Novel Surgical Classiﬁ cation and Treatment Strategy for ...

E1348 www.spinejournal.com October 2013

SURGERY

SPINE Volume 38 , Number 21 , pp E1348 - E1356 ©2013, Lippincott Williams & Wilkins

Novel Surgical Classifi cation and Treatment Strategy for Atlantoaxial Dislocations

Shenglin Wang , MD , Chao Wang , MD , Ming Yan , MD , Haitao Zhou , MD , and Gengting Dang , MD

DOI: 10.1097/BRS.0b013e3182a1e5e4

Study Design. Retrospective study of 904 patients with a diagnosis of atlantoaxial dislocation (AAD), using a novel surgical classifi cation and treatment strategy. Objective. To describe a novel surgical classifi cation and treatment strategy for AADs. Summary of Background Data. AADs can result from a variety of etiologies, yet no comprehensive classifi cation has been accepted that guides treatment. Because of the rarity of the cases, however, the treatment strategy has also been debated. Methods. During a period of 12 years, a total of 904 patients with a diagnosis of AAD were recruited from a single academic institution. According to the treatment algorithm that included preoperative evaluation using dynamic radiograph, reconstructive computed tomography, and skeletal traction test, the cases were classifi ed into 4 types: I to IV. Types I and II were fused in the reduced position from a posterior approach. Type III, which were irreducible dislocations, were converted to reducible dislocations using a transoral atlantoaxial release, followed by a posterior fusion. Type IV presented with bony dislocations and required transoral osseous decompressions prior to posterior fusion. Results. Four hundred seventy-two cases were classifi ed as type I, 160 as type II, 268 as type III, and 4 cases as type IV. Follow-up was in the range of 2 to 12 years (average: 60.5 mo). Eight hundred and ninety-nine cases (99.4%) achieved a solid atlantoaxial fusion. Anatomic atlantoaxial reduction was achieved in 892 cases (98.7%), whereas 12 cases had a partial reduction. Neurological improvement was seen in 84.1% (512/609) of the patients with myelopathy. The overall complication rate was 9.1% (82/949). Conclusion. Our surgical classifi cation and treatment strategy for AADs was applied in those 904 cases and associated with excellent clinical results with a minimal risk of complications.

From the Orthopaedic Department, Peking University Third Hospital, Beijing, China.

Acknowledgment date: October 1, 2012. First revision date: January 31, 2013. Second revision date: June 12, 2013. Third revision date: June 18, 2013. Acceptance date: June 21, 2013.

The device(s)/drug(s) that is/are the subject of this manuscript is/are exempt from FDA or corresponding national regulations.

No funds were received in support of this work.

No relevant fi nancial activities outside the submitted work.

Address correspondence and reprint requests to Chao Wang, MD, Orthopaedic Department, Peking University Third Hospital, 49 North Garden St, Haidian District, Beijing, China, 100191; E-mail: [email protected]

Atlantoaxial dislocations (AAD), although uncommon, can be associated with complex deformities of the craniovertebral junction and represent a potential for

serious clinical consequences, including cervical myelopathy, respiratory failure, or even death. 1 , 2 Loss of the normal ana-tomical articulation between the fi rst and second cervical verte-brae signifi es the presence of an atlantoaxial subluxation. The unique position of the odontoid process, rising between the ring of the atlas ventrally, together with the transverse atlantal liga-ment dorsally and the C1–C2 lateral mass joints are the major factors preventing dislocation of C1–C2, 3 with any disruption in the integrity of these structures predisposing to AAD devel-opment. 4 – 6 With regards to etiology, AAD can be divided into 3 main categories: traumatic, spontaneous, and congenital AAD. 3 On the basis of the reducibility, AAD can be further classifi ed as reducible and irreducible (fi xed). 3 , 7 , 8 However, the character-ization of reducibility is somewhat controversial because some surgeons depend on dynamic radiographs, 9 , 10 whereas others make this judgment based on preoperative traction. 3 , 11

AAD, particularly in the setting of clinical symptoms, typi-cally requires surgical intervention. Despite this consensus, the actual treatment rendered has been less consistent. Several authors have recommended preoperative skeletal traction to facilitate reduction, 12 , 13 whereas others have used skeletal trac-tion only during the surgical procedure. 14 , 15 Similarly, multiple surgical approaches and techniques have been described, with anterior, 9 , 16 , 17 posterior, 14 , 18 or combined approaches, 19 , 20 along with varying instrumentation techniques, all being supported in respective case series. Collectively, however, existing reports have had small sample sizes, lacked controls, and provided limited duration of follow-up. Because of these limitations of the existing literature, coupled with the rarity of these cases, the optimal surgical classifi cation and treatment strategy for AAD remains unclear.

