Spinal fractures commonly occur in the thoracolumbar segment because it is where the thoracic and lumbar vertebrae meet; this region is unprotected by the ribs and has remarkable mechanical stress and mobility [12]. Inappropriate treatment or even no treatment further aggravates kyphosis; thus, secondary kyphosis deformity is formed, causing intractable low back pain, neurological dysfunction, and other complications in patients and posing a threat to the physical and mental health of middle-aged and elderly people [13–15]. Some studies [16, 17] found that when the RKA is greater than or equal to 30°, patients are at increased risk of chronic pain in the kyphosis area. Pain first occurs at the location of the deformity. Some patients may also complain of pain near the kyphosis area because localized kyphosis alters the normal biomechanics of the patient’s spine, causing premature degeneration of the vertebral levels above and below the deformity. Surgery is usually required when a patient presents with intractable low back pain, remarkable sagittal imbalance, or symptoms of spinal nerve compression. Therefore, a surgical osteotomy is usually performed in the apical region of kyphosis to correct the regional kyphosis of the thoracolumbar segment; thus, the sagittal balance of patients can be obtained and maintained, and the neurological symptoms can be improved [18, 19].
The severity of kyphosis deformity and the presence or absence of sagittal imbalance in the standing position should be considered first when making a surgical plan. The flexibility of the kyphosis deformity must also be evaluated before surgery [1, 11, 13]. RKA, usually measured on a standing spinal lateral X-ray radiograph before surgery, is considered an important indicator of the severity of kyphosis deformity in the spine[20]. Patients are typically in a forward protective posture when standing because of the effects of gravity, persistent back pain, and nerve traction pain caused by spinal stenosis; some elderly patients with weak back muscles cannot even stand entirely, thereby leading to an excessive measured RKA [11, 15]. This measured value does not reflect the actual intraoperative degree of regional kyphosis in patients with kyphosis secondary to so-OTLF, and its clinical application relevance is greatly reduced. Similarly, we found that preoperative standing RKA and intraoperative prone RKA after general anesthesia were not the same in clinical practice. Under general anesthesia, the RKA was reduced to a certain extent; sometimes, it exceeded 50%. Although obtaining full-length X-ray radiographs of a standing spine in the lateral position was straightforward, affordable, and quick, the compensating angle brought on by discomfort and the effects of gravity was not easy to eliminate.
Putto et al. [21] took lateral radiographs with padding at the apex of kyphosis, which can measure the passive reducibility of kyphosis and the extent of the remaining correction (sometimes up to 50%), which may change the entire surgical approach. However, this finding was not investigated in further detail because the authors emphasized only the judgment of intervertebral stability rather than the assessment of flexibility. The X-ray radiographs of lumbar spine flexion and extension positions are often used in clinical diagnosis and treatment of lumbar degenerative diseases to evaluate the stability of the interbody, suggesting that the change in the sagittal curvature of the interbody is the embodiment of flexibility. However, the increase in the patient’s back muscle activity and the aggravation of back pain during the dynamic overextension and overflexion position film limit or prevent the activity of some patients; thus, using dynamic position radiographs to judge sagittal kyphosis flexibility becomes inaccurate [11]. When the patient is lying down during an MRI examination, the patient's back muscles can be completely relaxed, and the protective anteversion brought on by discomfort can also be reduced. Due to the short examination time, these factors cannot be completely eliminated in CT examination[22]. Kaiser R et al. [23] found a linear dependency between hyperextension radiograph and MRI flexibility with a mean difference of 9.3° (R2 = 0.61, p < 0.001) by using the imaging data of 18 Scheuermann kyphosis patients. They demonstrated that preoperative MRI has a similar predictive value to bolster-assisted hyperextension lateral radiograph for the Scheuermann kyphosis flexibility assessment. Sharma A et al. analyzed the imaging data from 138 patients with sagittal imbalance. They found that a mean difference of 2.9° exists between the LL measured on supine MRI and that measured on intraoperative X-rays, as opposed to the 5.53° mean difference between standing X-rays and intraoperative X-rays[24]. In patients with flexible deformities (n = 24), the lumbar lordosis on MRI measured a discrepancy of 3.08°, as compared to a discrepancy of 11.46° when measured with standing X-ray. This finding showed that preoperative MRI effectively identifies flexible sagittal abnormalities. Moreover, it indicated that the LL recorded on MRI more correctly predicts the intraoperative LL than that measured on standing X-ray in cases of flexible sagittal abnormalities. Although different researchers analyzed the predictive value of preoperative MRI in predicting TK and LL flexibility for adult spinal deformity, the reports about the predictive role of MRI in SK kyphosis flexibility assessment in patients with kyphosis secondary to so-OTLF are few[23, 24]. We hypothesized that the MRI examination in supine position could eliminate the factors of protective anteversion caused by intractable back pain, gravity, and nerve traction pain; thus, highly accurate RKA could be obtained.
