experimental study in monkeys using tetracycline as a bone marker, and it was observed that the lateral and medial pterygoid plates, hamulus, infratemporal surface of the sphenoid bone, zygomatic arch, major wing of the sphenoid bone, infratemporal surface, and alveolar process of the jaw showed an increase in the amount of tetracycline absorbed, suggesting the stimulation of bone growth or bone remodeling .

Shetty et al. analyzed the stress responses generated by the activation of the hyrax appliance adapted to a photoelastic analog of the human skull. The initial activation before the cuts, which simulates osteotomies, produces forces that are transmitted over the traditional midface pillars, crosses the pterygomaxillary articulation, and dissipates in the pterygoid plates. Tensions have been observed in the orbital portion of the major wing sphenoid. After the cuts were performed, a decrease in the resistance of the maxilla to transverse expansion was observed, indicating an increase in stress in other areas such as the pterygomaxillary articulation, frontonasal suture, and lateral nasal wall. When separation of the pterygomaxillary articulation occurred, there was an increase in the tensions in remote locations, including the zygomatic-frontal suture, lower part of lateral nasal wall, zygomatico-maxillary suture, zygomatic arch, and supraorbital and frontal regions.

Although this experimental study requires clinical proof of the disjunction of craniofacial surgery, the data are relevant if the complications discussed in the present case and in the case reported by Lanigan and Mintz are considered. They reported the case of a patient with a fracture in the posterior aspect of the left maxillary sinus that extended to involve the left body of the sphenoid bone with fractures of the floor and roof on the left sphenoid sinus, resulting in the development of ptosis and ophthalmoplegia due to oculomotor, abducens and facial nerve palsies, which resolved 4 months after the initial maxillary expansion procedure.

The present case is noteworthy because the fracture involved the body of the sphenoid bone and the right sphenoid sinus and also because the patient developed carotid-cavernous sinus fistula following SARME.

In the present study, experienced surgeons, who routinely perform osteotomies via the zygomatic-maxillary, intermaxillary, and pterygomaxillary approaches, performed the operation. The transpalatal distractor was activated before the osteotomies, which may have generated a strain that may have caused the fractures.

Lanigan and Mintz postulated that a curved pterygoid osteotome should no longer be used because its use can result in high incidences of pterygoid plate fractures, which have the potential to disrupt contents of the pterygopalatine fossa. Pterygomaxillary osteotomies using a curved chisel have been related to pterygoid plate, skull base, and orbit fractures .

Newhouse et al. reported that a patient who underwent a Le Fort I osteotomy suffered from a stroke. With pterygomaxillary disjunction, the right pterygoid complex is also suggested to detach and posteriorly force a sharp piece of bone into the vessels, lacerating the vessels and causing subsequent vascular injury.

A case of complete, isolated, and unilateral (right) abducens nerve palsy has been reported 5 days after a Le Fort I osteotomy. The CT of the patient showed a fracture of the right pterygoid plate and blood in the right sphenoid sinus. The patient completely recovered without intervention over a 6-week period. It was postulated that the cause may have been compression or distraction of the sixth cranial nerve that was caused by a suspected hairline fracture in the sphenoid sinus that extended to the orbit and resulted in temporary displacement.

The amount of force generated by the expansion device must also be taken into consideration during the procedure. For a stage 1 procedure, surgeons should not expand the maxilla to its maximum width intraoperatively . Instead, expansion should be achieved in a slow, controlled manner over days to weeks depending on the amount of expansion required. Attempting to achieve required expansion intraoperatively at once is not only a dangerous practice that greatly increases the chances of untoward fractures but also violates the concept of SARME as a technique used to achieve distraction osteogenesis .

The rapid palatal expander distractor was used in the present case. This distractor is supported at 2 points, and the load is not shared with the teeth unlike the case with the hyrax distractor, which is the most commonly used. The appliance was activated before performing the osteotomy; this activation, which is supported on the bone, may have caused the undesirable fractures in the skull base. The fractures may have also been because of pterygoid separation and the excessively wide expansion achieved at surgery.

The present case and the case reported by Lanigan and Mintz should alert surgeons to the complications involved in SARME, such as fractures of the skull base, aneurysms, arteriocavernous fistulas, and injuries involving the cranial nerves and eyelids. Biomechanical studies related to craniofacial SARME, especially those that compare the various types of equipment used in SARME, are few. Such studies are required to increase knowledge on the factors responsible for undesirable fractures associated with SARME.