Posterior Facet Distraction and Fusion Technique: Driving a Paradigm Shift in the Surgical Algorithm for Atlantoaxial Dislocation

Abstract

BACKGROUND AND OBJECTIVES:

Atlantoaxial dislocation (AAD) remains one of the most technically challenging disorders of the craniovertebral junction. Traditional combined anterior-posterior approaches, although effective, carry substantial morbidity. The posterior facet distraction and fusion (PFDF) technique achieves reduction through sequential facet release and distraction, offering a safer posterior-only alternative. This single-center retrospective study aimed to evaluate the reduction efficacy, clinical outcomes, and safety of a PFDF-centered surgical strategy, and to develop and validate a PFDF-based treatment algorithm for AAD.

METHODS:

From 2017 to 2024, a total of 434 patients with AAD were surgically treated following a PFDF-based algorithm. Based on preoperative imaging, patients were classified as distractable or nondistractable according to the atlantoaxial joint condition. PFDF was performed in distractable patients without vertebral artery contraindications. Combined posterior-anterior approaches were used for patients with contraindicated vertebral artery or inadequate reduction after PFDF. Patients with rigid fusion underwent transoral odontoidectomy for ventral decompression. Clinical and radiological outcomes, complications, and fusion status were analyzed.

RESULTS:

According to the algorithm, a total of 399 patients (91.9%) achieved satisfactory reduction and decompression using PFDF alone, 12 (2.8%) required posterior fixation with anterior decompression, and 23 (5.3%) underwent transoral odontoidectomy. Significant postoperative improvement was observed in neurological and radiological outcomes (P < .001). Anatomic reduction was achieved in 90.9% of PFDF patients, with complete bone fusion in all patients within 12 months. The overall complication rate was 5.1%, and no perioperative mortality occurred.

CONCLUSION:

The PFDF technique provides a safe, effective, and standardized posterior approach for the management of AAD. By achieving reduction through sequential joint release and distraction, PFDF achieves high reduction and fusion rates while minimizing surgical morbidity. This method simplifies the treatment algorithm for AAD, allowing most patients to avoid anterior surgery and its associated complications.

ABBREVIATIONS:

AAD

atlantoaxial dislocation

ADI

atlantodental interval

JOA

Japanese Orthopaedic Association

PF

posterior fixation

PFDF

Posterior facet distraction and fusion

SF-12

12-Item Short Form Health Survey

SSW

subarachnoid space width

SW

syrinx width

TO

transoral odontoidectomy

VA

vertebral artery.

The surgical management of atlantoaxial dislocation (AAD) is guided by the principles of reduction, neural decompression, and rigid fixation with fusion to restore craniovertebral stability.^1-3 Transoral approach, including odontoidectomy or ligament release, combined with posterior fixation (PF), has long been regarded as the mainstay of treatment for complex cases.^2,4-6 Although technically demanding, this combined approach has been widely adopted because of its reliable reduction efficacy.^7,8 Nevertheless, it is associated with a relatively high incidence of complications, including intraoperative spinal cord injury during position changes, pharyngeal edema, and infection.^9-12

In 2004, Goel et al¹³ introduced the concept of posterior intra-articular manipulation, in which reduction is achieved by distraction through the atlantoaxial joints. This technique proposed a paradigm shift in management of AAD and basilar invagination. Initially, limited surgical tools and implant designs hindered ideal reduction.

Over the past 2 decades, we have been committed to technical improvement for AAD.^14-16 Based on the philosophy of direct posterior reduction, we proposed a unique and standardized technique termed the posterior facet distraction and fusion (PFDF) technique and designed a dedicated set of instruments for clinical application.^17-19 This technique achieves reduction of AAD through posterior joint sequential release and distraction, shifting irreducible to reducible, thereby avoiding transoral surgery and its complications.

This technique has reshaped the conventional management paradigm for irreducible cases.⁸ As clinical experience accumulated, we formulated a comprehensive treatment algorithm for AAD based on PFDF technique. This study summarizes technical essentials, clinical, and radiological outcomes in 434 patients treated following the PFDF-based algorithm, aiming to provide a practical reference for the management of complex atlantoaxial disorders.

