Objective:To investigate the risk factors for severe distal curve progression after posterior hemivertebra (HV) resection and short-segment fixation in patients with congenital cervicothoracic scoliosis (CTS), and to analyze the surgical revision strategy.Methods:Imaging and clinical data of patients who underwent posterior HV resection and short-segment fixation for CTS between August 2012 and August 2021 at Nanjing Drum Tower Hospital were retrospectively analyzed. A total of 55 patients were recruited, including 27 females and 28 males with an average age of 8.5±3.6 years (range 3-15 years) at surgery and an average Risser grade of 0.7±1.4 (range 0-4). The number of fused segments averaged 6.9±1.6 (range 4-10), and the mean follow-up was 38.7±18.9 months (range 9-94 months). According to the severity of distal curve progression, the recruited patients were divided into three groups: non-progression group (NPG), mild progression group (MPG), and severe progression group (SPG). The latter two groups were collectively called the progression group (PG). The cervicothoracic Cobb angle, T1 tilt angle, coronal balance distance (CBD), neck tilt angle, clavicular angle, head tilt angle, head shift, and upper (UIV) and lower instrument vertebra (LIV) tilt angle on the standing whole spine X-ray were measured before and after surgery and at the last follow-up. The correction rate of the Cobb angle in the osteotomy area was measured and calculated on CT three-dimensional reconstruction, and the proportion of patients with Klippel-Feil syndrome (KFS) was recorded. Statistical analysis was conducted on the various parameters between the two groups. For factors with statistical significance in the single-factor analysis, binary logistic regression analysis was performed to identify the high-risk factors for distal curve progression.Results:There were 38 cases in the NPG, 11 in the MPG, and 6 in the SPG. Compared to the NPG, the PG showed more severe coronal imbalance preoperatively, with CBD of 35.6±22.3 mm and 11.6±7.1 mm respectively; more severe neck tilt and head shift, with neck tilt angle of 17.4°±8.3° and 12.4°±6.9° respectively, and head shift of 22.8±17.7 mm and 13.9±9.8 mm respectively; and a higher proportion of KFS, 65% (11/17) and 34% (13/38) respectively, all with statistical significance ( P<0.05). Postoperatively, the PG showed more severe coronal imbalance compared with the NPG, with 17.3±12.7 mm and 9.6±8.1 mm respectively; more evident residual deformity, with cervical tilt angles of 9.4°±4.6° and 6.4°±5.3° respectively, and head shift of 14.7±7.4 mm and 9.1±5.9 mm respectively; lower correction of Cobb angle in the apical osteotomy region, with rates of 40.1%±15.2% and 50.3%±19.9% respectively; more significant UIV and LIV tilt, with UIV tilt angles of 14.3°±7.4° and 9.8°±5.3° respectively, and LIV tilt angles of 8.1°±5.5° and 4.5°±3.6° respectively, all with statistical significance ( P<0.05). SPG showed only more severe coronal imbalance preoperatively compared with the MPG, with 50.7±31.3 mm and 27.3±9.6 mm respectively; and head shift, with 33.5±25.0 mm and 16.9±11.0 mm respectively, all with statistical significance ( P<0.05). Logistic regression analysis demonstrated a significant correlation between significant preoperative coronal imbalance and postoperative distal scoliosis progression [ OR=1.299, 95% CI (1.101, 1.531), P=0.002]. Five cases (83.3%) in SPG underwent revision surgery with an average follow-up of 25 months, and selecting the LIV down to the stable region was the major revision strategy. Conclusion:Combined KFS, residual cervicothoracic deformities, and tilting of UIV and LIV are key causes, whereas significant preoperative coronal imbalance is an independent risk factor predisposing to the distal curve progression.

