SAGITTAL BALANCE IN ADOLESCENT IDIOPATHIC SCOLIOSIS

Objective: The objective of this study is to achieve a better understanding of the parameters that influence sagittal balance in a population with adolescent idiopathic scoliosis (AIS). Methods: A retrospective study of 80 patients with adolescent idiopathic scoliosis (AIS) was conducted. The parameters evaluated were: age, sex, pelvic incidence (PI), sacral slop (SS), pelvic tilt (PT), sagittal balance (SB), coronal balance (CB), lumbar lordosis (LL), thoracic kyphosis (TK) divided into upper (between T1 and T5) and lower (between T5 and T12), cervical spine alignment (CSA), and Cobb’s coronal angle (CCA) of primary scoliotic curvature. Results: Regarding the sagittal balance, this study demonstrated a significant statistical positive correlation with cervical shape (p<0.01) and upper thoracic kyphosis (from T1 to T5) (p<0.05), but not with the other variables. LL had a strong influence on lower thoracic curvature (from T5 to T12) and was strongly influenced by the PI and SS. Conclusions: Sagittal balance is a parameter that is influenced by multiple factors. In fact, it is closely related to cervical shape and the upper thoracic curvature (from T1 to T5), which in turn, is closely linked to the severity of the scoliotic kyphosis. The Cobb angle of the lower thoracic spine (from T5 to T12) is more closely correlated with the angle of lumbar lordosis than with the upper thoracic kyphosis (from T1 to T5). Level of evidence IV; Case Series.


INTRODUCTION
The human standing posture involves a delicate balance between the spine and the pelvis. Although all primates are able to place themselves in the orthostatic position, Homo sapiens is the only able to travel long distances and steadily in this way. 1 The reason for this occurrence is that humans have a unique lumbar lordotic curvature. A balanced posture is obtained when these body segments are aligned in a way that minimizes energy expenditure. [2][3][4] The importance of sagittal spinopelvic organization in spine pathology, and the crucial need to restore it after surgical treatment, are now recognized in the literature. 5 Some authors have reported the normal values of sagittal spinal curves in children and adolescents. 6,7 Other studies have also characterized the pelvic orientation and morphology in the sagittal plane of paediatric subjects. 8,9 Upasani et al. presented a relation between lumbar spine lordosis and pelvic geometry. Effectively, the lumbar spine has a close relationship with the pelvic orientation, expressed in particular by the sacral slope, which is influenced by the pelvic incidence (PI). 10 The sagittal curvature of the spine and the pelvic balance swing together to maintain a stable posture and horizontal gaze. Once this sagittal alignment is lost, more energy is needed for the body to remain balanced without any external support. 11 The pelvic incidence is unique in each individual, and is independent of the spatial alignment of the pelvis. Therefore, it reflects the anatomy of the pelvis and does not modify with changes in pelvic or spinal position. [12][13][14][15][16] Several studies have shown that PI is closely related to the sacral slope (SS) and pelvic tilt (PT), and that it remains relatively constant during childhood. It increases significantly during adolescence, reaching its maximum value in adulthood. 17 Others defend the view that there is no correlation between the PT and SS in patients with AIS.
Idiopathic scoliosis is a three-dimensional deformity of the torso consisting of lateral curvature of the spine and vertebral rotation. 18 Satisfactory treatment for AIS includes maximum correction in the coronal plane, in addition to adequate restoration of the sagittal spinal alignment and vertebral rotation. Many studies [19][20][21][22] have demonstrated that sagittal balance, rather than coronal balance (CB), is significantly correlated with health-related quality of life (HRQOL). Others have found great impacts of lumbar lordosis (LL) and thoracic kyphosis (TK) on the maintenance and prediction of sagittal balance, which are considered novel regional predictors for sagittal balance. 23,24 The C7 plumb line is the most commonly used index of global balance. This parameter has been historically quantified by measuring the position of a vertical line originating in the center of the C7 vertebral body with respect to the posterior superior corner of S1. [25][26][27] In a systematic literature review, Kuntz et al. 28 noted that as a global parameter, the C7 plumb line was a stable, reliable index of sagittal balance, being maintained in narrow ranges for alignment of the spine over the pelvis and femoral heads. Furthermore, the morphology of the pelvis, and the position of the pelvis relative to the spine, may also influence sagittal spinal alignment and balance. [29][30][31][32][33] Glassman et al. 34 reported that a positive sagittal balance is related to symptomatology and quality of life in patients with spinal deformities. Berthonnaud et al. 2 have proposed the concept of a linear chain linking the head to the pelvis, in which the shape and orientation of each anatomic segment are closely related and influence the adjacent segment to maintain a stable posture with minimum energy expenditure. Accordingly, a change in the shape or orientation of any anatomic segment will modify the shape and/or orientation of adjacent segments of the spine and pelvis. The differences between normal and pathologic curvatures are less clear in the sagittal plane than in the coronal plane. 35

