Effect of the discrepancy between sacral and pelvic obliquity on postoperative disk wedging below the lower instrumented vertebra in patients with Lenke type 5 adolescent idiopathic scoliosis: a retrospective study in Japan

Ethical statement

This study received ethical approval from the Institutional Review Board of Keio University School of Medicine (20090042).

Study design

This single-center retrospective analysis enrolled 35 patients with Lenke type 5 AIS who underwent posterior correction and fusion surgery.

Participants

The study population was composed of 35 patients with Lenke type 5 AIS who underwent posterior correction and fusion surgery between 2010 and 2015. They provided informed consent. The inclusion criteria were female sex, Lenke type 5, LIV at L3, upper instrumented vertebra at T10–T12, and at least 5 years of follow-up. The mean age was 14.7±1.8 years (range, 11–17 years). The average Risser grade at the time of surgery was 3.9±1.2 (range, 0–5). All patients had a left convex lumbar curve. Patients with Richard’s disease or transitional lumbosacral vertebrae were excluded.

Radiographic measurements

The P angle, S angle, L4 tilt, sacral slope (SS), lumbar lordosis (LL), thoracic Cobb angle, lumbar Cobb angle, correction rate, height difference between both femoral heads (positive when the right side was higher), and wedging angle of the lower adjacent intervertebral disk of the LIV were measured (Fig. 1). These radiological parameters were measured using whole-spine standing posteroanterior radiographs. The P angle was defined as the intersection of the horizontal and connecting lines of both iliac crests. The S angle was defined as the intersection of the horizontal line and the upper endplate of the sacrum. To evaluate the extent to which the sacrum is tilted concerning the ilium, the angle was calculated by subtracting the S angle from the P angle to obtain the SP angle (SP angle=S angle–P angle). The postoperative distal adding-on phenomenon was defined as the progression of distal deformity with an increase of >5° in the LIV disk angle or a horizontal shift of LIV+1 exceeding 5 mm [4,7]. All values for the parameters were preceded by a positive sign if the apex of the lumbar curve was toward the left or if the left inclination was shown. As all the study patients had a left convex lumbar curve, both SP obliquity values were consistently positive. In this study, disk wedging was defined as the angle between the lower endplate of the LIV and the upper endplate of the LIV+1 (Fig. 1). All values for disk wedging were not negative because all cases exhibited a left-sided opening of the disk, with the lower endplate of the LIV tilted toward the left relative to the upper endplate of the LIV+1.

Fig. 1

(A) Schematic diagram showing pelvic coronal obliquity (P angle), sacral coronal obliquity (S angle), and the angle between the S and P angles (SP angle). (B) Whole-spine standing posteroanterior X-ray images of an 11-year-old female patient taken before and 5 years after surgery.

The patients were divided into two groups according to their preoperative SP angle. Patients with preoperative angles <5° and ≥5° were assigned to the <5° and ≥5° groups, respectively. The coronal radiographic parameters between the two groups were compared to investigate whether the preoperative SP angle affected postoperative coronal spinal alignment. The SP angle cutoff of 5° was chosen based on its clinical relevance and use in previous studies defining SP obliquities [11,14]. This threshold was deemed practical for analyzing its effect on postoperative outcomes.

Moreover, to determine the effect of preoperative coronal parameters on postoperative spinal alignment, the correlation coefficient was measured to assess the relationship between disk wedging 5 years postoperatively and the preoperative S, P, and SP angles.

Clinical assessment

At the final follow-up visit, the surgical outcomes were evaluated using the Scoliosis Research Society-22 item questionnaire.

Statistical analysis

All data were statistically evaluated using EZR (https://www.jichi.ac.jp/saitama-sct/SaitamaHP.files/statmedEN.html) [15], which is built on R (The R Foundation for Statistical Computing, Vienna, Austria) and utilizes the R commander. Paired t-tests were performed to analyze the preoperative radiographic and postoperative parameters 5 years postoperatively. The Mann-Whitney U test was performed to compare the two groups. Significance was set at p<0.05. Radiological parameters were obtained by the first author, who was not involved in treating patients with AIS in this study.

