Comparison of en-bloc direct vertebrae rotation and non-direct vertebrae rotation for the correction of adolescent idiopathic scoliosis Lenke 5C: a retrospective study in Changsha, China

Article information

Asian Spine J. 2024;18(6):803-812

Publication date (electronic) : 2024 December 31

doi : https://doi.org/10.31616/asj.2024.0318

Emmanuel Alonge^1,2, HongQi Zhang^1,2, Shaohua Liu^1,2, Yuxiang Wang^1,2

¹Department of Spine Surgery and Orthopaedics, Xiangya Hospital of Central South University, Changsha, China

²National Clinical Research Centre for Geriatric Disorders, Xiangya Hospital of Central South University, Changsha, China

Corresponding author: HongQi Zhang, Department of Spine Surgery and Orthopaedics, Xiangya Hospital of Central South University, Changsha, Hunan Province, China, Tel: +86-13204043733, Fax: +86-13707313601, E-mail: zhq9996@163.com

Received 2024 August 1; Revised 2024 October 5; Accepted 2024 October 13.

Abstract

Introduction

Adolescent idiopathic scoliosis (AIS) is a three-dimensional malformation of the spine that is frequently found in adolescents aged 10–18 years [1]. If not addressed early, the malformation could affect the physical and mental appearance of the patient. The three-dimensional malformation involves sagittal, coronal, and axial [2].

According to Lenke’s classification, the nonstructural thoracic and structural thoracolumbar/lumbar curves of Lenke’s 5C surgical interventions remain controversial [3]. Over the years, different techniques and approaches have been used to correct this malformation. However, the anterior approach was the most acceptable for correcting such malformations because it could preserve more segments for mobility and provide satisfactory follow-up outcomes for a long time [4]. The posterior approach is common with modern techniques, and many authors have presented different results [4,5].

However, this approach still has challenges because of the lower instrumented vertebra (LIV) selection and instrumentation length. Previous studies have focused on decreasing fusion levels to save more segments for mobility and efficient coronal balance and avert coronal decompensation to treat Lenke 5C [6,7].

Over the years, Lenke 5C correction has raised several unanswered questions. These include the selection of the preferred level of instrumentation with the posterior approach [8–10]. In addition, some questions remain about the cosmetic appearance based on the patient’s postsurgical outcomes [11].

In previous studies, the nonstructural thoracic curve progression and shoulder imbalance have been reported in Lenke 5C correction [12]. Rib hub is also one of the main concerns in patients with AIS.

However, one of the techniques to address this is thoracoplasty or direct vertebrae rotation (DVR) using all pedicle screw placement en-bloc or selective placement [13]. However, DVR increases surgical blood loss and surgical time despite studies about the effect of this procedure on the rib hub and radiological measurements [14]. Subsequently, given the increase in cosmetology awareness and effect on patient mental strength, no study has compared the effect of en-bloc DVR with non-DVR for Lenke 5C correction. Therefore, this study aimed to compare the clinical efficiency of en-bloc DVR to non-DVR for Lenke 5C correction.

Materials and Methods

A retrospective study of 200 patients with Lenke 5C AIS who underwent posterior-only approach spinal fusion in Xiangya Hospital Central South University between 2006 and 2020. The ethical committee for Xiangya Hospital Central South University approved the study (21017033559). After obtaining institutional review board approval and the patient’s informed consent, patient data were collected under the same surgeon’s directives. Furthermore, all methods were performed per our hospital’s relevant guidelines and regulations.

The inclusion selection criteria were as follows: patients with AIS Lenke 5C curves with the posterior-only approach, a minimum of 4 years of follow-up, age between 10 and 18 years, and history of a corrective procedure with en-bloc DVR (ED) and non-DVR (ND). The exclusion criteria were as follows: patients with a prior form of spinal fusion or spinal cord malformation, age <18 years, and use of different approaches.

During selection, 95 patients met the criteria above; however, 105 were excluded because they did not meet the inclusion criteria. To ensure equal representation of the two surgical techniques, a stratified random sampling approach was used. During the random selection, the LIV selection criteria for the ED group were LIV=LEV 1, and for the ND group, LIV=LEV. Both groups have their upper instrumented vertebra (UIV) at the upper-end vertebrae.

Surgical techniques

Patients were placed in a prone position after satisfactory anesthesia. Anatomical spine exposure was performed by a midline incision employing a subperiosteal dissection of the paraspinal muscles. The screws were positioned in an anatomical position once the location of the bilateral spinal pedicles was confirmed. All patients underwent freehand screw fixation. However, the level of instrumentation differed in both groups. All patients had apical posterior release before corrective procedures (rod derotation by 90° of the concave rods, minor undercontouring of the convex one): flavectomy and Ponte osteotomy. In the ED group, apical DVRs were administered to patients using the SmartLink Vertebral Manipulator device (SmartLink; Medtronic, Minneapolis, MN, USA) (Fig. 1). Over the apex screws, with the level above and below, the SmartLink manipulator device was mounted (three levels) at the levels undergoing DVR. After the SmartLink device construct was put together, a powerful derotation was carried out en bloc using three levels joined by a stiff construct, resulting in an even distribution of the derotation force across the entire apex.

