Management of Acute Odontoid Fractures: Operative Techniques and Complication Avoidance

Shushil Shilpakar SK, MD, Mark R. Mclaughlin, MD, Regis W. Haid, Jr. MD, Gerald E. Rodts, Jr., MD, Brian R. Subach, MD, Tribhuvan University Teaching Hospital, Kathmandu, Nepal; and The Emory Clinic, Emory University Hospital, Atlanta, Georgia

Neurosurg Focus. 2000;8(6) 

In This Article

Surgical Techniques

Single-Screw Anterior Odontoid Screw Fixation

We prefer to use a cannulated technique for anterior screw fixation.[6,19,35,55] After fiberoptic intubation with in-line stabilization, the lateral C-arm fluoroscope is brought into position and canted under the operating table (Fig. 2). Under fluoroscopic guidance, the patient is placed in the supine position with the neck extended. All attempts are made to minimize displacement of the odontoid to facilitate the ideal trajectory for screw placement into the apex of the odontoid process. A radiolucent mouth gag is used to keep the mouth open to provide a transoral AP view. Because biplanar C-arm fluorosopy is required to perform this procedure, the second unit is put into the AP position to obtain simultaneous projections prior to preparation of the patient.

Figure 2. Photograph of operating room showing the patient in the supine position prior to anterior odontoid fixation. Note that the lateral C-arm fluoroscope is canted under the table to allow for the second AP C-arm fluoroscope to be well positioned.

A small, transverse midcervical skin incision is made over the level of the cricothyroid junction (the C5-6 interspace). Similar to a standard Cloward-type approach[17] performed for anterior cervical discectomy, blunt dissection is performed down within the avascular plane to the prevertebral area and then extended cephalad to the C2-3 disc space. The trachea and esophagus are retracted medially, and the sternocleidomastoid muscle and the carotid sheath are mobilized laterally by using a Cloward hand-held retractor. The prevertebral fascia is bluntly dissected from the vertebral bodies of the midcervical spine by using a Kittner dissector. The dissection is then performed cephalad on the midline under fluoroscopic guidance to reach the C2-3 interspace. The hand-held retractors are then replaced by sharp-toothed, transverse self-retaining retractor blades, which are placed at C5-6 to obtain adequate exposure. The C2-3 interspace level is verified using the lateral C-arm fluoroscopic view. The acutely angled superior retractor blade is then placed in the submandibular area and anchored to the lateral retractor system to provide rostral exposure. We have found the Apfelbaum retractor system (Aesculap, San Francisco, CA) to be most suitable for this purpose. In this way, an adequate working channel to the C2-3 interspace is created.

For screw placement we prefer the Universal Cannulated Screw System (Medtronic Sofamor Danek, Memphis, TN). Using biplanar fluoroscopy, a 2-mm threaded K-wire is anchored into the midline of the anteroinferior edge of C-2. By applying a prying motion with the K-wire anchored into C-2 and levered on C-3 in the dorsal-to-ventral direction, the C-2 body is rocked slightly anteriorly, allowing for adequate trajectory to traverse the fracture line and incorporate the superior fragment.

After the drill guide and K-wire are anchored and the trajectory is deemed adequate, the K-wire is then drilled through the fracture line into the distal portion of the fractured dens. The K-wire should be drilled approximately 1 mm beyond the distal odontoid fragment to incorporate the outer cortex. The inner guide tube is then removed, and the K-wire is left in place. A pilot hole is then drilled into the C-2 body with the cannulated drill bit sliding over the K-wire across the fracture line and into the apical cortical surface of odontoid process.

Before the fracture line is crossed, additional head manipulation can be attempted to realign better the odontoid fracture prior to screw placement. Once the pilot hole is made, the depth of penetration of the drill can be determined by reading the markers on the drill bit or K-wire shaft. The drill is then withdrawn, and the drill hole is tapped over the K-wire to cut threads in the interface. A 4-mm nonself-tapping, partially threaded titanium cannulated lag screw is then gradually inserted through the guide tube over the K-wire, under direct fluoroscopic view, and countersunk 2 mm into the C-2 vertebral body. The tip of the screw should barely penetrate the apical cortex of the odontoid process, and it should be just adequately tightened to engage the distal odontoid fragment (Fig. 3). It is important to choose a slightly shorter screw length than is measured by approximately 2 mm because the lag effect derives only from final tightening of the screw to increase compression and to approximate the fracture line. The neck is manually flexed and extended, and alignment and screw placement are verified on biplanar fluoroscopy.

