The spontaneous correction of an unfused lumbar curve after instrumentation and fusion of a major thoracic curve was first reported by Moe in 1958.1 Since then the concept of selectively fusing the major structural curve in a double curve pattern in adolescent idiopathic scoliosis (AIS) has gained momentum. Numerous reports have investigated and defined the parameters, which may lead to a predictable correction of the unfused secondary curve.2–11 However, a majority of these reports have concentrated on the Cobb angle correction in the coronal plane and neglected the rotational deformity in the axial plane.
To our knowledge, there are no reports of axial plane derotation of an unfused secondary curve after a posterior spinal instrumentation. Given that the rotational deformity associated with AIS (the rib hump or lumbar prominence) is often the major cosmetic concern recognized by patients and families in the setting of scoliosis12,13; this is a shortcoming of previous reports. Thus, the purpose of this study was to evaluate spontaneous axial plane, rotational correction of the unfused curve in patients undergoing selective fusion for AIS. Both spontaneous lumbar prominence correction after selective thoracic fusion and spontaneous thoracic rib prominence correction after selective lumbar fusion were evaluated to enable some prediction of improvement in the clinical deformity after selective fusions.
Materials and Methods
A prospectively collected multicenter database of patients undergoing treatment for AIS was examined to analyze all patients who underwent selective fusion of double curve patterns. Only patients in whom the preoperative secondary curve’s angle of trunk rotation or angle of lumbar prominence was at least 5° (as measured by scoliometer) were included for review. In other words, patients who had no clinically significant rotational deformity of the secondary curve before surgery were excluded from this study. Scoliometer measurements were performed by placing the scoliometer at the level of the most prominent thoracic and lumbar hump during the Adam’s forward bend test.14
Two cohorts of patients were available for review–those who underwent selective thoracic fusion and those who underwent selective lumbar fusion. Selective thoracic fusion (STF) was defined as double curves treated with thoracic instrumentation and fusion in which the lowest instrumented vertebrae was L2 or above. Selective lumbar fusion was defined as double curves treated with selective lumbar/thoracolumbar fusion in which the upper instrumented vertebrae was at T9 or below. Measurements of both the primary fused and secondary unfused curves were recorded at both the preoperative and 2-year postoperative intervals. In addition, demographic data, Lenke classification, and pre and 2-year postoperative Cobb angle measurements and percent correction of both the primary (fused) and secondary (unfused) curves were recorded.
Statistical Package for the Social Sciences software (SPSS Inc., Chicago, IL) was used to conduct the analyses. The data were checked for normality and equal variances and the alpha level was set at 0.05. A univariate analysis of variance was used to compare age, gender, Risser sign, operative time, estimated blood loss, and percent coronal and axial correction, whereas a Pearson χ2 analysis was used to compare the categorical variables of Lenke curve type and surgical approach between the STF and selective lumbar fusion (SLF) groups. Additionally, a repeated measures analysis of variance was used to compare pre and 2-year postoperative Cobb measurements and scoliometer readings of the fused and unfused curves within each surgical group.
Demographic data for patients in the 2 surgical groups (STF and SLF) are presented in Table 1. There were no statistically significant differences found with regards to age, gender, Risser sign, operative time, and estimated blood loss (P > 0.17). However, Lenke curve type and surgical approach were found to be significantly different (P = 0.001). Only thoracic major curves (Lenke curve types 1 through 4) were included in the STF group, whereas thoracolumbar/lumbar major curves (Lenke curve types 5 and 6) were included in the SLF group. In the STF group, 9 patients had an open anterior spinal fusion, 26 had a thoracoscopic anterior spinal fusion, and 48 had a posterior spinal fusion. In the SLF group, 27 patients had an open anterior spinal fusion and 1 had a posterior spinal fusion.
The STF group comprised 83 patients who were instrumented to a lowest instrumented vertebrae of L2 or above with an average clinical and radiographic follow-up of 2.4 ± 0.5 years. Lenke curve types in the STF group comprised the following: 69 Lenke 1, 9 Lenke 2, 4 Lenke 3, and 1 Lenke 4. Preoperative and 2-year postoperative thoracic Cobb angles averaged 53° ± 12° and 18° ± 9°, with a mean correction of 65% for the instrumented thoracic segment. Preoperative and 2-year postoperative lumbar Cobb angles averaged 33° ± 10° and 16° ± 9° with a mean spontaneous correction of 53% for the uninstrumented lumbar segment. The 2-year postoperative correction for both the instrumented primary curve and uninstrumented secondary curve were statistically significant (P ≤ 0.001).