Fig. 1

(A, B) Showing the correction force of Smartlink vertebral manipulator device (Medtronic, USA) for the correction of Lenke 5C.

During derotation, motor-evoked potentials and somatosensory-evoked potentials were utilized. Distraction was also conducted on the concave side following single rod rotation in the non-DVR group. On the convex side, compression was performed after inserting the implant rod for the correction maneuvers. Tightening was done to the two-step securing caps. The laminae and transverse processes that were removed were used, as well as the allograft bone material, to fuse the bones in both groups.

Results

Ninety-five patients with AIS Lenke 5C met the inclusion criteria. The ED group consisted of 45 patients (10 males and 35 females), and the ND group had 50 patients (20 males and 30 females). The average ages at surgery were 15.0±1 years for the ED group and 14.3±1 years in the ND group. According to the comparison of these two groups, no significant differences were found in the demography of patient data, such as age, Risser sign, and follow-up duration (Table 1). Furthermore, no differences were found in the surgical duration, with the ED group having an average time of 264.7±29 minutes, compared with the ND group with a surgical time of 277.3±52 minutes (p=0.147). The ED group had approximately lesser blood loss, with an average mean of 460±87 mL compared with the ND group, with an average surgical blood loss of 468±88 mL (p=0.637). Moreover, no significant differences were found in the level of Ponte osteotomy; however, substantial differences were noted in the fused segments, with the ED group having an average mean of 5.6±0 and group ND with 6.3±0 (p=0.001). There were more extended hospital stays in the ND group than in the ED group, averaging (ED 9.7±1 days versus ND 10.8±1 days, p=0.006).

Table 1

Patient’s characteristics and surgical demography

Table 2 shows the comparison of the radiographic data between the two groups. Both groups had no significant differences in preoperative, postoperative, and last follow-up radiography measurements regarding TK, lumbar lordosis, C7–CSVL, and SVA.

Table 2

Comparison of radiography parameters

No significant differences were found in the preoperative nonstructural thoracic curves between the two groups (ED group, 16.6°±4°; ND group, 17.0°±3°; p=0.605). However, after 6 months of postoperative follow-up, significant differences were observed in the nonstructural thoracic curves between the two groups. The ED group demonstrated better nonstructural thoracic curve correction than the ND group (ED group, 3.7°±3°; ND group, 5.6°±5°; p=0.036). At the final follow-up, both groups exhibited significant differences (ED 4.6°±40° versus ND 6.9°±40°, p=0.002). However, no significant differences were noted in the major curve between the groups preoperatively (ED 52.2°±20° versus ND 52.8°±30°, p=0.327). Despite this, the ED group showed a superior correction rate during the follow-up evaluation compared with the ND group (ED 4.6°±20° versus ND 6.1°±20°, p=0.005).

Radiographic complications, such as AO, were significant at the 6-month follow-up, with the ED group having better stable end vertebrae; however, no significant differences were found between the two groups at the last follow-up. Furthermore, postjunctional kyphosis (PJA) was not significant in either group.

The preoperative LIV tilts were insignificant in both groups; however, 6 months after surgery, substantial differences were found in both groups (ED 0.3°±0° versus ND 0.8°±0°, p=0.001).

Axial vertebrae rotation (RAsag) evaluation showed significant maneuver in the ED group compared with the ND group. Preoperatively, no significant differences; however, ED showed compelling maneuver of RAsag compared with non-DVR, with significant differences in the ED group (8.4°±0°) and ND group (11.1°±1°) postoperatively (p=0.001).

Fig. 2 shows that RSH was minimal before surgery and at the end of the last follow-up, and neither group had apparent differences (ED 1.6±1 versus ND 1.8±1, p=0.671).

Fig. 2

(A, B) Showing the radiography shoulder height (RSH) of both groups. (A) Group ED: en-bloc direct vertebrae rotation (DVR) group. (B) Group ND: non-DVR group. Preop, preoperative; Postop, postoperative; FU, follow-up.

The results of the SRS-22 questionnaire given before surgery in both groups were not significant. However, at the last follow-up visitation, a considerable improvement was noted in both groups, and substantial differences in satisfaction were found in the ED group compared with the ND group (Table 3). Furthermore, no intraoperative complication was recorded, and during the last follow-up, no patient developed PJK and AO (Figs. 3, 4).

