Abstract
Background
Abnormal cervical sagittal alignment (CSA), typically cervical kyphosis, is more common in patients with adolescent idiopathic scoliosis (AIS) than in teenagers without AIS. Changes in CSA after posterior spinal fusion for AIS have been observed and may be associated with patient-reported clinical outcomes and age-related cervical disc degeneration. Previous studies have shown inconsistent postoperative reciprocal changes in CSA in patients with single structural thoracolumbar/lumbar (TL/L) AIS (Lenke 5C classification). However, little is known about the medium-term reciprocal changes in CSA after selective posterior TL/L fusion surgery.
Questions/purposes
We sought to determine the following: (1) What proportion of patients with Lenke 5C AIS have abnormal CSA before surgery? (2) What were the changes in CSA after selective posterior TL/L fusion surgery in the overall Lenke 5C AIS cohort and in subgroups classified by thoracic kyphosis? (3) What global sagittal parameters were associated with CSA preoperatively and at the latest follow-up? (4) What is the correlation between CSA and Scoliosis Research Society Outcomes Questionnaire (SRS-22) scores?
Methods
We queried our institutional database and identified 186 patients diagnosed with Lenke 5C AIS who underwent selective posterior TL/L fusion surgery from April 2010 to February 2018. Of these, 13% (25) of patients were lost to follow-up before 5 years, and 8% (15) of patients were excluded based on exclusion criteria, leaving 79% (146) of patients for analysis in this retrospective study. During this period, we typically offered selective posterior TL/L fusion surgery to patients with Lenke 5C AIS when the main TL/L Cobb angle exceeded 35°. All patients who were offered surgery for this diagnosis opted to have the procedure. Briefly, the surgical procedure consisted of pedicle screw insertion, multiple-level Ponte osteotomy, and segmental direct vertebral body derotation to correct the deformity. Ninety percent (132 of 146) of the patients were female, with a mean ± SD age of 15 ± 2 years. The mean follow-up time was 7 ± 1 years. All patients had a single structural TL/L curve, with a mean preoperative main TL/L Cobb angle of 43° ± 9°. Radiologic measurements included coronal deformity parameters, cervical sagittal parameters, and global sagittal parameters. SRS-22 scores were used to evaluate clinical outcomes. The preoperative cervical sagittal parameters were analyzed to assess the abnormal CSA proportion. Cervical sagittal parameters were compared preoperatively, postoperatively, and at the latest final follow-up. Based on the degree of thoracic kyphosis, patients were classified into a hypokyphotic group (thoracic kyphosis < 20°) and a normokyphotic group (thoracic kyphosis ≥ 20°), with further comparison of CSA within subgroups. Multiple linear regression analysis was performed to assess the correlation between CSA and global sagittal parameters. Finally, the SRS-22 scores at the latest follow-up were compared between cervical lordosis (defined as CSA > 0°) and cervical kyphosis (defined as CSA < 0°).
Results
Fifty-eight percent (84 of 146) of patients with Lenke 5C AIS had cervical kyphosis before surgery. After selective posterior TL/L fusion surgery, we observed an increase in cervical lordosis (from 5° ± 13° before surgery to 2° ± 12°, mean difference 4° [95% confidence interval (CI) 2° to 5°]; p < 0.001), thoracic kyphosis (from 19° ± 9° before surgery to 28° ± 10°, mean difference -9° [95% CI -11° to -8°]; p < 0.001), and lumbar lordosis (from -48° ± 11° before surgery to -51° ± 14°, mean difference 3° [95% CI 1° to 6°]; p = 0.005) at 5 years of follow-up. Subgroup analysis revealed an increase in CSA in the hypokyphotic group, while no increase was observed in the normokyphotic group. At the preoperative stage, after controlling for potentially confounding variables such as the C2-7 sagittal vertical axis and lumbar lordosis, a correlation was observed between thoracic kyphosis (β = -1.27 [95% CI -1.50 to -1.03]; p < 0.001) and the preoperative CSA. At the latest follow-up, after controlling for potentially confounding variables such as thoracic kyphosis, a correlation was observed between global thoracic kyphosis (β = -0.46 [95% CI -0.85 to -0.08]; p = 0.02) and the follow-up CSA. When we compared patients with cervical kyphosis at the most recent follow-up to those with cervical lordosis, we found no between-group differences in SRS-22 scores between those groups.
Conclusion
In light of our findings, surgeons should pay particular attention to preoperative CSA in these patients. For patients with cervical kyphosis and cervical discomfort, if there is concurrent reduction in thoracic kyphosis, it can be communicated that both CSA and thoracic kyphosis are likely to improve postoperatively. Future studies should use more specific outcome measures to assess the correlation between CSA changes and patient-reported clinical outcomes.
Level of Evidence
Level III, therapeutic study.
Introduction
Abnormal cervical sagittal alignment (CSA), especially cervical kyphosis, has been reported to occur in a higher proportion of the adolescent idiopathic scoliosis (AIS) population [18]. As early as 1995, Hilibrand et al. [10] found that abnormalities in CSA were common in patients with AIS. Previous studies have indicated that cervical kyphosis compromises health-related quality of life (HRQoL) and may be associated with the subsequent development of age-related cervical spine degenerative diseases [27]. As an important part of spinal sagittal alignment, CSA is also a paramount factor in maintaining a horizontal gaze. Changes in CSA in patients with AIS after posterior spinal fusion have also proven to have a significant impact on HRQoL [5]. Other studies have noted the reciprocal changes in CSA once global sagittal alignment was altered through spinal fusion surgery [3, 9]. Furthermore, a study observed that postoperative restoration of thoracic kyphosis played a central role in cervical sagittal compensation in patients with AIS [16]. A quantitative analysis found that every 10° of thoracic kyphosis gain was associated with a 6° increase in cervical lordosis after posterior thoracic selective fusion [3].