MATERIALS AND METHODS

Patient Recruitment During the period extending from June 1998 through December 2010, patients with a diagnosis of AAD were recruited from a

Key words: atlantoaxial dislocation , classifi cation , treatment strategy , transoral surgery , atlantoaxial release . Level of Evidence: 4 Spine 2013;38:E1348–E1356

Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

BRS205732.indd E1348BRS205732.indd E1348 03/09/13 10:30 PM03/09/13 10:30 PM

drwang

高亮

SURGERY Novel Treatment Strategy for AAD • Wang et al

Spine www.spinejournal.com E1349

single tertiary care referral center. The sole criterion for AAD was an atlantodental distance greater than 3 mm among adults ( ≥ 18 yr), whereas in younger individuals, a distance greater than 5 mm was considered abnormal. For patients with an incomplete odontoid present (os odontoideum or evidence of a prior odontoid fracture), the same criteria were adopted by measuring the distance between the inferior rim of the C1 anterior arch and the remaining attached part of the odontoid or C2 anterior-superior edge. 21 Patients were excluded if they had evidence of a tumor, acute fracture, or previous surgery at the craniovertebral junction. Institutional board approval was obtained prior to initiating this retrospective review. Before surgery, the patients underwent dynamic lateral radiography, and reconstructive computed tomography (CT) and magnetic resonance imaging of the cervical spine.

Classifi cation The classifi cation system used for grading severity in this series was based on the after 3 consecutive diagnostic tests ( Figure 1 ):

Dynamic Lateral Radiograph If an anatomical atlantoaxial reduction could be achieved on the fl exion or extension radiographs, the patients were judged as “type I: instability” (IN) ( Figure 2A, B ). Those who could not achieve full reduction on the dynamic radiographs ( Figure 2C, D ) were evaluated by the following steps.

Reconstructive CT If the presence of an osseous fusion between C1 and C2 was found on the reconstructive CT scan ( Figure 3 ), a classifi ca-tion of type IV was given, that is, a “bony dislocation.” Only those cases without evidence of a C1–C2 osseous fusion would then undergo the following step.

Skeletal Traction Test Under General Anesthesia In the operating room, cranial traction (using Gardner-Wells tongs) was performed under general anesthesia ( Figure 4C, D ) and fl uoroscopic guidance, with gradual incremental increases in the applied traction weight. Before October 2003, the trac-tion weight used was typically in the range of 6 to 10 kg, with a maximum of no more than one-tenth of the patients’ weight. 19 After that time interval, the traction weight used was modifi ed to a maximum of one-sixth of the patient’s body weight. The patient was placed in 15 ° to 20 ° of reverse Trendelenburg posi-tion to provide countertraction. Ten minutes after complete muscular paralysis, the reducibility was assessed radiographi-cally. Those with a complete reduction were classifi ed as “type II: reducible dislocation” (defi ned as “RD”; Figure 4E ). If not, “type III: irreducible dislocation” was given (defi ned as “ID”; Figure 4F ). After the test was performed, the soft tissues and ligaments of the craniovertebral junction were relaxed by the traction, and the atlantoaxial joint was further destabilized. Therefore, the following treatment procedures were per-formed immediately, with the traction carefully maintained while going from the supine to prone position.

Surgical Treatments

Posterior Reduction and Fixation Types I and II AAD were relatively easily fi xed and fused in the reduced position from a posterior approach. For these 2 categories, the following 4 techniques of short segment pos-terior fi xation were used:

Transarticular C1–C2 Fixation (the Modifi ed Magerl Tech-nique 15 ) . Cases without occipitalization of the atlas and/or swan neck deformities 22 underwent this procedure (78 cases).

C1 Lateral Mass Screw and C2 Pedicle Screw Fixation (the Goel 14 and Harms 18 Procedure). The cases without


Atlantoaxial dislocation: Loss of C1–C2 stability

Flexion and extension radiograph

Dislocation

I: Instability

Reconstructive CT

Skeletal traction under general anesthesia (OR)

Without C1–C2 bony fusion

II: Reducible dislocation

III: Irreducible dislocation

IV: Bony dislocation

With C1–C2 bony fusion

Not reduced

Dynamic reduction

IN RD ID BD

Figure 1. The classifi cation tree. ID indicates instability; RD, reducible dislocation; ID, irreducible dislocation; BD, bony dislocation; CT, computed tomography; OR, operation room.


drwang

高亮

drwang

高亮



occipitalization of the atlas, including those with swan neck deformity, underwent this procedure (404 cases).