Radiographic changes after posterior correction and fusion surgery

The mean S angle decreased significantly 5 years postoperatively (5.2°±4.2°) compared with the preoperative value (7.1°±4.2°) (p<0.05). Conversely, no significant difference was found in the mean P angles preoperatively and those 5 years postoperatively (2.8°±1.8° versus 2.4°±2.0°, p=0.11). The mean SP angle was significantly lower 5 years postoperatively than preoperatively (2.8°±3.4° versus 4.3°±3.7°, p<0.05). Five years postoperatively, disk wedging below the LIV was 6.2°±3.7°, and no correlation was found between the preoperative P angle and postoperative disk wedging (Fig. 2). No significant difference was noted in the height difference of both femoral heads preoperatively and those 5 years postoperatively (2.3°±4.6° versus 1.8°±4.0°, p=0.65) (Table 1).

Fig. 2

Correlation between preoperative pelvic coronal obliquity angle (P angle) and disk wedging 5 years post-surgery on the left (A); correlation between preoperative angle between the sacral coronal obliquity angle and P angle (SP angle) and disk wedging 5 years post-surgery on the right (B). The solid lines represent the regression lines for each correlation.

Table 1

Mean preoperatively and 5 years radiographic parameters

Among the 35 patients included in this study, 4 (11.4%) met the criteria for postoperative distal adding-on. However, no cases demonstrated worsening postoperative coronal imbalance that required revision surgery.

Correlation coefficient between radiographic parameters pre- and postoperatively

In our analysis comparing the correlation coefficients of the SP, S, and P angles with disk wedging observed 5 years postoperatively, the SP angle had the highest correlation at 0.69, followed by the S angle at 0.68 and the P angle at 0.2 (Table 2, Fig. 2). Accordingly, the SP angle was chosen as the criterion for dividing the patients into two groups.

Table 2

Correlation coefficients between preoperative and postoperative radiographic parameters

Differences in preoperative radiographic parameters between the <5° and ≥5° groups

A total of 23 and 12 patients were included in the <5° and ≥5° groups, respectively. No significant differences in age at the surgery, Risser grade, height difference of both femoral heads, P angle, SS, and LL preoperatively were observed between the <5° and ≥5° groups. However, the mean S angle (4.9°±2.7° versus 11.4°±3.3°, p<0.05), lumber Cobb angle (42.6°±8.4° versus 49.1°±8.6°, p<0.05), thoracic Cobb angle (22.9°±6.7° versus 28.7°±7.1°, p<0.05), and L4 tilt (21.6°±5.7° versus 26.5°±6.3°, p<0.05) were significantly larger in the ≥5° group (Table 3).

Table 3

Mean preoperative radiographic parameters in the <5° and ≥5° groups

Differences in postoperative radiographic parameters between the <5° and ≥5° groups

Disk wedging below the LIV was significantly greater in the ≥5° group than in the <5° group (9°±2.6° versus 4.7°±3.4°, p<0.05) 5 years postoperatively. The ≥5° group exhibited a significantly lower L-Cobb angle (23.5°±9.6°) than the <5° group (14.4°±4.9°) (p<0.05) 5 years postoperatively.

The ≥5° group exhibited a significantly lower correction rate of the L-Cobb angle (52.1%±17.8%) than the <5° group (65.5%±12.7%) (p<0.05) 5 years postoperatively. The mean age, preoperative Risser grade, correction rate of the T-Cobb angle, and C7–CSVL were comparable between the two groups. No significant difference in the height difference of both femoral heads was noted between the <5° and ≥5° groups; however, a trend toward smaller differences was noted in the ≥5° group (Table 4).

Table 4

Mean radiographic parameters 5 years post-surgery in <5° and ≥5° groups

Comparison of sagittal parameters between the <5° and ≥5° groups

The sagittal radiographic parameters (SS and LL) were also evaluated 5 years postoperatively. However, mean SS and LL values were not significantly different between the two groups (Table 4).

Comparison of clinical outcomes between the <5° and ≥5° groups

Clinical outcomes were compared using the Scoliosis Research Society 22-item questionnaire. No significant differences in any of the domains were found between the two groups (Table 5).

Table 5

Scoliosis Research Society 22-item scores in <5° and ≥5° groups during last follow-up

Pelvic and sacral obliquity in patients with AIS

Chan et al. analyzed 311 patients with AIS who underwent corrective surgery at their institution and reported that 22% of these had a symmetrical pelvis [8]. However, few studies have focused on pelvic or sacral obliquity, particularly how it is altered by corrective surgery. Initially, SP obliquities were not clearly defined. Lee et al. [14] reported that in patients with AIS, the frequencies of sacral obliquity were 19.5%, 29.6%, and 40.6% according to the criteria of 5°, 4°, and 3°, respectively. Ploumis et al. [11] defined pelvic obliquity as a sacral angle of >5° or an iliac crest height difference of >10 mm. Joo et al. [3] defined sacral obliquity as a sacral takeoff angle >3°. Because pelvic and sacral obliquity is primarily observed in patients with extensive lumbar curves, such as Lenke types 5 and 6 [8], the effect of preoperative pelvic and sacral obliquity on postoperative coronal radiographic parameters was analyzed in patients with Lenke type 5 AIS.

