Abstract
Objective
To evaluate the association between spinal canal elongation during deformity correction and distraction spinal cord injury (DSCI) in severe adolescent idiopathic scoliosis.
Methods
We retrospectively analyzed 79 patients with severe AIS (Major Cobb angle > 80°) who underwent posterior correction and had both preoperative and postoperative computed tomography scans available for three-dimensional reconstruction. Spinal canal length (SCL) was measured preoperatively and postoperatively at the concave, central, and convex aspects of the canal. The cross-sectional area of the spinal canal (CSA) was also recorded. Logistic regression was used to assess associations between spinal canal parameters and postoperative neurological injury. Receiver operating characteristic (ROC) analysis was performed to evaluate discriminative performance and to derive cohort-specific candidate cut points using the maximum Youden index.
Results
Among 79 included patients, 13 (16.5%) developed DSCI. The DSCI group showed a smaller CSA (151.3 ± 19.8 vs. 181.1 ± 40.7 mm2, p = 0.010) and greater SCL elongation at the concave side (10.1 ± 5.6 vs. 7.5 ± 3.6 mm, p = 0.008), central line (8.8 ± 3.9 vs. 5.0 ± 3.0 mm, p < 0.001), and convex side (7.6 ± 3.3 vs. 3.4 ± 2.7 mm, p < 0.001). Multivariate analysis identified convex SCL elongation (OR = 2.230, 95% CI 1.329–3.739, p = 0.002) and CSA (OR = 0.950, 95% CI 0.916–0.985, p = 0.006) as independent factors associated with DSCI. Convex SCL showed the highest discriminative ability on ROC analysis (AUC = 0.86).
Conclusion
Excessive convex-side SCL elongation combined with reduced CSA represents a high-risk morphological pattern for DSCI in severe AIS correction. Postoperative neurological risk is multifactorial—driven by complex three-dimensional corrective forces, osteotomies, and baseline anatomical reserve, rather than simple longitudinal distraction. Preoperative evaluation of canal reserve and anticipated elongation may provide a morphological framework for risk stratification in complex deformity surgery.
Keywords: Adolescent idiopathic scoliosis, Deformity correction, Spinal canal length, Distraction spinal cord injury
Introduction
Adolescent idiopathic scoliosis (AIS) is primarily considered an organic disorder of the spine rather than a systemic myopathic disease, as histochemical and physiological studies have shown no evidence of primary myopathic changes in the paravertebral muscles [1, 2]. However, the severe three-dimensional deformity and its subsequent surgical correction carry a significant risk of neurological injury due to spinal cord traction, compression, or ischemia during corrective maneuvers. Recent reports indicate that neurological complications occur in approximately 0.5–1% of scoliosis correction procedures despite modern surgical techniques and monitoring strategies [3, 4]. Intraoperative neurophysiological monitoring (IONM), including somatosensory-evoked potentials and motor-evoked potentials, is widely used during scoliosis surgery to detect early signs of spinal cord compromise and allow timely intraoperative intervention [5, 6]. However, IONM functions primarily as an intraoperative warning system for real-time detection of evolving spinal cord compromise and does not directly quantify the spinal cord’s preoperative tolerance to elongation during deformity correction [7]. Therefore, identifying morphologic parameters associated with traction-related neurological injury may improve our understanding of correction-related risk and provide a basis for future translational studies.
Spinal canal elongation during deformity correction reflects the traction imposed on the spinal cord as the spine is straightened [8]. Imaging studies have demonstrated that spinal canal length increases during scoliosis correction, indicating longitudinal stretching of the spinal cord [9, 10]. However, this elongation may not be uniformly distributed along the spinal canal, and asymmetric changes between the convex and concave sides may occur during deformity correction [11]. Such asymmetric elongation may lead to uneven mechanical stress on the spinal cord and potentially contribute to neurological injury. Despite these observations, the relationship between spinal canal elongation and postoperative neurological complications remains poorly characterized in clinical studies.