Most studies have focused on analyzing the cervical sagittal profile in patients with AIS in the thoracic region. Researchers have found cervical kyphosis not only in Lenke Types 1 and 2 but also in Lenke Types 5C and 6 (thoracolumbar/lumbar [TL/L] curves) [18], where the incidence of cervical kyphosis has ranged from 43% to 65% [21, 26, 29, 31]. Prior clinical studies of thoracolumbar fusion [14, 23] and a statistical modeling test [6] have shown alterations in the cervical sagittal profile after surgery, and the possible mechanism may be synchronous restoration of lumbar lordosis and thoracic kyphosis. For patients with Lenke 5C AIS, spontaneous thoracic kyphosis improvement after selective posterior TL/L fusion has been observed [28]; however, mixed results of CSA change have been reported [21, 26, 29, 31]. The conflicting results of CSA may be due to reports with a relatively low number of patients and short follow-up.
We therefore asked the following questions: (1) What proportion of patients with Lenke 5C AIS have abnormal CSA before surgery? (2) What were the changes in CSA after selective posterior TL/L fusion surgery in the overall Lenke 5C AIS cohort and in subgroups classified by thoracic kyphosis? (3) What global sagittal parameters were associated with CSA preoperatively and at the latest follow-up? (4) What is the correlation between CSA and Scoliosis Research Society Outcomes Questionnaire (SRS-22) scores?
Patients and Methods
Study Design and Setting
We conducted a single-center, retrospective study that enrolled patients diagnosed with Lenke 5C AIS who underwent posterior spinal fusion from April 2010 to February 2018. The study center is a nationwide comprehensive tertiary hospital located in the capital of China. Its orthopaedic department includes subspecialties such as joints, trauma, and spine. The center accepts referred patients with spinal deformities from across the country for surgical treatment. Two surgeons (JZ, SW) serve as the lead surgeons. Additionally, five other spine surgeons are involved in the operations, participating on a rotating basis. Written informed consent was acquired from all participants.
Patients
Patients who were 18 years or younger at surgery and who had undergone selective posterior TL/L fusion surgery were potentially eligible. These patients presented with moderate spinal deformities (main TL/L Cobb angle > 35°) and visible lumbar asymmetry, and some also experienced low back pain; all of them were willing to undergo surgery. During this period, we offered selective posterior TL/L fusion surgery to all patients with a confirmed diagnosis of Lenke 5C AIS, as the procedure was known for its effective deformity correction and low complication rates. Previously, our center also performed anterior spinal fusion surgery, but it was discontinued because of complications associated with the anterior approach and inadequate improvement in compensatory thoracic curve correction. Of the 186 patients diagnosed with Lenke 5C AIS in our database, 13% (25) were lost to follow-up before 5 years and 8% (15) were excluded based on exclusion criteria, leaving 79% (146) of patients for analysis in this retrospective study (Fig. 1). The exclusion criteria included: (1) a history of prior spinal surgery, (2) a lack of preoperative and/or postoperative radiographic data, (3) implant failure and/or revision surgery during follow-up, and (4) improper patient position (patient’s head failed to maintain a horizontal gaze) when radiographs were taken.
Fig. 1.
Flow chart showing the recruitment process for the current study.
Descriptive Data
Ninety percent (132 of 146) of the patients were female, with a mean ± SD age of 15 ± 2 years. The mean follow-up period was 7 ± 1 years (range 5 to 11 years). All patients had single structural TL/L scoliosis, with a mean preoperative main TL/L Cobb angle of 43° ± 9°. Seventy-six patients were classified in the hypokyphotic group (thoracic kyphosis < 20°), and 70 in the normokyphotic group (thoracic kyphosis ≥ 20°), based on criteria proposed by Garg et al. [7]. No differences were observed in baseline characteristics such as age, sex, and coronal plane deformity parameters (Table 1).
Table 1.
Baseline characteristics
| Characteristic | Total (n = 146) | Hypokyphotic group (n = 76) | Normokyphotic group (n = 70) | p value |
| Sex, male | 10 (14) | 7 (5) | 13 (9) | 0.26 |
| Age at index surgery in years | 15 ± 2 | 15 ± 2 | 15 ± 2 | 0.73 |
| Follow-up in years | 7 ± 1 | 7±1 | 7 ± 1 | 0.62 |
| Upper instrumented vertebra | 0.46 | |||
| T8 | 3 (5) | 5 (4) | 1 (1) | |
| T9 | 5 (7) | 7 (5) | 3 (2) | |
| T10 | 38 (55) | 39 (30) | 36 (25) | |
| T11 | 49 (71) | 43 (33) | 54 (38) | |
| T12 | 5 (8) | 5 (4) | 6 (4) | |
| Lower instrumented vertebra | 0.80 | |||
| L3 | 40 (58) | 39 (30) | 40 (28) | |
| L4 | 58 (85) | 59 (45) | 57 (40) | |
| L5 | 2 (3) | 1 (1) | 3 (2) | |
| Coronal deformity | ||||
| Main TL/L Cobb angle in ° | 43 ± 9 | 41 ± 9 | 44 ± 9 | 0.12 |
| Minor thoracic Cobb angle in ° | 17 ± 14 | 17 ± 14 | 16 ± 14 | 0.93 |
| Apical vertebral translation in mm | 41 ± 12 | 41 ± 12 | 40 ± 12 | 0.66 |
| Trunk shift in mm | 15 ± 12 | 15 ± 12 | 15 ± 13 | 0.92 |
Data presented as % (n) or mean ± SD.
Surgical Procedure
Patients were placed in a prone position under general anesthesia. After a midline incision, subperiosteal dissection of the posterior soft tissues was performed, and the tips of the transverse processes were exposed. Pedicle screws were placed bilaterally by the freehand technique and confirmed by fluoroscopic C-arm. The upper and lower instrumented vertebrae were selected as the upper and lower end vertebrae, respectively. Unless a developmental anomaly of the pedicle was present, bilateral screws were placed in each vertebra. Monoaxial screws were used both at the lower instrumented vertebra and on the convex side, while polyaxial screws were employed on the concave side. Multiple-level Ponte osteotomies (Schwab Grade 2) were performed as described by Shufflebarger et al. [24], with no intervertebral disc substitution conducted. After the insertion of precontoured cobalt-chromium-molybdenum alloy rods, bilateral rod rotation and a segmental direct vertebral body derotation maneuver [15] were used to correct the deformity. Compression, distraction, and in situ bending maneuvers were applied if necessary. The lamina and transverse processes were decorticated, and allograft bone material was used for fusion.
