O-Arm Assisted Anterior Odontoid Screw Fixation in Type II and Rostral Type III Odontoid Fractures: Single Center Surgical Series of 50 Patients
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.360924
Source of Support: None, Conflict of Interest: None
Keywords: Anterior screw fixation, fusion, O-arm, odontoid fracture, trauma
Odontoid fractures comprise 9–15% of all cervical spine fractures. Clinical presentation is variable from on-sight death to having only neck pain or neurological deficit., Classification given by Anderson and D'Alonzo is the most accepted one. Type II and rostral type III fractures are potentially unstable. Important problems associated with management of these fractures are the risk of nonunion and delayed development of myelopathy.,,
Surgical intervention gives instant fixation and helps in early mobilization of patients. Moreover, many authors have reported better fusion rates with surgical fixation than conservative management., Anterior approach with odontoid screw (OS) fixation is recommended as procedure of choice if fracture anatomy is suitable and transverse ligament is intact and it is associated with fusion rate of 80–100% in various series.,,,,,[5s4] Anterior OS fixation allows anatomical correction with preservation of C1-C2 motion however; this is a challenging procedure.
Several authors have reported the use of navigation with virtual fluoroscopy and Isocentric C-arm three-dimensional (Iso-C 3D) fluoroscopy to increase the precision of anterior odontoid screw placement.,,,,, Recent introduction of O-arm (Medtronic®, Inc, Minneapolis, MN) allow the intraoperative real time assessment of C1-C2 anatomy with intraoperative scan, have better image quality and also allows multi-planar 2D imaging with ease., Moreover, the image quality of the O-arm system is quite superior and is based on the use of distortion-free flat panel technology. Its 30 cm × 40 cm flat panel detector with 3 Mega-pixels provides a high dynamic range for greater image accuracy.
In this study, we have evaluated the precision with which anterior odontoid screw can be placed with help of intraoperative O-arm and ultimately the outcome for anterior odontoid screw fixation in acute odontoid fractures at our institution (level one trauma center).
With our technique of anterior OS fixation under assistance of O-arm intraoperative imaging system (Medtronic®, Inc, Minneapolis, MN), we have treated 50 patients with odontoid fractures from 15 April 2010 to 15th September 2015. This was a retrospective study, patient's hospital records were reviewed for demographic profile, risk factors, injury characteristics (type, displacement etc.), neurological status, interval between injury and surgery, surgical details, clinical and radiographic outcome, morbidity and mortality.
Pre-operative evaluation was done with X-ray, CT scan and MRI to get complete details of fracture anatomy. Transverse ligament disruption was absolute contraindication for anterior odontoid screw fixation and integrity of the same was confirmed with pre-operative MRI in all these patients.
Patient was positioned in supine position on Allen's table (Allen Medical Systems, A Hill-Rom Company, MA) after careful intubation using fiber-optic assistance. Gardner-Wells skull traction tongs were applied in all patients in slight extension to achieve the reduction. We prefer to keep head on table with traction rather than Mayfield head holder as this allows us to subsequently manipulate the head during surgery. O-arm (Medtronic®, Inc, Minneapolis, MN), intraoperative imaging system along with its workstation (Stealth Station) was then brought in place and base line 2D image was taken to check the status of fracture alignment. Proper alignment of fracture segments is prerequisite for the anterior OS fixation. Subsequently patient was prepped and draped along with draping of O-arm with sterile O-arm cover making whole operative field sterile. This motorized O-arm System and its user-friendly controls support quick and easy handling and a smooth workflow. Its gantry rotate by 360° around the patient and it is comfortable to do multi-planar imaging and intraoperative scan in reasonably rapid time without compromising the sterility of operating field. Though O-arm additionally enables one to perform navigated surgery using its navigation system (Medtronic, StealthStation), we did not feel the need of it as it requires rigid head fixation on Mayfield head holder limiting any further head manipulation intraoperatively. Moreover, we were concerned as there is potential danger with navigation with fixed head, as shift of fracture segment may not get detected. Also patient will need additional preoperative 3D scan for navigation.