Occiput to C2 Fixation Using C2 Pedicle Screws (the Abumi Technique 23 ). Patients with occipitalization of the atlas or dysplasia of the C1 posterior arch underwent this proce-dure (368 cases).

C1–C2 or Occiput to C2 Fixation Using C2 Laminar Screws . If patients had a vertebral artery (VA) that was con-sidered to be high riding, or C2 a pedicle osseous abnormality, C2 laminar screws (LS) were used for C2 fi xation in those undergoing C1–C2 or occipitocervical fi xation. In total, this was performed in 50 cases, including 46 cases with unilateral LS and 4 cases with bilateral LS. The technique of Wright 24 was used for these LS. Because of the polyaxial head of the screws and acute bending of titanium rods required to con-nect the laminar fi xation points, the instrumentation was con-

sidered to be biomechanically inferior than that of C2 pedicle screws. 25 Therefore, we used halo-vests postoperatively for supplemental stability in the 4 cases with bilateral C2 LS.

Transoral Atlantoaxial Joint Release: Conversion From Irreducible to Reducible AAD (268 Cases) Type III patients (irreducible dislocation) underwent the fol-lowing procedure. Continuous skull traction was maintained during the procedure. The surgical bed was tilted 15 ° to 30 ° with the head elevated. The bilateral longus coli and longus capitis muscles, and the anterior longitudinal ligament were dissected just caudal to the anterior ring of C1. In the setting of occipitalization of C1, the caudal half of the anterior ring of the atlas was also removed. The anterior joint capsules were excised, and the cartilage of the bilateral C1–C2 lateral joints were curetted and removed. The apical and alar ligaments

Figure 3. A bony fusion between C1 and C2 was found on the reconstructive CT scan, a classifi cation of type IV was given. CT indicates computed tomography.

Reconstructive CT showed bony fusion between left C1–C2 lateral mass joint


Figure 2. Dynamic radiographs. Flexion radiograph showed anterior AAD ( A ); Extension radiograph showed posterior subluxation of C1, and the radio-graphs demonstrated instability in both fl exion and extension (type I) ( B ); Flexion radiograph showed severe anterior AAD ( C ); Extension radiograph showed complete reduction was not achieved ( D ) (the case was subsequently evaluated by following the steps described below). AAD indicates atlanto-axial dislocation.

Flexion Extension

FlexionExtension

A B

C D

One patient

Another patient




were detached from the odontoid process. At this point, the curette could be placed at the cephalic portion of the odontoid process to reduce it inferiorly and anteriorly. The release pro-cedure was deemed complete when the dens could be easily levered up to contact the C1 anterior arch using a curette, sig-nifying anatomic reduction of the AAD. The procedure was confi rmed using C-arm.

After the conversion from an irreducible to a reducible AAD, the previous posterior techniques were performed immediately.

Transoral Odontoidectomy If a patient had been classifi ed as a type IV (bony disloca-tion), then a odontoidectomy was performed via a transoral approach, as has been described by Menezes et al. 26 Four cases were classifi ed as type IV dislocations. Two of them under-went odontoidectomies, the other 2 cases declined surgery. Two of the type III dislocations also underwent odontoidec-tomies during the earlier years of our series.

Postoperative Treatments Drains were removed at 48 hours. After these patients were mobilized out of bed, initially with assistance. No external fi x-ation, such as a cast or orthosis, was used after surgery, except the 4 cases with bilateral C2 laminar screws. These 4 were placed in a halo vest. For those patients who underwent a tran-soral approach, oral feeding was not allowed until the seventh postoperative day. Early on in our series, a tracheotomy was performed before the transoral approach and kept in place for 5 days (191 cases). After 2008, tracheotomies were aban-doned, and the endotracheal tube was removed soon after recovery from general anesthesia (79 cases).

All patients underwent postoperative radiographs, recon-structive CT scanning, and magnetic resonance imaging of the cervical spine at 5 to 7 days and 4 to 6 months. Fusion

was confi rmed using postoperative reconstructive CT scans, with a successful fusion defi ned as continuous osseous union between the C2 lamina and the C1 posterior arch (or occiput in occipitocervical fusions). 27 Then, radiographs were taken at 12 months and annually, thereafter. An excep-tion to this was in the patients with myelopathy, who under-went magnetic resonance imaging at 12 and 24 months postoperatively.

RESULTS A total of 904 patients were retrospectively reviewed. Among them, 502 were males and 402 were females. Ages were in the range of 4 to 86 years, with a mean of 37.0 years. Among the 904 cases, 601 of them presented with cervical myelopathy. Of note, 65 cases had an associated diagnosis of syringomy-elia, and 52 had Chiari type I deformities. Details regarding the etiologies are summarized in Table 1 , whereas the age dis-tribution is shown in Table 2 .

