Ever since the importance and influence of spinopelvic alignment on health-related quality of life was recognized, avid research has deepened our understanding of the normal values and pathophysiology related to abnormal alignment. This topic has not been scrutinized in the adolescent scoliosis population, who have a very unique sagittal profile due to the abnormal curve and rotation of the spine.
Thoracic AIS has been reported to have small TK. This trend can be universally found in black, white, and Asian populations, but especially in the Asian population [10, 12,13,14,15]. Among AIS patients, those with a primary curve in the thoracic spine had smaller TK than those with a primary curve in the lumbar spine [13, 14]. In our study, in which all patients were Japanese, the average TK was as low as 16.1°, which, interestingly, is close to that reported in Chinese girls with T-AIS by Yong (Table 3) .
Clement found that half of his included cohort had very small TK, with an average of 8.2° . Similar to his report, a cluster of patients in our cohort had very small TK and was classified as Type 2. These patients are reported to be at risk for several clinical issues, including pulmonary impairment, lumbar disc degeneration, and possibly neck pain, many years after fusion [3, 4, 6, 16,17,18]. In particular, the negative impact of thoracic hypokyphosis on pulmonary function has been considered clinically important and requires the appropriate choice of treatment. In this regard, Winter et al. recommended in his early case report on thoracic lordoscoliosis that these patients should undergo early fusion surgery rather than continue brace treatment if the lordosis increases under brace treatment . The flattening effect of brace treatment for AIS patients on the sagittal profile has been reported, and this effect from orthotics understandably exerts a negative impact on primarily hypokyphotic patients with decreased pulmonary function [20, 21]. Hence, for this group of patients, earlier surgical correction should be considered with the objective to restore thoracic kyphosis. Several techniques have been introduced for the correction of thoracic kyphosis, including meticulous facetectomy for spinal column release, surgery with a higher implant density and a pedicle screw system, and the vertebral coplanar alignment method [22, 23].
The correlation between lumbar and sacropelvic alignment has been advocated for in various populations, including healthy children, healthy adults, and scoliotic adult scoliotic patients [1, 2, 24]. Thoracic AIS can be expected to have some correlation to lumbosacropelvic parameters, similar to that seen in the normal population. The correlation between thoracic and sacropelvic parameters remains unclear. However, this topic is reported less clearly, and these parameters seem to have weaker correlations with each other. The spinopelvic parameters of healthy children in a similar age group to those in this study were reported as follows: TK 20.8 ~ 46°, LL 48 ~ 57°, PI 44.6 ~ 46.9°, PT 7.7 ~ 11.3°, and SS 33.3 ~ 39.1° [2, 10, 25,26,27]. Compared to these values, the parameters in this study were all similar except for TK, which was smaller. This can be interpreted as thoracic scoliosis influencing only TK in the sagittal spinopelvic parameters. This is consistent with a report of healthy children and adolescents that found that PI regulated SS and PT, but there was a weaker correlation between TK and PI . The lumbar spine seems to function as the absorber between the thoracic spine and pelvis to lessen the influence of the alignment on its counterpart.
Three types of thoracic AIS by cluster analysis
In our cohort, the significant factors for a decrease in TK were an increase in SS and decrease in max-LL, and thoracic AIS could be classified into three types based on SS and max-LL. (Fig. 3).
Type 1 is thoracic AIS that manifests as flat global sagittal alignment with low SS and low max-LL. This type is hypokyphotic thoracic scoliosis with an average TK of 15°. The LL is small, seemingly adapting to this small TK. Type 2 is thoracic AIS that manifests with high SS and low max-LL and has very small TK, with an average of 6°. This type corresponds to the sagittal thoracic modifier “minus” in the Lenke classification. Type 3 is thoracic AIS that manifests with high SS and high max-LL and exhibits undulating sagittal alignment at an average TK of 23°. This type has a large LL that looks well balanced with a large SS. Its larger TK seems to adapt to the large LL. Abelin-Genevois classified AIS sagittal alignment into four categories and reported that a fraction of patients had the curve pattern of thoracolumbar lordosis with a proximally shifted inflection point . The reason why we did not find this type among our patients is unclear, but the reason may be attributable to racial differences, minor but nonetheless present difference in the pelvic shape between the two studies, or possibly a prevalence too low to form another cluster category in our study.