Objective:To compare the clinical outcomes between three-column osteotomy and posterior-column osteotomy for correcting dystrophic kyphoscoliosis secondary to neurofibromatosis type 1 (DKS-NF1).Methods:ALL of 84 patients with DKS-NF1 were retrospectively analyzed, and the average age was 17.7±6.9 years. There were 50 cases with single curve, 18 cases with double curves, and 16 cases with triple curves; kyphosis was found in 42 cases in the thoracic area, 31 cases in the thoracolumbar area, and 11 cases in the lumbar area. The patients were divided into two groups: posterior column osteotomy group and three column osteotomy group based on surgical strategy. The radiographic parameters (including the magnitude of kyphosis, scoliosis, coronal balance distance, etc.) were compared between the two groups before and after surgery, and during the follow-up. The surgical efficacy was also compared based on the spinal correction and complications (such as cerebrospinal fluid leakage, pneumothorax, rod breakage, etc.).Results:The posterior column osteotomy group consisted of 74 patients and the column osteotomy group consisted of 10 patients. The age of patients in the posterior column osteotomy group was significantly younger than that in the three-column osteotomy group (15.8±4.8 years vs. 29.4±10.2 years, t=7.088, P<0.001), and the proportion of preoperative traction in this group was significantly higher than that in the three column osteotomy group (26/74 vs. 0, P=0.027). The apex of kyphosis in the three-column osteotomy group mainly located in the thoracolumbar and lumbar area, significantly higher than that in the posterior column osteotomy group (10/10 vs. 32/74, P=0.001). The magnitude of kyphosis in the two groups were 73.8°±20.9° and 63.1°±21.4° before surgery, respectively ( t=1.506, P=0.136). After surgery, they were corrected to 43.1°±20.9° and 21.1°±22.8°, respectively ( t=3.066, P=0.003), with correction rates of 43.7% ±19.6% and 84.1% ±78.7%, respectively ( t=3.677, P<0.001). At the last follow-up, they were maintained at 46.5°±20.9° and 24.6°±25.5°, respectively ( t=3.016, P=0.003). The Cobb angle of the main curve was corrected from preoperative 83.0°±29.0° and 66.3°±17.7° ( t=1.766, P=0.081) to postoperative 50.6°±20.8° and 40.8°±15.6° ( t=1.436, P=0.155), with correction rates of 38.3% ±16.6% and 39.3% ±12.7% ( t=0.191, P=0.849), respectively. At the last follow-up, they were maintained at 52.3°±20.5° and 43.1°±18.2°, respectively ( t=1.339, P=0.185). The proportion of multi-rod system application and screw density in three column osteotomy group was significantly higher than that in posterior column osteotomy group (8/10 vs. 20/74, P=0.002; 72.0% ±11.3% vs. 61.4% ±14.6%, t=2.173, P=0.033). The incidence of complications in the two groups was 12.2% (posterior column osteotomy group, 9/74) and 20% (three column osteotomy group, 2/10), respectively, with no statistically significant difference ( P=0.613). Conclusion:Three-column osteotomy is mainly used to treat adult kyphosis in DKS-NF1 patients. While the posterior column osteotomy methods were mainly applied in young patients. Most patients can achieve the purpose of deformity correction by posterior column osteotomy alone or combined with anterior complementary fusion. For patients with severe kyphosis, preoperative Halo gravity traction can help to further correct the intraoperative deformities.