1. Classifi cation and treatment rendered: Four hundred seventy-two cases were classifi ed as type I, 160 as type II, 268 as type III, and 4 cases as type IV ( Table 3 ). The surgical strategy is summarized in Figure 5 .

2. Clinical outcomes: The postoperative follow-up duration of 904 cases was in the range of 2 to 12 years (mean: 60.5 mo). Among them, 899 cases (99.4%) achieved solid atlantoaxial fusion. On the basis of the postop-erative radiographs, anatomical atlantoaxial reduction was achieved in 892 cases (98.7%, Figure 6 ), whereas 12 cases underwent an incomplete transoral release in the early stages of this series. Among these 12 partial reduc-tions, 2 underwent subsequent odontoidectomies. The preoperative JOA scores of the 601 cases that presented with myelopathy were in the range of 5 to 16, with a mean of 12.1. Neurological improvement was seen in

1/6 body weight

Flexion Extension Skeletal traction

After full curarization

for ten minutes, fully reduced.

Another patient,

not reduced

A B C

D E F

Figure 4. The skeletal traction test. The 30-year male could not achieve full reduction on the dynamic radio-graphs ( A , B ); The cranial traction (Gardner-Wells) was performed under general anesthesia ( C ); One-sixth of the body weight was used ( D ); After 10 minutes, C-arm confi rmed that full reduction was achieved (type II) ( E ); Another patient without full reduction after the traction test was judged as type III ( F ).





85.2% (512/601) of the patients with myelopathy, with a mean postoperative JOA score of 14.8 (range: 5–17).

3. Complication: The complications seen in this series are summarized in Table 4 . The overall complication rate was 9.1% (82/904). Two patients experienced VA in-juries while undergoing a transoral release. One case (a 37-year-old male with AAD and atlas occipitaliza-tion) experienced severe bleeding when we removed the posterior-lateral capsules of the left C1–C2 lateral mass joint, and the estimated blood loss was 2000 mL for the procedure. In this case, the bleeding was controlled by immediate hemostatic tamponade, and an intraoperative VA angiography confi rmed a left VA injury. Subsequently, a left VA embolization using spring coils was then per-formed in the operating room. After this, the bleeding in

the transoral incision was completely controlled, and the remaining portions of the procedure were performed. Af-ter completion of the surgical procedure; however, upon weaning of sedation and attempt at extubation, the pa-tient was found to be unarousable without independent breathing activity. Postoperative CT found a large in-farction in the brain stem, and the patient died 35 hours after the surgery. Another patient, a 50-year-old male, experienced severe bleeding during dissection of the right longus coli muscle, with an estimated blood loss of 700 mL. After hemostatic tamponade, a VA angiography confi rmed a right VA injury, and spring coils were placed in the operating room. This patient awoke and recovered uneventfully after surgery, with no symptoms of cerebral ischemia. Among the complications, 4 cases experienced postoperative respiratory failure. Of note, all 4 cases had symptoms of preoperative respiratory diffi culty, includ-ing dyspnea on exertion and less so at rest, and were un-able to work as result. Among them, 1 case (a 24-year-old female) had preoperative dyspnea, high carbon dioxide levels (arterial blood sample revealed PaCO2 more than 50 mmHg) and required intermittent respirator support. Postoperatively, these 4 patients required prolonged ven-tilation using a respirator. Three of them died postopera-tively from pneumonia and carbon dioxide retention at 3 days, 2 weeks, and 6 weeks, respectively.

DISCUSSION In 1968, Greenberg 3 initially divided AAD into 2 subcatego-ries, reducible and irreducible, and further devised a treatment strategy based on this factor together with the etiology of the dislocation. For irreducible AAD, Greenberg specifi cally stated that the treatment must be aimed at immediate decom-pression and achieving stabilization. Since that time, this has been considered a landmark publication, and the algorithm has been considered to be a “gold standard” for AAD treat-ment in the subsequent literature. 8 – 11

Despite chronic AAD arising from a diverse collection of etiologies, anterior dislocation of the atlas is the most common directional endpoint. Loss of the integrity of the odontoid or the transverse atlantal ligament results in an unstable atlan-toaxial joint. Initially, because of repetitive fl exion of head during daily activities, the atlas translates anteriorly along the relatively horizontal C1–C2 facets. Because of progressive anterior translation, the atlas eventually loses its support from the superior C2 facets, and migrates further anteriorly. The C1–C2 facets gradually remodel with the articular surface becoming increasingly vertically sloped ( Figure 6C ). The mus-cles, ligaments, and capsules of atlantoaxial joint then become shortened and eventually contracted, resulting in an irreduc-ible AAD. We hypothesize that the pathological development of chronic AAD represents the culmination of a consecutive series of gradual events. Despite the diverse etiologies, the developmental process of an irreducible AAD seems to origi-nate from an unstable C1–C2 joint, and in some cases to progress through various stages of reducible dislocations, and