The preoperative and 2-year postoperative angle of trunk rotation (as measured by scoliometer) of the instrumented thoracic rib hump were 15° ± 5° and 7° ± 4°, respectively with a mean correction of 51% (P ≤ 0.001). Similarly, the preoperative and 2-year postoperative angle of trunk rotation (as measured by scoliometer) of the uninstrumented lumbar prominence were 9° ± 3° and 4° ± 3°, respectively with a mean correction of 49% (P ≤ 0.001) (Table 2). Of note, 29 of these 83 patients (35%) had a concomitant thoracoplasty at the time of the STF. The average 2-year postoperative thoracic rib hump values, however, were found to be identical between those patients that had a thoracoplasty and those that did not (7° ± 4°); with a statistically nonsignificant difference in average percent correction of the thoracic rib hump (51% with a thoracoplasty vs. 49% without a thoracoplasty; P = 0.82).
The SLF group comprised 27 patients who were instrumented to an upper instrumented vertebrae of T9 or below who had an average clinical and radiographic follow-up of 2.3 ± 0.5 years. Twenty-three patients in the SLF group had a Lenke 5C curve patterns, whereas 4 patients had a Lenke 6C curve type. Preoperative and 2-year postoperative lumbar Cobb angles were 48° ± 10° and 13° ± 9° with a mean correction of 73% for the instrumented lumbar/thoracolumbar segment. Preoperative and 2-year postoperative thoracic Cobb angles were 31° ± 12° and 18° ± 9° with a mean correction of 41% for the uninstrumented thoracic segment. The 2-year postoperative correction for both the instrumented primary curve and uninstrumented secondary curve were statistically significant (P ≤ 0.001).
The preoperative and 2-year postoperative angle of trunk rotation (as measured by scoliometer) of the instrumented lumbar prominence were 11° ± 5° and 3° ± 3°, respectively with a mean coronal plane correction of 66% (P ≤ 0.001). The preoperative and 2-year postoperative angle of trunk rotation (as measured by scoliometer) of the uninstrumented thoracic rib hump were 8° ± 3° and 6° ± 3°, respectively with a mean correction of only 26%. This was not a statistically significant change (P = 0.14) (Table 2). Five of these 27 patients (19%) had a concomitant thoracoplasty performed at the time of the SLF. However, the average 2-year postoperative scoliometer measurement and average percent correction of the uninstrumented thoracic rib hump were not different in those patients with and without the thoracoplasty (6° ± 3° and 26%, in both groups, respectively) (P = 0.93).
Despite similar coronal plane (Cobb angle) correction of the uninstrumented secondary curve (53% spontaneous lumbar curve coronal plane correction in the STF group and 41% spontaneous thoracic curve coronal plane correction in the SLF group; P = 0.12), the spontaneous axial plane rotational correction of the unfused lumbar prominence in patients undergoing STF (49%) was significantly greater than the spontaneous axial plane rotational correction of the unfused thoracic rib hump in patients undergoing SLF (26%) (P = 0.04).
A substantial body of literature exists validating the phenomenon of spontaneous coronal plane correction of a secondary uninstrumented curve in selective fusions for AIS.2–11,15 Most of these reports emphasize the coronal plane and do not address the axial plane rotational correction of the rib or lumbar prominence. This emphasis on coronal plane correction does not completely mirror patients’ perception of successful surgery, as trunk shape improvement afforded by axial plane correction of the rib hump is also considered important to patients.12,13 Accordingly, it is valuable to provide insight into the anticipated spontaneous axial plane correction imparted to the unfused secondary curve during selective fusions of double curve patterns in AIS. These data should facilitate communication of postoperative expectations with patients and their families and may, in some cases, facilitate decision-making in determining whether to include or exclude a secondary curve in a fusion.