In our previous work, three-dimensional CT reconstruction was used to quantify spinal canal elongation, demonstrating that asymmetric changes in spinal canal length occurred during scoliosis correction [12, 13]. Building upon these findings, we hypothesized that greater spinal canal elongation may be associated with an increased risk of neurological injury following scoliosis correction. Therefore, the purpose of the present study was to evaluate the association between spinal canal elongation during scoliosis correction and distraction spinal cord injury (DSCI), with particular emphasis on whether convex-side canal elongation and reduced canal reserve jointly identify a high-risk morphologic pattern within this cohort.
Methods
Study design and patient selection
This retrospective single-center study was approved by the local ethics committee. Between January 2020 and December 2025, approximately 700 scoliosis correction surgeries were performed at our institution, including more than 320 patients with AIS. For the present study, we focused on severe AIS patients treated with posterior correction surgery who had complete clinical records and preoperative and postoperative CT data suitable for three-dimensional reconstruction of the spinal canal. Inclusion criteria were: (1) diagnosis of AIS; (2) severe scoliosis meeting the predefined radiographic criteria (Major Cobb angle > 80°); and (3) availability of complete preoperative and postoperative CT-based three-dimensional reconstruction data. Postoperative CT scans were not routinely performed for all AIS patients but were specifically indicated for those who developed postoperative neurological deficits or those with extremely rigid and severe deformities requiring complex osteotomies where routine radiographs were insufficient to assess postoperative spinal alignment and canal patency. Exclusion criteria were: (1) congenital, syndromic, or neuromuscular scoliosis; (2) previous spinal surgery; and (3) incomplete clinical or imaging data. After applying these criteria, 79 patients were included in the final analytic cohort. This cohort therefore represented a selected subgroup of severe AIS patients for imaging-based morphologic analysis rather than the overall AIS surgical population. At our institution, preoperative thin-slice CT is routinely indicated for severe and highly rigid deformities. This is a standardized protocol required to evaluate atypical pedicle morphology, rule out occult intraspinal anomalies, and plan complex posterior osteotomies prior to surgery.
DSCI was defined as any new postoperative neurological deficit involving motor (and/or sensory) function in the lower extremities, assessed using the American Spinal Injury Association (ASIA) Impairment Scale. Intraoperative neurophysiological monitoring (IONM) changes (signal amplitude reduction > 50%) were considered supportive evidence of spinal cord compromise. Postoperative deficits were defined as transient if significant recovery was observed within 6 months; otherwise, deficits persisting without meaningful improvement at the final follow-up were categorized as permanent [14].
This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Beijing Chao-Yang Hospital. Written informed consent was obtained from all individual participants or, in the case of adolescent participants, from their parents or legal guardians.
Surgical procedure and neuromonitoring
All patients underwent posterior spinal deformity correction with pedicle screw instrumentation performed by the same experienced spine surgery team. All patients underwent wide posterior releases, facetectomies, and multi-level osteotomies (e.g., Ponte osteotomies or asymmetric pedicle subtraction osteotomies) as indicated by curve rigidity. Deformity correction was achieved through a combination of rod derotation, translation, and in situ contouring, rather than isolated longitudinal distraction. Once an IONM alert occurred, corrective maneuvers were immediately suspended. To restore spinal cord perfusion and neural function, we released the corrective force (distraction/traction), elevated the mean arterial pressure (MAP), and administered high-dose corticosteroid pulse therapy.
Three-dimensional reconstruction
Thin-slice CT scans of the whole spine were obtained preoperatively and postoperatively with a slice thickness of approximately 1 mm. CT images were exported in DICOM format and imported into 3D Slicer (version 4.6.3) to reconstruct three-dimensional models of the spine. The reconstructed models (STL files) were subsequently exported to Rhinoceros (Robert McNeel and Associates, USA version 7.34) for geometric analysis and length measurements (Fig. 1a). To standardize measurements, the analyzed region of the spinal canal was defined between T2 and L2, corresponding to the typical instrumented region and the anatomical extent of the spinal cord.