Changes in pelvic and sacral obliquity following corrective surgery

The S and SP angles decreased significantly, whereas the P angle exhibited no significant change 5 years postoperatively. These data indicate that the discrepancy between the sacral and pelvic obliquity (SP angle) is reduced owing to the decreased S angle following corrective surgery. The significant decrease in the S angle observed 5 years postoperatively may be attributed to the direct effect of corrective fusion surgery on the sacrum. In contrast, changes in the P angle were less pronounced, likely because the pelvis is not directly fused to the sacrum, despite their anatomical connection. To achieve correction and fixation, surgical correction was primarily performed using a rod rotation technique from the convex side. Intraoperative alignment was evaluated using frontal X-ray images that included the pelvis and compared with preoperative standing full-length radiographs. Adjusting the LIV tilt to a horizontal position served as a useful target; however, the final correction and fixation prioritized achieving optimal overall spinal balance, considering pelvic obliquity and other preoperative alignment factors. Lee et al. [14] also reported the reduction of sacral obliquity in some patients following corrective surgery. The ≥5° group had significantly larger lumbar Cobb angle, thoracic Cobb angle, and L4 tilt at the time of surgery. These results are presumed to be due to the immaturity of the patient’s bones. They may also indicate that the difference between SP obliquity occurs as compensation for the lumbar curve [13,14].

In this study, no patients had sacral obliquity and pelvic obliquity that were inclined in opposite directions. This consistent alignment of obliquities reflects the uniformity of the cohort given that all patients presented with a left convex lumbar curve.

Disk wedging following corrective surgery in patients with Lenke type 5 AIS

The ≥5° group displayed a significantly larger wedging angle of the lower adjacent intervertebral disk 5 years postoperatively. The ≥5° group exhibited a substantially lower correction rate of the L-Cobb angle 5 years postoperatively. Because the LIV was L3 in cases with a large lateral tilt of the sacrum in the pelvis, the L4 tilt remained significant, and the surgical correction rate was speculated as low.

Moreover, determining the appropriate LIV in patients with AIS and large lumbar curves is difficult [16,17]. The preoperative discrepancy between the SP obliquity could be crucial in selecting appropriate distal and proximal fusion levels in the surgical treatment of patients with AIS. As practical surgical countermeasures, reevaluating the fixation range or adjusting the LIV tilt during intraoperative correction to align with SP obliquity could be considered. However, this study could not provide definitive conclusions regarding these approaches owing to its retrospective design and relatively small sample. Further prospective studies are necessary to validate these strategies and assess their potential effect on long-term outcomes.

This study had some limitations. Standing radiographs were obtained, which could have affected the accuracy of the angles. Computed tomography may enable more accurate results, and a three-dimensional evaluation should be performed. In addition, leg length discrepancy could contribute to SP obliquity. However, in this study of 35 cases, radiography could not accurately measure leg length differences. Therefore, the full extent of leg length discrepancies, which may have affected our overall findings, could not be determined. We attempted to estimate the discrepancies by assessing the differences in the heights of the femoral heads using standing frontal radiographs; however, this method cannot substitute for direct leg length measurements. Future studies using more precise imaging techniques, such as standing computed tomography, are recommended to explore this potential correlation. Furthermore, the number of cases may have been inadequate. To minimize bias and ensure the reliability of the results despite the small sample, clear and consistent patient selection criteria, adhering to rigorous data collection processes were established, and appropriate statistical methods were applied.

Despite the abovementioned limitations, this study has demonstrated that the group with SP angle ≥5° exhibited a larger wedge angle below the LIV and a lower lumbar Cobb angle correction rate than the <5° group 5 years postoperatively. Preoperative discrepancies between SP obliquity should be considered when devising surgical strategies for patients with Lenke type 5 AIS. However, the long-term results of the presented data on disk wedging are not still unavailable. If future long-term data reveal poor prognoses in patients with a significantly large preoperative SP angle, surgery should be avoided. The use of lower-limb orthoses to address pelvic tilt may also be considered in such cases.