METHODS

Patient Selection

This was a retrospective cohort study approved by the Institutional Review Board, and written informed consent was obtained from all patients. Between January 2017 and January 2024, a total of 434 consecutive patients with AAD were surgically treated in our institution and included. AAD was defined using the criteria of an increased atlantodental interval (ADI) >3 mm in adult patients and 5 mm in children. For os odontoideum, the modified ADI was calculated as the distance from the posterior border of the C1 anterior tubercle to the extended anterior border of the C2 vertebral body.²⁰ Patients with severe osteoporosis, AAD secondary to tumor, infection, tuberculosis, or acute fracture, or with comorbidities precluding surgery were excluded. All patients underwent detailed neurological evaluation and symptom documentation. Neurological function was assessed using the Japanese Orthopaedic Association (JOA) score, and general health status was evaluated by the 12-item Short Form Health Survey (SF-12).

Radiological parameters were measured on computed tomography (CT) and MRI images and included the following:

• ADI, atlantodental interval;

• DCL, distance from the odontoid to Chamberlain line;

• CCA, clivus-canal angle;

• CMA, cervicomedullary angle;

• CVJTA, craniovertebral junction triangle area;²¹

• SSW, subarachnoid space width anterior to the odontoid process;²² and

• SW, syrinx width (if present, in the maximum percentage of the spinal cord).²³

Classification and Treatment Algorithm

In our treatment algorithm (Figure 1), patients with AAD are categorized into distractable and nondistractable, based on the feasibility of posterior atlantoaxial facet joint distraction.

FIGURE 1.

Flow chart for the treatment algorithm of AAD based on the PFDF technique. CT, computed tomography; PF, posterior fixation; PFDF, posterior facet distraction and fusion; TO, transoral odontoidectomy; VA, vertebral artery.

Nondistractable Dislocation

Patients exhibiting solid bony fusion between C1 and C2 vertebra, whether at the atlanto-odontoid joint, lateral facets, or posterior elements, are classified as nondistractable dislocations, also termed PFDF-irreducible dislocations. For nondistractable dislocations, we do not attempt posterior reduction through osteotomy at the fusion site, because this procedure is highly risky and complex. Instead, we perform transoral odontoidectomy (TO) from anterior approach, which is relatively straightforward and safe.

Distractable Dislocation

Patients without solid bony fusion at the atlantoaxial joint are considered as distractable dislocations, also termed PFDF-reducible dislocations, which constitute most cases. Even in patients with joint sclerosis and fibrous connection observed on CT, posterior distraction remains feasible as long as the joint surface remains visible.

With Contraindicated VA

Vertebral artery (VA) variations are commonly seen in patients with AAD. According to previous studies, unilateral VA coursing posterior to the facet joint, or VA coursing into the joint are considered contraindicated.²⁴ In such cases, attempting joint manipulation may result in VA occlusion, leading to potentially fatal posterior circulation ischemia. For these cases, we use a sole posterior fixation and fusion without intra-articular manipulation, and combined with anterior odontoidectomy for decompression.

Without Contraindicated VA

For distractable dislocation with safe VA, the PFDF technique was used. The detailed surgical steps with demonstration video have been mentioned in previous publications and will be summarized above.^17,18 Postoperatively, CT and MRI scans are conducted to assess the extent of reduction. If CT imaging indicates incomplete reduction and MRI reveals spinal cord compression, supplementary TO is performed.

PFDF Technique

The philosophy of PFDF technique is to achieve reduction through posterior joint release and sequential distraction, which relaxes the anterior tension band and remodels the articular surfaces, rather than by direct joint distraction or forced reduction through cage insertion. The intraarticular cages in this technique serve primarily to provide support and facilitate fusion (Figure 2).

FIGURE 2.

A, A 48-year-old man presented with atlantoaxial dislocation, basilar invagination, atlas assimilation, and congenital C2-C3 fusion. B and C, Parasagittal computed tomography scans showed an oblique facet on the right side and a vertical facet with a pseudo joint on the left side. D, Preoperative MRI revealed ventral compression of the cervicomedullary junction, Chiari malformation, and syringomyelia. E, Preoperative computed tomography angiography identified an abnormal right vertebral artery coursing posterior to the facet joint. F, The patient underwent the posterior facet distraction and fusion technique and achieved anatomic reduction. G and H, Parasagittal computed tomography scans demonstrated intra-articular distraction and cage placement. The tip of the C1 inferior articular process was remodeled and horizontalized during joint release to facilitate cage placement. The vertical facet was thoroughly released to achieve optimal reduction. I, Postoperative MRI showed decompression of the cervicomedullary junction and regression of the syringomyelia. J, Postoperative x-ray demonstrated ideal positioning of the implants.