Fig. 1.
a Illustration of spinal canal three-dimensional reconstruction and measurement procedures. A plane parallel to the superior endplate was defined using the line connecting the bilateral pedicles, and the vertebra was sectioned to expose the inner wall of the spinal canal. b The canal boundary was outlined with a smooth curve on the axial plane to calculate the cross-sectional area (CSA). Additional CSA measurements were obtained on planes 1 mm above and below this reference plane. c The centroid of the canal area was identified as the canal center, from which orthogonal lines were drawn to intersect the canal boundary, defining the concave and convex points. d This procedure was repeated across all levels, and the corresponding points were connected with smooth curves to reconstruct the concave and convex canal walls. The lengths of these curves were recorded as the concave and convex spinal canal lengths (SCL)
Measurement of spinal canal parameters
For each vertebral level, a cross-sectional plane parallel to the superior endplate and passing through the centers of both pedicles was generated. The boundary of the spinal canal was traced as CSA. To ensure measurement reliability, the CSA for each segment was calculated as the average of three consecutive axial slices at 1 mm intervals, centered at the mid-pedicle level (Fig. 1b). The centroid of the canal area was defined as the central reference point. Based on anatomical landmarks, reference points were defined at the central, concave, and convex aspects of the canal (Fig. 1c). Corresponding points across vertebral levels were connected using smooth three-dimensional spline curves to reconstruct the canal trajectory (Fig. 1d). The lengths of these curves were measured to obtain concave, central, and convex SCL. Measurements were performed on both preoperative and postoperative models. Spinal canal elongation was calculated as: ΔSCL = postoperative SCL − preoperative SCL.
To minimize errors in measurements, all measurements were independently performed by two observers who were blinded to the clinical outcomes. Each measurement was repeated twice, and the mean value was used for statistical analysis [15, 16].
Statistical analysis
Statistical analyses were performed using SPSS (version 22.0). Continuous variables were compared using the t-test or Mann–Whitney U test, and categorical variables using the chi-square or Fisher’s exact test. Intra- and inter-observer reliability for the morphological measurements (SCL and CSA) were assessed using the intraclass correlation coefficient (ICC) based on a two-way mixed-effects model for absolute agreement. An ICC value greater than 0.80 was considered to indicate excellent reliability. Logistic regression was used to identify factors associated with DSCI. ROC analysis was used to evaluate discriminative performance and to derive cohort-specific candidate cut points based on the maximum Youden index. Patients were further stratified into four exploratory groups according to the cohort-derived candidate cut points of convex SCL and CSA. Group differences were initially assessed using Pearson’s chi-square test, and Fisher’s exact test was additionally performed because of the small expected cell counts in the four-group comparison. A two-sided p < 0.05 was considered significant.
Results
During the study period, more than 320 patients underwent surgical treatment for AIS at our institution, among whom 13 developed distraction spinal cord injury (DSCI), corresponding to an overall institutional rate of approximately 4.1%. For the present imaging-based analysis, 79 patients with severe AIS who met the predefined eligibility criteria and had complete three-dimensional reconstruction data were included in the final analytic cohort. Within this cohort, 13 patients had DSCI and 66 did not; therefore, the 16.5% event proportion reflects the characteristics of this selected severe AIS subgroup rather than the overall incidence of DSCI in the broader AIS surgical population. Patients’ demographic and surgical characteristics were comparable between the two groups (Table 1). Reliability analysis demonstrated excellent agreement for the three-dimensional measurements. The intra-observer ICCs for all spinal canal parameters (including CSA, and concave, central, and convex SCL) ranged from 0.902 to 0.965, and the inter-observer ICCs ranged from 0.890 to 0.953, indicating high reproducibility of the measurement protocol. Patients with DSCI had a significantly smaller spinal canal cross-sectional area (CSA) compared with those without neurological injury (151.3 ± 19.8 vs. 181.1 ± 40.7 mm2, p = 0.010). In addition, SCL elongation was significantly greater in the DSCI group, including the concave side (10.1 ± 5.6 vs. 7.5 ± 3.6 mm, p = 0.008), central line (8.8 ± 3.9 vs. 5.0 ± 3.0 mm, p < 0.001), and convex side (7.6 ± 3.3 vs. 3.4 ± 2.7 mm, p < 0.001).
Table 1.