TABLE 2. Distribution of the Age Ages Cases Percentage

4–9 55 6.1

10–19 138 15.3

20–29 89 9.8

30–39 176 19.5

40–49 188 20.8

50–59 178 19.7

60–69 73 8.1

70–79 6 0.7

Over 80 1 0.1

Total 904 100

TABLE 1. Etiologies of AAD in 904 Cases During 12 yr

Diagnosis Cases

Os odontoideum 429

Atlas occipitalization 207

Transverse ligament loosen 84

Atlas occipitalization and block vertebra C2–C3 58

Rheumatoid arthritis 36

Old odontoid fracture, nonunion 25

Ankylosis spondylitis 11

Down syndrome 7

Block vertebra C2–C3 8

Atlas occipitalization and os odontoideum 7

Block vertebra C2–C3 and os odontoideum 5

Unknown 27

Total 904




culminate in an irreducible dislocation. Although not every AAD evolves from reducible to irreducible, we recommend C1–C2 arthrodesis in the setting of even early stages of AAD given the potential for deterioration.

This series demonstrates the extreme importance of the skeletal traction test, under general anesthesia, to differenti-ate between AAD types II and III and further delineate the optimal treatment strategy. Because of the presence of neck pain, muscle tension, and/or positional restrictions, the reduc-ibility of AAD could not be reliably assessed with dynamic radiographs. In this report, 37% of patients with an incom-pletely reducible atlantoaxial joint on dynamic radiographs were able to be completely reduced following a short dura-tion of skeletal traction under general anesthesia. This test is more likely to achieve anatomical reduction as a result of application of substantial traction (up to one-sixth body weight) with total muscle curarization, thus eliminating any muscular resistance, which is unable to be performed safely in the absence of general anesthesia. In a report of 23 pedi-atric patients with congenital AAD, Kumar and Nayak 28 car-ried out conscious cervical traction with the applied weight

ranging from 7% of the body weight to a maximum of 4 kg, with no patient achieving anatomical reduction. Simi-larly, Salunke et al 29 initiated conscious cervical traction with 7% to 8% of body weight, with gradual incremental increases to a maximum of 7 kg. In 57 pediatric patients with IAAD, reduction was achieved in only one patient. We have previ-ously documented our experience of 3 cases undergoing an unsuccessful attempt at conscious traction in other hospitals, where reduction was easily achieved under general anesthesia using our described technique. 30 Consistent with the litera-ture, it has also been our experience that the effect of con-scious skull traction is largely nullifi ed by the counteractive retractions of the neck muscles, with further increase in the traction weight necessary to overcome these forces typically considered unbearable to the conscious patient. This series revealed the traction test to be safe when a maximum of one-sixth of body weight was used, with no recognized traction related complications among 428 cases.

Despite its uses, the traction test does not allow for ana-tomic reduction in all cases. In such cases, a transoral atlanto-axial joint release can provide a conversion from irreducible


TABLE 3. Results of the Classifi cation in 904 Cases

TypeDynamic

RadiographTraction

Test Reconstructive CTROM in

C1–C2 Joint Cases Percentage

I (IN) Reduction NA Normal C1 and C2 lateral facets (horizontal) Increased 472 52.2

II (RD) No reduction Reduction C1 and C2 facets slightly collapsed and sloped Decreased 160 17.7

III (ID) No reduction No reduction C1 and C2 facets severely collapsed and sloped, even in the vertical orientation Little 268 29.6

IV (BD) No reduction NA Bony fusion between C1 and C2 facets No movement 4 0.4

ID indicates instability; RD, reducible dislocation; ID, irreducible dislocation; BD, bony dislocation; CT, computed tomography.