Position and Exposure

The patient is placed prone and continuous skull traction equivalent to 1/6 of the body weight is applied to assist in joint distraction. A midline subperiosteally dissection is performed from the external occipital protuberance to the inferior border of the C2. For revision cases with existing instrumentation, fixation devices should first be removed.¹⁸

Facet Joint Release and Distraction

Subperiosteally, the C2 nerve root is gently elevated, and the posterior rim of the atlantoaxial joint surface is identified. A 2-mm osteotome is inserted into the joint space, with rotation to remove the articular cartilage. Specially designed distractors of varying widths (4-9 mm, Wego Ortho) are sequentially inserted into the joint space from narrow to wide. The joint distractor is initially advanced with firm pressure to penetrate the anterior joint capsule to release the anterior longitudinal ligament and longus colli muscle. Two maneuvers—levering and rotation—is used for joint release. The distraction process should be gradual and excessive force may cause articular surface fracture and collapse, which would hinder further procedures.

Distraction is first performed on one side, and a trial cage is placed in the joint to maintain distraction. The contralateral joint is distracted using the same technique. This maneuver can loosen the anterior tension band, allowing longitudinal separation of the joint space and caudal movement of the axis.

For patients with VAs coursing dorsally to the C1-C2 facet, the vessel is gently freed from the facet to improve mobility, allowing facet manipulation without injury or occlusion. The VA was carefully protected and mobilized cranially using a Penfield dissector, followed by facet joint release and distraction with a specially designed distractor. During distraction, the rounded bar of the custom dissector minimized the risk of iatrogenic vascular injury on vessel contact; upward displacement of the bar was strictly avoided to prevent impingement of the cranial segment of the VA.

Fixation and Reduction

We used a specially designed occipital plate with long rod connection ends to perform cantilever reduction. For patients with atlas assimilation, an occipital plate was fixed, followed by insertion of C2 pedicle screws. For patients without atlas assimilation, C1 lateral mass screws were inserted to preserve the mobility of atlanto-occipital joint. Precontoured titanium rods were then secured to the pedicle screws, allowing the cranial end to elevate. By tightening the cranial screws, the C2 vertebra was pushed anteriorly and inferiorly through the cantilever mechanism.

Perioperative Evaluation and Follow-Up

Intraoperative data, such as operation time, estimated blood loss, and surgical approach, were recorded for all patients. Postoperative evaluation involved both clinical and radiological assessments and was performed immediately after surgery and at 3, 6, and 12 months postoperatively and then annually. Bony fusion was confirmed by CT when continuous trabecular bone bridging across the cage was observed, and no hardware loosening or motion was detected. Any intraoperative and postoperative complications were documented. Patients requiring reoperation were specifically analyzed.

Statistical Analysis

Statistical analysis was performed using the SPSS 26.0 software (IBM). The paired-samples t-test was used to analyze any statistical difference between the preoperative and postoperative measurements. The level of significance was set at P < .05. Data are presented as mean ± SD unless stated otherwise.

RESULTS

Patient Characteristics

Patient characteristics are summarized in Table 1. A total of 434 patients were included in this study, comprising 157 male individuals (36.2%) and 277 female individuals (63.8%). Basilar invagination was present in 354 patients (81.6%), Atlas assimilation was presented in 313 (72.1%) patients. Other associated conditions included syringomyelia in 125 patients (28.8%), congenital C2-C3 fusion in 172 (39.6%) patients, craniocervical coronal imbalance in 98 (22.6%) patients, os odontoideum in 44 (10.1%) patients, and atlantoaxial rotatory fixation in 7 (1.6%) patients. An abnormal VA course was identified in 228 patients (52.5%), and 78 patients (18.0%) were revision cases.

TABLE 1.