Evaluation of Radiographic Parameters
All patients were asked to stand facing straight ahead (with chin up) with knees fully extended when taking full-length spine radiographs preoperatively, postoperatively, and at the latest follow-up. Radiographic parameters were measured by Surgimap Software (Nemaris), including coronal deformity parameters, cervical sagittal parameters, and global sagittal parameters [22, 32] (Fig. 2). For sagittal parameters, negative (-) values indicate lordosis. Each radiograph was independently measured by two experienced spine surgeons (YZ, HZ), and the average value of both was used as the final data.
Fig. 2.
Measurement of spinal sagittal parameters. CSA is the angle between the lower endplates of the C2 and C7 vertebrae. The C2-7 sagittal vertical axis (SVA) is the distance from the posterosuperior corner of the C7 vertebra and the vertical line from the center of the C2 body. The T1 slope is the angle between the T1 upper endplate and the horizontal line to the ground in a standing lateral radiograph. Upper thoracic kyphosis (UTK) is the Cobb angle between the upper endplate of the T2 vertebra and the lower endplate of the T5 vertebra. Thoracic kyphosis (TK) is the Cobb angle between the upper endplate of the T5 vertebra and the lower endplate of the T12 vertebra. Global thoracic kyphosis (GTK) is the Cobb angle between the upper endplate of the T2 vertebra and the lower endplate of the T12 vertebra. The thoracolumbar junction (TLJ) is the Cobb angle between the upper endplate of the T10 vertebra and the lower endplate of the L2 vertebra. Lumbar lordosis (LL) is the Cobb angle between the upper endplates of the L1 and the S1 vertebrae. Pelvic incidence (PI) is the angle between the line perpendicular to the sacral plate at its midpoint and the line connecting this point to the axis of the femoral heads. Pelvic tilt (PT) is the angle between the line connecting the midpoint of the sacral plate to the axis of the femoral heads and the vertical reference line. Sacral slope (SS) is the angle between the upper endplate of the sacrum and the horizontal line to the ground in standing radiographs. Sagittal vertical axis (SVA) is the horizontal offset from a plumbline dropped from C7 to the posterosuperior corner of S1. A color image accompanies the online version of this article.
Evaluation of Clinical Outcomes
The clinical outcomes were evaluated using the SRS-22, which was administered before surgery and at each regular postoperative follow-up period (3 months, 6 months, and 1 year after surgery, and then at 1-year intervals). The results from the preoperative and the most recent follow-up were included in the analysis. In all, 76% (111 of 146) of patients completed the SRS-22 questionnaire at both the preoperative and the most recent follow-up.
Primary and Secondary Study Outcomes
Our primary study goal was to explore the proportion of abnormal CSA values in patients with Lenke 5C AIS. To achieve this, we analyzed preoperative cervical sagittal parameters. Additionally, we aimed to clarify changes in CSA after selective posterior TL/L fusion surgery in the overall Lenke 5C AIS cohort and in subgroups classified by thoracic kyphosis. This was accomplished by conducting intragroup comparisons of CSA between preoperative and the latest follow-up measurements in all patients, as well as in the hypokyphotic and normokyphotic subgroups.
Our secondary study goal was to investigate the relationship between CSA and global sagittal parameters. To achieve this, we performed multiple linear regression analyses to evaluate the association between CSA and global sagittal parameters, controlling for potential confounders (preoperative stage: C2-7 sagittal vertical axis, thoracic kyphosis, global thoracic kyphosis, and lumbar lordosis; latest follow-up: thoracic kyphosis and global thoracic kyphosis) both preoperatively and at the latest follow-up. Another secondary goal was to explore the correlation between CSA and SRS-22 scores. This was achieved by comparing SRS-22 scores at the latest follow-up between patients with cervical lordosis (defined as CSA > 0°) and cervical kyphosis (defined as CSA < 0°).
Ethical Approval
We obtained ethical approval for this study from the Peking Union Medical College Hospital Institutional Review Board (K3210).
Statistical Analysis
The minimum number of patients was calculated using PASS15 (Numerical Computing and Statistical Systems), and a sample size of 90 patients achieved 80% statistical power. Mean imputation was used for missing values of radiographic data and SRS-22 scores with SAS software. Continuous variables were tested for normal distribution by the Shapiro-Wilk test. For data that followed a normal distribution (main TL/L Cobb angle, apical vertebral translation, CSA, thoracic kyphosis, global thoracic kyphosis, lumbar lordosis, pelvic tilt, and sacral slope), we applied the paired t-test and Student t-test. For data that did not follow a normal distribution, we used the Mann-Whitney U test. The chi-square test or Fisher exact test was used for categorical variables. A correlation analysis was performed using a Spearman test, all cervical sagittal parameters and global sagittal parameters were included in the univariate analysis, and factors with p < 0.05 in the univariate analysis were entered into the multiple linear regression analysis (for the analysis at the preoperative stage, C2-7 sagittal vertical axis, thoracic kyphosis, global thoracic kyphosis, and lumbar lordosis were included; for the analysis at the latest follow-up, thoracic kyphosis and global thoracic kyphosis were included). Statistical analyses were performed using SPSS Statistics for Windows, version 22.0 (IBM). Alpha was set at 0.05.
Results
Proportion of Patients With Abnormal CSA Before Surgery
Fifty-eight percent (84 of 146) of patients with Lenke 5C AIS had cervical kyphosis before surgery (Fig. 3). The mean ± SD preoperative CSA was 5° ± 13°, and the preoperative C2-7 sagittal vertical axis was 9 ± 8 mm (Table 2).
Fig. 3.
A 13-year-old female with Lenke 5C AIS underwent selective TL/L fusion surgery. (A-C) Coronal deformity was corrected from 44° to 3° after surgery and maintained at 4° at the 6-year follow-up. (D-F) The sagittal plane showed good restoration of CSA and other relevant parameters.
Table 2.