We evaluated 37 patients (8 males and 29 females) with kyphosis secondary to so-OTLF, with an average age of 66.3 years. The RKAs measured on standing X-ray radiographs, supine MRI, and prone X-ray radiographs before the operation were 48.03°, 34.40°, and 31.95°. Compared with the RKA measured in standing position, the RKA measured on supine MRI decreased by 13.63° (95% confidence interval 11.44°–15.82°), whereas the RKA measured on intraoperative prone radiographs decreased by 16.08° (95% confidence interval 13.66°–18.50°). In the present study, the RKA measured on intraoperative prone radiographs was significantly decreased compared with that measured on standing radiographs(p < 0.001). This finding suggests that postural reduction of the surgical frame during surgery significantly reduced the RKA measured on intraoperative prone radiographs. The SK flexibility measured on supine MRI was significantly linearly related to the actual SK flexibility during operation, with an average difference of 2.44° (R2 = 0.912, p < 0.001). This finding indicates that supine MRI can be used to assess SK flexibility in patients with kyphosis secondary to so-OTLF. Moreover, the difference between the RKA measured with preoperative MRI and the actual RKA during the operation was only 2.44°.
Although supine MRI cannot perfectly reflect all the information of the prone position on the operating frame, as a non-weight-bearing imaging examination, it can convey the relevant information about SK flexibility. We used supine MRI to assess the SK flexibility of patients with kyphosis secondary to so-OTLF to develop surgical protocols that can benefit individual patient management. With a flexible deformity, a large portion of the correction was obtained by positioning only a surgical frame. Moreover, aggressive corrective maneuvers (i.e., pedicle subtraction osteotomy [PSO]) may not be necessary. Although PSO is a useful procedure for effectively correcting certain deformities of the spine, it has notable complications, such as massive blood loss and complications of nerve damage. The research of Boachie-Adjei et al. on PSOs revealed that PSOs are associated with a 2700 cc hemorrhage and a 100% complication rate, with 9 of 14 patients needing a revision operation within 2 years[25]. The acceptance of such aggressive surgery by surgeons has increased in recent years; however, mortality and complication rates have not decreased considerably[26]. Additionally, proper anesthetic assessment, electrophysiological detection, autologous blood transfusion, and adequate blood preparation are required when adopting such an aggressive surgical osteotomy plan [27]. Therefore, surgeons must determine whether an aggressive osteotomy like a PSO is required preoperatively. Additionally, preoperative MRI, routinely performed in such patients to exclude neuroaxis pathologies or any compressive cord lesion before, does not increase the economic burden of the patient and the absorption of ionizing radiation.
Given the small sample size, the present study only makes some preliminary discussions on the differences between different positions. It cannot analyze the the apex of the deformity in groups. Thus, a large sample size is required. During the operation, we used the C-arm to perform fluoroscopy with the posterior convex apex as the center. Some images of multiple vertebral fractures need to be spliced, which may have some errors. Furthermore, although different spinal frames might affect RKA in the prone position, we only evaluated one type of frame, and validation using various frames will be required.