Standard surgical procedure was used in all patients. Trajectory of screw was planned on lateral X-ray and then right horizontal or vertical skin incision was given at appropriate level (generally C5/C6). Platysma was divided and undermined to allow comfortable retraction. Then pre-vertebral fascia is reached through avascular plane between trachea and esophagus medially and sternocleidomastoid (SCM) muscle and carotid sheath structures laterally. Pre-vertebral fascia was then opened in midline at C2-C3 junction and self-retaining Caspar retractor was used after this stage. To achieve optimum trajectory for screw placement, small anterior and superior part of C3 vertebral body in midline was drilled which will help in maintaining midline trajectory and related part of C2-C3 disc was removed for entry point at base of C2 body and slightly posterior to the anterior margin of inferior endplate of C2 to have better screw hold [Figure 1], [Figure 2], [Figure 3].
Trajectory of screw placement was then confirmed with Kirschner (K) wire over guide using AP and lateral 2D X-ray image. Using drill, K wire was placed across fracture segment toward tip of dense making it bi-cortical under AP and lateral O-arm images. In case of any doubt, final position of K wire was then confirmed by doing intraoperative scan using O-arm. After acquisition, images were then transferred to workstation where 3 D reconstruction of images was performed to analyze the adequacy of its placement in all planes (sagittal, coronal, and axial). This time window allows the subsequent re-adjustment of K wire before placement of final odontoid screw. [Figure 2]c After achieving the proper placement of K wire, single, partially threaded, cannulated lag screw (4 mm) of appropriate length was placed over it achieving lag effect. [Figure 2]d Accuracy of screw placement was then verified with either 2D X-ray or if necessary second intra-operative CT scan if there is doubt about screw position. There after adequate hemostasis is achieved and wound is closed in layers.
We preferred to immobilize the neck in Philadelphia cervical hard collar for period of 12 weeks.
On follow-up CT scan was performed at 3 month and “bony fusion” was confirmed if we found bony trabeculations crossing the fracture site and absence of sclerotic borders adjacent to the fracture site. In doubtful cases dynamic X-ray imaging was performed to confirm the absence of motion on flexion/extension views. In patients where there was no satisfactory fusion, imaging were repeated after 3 months intervals and declared as “non-union” if no fusion occurred even after 9 month. Also, if patient had loss of reduction due to screw cut-through requiring posterior fixation within 30 days after operation, it was labeled as “failure of fixation”.
Since the installment O-arm at our center in 2010, 50 patients (Age- mean 34.6 years, SD 14.10; Range 7–70 years) underwent anterior OS fixation using O-arm for intraoperative imaging and image acquisition. Most of the patients were male (male 44; female 6) and common causes were motor vehicle accidents (66%) and fall (30%). Median waiting period before surgery was 10 days (mean 12.28, SD- 12.85, range 1–65 days). Demography of patient is given in [Table 1]. Most of these patients had type II fractures (Type II-42; Type IIA-3; Type III-5). Details of fracture and clinical finding are given in [Table 2].
Most patients had no displacement (54%) or displacement less than 4 mm (36%) but five patients (10%) also had displacement more than 4 mm. All patients had reduction on traction before surgery. Solitary odontoid fracture was present in 38 patients (76%) while remaining 12 patients (24%) also had associated other spine fractures. 15 patients (30%) also had some neurological deficits which was categorized using ASIA Impairment scale (AIS). Perioperative and follow-up details are mentioned in [Table 3].
Radiological outcome, neurological outcome and complications
Accurate placement of screw was performed in all patients using O-arm resulting in 100% successful screw placement [Table 4]. One patient with high type, type III fracture had failure of fixation at 3rd postoperative day due to screw cut through even after successful screw placement [Figure 4]. Presence of small anterior fractured chip fragment resulting in poor bone purchase could be the reason for screw cut through. Eventually this patient underwent posterior fixation with Magerl's technique complimented by posterior C1-C2 sublaminar wiring resulting in good fusion at 3 month [Figure 5]. Therefore overall 48 patients (96%) had successful OS fixation barring two patients (one had failure of fixation requiring posterior salvage surgery and other one had procedure related mortality).
Out of 48 patients (96%) who had successful OS fixation, follow-up records of 46 patients were available for assessing outcome (mean 26.4 months, range 6 month–54 month, SD 13.75). At 3 months follow-up, 43 patients (93.5%) had good bony union and 2 patients (4.3%) had”'fibrous union: as though there was no evidence of bony union, alignment was maintained on dynamic X-ray. These two patients were treated conservatively with Philadelphia cervical collar and showed delayed fusion. Remaining one patient (2.2%) had nonunion as there was no bony union and also there was secondary loss of reduction. This patient was successfully salvaged with posterior approach using Magerl's fixation complimented by C1-C2 sublaminar wiring and fusion with good results. Therefore overall acceptable fusion rates were 97.8% at their last follow up with nonunion of 2.2%.
In one patient, there was proximal migration of K wire during procedure resulting subarachnoid hemorrhage (SAH) revealed on postoperative CT head. [Figure 6] Due to poor clinical condition, this patient could not be evaluated further and died probably due to neurovascular injury. Another patient had K wire breakage during drilling but successfully managed by odontoid screw placement without any complication [Figure 7]. Therefore, finally OS fixation was successful in 94% patients (47 patients) with overall morbidity and mortality in our series was 4% (2/50) and 2% (1/50), respectively.
Out of 50 patients, one patient deteriorated and had procedure related mortality (2%). On follow-up (n-46), successful preservation of cervical motion could be achieved in 97.8% patients who had successful OS fixation and also 14 patients who had neurological deficits, showed improvement in their neurological deficits.
Odontoid fractures comprise 9–15% of all cervical spine fractures. Important problem associated with management of these fractures is the risk of nonunion and delayed development of myelopathy especially in type II and rostral type III odontoid fractures.,
Literature reveals increasing trend toward surgical fixation of odontoid fractures rather than conservative management as it provide better fusion rates, with acceptable complication rates and with early mobilization of the patients., Since its initial description by Nakashini (1980) and Bohler (1982), anterior approach with OS fixation is now procedure of choice if there is an intact transverse ligaments and good alignment of fracture segments can be achieved resulting in fusion rate of 80–100% in most of the series.,,,, Additionally anterior OS fixation allows physiological correction with preservation of C1-C2 motion.,
Technological advances in various factors like development of better retractor system, lag screws, and intraoperative imaging modality were instrumental in rising popularity of this fixation procedure.
Though there is no doubt that union rate are far better with anterior odontoid fixation as compared to conservative management, still it has 0–20% of nonunion rate reported in various series but can be as high as 62% as reported by Omeis et al. (2009) and hence revealing the scope for further improvement. Regarding long-term outcome, Fountas et al. reported 9.5% failure rate over 24 month follow-up period after anterior fixation. Various factors either related to patient and fracture anatomy or procedure were implicated in etiology of nonunion. While patient related factors are difficult to control, procedure related issue especially of accuracy of screw placement could be controlled which ultimately increases the prospects of successful fusion.
Misplacement of screw is often underreported and is one of the major factors determining the fusion after surgery. [Table 5] summarizes the nonunion rates, re-operation rates, and screw malposition rates in various series over last 15 years showing the scope for the improvement.,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, This table shows the variable reported rate of malposition of screw from 0% to as high as 27.2%.,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Lee et al. (2004) reported the misplacement of screw in 6% of the patients. Apfelbaum (2000) reported the misplacement of screw in 4.5% of the patients, followed by Lee et al. (6%) in 2004; Platzer et al. (2.7%) in 2007; Osti et al. (27.2%) in 2011; Rizvi et al. (20%) in 2012 and steltzlen (14.3%) in 2013.
Traditionally C-arm is used for OS fixation. Proper OS fixation ideally requires bi-planar imaging with either use of two C-arm or when it is not available, single C-arm which has to be adjusted repeatedly in AP and lateral view. But this exercise is very cumbersome, time consuming and it does not provide complete 3Dl real time pictures. Its image quality is also inferior to formal CT scan.,,
Addressing this issue various techniques like use of Virtual fluoroscopy, Iso-C 3D fluoroscopy and cone beam intraoperative CT using O-arm with navigation have been developed in last decade to give real time 3D picture of fracture anatomy to allow accurate screw placement as well as decrease the radiation exposure.,,,,,,
Chibbaro et al. used virtual fluoroscopy for anterior OS fixation in 10 patients and recommended that anterior odontoid screw fixation using image guided surgery with virtual fluoroscopy using FluoroNav (Medtronic, Minneapolis, USA) is a safe, effective, less time consuming and a low X-ray exposure technique. Disadvantage of this procedure is that it don't provide the real time images and he cautioned regarding the shift of fracture segment during screw placement could lead to malposition of screw.