Table 3

The Scoliosis Research Society-22 questionnaire

Fig. 3

A 17-year-old patient with adolescent idiopathic scoliosis Lenke 5C. (A, B) Preoperative posteroanterior and lateral standing radiograph. (C, D) Immediate postoperative posteroanterior and lateral standing radiograph. (E, F) 48 months postoperative posteroanterior and lateral radiograph. She was treated with posterior en-bloc direct vertebrae rotation. RSH, radiography shoulder height; C.Align, coronal alignment; TL/L, thoracolumbar/lumbar; AO, adding-on; PO, pelvic oblque; LL, lumbar lordosis; TK, thoracic kyphosis; SVA, sagittal vertical axis.

Fig. 4

A 15-year-old patient with adolescent idiopathic scoliosis Lenke 5C. (A, B) Preoperative posteroanterior and lateral standing radiograph. (C, D) Immediate postoperative posteroanterior and lateral standing radiograph. (E, F) 50 months postoperative posteroanterior and lateral radiograph. She was treated with posterior non-direct vertebrae rotation. RSH, radiography shoulder height; C.Align, coronal alignment; AO, adding-on; PO, pelvic oblique; LL, lumbar lordosis; TK, thoracic kyphosis; SVA, sagittal vertical axis; DJA, distal junctional angle; TL/L, thoracolumbar/lumbar.

Discussion

Posterior-only spinal instrumentation is one of the most popular techniques for AIS correction due to the intervention of modern technology equipment [10]. Although surgery for adolescent scoliosis generally aims to achieve balanced sagittal and coronal alignment, shoulder asymmetry, which may reduce patient happiness, is connected to a residual imbalance in the coronal plane and is a crucial aspect of evaluating the functional outcomes of patients with AIS [15]. Shoulder imbalance in AIS Lenke 1 and 2 is common, and massive studies have been published about the management and improvement of the outcomes [16,17]. On the contrary, AIS Lenke 5C shoulder imbalance is little known; few papers have reported postsurgical functional outcome [18]. Shoulder balance following surgical correction affects self-image and strongly correlates with subjective postoperative outcomes [15]. Achieving shoulder balance in patients with AIS is considered one of the key indicators of a good surgical outcome based on radiographic measures, aside from a potential association with clinical outcomes [15].

AIS Lenke 5C is a nonstructural thoracic curve and structural thoracolumbar/lumbar curve; its progression is massive because of the significant effect on patient mobility and stability. Over the years, the DVR technique has been used for the correction of AIS deformity [6,14,19]. This technique complements AIS correction. From a technical point of view, the DVR can be used for either selective or en-bloc instrumentation, contrary to previous studies, which highlighted the effect of DVR derotation radiography outcomes [6,19,20]. Despite studies focusing on the efficiency of these techniques on different AIS types, few studies have focused mainly on AIS Lenke 5C DVR efficiency. Previous studies combined both selective and en-bloc DVR instrumentation maneuvers [20–22]. Nevertheless, this is the first study that fully deployed and compared the radiography and clinical efficiency of ED versus non-DVR techniques to correct AIS Lenke 5C.

This study aimed to determine the influence of DVR on AIS Lenke 5C correction and identify the conditions under which it may provide optimal spinal deformity correction. The focus on patients aged 10–18 years is crucial, as this period encompasses the growth spurt before skeletal maturation, allowing for a valuable assessment of the efficacy of both surgical techniques. The study included 45 patients in the ED group and 50 patients in the non-DVR (ND) group, making it one of the largest comparative studies on these techniques for AIS Lenke 5C correction. By utilizing a comprehensive set of radiographic measures, this study seeks to provide insights into the effectiveness of en-bloc DVR relative to non-DVR methods in achieving spinal correction.

The pedicle screw is the hallmark of the posterior approach for AIS correction [23]. While pedicle screws appear to improve the correction in the coronal plane, most patients with AIS have intrinsic hypokyphosis in the thoracic region, which appears to have a negative effect on the sagittal profile—decreased TK. In a previous study, the use of en-bloc DVR for scoliosis correction was attributed to hypokyphosis after surgery [14,20]. Conversely, in this study, the patient’s TK with an average mean of (preoperative 26.9°±19° and last follow-up 30.0°±7°) and group ND with an average mean of (preoperative 25.7°±8° and last follow-up 30.9°±9°). No differences were found between the two groups. Moreover, a recent study reported that DVR increased TK [20]. In contrast, in this study of en-bloc DVR and non-DVR, no significant differences were found between the two groups.