Patient Characteristics

	Value
Sex
Male	157 (36.2)
Female	277 (63.8)
Basilar invagination	354 (81.6)
Atlas assimilation	313 (72.1)
Syringomyelia	125 (28.8)
Congenital C2-C3 fusion	172 (39.6)
Craniocervical coronal imbalance	98 (22.6)
Os odontoideum	44 (10.1)
Atlantoaxial rotatory fixation	7 (1.6)
Abnormal vertebral artery course	228 (52.5)
Revision cases	78 (18.0)

Values are presented as the number of patients (%).

Preoperative Symptoms, Clinical, and Radiological Findings

Preoperative symptoms are summarized in Table 2. The most common preoperative symptoms were limb numbness (45.2%) and limb weakness (41.0%), followed by head or neck pain (24.9%) and gait instability (18.2%). Other symptoms included dizziness (11.8%), trunk numbness (7.6%), trunk or limb pain (7.1%), dysphagia (4.8%), choking on drinking (3.5%), and muscular atrophy (2.3%).

TABLE 2.

Preoperative Symptoms, Clinical, and Radiological Assessments

	Value
Symptoms
Limb numbness	196 (45.2)
Limb weakness	178 (41.0)
Trunk numbness	33 (7.6)
Head or neck pain	108 (24.9)
Trunk or limb pain	31 (7.1)
Dizziness	51 (11.8)
Gait instability	79 (18.2)
Dysphagia	21 (4.8)
Choking on drinking	15 (3.5)
Muscular atrophy	10 (2.3)
Clinical scores
JOA	13.4 ± 2.8
SF-12	82.6 ± 11.6
Radiological assessments
ADI, mm	3.5 ± 2.0
DCL, mm	9.6 ± 4.7
CCA, °	132.1 ± 14.7
CMA, °	129.9 ± 14.0
CVJTA, cm2	2.3 ± 0.7
SSW, mm	1.1 ± 0.6
SW, %	67.9 ± 25.1

ADI, atlantodental interval; CCA, clivus-canal angle; CMA, cervicomedullary angle; CVJTA, craniovertebral junction triangle area; DCL, distance from the odontoid to Chamberlain line; JOA, Japanese Orthopaedic Association; SF-12, 12-Item Short Form Health Survey; SSW, subarachnoid space width; SW, syrinx width.

Values are presented as the number of patients (%) or mean ± SD.

The mean preoperative JOA score was 13.4 ± 2.8, and the mean SF-12 score was 82.6 ± 11.6 (Table 2). Radiographically, the average ADI was 3.5 ± 2.0 mm, distance from the odontoid to Chamberlain line was 9.6 ± 4.7 mm, the clivus-canal angle measured 132.1° ± 14.7°, and the cervicomedullary angle measured 129.9° ± 14.0°. The craniovertebral junction triangle area was 2.3 ± 0.7 cm², the mean SSW was 1.1 ± 0.6 mm, and. The average maximum SW was 67.9 ± 25.1% (Table 2).

Surgical Data

Surgical data are shown in Table 3. According to the proposed treatment algorithm, a total of 399 patients (91.9%) underwent posterior-only reduction using the PFDF technique. Twelve patients (2.8%) required combined posterior fixation with anterior odontoidectomy, including 8 patients who underwent supplementary anterior decompression because of insufficient ventral decompression after PFDF and 4 patients with contraindicated VA. Twenty-three patients (5.3%) were considered nondistractable and underwent TO alone for ventral decompression. The mean operative time was 170.2 ± 62.2 minutes, with a mean intraoperative blood loss of 116.5 ± 162.1 mL and an average hospital stay of 10.4 ± 4.2 days.

TABLE 3.

Surgical Data

	Value
Surgical approach
PFDF	399 (91.9)
PFDF + TO	8 (1.8)
PF + TO	4 (1.0)
TO	23 (5.3)
Surgery details
Operation time, min	170.2 ± 62.2
Blood loss, mL	116.5 ± 162.1
Length of stay, d	10.4 ± 4.2

PF, posterior fixation; PFDF, posterior facet distraction and fusion; TO, transoral odontoidectomy.

Values are presented as the number of patients (%) or mean ± SD.

Postoperative Clinical and Radiological Findings

Both clinical and radiological outcomes improved significantly after surgery, and the patients who underwent TO were not included in radiological analysis (Table 4).