Iso-C 3D fluoroscopy with navigation obviates the need of cumbersome biplanar imaging as well as reduces the radiation exposure has shown to enhance the safety and efficacy of screw placement in anterior and posterior spinal fixation. It was successfully applied for anterior odontoid screw placement by Summers et al. (2008, 9 patients), Martirosyane et al. (2011, 26 patients), Yang et al. (2011, 13 patients) and Kantelhartdts et al. (2012, 6 patients). Martirosyane et al. and Yang et al. compared their results with patients who underwent OS fixation under C arm and found that results are better with Iso-C 3D group though not statistically significant and is the preferred option when available. They recommended Iso-C 3D fluoroscopy as an important tool in increasing the safety of procedure especially in technically difficult cases.,,,
Latest development in this aspect is the introduction of O-arm which has ability to acquire multiplanar 2D as well as intraoperative CT scan at ease and provides real time images with better image quality as compared to 3D fluoroscopy., O-arm along with attached workstation makes it possible to rapidly acquire the images and reconstruct its coronal, sagittal and axial views and plan the trajectory. Also preparation time is less for O-arm. Use of O-arm is widely reported recently for posterior cervical and thoracolumbar fixation and shown to have increased the accuracy of screw placement as well as reduced complication rates.,, However, only few authors have reported its use for anterior OS fixation and there is no standard protocol available regarding how to use it to have optimum benefit. Also, O-arm affords better image quality, allows fast reconstruction of 3D images, and is easy to use when compared to 3D C arm fluoroscopy devices.
Castro-Castro J. recently reported five patients, who underwent anterior OS placement under O-arm guidance. They have performed initial baseline CT scan followed by navigation guided screw placement and status of final screw was then confirmed with second CT scan. They could achieve accurate screw placement in 100% patients using O-arm navigation system. Four out of five patient had good bony fusion and one had fibrous union, which was then treated with posterior fixation in their series.
Since installation of O-arm in our institute in 2010, we have published our series regarding its utility on various spinal procedures including posterior cervical fixation. In this series we have evaluated its use in anterior odontoid fracture surgery. In this series of 50 patients, we could precisely place the anterior odontoid screw in all the type II and type III fracture patients (100%). Verification of the position of K wire with intraoperative imaging with 3D reconstruction in coronal, sagittal, and axial plane before placement of final odontoid screw was important part of our technique whenever there is any doubt about trajectory and bone purchase. This gave us time window to adjust the screw trajectory, as repositioning of K wire is safe and not associated with bone loss as it may occur in repositioning of screw itself and ultimately increasing the prospects of success. Once the final trajectory was confirmed, bone around K wire is drilled and then cannulated screw is placed over K wire and final position is confirmed with 2D image. With this protocol using intraoperative imaging, purpose of precise placement of screw is successfully addressed.
In our series, major complications occurred in two patients. In one patient with type III fracture, developed failure of fixation due to screw cut through on 3rd postoperative day. This patient was subsequently treated with Magerl's posterior fixation complimented by C1-C2 sublaminar wire –iliac crest bone graft fusion. Patient showed good fusion at 3 month follow-up. In another patient, there was significant diffuse SAH in postoperative period resulting in patient death. Intra-operatively there was accidental proximal migration of K wire during K wire placement across fracture segments. This event probably caused the neurovascular injury, which we could not evaluate due to patients poor condition. Occurrence of vascular injury after anterior OS fixation is very rare and only few cases are reported in literature, which can either occurs acutely or remotely due to proximal screw migration. This shows that one should be very careful during the K wire placement especially while achieving bicortical purchase and one should be very slow at that time to avoid this major complication. There was occurrence of K wire breakage during odontoid drilling in one patient. In this tricky case we could place another screw anteriorly using O arm based intraoperative image guidance with successful fusion at 3 month.