Population
The sample was composed by 80 patients. Inclusion criteria were patients with a diagnosis of AIS, a PA, and lateral full-length radiographs. Exclusion criteria were having undergone surgery, having initiated orthotic treatment, and a low quality radiography (without inclusion from C1 to S1 and without inclusion of femoral heads or incorrect positioning, mainly of the upper limbs).

Radiography
All the radiographs were performed in the same hospital and requested by the same doctor. As recommended by the Scoliosis Research Society, plain full-length radiographs were performed in the upright standing position, with the anterior superior iliac spines and hips parallel to the cassette and the beam aimed at T10. Lateral radiographs were performed with the patient in as similar a position as possible to the PA radiograph, with the beam at 90º to that used for the PA radiography, anterior superior iliac spines and hips perpendicular to the film, and the right side of the patient parallel to the cassette. Both feet were on the same alignment, 20 to 25 cm between each other, with the fingertips resting on the collarbones.

Parameters
The parameters evaluated were: age, sex, pelvic incidence, sacral slop, pelvic tilt, global sagittal and coronal balance, sagittal Cobb angle between T1 and T5, and sagittal Cobb angle between T5 and T12. Regarding the parameters evaluated, and their description, the shape of the cervical spine was obtained through the angle between two lines parallel to the posterior margins of the vertebral bodies of C2 and C7 (Jackson physiological stress lines). Thoracic kyphosis was evaluated from the Cobb angle between two lines; one along the upper platform of the body of T1 and the other along the lower platform of the body of T5 that measured the upper thoracic kyphosis, and two other lines; one along the superior end plate of T5 vertebral body and the other along the inferior end plate of T12 vertebral body (19), which measured the inferior thoracic kyphosis. Lumbar lordosis was assessed through the Cobb angle between a line along the superior L1 end plate and another along the upper sacral plate. Regarding the pelvic parameters, the pelvic incidence is the angle between the perpendicular to the sacral plate and the line connecting the sacral end plate midpoint to the hip axis. The sacral slope is the angle between the superior end plate of S1 and the horizontal plane. The pelvic tilt is the angle between the line connecting the midpoint of the sacral plate to the hip axis and the vertical plane. The sagittal balance is assessed by the horizontal distance or misalignment of a vertical line passing from the center of C7 vertebral body to the posterolateral part of the S1 superior end plate. It was considered positive vs negative as it presented an anterior vs posterior deviation from the previously stipulated line. To assess coronal balance, a vertical line is drawn downwards from the midpoint of the C7 vertebral body. The horizontal distance between this plumb line and the midline of the sacrum or central sacral vertical line is measured and the position of this line is then named positive, neutral or negative, depending on the distance and direction from the midline.

Statistical analysis
SPSS Statistics V21 MacOSX was used to perform this study.
A descriptive statistical analysis of all variables was collected. Subsequently, a separate analysis of the sagittal balance and the remaining variables under study was performed using the Spearman test. The assessment of the difference between the positive vs negative sagittal balance relative to the different variables was obtained through an independent-sample T test (Levene's Test for Equality of Variances and t-test for Equality of Means). The differences between sex were also obtained from this last test.
As this work is retrospective, based solely on imaging study without direct contact with patients, interviews, or application of scores, the hospital where the study was performed does not require approval by the ethics committee.

