Prognostic Factors in Craniocervical Realignment for Crainovertebral Junction Kyphosis With Negative Cervical Imbalance: A Comprehensive Study
Article information
Abstract
Objective
To elucidate the clinical outcomes of craniocervical realignment surgery in patients with craniovertebral junction (CVJ) kyphosis accompanied by negative sagittal imbalance, and to identify radiological predictors associated with favorable outcomes.
Methods
A retrospective analysis was performed on 28 patients who underwent craniocervical realignment between 2014 and 2022 for CVJ kyphosis with accompanying negative sagittal imbalance. Clinical outcomes were evaluated using the Neck Disability Index (NDI), visual analogue scale for neck pain, and the Japanese Orthopaedic Association (JOA) score. Radiographic parameters included the C0–2 angle and the C2–7 sagittal vertical axis (SVA). Favorable outcomes were defined as an improvement of more than 20 points in the NDI and a JOA recovery rate exceeding 50%. Multiple linear regression and receiver operating characteristic (ROC) curve analyses were conducted to identify independent predictors and to determine optimal threshold values.
Results
Significant improvements in both clinical outcomes and radiographic alignment were observed in association with craniocervical realignment surgery. Patients who achieved favorable outcomes exhibited greater postoperative changes in the C0–2 angle and the C2–7 SVA. Multivariate analysis identified changesm in the C0–2 angle (p=0.019) and C2–7 SVA (p=0.010) as independent predictors of NDI improvement, while age (p=0.033) and C2–7 SVA change (p=0.037) were independently associated with the JOA recovery rate. ROC curve analysis determined optimal cutoff values of ≥10.65° for C0–2 angle change and ≥19.2 mm for C2–7 SVA change, with corresponding area under the curve values of 0.872 and 0.802, respectively.
Conclusion
Craniocervical realignment appears to be a viable surgical option for patients with CVJ kyphosis and negative sagittal imbalance. Postoperative changes in C0–2 angle and C2–7 SVA were found to be associated with favorable clinical and functional outcomes, suggesting their potential role as prognostic factors.
INTRODUCTION
The craniovertebral junction (CVJ), which comprises the occiput, atlas (C1), and axis (C2), serves as the transitional zone between the skull and the cervical spine and plays a critical role in both stability and mobility of the upper cervical region [1-3]. CVJ disorders encompass a spectrum of rare but clinically significant conditions that arise from various etiologies, including congenital anomalies, trauma, inflammatory diseases such as rheumatoid arthritis (RA), and infections [4,5]. These disorders can lead to structural instability or kyphotic deformity at the CVJ, resulting in compression of the cervicomedullary junction and upper cervical spinal cord. Clinical manifestations include myelopathy, axial neck pain, restricted cervical motion, and impaired horizontal gaze. Due to the anatomical complexity and proximity to critical neurovascular structures, CVJ pathologies can have serious neurological consequences, making early recognition and appropriate surgical intervention essential [1,6-8].
Epidemiological studies have reported variable prevalence depending on the underlying etiology: atlantoaxial instability occurs in up to 20% of patients with Down syndrome, RA affects the cervical spine in 25%–86% of cases with craniocervical settling observed in approximately 8%, and foramen magnum stenosis is a common feature in achondroplasia with surgical decompression required in up to 40% of infants [9-13]. However, to the best of our knowledge, comprehensive epidemiological data encompassing the overall spectrum of CVJ disorders are lacking, largely due to their heterogeneity and rarity.
The sagittal balance of the subaxial cervical spine has been the focus of extensive research, with numerous studies demonstrating a strong correlation between sagittal alignment and clinical outcomes, thereby enhancing the understanding of cervical spine pathology [14-17]. Kyphosis and sagittal malalignment have been recognized as key determinants of surgical outcomes in both the thoracolumbar and cervical regions. A growing body of evidence has highlighted the detrimental effects of positive sagittal imbalance on postoperative results, underscoring the clinical importance of correcting this condition [18-24]. However, an equally critical but less thoroughly explored issue is the impact of negative sagittal imbalance and malalignment, particularly in patients with CVJ kyphosis (Fig. 1). Despite its potential relevance, the clinical implications of negative sagittal imbalance in this population—especially in the context of surgical intervention—remain inadequately investigated.
Illustration of negative sagittal balance in the subaxial spine with craniovertebral junction (CVJ) kyphosis in a 48-year-old female. CVJ kyphosis can lead to negative sagittal imbalance, a less common presentation.
This study aims to address this gap in the literature by evaluating the outcomes of craniocervical realignment procedures in patients with CVJ kyphosis and negative sagittal imbalance. Through a comprehensive analysis of radiographic and clinical parameters, the authors seek to assess the effectiveness of surgical intervention and to identify prognostic factors associated with favorable outcomes. The ultimate objective is to establish predictive indicators that may guide clinical decision-making and improve management strategies for this complex and underrecognized pathology.
MATERIALS AND METHODS
1. Study Design
This retrospective study, approved by the Institutional Review Board of the Catholic University of Korea (No. PC23DASS0107, approved on July 3, 2025), investigated the clinical outcomes of craniocervical realignment procedures in patients with CVJ kyphosis and negative sagittal imbalance over a study period spanning from January 2014 to December 2022. Written informed consent was waived due to the retrospective nature of the study, which was conducted using anonymized data.
2. Patient Selection
The inclusion criteria were carefully defined to ensure a homogeneous study population:
(1) C0–2 angle: Patients with a C0–2 angle less than 0 de-grees were included. In this study, CVJ lordosis was defined as a positive value and kyphosis as a negative value for the C0–2 angle.
(2) C2–7 sagittal vertical axis (SVA): Those with a C2–7 SVA less than 0 mm were eligible.
(3) Follow-up: A minimum follow-up period of one year after craniocervical realignment was mandated.
Patients with incomplete radiographic or clinical data, or those presenting with concurrent thoracolumbar deformity, were excluded from the analysis.
3. Surgical Techniques
CVJ instrumentation was primarily based on the Goel-Harms technique [25-28], with case-specific modifications applied as needed. The following section outlines the distinct surgical modifications applied in this study.
In patients with severe CVJ kyphosis and oblique C1–2 facet joint orientation, partial resection of the posterior superior corner of the C2 facet and the anterior inferior portion of the C1 facet was performed to flatten the oblique C1–2 facet joint. Additional procedures—such as drilling, reaming, and curettage— were utilized to release the fixed C1–2 joint and to reduce the facet inclination. To widen the joint space and achieve both vertical and horizontal reduction of the subluxation, an osteotome was employed as a lever, using the posterior edge of the C2 facet as a fulcrum [1,29].
The rod was contoured beyond the patient’s native CVJ alignment. The head was then extended using the Mayfield head holder to facilitate rod placement and achieve CVJ kyphosis correction, either to the occipital plate or to the uppermost cervical anchor point depending on the construct. Longitudinal distraction between the screws facilitated anterior and inferior translation of the odontoid process. Once adequate distraction of the C1–2 joint was achieved, bilateral insertion of facet cages or bone blocks was performed. With the head placed in extension, the occipital and C1 screws were subsequently tightened to further reduce C1–2 subluxation and correct CVJ kyphosis. In severe cases with marked spinal cord compression, both vertical reduction of the C1–2 joint and head extension were essential to achieve 3-dimensional realignment.
Caution is necessary to prevent excessive spinal cord stretching and potential myelopathy, particularly due to a ventral bony mass and kyphotic angulation. Intraoperative neuromonitoring (IONM) is useful for preventing overdistraction and reducing the risk of spinal cord or vertebral artery injury.
4. Radiological and Clinical Assessment
Radiological parameters—including the C0–2 and C2–7 angles, SVA, both C2 and C7 slopes, and the cervicomedullary angle (CMA)—were measured preoperatively and at one year postoperatively. Clinical outcomes were evaluated using Japanese Orthopaedic Association (JOA) score, Neck Disability Index (NDI), and visual analogue scale (VAS) for neck pain.
Lateral, flexion, and extension radiographs of the cervical spine were obtained both preoperatively and postoperatively. Wholespine radiographs were also acquired to evaluate spinopelvic parameters. Lateral cervical radiographs were taken with the patient maintaining a horizontal gaze. Radiographic assessments were performed by 3 independent spine surgeons who were blinded to the clinical information, and the average of their measurements was used for analysis.
The radiographic parameters analyzed included the C0–2 angle, C0–2 range of motion (ROM), C2–7 angle, C2–7 ROM, C7 SVA, C2–7 SVA, C2 slope, C7 slope, CMA, C1–2 facet inclination angle (FIA), thoracic kyphosis (TK), pelvic incidence (PI), and lumbar lordosis (LL).
The C0–2 and C2–7 Cobb angles were used as indicators of upper and lower cervical alignment, respectively, and were defined by the angles formed between the McGregor line and the inferior endplate of C2, and between the inferior endplates of C2 and C7. The C2–7 SVA was defined as the horizontal distance between the posterosuperior corner of C7 and a vertical plumb line drawn from the center of C2. The C2 slope was measured as the angle between the horizontal reference line and the inferior endplate of C2, and the C7 slope as the angle between the horizontal reference and the superior endplate of C7 (Fig. 2A). Forward-sloping lines were assigned positive values, whereas backward-sloping lines were assigned negative values. In cases where the C7 vertebra was obscured by the shoulders and not clearly visible on lateral radiographs, the C6 vertebra was used as a substitute reference point for radiographic measurements. The CMA was defined as the angle between the medulla oblongata and the cervical spinal cord on midsagittal T2-weighted magnetic resonance imaging (Fig. 2B). C1–2 FIA was measured as the angle between the C1–2 facet joint and a horizontal reference line on sagittal reconstructed computed tomography images (Fig. 3A–D).
Cervical alignment and cervicomedullary angle measurements. (A) Radiographic measurements of cervical alignment parameters in a 59-year-old female. C0–2 lordosis: angle between McGregor line (connecting hard palate to inferior occiput margin) and C2 inferior endplate. C2–7 lordosis: Cobb angle between C2 and C7 inferior endplates. C2–7 sagittal vertical axis: horizontal distance between C2 plumb line and C7 posterior superior endplate. Positive values indicate anterior deviation; negative values indicate posterior deviation. The yellow line indicates the horizontal reference line. (B) Changes in cervicomedullary angle (CMA) in a 50-year-old female. Sagittal magnetic resonance imaging (MRI) images showing CMA. The CMA is defined as the angle between the medulla oblongata’s midline and the cervical cord’s midline in the sagittal MRI of the cervical spine
Craniovertebral junction (CVJ) kyphosis & facet inclination. Representative sagittal computed tomography (CT) images illustrating the relationship between CVJ kyphosis and C1–2 facet joint inclination. Preoperative sagittal CT images from a 63-year-old male (A) and a 68-year-old male (B) showing oblique orientation of the C1–2 facet joints, contributing to CVJ kyphosis. Arrows indicate the C1–2 facet joints, highlighting their oblique orientation in patients with CVJ kyphosis. Postoperative sagittal CT images of the same patients, with panel C corresponding to the postoperative image of patient panel A and panel D corresponding to the postoperative image of patient panel B, demonstrating changes in facet joint inclination after surgical realignment, with improved alignment of the C1–2 facets (black arrows). The yellow line indicates the horizontal reference line.
Alignment changes were calculated as follows:
Change in value=postoperative value–preoperative value
Neurological outcomes were primarily assessed using the JOA score, while axial symptom severity was evaluated using the NDI and VAS for neck pain [30,31]. The JOA recovery rate (RR) was calculated using the following formula:
JOA RR (%)=[(postoperative JOA score–preoperative JOA score)(/17–preoperative JOA score)]×100
NDI was expressed as a percentage (maximum score of 100%), with higher scores indicating greater disability. VAS neck scores ranged from 0 (no pain) to 10 (maximum pain).
5. Patient Stratification
Patients were categorized into groups based on their neurological recovery outcomes:
(1) JOA RR: A JOA RR greater than or equal to 50% was classified as a favorable outcome, while a JOA RR less than 50% was considered unfavorable [32].