Comparison of preoperative, postoperative, and the latest follow-up of radiologic parameters among all patients
| Parameter | Preoperative | Postoperative | Follow-up | p value |
| Main TL/L Cobb angle in ° | 43 ± 9 | 8 ± 5 | 11 ± 6 | < 0.001 |
| Minor thoracic Cobb angle in ° | 17 ± 14 | 8 ± 8 | 10 ± 10 | < 0.001 |
| Apical vertebral translation in mm | 41 ± 12 | 14 ± 7 | 12 ± 6 | < 0.001 |
| Trunk shift in mm | 15 ± 12 | 8 ± 9 | 4 ± 4 | < 0.001 |
| CSA in ° | 5 ± 13 | 2 ± 13 | 2 ± 12 | < 0.001 |
| CSA < 0° | 58 (84) | 46 (68) | 40 (59) | < 0.001 |
| C2-7 sagittal vertical axis in mm | 9 ± 8 | 8 ± 8 | 8 ± 8 | 0.78 |
| T1-slope in ° | 14 ± 6 | 14 ± 7 | 14 ± 6 | 0.54 |
| Upper thoracic kyphosis (T2-5) in ° | 13 ± 6 | 12 ± 5 | 12 ± 5 | 0.05 |
| Thoracic kyphosis (T5-12) in ° | 19 ± 9 | 24 ± 9 | 28 ± 10 | < 0.001 |
| Global thoracic kyphosis (T2-12) in ° | 28± 10 | 33 ± 9 | 35 ± 9 | < 0.001 |
| Thoracolumbar junction (T10-L2) in ° | 7 ± 6 | -3 ± 4 | -1 ± 6 | < 0.001 |
| Lumbar lordosis in ° | -48 ± 11 | -51 ± 10 | -51 ± 14 | 0.005 |
| Pelvic incidence in ° | 42 ± 12 | 42 ± 12 | 42 ± 12 | 0.97 |
| Pelvic tilt in ° | 9 ± 10 | 10 ± 11 | 10 ± 10 | 0.36 |
| Sacral slope in ° | 33 ± 9 | 33 ± 8 | 33 ± 8 | 0.39 |
| Sagittal vertical axis in mm | -11 ± 26 | -11 ± 22 | -11 ± 20 | 0.87 |
Data presented as mean ± SD or % (n). p values are for the latest follow-up compared to preoperative.
Postoperative Changes of CSA
After selective posterior TL/L fusion surgery, we observed an increase in CSA (from 5° ± 13° before surgery to 2° ± 12° (Fig. 4), mean difference 4° [95% confidence interval (CI) 2° to 5°]; p < 0.001), thoracic kyphosis (from 19° ± 9° before surgery to 28° ± 10°, mean difference -9° [95% CI -11° to -8°]; p < 0.001), and lumbar lordosis (from -48° ± 11° before surgery to -51° ± 14°, mean difference 3° [95% CI 1° to 6°]; p = 0.005) at 5 years of follow-up (Table 2). In the subgroup analysis, we found that CSA and some (but not all) global sagittal parameters were changed between preoperative and the latest follow-up in the hypokyphotic group, while no changes in CSA were noted in the normokyphotic group. In addition, all coronal deformity parameters were changed at the 5-year follow-up compared with the preoperative values (Table 3).
Fig. 4.
A 12-year-old female with Lenke 5C AIS underwent selective TL/L fusion surgery. (A-C) Coronal and axial deformities were well corrected from 42° to 3° after surgery and maintained at 8° at the 7-year follow-up. (D-F) The cervical alignment changed from straight to lordosis after the reconstruction of lumbar lordosis, thoracolumbar junction, and thoracic kyphosis at 7-year follow-up.
Table 3.
Comparison of sagittal parameters between the hypokyphotic group and normokyphotic group
| Hypokyphotic group, thoracic kyphosis < 20° | Normokyphotic group, thoracic kyphosis ≥ 20° | |
| CSA in ° | ||
| Preoperative | 12 ± 11 | -3 ± 11 |
| Postoperative | 7 ± 12 | -4 ± 11 |
| Follow-up | 7 ± 12 | -4 ± 11 |
| p value (follow-up versus preoperative) | < 0.001 | 0.11 |
| C2-7 sagittal vertical axis in mm | ||
| Preoperative | 8 ± 9 | 9 ± 8 |
| Postoperative | 8 ± 7 | 8 ± 8 |
| Follow-up | 8 ± 8 | 8 ± 8 |
| p value (follow-up versus preoperative) | 0.94 | 0.66 |
| T1-slope in ° | ||
| Preoperative | 14 ± 6 | 14 ± 6 |
| Postoperative | 13 ± 7 | 14 ± 7 |
| Follow-up | 14 ± 6 | 14 ± 6 |
| p value (follow-up versus preoperative) | 0.60 | 0.75 |
| Upper thoracic kyphosis in ° | ||
| Preoperative | 13 ± 6 | 12 ± 6 |
| Postoperative | 13 ± 5 | 12 ± 5 |
| Follow-up | 12 ± 5 | 11 ± 6 |
| p value (follow-up versus preoperative) | 0.23 | 0.13 |
| Thoracic kyphosis in ° | ||
| Preoperative | 12 ± 5 | 27 ± 6 |
| Postoperative | 20 ± 7.6 | 29 ± 7 |
| Follow-up | 25 ± 9 | 32 ± 9 |
| p value (follow-up versus preoperative) | < 0.001 | < 0.001 |
| Global thoracic kyphosis in ° | ||
| Preoperative | 21 ± 7 | 35 ± 8 |
| Postoperative | 30 ± 8 | 36 ± 8 |
| Follow-up | 33 ± 9 | 38 ± 9 |
| p value (follow-up versus preoperative) | < 0.001 | < 0.001 |
| Thoracolumbar junction in ° | ||
| Preoperative | 6 ± 6 | 7 ± 7 |
| Postoperative | -3 ± 4 | -2 ± 4 |
| Follow-up | -2 ± 6 | 0 ± 6 |
| p value (follow-up versus preoperative) | < 0.001 | < 0.001 |
| Lumbar lordosis in ° | ||
| Preoperative | -45 ± 11 | -50 ± 11 |
| Postoperative | -49 ± 10 | -52 ± 9 |
| Follow-up | -50 ± 11 | -52 ± 17 |
| p value (follow-up versus preoperative) | < 0.001 | 0.45 |
| Pelvic incidence in ° | ||
| Preoperative | 44 ± 12 | 41 ± 12 |
| Postoperative | 44 ± 12 | 41 ± 12 |
| Follow-up | 44 ± 12 | 41 ± 12 |
| p value (follow-up versus preoperative) | 0.91 | 0.90 |
| Pelvic tilt in ° | ||
| Preoperative | 10 ± 10 | 8 ± 10 |
| Postoperative | 11 ± 11 | 8 ± 12 |
| Follow-up | 11 ± 10 | 8 ± 10 |
| p value (follow-up versus preoperative) | 0.21 | 0.99 |
| Sacral slope in ° | ||
| Preoperative | 34 ± 9 | 33 ± 9 |
| Postoperative | 32 ± 8 | 33 ± 7 |
| Follow-up | 32 ± 9 | 33 ± 8 |
| p value (follow-up versus preoperative) | 0.24 | 0.98 |
| Sagittal vertical axis in mm | ||
| Preoperative | -13 ± 26 | -9 ± 27 |
| Postoperative | -15 ± 19 | -8 ± 24 |
| Follow-up | -15 ± 19 | -7 ± 21 |
| p value (follow-up versus preoperative) | 0.52 | 0.58 |
Data presented as mean ± SD.
Global Sagittal Parameters Associated With CSA
At the preoperative stage, after controlling for potentially confounding variables such as the C2-7 sagittal vertical axis and lumbar lordosis, a correlation was observed between thoracic kyphosis (β = -1.27 [95% CI -1.50 to -1.03]; p < 0.001) and the preoperative CSA. At the latest follow-up, after controlling for potentially confounding variables such as thoracic kyphosis, a correlation was observed between global thoracic kyphosis (β = -0.46 [95% CI -0.85 to -0.08]; p = 0.02) and the follow-up CSA (Table 4).
Table 4.
Multiple linear regression analysis of CSA
| Factor | β (95% CI) | p value |
| Preoperative CSA | ||
| C2-7 sagittal vertical axis | -0.14 (-0.28 to 0.01) | 0.06 |
| Thoracic kyphosis | -1.27 (-1.50 to -1.03) | < 0.001 |
| Global thoracic kyphosis | 0.09 (-0.13 to 0.31) | 0.41 |
| Lumbar lordosis | -0.01 (-0.13 to 0.11) | 0.83 |
| Follow-up CSA | ||
| Thoracic kyphosis | 0.01 (-0.36 to 0.37) | 0.96 |
| Global thoracic kyphosis | -0.46 (-0.85 to -0.08) | 0.02 |
The only variable correlated with preoperative CSA after controlling for confounding variables like C2-7 sagittal vertical axis and lumbar lordosis was thoracic kyphosis; as preoperative cervical alignment became more lordotic, thoracic kyphosis increased. The adjusted R2 for our multiple linear regression analysis of preoperative CSA was 0.69, which means that 69% of the variability in preoperative CSA could potentially be attributed to the other factors listed in the left-hand column of that portion of the table. The only variable correlated with follow-up CSA after controlling for confounding variables like thoracic kyphosis was global thoracic kyphosis; as preoperative cervical alignment became more lordotic, global thoracic kyphosis increased. The adjusted R2 for our multiple linear regression analysis of follow-up CSA was 0.11, which means that 11% of the variability in follow-up CSA could potentially be attributed to the other factors listed in the left-hand column of that portion of the table.
Correlation Between CSA and SRS-22 Scores
When we compared patients with cervical kyphosis at most recent follow-up with those with cervical lordosis (cervical lordosis was defined as CSA > 0° and cervical kyphosis was defined as CSA < 0°), we found no between-group differences in SRS-22 scores (Supplemental Table 1; http://links.lww.com/CORR/B407). Two patients experienced superficial wound infections and one developed a keloid scar. No cases of deep wound infections, intraoperative monitoring loss, paralysis, or pseudarthrosis were observed.
Discussion
Abnormal cervical sagittal profile, particularly cervical kyphosis, is common in patients with AIS and has been associated with unsatisfactory patient-reported clinical outcomes and age-related cervical spine degeneration [18, 27]. Changes in cervical sagittal profile after posterior spinal fusion surgery significantly affect HRQoL and are influenced by global sagittal alignment [5]. However, several questions remain unclear in patients with Lenke 5C AIS, including the proportion of abnormal CSA, the changes in CSA after surgery, the association between CSA and global sagittal parameters, and the correlation between CSA and outcome scores. We aimed to address these questions by analyzing and comparing cervical sagittal parameters, global sagittal parameters, and SRS-22 scores at the preoperative and the latest follow-up. The findings revealed a high proportion of abnormal CSA in patients with Lenke 5C AIS. An increase in cervical lordosis was observed following selective posterior TL/L fusion surgery, and most of them were found in the hypokyphotic group. A correlation was found between thoracic kyphosis and CSA preoperatively and between global thoracic kyphosis and CSA at the latest follow-up. We found no correlation between CSA and SRS-22 scores. Based on our results, we recommend that preoperative cervical sagittal parameters be routinely assessed for all patients with Lenke 5C AIS. Surgeons can be confident in the midterm reciprocal changes in CSA after selective posterior TL/L fusion surgery, especially in patients with thoracic hypokyphosis.
Limitations
First, in this study, all patients with Lenke 5C AIS underwent selective posterior TL/L fusion surgery, and the deformity of the included patients would be considered moderate in severity rather than being severe or complex. Therefore, the results of this study are most applicable to such patients. Additionally, 90% of the patients in this study were women. Although we hypothesize that gender may not significantly affect the results, we still have concerns about the applicability of our findings to men. Second, the sagittal alignment of the uninstrumented spinal segment may continue to change over a longer follow-up period, making it essential to continue following these patients. However, our study had relatively complete follow-up, with a minimum follow-up period of 5 years. We believe that patients with this length of follow-up would not experience significant interference in spinal alignment because of factors such as postoperative pain, muscle recovery, or spinal fusion. We also believe that reciprocal changes in spinal alignment would be adequately completed within the 5-year postoperative period. Third, this single-center study was conducted at a large, urban center in China by a high-volume, specialized scoliosis surgeon team, and therefore, our findings are most applicable to other centers that treat referral patients with spinal deformity where the team has extensive experience in spinal deformity surgery. Fourth, during the study period, we used the SRS-22 score as our primary patient-reported outcome measure, which does not specifically evaluate cervical spine discomfort. Therefore, our finding of no difference in our fourth research question may represent a false negative. Future studies should incorporate specific assessment tools such as the Neck Disability Index or Japanese Orthopaedic Association Neck Scores to better evaluate cervical discomfort. Fifth, our study used cobalt-chromium-molybdenum rods and a high implant density with an all-pedicle screw construct, which may have impacted the sagittal alignment of the fused segments. Therefore, the conclusions of this study are only applicable to patients who undergo the same surgical technique and fixation method. Sixth, the positioning of patients during radiographic imaging did not strictly follow the prescribed guidelines, which may have introduced errors into the results. In future studies, we will apply the method described in a previous study [19], which places a mirror in front of the patient to standardize the head’s position and unify the line of sight. Additionally, with regard to the measurement method, these were not repeated, and no consistency analysis such as kappa or intraclass correlation coefficient was performed, which was another limitation of this study.
Proportion of Patients With Abnormal CSA Before Surgery
A large proportion of patients with Lenke 5 AIS presented with cervical kyphosis prior to surgery, which we defined as an angle between the lower endplates of C2 and the C7 vertebrae of < 0°. This is potentially important because cervical kyphosis may lead to a range of debilitating symptoms, such as neck pain, difficulty swallowing, and compromised posture, all of which can further impair a patient’s quality of life. It may also help explain why some patients complain of neck discomfort and easy fatigue. Compared with age-matched controls, patients with AIS have a higher proportion of cervical kyphosis [13]. In a study of 318 patients with AIS, 67% (213 of 318) had preoperative cervical kyphosis [2]. As for patients with Lenke 5C AIS, only two studies with small sample sizes directly reported a 41% to 43% incidence of cervical kyphosis [29, 31]. In contrast, our findings indicated a 58% incidence of cervical kyphosis. Given that there were no racial differences and that age and preoperative deformity severity appeared to be consistent, we hypothesize that the discrepancy may be attributed to sample size. Future research should continue to investigate this and further explore the proportion of cervical kyphosis in different types of scoliosis.
Postoperative Changes of CSA
After selective posterior TL/L fusion surgery, we observed an increase in CSA, thoracic kyphosis, and lumbar lordosis at the latest follow-up, and the increase of CSA was primarily observed in patients with a thoracic hypokyphosis. This is important because it highlights the spine’s strong ability to self-correct, not just in the coronal plane, but also in the sagittal plane. Previous studies have confirmed that in patients with major thoracic curve, the CSA could change spontaneously after posterior spinal fusion surgery [11]. However, studies on CSA in patients with Lenke 5C AIS are scarce, and the existing results are controversial. A retrospective study that enrolled 39 patients with a mean follow-up time of 2 years found that cervical lordosis increased after selective posterior TL/L fusion surgery. With respect to the Lenke sagittal modifier, CSA was more likely to be affected in the sagittal minus group (thoracic kyphosis < 10°) than in the sagittal normal group (10° < thoracic kyphosis ≤ 40°) [26]. A 5-year follow-up study found that the postoperative cervical lordosis angle was significantly increased compared with preoperative values [20]. Another study also demonstrated an increase and subsequent maintenance of CSA after surgery with a minimum of 2 years of follow-up [31]. In contrast, some studies have reported inconsistent results. A retrospective clinical and radiographic study of 30 patients with Lenke 5C AIS, with a mean ± SD follow-up of 27 ± 22 months, found that cervical lordosis decreased after surgery but then increased over time [29]. The authors suggested that the spontaneous adjustment of cervical alignment might be limited, possibly due to the inherent rigidity of the cervical spine. A further study including 66 patients with a 5-year follow-up period also found that CSA was not significantly affected by surgery in most patients. However, their subgroup analysis indicated that an upper instrumented vertebra located at the T9 or higher segment could significantly improve CSA [21]. Our study may help explain the discrepancies observed in previous reports. We found that 58% (84 of 146) of patients had cervical kyphosis before surgery, which decreased to 40% (59 of 146) at the latest follow-up. We performed a minimum sample size calculation, which showed that a sample size of 90 achieves 80% statistical power, surpassing the sample sizes of all previous studies focused on Lenke 5C AIS. Therefore, the smaller sample sizes in earlier studies may have limited their ability to detect statistically significant differences. Additionally, our subgroup analysis revealed that CSA changed in the hypokyphotic group but did not change in the normokyphotic group. This suggests that factors such as preoperative thoracic kyphosis may influence the results in patients with Lenke 5C AIS, yet these factors were not fully explored in previous studies. Furthermore, the reciprocal changes in CSA after surgery occur over time, meaning that inadequate follow-up duration may not give enough time for CSA to exhibit compensatory changes.
Global Sagittal Parameters Associated With CSA
At the preoperative stage, after controlling for potentially confounding variables such as the C2-7 sagittal vertical axis and lumbar lordosis, a correlation was observed between thoracic kyphosis and the preoperative CSA. Many previous studies have explored the relationship between CSA and global sagittal alignment in patients with AIS and the potential reciprocal interactions of the spine that exist [1, 25]. However, most previous studies have focused on patients with thoracic AIS, and few studies have focused on TL/L curves (Lenke 5C). In patients with Lenke 5C AIS, cervical lordosis was found mainly associated with the T1 slope, thoracic kyphosis, and global thoracic kyphosis, while there was no significant correlation between cervical lordosis and upper thoracic kyphosis, lumbar lordosis, sagittal vertical axis, and pelvic parameters [29]. A similar significant relationship between cervical lordosis and T1 slope, as well as thoracic kyphosis, was noticed in 39 patients with a major lumbar curve [31]. Our findings were consistent with those results reported in the comprehensive review of selective posterior TL/L fusion [8], and our study provided a more robust validation, ultimately identifying thoracic kyphosis as a key parameter associated with CSA in multiple linear regression analysis. We proposed that reciprocal changes in CSA are present in Lenke 5C AIS to maintain horizontal gaze, with sagittal parameters influencing one another. Specifically, CSA was directly and closely correlated with thoracic kyphosis, while other sagittal parameters, such as lumbar lordosis, may have indirectly influenced CSA through their effects on thoracic kyphosis.
At the latest follow-up, after controlling for potentially confounding variables such as thoracic kyphosis, a correlation was observed between global thoracic kyphosis and the follow-up CSA. Restoring appropriate sagittal alignment is widely regarded as a critical objective of surgery for patients with AIS [7, 12]. Numerous studies focusing on main thoracic curves have demonstrated that the restoration of thoracic kyphosis is a prerequisite for improving cervical lordosis. In patients with Lenke 5C AIS, selective posterior TL/L fusion surgery directly reconstructs lumbar lordosis, and thoracic kyphosis is usually changed after lumbar lordosis restoration [17]. However, there is currently limited research on the relationship between CSA and global sagittal parameters after surgery. Wang et al [29] inferred that overcorrection of lumbar lordosis led to an increase in thoracic kyphosis, which then increased cervical lordosis after surgery in 30 patients with Lenke 5 AIS. Other researchers have also observed changes in thoracolumbar junction, thoracic kyphosis, global thoracic kyphosis, and cervical lordosis after surgery, emphasizing that achieving a smooth transition from lumbar lordosis to thoracic kyphosis is critical for influencing cervical lordosis [26]. Additionally, a subgroup analyses in a 5-year follow-up study also found that changes in thoracolumbar junction could contribute to alterations in both thoracic kyphosis and CSA [21]. According to our findings, we believe that after selective posterior TL/L fusion surgery, the lumbar lordosis and thoracolumbar junction were directly reconstructed by surgery, and then reciprocal interactions in the proximal sagittal plane of the spine followed, appearing as restoration of global thoracic kyphosis and CSA.
Correlation Between CSA and SRS-22 Scores
When we compared patients with cervical kyphosis at the most recent follow-up with those with cervical lordosis, we found that the groups did not differ in terms of SRS-22 scores. But this may be attributed to the fact that the outcome tool we used does not directly assess cervical discomfort or related function. Wang et al. [30] proposed that neck pain was significantly more common in the cervical imbalance group than in the control group. Furthermore, approximately 18% to 57% of patients with AIS reported experiencing neck pain after posterior spinal surgery, a finding that warrants the attention of spine surgeons [4]. However, the relationship between CSA and patient-reported clinical outcomes in Lenke 5C AIS remains poorly understood. Only one study [21] has shown comparable SRS-22 outcomes between groups divided according to the upper end vertebra chosen in selective posterior TL/L fusion surgery. Future research should incorporate more specific assessment tools, such as the Neck Disability Index, Japanese Orthopaedic Association Neck Score, or other HRQoL questionnaires, to more accurately evaluate cervical discomfort.
Conclusion
Our findings revealed a high proportion of abnormal CSA in patients with Lenke 5C AIS. After selective posterior TL/L fusion surgery, an increase in cervical lordosis was observed. We found reciprocal interactions between CSA and global sagittal alignment; however, no correlation was observed between CSA and SRS-22 scores. Based on these results, we recommend that preoperative cervical sagittal parameters be routinely assessed in patients with Lenke 5C AIS. Surgeons can be confident in the medium-term reciprocal changes in CSA after selective posterior TL/L fusion, especially in patients with thoracic hypokyphosis. Future studies should employ more specific tools to further investigate the correlation between CSA changes and patient-reported clinical outcomes.
Acknowledgments
We thank Jing Chen and Yushan Xue for their help and statistical guidance in our data analysis.
Footnotes
The institution of one or more of the authors (JZ) has received, during the study period, funding from the National Natural Science Foundation of China (82172382 and 81972037).
Each author certifies that there are no funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article related to the author or any immediate family members.
All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.
Ethical approval for this study was obtained from the Peking Union Medical College Hospital Institutional Review Board (K3210).
Contributor Information
Yiwei Zhao, Email: zywspine@163.com.
You Du, Email: duyou1994@qq.com.
Yang Yang, Email: kaido137@hotmail.com.
Haoran Zhang, Email: zhanghaoran96@126.com.
Chenkai Li, Email: leechenkai@163.com.
Dihan Sun, Email: sundihan1998@126.com.
Ziquan Li, Email: 13953138377@163.com.
Jianguo Zhang, Email: jgzhang_pumch@yahoo.com.
References
- 1.Charles YP, Sfeir G, Matter-Parrat V, Sauleau EA, Steib JP. Cervical sagittal alignment in idiopathic scoliosis treated by posterior instrumentation and in situ bending. Spine (Phila Pa 1976). 2015;40:E419-E427. [DOI] [PubMed] [Google Scholar]
- 2.Cho JH, Hwang CJ, Choi YH, Lee DH, Lee CS. Cervical sagittal alignment in patients with adolescent idiopathic scoliosis: is it corrected by surgery? J Neurosurg Pediatr. 2018;21:292-301. [DOI] [PubMed] [Google Scholar]
- 3.Clement JL, Le Goff L, Oborocianu I, et al. Surgical increase in thoracic kyphosis predicts increase of cervical lordosis after thoracic fusion for adolescent idiopathic scoliosis. Eur Spine J. 2021;30:3550-3556. [DOI] [PubMed] [Google Scholar]
- 4.Edgar MA, Metha MH. Long-term follow-up of fused and unfused idiopathic scoliosis. J Bone Joint Surg Br. 1988;70:712-716. [DOI] [PubMed] [Google Scholar]
- 5.Fan Y, Wang J, Cai M, Xia L, Wang X. The correlation between postoperative cervical sagittal alignment and spine sagittal alignment in adolescent idiopathic scoliosis: a meta-analysis. World Neurosurg. 2020;134:e311-e316. [DOI] [PubMed] [Google Scholar]
- 6.Gardner A, Berryman F, Pynsent P. Statistical modelling of how the sagittal alignment of the cervical spine is affected by adolescent idiopathic scoliosis and how scoliosis surgery changes that. J Anat. 2022;241:437-446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Garg B Mehta N Gupta A, et al. Cervical sagittal alignment in Lenke 1 adolescent idiopathic scoliosis and assessment of its alteration with surgery: a retrospective, multi-centric study. Spine Deform. 2021;9:1559–1568.34097297 [Google Scholar]