In regard to coronal plane correction, this study’s outcomes correlate favorably with published data for both selective thoracic fusions (65% vs. 38%–68% Cobb angle correction of instrumented thoracic curve, and 53% vs. 38%–56% spontaneous Cobb angle correction of the uninstrumented lumbar curve)1,3,5,7,9,11 and selective lumbar fusions (73% vs. 54%–83% Cobb angle correction for the instrumented lumbar curve, and 41% vs. 19%–42% spontaneous Cobb angle correction of the uninstrumented thoracic curve).4,6,8,10,11,16 In doing so, these results further substantiate the phenomenon of spontaneous coronal plane correction of an unfused compensatory curve after a selective fusion. Additionally, this study’s results also compare favorably with those published reports that do address axial plane correction of the instrumented segment (51% vs. 22%–70% rib hump correction in STF, and 66% vs. 51%–87% lumbar prominence correction in SLF).4,8,9,11,16
The emphasis of this series, however, was to demonstrate that spontaneous compensatory curve correction is not only limited to the coronal plane, but that spontaneous axial plane derotation of the uninstrumented curve also occurs during selective fusions. In selective thoracic fusions, the axial plane correction of the instrumented thoracic rib prominence nearly equaled that of the uninstrumented lumbar prominence (51% and 49%, respectively). In selective lumbar fusions, however, significantly less compensatory axial plane correction was imparted on the uninstrumented thoracic rib prominence than that achieved in the instrumented lumbar segment (26% and 66%, respectively). Thus, significantly less axial plane derotation is transmitted from the instrumented to the uninstrumented segment in selective lumbar fusions than in selective thoracic fusions.
The explanation for this discrepancy is hypothesized to be related to the rigidity of the thoracic spine and rib cage in comparison with the lumbar spine. In the case of selective thoracic fusions, the rigidity of the rib-vertebral complex, when instrumented, serves as a powerful construct that imparts corrective forces on the less constrained lumbar spine. On the other hand, the less constrained lumbar spine, when instrumented, imparts less corrective forces on the more rigid rib-vertebral complex of the thoracic spine. Of note, this discrepancy is also present in the coronal plane correction—spontaneous lumbar curve correction was 53% of that attained in the instrumented thoracic spine in STF, whereas spontaneous thoracic curve correction was only 41% of that attained in the instrumented lumbar spine in SLF.
Numerous methods have been developed to assess axial plane rotation in scoliosis.17–21 This prospective study, however, relied solely on clinical examination determination of the angle of trunk rotation during the Adam forward bend test14 using a scoliometer measurement. Although this is a possible criticism of this report as there is some potential for clinician error in attaining these values, scoliometer assessment is a frequently used, reproducible, and simple technique generally accepted and reported in scoliosis literature. Plain radiographic techniques such as Nash-Moe and Perdriolle are difficult to apply after posterior spinal surgery as instrumentation may obscure visualization of the vertebral landmarks. Computed tomography and magnetic resonance assessments of vertebral rotation are very accurate; however, because of expense, high radiation doses (computed tomography), and the requirement for pre and postoperative studies for comparison, these studies were not routinely obtained in the prospective AIS protocol. Additionally, although digital radiometric rotation analysis and rasterstereography both have merit in the assessment of rotation, Schulte et al 11 demonstrated that both of these methods correlated well with clinical scoliometer measurements. Thus, because of widespread acceptance, ease of application, and documented correlation with other methods, we chose to use clinical scoliometer measurements to evaluate rib and lumbar prominence derotation after surgery.
The inclusion of patients who underwent thoracoplasty at the time of selective fusion could potentially impart some bias to our results. In the case of STF, thoracoplasty of the instrumented thoracic segment would have no biomechanical sequelae on the spontaneous coronal or axial plane correction of the unfused lumbar curve; therefore these cases were not excluded. Interestingly, the additional morbidity of a thoracoplasty imparted no significant improvement in the rib hump deformity of the instrumented segment when compared with those patients who underwent STF with (51%) and without a thoracoplasty (49%) (P = 0.82). In the case of SLF, the addition of a thoracoplasty to the lower thoracic levels included in the fusion also imparted no measurable improvement in the uninstrumented thoracic rib hump (26% correction in patients with and without a thoracoplasty) (P = 0.93). Accordingly, the 5 patients in the SLF group were not excluded from this study as no difference was measured; therefore no bias imparted. Thus, in this series, thoracoplasty, a procedure with primary cosmetic goals and significant morbidity, did not achieve measurable improvements in the axial plane rib hump deformity raising question regarding its value.
Spontaneous correction of the uninstrumented secondary curve occurs in both the coronal and axial plane in selective fusions of double curve patterns in adolescent idiopathic scoliosis. Significantly more correction is transmitted in both of these planes from the instrumented thoracic spine to an uninstrumented, compensatory lumbar curve with a STF than is transmitted from the instrumented lumbar spine to an uninstrumented, compensatory thoracic curve after a SLF. In the case of selective thoracic fusions, 53% and 49% correction of the uninstrumented lumbar curve occurred in the coronal plane (Cobb angle) and axial plane (lumbar prominence), respectively. In the case of selective lumbar fusions, less correction was imparted to the uninstrumented thoracic curve as only 41% and 26% correction was spontaneously achieved in the coronal plane (Cobb angle) and axial plane (rib prominence), respectively.