Objectives:To propose a coronal imaging classification of neurofibromatosis type Ⅰ with dys-trophic upper thoracic scoliosis(DUTS-NF1),to verify its reliability and reproducibility,and to explore its clinical significance.Methods:We retrospectively analyzed the data of patients diagnosed with neurofibromato-sis type Ⅰ with dystrophic spinal scoliosis in the database of our hospital from June 2009 to December 2023.The patients whose main curve apex located in the upper thoracic vertebrae(T1-T5)were screen out and were classified according to the coronal scoliosis pattern on standing spinal X-rays:Type A,shoulder-neck type;Type B,distal thoracic compensatory curve type;Type C,trunk inclination type.The Cobb angle,upper thoracic kyphosis angle,clavicle angle(CA),T1 tilt,neck tilt(NT),head shift(HS),coronal balanced dis-tance(CBD),and deformity angular ratio(DAR)were measured and calculated for each type of patients.Three spine surgeons,who had been specially trained,independently performed two classifications according to the method.The results of the two classifications of the same observer were analyzed for reproducibility by apply-ing the Kappa value,and the results of the classification were analyzed for confidence between different ob-servers.Results:29 patients(7.9％)with DUTS-NF1 were identified from the 367 NF1 patients with dystrophic scoliosis,and the main Cobb angle was 78.7°±12.9°.The classification results of the designer were 16 patients of type A,8 of type B,and 5 of type C.The DAR of type B and type C was significantly higher than that of type A(20.6±2.2,20.0±3.0 vs 13.2±1.8,P＜0.001);Type C had more significant head and coronal shift distances than type A(HS:27.6±11.7mm vs 13.5±6.7mm,P＜0.001;CBD:34.8±20.5mm vs 13.9±10.9mm,P＜0.001);The T1 tilt of type C was significantly higher than that of type A(P＜0.05);The other imaging parameters were not statistically different between the three types(P＞0.05).174 times of classifications were performed by the 3 observers using the DUTS-NF1 coronal imaging classification method,including 96 classifications of type A,45 classifications of type B,and 33 classifications of type C.The intra-observer classification concordance rate was(82.57±8.44)％,with a Kappa value of 0.771-0.81,which was considered to be"Basically reliable";The inter-observer classification concordance rate was(84.19±8.65)％,with a Kappa value of 0.884-0.886,which was considered to be"Completely reliable".Conclusions:Based on the coronal imaging features,patients with DUTS-NF1 can be classified into three types of shoulder-neck type,distal thoracic compensatory curve type,and trunk inclination type.This classification system has satisfied reproducibility and reliability,which is helpful for assessing the coronal compensation of DUTS-NF1 patients,and thus guides the clinical diagnosis and treatment.

Objective:To analyze the efficacy of posterior staged correction in the treatment of severe kyphoscoliosis.Methods:Retrospective analysis was conducted on 61 patients with severe kyphoscoliosis who underwent one-stage posterior Ponte osteotomy followed by Halo-femoral traction and two-stage posterior correction in Nanjing Drum Tower Hospital, the Affiliated Hospital of Nanjing University Medical School between January 2010 and January 2020. There were 23 males and 38 females with an average age of 22.0(18.0, 25.5) years. The etiologies were idiopathic in 26 cases, congenital in 17 cases, neuromuscular in 16 cases and Marfan syndrome with kyphoscoliosis in 2 cases. The curves were thoracic in 49 cases, thoracolumbar in 3 cases and double major in 9 cases. The apical vertebrae were T 5 level in 1 case, T 7 level in 2 cases, T 8 level in 9 cases, T 9 level in 15 cases, T 10 level in 23 cases, T 11 level in 8 cases, T 12 level in 1 case, and L 1 level in 2 cases. The flexibility of main curve was 13.5%±8.6%. The Cobb angle of main curve, global kyphosis (GK), coronal trunk shift (CTS), sagittal vertical axis (SVA), thoracic kyphosis (TK), lumbar lordosis (LL), pelvic incidence (PI), pelvic tilt (PT), and sacral slope (SS) were assessed at pre-operation, post-traction, post-operation and the last follow-up. The quality of life was evaluated using the MOS item short form health survey (SF)-36 questionnaire, and the complications during peri-operation and long-term follow-up were recorded. Results:All 61 patients were followed up for 25.0 (24.0, 26.5) months. The Cobb angle of main curve and GK were 121.4°±13.9° and 86.8°±20.0° at pre-operation, 94.1°±18.7° and 66.9°±15.3° at post-traction, 78.5°±20.3° and 54.7°±13.6° at post-operation and 79.5°±20.1° and 53.2°±11.3° at the last follow-up, respectively. The differences were statistically significant ( F=210.54, P<0.001; F=93.74, P<0.001). There was no significant difference between the last follow-up and post-operation ( P>0.05). There was no significant correction loss of SVA, TK, LL, PI, PT or SS at the last follow-up when compared with those at post-operation ( P>0.05). The CTS was 17.1±9.8 mm at pre-operation, 17.5±11.4 mm at post-operation, 11.1 (5.9, 23.3) mm at the last follow-up and the difference was statistically significant (χ 2=6.70, P=0.035). The difference between the last follow-up and post-operation was statistically significant ( P=0.032). The scores of physical functioning 80.0 (75.0, 85.0), general health 82.0 (69.5, 87.0) and social functioning 75.0 (62.5, 75.0) in SF-36 at the last follow-up were significantly improved compared with those at pre-operation ( Z=-2.11, P=0.035; Z=-3.64, P<0.001; Z=-2.07, P=0.039). During the traction process, the complications included pin track infection in 1 case, deep vein thrombosis of the lower extremity in 2, misplacement of pedicle screws in 3, coronal decompensation at immediate post-operation in 2, sagittal decompensation at immediate post-operation in 1, and 1 patient had broken rod at 3 years follow-up, respectively. Conclusion:The posterior staged correction could provide satisfying radiographic and clinical outcomes in patients with severe kyphoscoliosis, which could be well maintained during 2 years follow-up. Therefore, the posterior staged correction is a safe and effective treatment for severe kyphoscoliosis.