(2) NDI Improvement: The minimum clinically important difference (MCID) was evaluated by calculating the change between postoperative and preoperative NDI scores. This difference was compared against a previously established MCID threshold of -20 points [33]. Based on this criterion, patients were further stratified into favorable (NDI improvement ≥20, achieving MCID) and unfavorable (NDI improvement ≤20) outcome groups with respect to axial symptom improvement.
6. Statistical Analysis
Data analyses were conducted using IBM SPSS Statistics ver. 20.0 (IBM Co., USA). Continuous variables are presented as mean±standard deviation, while categorical variables are expressed as the number of subjects in each group.
The normality of continuous variables was assessed using the Shapiro-Wilk test. Based on the results, parametric tests (paired t-tests) were applied to normally distributed variables for preand postoperative comparisons, whereas nonparametric tests (Wilcoxon signed-rank test) were used when normality could not be assumed. Differences between the 2 groups were analyzed using the Mann-Whitney U-test for continuous variables and Fisher exact test for categorical variables.
To identify independent prognostic factors for clinical outcomes, multivariate linear regression analysis was performed. The receiver operating characteristic curve was plotted, and the Youden index was used to determine the optimal cutoff value.
A p-value of <0.05 was considered statistically significant.
7. Illustrative Cases
1) A case of favorable outcome in a 55-year-old female
The patient, a 55-year-old female, had experienced severe neck pain, gait disturbance, and impaired fine motor function of the hands for 6 months prior to presentation. Preoperative radiographs demonstrated severe CVJ kyphosis with negative sagittal imbalance (C0–2 angle=-19°, C2–7 SVA=-26 mm) and compensatory lower cervical hyperlordosis (C2–7 angle=46°). Following surgical correction, postoperative imaging revealed restoration of neutral CVJ alignment (C0–2 angle=0°, C2–7 SVA=0 mm) and reduction of hyperlordosis (C2–7 angle=30°). Clinically, neck pain improved dramatically (VAS: 8→1), with marked recovery in functional status as reflected by improvements in NDI: 52→8 and JOA: 7→14, accompanied by substantial gains in both gait and hand dexterity (Fig. 4A and B).
A case of favorable outcome in a 55-year-old female. (A) Preoperative lateral cervical x-ray showing severe craniovertebral junction (CVJ) kyphosis with negative sagittal imbalance (C0–2=-19°, C2–7 SVA=-26 mm) and compensatory lower cervical hyperlordosis (C2–7=46°). (B) Postoperative x-ray demonstrating correction of CVJ kyphosis (C0–2=0°, C2–7 SVA=0 mm) and reduced hyperlordosis (C2–7=30°), with significant clinical improvement (VAS neck: 8→1, NDI: 52→8, JOA: 7→14). A case of unfavorable outcome in a 43-yearold female. (C) Preoperative lateral cervical x-ray showing severe CVJ kyphosis with negative sagittal imbalance (C0–2=-21°, C2–7 SVA=-33 mm) and compensatory hyperlordosis (C2–7=35.3°). (D) Postoperative x-ray showing minimal correction of CVJ kyphosis (C0–2=-18.5°, C2–7 SVA=-28 mm), with limited clinical improvement (VAS neck: 3→5, NDI: 23→25, JOA: 14→15). The yellow line indicates the horizontal reference line. SVA, sagittal vertical axis; VAS, visual analogue scale; JOA, Japanese Orthopaedic Association; NDI, Neck Disability Index.
2) A case of unfavorable outcome in a 43-year-old female
The patient, a 43-year-old female, presented with gait disturbance, bilateral hand weakness, and urinary dysfunction that had developed over the preceding year. Neck pain at presentation was manageable with medication. Preoperative radiographs demonstrated severe CVJ kyphosis with negative sagittal imbalance (C0–2 angle=-21°, C2–7 SVA=-33 mm) and compensatory hyperlordosis (C2–7 angle=35.3°). She underwent C1 laminectomy followed by fixation without correction of CVJ kyphosis or cervical negative balance. Postoperative imaging revealed minimal improvement in alignment (C0–2=-18.5°, C2–7 SVA=-28 mm, C2–7 angle=32.1°), and clinical outcomes were unsatisfactory, with deterioration in neck pain (VAS: 3→5) and no meaningful improvement in functional scores (NDI: 23→25, JOA: 14→15) (Fig. 4C and D).
RESULTS
1. Patient Demographics
Of the 260 patients who underwent upper cervical spine surgery between January 2014 and December 2022, 227 were excluded due to having a C0–2 angle≥0° or a C2–7 SVA≥0 mm. An additional 5 patients were excluded due to insufficient follow-up data up to 12 months postoperatively. Consequently, a total of 28 patients were included in the study (Fig. 5).
Flowchart of patient selection. SVA, sagittal vertical axis; CVJ, craniovertebral junction; F/U, follow-up.
A total of 28 consecutive patients diagnosed with CVJ kypho-sis and negative cervical imbalance underwent craniocervical realignment surgery (male:female=6:22, mean age=57.1± 17.3 years). The most common etiology of CVJ kyphosis was congenital anomalies (n=16, 57.1%), followed by RA (n=5, 17.9%) and trauma (n=2, 7.1%). Given the diversity and potential ambiguity of congenital anomalies, their detailed subtypes have been presented separately (Supplementary Table 1).
The upper instrumented vertebrae were either between occiput (C0) (n=15, 53.6%) and C1 (n=13, 46.4%). The lower instrumented vertebrae varied across cases, with the majority involving C2 (n=14, 50%), followed by C4 (n=6, 21.4%), C3 (n=5, 17.9%), and a smaller number extending to C5 or C7 (Table 1).
Demographics of patients (n=28)
2. Follow-up Analysis
Follow-up analysis demonstrated significant improvements in both clinical and radiological parameters following craniocervical realignment surgery. Statistically significant postoperative changes were observed in the C0–2 angle, C2–7 angle, C2 slope, CMA, and C2–7 SVA (p<0.001). In contrast, parameters such as C7 SVA, TK, LL, and PI did not show significant postoperative differences (p=0.998, p=0.292, p=0.641, and p=0.608, respectively).
In addition to radiological improvements, clinical outcomes also showed substantial postoperative enhancement. Scores for the NDI, JOA, and VAS for neck pain all improved significantly (p<0.001). These results underscore the association between craniocervical realignment surgery and improvements in both anatomical alignment and patient-reported outcomes (Table 2).
Preoperative and postoperative clinical and radiological parameters
To ensure measurement consistency, the intra- and interobserver intraclass correlation coefficients (ICCs) for both preoperative and postoperative measurements generally ranged from 0.763 to 0.999, indicating good to excellent reliability across variables. However, the postoperative interobserver ICC for PI was 0.452 (Supplementary Table 2).
3. NDI Outcomes
Among the 28 patients, 16 were classified into the favorable NDI outcome group (NDI improvement≥20), while 12 exhibited unfavorable outcomes (NDI improvement <20).
Patients with favorable outcomes demonstrated a significant postoperative increase in both the C2–7 SVA (p=0.006) and the C0–2 angle (p<0.001). Improvement in NDI scores was significantly correlated with correction of CVJ kyphosis (ΔC0–2 angle, p=0.002) and reduction of negative sagittal imbalance (ΔC2–7 SVA, p=0.003) following surgery (Table 3). In contrast, age and other radiological parameters did not show statistically significant associations with NDI improvement.
Comparison of clinical and radiological changes between favorable and unfavorable NDI outcome groups
4. JOA RR Analysis
Among the 28 patients, 20 were categorized into the favorable outcome group (JOA RR≥50%), while 8 patients exhibited unfavorable outcomes (JOA RR<50%). Comparative analysis between the 2 groups revealed significantly greater changes in the C0–2 angle change (12.5° vs. 7.9°, p=0.049) and the C2–7 SVA change (18.8 mm vs. 8.6 mm, p=0.013) in the favorable group. Although the mean age tended to be higher in the unfavorable group, the difference did not reach statistical significance (p=0.055). JOA RR was significantly correlated with the degree of correction in negative sagittal imbalance (ΔC2–7 SVA, p=0.001) following surgery (Table 4).
Comparison of clinical and radiological changes between favorable and unfavorable JOA recovery rate groups
5. Multivariate Regression Analysis
To determine the prognostic factors for NDI improvement and neurological recovery in patients with CVJ kyphosis, and to account for potential confounding variables, a multivariate regression analysis was performed. This analysis revealed independent associations between specific radiological parameters and clinical outcomes (Table 5). Specifically, changes in the C0–2 angle and C2–7 SVA were independently associated with improvements in NDI scores. Although age did not reach statisti-cal significance in the univariate group comparison (p=0.055), both age and C2–7 SVA change were independently associated with JOA RR improvement in the multivariate analysis (p=0.033 and p=0.037, respectively).
Multivariate analysis of prognostic factors for clinical outcomes
Furthermore, the optimal cutoff points for predicting favorable NDI outcome were calculated using the Youden index (Fig. 6). The cutoff values were identified as 10.65° for the change in C0–2 angle and 19.2 mm for the change in C2–7 SVA. Both parameters demonstrated strong predictive ability for favorable NDI outcomes, as indicated by area under the curve values of 0.872 and 0.802, respectively.
Receiver operating characteristic (ROC) curve for predicting favorable Neck Disability Index (NDI) outcomes. The ROC curve illustrates the predictive ability of changes in radiological parameters, specifically ΔC0–2 angle (red line) and ΔC2–7 sagittal vertical axis (SVA) (blue line), for favorable NDI outcomes. The area under the curve (AUC) for ΔC0–2 angle is 0.872 (95% confidence interval [CI], 0.730–1.000; p=0.001), indicating very good predictive ability. The AUC for ΔC2–7 SVA is 0.802 (95% CI, 0.641–0.963; p=0.007), also demonstrating good predictive ability. The optimal cutoff values, determined using the Youden index, are 10.65° for ΔC0–2 angle and 19.2 mm for ΔC2–7 SVA. Sensitivity and specificity at these cutoff points highlight their clinical relevance in predicting favorable NDI outcomes after craniocervical realignment surgery.
6. C2 Nerve Resection and Complications
Of the 28 patients analyzed in this series, the C2 nerve roots were preserved in 6 cases (21.4%), resected bilaterally in 18 cases (64.3%), and resected unilaterally in 4 cases (14.3%) (Table 6). Among the 22 patients who underwent C2 nerve resection, postoperative sensory sequelae were noted in the majority. Numbness in the C2 dermatome without associated pain developed in 17 patients (77.3%), while only 1 patient (4.5%) experienced persistent C2 neuropathic pain at 1.5 years postoperatively. In contrast, 4 patients (18.2%) demonstrated improvement in preoperative C2-related neuropathic pain following surgery.
Extent of C2 nerve resection and perioperative complications
With respect to general surgical complications, adverse events were relatively uncommon. One patient (3.6%) experienced a transient intraoperative reduction in motor evoked potential (MEP), but no lasting neurological deficit was observed after the surgery. Another patient (3.6%) required revision surgery for postoperative hematoma evacuation. Two patients (7.1%) developed transient dysphagia, likely related to retropharyngeal swelling after C1–2 facet manipulation, which resolved within several days with conservative management. No cases of permanent neurological injury or instrumentation-related failure were identified during the minimum 1-year follow-up period.
DISCUSSION
1. Summary of Key Findings
In this study, postoperative changes in radiological parameters were associated with improvements in clinical outcomes among patients undergoing craniocervical realignment for CVJ kyphosis with negative cervical imbalance. Substantial changes were observed in C0–2 angle, C2–7 angle, C2 slope, and C2–7 SVA, alongside improvements in JOA, NDI, and VAS scores. These findings suggest that surgical realignment may help address the complex anatomical distortions and compensatory mechanisms observed in CVJ kyphosis with negative cervical balance.
2. Correlation Between Radiological Changes and Clinical Outcomes
The study demonstrates a strong correlation between the correction of CVJ kyphosis (ΔC0–2 angle) and the resolution of negative sagittal imbalance (ΔC2–7 SVA) with improvements in NDI scores, underscoring the importance of addressing both CVJ alignment and cervical sagittal balance to optimize clinical outcomes. Additionally, data analysis revealed a correlation between the change in C0–2 angle and C2–7 SVA, suggesting that hyperlordosis of the lower cervical spine, reversed C2 slope, and negative sagittal balance serve as compensatory mechanisms for CVJ kyphosis.