RESULTS
The idiopathic scoliosis group showed a mean age of 12.44 years, a mean sagittal balance of -1.14±3.30 cm and 29 patients with a negative sagittal deviation (<-2 cm), only 9 with a positive deviation (>2 cm) and the remainder were considered balanced (with SB greater than -2 cm and less than 2 cm).The mean coronal balance was -0.49±1.34 cm and LL 58.08±12.47º. (Table 1) Thoracic kyphosis presented a mean value between T1-T5 of 19.46±9.26º and between T5-T12 of 26.99±11.13º. (Table 1) The main coronal curvatures had a mean value of 21.83±15.68º and were later divided into 3 groups: patients with scoliotic angles of 10-20º, 20-45º and> 45º, each presenting a prevalence of 72.5%, 15.0% and 12.5%, respectively. Regarding the location of the main curvatures, 32.5% were thoracic, 53.8% thoracolumbar and 13.8% lumbar. The spinopelvic parameters presented mean values similar to those found in previous studies, with a PI of 50.79±14.31º, an SS of 42.48±10.51º and a mean PT of 8.68±9.94º. (Table 1) The sagittal Cobb angle of T1-T5 showed a statistically significant (p<0.05) and positive correlation with the sagittal balance, indicating that an increase in the sagittal Cobb angle in the upper portion of the thoracic spine would lead to an increase in sagittal decompensation in a positive direction, which did not occur with the lower portion of the thoracic spine (T5-T12), lumbar lordosis curvature, or other variables. (Table 2) Moreover, a detailed analysis of these two variables showed that AIS patients with negative sagittal balances had a lower kyphotic Cobb angle in both the upper and lower thoracic spine compared to those with positive sagittal balances, although the difference was only significantly higher (p<0.05) in the upper thoracic region (T1 to T5). On the other hand, scoliotic patients with sagittal balance <2 cm vs>/=2 cm presented, respectively, upper thoracic spine (T1-T5) angles of 18.56±9.32º vs 25.78±6.05º. (Table 3) Regarding the curvature of the cervical spine, 38.8% of the AIS patients presented lordotic curvature, 28.8% were straight, and 23.8% were kyphotic. Five patients were excluded from the data set in relation to this last variable because they were at the borderline between two cervical shape groups. The data were also crosschecked, to verify whether patients with a positive sagittal balance had spinopelvic parameters, LL, or thoracic sagittal curvature that differed significantly from those with a negative (<-2 cm) or normal (between -2 cm and 2 cm) SB. We used an independent sample test (Levene's Test for Equality of Variances and T-test for Equality of Means), which concluded that patients with SB>2 cm presented significantly different Cobb values of the upper thoracic portion (T1 to T5) (p<0.05) and SS (p<0.05) from those of the remaining population (with SB <2cm). Using the same test, we verified that scoliotic patients with kyphotic vs lordotic cervical curvatures presented with     significantly different SB (p<0.01), -3.66±3.15 cm vs 0.73±2.24 cm, respectively. (Table 4 and 5) This correlation did not occur in the comparison of SB and the remaining variables, including the lower thoracic spine Cobb angle (from T5 to T12) or LL. The coronal Cobb angle of the main scoliotic curvature showed a significant and inverse correlation with the kyphotic angle of the upper and lower portion of the thoracic spine (p<0.05) ( Table 2), demonstrating that an increase in the severity of scoliosis was correlated with an increase in the severity of the hypokyphosis itself. This result is in agreement with previous studies that support the relationship between scoliosis and hypokyphosis. The coronal Cobb angle presented a positive and significant correlation (p<0.05) with PI, as demonstrated by the higher PI values in those with a higher Cobb angle of primary scoliotic curvature. PI did not show a significant correlation with thoracic kyphosis or sagittal balance.
Lumbar lordosis presented a positive and significant correlation (p <0.05) with the lower thoracic spine (T5 to T12), PI and SS, but was not correlated with the remaining variables.
The spinopelvic parameters were strongly related to each other (p<0.01). Interestingly, the coronal balance had no significant correlation with the variables analyzed, including the coronal Cobb angle of the main scoliotic curve. This may have been due to the small discrepancy in the results for CB found in our group. diagnosis or with the risk of progression of scoliosis. 42 Our sample showed that AIS is related to thoracic hypokyphosis, both in its upper and lower portions. On the other hand, the sagittal balance seems to be much more influenced by the upper spine, specifically by the cervical spine shape and by the first 5 thoracic vertebrae. Idiopathic scoliosis correlates with hypokyphosis and a decrease in cervical lordosis, two parameters with a significant and negative influence on SB, which consequently decreases its value in a population with AIS. On the other hand, our results did not present a statistically significant correlation between the sagittal balance and the lumbar lordosis or spinopelvic parameters. This may be because there is no relationship at all between these variables, but it is more likely that the LL and spinopelvic parameters may have influenced the SB, mainly by indirect mechanisms and compensations, rather than direct ones, which is why they did not reach statistical significance. On the other hand, LL was closely related to the lower portion of the thoracic spine (from T5 to T12) (p<0.05), which did not occur with respect to the upper thoracic portion (from T1 to T5). When evaluating patients with structural deformities of the cervical spine, an in-depth understanding of the spinopelvic parameters is crucial. Pelvic morphology may influence sagittal spinal alignment and balance. 43,44 In a well-balanced spine, lumbar lordosis is organized according to SS, which in turn, depends on the shape of the pelvis quantified by the PI. 45 Mac-Thiong et al. 46 evaluated the spinopelvic parameters in patients with AIS and concluded that lumbar lordosis was strongly related to the pelvic configuration. Legaya et al. report that PI is a fundamental pelvic parameter for the three dimensional regulation of spinal sagittal curves, and is correlated with LL. 13 In addition, Stagnara et al. 35 found a strong relationship of lumbar curvature with SS and PT.
In the present sample, LL showed a positive and significant correlation (p <0.01) with PI and SS and the spinopelvic parameters demonstrated a significant statistical correlation between each other (p <0.01).Thereby, it was possible to confirm a strong connection between the pelvic geometry and lumbar lordosis. On the other hand, the severity of the scoliotic curvature presented a significant direct correlation with the PI, with a tendency to worse scoliotic curvature Cobb angles in patients with higher PI values (p<0.05).
The impact of sex on spinopelvic parameters remains controversial. Vialle et al. reported significant differences in LL and PI between male and female subjects. 47 In addition, Zhu et al. found a significant sex difference in LL. 48 Conversely, as in this sample, other researchers did not demonstrate significant sex differences in any spinopelvic parameter. [49][50][51] Interestingly, the coronal balance did not correlate with the variables under study, possibly due to the small discrepancy of results found in the main population with idiopathic scoliosis, in relation to this variable.
The results of this study may be of paramount importance in the analysis and planning of the surgical treatment of patients with spinal adolescent idiopathic scoliosis deformity, differentiating some of the most influential parameters in the SB.