Type 2 has a unique sagittal profile. The lumbar lordosis seems disproportionately small compared to the large sacral slope. The IP SVA was significantly anterior, while the levels of the IP and C7 SVA remained similar compared to those in the other types (Fig. 4). In the comparison between Type 2 and Type 3, both had similar PI, SS, PT, level of IP, coronal Cobb angle, and C7 SVA, but significant differences were found in LL, max LL, TK, and IP SVA (Fig. 4). Interestingly, despite similar sacropelvic alignment and C7 SVA values, Type 2 patients presented with largely different sagittal profiles from Type 3 patients. We tried to find the theoretical reason for this difference, which may lie in the flexibility between the thoracic spine and thoracolumbar/lumbar spine. The more flexible segment should function as the compensator for ergonomic alignment. Here, two possible hypotheses for this mechanism are presented. One is due to the stiffer thoracic spine, and the other is due to the stiffer lumbar spine.
One hypothesis is that stiff lordoscoliosis or hypokyphotic scoliosis in the thoracic spine influences the shape of LL or IP SVA. SS is said to develop and stabilize as the individual starts to walk in his or her early childhood, while PI and LL continue developing until the teenage years . As the child grows, sufficient LL for the SS should develop in the patients of Type 2. However, stiff lordoscoliosis manifests in adolescence, and this change would shift the C7 SVA posteriorly. Between the stabilized SS and the stiff TK, the more flexible thoracolumbar/lumbar spine reacts as a buffer to shift the IP-SVA anteriorly by becoming less lordotic.
The other hypothesis is that a sufficiently large LL cannot develop, so TK compensates for the anterior shift of the IP SVA by becoming as flat as possible to achieve the standard C7 SVA. If LL is defined by PI far more strongly than by SS, the aforementioned gap in the timing of maturity between LL/PI and SS may be the key to explaining why a sufficient large LL cannot be achieved. That is, if the immature small PI defines the small LL, this LL would not be large enough for the SS, which has already fully developed in early childhood. This insufficient LL would result in an anterior shift of the IP SVA, and this shift would require compensation by the uniquely flat TK to hold the C7 SVA within the standard range.
In both of these hypotheses, the chronological gap of the maturity of PI/LL and SS plays a role and should be considered. This gap was mentioned by MacThiong et al. [2, 28]. According to them, SS stabilizes just after the initiation of bipedal walking in early childhood and does not significantly change toward adulthood. In contrast, he also states that PT gradually enlarges to manage the growing and posteriorly shifting mechanical load from the center of gravity over the growth period. Following this development of PT, PI also grows larger under the formula PI = PT + SS, driving the increase in LL as well. As such, this gap in maturity among several sagittal parameters seems to occur over a decade. This would be long enough to induce a unique compensatory cascade for balanced sagittal alignment in developing children, as proposed in our hypotheses.
The limitations of this study are as follows. The included patients ranged in age from preadolescence to young adulthood. This age range varies in many aspects, including growth phase and spinal stiffness. For example, spinal stiffness may influence radiographic appearance, especially when the compensatory mechanism is thought to influence the deformity. The number of total cases included was not large enough to stratify the cases based on age, but ideally, we should perform age stratification as more cases become available. The other limitation is that we assessed the curve only on lateral X-ray images. It has been reported that thoracic kyphosis is overestimated on 2D images compared to 3D images because of vertebral rotation . This reflects that our thoracic AIS patients may have less TK or TLK or a higher inflection point than if these parameters were measured on 3D images. This susceptibility to rotation may also suggest that the angular measurements made with lateral X-ray images can easily vary, even in the same subject, if the X-ray beams are not shot exactly in the true lateral direction.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.