At mean follow up of 26.4 month (range 9–54 month) in 46 patients with OS fixation who were available for follow up, overall acceptable fusion rate was 97.8% with nonunion rate of 2.2%. One patient had secondary loss of reduction along with nonunion (2.2%) and was managed with Magerl's posterior fixation. This patient showed good fusion after salvage surgery but had compromised neck movements. This patient had only uni-cortical placement of odontoid screw and this could be reason for his nonunion. So overall 94% patients (47/50) had good outcome after odontoid screw fixation excluding 3 patients (one had fixation failure, one expired and one patient had nonunion) and acceptable fusion rate was 97.8% among those available for follow up.
We also compared our data with our previous cases of anterior odontoid fixation where we used C-arm. Procedural failure rates and union rates in cases done with C-arm and O arm were 4.6% and 95% to 2% and 97.8%, respectively. These results were also better when we compared it with many other published series with either C-arm or with 3D fluoroscopy [Table 5].,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
Average time taken for our procedure was 135 min (45–220 min) which was higher in comparison with what was reported by Martirosyan et al. (83 min), Yang et al. (91.5 min) and Kantelhardt et al. (108 min) as it was overall time of the procedure rather than surgical time. Since it was retrospective study we did not have the accurate surgical time and could not calculated the time required for setting the O-arm in these cases however, according to our experience its not more than that required while setting the C-arm as very user friendly and intraoperative image transfer and reconstruction is fast. Also with the experience and familiarity of OT personnel, its very easy to use but require little more space than C-arm.
Special circumstances: Type III fractures and anterior oblique type II fracture
This technological advancement also allowed us to surgically treat the technically difficult odontoid fracture patients having Type III fractures and anterior oblique Type II fracture.
Type III fracture
We could accurately place the odontoid screw in all the five patients (100%) with type III fracture confirmed by intraoperative imaging. It was successful in 4 patients (80%) while one patient developed fixation failure (20%). This patient had small additional anterior chip fracture and hence he developed screw cut through on 3rd postoperative day. The reason for screw failure was additional chip fracture resulting in bone loss with poor screw holding strength. This patient was managed with Magerl's posterior fixation with good bony fusion at three month. This implies that for successful anterior odontoid screw fixation in type III odontoid fracture cases require proper selection. This case again stressed the importance of availability of two large fracture segments to strongly hold the screw in place to avoid fixation failure.
Anterior oblique type II fracture
Type II fractures with anterior oblique fracture line are generally not recommended by many for anterior fixation due to difficulty in planning the trajectory with subsequent high nonunion rates. Posterior fixation techniques in these patients again have morbidity of restricted neck motion. However, with O-arm assistance we could negotiate the accurate trajectory in selected three patients having less steep anterior oblique fracture with 100% fusion rate. But again proper preoperative planning and case selection is important here.
Potential disadvantage of using O-arm is its cost but beside odontoid fracture various spinal pathology can be safely and effectively dealt with as per recent literature and surely it's a very good adjunct to provide quality care in institutional establishment and we have explored its benefits in anterior odontoid screw fixation. Also its is very good tool for the novice surgeon giving better intraoperative information.
Our study has some limitations. First it is a retrospective study but considering large number of patients treated by this innovative technology, it will definitely enlighten us further regarding increasing the safety of fixation as well as adding more options in one's armamentarium to provide more advanced and more effective care in almost all cases of odontoid fractures including technically difficult type II and rostral type III odontoid fractures. Second one is that there is no control group for comparison hence we have to rely on historical controls. However besides good results, this study also highlights the probable grave complications that one should be careful about.
Following on from this study, further research on effective use of navigated O arm guided surgery and its comparison with C arm in prospective manner for set up time, operative time, overall safety including radiation exposure and accuracy of anterior odontoid surgery would be valuable.
We conclude that use of intraoperative three-dimensional imaging using O arm for anterior odontoid screw fixation adds precision as well as increases the success of surgical techniques by increasing accuracy of anterior odontoid screw placement. It enables us to further extend the indications of odontoid screw fixation to selected complex Type II and rostral Type III odontoid fractures and therefor allowing us to do the more physiological neck motion preserving surgery in these subsets of patients.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]