Figure 3. Lateral fluoroscopic C-arm view of final position of single-screw placement engaging the superior cortex of the odontoid tip.

Surgical Techniques

Posterior C1-2 Transarticular Screw Fixation With Supplemental Tension-Band Fixation

The patient is positioned prone in a rigid three-point pin fixation system. The neck should be slightly flexed in a "military tuck" position, creating posterior translation and reduction, as well as allowing the surgeons to achieve the desired C1-2 trajectory for the instruments. Lateral C-arm fluoroscopy is used to assess planned screw trajectory prior to preparation. The entire posterior cervical region down to the mid-thoracic area is prepared and draped. After a midline incision is made to expose the posterior elements of C1-3, careful attention is directed toward exposure of the C2-3 facet joints. The C-2 pars is exposed and palpated with a Penfield No. 4 dissector, which is used to dissect the C-2 nerve root with gentle sweeping movements in a rostral direction to expose the medial pars. The roof of the C-2 pars is carefully followed into the C1-2 facet joint. After obtaining this rostral dissection, the No. 4 Penfield dissector is directed medially to palpate the medial border of the C-2 pars. This maneuver allows the surgeon to assess the pedicle angulation and screw direction. Venous bleeding can be minimized by centering the dissection on the medial aspect of the pars. Careful dissection and bipolar coagulation of the medial venous complex can also minimize bleeding. Any venous bleeding that is difficult to manage is tamponaded with Gelfoam (Pharmacia Upjohn, Berwyn, PA).

For the drill guide a percutaneous entrance is typically required to achieve the desired trajectory for screw placement. To choose the optimum entry point a long instrument is placed adjacent to the neck, outside of the wound, paralleling the ideal trajectory. This allows the surgeon to judge proper visual alignment of the drill and screw across the C1-2 facet joint with fluoroscopic imaging, and it also demonstrates the optimum entry point for the drill guide through the skin, which is typically two finger-breadths paramedian to the T-1 spinous process. Bilateral stab incisions are made at the paramedian entrance site of the drill guide.

The guide tube is then passed through the stab incision and directed toward the exposed C2-3 facet joint into the open operative site. The tip of the guide tube is docked at the C-2 entry site.

Reduction of C1-2 space must be accomplished prior to drilling the K-wire. In patients with significant subluxation a cable is placed beneath C-1, and the posterior arch is pulled up. Application of manual pressure on the C-2 spinous process can improve alignment. This technique can be used temporarily until the first K-wire is drilled into position, and then the tension can be released. Once the K-wire is placed, often the reduction will remain in place. Alternatively, the interspinous tension band construct can be fixated first to achieve reduction of the fracture prior to undertaking transarticular screw fixation. If the tension band construct is placed prior to implanting the transarticular screws to improve alignment, we recommend only provisional clamping of the cable because the construct can sometime loosen after transarticular screw fixation. After alignment is achieved and the transarticular screws are in place, the posterior tension band then can be permanently tightened.

The inferomedial edge of the C2-3 facet joint is identified, and the C-2 entry site is marked approximately 3 to 4 mm rostral and 3 to 4 mm lateral to this point (that is, from the inferomedial facet joint, up 3 mm and out 3 mm). Using a high-speed drill and mini-match stick bit, the cortical bone is pierced to mark the K-wire entry site. The K-wire trajectory is visualized on lateral fluoroscopy. Drill angles typically vary from 5 to 15° medial in the saggital plane. The K-wire is drilled into the C-2 pars and across the C1-2 joint, aimed at the anterior tubercle of C-1. The ideal K-wire tip terminus is a point 3 to 4 mm dorsal to the tip of the anterior C-1 tubercle, barely penetrating the cortical bone of the C-1 lateral mass.