Patients’ demographic and surgical characteristics
| No neurological injury n = 66 |
DSCI n = 13 |
P value | |
|---|---|---|---|
| Age | 18.8 ± 4.5 | 20.3 ± 4.0 | 0.140 |
| Height | 155.0 ± 11.1 | 153.6 ± 7.9 | 0.347 |
| Weight | 47.2 ± 8.7 | 52.1 ± 12.9 | 0.217 |
| Pre-Cobb* | 99.9 ± 21.0 | 113.1 ± 26.4 | 0.116 |
| Pre-kyphosis* | 57.2 ± 26.5 | 71.4 ± 32.1 | 0.104 |
| Correction rate (%) | 64.2 ± 10.9 | 61.8 ± 7.9 | 0.466 |
| CSA (mm2) | 181.1 ± 40.7 | 151.3 ± 19.8 | 0.010 |
| △Concave (mm) | 7.5 ± 3.6 | 10.1 ± 5.6 | 0.008 |
| △Central (mm) | 5.0 ± 3.0 | 8.8 ± 3.9 | < 0.001 |
| △Convex (mm) | 3.4 ± 2.7 | 7.6 ± 3.3 | < 0.001 |
In univariate analysis, CSA was inversely associated with DSCI (OR = 0.97, 95% CI 0.94–0.99, p = 0.011), whereas greater SCL elongation at all measured regions was significantly associated with higher odds of DSCI (all p < 0.05) (Table 2). In multivariable analysis, convex SCL (OR = 2.230, 95% CI 1.329–3.739, p = 0.002) and CSA (OR = 0.950, 95% CI 0.916–0.985, p = 0.006) remained independently associated with DSCI (Table 3).
Table 2.
Univariate logistic regression analysis of spinal canal parameters associated with distraction spinal cord injury (DSCI)
| Odds ratio (OR) | 95% CI | P value | |
|---|---|---|---|
| CSA | 0.97 | 0.94–0.99 | 0.011 |
| △Concave SCL | 1.26 | 1.05–1.52 | 0.015 |
| △Central SCL | 1.44 | 1.15–1.81 | 0.002 |
| △Convex SCL | 1.72 | 1.25–2.37 | 0.001 |
Table 3.
Multivariate logistic regression analysis of factors associated with DSCI
| Variable | B | SE | OR | 95% CI | P value |
|---|---|---|---|---|---|
| Convex SCL | 0.802 | 0.264 | 2.230 | 1.329–3.739 | 0.002 |
| CSA | − 0.051 | 0.018 | 0.950 | 0.916–0.985 | 0.006 |
ROC analysis demonstrated that convex SCL had the highest discriminative ability for distinguishing patients with and without DSCI within this cohort (AUC = 0.862), followed by central SCL (AUC = 0.790), CSA (AUC = 0.735), and concave SCL (AUC = 0.711) (Table 4, Fig. 2). The cohort-derived candidate cut points were 5.6 mm for convex SCL and 173.7 mm2 for CSA. DeLong testing confirmed that convex SCL performed significantly better than concave SCL (p = 0.049), while no significant differences were observed between other pairs. Based on these cohort-derived cut points, patients were stratified into four exploratory groups according to convex SCL and CSA. The event proportion of DSCI was markedly higher in patients with both high convex SCL and low CSA (68.8%) than in the other three groups (0–4%) (Table 5, Fig. 3). Both Pearson’s chi-square test (χ2 = 39.982, p < 0.001) and Fisher’s exact test (p < 0.001) showed significant differences among the four groups, suggesting that DSCI events clustered in patients with greater canal elongation and reduced canal reserve.
Table 4.
Receiver operating characteristic (ROC) analysis of spinal canal parameters for distinguishing patients with and without distraction spinal cord injury (DSCI)
| Variable | AUC | 95% CI | P value | Cutoff | Sensitivity (%) | Specificity (%) |
|---|---|---|---|---|---|---|
| CSA mm2 | 0.735 | 0.61–0.86 | 0.008 | 173.7 | 92.3 | 51.5 |
| Concave SCL mm | 0.711 | 0.56–0.87 | 0.016 | 9.5 | 69.2 | 71.2 |
| Central SCL mm | 0.790 | 0.67–0.91 | 0.001 | 6.1 | 84.6 | 68.2 |
| Convex SCL mm | 0.862 | 0.77–0.95 | < 0.001 | 5.6 | 84.6 | 81.8 |
Candidate cut points were derived using the maximum Youden index. Because CSA showed an inverse association with DSCI (OR < 1), lower CSA values were associated with higher risk
Fig. 2.