Atlantoaxial dislocation : Loss of C1–C2 stability (904 cases)

Flexion and extension radiograph

Dislocation (432 cases)

I: Instability (472 cases)

Reconstructive CT

Skeletal traction under general anesthesia (OR)

Transoral release

Without C1–C2 bony fusion

II: Reducible dislocation (160)

III: Irreducible dislocation (268)

IV: Bony dislocation (4 cases)

With C1–C2 bony fusion

Successful

Incomplete reduction

Ventral decompression(2 cases)

Posterior fusion

(898 cases)

Not reduced

Dynamic reduction (472 cases)

Ventral decompression and posterior fusion

(2 cases)

(266 cases)2 cases declined surgery

Figure 5. Surgical strategy for 904 cases. CT indicates computed tomography.




cranial traction facilitates reduction when transoral release is required by allowing us to determine which structures are preventing reduction. Because the soft tissues are released, reduction of the atlantoaxial joint is progressively achieved until the tip of the odontoid is clearly exposed after partial removal of C1 anterior arch. Along with the reduction of the atlantoaxial joint, the tip of the odontoid process has shifted away from the dura, minimizing the risk of resecting the liga-ments surrounding the odontoid tip. Recently, several authors have described modifi cations of the approach and technique for anterior atlantoaxial release. Wu et al 31 performed atlan-toaxial joint releases to treat irreducible AAD using an endo-scopic anterior approach, whereas Liu et al 16 and Wang et al 32 used a video-assisted anterior transcervical approach. Col-lectively, these modifi ed techniques used a sloped approach to the atlantoaxial joint and reported diffi culties releasing the lateral mass joints and the apical ligament, particularly in severe AAD cases. In contrast, the transoral approach can directly expose the contracted ligaments and shorten liga-ments and muscles inhibiting reduction, as well as the facets and the odontoid when necessary. In the current series, 268 irreducible AAD cases underwent transoral releases, with 266 of these being successfully converted to reducible AAD.

In the setting of an irreducible AAD, the mechanism of spinal cord compression is the osseous dislocation itself. In the majority of irreducible dislocations, obtaining a reduction without performing a supplemental osteoectomy can provide direct and complete decompression. In contrast, fusion in situ typically necessitates osseous decompression, particularly in cases with high-grade dislocations. 3 In this setting, extensive removal of involved osseous structures, including the C2 body, can be required for complete decompression. 19 , 33 As a result, longer fi xation constructs are required to compensate for the iatrogenic instability created by the decompression of cervical segments. In addition, we have documented previously that AAD can also infl uence the alignment of subaxial cervical spine, and even accelerate the degenerative process, particu-larly in those cases with an associated swan neck deformity. 22

to reducible dislocation. 19 On the basis of the results of this series, we have abandoned transoral odontoidectomies in the setting of irreducible AAD, and propose the idea of neural decompression by odontoid reduction. Continuous


Figure 6. A 46-year female with AAD (type III). Pre-operative radiograph showed severe AAD ( A ); Pre-operative CT, sagittal plane ( B ); Parasagittal CT plane showed the facet orientation becomes sloped severe-ly, near vertically (yellow arrow) ( C ); Preoperative MRI showed severe cervical-medullary compression and syringomyelia ( D ); Four months later after sur-gery, full reduction was maintained ( E ); CT identifi ed anatomical reduction and solid fusion ( F ); After the reduction, the facet joints of C1 and C2 were open forward ( G ); Postoperative MRI showed complete de-compression ( H ). Syringomyelia was reduced mostly. CT indicates computed tomography; AAD, atlanto-axial dislocation; MRI, magnetic resonance imaging.

A B C D

F HGE4 mo

TABLE 4. Complications in 904 Cases Complications Cases

During the surgery

Vertebral artery injury during the transoral approach 2 (1 died)

Vertebral artery injury during the posterior approach 4

Cerebrospinal fl uid leakage 3

Postoperative: early stage

Incision infection 13

Hematosepsis* 5*

Neurological deterioration 6

Hyposmia 5

Swallow dysfunction 10

Deep vein thrombosis 5

Pulmonary embolism 2 (1 died)

Ischemic stroke 2

Respiratory distress 4 (3 died)

Postoperative: late stage

Instrumentation failure or fusion failure 10

Revision 8

Delayed deep infection 5

Undesired addition fusion (Oc–C1 or C2–C3) 4

Subaxial kyphosis 2

Total 82

*All these 5 cases underwent transoral approach, experiencing hematosepsis. Among them 1 had delayed meningitis at 1 mo after surgery.




➢ Key Points

A novel surgical classifi cation and treatment strategy for AAD were described in this report.

A total of 904 patients treated by the surgical classifi cation and treatment strategy were retrospectively studied.

Anatomic atlantoaxial reduction was achieved in 892 cases (98.7%).

Neurological improvement was seen in 84.1% of the patients with myelopathy. The overall complication rate was 9.1%.

Recreation of an anatomic alignment in the atlantoaxial joint in this situation may allow for restoration of global balance in the cervical spine.

With regards to major complications experienced in this series, 2 (0.7%, 2/268) experienced VAI during the transoral approach. We share the opinion with Peng et al 34 that imme-diate angiography should be performed and endovascular techniques, such as coil embolization, be used for controlling the potentially fatal bleeding. Considering that poor preop-erative pulmonary function was associated with fatality in 3 of the 4 instances, patients with preoperative respiratory diffi culty should be carefully assessed prior to surgery.