Objective:To evaluate whether pelvic fixation is needed in patients undergoing posterior lumbosacral hemivertebra (LSHV) resection and long fusion.Methods:All 32 adult spinal deformity patients with posterior hemivertebra (HV) resection and long segment fixation treated from April 2005 to August 2019 were analyzed retrospectively, including 12 males and 20 females with a mean age of 32.9±8.8 years. According to the state of coronal balance distance (CBD), there were 15 cases of type A (preoperative CBD≤ 30 mm), 1 case of type B (preoperative CBD>30 mm and C 7 plumb line offset to the concave side), and 16 cases of type C (preoperative CBD>30 mm and C 7 plumb line offset to the convex side). The clinical and imaging data before operation, immediately after operation and at the last follow-up were collected, and the short-term and long-term complications related to operation were recorded. The improvement of Cobb angle and coronal balance of primary curve and compensatory curve were evaluated on the whole spine frontal and lateral X-ray films, and the change of coronal balance type after operation was evaluated. According to the mode of distal internal fixation, the patients were divided into two groups: PF group (pelvic fixation): distal fixation to iliac or sacroiliac; NPF group (non-pelvic fixation): distal fixation to L 5 or S 1. Results:All 32 patients were followed up with an average time of 3.9±2.6 years (range 2-11 years). The Cobb angle of primary curve in PF and NPF groups were 42.6°±13.5° and 41.3°±10.9° respectively before operation, and corrected to 13.1°±5.4° and 17.7°±5.8° respectively after operation. It maintained at 13.4°±5.1°and 18.5°±6.7° in the two groups at the last follow-up, respectively ( FPF=32.58, FNPF=28.64, P<0.001). The correction rates were 69.3%±11.8% and 57.6%±10.3%, respectively ( t=2.14, P=0.012). The compensatory curves of in the two groups were corrected from 54.9°±14.8° and 46.8°±13.6° before operation to 17.3°±9.6° and 15.4°±8.4° after operation. It also maintained at 18.5°±8.8°and 17.6°±9.5° in the two groups at the last follow-up, respectively ( FPF=42.97, FNPF=38.56, P<0.001). The correction rates were 68.4%±16.7% and 67.2%±14.9%, respectively ( t=0.17, P=0.849) in the two groups. In PF group, the primary and compensatory curve were similar (69.3%±11.8% vs. 68.4%±16.7%, t=0.15, P=0.837), while the correction rate of compensatory curve in NPF group was significantly higher than that of the primary curve (67.2%±14.9% vs. 57.6%±10.3%, t=2.13, P=0.013). Coronal decompensation occurred in 12 patients (12/32, 37.5%). The CBD in PF and NPF groups was corrected from 33.3±11.2 mm and 28.8±8.1 mm preoperatively to 18.5±3.5 mm and 27.1±6.8 mm postoperatively, respectively, and it showed no significant change at the last follow-up ( FPF=41.61, P<0.001; FNPF=0.38, P=0.896). While the CBD in PF group was significantly better than that in NPF group ( t=3.23, P=0.002; t=2.94, P=0.008). The incidence of coronal decompensation in PF group was 0%, which was significantly lower than 50% (12/24) in NPF group (χ 2=6.40, P=0.014). In addition, 6 cases in PF group were type C coronal decompensation before operation, and the coronal balance was corrected to type A after surgery (100%). Among 10 patients with type C coronal decompensation in NFP, 4 (40%) patients returned to type A after operation, and the difference was statistically significant (6/6 vs. 4/10, χ 2=5.76, P=0.034). Conclusion:Coronal decompensation (12/32, 37.5%) is not rare in patients after posterior LSHV resection and long fusion. Attention should be paid to the match of the corrections between lumbosacral deformity and compensatory curve, which is of great significance in coronal balance reconstruction. Pelvic fixation is helpful to reduce the incidence of postoperative coronal decompensation, especially for the type C patients.

Objective:To investigate the risk factors of long-term shoulder imbalance in patients presented postoperative shoulder imbalance who underwent single segment hemivertebra resection, and the role of postoperative trunk shift in shoulder imbalance.Methods:All of 30 patients who presented shoulder imbalance after hemivertebrae resection and short fusion from July 2006 to December 2018 were reviewed in this study, including 16 males and 14 females, aged 4.53±2.05 years (range, 2-8 years). Among them, 10 cases were thoracic hemivertebra, 12 thoracolumbar hemivertebra and 8 lumbar hemivertebra. According to the vertical height difference at the highest point of soft tissue shadows on both shoulders in the final follow-up upright posteroanterior radiograph, which was shoulder imbalance (SI), they were divided into two groups: Group B (balance, shoulder imbalance less than 10 mm) and Group IB (imbalance, shoulder imbalance more than 10 mm). Several radiographic parameters were measured preoperatively, 3 months after surgery and at the final follow-up, such as SI, distance between C 7 plumbline and center sacral vertical line (C 7PL-CSVL), Cobb angle of main curve, cobb angle of proximal curve (CAPC), Cobb angle of distal curve (CADC), upper instrumented vertebra offset (UO), lower instrumented vertebra offset (LO), upper instrumented vertebra slope (US), lower instrumented vertebra slope (LS), T1 tilt and sagittal vertical axis (SVA). Results:The mean follow-up period was 54.3±33.7 months (range, 24-132 months). A mean of 3.1 segments were fused. 7 cases (70%) of thoracic, 6 cases (50%) of thoracolumbar and 3 cases of lumbar hemivertebrae (37.5%) with shoulder imbalance at 3 months after surgery remained imbalanced at the last follow-up. Thirteen cases presented coronal imbalance postoperative (C 7PL-CSVL>2 cm), among 6 cases whose trunk shafted to the side of the higher shoulder postoperatively, 5 cases presented aggravated SI at final follow-up, and among 7 cases whose trunk shafted to the side of the lower shoulder postoperatively, 6 cases presented aggravated SI at final follow-up, while the difference had statistical significance ( P=0.029). The static analysis indicated that postoperative and long-term C 7PL-CSVL, long-term lowest instrumented vertebra and long-term T 1 tilt were risk factors of shoulder imbalance at final follow-up. Conclusion:A proportion of congenital scoliosis patients who presented shoulder imbalance after hemivertebra resection plus short fusion are less likely to achieve shoulder balance at the final follow-up. Long-term shoulder imbalance is often presented in the patients whose trunk shafted to the side of the higher shoulder postoperatively.

Objective:To evaluate the radiographic and clinical outcomes of pre-operative Halo-gravity traction (HGT) and posterior correction surgery in treating patients with neurological deficits secondary to severe focal angular kyphosis of the upper thoracic spine.Methods:A total of 16 patients (11 males, 5 females) with neurologic deficits secondary to severe focal angular kyphosis of the upper thoracic spine undergoing preoperative HGT and posterior correction operation from January 2010 to December 2019 were retrospectively analyzed. The average age was 12.9±5.6 years (range 6-27 years). The standing X-ray of whole spine was taken at pre-, post-traction and post-operation. The Cobb angles of main curve at both sagittal and coronal planes were measured accordingly. The neurologic function at pre-traction, post-traction and post-operation was assessed according to the American Spinal Injury Association (ASIA) grading. The complications during HGT, operation and post-operative follow-up were recorded for each patient.Results:The average values of focal kyphosis and scoliosis were 96.1°±16.0° (71°-128°) and 75.5°±20.5° (40°-107°) at pre-traction respectively. The spinal cord function graded by ASIA criteria at pre-traction was B in 1 patient, C in 6 and D in 9, respectively. The correction rates of focal kyphosis and scoliosis were 32.8%±15.0% (18.0%-65.9%) and 22.9%±8.0% (14.1%-38.6%) after traction, which were further improved to 45.4%±14.9% (29.0%-69.0%) and 33.6%±8.6% (23.3%-49.3%) at post-operation without significant correction loss during 35.6±14.2 (24-72) months follow-up. After traction, the spinal cord function improved to grade D in 4 patients and grade E in 12 patients. At the last follow-up, the spinal cord functions were grade E in 15 patients and grade C in 1 patient. No neurologic monitor events occurred during operation. One patient suffered from transient left brachial plexus after operation. Further, proximal hook loosening was observed in 2 patients during follow-up. The spinal cord function was ASIA grade C pre-operatively in one patient, who recovered to ASIA grade E after operation and significantly deteriorated to ASIA C at 4 years follow-up.Conclusion:The correction of spinal kyphoscoliosis was satisfactory in this cohort. Preoperative HGT followed by posterior spinal correction surgery is an effective and safe procedure in treating neurological deficits secondary to focal angular kyphosis in the upper thoracic spine.

Objective:To explore the efficacy and correction mechanism of posterior column osteotomy for treatment of "Lenke 5-like" lumbar congenital scoliosis.Methods:From April 2008 to September 2019, 16 patients with lumbar congenital scoliosis underwent posterior column osteotomy were retrospectively reviewed including 6 males and 10 females, aged 23.9±11.7 years (range, 14-48 years). Among them, 8 cases were unsegmentation, 5 malformation and 3 mixed type. The average segments of PCO were 5.3. Posterior column osteotomy was adapted after the insertion of pedicle screws, then removed spinous process, ligaments, superior and inferior facet and corrected the deformity with the rods. The pre- and post-operative and last follow-up radiographic parameters were measured: Cobb angle of lumbar curve, distance between C 7 plumbline and center sacral vertical line (C7PL-CSVL), deformity angle (DA), disc correction angle (DCA) of instrumented segments, disc angle above upper instrumented vertebra (DAAU), lower instrumented vertebra disc angle (LDA), upper instrumented vertebra slope (US), lower instrumented vertebra slope (LS), lower instrumented vertebra offset (LO) and sagittal parameters such as thoracic kyphosis (TK), lumbar lordosis(LL) and thoracic junctional kyphosis (TJK). The Scoliosis Research Society-22 questionnaire (SRS-22) were conducted at preoperation and the final follow up to evaluate the clinical outcomes. Results:The mean follow-up period was 16.69±7.65 months (range, 12-36 months). The coronal DA was 26.74°±10.59° while the sagittal DA was 14.70°±11.63°. The pre- and post-operative Cobb angle were 51.19°±12.91° and 23.25°±12.86° while the correction rate was 57.17%±16.31% and reached 24.26°±13.19° in the last follow-up. The improvement