Receiver operating characteristic (ROC) curves of CSA and concave, central, and convex spinal canal length (SCL) for distinguishing patients with and without distraction spinal cord injury (DSCI)
Table 5.
Exploratory risk stratification of DSCI based on combined convex SCL and CSA categories
| Group | Definition | Non-DSCI | DSCI | Event proportion (%) |
|---|---|---|---|---|
| 0 | Low Convex + High CSA | 24 | 1 | 4 |
| 1 | Low Convex + Low CSA | 29 | 1 | 3.3 |
| 2 | High Convex + High CSA | 8 | 0 | 0 |
| 3 | High Convex + Low CSA | 5 | 11 | 68.8 |
Convex SCL was dichotomized using the cohort-derived exploratory cut point from ROC analysis. Group differences were significant by both Pearson’s chi-square test (χ2 = 39.982, p < 0.001) and Fisher’s exact test (p < 0.001). The cohort-derived candidate cut points were 173.7 mm2 for CSA and 5.6 mm for convex SCL
Fig. 3.
Event proportion of DSCI according to combined convex SCL and CSA categories. Patients were stratified into four groups based on cohort-derived exploratory cut points from ROC analysis (convex SCL: 5.6 mm; CSA: 173.7 mm2). The event proportion of DSCI was significantly higher in patients with high convex SCL and low CSA (p < 0.001)
Detailed clinical characteristics of patients with postoperative DSCI are presented in Table 6. Most patients exhibited intraoperative neuromonitoring changes, and neurological outcomes varied at final follow-up.
Table 6.
Clinical characteristics and neurological outcomes of patients with DSCI
| Case | Age | Cobb angle | Convex SCL (mm) | CSA (mm2) | Intraoperative IONM change | Postoperative ASIA Scale | Final follow-up (yrs) | Final ASIA Scale |
|---|---|---|---|---|---|---|---|---|
| 1 | 25 | 115.5 | 3.8 | 175.3 | MEPs ≥ 50% decrease | B | 1 | C |
| 2 | 23 | 101.5 | 5.7 | 118.6 | MEPs ≥ 50% decrease | B | 1 | D |
| 3 | 17 | 80.5 | 15.2 | 163.1 | Complete loss of MEPs | A | 2 | A |
| 4 | 18 | 149.4 | 6.6 | 161.1 | MEPs ≥ 50% decrease | B | 2 | B |
| 5 | 24 | 83.5 | 12.6 | 145.4 | Complete loss of MEPs | A | 2 | B |
| 6 | 22 | 123.8 | 6.3 | 167.3 | Complete loss of MEPs | A | 2 | A |
| 7 | 22 | 158.2 | 6.0 | 136.4 | MEPs ≥ 50% decrease | C | 3 | D |
| 8 | 23 | 108.2 | 7.4 | 158.4 | Complete loss of MEPs | A | 3 | C |
| 9 | 14 | 104.7 | 4.9 | 167.5 | MEPs ≥ 50% decrease | C | 3 | D |
| 10 | 13 | 81.4 | 9.9 | 168.5 | MEPs ≥ 50% decrease | C | 3 | D |
| 11 | 23 | 101.4 | 6.2 | 119.6 | Complete loss of MEPs | A | 3 | A |
| 12 | 19 | 95.5 | 6.1 | 134.1 | Complete loss of MEPs | A | 3 | A |
| 13 | 25 | 165 | 8.8 | 132.4 | Complete loss of MEPs | A | 3 | C |
Discussion
Scoliosis is fundamentally a complex three-dimensional deformity characterized by lateral deviation and generalized generalized torsion [17]. The management of spinal deformities carries inherent risks that vary by technique and patient population. In early-onset scoliosis, distraction-based techniques such as magnetically controlled growing rods are effective but commonly associated with implant-related or mechanical complications [18]. In contrast, acute corrective surgery for severe AIS poses a distinct risk for DSCI due to acute morphologic changes. In this study, we examined the relationship between spinal canal morphological changes during deformity correction and DSCI. Specifically, we evaluated longitudinal canal elongation using SCL and canal capacity using cross-sectional area (CSA) to explore their association with DSCI.