One major criticism of this report is that several aspects of the treatment strategy were adjusted during the 12-year study period. Examples of alterations in the treatment dur-ing the course of this series include the use of tracheoto-mies being abandoned after 2008, and changes in the trac-tion weight applied. In the earlier stage of this series (before October 2003), we used one-tenth of the patients’ weight, which was subsequently increased to one-sixth of the body weight applied during the traction test. Another potential limitation is that the case series presented is derived from the experience of a single institution and team of surgeons. In addition, there is no patient-based outcome data provided in the report because it is a retrospective review. Lastly, defi ni-tive conclusions on the development and progression of AAD cannot be made based on this series, and future research is required to elucidate our understanding of this process.

CONCLUSION Despite the limitations, the authors anticipate that the data and experience in this series will enable us to develop a novel and comprehensive surgical classifi cation and treat-ment strategy for this complex problem, and can help guide physicians in their treatment and allow for improved com-parative literature.

References 1. Kumar R , Kalra SK , Mahapatra AK. A clinical scoring system for

neurological assessment of high cervical myelopathy: measurements in pediatric patients with congenital atlantoaxial dislocations . Neu-rosurgery 2007 ; 61 : 987 – 93 ; discussion 93–4.

2. Reddy KR , Rao GS , Devi BI , et al. Pulmonary function after surgery for congenital atlantoaxial dislocation: a comparison with surgery for compressive cervical myelopathy and craniotomy . J Neurosurg Anesthesiol 2009 ; 21 : 196 – 201 .

3. Greenberg AD. Atlanto-axial dislocations . Brain 1968 ; 91 : 655 – 84 . 4. Gholve PA , Hosalkar HS , Ricchetti ET , et al. Occipitalization of

the atlas in children. Morphologic classifi cation, associations, and clinical relevance . J Bone Joint Surg Am 2007 ; 89 : 571 – 8 .

5. Cremers MJ , Bol E , de Roos F , et al. Risk of sports activities in chil-dren with Down’s Syndrome and atlantoaxial instability . Lancet 1993 ; 342 : 511 – 4 .

6. Lourie H , Stewart WA. Spontaneous atlantoaxial dislocation. A complication of rheumatoid disease . N Engl J Med 1961 ; 265 : 677 – 81 .

7. Menezes AH. Honored guest presentation: conception to implica-tion: craniocervical junction database and treatment algorithm . Clin Neurosurg 2005 ; 52 : 154 – 62 .

8. Behari S , Bhargava V , Nayak S , et al. Congenital reducible atlan-toaxial dislocation: classifi cation and surgical considerations . Acta Neurochir (Wien) 2002 ; 144 : 1165 – 77 .

9. Subin B , Liu JF , Marshall GJ , et al. Transoral anterior decompres-sion and fusion of chronic irreducible atlantoaxial dislocation with spinal cord compression . Spine (Phila Pa 1976) 1995 ; 20 : 1233 – 40 .

10. Menezes AH. Craniovertebral junction database analysis: incidence, classifi cation, presentation, and treatment algorithms . Childs Nerv Syst 2008 ; 24 : 1101 – 8 .

11. Dai LY , Yuan W , Ni B , et al. Surgical treatment of nonunited fractures of the odontoid process, with special reference to occipitocervical fusion for unreducible atlantoaxial subluxation or instability . Eur Spine J 2000 ; 9 : 118 – 22 .

12. Nordt JC , Stauffer ES. Sequelae of atlantoaxial stabilization in two patients with Down's syndrome . Spine (Phila Pa 1976) 1981 ; 6 : 437 – 40 .

13. Hedequist D , Bekelis K , Emans J , et al. Single stage reduction and stabilization of basilar invagination after failed prior fusion sur-gery in children with Down's syndrome . Spine (Phila Pa 1976) 2011 ; 35 : E128 – 33 .

14. Goel A , Desai KI , Muzumdar DP. Atlantoaxial fi xation using plate and screw method: a report of 160 treated patients . Neurosurgery 2002 ; 51 : 1351 – 6 ; discussion 6–7.

15. Wang C , Yan M , Zhou H , et al. Atlantoaxial transarticular screw fi xation with morselized autograft and without additional internal fi xation: technical description and report of 57 cases . Spine (Phila Pa 1976) 2007 ; 32 : 643 – 6 .

16. Liu T , Li F , Xiong W , et al. Video-assisted anterior transcervical approach for the reduction of irreducible atlantoaxial dislocation . Spine (Phila Pa 1976) 2011 ; 35 : 1495 – 501 .

