LKCA and TK significantly reduced from standing to the prone position before surgery. The LKCA, TK, and SVA of immediate postoperative and at last follow-up results were statistically different compared with those before surgery; the differences in LL, PT, and PI-LL in the postoperative period and at the last follow-up were not statistically significant compared with those before surgery. No complications of infections or nerve injuries occurred, and there was a significant decrease in ODI and VAS scores at two months and at the last follow-up after surgery.
The etiology of kyphosis in old traumatic spinal fractures is diverse. Furthermore, the key factors remain unclear, exacerbated by misdiagnosis or delayed diagnosis, treatment failure (regardless of surgical or non-surgical treatment), disc injury, and other diseases affecting bone density [27]. Patients are severely affected and require surgical treatment; and the indications for surgery include pain, nerve damage, and worsening of progressive kyphosis [4, 27].
The selected fusion segment of posttraumatic kyphosis is generally 2–3 vertebrae above and below the fracture site [6, 28]; therefore, we focused on the posterior convex Cobb angle of the surgical segment. In our cases, LKCA was significantly reduced from 45.26° in the standing position to 32.27° in the prone position with a flexibility of 42.25%. This change may be due to the effect of gravity [29, 30]. Fei et al. compared the MRI results of patients with degenerative scoliosis in standing and supine positions and the kyphosis Cobb angle of the surgical segment in the prone position, and found that the kyphosis Cobb angle of the surgical segment in the prone position was significantly reduced [22]. In the prone position, the shoulder and pelvis are cushioned, the trunk collapses forward due to gravity, the anterior disc opens, the posterior column closes, and the kyphosis Cobb decreases. Another reason for this is the average age of 63.67 years in our group of patients. The decrease in muscle strength in elderly patients results in insufficient bony strength to maintain trunk balance in the standing position [31]. The anteversion of the spine above the fracture also led to a large kyphosis Cobb angle, which is eliminated in the prone position. In addition, spinal flexion is also a way for elderly patients to compensate for other diseases, such as spinal stenosis and pain [32]. Pseudarthrosis may also be responsible for the reduction of kyphosis. Two patients in this group had combined pseudarthrosis, and an intraoperative PKP was performed in the injured vertebra to stabilize the anterior column.
LL in the prone position was affected by the hip joint and increased when the hip joint was hyperextended, decreased when the hip joint was hyper flexed, and was less affected by the hip joint when the hip joint was within 30° of flexion [33, 34]. Compared with the standing position, there was no significant change in LL in the prone position in this group, probably due to the pelvic cushion causing hip flexion within 30°.
The main parameters considered in the current surgical plan are the standing kyphosis Cobb angle and neurological function. The normal range of thoracolumbar kyphosis Cobb angle is 2.2 ± 8.0 ° [35], and increases with age [36]. When the patient’s LKCA is < 20° in the prone position, using the corrective force of the rod can corrected the alignment to the normal range. In our group, six patients were not osteotomized.
In patients with good flexibility and no neurological compression, the deformity was self-corrected in the prone position. When the kyphosis Cobb angle is 20 to 30° in the prone position, a simple posterior column osteotomy can be used to complete the correction. The non-osteotomy technique and simple posterior column osteotomy, which does not destroy the anterior column, has comparatively better stability of the osteotomy end and less possibility of nerve injury [20].
Therefore, the surgical plan based on the prone sagittal parameters has the potential to reduce the operation time and blood loss. In this group, the operation time was 152.7 min and the mean bleeding was 408.3 ml. SVA, ODI scores, and VAS scores at two months postoperatively and at the final follow-up were significantly improved compared with the preoperative period.
However, non-osteotomy or single posterior column osteotomy has limited correction ability and cannot decompress the compressed nerve. Thus, the technique in the study is not suitable for patients with nerve compression. For patients with rigid or combined anterior nerve compression or kyphosis Cobb angle still over 30° in the prone position, an anterior column osteotomy is required [13, 16, 24, 37], including PSO, VCR osteotomy, etc. Although researchers have refined these osteotomy techniques [11, 12, 14], three-column osteotomy often decreases spinal stability, and neurological complications, operation time, and complication rates increase.
Rigid fixation and bony fusion is the key to prevent correction loss. During PSO and VCR procedures, both the anterior and posterior column could be bone grafted [37, 38], with good long-term outcomes [15, 17]. In our opinion, the simple posterior column fusion used in this study did not destroy the anterior column structure and the posterior column fusion showed a good therapeutic effect in flexile deformity, unlike that in Scheuermann disease, Mucopolysaccharidosis syndrome, etc. [26, 39]. Our patients were followed up for at least two years, and no obvious correction loss was found.
Changes in postoperative imaging parameters in patients with kyphosis are related to compensatory mechanisms. Different from degenerative kyphosis in the elderly, posttraumatic kyphosis is characterized by a decrease in thoracic kyphosis and an increase in lumbar lordosis [40, 41], so theoretically, after the correction of the deformity, TK increases and LL decreases [42], but the compensatory mechanism is not simple. Li et al. found an increase in LL after traumatic kyphosis correction surgery [10], while Olivares et al., who used PSO osteotomy for old traumatic kyphosis, showed that TK and LL were unchanged after surgery [21]. The immediate postoperative TK and LCKA were significantly reduced compared to the preoperative period in our patients, which is similar to previous studies [15, 17]; the lack of change in postoperative LL may be explained by the small SVA and no severe sagittal imbalance in our patients [43]. The mean age of the patients in this group was 63.67, and the results suggest that the compensatory mechanism in the elderly may occur first in the thoracic spine and later in the lumbar lordosis area. There was a tendency for the LL to decrease at the final follow-up, although the difference was not statistically significant (P > 0.05).
PT is associated with LL, as well as PT increases and LL decreases [44]. There was no change in the preoperative standing position, prone position, or postoperative LL in our patients; therefore, there was no change in PT, similar to other studies [21, 42]. PI is a fixed value that does not change with position, so PI-LL does not change [21, 42]. SS reflects the degree of pelvic tilt. PI has a geometric relationship with PT and SS, and PI equals PT and SS [25]. PT and SS did not change in this group of patients.
There were some limitations to this study. First, we had a small sample size and no severe sagittal imbalance, which may cause selective bias, and we still need to observe the long-term efficacy. Second, we did not study the factors influencing the flexibility of kyphosis, and require a further large sample, multi-center study. Third, regarding how many degrees were required to choose between an SP or Ponte osteotomy, we did not stipulate any strict criteria, and further studies are required to quantify the effects thereof.