DSCI is widely considered to be a traction-related form of spinal cord injury. Experimental studies have demonstrated that excessive longitudinal distraction can impair spinal cord perfusion and increase intramedullary pressure, ultimately resulting in neurological dysfunction [19, 20]. Busch et al. [21] reported that spinal distraction significantly decreased spinal cord blood flow at the distraction site by approximately 40%, supporting ischemia as a key mechanism of distraction-related spinal cord injury. Yang et al. [8] showed that vertebral column distraction of about 20 mm (≈3.6% of thoracolumbar spinal length) was sufficient to induce spinal cord injury and histological axonal damage in a porcine model. Although experimental studies have established that exceeding the biomechanical tolerance of the spinal cord can lead to irreversible neural injury, the degree of longitudinal strain that can be tolerated by the human spinal cord remains difficult to determine in clinical practice.
Against this background, our findings provide clinical support for the relevance of traction-related mechanisms in DSCI. In the present cohort, the extent of convex-side SCL elongation in the injury group was significantly higher than the elongation typically observed in asymptomatic mild-to-moderate scoliosis correction. During three-dimensional deformity correction, the shorter concave SCL may exhibit greater tolerance to stretching, whereas the inherently longer convex-side canal is subjected to peak tensile stress. This asymmetric longitudinal strain, combined with decreased local perfusion, exacerbates the vulnerability of the spinal cord [12, 13].
Reduced CSA may reflect diminished canal reserve rather than a simple morphologic variation. Prior deformity studies have shown that spinal cord morphology and the surrounding cerebrospinal fluid space are clinically relevant to neurological risk during correction, suggesting that the reserve space around the cord may influence its tolerance to operative stress [22, 23]. A narrowed canal reduces this reserve and may indicate a relatively pre-compressed cord with less buffering capacity. In addition, compression may impair local cerebrospinal fluid dynamics and perfusion [24], while distraction itself can further decrease spinal cord blood flow [25]. Together, these factors may render the cord less tolerant to elongation, which may explain why the highest neurological risk in our cohort was observed in patients with both greater convex-side SCL elongation and lower CSA.
It is important to emphasize that neurological compromise during severe AIS correction is rarely the result of simple longitudinal distraction. It emerges from a complex interplay of preexisting anatomical constraints (such as reduced CSA), altered biomechanics from wide osteotomies, and multi-planar corrective forces including derotation and translation. While convex SCL elongation serves as a quantifiable morphological surrogate for cord stretching, the ultimate risk is determined by the cumulative mechanical and ischemic burden on the spinal cord during these multi-step maneuvers. Previous biomechanical studies have demonstrated that strategically modifying instrumentation patterns, such as altering the density of sublaminar wiring, can maintain construct stability under torsional strain while actively reducing the surgical risk of direct cord damage [26]. Similarly, in modern screw-based corrections, recognizing high-risk morphological patterns preoperatively may help surgeons tailor their osteotomy and distraction strategies to avoid exceeding the biomechanical tolerance of the spinal cord.
Beyond its mechanistic implications, this finding may provide a preliminary morphologic framework for future translational research of correction-related risk. Although convex-side SCL elongation was quantified from postoperative reconstruction, it reflects the net longitudinal distraction imposed during correction and may therefore serve as a preliminary morphologic reference for future preoperative three-dimensional planning and simulation, rather than a validated safety threshold [27]. Accordingly, the 5.6 mm value identified in this cohort should be interpreted as an exploratory reference derived from retrospective analysis, rather than as a validated planning threshold or safety limit. Future studies are needed to determine whether similar morphologic information can be incorporated into preoperative simulation models [28].