17. Yin Q , Ai F , Zhang K , et al. Irreducible anterior atlantoaxial dislo-cation: one-stage treatment with a transoral atlantoaxial reduction plate fi xation and fusion. Report of 5 cases and review of the litera-ture . Spine (Phila Pa 1976) 2005 ; 30 : E375 – 81 .

18. Harms J , Melcher RP. Posterior C1-C2 fusion with polyaxial screw and rod fi xation . Spine (Phila Pa 1976) 2001 ; 26 : 2467 – 71 .

19. Wang C , Yan M , Zhou HT , et al. Open reduction of irreduc-ible atlantoaxial dislocation by transoral anterior atlantoaxial release and posterior internal fi xation . Spine (Phila Pa 1976) 2006 ; 31 : E306 – 13 .

20. Crockard HA , Calder I , Ransford AO. One-stage transoral decompression and posterior fi xation in rheumatoid atlanto-axial subluxation . J Bone Joint Surg Br 1990 ; 72 : 682 – 5 .

21. Wadia NH. Myelopathy complicating congenital atlanto-axial dislocation. (A study of 28 cases) . Brain 1967 ; 90 : 449 – 72 .

22. Passias PG , Wang S , Kozanek M , et al. Relationship between the alignment of the occipitoaxial and subaxial cervical spine in patients with congenital atlantoxial dislocations . J Spinal Disord Tech 2013 ; 26 : 15 – 21 .

Acknowledgment The authors thank Peter G. Passias, MD, for his help in English language editing. The article was presented at Cervi-cal Spine Research Society 41st Meeting in December 2012 as “3rd place of the Best Clinical Research Award.”





23. Abumi K , Takada T , Shono Y , et al. Posterior occipitocervical reconstruction using cervical pedicle screws and plate-rod systems . Spine (Phila Pa 1976) 1999 ; 24 : 1425 – 34 .

24. Wright NM. Posterior C2 fi xation using bilateral, crossing C2 lam-inar screws: case series and technical note . J Spinal Disord Tech 2004 ; 17 : 158 – 62 .

25. Finn MA , Fassett DR , McCall TD , et al. The cervical end of an occipitocervical fusion: a biomechanical evaluation of 3 con-structs. Laboratory investigation . J Neurosurg Spine 2008 ; 9 : 296 – 300 .

26. Menezes AH. Surgical approaches: postoperative care and compli-cations “transoral-transpalatopharyngeal approach to the cranio-cervical junction ”. Childs Nerv Syst 2008 ; 24 : 1187 – 93 .

27. Wang S , Wang C , Wood KB , et al. Radiographic evaluation of the technique for C1 lateral mass and C2 pedicle screw fi xation in three hundred nineteen cases . Spine (Phila Pa 1976) 2011 ; 36 : 3 – 8 .

28. Kumar R , Nayak SR. Management of pediatric congenital atlanto-axial dislocation: a report of 23 cases from northern India . Pediatr Neurosurg 2002 ; 36 : 197 – 208 .

29. Salunke P , Behari S , Kirankumar MV , et al. Pediatric congenital atlantoaxial dislocation: differences between the irreducible and reducible varieties . J Neurosurg 2006 ; 104 : 115 – 22 .

30. Wang C , Wang S. , Visocchi M , et al. Pre-operative irreducible C1-C2 dislocations: intra-operative reduction and posterior fi xation. The “always posterior strategy” . Acta Neurochir 151 : 551 – 560 ; discussion. Acta Neurochir (Wien) 2009;151:1329–31; author reply 33–6.

31. Wu YS , Chi YL , Wang XY , et al. Microendoscopic anterior approach for irreducible atlantoaxial dislocation: surgical techniques and pre-liminary results . J Spinal Disord Tech 2010 ; 23 : 113 – 20 .

32. Wang B , Lu G , Deng Y , et al. Anterior endoscopically assisted tran-scervical reconstruction of the upper cervical spine . Eur Spine J 2011 ; 20 : 1526 – 32 .

33. Youssef AS , Guiot B , Black K , et al. Modifi cations of the transoral approach to the craniovertebral junction: anatomic study and clini-cal correlations . Neurosurgery 2008 ; 62 : 145 – 54 ; discussion 54–5.

34. Peng CW , Chou BT , Bendo JA , et al. Vertebral artery injury in cer-vical spine surgery: anatomical considerations, management, and preventive measures . Spine J 2009 ; 9 : 70 – 6 .



Novel Surgical Classiﬁ cation and Treatment Strategy for ...

Documents

Transcript of Novel Surgical Classiﬁ cation and Treatment Strategy for ...