3. Compensatory Mechanisms in CVJ Kyphosis
The data indicate that hyperlordosis of the lower cervical spine, reversed C2 slope, and negative sagittal balance serve as compensatory mechanisms for CVJ kyphosis to maintain horizontal gaze. Correction of CVJ kyphosis resulted in the resolution of these compensatory adaptations, thereby restoring sagittal balance. This finding underscores the interdependent nature of cervical spine alignment and emphasizes the importance of addressing the primary deformity.
4. Prognostic Factors for Clinical Improvement
Multivariate regression analysis identified changes in C0–2 angle and C2–7 SVA as independent predictors of NDI improvement. For neurological improvement, age and C2–7 SVA change were identified as potential independent factors associated with JOA RR. This relationship may be better understood when considering the nature of CVJ kyphosis and its long-term effects on patients.
CVJ kyphosis is frequently associated with congenital anomalies [8], suggesting that many patients may have experienced prolonged myelopathic symptoms prior to surgical intervention. This extended symptom duration—often correlated with patient age—may have a substantial impact on the JOA RR.
5. Complications and C2-Related Outcomes
In our cohort, the overall rate of perioperative complications was relatively low and comparable to previously published series of craniocervical reconstruction [34]. The most frequent C2-related consequence was postoperative numbness, occurring in more than three-quarters of patients who underwent C2 nerve resection. This finding is consistent with existing literature, which has noted high rates of sensory disturbance following C2 manipulation or sacrifice, although these symptoms are generally well tolerated and rarely functionally disabling [35,36]. Notably, only 1 patient (4.5%) developed persistent neuropathic pain, while 4 patients (18.2%) experienced improvement of preoperative radicular pain, suggesting that C2 resection does not invariably worsen quality of life and may, in selected cases, provide symptomatic benefit.
Perioperative complications were infrequent and included transient intraoperative MEP changes (3.6%) without postoperative neurological deficit, hematoma requiring revision (3.6%), and transient dysphagia (7.1%). Notably, no cases of long-lasting dysphagia, permanent neurological deficit, or hardware-related failure occurred during follow-up. Achieving lordotic craniocervical realignment in patients with CVJ kyphosis appears to play an important role in minimizing such complications. In particular, restoration of a more physiological craniocervical alignment may reduce the risk of postoperative dysphagia and dyspnea by enlarging the oropharyngeal space and alleviating mechanical compression on swallowing and airway structures. These findings support the concept that alignment correction is essential not only for neurological recovery but also for optimizing airway and swallowing function [37,38]. Furthermore, our complication rates compare favorably with prior series of occipitocervical and atlantoaxial fusion, in which wound-related, neurological, or instrumentation-related complications have been reported in up to 10%–20% of cases [27,39]. Taken together, our results suggest that craniocervical reconstruction, despite its technical and anatomical challenges, can be performed safely with acceptable and manageable complication profiles in appropriately selected patients.
6. Clinical Implications and Surgical Strategy
The degree and reducibility of the deformity largely determine whether posterior fixation alone is adequate or whether formal realignment should be pursued. Realignment is generally indicated for patients with functional impairment, progressive deformity, neurological compromise, or sagittal imbalance unresponsive to conservative management, and may also be warranted when kyphosis impairs horizontal gaze or swallowing. In congenital anomaly–associated cases, extensive correction is not always necessary, and gradual, well-monitored adjustments— rather than single, large corrections—should be favored to minimize risk. These general principles of case selection and safety are further supported by the present findings, which provide quantitative thresholds to guide realignment goals.
The present study establishes cutoff values for C0–2 angle change (10.65°) and C2–7 SVA change (19.2 mm) as predictors of favorable NDI outcomes. During craniocervical realignment surgery, the upper cervical spine is the principal segment amenable to direct manipulation, rendering C0–2 angle correction a critical determinant of surgical efficacy. The subsequent change in C2–7 SVA likely reflects a compensatory adjustment following C0–2 correction.
The procedure primarily entails reduction of CVJ kyphosis through craniocervical realignment and C1–2 facet remodeling techniques [1]. Adequate correction of the C0–2 angle requires meticulous execution of C1–2 facet release and head extension, which directly impact upper cervical alignment and overall sagittal balance.
The identified thresholds provide practical guidance in determining the degree of correction necessary to achieve satisfactory clinical outcomes. Targeting a C0–2 angle change of at least 10.65° may significantly increase the likelihood of favorable NDI improvement.
Although these cutoff values offer valuable reference points, anatomical variability and individual patient factors must be carefully considered. The thresholds should serve as general guidelines rather than absolute criteria, allowing intraoperative flexibility based on case-specific requirements.
Emphasizing C0–2 angle correction as the primary surgical objective may streamline the procedure, reduce operative time, and minimize the risks associated with extensive lower cervical manipulation. Focusing on the principal deformity segment can enhance surgical effectiveness while reducing the likelihood of complications.
IONM plays a critical role in CVJ realignment procedures due to the region’s complex neurovascular anatomy and the elevated risk of neurological injury [40]. IONM facilitates real-time assessment of spinal cord integrity, thereby improving surgical precision and patient safety. Furthermore, it contributes to preoperative positioning strategies by enabling baseline evaluations, particularly in cases requiring controlled neck flexion or rotation. When integrated into a surgical approach centered on C0–2 angle correction, IONM supports safe and effective realignment.
7. Comparison With Existing Literature
While numerous studies have focused on positive sagittal imbalance and lower cervical kyphosis [18,22-24], the present study addresses an important gap in the literature by investigating negative sagittal imbalance in the context of CVJ kyphosis. These findings are consistent with prior research emphasizing the significance of cervical sagittal balance, while also extending this understanding to CVJ deformities.
8. Limitations and Future Directions
This study has several inherent limitations that should be acknowledged. First, the sample size was relatively small, reflecting the rarity of craniocervical junction pathologies requiring reconstruction, and the retrospective, single-center design inevitably introduces selection bias. Second, the patient cohort was heterogeneous, including congenital anomalies, RA, and other etiologies, which may affect the generalizability of the results. Third, the minimum follow-up period was set at one year; while some patients had extended follow-up, longer-term evaluation will be necessary to fully assess radiographic fusion durability and late complications. In addition, all surgeries were performed by a single surgeon, which may limit reproducibility and external validity due to surgeon-specific factors such as technique, experience, and decision-making. Furthermore, the use of multiple linear regression in a relatively small sample carries a risk of overfitting. Logistic regression or automated selection procedures were not applied; instead, predictors were chosen based on clinical relevance and statistical significance in univariate analysis, which may reduce the robustness of the multivariate findings.
Despite these shortcomings, this study contributes valuable clinical information regarding surgical outcomes, complication profiles, and C2-related sequelae in an area where large-scale prospective data remain scarce. Prospective multicenter studies with larger cohorts and longer follow-up periods are warranted to validate our findings and better define the long-term clinical significance of craniocervical realignment surgery.
CONCLUSION
Craniocervical realignment surgery was associated with improvements in both radiological parameters and clinical outcomes in patients with CVJ kyphosis and negative cervical imbalance. Postoperative changes in C0–2 angle and C2–7 sagittal alignment may serve as prognostic factors for neurological and functional recovery. These findings may enhance the understanding of CVJ alignment and its compensatory mechanisms in the lower cervical spine, offering valuable guidance for clinical decision-making and surgical planning in this patient population.
Supplementary Materials
Supplementary Tables 1-2 are available athttps://doi.org/10.14245/ns.2550990.495.
Distribution of congenital anomalies
Intra- and interobserver reliability of radiological measurements
Notes
Conflict of Interest
The authors have nothing to disclose.
Funding/Support
This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author Contribution
Conceptualization: DHK, JTH, JYK, KBK, HJL, ISK; Data curation: DHK, JTH, JWH, DHL, ISK; Formal analysis: DHK, JTH, JYK, JWH, DHL, ISK; Methodology: DHK, JTH, JWH, HJL; Project administration: DHK, JTH, JYK, KBK, HJL, ISK; Visualization: DHK, JTH, JWH; Writing – original draft: DHK, KBK; Writing – review & editing: JTH, HJL, ISK.