CONCLUSIONS
Based on this study, we can infer several relevant factors that influence sagittal balance. We can also conclude that the spine of a population with AIS presents a series of spinal compensations and mechanisms that lead to structural alterations that directly influence the vertical axis.
Sagittal balance is a parameter that is influenced by multiple factors. In fact, it is closely related to the cervical shape and the upper thoracic spine (from T1 to T5), which in turn, is in symbiosis with the severity of scoliotic curvature. The Cobb angle of the lower portion of the thoracic spine (from T5 to T12) presents a closer relation with the angle of lumbar lordosis than with the upper thoracic curvature (from T1 to T5).
All authors declare no potential conflict of interest related to this article.

DISCUSSION
The spine is a complex structure balanced by multiple forces that implement structural changes in an attempt to compensate the sagittal and coronal vertical axis, so that the human being maintains as balanced a movement as possible. Sagittal balance can be maintained through three main compensatory mechanisms, which may occur in the spine, pelvis and/or lower limb areas, including reduction of TK/ hyperextension of adjacent segments, pelvis retroversion (increase of PT and rotation of the pelvis), knee flexion and ankle extension. 36,37 Hyperextension of the adjacent segments is a common compensatory mechanism in retaining sagittal balance, as pelvis retroversion, knee flexion and ankle extension may occur secondary to hyperextension of the adjacent segments of these are too rigid to extend or reach their limits. [36][37][38] Yang C et al. 23,24 Also advocated an analogy between LL, TK and sagittal balance. Thoracic hypokyphosis may be a risk factor for scoliosis curve progression based on the theory of "anterior column overgrowth", and the results of histomorphometric and magnetic resonance imaging studies that showed disproportionate growth of the anterior and posterior vertebral columns in patients with idiopathic scoliosis. 39,40 The progression velocity of scoliosis curves may be greater in patients with thoracic hypokyphosis than in those with normal thoracic kyphosis. 41 In another study, thoracic kyphosis was not associated with either the severity of scoliosis at the time of