Using a Penfield No. 4 dissector, the articular surfaces of the C1-2 facet joint and the dorsal cortex of the C-1 lateral mass are palpated. Often the K-wire can be visualized as it traverses the C1-2 facet joint. While drilling the K-wire, the surgeon can perceive subtle changes in resistance as the wire traverses the cortical surfaces along its pathway into the C-1 lateral mass. The surfaces include the superior articular surface of C-2, the inferior articular surface of C-1, and the anterior cortex of the C-1 ring. Intermittent pulses with the drill, rather than continuous drilling, will give the surgeon better proprioceptive feedback and control as the K-wire is passed through the C1-2 facet joint. Once the K-wire is drilled into position, a cannulated drill bit is then placed over the K-wire and drilled to the depth of the K-wire fixation point. An assistant holds the K-wire with a needle driver as the drill bit is advanced under fluoroscopic guidance to prevent advancement of the K-wire.

The pilot hole is then tapped over the K-wire into the lateral mass of C-1. A fully threaded 3.5- or 4-mm cortical screw of premeasured length is used. We prefer 4-mm screws, if possible, and attempt to place this sized screw on all of our patients in whom preoperative Stealth scan (Medtronic Sofamor Danek) measurements have been obtained. Screw length can be measured directly from the drill or K-wire inserted under C-arm fluoroscopic control, or it can be premeasured. Screws are typically 38 to 42 mm in length. The technique is repeated on the contralateral side if anatomy permits. It is important to use nonlag screws when performing transarticular screw fixation because the screw functions more as a nail-stabilizing rotation than as a compressive force. Lag screws are not necessary for this construct because the goal of the screws is to prevent rotation at the facet joints. The screws in no way need to lag the facet joints together. In addition fully threaded screws have more surface area and, thus, more screw-bone interface than partially threaded lag screws. Therefore, fully threaded cortical screws increase the rigidity of the total construct compared with lag screws. Data obtained from laboratory investigations have shown equal efficacy and strength of cannulated and noncannulated screws of similar diameter (Dickman, personal communication).

In all cases of transarticular screw fixation we use supplemental posterior tension-band constructs by performing the Sonntag modification of the Gallie-type fusion with Atlas titanium cables (Medtronic Sofamor Danek). This technique provides added stability, particularly in flexion and extension, without the potential morbidity associated with passing sublaminar wires under C-2. We have also used a posterior clamp fixation system (Apofix, Medtronic Sofamor Danek) for tension-band constructs, which has yielded excellent results (Fig. 4). Decortication of the inferior C-1 ring prior to final placement of the graft is critical, as it has been our experience that this region is the region in which a pseudarthrosis will most commonly occur. The graft for the posterior construct is harvested more laterally along the posterior iliac crest, approximately five finger-breadths lateral to the posterosuperior iliac spine. In this region the pelvis takes on an anatomical configuration more suited for the anatomy of the C1-2 region. The curve of this portion of the ilium matches that of the C1-2 interspace. Although the incision necessary for harvesting a more lateral graft (five finger-breadths off of the midline) is slightly more lateral than the standard incision, it is important to stay within 8 cm medial to the posteroinferior iliac crest to avoid injury to the cluneal nerves. A tricortical autologous iliac crest graft is harvested, and the superior cortex is removed using an oscillating saw. The remaining curved rectangular bicortical graft is placed with medullary bone interfacing both the decorticated inferior surface of C-1 and the decorticated superior surface of C-2.

Figure 4. Postoperative radiographic studies. Left: Lateral cervical spine x-ray film obtained in a patient 2 years after undergoing C1-2 transarticular screw fixation and posterior tension banding with Apofix for treatment of an odontoid fracture with transverse ligament disruption. Note the solid arthrodesis with confluent bone incorporating the posterior graft, as well as the engaged superior cortex of the C-1 lateral mass. Right: Cervical spine x-ray film (AP view) obtained in the same patient, demonstrating optimum screw position.

Postoperative Management

Postoperative management is similar for patients who have undergone either an anterior odontiod screw fixation or a transarticular screw fixation procedure. The patients are monitored closely for neurological and respiratory changes in the recovery room and are then transferred to the ward. Almost all patients are discharged from the hospital within the first few days of surgery regardless of which surgical approach was undertaken.

Patients in whom either of these procedures are performed undergo rigid cervical collar for 12 weeks postoperatively. Outpatient follow up is typically conducted at 6, 12, and 24 weeks as well as at 1 and 2 years postoperatively. Clinical assessment of functional status and postoperative flexion-extension radiographic evaluation are performed during these sessions. We also obtain CT scans with sagittal reconstructions if the adequacy of fusion can not be determined on plain x-ray films.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: