INTRODUCTION
The atlantoaxial segment is an anatomically complex and biomechanically challenging region of the cervical spine. Its high degree of mobility, thin cortical and trabecular bone, and the intricate course of critical neurovascular structures, such as the vertebral artery (VA) and spinal nerve roots, make surgical intervention at this level particularly challenging [1]. Instability of the C1–2 complex may result from various pathological conditions, including traumatic disruption, inflammatory diseases such as rheumatoid arthritis, and neoplastic lesions. In most cases, progressive neurological deterioration necessitates surgical stabilization [2]. Since Gallie [3] introduced posterior wiring with autologous bone graft in 1939, various techniques have emerged, including transarticular screw fixation, C1 lateral mass-C2 pedicle screw system, and C1–2 pedicle screw system.
Although each technique presents distinct biomechanical characteristics, a consensus on the optimal surgical technique for C1–2 reconstruction remains lacking. Given the anatomical variability of the C1–2 complex and the diverse etiologies of instability, the selection of an appropriate fixation strategy should aim to achieve rigid stabilization while minimizing the risk of injury to adjacent neurovascular structures [4]. Attaining this goal necessitates a comprehensive understanding of the historical evolution of surgical techniques. This review aims to explore the various surgical methods introduced to date and identify the most effective approach to atlantoaxial reconstruction.
MATERIALS AND METHODS
An online search of the PubMed database was performed to identify articles published in the English language only from 1900 to June 1, 2025, using the following keywords: “cervical atlas,” “axis,” “cervical vertebra,” “C1,” “C2,” and related terms. The complete search syntax is presented in Table 1. In total, 224 records were retrieved. Among these, 101 were excluded using automation tools before screening. The remaining 123 records were screened manually by 7 board-certified neurosurgeons based on their titles and abstracts. Of these, 42 were excluded owing to irrelevance, leaving 81 articles for full-text retrieval. Full texts were successfully obtained for 78 articles, of which 3 articles were excluded because of unavailability. Among the 78 full-text articles, 17 were excluded based on predefined criteria, such as pediatric, Nonposterior approach and revision surgeries. Ultimately, 61 studies were included in the final analysis. Of these, 41 were narratively described and cited in the text to provide a comprehensive overview of the historical evolution and conceptual development of C1–2 fixation and fusion techniques. The remaining 20 studies, published after 2000 and focusing on updated or modified surgical techniques, were systematically reviewed. Table 2 summarize the key technical features and reported clinical outcomes [5-24]. A flow diagram based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines was constructed to illustrate the study selection process (Fig. 1).
HISTORICAL EVOLUTION OF C1-2 POSTERIOR INSTRUMENTATION AND FUSION
Early attempts at atlantoaxial stabilization began with Mixter and Osgood [25] in 1910, who used stout braided sutures between the posterior arch of C1 and the spinous process of C2, securing the construct over the C2 spinous process to treat a chronic, nonunion odontoid fracture following a fall in a 15-year-old patient (Fig. 2A). In 1939, Gallie3 introduced a comparable technique, substituting braided silk sutures with steel wires. This method involves threading the wire beneath the posterior arch of C1 and looping it around the spinous process of C2, followed by bone graft fusion using an H-shaped single corticocancellous bone graft harvested from the iliac crest (Fig. 2B). This technique was further advanced by Brooks and Jenkins [26] in 1978, who introduced double wedge-shaped bone graft compression between the laminae of C1 and C2 using sublaminar stainless-steel wires (Fig. 2C). However, this technique requires an intact posterior arch of C1 and is, thus, not usually recommended for patients with rheumatoid arthritis. Moreover, early wiring techniques demonstrated relatively low fusion rates compared to modern instrumentation methods, which now achieve fusion rates exceeding 95% [27].
To overcome these limitations, Magerl and Seemann [28] introduced the transarticular screw fixation technique in 1987, which employs a screw trajectory originating from the C2 pars interarticularis, traversing the C1–2 facet joint, and terminating in the lateral mass of C1 (Fig. 2D). This technique enables fixation even in cases with a posterior arch defect of C1 by packing can-cellous bone graft into the subchondral bone of the posterior C1–2 facet joint before screw fixation. It offers superior biomechanical stability compared to previous methods and enables posterior fusion even in unstable Jefferson fractures. However, this technique poses technical challenges in patients with a narrow isthmus width or height and requires complete preopera tive reduction of the C1–2 complex [29-31].
In 1994, Goel and Laheri [32] introduced a method utilizing C1 lateral mass and C2 pedicle screws connected with a plate, providing superior screw pullout strength through its mechanical advantage. The screws were inserted into the C1 lateral mass in a posterior-to-anterior and slightly upward direction (Fig. 2E). This technique was further refined by Harms and Melcher [33] in 2001, who popularized the use of polyaxial screws and rods for fixation of the C1 lateral mass and C2 pedicle. These methods do not rely on the integrity of the C1 posterior arch or the C2 lamina, making them particularly useful in cases of C1 arch disruption or when posterior element removal is necessary for decompression. They also serve as a reliable salvage option when sublaminar wiring has failed, by offering robust lateral mass fixation.
In 2002, Resnick and Benzel [34] introduced the C1–2 pedicle screw fixation technique, which is biomechanically distinct from traditional transarticular screw fixation. While transarticular screws function as rigid cantilever beams that resist C1–2 translation independently, the pedicle screw–rod construct provides stability through a rigid screw–rod interface and the pullout strength of the screws. This construct supports axial loading in the rostrocaudal direction. The entry point of the C1 pedicle screw was located at the rostrocaudal midpoint of the C1 lamina (Fig. 2F). Screws were inserted at approximately 10° medial angulation to avoid violation of the vertebral foramen (Fig. 2F-1).
In 2003, Tan et al. [35] proposed a C1 posterior arch screw fixation technique based on an anatomical study of 50 dried atlas specimens, which they considered the functional equivalent of the pedicle in typical vertebrae. Anatomical analysis showed that 3-mm diameter, 24-mm length screws could be placed in C1 posterior arch of most specimens, except when the posterior arch was less than 4 mm thick. The entry point was located 19 mm lateral to the midline and 2 mm above the inferior border of the posterior arch, with a trajectory perpendicular to the coronal plane and angled 5° cephalad to the axial plane (Fig. 2F-2). This technique was safely applied in 5 patients with atlantoaxial instability without neurovascular injury.
CURRENT CONCEPTS AND TECHNICAL CONSIDERATIONS IN ATLANTOAXIAL POSTERIOR FIXATION
1. Comparison of 2 Most Popular Techniques: Cervical Pedicle Screws Versus Lateral Mass Screws
For fixation of the C1 vertebra, both conventional C1 lateral mass screws, inserted beneath the posterior arch, and modified C1 pedicle screws, inserted through the posterior arch, are commonly used. In 2012, Yeom et al. [14] reported a prospective study in which a single surgeon attempted posterior arch visually guided freehand insertion of 102 C1 lateral mass screws in 52 patients. Among these, 7 screws breached cranially, 30 breached caudally, 3 breached craniocaudally, and 14 resulted in vertical splitting. Among the 33 patients without preoperative occipital neuralgia, 7 developed new-onset symptoms postoperatively. Of these, 3 underwent C2 root resection and 4 underwent C2 root dissection for interarticular fusion. At the final follow-up, 5 of the 7 patients experienced complete resolution of symptoms, while 2 reported only mild residual discomfort.
A notable advantage of using C1 pedicle screws is the reduced risk of occipital neuralgia caused by C2 nerve root irritation or resection. In addition, they offer superior pullout strength and do not require dissection of the venous plexus [36]. However, concerns have been raised that the small diameter of the C1 pedicle may lead to cortical disruption or fracture during screw insertion. To address the potential risk associated with the small height of the C1 posterior arch, Lee et al. [22] in 2020 showed that safe pedicle screw placement is feasible even in C1 pedicles with a diameter of less than 4 mm, through direct visualization of the superior pedicle wall combined with serial dilation techniques. In this technique, a 2.5-mm drill bit is initially used to create a pilot hole, followed by serial enlargement with a 3.0-mm drill bit to accommodate the screw. The screw is inserted as the final step of fixation (Fig. 3). This stepwise approach minimizes the risk of cortical breach and facilitates secure screw placement within the narrow posterior arch corridor.
2. Accuracy-Enhancing Technology for Atlantoaxial Screw Placement
Owing to the complex anatomical structure of the C1–2 region, several adjunctive techniques have been employed to im-prove the accuracy of screw fixation. With advancements in imaging technology and a growing understanding of regional anatomy, these techniques now include fluoroscopy-assisted insertion, 3-dimensional (3D) image-guided navigation systems, 3D template-guided screw placement, freehand techniques, and robot-assisted instrumentation.
Tatter et al. [37] reported a single-center case series of 78 patients who underwent fluoroscopy-assisted posterior C1–2 fixation for atlantoaxial instability, in which postoperative computed tomography (CT) imaging was used to assess C1–2 screw trajectory accuracy. Among 253 screws placed, the accuracy was 98.0% for polyaxial screws and 94.2% for transarticular screws, with VA injury observed in only 2.6% of cases. These findings indicate that fluoroscopy-guided instrumentation achieves high placement accuracy with a low complication rate, and highlight the importance of postoperative CT imaging for verifying proper screw placement.
Recent advancements include intraoperative 3D image-guided navigation systems, such as the SIREMOBIL Iso-C3D (Siemens Medical Solutions, Germany) and O-arm Surgical Imaging System (Medtronic Inc., USA), which provide real-time updates without the need for anatomical registration. These systems have been shown to offer higher screw placement accuracy than conventional techniques. In one study, O-arm navigation was used to place all 44 pedicle screws successfully in the intrapedicular area without cortical breach, achieving a 100% placement accuracy [38]. In the same study, the 3D template-guided group achieved an accuracy of 98.3 percent, while the fluoroscopy-assisted group demonstrated a lower accuracy of 85.7%.
Custom navigation instruments, such as 3D templates, enable precise preoperative planning and accurate screw trajectory selection. This method reduces the operative time and radiation exposure; however, the production of templates is expensive and requires several days. In one study, 80.7% of 88 pedicle screws achieved complete intrapedicular placement without cortical breach, while 15.9% had minor cortical violations measuring less than 2 mm or under half the screw diameter [39].
Despite being technically demanding, the freehand technique has been shown by Lee et al. [22] to be effective, safe, and accurate in 2020. It offers advantages such as reduced surgical time, minimal radiation exposure, less soft tissue injury, and lower cost; however, it remains challenging for both inexperienced and experienced surgeons. In this study, 28 of 29 C1 pedicle screws inserted using the freehand technique were accurately placed within the safe zone without cortical breach (96.6 % accuracy).
In 2016, Tian [40] reported the first case of robot-assisted posterior C1–2 transarticular screw fixation using the TiRobot system (Tinavi Medical Technologies, China). This procedure was performed in a 43-year-old man with atlantoaxial instability and deformity. Intraoperative 3D imaging was used to plan the screw trajectory, and the robotic arm guided the screw insertion. The screw was accurately placed with a deviation of only 0.8798 mm, and no intraoperative complications occurred. This study demonstrated the feasibility, safety, and precision of robotic guidance in C1–2 transarticular screw fixation.
3. VA Anomalies
Anatomically, the third segment of the VA travels around the atlantoaxial segment. It passes through the C2 vertebral foramen in the transverse process, exits laterally over the C1 lateral mass, and runs posteromedially along the groove of the C1 posterior arch before entering the cranium. Anomalies such as a persistent first intersegmental artery, VA fenestration, posterior inferior cerebellar artery variations, and ponticulus posticus have been reported in this region. These anomalies require careful consideration during C1–2 posterior screw fixation procedures [41]. With advancements in C1–2 posterior screw fixation techniques, proper surgical planning is essential to prevent VA injury. Posterior occipitocervical fixation, which bypasses C1, has been advocated as an alternative [42]. Although traditionally recommended to avoid critical VA injury, this approach has disadvantages such as greater loss of motion in the C1–2 segment, increased surgical complexity, higher risk of neurological injury, increased adjacent segment degeneration, and hardware-related complications [43]. As another alternative, a lateral mass screw may be considered in the presence of ponticulus posticus to avoid VA injury [44]. However, pedicle screw placement with VA mobilization may be required used in special cases, such as in the presence of VA duplication [45]. Yeom et al. [14] demonstrated that routine dissection of the VA off the C1 posterior arch, coupled with protection using a Penfield elevator and cottonoid, effectively prevented vascular injuries. Even in cases of ponticulus posticus, the VA could be protected using this method. Byun et al. [46] observed that VA anomalies are more common in patients with atlantoaxial instability, particularly in cases of long-standing deformity rather than congenital deformity, presenting additional challenges for posterior instrumentation insertion owing to a more tenuous isthmus and pedicle. Such anatomical variations underscore the importance of preoperative radiological assessment of the VA course and surrounding bony structures to avoid complications from VA injuries. Lin et al. [47] reported that patients with bone abnormalities are at higher risk for VA anomalies, highlighting the importance of CT angiography for evaluating variant VA anatomy preoperatively.
BONE GRAFT FUSION TECHNIQUES
Advancements in posterior C1–2 fixation techniques have significantly improved bone fusion rates. However, achieving solid fusion at this C1–2 junction remains challenging owing to its complex anatomy and limited fusion surface area. To overcome these limitations, various fusion techniques have been introduced, including wiring with a bone block graft, onlay bone grafts, and interarticular bone grafts. The following section reviews these techniques, focusing on their surgical characteristics, biomechanical advantages, and limitations to guide the selection of optimal bone fusion strategies (Table 3).
1. Wiring With Bone Block Graft
In 1939, Gallie [25] proposed a posterior fusion technique using an H-shaped corticocancellous bone block graft harvested from the iliac crest, and placed over the dorsal aspect of the C1 posterior arch and upper part of the C2 spinous process (Fig. 2B). The graft was secured at the midline with a single sublaminar wire. However, this technique was limited in cases of posteriorly displaced and irreducible C1 rings and showed inadequate rotational stability. To overcome these limitations, Brooks and Jenkins [26] introduced the wedge-compression technique in 1978. They inserted 2 corticocancellous bone grafts bilaterally between the posterior arch of C1 and the laminae of C2 (Fig. 2C). In 1998, Dickman and Sonntag [48] proposed a simplified posterior C1–2 fusion technique using a single bicortical bone graft. A central notch was created at the inferior aspect of the graft to seat it against the C2 spinous process, functioning as a structural strut to enhance segmental stability. The construct was reinforced with a double-stranded twisted interspinous wire that passed through notches created bilaterally at the inferior aspect of the C2 spinolaminar junction.
However, with the advancement of segmental instrumentation and improvements in biomechanical stability, the necessity of wire-based fusion techniques has increasingly been questioned. Wires may be regarded as additional implants that carry the risk of neural irritation and can prolong the operative time. Furthermore, when decompression requires resection of the C1 posterior arch, such techniques become inapplicable. For these rea sons, the use of wire-based fusion techniques is declining in current clinical practice [49].
2. Onlay Bone Graft Fusion Techniques
With advancements in fixation techniques, onlay bone grafting has become a widely used to achieve C1–2 fusion. In 2012, Ni et al. [50] proposed a modified atlantoaxial fusion technique using a bicortical iliac crest bone graft placed between the C1 posterior arch and the C2 lamina under load-bearing conditions to achieve bone-to-bone continuity. This method eliminates the need for supplemental wiring and may improve resistance to extension in screw–rod constructs. Postoperative CT demonstrated a 94.3% fusion rate within 3 months. However, this technique remains feasible only when the posterior C1–2 elements are anatomically preserved. Moreover, reports indicate that autologous bone grafts placed directly on the bony fusion bed can be absorbed due to micromobility and local inflammatory reactions [51]. To overcome these limitations, locally harvested autologous bone blocks can be carefully placed between the inferior wall of the C1 pedicle and the superior wall of the C2 pedicle (Fig. 4). This technique increases bone-to-bone connectivity and provides a larger fusion bed for additional onlay bone grafts. The structural bone graft is stabilized under load-bearing conditions, increasing the chance of successful fusion. It also reduces graft resorption caused by micromotion and improves overall mechanical stability. In addition, combining osteoconductive and osteoinductive biomaterials around the graft site may promote early bone healing and improve the overall fusion outcomes.
3. Interarticular Fusion Techniques
Numerous studies have demonstrated that the facet joints, as regions subjected to concentrated axial loading in the spine, experience significant compressive forces and have traditionally served as reliable fusion beds in spinal arthrodesis [52,53]. In 2007, Goel [54] reported a joint-jamming technique in which a titanium spacer measuring 6 mm in height, featuring a spiked and multiholed design, was combined with autologous bone graft and impacted into the C1–2 facet joint space. Solid fusion was achieved in all 4 patients during a 16-month follow-up without screw– rod fixation. In 2017, Turel et al. [55] applied a technique involving the insertion of allograft spacers into the C1–2 facet joints followed by posterior screw–rod fixation in 19 patients with atlantoaxial instability, achieving solid arthrodesis in 94% of patients at a mean follow-up of 12.1 months. More recently, in 2022, Oh et al. [23] described a technique combining unilateral C1–2 interarticular fusion using local autologous bone graft with C1–2 pedicle screw fixation (Fig. 5). In their series of 25 patients, radiological follow-up demonstrated solid arthrodesis in all cases, with no instances of postoperative occipital neuralgia or significant complications. Considering the comparable fusion rates achieved in the atlantoaxial region, the location of the fusion bed appears to be more critical than that of the graft material used. This approach allows access below the C2 nerve root, thereby avoiding disruption of the C2 venous plexus capsule. This technique requires only a small amount of bone graft, that can be sufficiently harvested from the C2 spinous process. Additionally, it enables slight distraction of the C1–2 joint space, facilitating the restoration of facet joint height.
4. Use of Graft Materials: Iliac Bone, Autologous Bone, Allograft, and Synthetic Bone
There are both advantages and disadvantages to interlaminar bone block and wiring fixation techniques. One significant limitation is their reliance on the integrity of the C1 posterior arch. In cases involving traumatic fractures, congenital anomalies such as axis assimilation, or posterior decompression, alternative methods may be required. Another important aspect influencing fusion success is the choice of graft material. Both allograft and autologous bone are widely used for spinal fusion [50]. Despite concerns regarding infectious complications and cost, allograft cancellous bone provides confirmed osteoconductive properties. In cases with neoplastic or osteomyelitis etiologies, allograft cancellous bone may serve as an appropriate fusion material. Harvesting iliac crest tricortical autologous bone harvest is generally safe and feasible, offering osteoinductive, osteoconductive, and osteogenic properties. However, complications such as donor-site infection, pain, and lateral femoral cutaneous nerve injury have been reported. Local autologous bone harvested from the posterior elements (C2 spinous process or C1 lamina) during posterior decompression could also be used as graft material. Although the available quantity is limited, these small fixed bone blocks have shown promising fusion rates in interarticular fusion procedures. As an alternative technique, Aryan et al. [9] reported that the use of allograft spacers supplemented with recombinant human bone morphogenetic protein (rhBMP) achieved fusion rates comparable to those observed using iliac bone crest grafts while reducing complications. However, owing to cost considerations, routine use of rhBMP is not recommended. Synthetic bone grafts, including β-tricalcium phosphate and hydroxyapatite, have been developed in various formulations for spinal fusion [51]. These materials primarily serve as osteoconductive scaffolds and may be combined with osteoinductive adjuncts to enhance bone regeneration. According to recent research, a randomized clinical trial of posterolateral lumbar fusion demonstrated that patient-reported outcomes were comparable between the ABM/P-15 and allograft groups, even though the fusion rate in the ABM/P-15 group was significantly higher, 50 percent compared with 20 percent in the allograft group [56]. However, clinical evidence in the setting of atlantoaxial fusion remains limited, and further well-designed clinical studies are warranted to validate its efficacy and safety.
REDUCTIONS
Although traditionally considered standard treatments for basilar invagination, in which the tip of the odontoid process is positioned into the foramen magnum above the McRae and Chamberlain lines [57,58], transoral decompression and/or posterior occipitocervical fixation are limited by narrow indications and high morbidity rates [59,60]. These limitations have prompted the development of alternative approaches aimed at reducing basilar invagination by distracting the C1–2 joint. The rationale for this approach is that distraction of the C1–2 joint may allow the descent of the odontoid process from the foramen magnum.
In 2004, Goel [61] reported a series of 22 patients with basilar invagination who underwent C1–2 joint distraction. Large corticocancellous bone grafts, along with strut grafts composed of hydroxyapatite blocks or metal spacers, were packed into the distracted articular cavity to achieve reduction. The mean height of the spacers used was 3 mm. However, during exposure and spacer insertion, bilateral sacrifice of the C2 nerve roots is often required, which may result in postoperative occipital neuralgia as a form of neurogenic pain.
More recently, Oh et al. [23] introduced an interarticular fusion technique using small autologous bone chips harvested from the C2 spinous process. This technique does not require routine spacer insertion for reduction, except in cases of severe basilar invagination. The authors recommend this technique as a treatment protocol (Fig. 6).
DISCUSSION
With the accumulation of extensive research and the advancement in modern medical technology, atlantoaxial fusion surgery has undergone artful evolution over time. It has progressed from wiring fixation to screw fixation, fluoroscopy-guided techniques to navigation-guided techniques integrated with robotic systems, large iliac bone block grafting to synthetic bone grafting, and wiring-based large bone block grafting to interarticular bone block grafting. As part of this evolution, preoperative CT angiography has become essential for identifying vascular anomalies and preventing VA injury, particularly in patients with anatomical variations [41]. However, some of these techniques remain controversial. For example, interarticular fusion using local autologous bone chips is known to provide an effective and biologically favorable environment for arthrodesis [23]; however, fusion with onlay grafts remains debated because of the high rate of graft resorption. In the future, as medical technology advances and surgeons gain more experience, these controversies are likely to decrease, and C1–2 fusion will become more precise, refined, and tailored to each patient’s anatomy.
Since the 2000s, numerous clinical and biomechanical studies have emphasized individualized surgical strategies, along with advances in minimally invasive techniques, and biological fusion enhancement. Maximizing the clinical benefits of these technologies requires a systematic and well-defined decisionmaking framework. We developed a decision-making protocol that guides surgical instrumentation and fusion strategies based on VA anomalies, osseous variants such as ponticulus posticus, and the presence of basilar invagination. The complete protocol is illustrated in Fig. 6, which outlines the surgical algorithm for atlantoaxial instrumentation, reduction, and fusion. The initial consideration is the presence of a VA anomaly. If the VA course is abnormal, either a pedicle or lateral mass screw is selected, depending on its trajectory. Once instrumentation is determined, the presence of basilar invagination guides the decision on whether reduction is required. If basilar invagination is absent, direct posterior C1–2 fusion is performed using a local bone graft harvested from the C2 spinous process. If basilar invagination is identified, vertical reduction and distraction are achieved using a cage or C1 laminectomized bone graft. Before interarticular reduction or fusion, we should determine whether C1 laminectomy is necessary. If laminectomy is indicated, the excised C1 lamina is fashioned into a structural graft and inserted into the facet joint space. If laminectomy is not required, interarticular fusion is performed by using an interbody cage filled with autologous bone harvested from the C2 spinous process. We strongly against the routine use of interarticular artificial cage insertion as well as C2 nerve root sacrifice. This algorithm provides a structured, pathology-specific framework for atlantoaxial instrumentation and fusion. Incorporating anatomical variants and the presence of basilar invagination facilitates personalized surgical planning and enhances procedural safety. As surgical technologies evolve, decision-making protocols will remain critical for achieving stable fixation and successful arthrodesis at the craniovertebral junction.
CONCLUSION
Atlantoaxial fusion surgery continues to advance toward greater precision and individualization, based on patient-specific anatomical characteristics with a technical development. The major artful evolutions were the use of screws for segmental rigid fixation, interarticular reduction, and interarticular fusion. C1 lateral mass, posterior arch, pedicle screws and C2 pedicle, lamina screws give us much stronger stability and higher fusion rates. Interarticular fusion using local bone also gives us technical easiness guaranteeing high fusion rate overcoming inconvenience of wiring and iliac bone harvest. Interarticular height reduction and interarticular fusion should be discriminated.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Print