- Spontaneous correction of the uninstrumented secondary curve occurs in both the coronal and axial plane in selective fusions of double curve patterns in adolescent idiopathic scoliosis.
- Selective thoracic fusions result in an approximate 50% spontaneous reduction in the clinical lumbar prominence.
- Significant spontaneous correction of a thoracic rib hump should not be anticipated after a selective lumbar fusion.
- These data could facilitate communication of postoperative expectations with patients and their families in regards to correction of the truncal deformity, and may, in some cases facilitate decision-making in determining whether to include or exclude a secondary curve in a fusion.
1. Moe J. A critical analysis of methods of fusion for scoliosis: an evaluation in two hundred and sixty-six patients. J Bone Joint Surg Am
2. Lenke L, Edwards C, Bridwell K. The Lenke classification of adolescent idiopathic scoliosis
: how it organizes curve patterns as a template to perform selective fusions
of the spine. Spine
3. Bullmann V, Halm H, Lepsien U, et al. Selective ventral derotation spondylodesis in idiopathic thoracic scoliosis: a prospective study. Z Orthop Ihre Grenzgeb
4. Burwell R, Aujla K, Cole A, et al. Anterior universal spine system for adolescent idiopathic scoliosis
: a follow-up study using scoliometer, real-time ultrasound and radiographs. Stud Health Technol Inform
5. Edwards C, Lenke L, Peelte M, et al. Selective thoracic fusion of adolescent idiopathic scoliosis
with C modifier lumbar curve: 2- to 16-year radiographic and clinical results. Spine
6. Halm H, Niemeyer T, Link T, et al. Segmental pedicle screw instrumentation in idiopathic thoracolumbar and lumbar scoliosis. Eur Spin J
7. Lenke L, Betz R, Bridwell K. Spontaneous lumbar curve coronal correction after selective anterior or posterior thoracic fusion in adolescent idiopathic scoliosis
8. Min K, Hahn F, Ziebarth K. Short anterior correction of the thoracolumbar/ lumbar curve in King I idiopathic scoliosis: the behaviour of the instrumented and non-instrumented curves and the trunk balance. Eur Spin J
9. Potter B, Kuklo T, Lenke L. Radiographic outcomes of anterior spinal fusion versus posterior spinal fusion with thoracic pedicle screw for treatment of Lenke type I adolescent idiopathic scoliosis
10. Sanders A, Baumann R, Brown H, et al. Selective anterior fusionof thoracolumbar/lumbar curves in adolescents: When can the associated throacic curve be left unfused? Spine
11. Schulte T, Liljenquvist U, Hierholzer E, et al. Spontaneous correction and derotation of secondary curve after selective anterior fusion of idiopathic scoliosis. Spine
12. Theologis T, Jefferson R, Simpson A, et al. Quantifying the cosmetic defect of adolescent idiopathic scoliosis
13. Weatherley C, Draycott V, O’Brien J, et al. The rib deformity in adolescent idiopathic scoliosis
. A prospective study to evaluate changes after Harrington distraction and posterior fusion. J Bone Joint Surg Br
14. Adams W. Lectures on the Pathology and Treatment of Lateral and Other Forms of Curvatures of the Spine
. London, 1865.
15. Willers U, Transfeldt E, Hedlund R. The segmental effect of Cotrel-Dubousset instrumentation on vertebral rotation, rib hump, and the thoracic cage in idiopathic scoliosis. Eur Spin J
16. Pratt R, Webb J, Burwell R, et al. Changes in surface and radiographic deformity after universal spine system for right thoracic adolescent idiopathic scoliosis
: is rib-hump reassertion a mechanical problem of the thoracic rib cage rather than an effect of relative anterior spinal overgrowth. Spine
17. Nash C, Moe J. A study of vertebral rotation. J Bone Joint Surg Am
18. Perdriolle R, Vidal J. Morphology of scoliosis: three-dimensional evolution. Orthopaedics
19. Kuklo T, Potter B, Lenke L. Vertebral rotation and thoracic torsion in adolescent idiopathic scoliosis
: what is the best radiographic correlate? J Spinal Disord Tech
20. Aaro S, Dahlborn M. Estimation of vertebral rotation and the spinal and rib cage deformity in scoliosis by computer tomography. Spine
21. Birchall D, Hughes D, Hindle J, et al. Measurement of vertebral rotation in adolescent idiopathic scoliosis
using three-dimensional magneteic resonance imaging. Spine