This study has several limitations. First, as a retrospective single-center study with a limited number of DSCI events, this analysis may be subject to selection bias and limited statistical stability. Second, the analytic cohort was restricted to a specific subpopulation: severe, highly rigid AIS patients requiring preoperative CT for complex surgical planning. Consequently, this study carries an inherent selection bias. The high DSCI event proportion reflects the extreme end of the deformity spectrum and must not be extrapolated as the overall incidence of neurological complications in general AIS practice. Third, because convex-side SCL elongation was quantified from postoperative reconstruction, the 5.6 mm value should be regarded as an exploratory cohort-specific reference derived from retrospective analysis rather than a planning threshold or safety limit. Fourth, regarding the statistical methodology, the low absolute number of DSCI events limits the power of our multivariable logistic regression. Specifically, the model yields an Events Per Variable (EPV) ratio of 6.5, which falls below the traditional threshold of 10, posing a risk of statistical overfitting. To mitigate this, we strictly restricted our multivariable model to the two most robust clinical predictors. Given this limited statistical stability, the identified cut-offs must be strictly interpreted as exploratory, hypothesis-generating references for this specific cohort, rather than validated predictive safety thresholds.
Conclusion
Excessive convex-side SCL elongation combined with reduced CSA represents a high-risk morphological pattern for DSCI in severe AIS correction. Postoperative neurological risk is multifactorial—driven by complex three-dimensional corrective forces, osteotomies, and baseline anatomical reserve, rather than simple longitudinal distraction. Preoperative evaluation of canal reserve and anticipated elongation may provide a morphological framework for risk stratification in complex deformity surgery.
Acknowledgements
None.
Funding
The authors did not receive support from any organization for the submitted work. No funding was received to assist with the preparation of this manuscript.
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Beijing Chaoyang Hospital. Written informed consent was obtained from all individual participants or, in the case of adolescent participants, from their parents or legal guardians.
Consent for publication
Written informed consent was obtained from the patients’ legal guardians for publication of this study and any accompanying images.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Chaofan Han and Honghao Yang have contributed equally to this work.
References
- 1.Maffulli N. Histochemical and physiological studies in idiopathic scoliosis. Ital J Orthop Traumatol. 1990;16(1):61–71. [PubMed] [Google Scholar]
- 2.Gargano G, Oliva F, Migliorini F, Maffulli N. Melatonin and adolescent idiopathic scoliosis: the present evidence. Surgeon. 2022;20(6):e315–21. [DOI] [PubMed] [Google Scholar]
- 3.Qiu Y, Wang S, Wang B, Yu Y, Zhu F, Zhu Z. Incidence and risk factors of neurological deficits of surgical correction for scoliosis: analysis of 1373 cases at one Chinese institution. Spine. 2008;33(5):519–26. [DOI] [PubMed] [Google Scholar]
- 4.Li H, Gao S, Li R, et al. Identifying Intraoperative spinal cord injury location from somatosensory evoked potentials’ time-frequency components. Bioengineering (Basel). 2023. 10.3390/bioengineering10060707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Angelliaume A, Alhada TL, Parent HF, Royer J, Harper L. Intraoperative neurophysiological monitoring in scoliosis surgery: literature review of the last 10 years. Eur Spine J. 2023;32(9):3072–6. [DOI] [PubMed] [Google Scholar]
- 6.Zanin L, Broglio L, Panciani PP, et al. Intraoperative neurophysiological monitoring in contemporary spinal surgery: a systematic review of clinical outcomes and cost-effectiveness. Brain Sci. 2025. 10.3390/brainsci15070768. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Laratta JL, Ha A, Shillingford JN, et al. Neuromonitoring in spinal deformity surgery: a multimodality approach. Global Spine J. 2018;8(1):68–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yang JH, Suh SW, Modi HN, et al. Effects of vertebral column distraction on transcranial electrical stimulation-motor evoked potential and histology of the spinal cord in a porcine model. J Bone Joint Surg Am. 2013;95(9):835–42. [DOI] [PubMed] [Google Scholar]