TABLE 4.

Postoperative Clinical and Radiological Data

	Preoperative	Postoperative
Clinical scores
JOA	13.4 ± 2.8	15.2 ± 2.0a
SF-12	82.6 ± 11.6	97.5 ± 11.8a
Radiological assessments
ADI, mm	3.5 ± 2.0	2.0 ± 1.2a
DCL, mm	9.6 ± 4.7	4.5 ± 5.4a
CCA, °	132.1 ± 14.7	140.4 ± 13.8a
CMA, °	129.9 ± 14.0	149.1 ± 12.6a
CVJTA, cm2	2.3 ± 0.7	1.8 ± 0.5a
SWW, mm	1.1 ± 0.6	2.0 ± 1.0a
SW, %	67.9 ± 25.1	47.9 ± 23.9a

ADI, atlantodental interval; CCA, clivus-canal angle; CMA, cervicomedullary angle; CVJTA, craniovertebral junction triangle area; DCL, distance from the odontoid to Chamberlain line; JOA, Japanese Orthopaedic Association; SF-12, 12-Item Short Form Health Survey; SW, syrinx width.

^aP < .001 compared with preoperative parameters.

The JOA score increased from 13.4 ± 2.8 preoperatively to 15.2 ± 2.0 postoperatively (P < .001), corresponding to a 50.6% improvement rate. The SF-12 score improved from 82.6 ± 11.6 to 97.5 ± 11.8 (P < .001).

Radiographically, ADI and distance from the odontoid to Chamberlain line significantly decreased from 3.5 ± 2.0 mm to 2.0 ± 1.2 mm and from 9.6 ± 4.7 mm to 4.5 ± 5.4 mm, respectively (P < .001). The clivus-canal angle and cervicomedullary angle increased from 132.1° ± 14.7° to 140.4° ± 13.8° and from 129.9° ± 14.0° to 149.1° ± 12.6°, respectively (P < .001). The SSW expanded from 1.1 ± 0.7 mm to 2.0 ± 1.0 mm (P < .001), whereas the SW decreased from 67.9% ± 25.1% to 47.9% ± 23.9% (P < .001).

Bone fusion was achieved in 35.3% of the patients at 3 months, 73.7% at 6 months, and 100% at 12 months and beyond (Figure 3).

FIGURE 3.

A, A 13-year-old boy presented with atlantoaxial rotatory fixation and dislocation with os odontoideum. B, The patient initially underwent posterior atlantoaxial fixation by using Harm's technique, and postoperative computed tomography at 1 week showed satisfactory reduction. C, However, follow-up at 6 months revealed loss of reduction and recurrent rotatory dislocation. D, A revision surgery using the posterior facet distraction and fusion technique was then performed, with an intra-articular cage implanted only on the right side to restore coronal balance. E-G, At 1-year follow-up after revision, satisfactory reduction was maintained, with solid bony fusion achieved at both the atlanto-odontoid and lateral facet joints. H, The rotatory dislocation was completely corrected, fusion was firm, and the instrumentation was removed at the patient's request.

Complications

A total of 22 patients experienced complications, yielding an overall complication rate of 5.1% (Table 5). No perioperative mortality occurred in this series.

TABLE 5.

Summary of Complications in 434 Cases

Complications	No.
Intraoperative and postoperative
Vertebral artery injury	4
Cerebrospinal fluid leakage	1
Respiratory distress	1
Dysphagia	1
Incision infection	1
Instrumentation failure	2
Pulmonary infection	5
Limb weakness	2
Severe pain	3
Revision surgery after PFDF
Refixation	2
Posterior fossa decompression	2

PFDF, posterior facet distraction and fusion.

VA injury occurred in 4 patients. One VA laceration occurred during facet joint exposure and release, and was managed with direct microsurgical suturing. In the remaining 3 patients, VA occlusion was caused by intraoperative VA traction or facet manipulation, and was confirmed by postoperative computed tomography angiography, which was managed with antiplatelet therapy. Owing to adequate collateral circulation, none of these patients developed neurological sequelae.

Other complications included cerebrospinal fluid leakage in 1 patient, respiratory distress in 1, dysphagia in 1, incision infection in 1, instrumentation failure in 2, pulmonary infection in 5, limb weakness in 2, and severe pain in 3.

A total of 4 patients underwent revision surgery after PFDF: 2 for fixation failure and 2 for residual foramen magnum stenosis requiring posterior fossa decompression.

DISCUSSION

The surgical goal for AAD is to achieve anatomic reduction, rigid fixation, and solid fusion, thereby decompressing the neural elements and restoring long-term stability.^2,25 However, irreducible AAD remains one of the most challenging conditions at the craniovertebral junction.²⁶ Traditionally, TO followed by posterior fixation has been the standard approach for irreducible AAD.² Despite its effectiveness, its morbidity is considerable.²⁷

Recognition of the importance of facet release refined surgical concepts. Early efforts focused on achieving effective facet release through anterior approaches. Wang et al⁵ modified the TO into a less invasive transoral release procedure followed by posterior reduction and fixation. This approach allows adequate ventral and dorsal decompression and significantly improves reduction rates, with approximately 30% of patients requiring anterior release according to their treatment algorithm. Yin et al²⁸ subsequently introduced the transoral atlantoaxial reduction plate system, which enables anterior release, reduction, and fixation through a single anterior approach. Although effective, its technical complexity limits broad adoption.

The technical complexity and inherent risks of anterior procedures have prompted the pursuit of safer posterior reduction alternatives. In 2004, Goel et al¹³ pioneered the concept of intra-articular distraction and fixation for AAD with basilar invagination by inserting spacers into the C1-C2 joints. Although limited by spacer's height and screw proximity, the reduction capability was suboptimal. Despite these constraints, the work of Goel laid the foundation for posterior intraarticular distraction as a viable alternative for AAD reduction. Building on this concept, Chandra et al²⁹ proposed the Distraction-Compression-Extension-Reduction technique, which uses bullet-shaped biconvex spacers as fulcrums to facilitate reduction through posterior compression. Although Distraction-Compressio-Extension-Reduction improved reducibility, in patients with vertical facets and pseudo joints, this technique does not achieve thorough intra-articular release.³⁰ Distraction is limited between the occiput and C2 pseudo joint, hindering optimal reduction, and weak lamina fixation that may require extending fusion levels.

Over the past 2 decades, we have focused on refining posterior surgical techniques for AAD. Early work by Jian et al¹⁴ introduced posterior occipito-cervical distraction and reduction technique, treating irreducible AAD. However, distraction without intra-articular support risked resettling. To address these issues, we introduced the novel PFDF technique, achieving reduction through sequential posterior joint release and distraction, rather than forced spacer insertion.¹⁷

Sequential release was performed using specially designed instruments. The distractor was first used to penetrate the anterior joint capsule and through rotation and levering, the anterior tension band was released. Trial spacers were then inserted for distraction, alternating bilaterally with gradually increasing heights until optimal reduction was achieved. During this process, the C1 inferior facet was remodeled and horizontalized to facilitate cage placement. Intra-articular cages were then implanted, followed by a cantilever maneuver to angulate C2 and close the ADI. All procedures were verified intraoperatively using 3-dimensional imaging for reduction and screw trajectory confirmation.

Using the PFDF technique, we have treated a large series of patients with satisfactory outcomes—an anatomic reduction rate of 90.9% (370/407) and a complication rate of 5.1%. Even in patients with vertical facets and pseudo joints, adequate reduction was achievable by using thorough joint release. We have subsequently published a series of studies detailing the technical nuances of PFDF, including VA protection,²⁴ C2 screw insertion strategies,³¹ revision techniques,¹⁹ and management of bony fusion cases³² and pediatric cases.³³ We also proposed individualized quantitative reduction using cages of varying heights and angles.^34,35 Finite element analysis confirmed superior biomechanics and stress distribution, promoting rapid fusion consistent with Wolff's law.³⁶ Solid fusion was achieved in all patients within 1 year. The PFDF technique has since been widely adopted by other centers, consistently yielding excellent reduction and fusion outcomes.^37-40 Xu et al⁸ recently reported that this technique rendered nearly 40% of previously irreducible dislocations reducible, effectively eliminating the need for transoral release and updating their treatment paradigm.

Based on our accumulated experience, we have established a treatment algorithm centered on PFDF (Figure 1). Reducibility is determined preoperatively using CT imaging rather than intraoperative traction. Patients without rigid bony fusion or a contraindicated VA are candidates for PFDF. Theoretically, once the interarticular cage is successfully inserted, vertical dislocation will inevitably improve and ventral compression will be alleviated. If, despite repeated attempts, intraoperative 3-dimensional imaging still shows incomplete reduction, a posterior bony decompression is performed. Postoperative MRI was routinely performed to evaluate the extent of decompression. In patients with partial reduction and no residual spinal cord compression, further decompression was not required. Secondary TO was performed only when postoperative MRI demonstrated persistent ventral compression (Figure 4). In our algorithm, the PFDF technique achieved satisfactory decompression in 91.9% patients (399/434), whereas only 1.8% patients (8) required subsequent transoral surgery because of residual ventral compression. Thus, the PFDF-based surgical algorithm markedly reduced the need for transoral procedures and minimized associated risks. In patients with fibrous or partial bony fusion but visible joint surfaces, reduction through PFDF remains feasible;³² however, in cases of firm bony fusion, we no longer attempt dissecting fused atlantoaxial lateral joints, and anterior decompression is indicated.

FIGURE 4.

A, A 51-year-old man presented with atlantoaxial dislocation, basilar invagination, and atlas assimilation. B, Preoperative MRI revealed ventral compression of the cervicomedullary junction. C, Because this is a distractable case without contraindicated vertebral artery, the patient underwent the posterior facet distraction and fusion technique. D and E, The patient failed to achieve anatomic reduction because of parodontoid osteophytes obstruction and the atlantodental interval remained 3.8 mm postoperatively. F, MRI demonstrated residual upper cervical cord compression with unsatisfactory clinical improvement. G, According to our treatment algorithm, a secondary transoral odontoidectomy was performed. Follow-up computed tomography confirmed adequate resection. H, Postoperative MRI showed sufficient decompression.

Limitations

This study has several limitations. It is a retrospective cohort study without a control group, which inherently introduces selection bias. As a retrospective series, the absence of a control group limits the ability to compare outcomes directly with alternative surgical approaches. However, because of the rarity of this procedure, the large cohort of 434 patients offers valuable insights and serves as a useful reference for surgeons who may have limited experience with the PFDF technique. Although all surgeries were performed by one senior surgeon, the large sample size enhances the clinical relevance and applicability of the results. A prospective, multicenter trial with a control group is underway to further validate the safety, efficacy, and generalizability of the PFDF technique across diverse patient populations and surgical settings.

CONCLUSION

The PFDF technique provides a safe, effective, and standardized posterior solution for AAD. Through sequential joint release and distraction, PFDF achieves high reduction and fusion rates while minimizing surgical morbidity. This method simplifies the treatment algorithm for AAD, sparing most patients from anterior surgery and related complications when key technical aspects are properly executed.

Funding

This research was supported by National Key R&D Program of China-Young Scientist Project, Grant No. 2023YFC2509700 (W.D.), Beijing Hospitals Authority Clinical Medicine Development of Special Funding Support, code: ZLRK202516 (Z.C.), and Beijing Hospitals Authority “Dengfeng” Leading Talent Program, code: 20250043 (Z.C.).

Disclosures

The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.

COMMENTS

Craniocervical pathologies are relatively uncommon yet frequently associated with substantial morbidity. Historically, many of these conditions have been managed with transoral decompression, an effective but morbidity-laden approach. Despite the low prevalence of these disorders, the authors present an impressive series of 434 patients treated for craniocervical pathology. Leveraging this extensive experience, they propose a treatment algorithm informed by patients' preoperative osseous and vascular anatomy. Remarkably, 92% of patients achieved adequate reduction and fixation without the need for a transoral approach. Moreover, several pathologies traditionally deemed irreducible were successfully treated using a posterior-only strategy in this cohort. The authors are to be commended for their rigorous efforts to better characterize, understand, and manage these complex disorders. This work meaningfully advances the reader's understanding of craniocervical pathology and represents a paradigm shift away from historic management principles. Future technical refinements can further increase the safety and efficacy of the treatment principles provided herein.

Vince Alentado

Noblesville, Indiana, USA