of DAAU, LDA, US, LS, and LO pre- and post-operative had statistical significance ( P<0.001). The pre- and post-operative and the last follow-up C 7PL-CSVL were 27.13±17.08 mm, 21.81±12.80 mm and 20.24±15.02 mm. The pre-operative, postoperative and last follow-up DAAU were -4.35°±2.12°, 1.36°±2.34° and 1.60°±2.45°. The pre- and post-operative LDA were -7.03°±4.40° and 2.42°±3.39°, and the last follow-up LDA was 2.81°±2.98°. US and LS decreased from pre-operative 12.01°±8.33° and 21.46°±5.79° to 2.84°±7.52° and 11.64°±6.06°. The mean US and LS were 4.22°±6.56° and 11.56°±6.02° in the last follow-up. LO decreased after surgery and keep unchanged in the last follow-up, which were 12.71°±6.43°, 6.31°±5.17° and 7.01°±4.73°, respectively. For the sagittal plane parameters, the changes of TK, LL, and TJK reached statistical significance through the surgery. LL increased from 33.69°±14.01° to 44.28°±10.07° through the surgery and reached 41.97°±6.69° at the last follow-up while TK increased from 13.41°±12.37° to 23.52°±8.10°, TJK decreased from 29.02°±20.74° to 16.20°±12.62° after the surgery and reached 16.07°±13.33° at the last follow-up. The pre-operative, post-operative and last follow-up thoracic kyphosis were 13.41°±12.37°, 23.52°± 8.10°, and 24.21°±7.39°. There was no statistical significance of the change of C 7PL-CSVL, SVA, SSA and PI-LL through the surgery ( P>0.05). At the final follow-up, the self-image and psychologic status scores of SRS-22 were significantly higher than that before surgery ( t=15.457, P<0.001; t=14.726, P<0.001), and there was no significant difference in the rest of the domain ( P>0.05). Conclusion:"Lenke5-like" lumbar congenital scoliosis could obtain satisfactory correction of coronal and sagittal deformities with the treatment of posterior column osteotomy, while there was no significant loss of correction during follow-up. The incidence of surgical complications is low.

Objective:To evaluate the safety and efficacy of one-stage posterior-only jumping hemivertebra (HV) resection combined with respective short fusions in the treatment of congenital scoliosis (CS) caused by multiple HVs.Methods:All of 13 consecutive patients with multiple HVs treated surgically from January 2010 to December 2017 were retrospectively reviewed, including 4 males and 9 females with a mean age of 3.7±1.2 years. One child had 4 HVs, and the rest had 2 HVs. The responsible HVs causing local scoliosis/kyphosis deformity or coronal plane deviation were selected as the target of resection. The distal HV was removed firstly and then the proximal one was resected; both of the fixation vertebraes were horizontalized during surgery. The clinical and imaging data of the children before the initial operation, immediately after the operation and at the latest follow-up were collected, and the short-term and long-term complications related to surgery were recorded. The data were evaluated on the whole-standing spine anteroposterior and lateral films, including the corrections of proximal and distal main curves, coronal balance, local kyphosis, and the improvement of spinal growth height (upper and lower internal fixation length, T 1-S 1 length). At the same time, the re-progression of coronal and sagittal deformities of the spine during growth was recorded (coronal decompensation: emerging postoperative curve progression more than 20°; kyphosis progression: kyphosis aggravation between upper and lower internal fixation more than 40°) and internal-fixation-related complications (screw cutting, screw malposition) were recorded. Results:Dual HVs were resected in each child, of which 8 (61.5%) were located on contralateral side of the spine, and 5 (38.5%) were located on ipsilateral side of the spine. The follow-up time was 6.2±3.3 years (range 2.0-10.5 years) after surgery. The Cobb angles of proximal and distal main curves were 36.7°±11.8° and 35.2°±7.8° respectively before surgery and were corrected to 9.7°±6.6° and 6.1°±4.1° respectively after surgery ( F=31.249, F=93.83, P< 0.001) ( t=6.888, t=10.954, P<0.001), and the correction rates was 73.6%±19.6% and 82.7%±11.7%, respectively. They were maintained at 14.3°±5.4° and 8.0°±4.6° at the latest follow-up, showing the