- 9.Yahara Y, Seki S, Makino H, et al. Three-dimensional computed tomography analysis of spinal canal length increase after surgery for adolescent idiopathic scoliosis: a multicenter study. J Bone Joint Surg Am. 2019;101(1):48–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Glaser DA, Doan J, Newton PO. Comparison of 3-dimensional spinal reconstruction accuracy: biplanar radiographs with EOS versus computed tomography. Spine. 2012;37(16):1391–7. [DOI] [PubMed] [Google Scholar]
- 11.de Reuver S, de Block N, Brink RC, et al. Convex-concave and anterior-posterior spinal length discrepancies in adolescent idiopathic scoliosis with major right thoracic curves versus matched controls. Spine Deform. 2023;11(1):87–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Han C, Hai Y, Yin P, Cha T, Li G. In vivo deformation of the spine canal before and after surgical corrections of severe and rigid kyphoscoliosis. J Orthop Translat. 2020;23:1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Han C, Hai Y, Zhou C, et al. Investigation of in vivo three-dimensional changes of the spinal canal after corrective surgeries of the idiopathic scoliosis. JOR Spine. 2021;4(3):e1151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Li J, Hu Z, Qian Z, et al. The prognosis and recovery of major postoperative neurological deficits after corrective surgery for scoliosis: an analysis of 65 cases at a single institution. Bone Joint J. 2022;104(1):103–11. [DOI] [PubMed] [Google Scholar]
- 15.Capasso G, Testa V, Maffulli N. Variability in Cobb angle measurements. J Bone Joint Surg Br. 1996;78(2):339. [PubMed] [Google Scholar]
- 16.Capasso G, Maffulli N, Testa V. The validity and reliability of measurements in spinal deformities: a critical appraisal. Acta Orthop Belg. 1992;58(2):126–35. [PubMed] [Google Scholar]
- 17.Rahmatalla A, Chockalingam N, Dangerfield P, et al. Movement analysis of scoliotic subjects using Fastrak. Stud Health Technol Inform. 2002;91:162–6. [PubMed] [Google Scholar]
- 18.Migliorini F, Chiu WO, Scrofani R, et al. Magnetically controlled growing rods in the management of early onset scoliosis: a systematic review. J Orthop Surg Res. 2022;17(1):309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hu X, Xu W, Ren Y, et al. Spinal cord injury: molecular mechanisms and therapeutic interventions. Signal Transduct Target Ther. 2023;8(1):245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Han B, Liang W, Hai Y, et al. Elucidating the potential mechanisms underlying distraction spinal cord injury-associated neuroinflammation and apoptosis. Front Cell Dev Biol. 2022;10:839313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Busch DR, Lin W, Cai C, et al. Multi-site optical monitoring of spinal cord ischemia during spine distraction. J Neurotrauma. 2020;37(18):2014–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Meves R, Avanzi O. Correlation between neurological deficit and spinal canal compromise in 198 patients with thoracolumbar and lumbar fractures. Spine (Phila Pa 1976). 1976;30(7):787–91. [DOI] [PubMed] [Google Scholar]
- 23.Fehlings MG, Kato S, Lenke LG, et al. Incidence and risk factors of postoperative neurologic decline after complex adult spinal deformity surgery: results of the Scoli-RISK-1 study. Spine J. 2018;18(10):1733–40. [DOI] [PubMed] [Google Scholar]
- 24.Bae YJ, Lee JW, Lee E, Yeom JS, Kim KJ, Kang HS. Cervical compressive myelopathy: flow analysis of cerebrospinal fluid using phase-contrast magnetic resonance imaging. Eur Spine J. 2017;26(1):40–8. [DOI] [PubMed] [Google Scholar]
- 25.Dolan EJ, Transfeldt EE, Tator CH, Simmons EH, Hughes KF. The effect of spinal distraction on regional spinal cord blood flow in cats. J Neurosurg. 1980;53(6):756–64. [DOI] [PubMed] [Google Scholar]
- 26.Gadgil A, Ahmed EB, Rahmatalla A, Dove J, Maffulli N. A study of the mechanical stability of scoliosis constructs using variable numbers of sublaminar wires. Eur Spine J. 2002;11(4):321–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Majdouline Y, Aubin CE, Sangole A, Labelle H. Computer simulation for the optimization of instrumentation strategies in adolescent idiopathic scoliosis. Med Biol Eng Comput. 2009;47(11):1143–54. [DOI] [PubMed] [Google Scholar]
- 28.Cho W, Job AV, Chen J, Baek JH. A review of current clinical applications of three-dimensional printing in spine surgery. Asian Spine J. 2018;12(1):171–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.