- 8.Ghandhari H, Ameri E, Nikouei F, Mahdavi SM, Chehrassan M, Motalebi M. Selective thoracolumbar/lumbar fusion in adolescent idiopathic scoliosis: a comprehensive review of the literature. Arch Bone Jt Surg. 2023;11:313-320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hayashi K, Toyoda H, Terai H, et al. Cervical lordotic alignment following posterior spinal fusion for adolescent idiopathic scoliosis: reciprocal changes and risk factors for malalignment. J Neurosurg Pediatr. 2017;19:440-447. [DOI] [PubMed] [Google Scholar]
- 10.Hilibrand AS, Tannenbaum DA, Graziano GP, Loder RT, Hensinger RN. The sagittal alignment of the cervical spine in adolescent idiopathic scoliosis. J Pediatr Orthop. 1995;15:627-632. [DOI] [PubMed] [Google Scholar]
- 11.Hiyama A, Sakai D, Watanabe M, Katoh H, Sato M, Mochida J. Sagittal alignment of the cervical spine in adolescent idiopathic scoliosis: a comparative study of 42 adolescents with idiopathic scoliosis and 24 normal adolescents. Eur Spine J. 2016;25:3226-3233. [DOI] [PubMed] [Google Scholar]
- 12.Ilharreborde B. Sagittal balance and idiopathic scoliosis: does final sagittal alignment influence outcomes, degeneration rate or failure rate? Eur Spine J. 2018;27(Suppl 1):48-58. [DOI] [PubMed] [Google Scholar]
- 13.Iyer S, Lenke LG, Nemani VM, et al. Variations in sagittal alignment parameters based on age: a prospective study of asymptomatic volunteers using full-body radiographs. Spine (Phila Pa 1976). 2016;41:1826-1836. [DOI] [PubMed] [Google Scholar]
- 14.Lee CS, Noh H, Lee DH, Hwang CJ, Kim H, Cho SK. Analysis of sagittal spinal alignment in 181 asymptomatic children. J Spinal Disord Tech. 2012;25:E259-E263. [DOI] [PubMed] [Google Scholar]
- 15.Lee SM, Suk SI, Chung ER. Direct vertebral rotation: a new technique of three-dimensional deformity correction with segmental pedicle screw fixation in adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2004;29:343-349. [DOI] [PubMed] [Google Scholar]
- 16.Li J, Deng K, Tang Y, et al. Postoperative alterations of sagittal cervical alignment and risk factors for cervical kyphosis in 124 Lenke 1 adolescent idiopathic scoliosis patients. BMC Musculoskelet Disord. 2021;22:1001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Li J, Lin Z, Ma Y, Li W, Yu M. How to make a more optimal surgical plan for Lenke 5 adolescent idiopathic scoliosis patients: a comparative study based on the changes of the sagittal alignment and selection of the lowest instrumented vertebra. J Orthop Surg Res. 2023;18:224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Norheim EP, Carreon LY, Sucato DJ, Lenke LG, Glassman SD. Cervical spine compensation in adolescent idiopathic scoliosis. Spine Deform. 2015;3:327-331. [DOI] [PubMed] [Google Scholar]
- 19.Oe S, Togawa D, Yoshida G, et al. Effects of mirror placement on sagittal alignment of the spine during acquisition of full-spine standing X-Rays. Eur Spine J. 2018;27:442-447. [DOI] [PubMed] [Google Scholar]
- 20.Okubo T, Suzuki S, Ozaki M, et al. Influence of thoracolumbar kyphosis on postoperative spinal alignment in patients with Lenke type 5C adolescent idiopathic scoliosis. Spine Deform. 2024;12:411-422. [DOI] [PubMed] [Google Scholar]
- 21.Okubo T, Yagi M, Suzuki S, et al. Does selective posterior correction and fusion surgery influence cervical sagittal alignment in patients with Lenke type 5 adolescent idiopathic scoliosis? A 5-year follow-up retrospective cohort study. Spine (Phila Pa 1976). 2021;46:E976-E984. [DOI] [PubMed] [Google Scholar]
- 22.Pepke W, Almansour H, Lafage R, et al. Cervical spine alignment following surgery for adolescent idiopathic scoliosis (AIS): a pre-to-post analysis of 81 patients. BMC Surg. 2019;19:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Roussouly P, Labelle H, Rouissi J, Bodin A. Pre- and postoperative sagittal balance in idiopathic scoliosis: a comparison over the ages of two cohorts of 132 adolescents and 52 adults. Eur Spine J. 2013;22(Suppl 2):S203-S215. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Shufflebarger HL, Geck MJ, Clark CE. The posterior approach for lumbar and thoracolumbar adolescent idiopathic scoliosis: posterior shortening and pedicle screws. Spine (Phila Pa 1976). 2004;29:269-276. [DOI] [PubMed] [Google Scholar]
- 25.Tang Y, Xu X, Zhu F, et al. Incidence and risk factors of cervical kyphosis in patients with adolescent idiopathic scoliosis. World Neurosurg. 2019;127:e788-e792. [DOI] [PubMed] [Google Scholar]
- 26.Tauchi R, Kawakami N, Ohara T, et al. Sagittal alignment profile following selective thoracolumbar/lumbar fusion in patients with Lenke type 5C adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2019;44:1193-1200. [DOI] [PubMed] [Google Scholar]
- 27.Teo AQA, Thomas AC, Hey HWD. Sagittal alignment of the cervical spine: do we know enough for successful surgery? J Spine Surg. 2020;6:124-135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wang F, Xu XM, Wei XZ, Zhu XD, Li M. Spontaneous thoracic curve correction after selective posterior fusion of thoracolumbar/lumbar curves in Lenke 5C adolescent idiopathic scoliosis. Medicine (Baltimore). 2015;94:e1155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wang F, Zhou XY, Xu XM, et al. Cervical sagittal alignment limited adjustment after selective posterior thoracolumbar/lumbar curve correction in patients with Lenke type 5C adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2017;42:E539-E546. [DOI] [PubMed] [Google Scholar]
- 30.Wang L, Liu X. Cervical sagittal alignment in adolescent idiopathic scoliosis patients (Lenke type 1-6). J Orthop Sci. 2017;22:254-259. [DOI] [PubMed] [Google Scholar]
- 31.Yan P, Zhang Y, Liu S, et al. Sagittal profile response of cervical spine after posterior correction in thoracic and lumbar adolescent idiopathic scoliosis: correlation with thoracic kyphosis? World Neurosurg. 2018;120:e333-e341. [DOI] [PubMed] [Google Scholar]
- 32.Zhuang Q, Zhang J, Wang S, Yang Y, Lin G. How to select the lowest instrumented vertebra in Lenke type 5 adolescent idiopathic scoliosis patients? Spine J. 2021;21:141-149. [DOI] [PubMed] [Google Scholar]