Recent studies have highlighted an increased complication rate in patients undergoing DVR [21]. Research has shown that the substantial forces applied to the screws during DVR can potentially lead to screw loosening or even pedicle fracture, resulting in screw pull-out and implant migration [24]. In contrast to these findings, our studies on en-bloc DVR were designed to reduce the stress on the vertebral pedicle during rotation and limit the mechanical stress on the patient’s vertebrae. This approach is particularly advantageous for patients with AIS, particularly women, who often have a low bone mass density [25,26]. En-bloc DVR strengthens the vertebrae and minimizes mechanical stresses during deformity correction. Nevertheless, in this study, no complications were reported in either group.

A biomechanical study using DVR compared with non-DVR reported that DVR is more efficient with axial rotation than non-DVR. According to biomechanical research, axial rotation in scoliosis is a fundamental component of the deformity and contributes to the coronal and sagittal components. The “coupling” of rotation and translation between the anatomical axes describes this phenomenon. Based on coupled spine motions, three-dimensional correction with DVR appears to be an apparent part of correcting scoliosis and should result in a better overall result [14,20]. In this study with en-bloc DVR and non-DVR, the EV group had better derotation than the ND group. In contrast to previous studies without significant differences [20], our study produced superior axial rotation with differences postoperatively.

In this study, the preservation of more lumbar segments for mobility and flexibility is one of the essential parts for AIS Lenke 5C correction; thus, the selection for LIV for the ED group LEV-1 compared with the ND group LIV selection of LEV=LIV. Consequently, there was lesser instrumentation, and more segments persevered for mobility. Therefore, the fusion level in the ND group was significant to that in the ED group, with an average fusion level of 5.6±0 in the ED and ND 6.3±0 (p=0.001). In the ED group, en-bloc fusion is more convenient and less risky than that in the ND group, preserving more lumbar motion with less trauma. Despite the selection criteria for the LIV in both groups, preoperative films with posteroanterior, sagittal, and bending films during preoperative preparation. We are confident that the ED group could generate more force for derotation and correction during the operation than the ND group.

Coronal and sagittal balance is the fundamental aim of AIS correction; in this study, ED provided better coronal improvement than the non-DVR. Previous authors reported a similar correction rate with the use of DVR [6,14,20]. In the present study, we also tried to evaluate the correlation between the coronal correction and major curve derotation although no correlation was established.

The correlation was evaluated to check the effect of derotation and coronal correction on the shoulder balance of patients with Lenke 5C after surgery. Shoulder imbalances in Lenke 5C have been reported, and previous studies have also reported different ways to counter this postsurgical complication [22,27]. Some studies have even proposed the extension of UIV beyond the UEV level, and some studies in AIS have shown that, on average, after selective fusion of structural curves and spontaneous correction of thoracic compensatory curves will take place, leading to improved shoulder balance [6,20,28]. Some studies have even stated the lack of differences in the ED and non-DVR groups for the correction of the nonstructural thoracic curve even though their study was at a minimum of 2 years follow-up [14,20,21,29]. However, contrary to their early report, this study showed a significant difference in the compensatory thoracic curve in the ED group than in the ND group at 6-month postoperative visitation. However, the compensatory thoracic curve was maintained in the ED group more than that in the ND group during the follow-up visit, with a minimum of 4 years of follow-up.

Cases of increases in PJA were reported, which could lead to postjunctional kyphosis in patients with an increase in the UIV level in Lenke 5C [30]. However, no postoperative complications such as PJK were reported in either group. The PJA of both groups at the preoperative, postoperative, and last follow-ups were insignificant.

The clinical importance of shoulder imbalance in AIS Lenke 5C has received circumscribed attention [6,19,31]. In this study, the RSH levels in both groups were compared; nevertheless, both groups had no significant differences at the last follow-up.

Spinal balance is imported because it goes a long way to ensure a better mental state for patients, particularly adolescents, who are afraid of how people around them see them and fear being unable to participate in what their friends do. Due to this factor, the SRS-22 questionnaire was carried out preoperatively and at the last follow-up visitation; however, no differences were found in both groups apart from the last follow-up satisfaction, that is, the ED group had a better satisfaction outcome than the ND group.

This study is limited by its retrospective nature and short follow-up period of 4 years. Conversely, because the radiographic results are primarily of interest to surgeons but of less significance to patients, the study’s inclusion solely of radiographic data may be considered a constraint. Furthermore, prospective studies are needed to further evaluate the clinical efficiency of en-bloc DVR to correct patients with patients AIS Lenke 5C.

Conclusions

Compared with non-DVR methodologies, the ED exhibits superior efficacy in correcting the RAsag, thoracolumbar/lumbar curves, and thoracic compensatory curves associated with AIS Lenke 5C. Furthermore, the data indicated that the ED group had reduced instrumentation and experienced shorter hospitalization than the ND group, all while avoiding postoperative complications.

Article information Continued