correction rates loss of 15.8%±26.9% and 6.9%±7%, respectively. The coronal balance improved from 17.2±14.8 mm pre-operatively to -0.2±15.7 mm postoperatively ( t=2.703, P=0.008), and it remained at 0±18.4 mm at the final follow-up ( F=4.137, P=0.024). The T 1-S 1 length was corrected to 273.8±27.3 mm postoperatively, slightly increased compared with pre-operation 256.3±24.0 mm, ( t=0.680, P=0.527), and significantly increased to 333.2±33.4 mm at the latest follow-up ( t=2.986, P<0.001; F=6.704, P=0.003). Seven patients had local kyphosis before operation, which was significantly improved from 32.2°±13.6° to 6.1°±9.8° with a correction rate of 93.4%±27.0% after surgery ( t=3.355, P=0.004), which showed no significant loss of correction at the latest follow-up (5.4°±10.4°) ( F=11.187, P=0.002). Six patients (46.2%) developed coronal decompensation (Curve magnitude >20 °), with an average of 21.7°±1.9°. Two cases (15.4%) had progressive kyphosis between the thoracic regional internal fixations at 3 months after surgery, which were 68° and 58° respectively. After bracing, both coronal decompensation and sagittal kyphosis were improved. At the last follow-up, the coronal decompensation was improved to 14.7±8.9° and the kyphosis was alleviated to 55° and 46°, respectively. Conclusion:Posterior-only skipping hemivertebra resection and short fusion is a safe, effective procedure yielding significantly improvement of the growth imbalance and reginal spinal deformities of CS with multiple HVs. The mid-term follow-up results showed that the progress of the scoliosis was common during the growth period, which could be further controlled by supplementary brace treatment.

Objective:To compare the incidence of intraoperative neurological complications during correction surgery of spinal deformities in patients with or without rotatory subluxation (RS), and to analyze the correlation between RS and intraoperative neurological complications.Methods:From January 2012 to August 2017, a total of 37 patients with RS undergoing correction surgery, whom was excluded with preoperative spinal traction or three-column osteotomy during operation were retrospectively reviewedin our hospital. Thirty-seven patients without RS undergoing correction surgery were included asthe control group. The radiographic parameters included Cobb angle of main curve, coronal trunk balance (CTB) which was the distance between C 7 plumb line and center sacral vertical line, global kyphosis (GK), sagittal vertical axis (SVA), RS at coronal plane (CRS), RS at sagittal plane (SRS) and axial rotation (AR). The abnormal intraoperative neurophysiological monitoring events and positive wake-up test were recorded. Results:The mean age was 42.4±17.9 years (12-74 years) in the RS group and 42.7±18.3 years (12-74 years) in the control group( t=0.0713, P=0.943). The mean preoperative Cobb angle of main curve, CTB, GK, and SVA was 75.4°±29.7°, 38.4±28.4 mm, 52.8°±25.2°, and 40.3±36.8 mm respectively in the RS group, which was 75.1°±27.6°( t=0.045, P=0.964), 34.8±24.4 mm( t=0.584, P=0.560), 49.8°±22.5°( t=0.540, P=0.591), and 38.7±25.3 mm ( t=0.219, P=0.828) respectively in the control group. There was no significant difference between the two groups among the above preoperative index. Significant improvements in Cobb angle of main curve, CTB, GK, SVA,CRS, SRS and ARA were found between preoperation and postoperation ( P<0.05 for all), while no significant correction loss was observed during follow-up ( P>0.05 for all). There were 6 patients (16.2%) in the RS group and 4 patients (10.8%) in the control group with preoperative neurological deficit of Frankel grade D ( F=0.463, P=0.496). The abnormal intraoperative neurophysiological monitoring events were observed in 5 patients (13.5%) of the RS group and 1 patient (2.7%) of the control group ( F=2.902, P=0.088). Positive wake-up test was found in 2 patients of RS group (5.4%) ( F=2.056, P=0.493). Conclusion:Patients with RS had higher risks of preoperative neurological deficit, abnormal intraoperative neurophysiological monitoring events and deteriorative neurological deficit at postoperation. The RS at preoperation may be a risk factor for intraoperative neurological deficit.