The decision of whether to treat a TL vertebral fracture either conservatively or with surgery can be challenging. The validated TLICS scoring system is an objective and effective tool for guiding operative versus non-operative decisions for management of vertebral fractures; however, a score of 4 is indeterminate [20]. The decision to proceed to surgery for vertebral fractures with indeterminate stability (TLICS score 4) and potentially stable fractures (TLICS score ≤ 3) is affected by a combination of several different factors. Radiographic factors that are used to guide management decisions include kyphotic angle, severity of vertebral fracture, vertebral body height loss, fracture location, integrity of posterior column structures, and remaining canal area [7, 21]. When the TLICS score is indeterminate or there is cause for uncertainty as to whether surgery is indicated for instability, standing radiographs of the TL spine are taken at 1, 3, 6, and 12 weeks to monitor the healing process and identify occult instability not recognized on initial evaluation [5].
As shown in this study, monitoring of vertebral fractures with sequential upright radiographs can lead to a change from conservative to surgical management. However, vertebral fracture management guidelines are inconsistent and there is no general consensus as to what factor is weighted more when selecting the appropriate treatment plan [22, 23]. This is why the same type of fracture and/or TLICS score can have different treatment paths depending patient injury and/or circumstances. In this study there were a total of nine neurosurgeons and seven orthopaedic surgeons who were involved in vertebral fracture management decisions. There could be differences in training and practices, but there were comparable amounts of neurosurgeons and orthopaedic surgeons in the non-operative and operative groups. In this study, it was discovered that decisions to follow conservative or operative pathways were based on findings on upright radiographs and patient situations. Two patients in this study with TLICS scores of 6 and 7 had flexion-distraction fractures, which would typically suggest surgery due to a higher degree of injury. However, both were treated conservatively with a TLSO brace. Despite having an injured PLC, the first patient had limited social support, active IV drug use, and poor nutritional status, which did not make them a good candidate for surgery. The second patient had an indeterminate PLC injury, but was determined to have a stable fracture after the first upright radiograph confirmed fracture stability. This decision to manage conservatively was supported by the anterior body height loss being 21% on initial CT scan and 16% on endpoint X-ray at the 3 month follow up. Additionally, the kyphotic increase from CT scan to endpoint X-ray was only 0.8 degrees. This is an example of how upright radiographs can be useful in guiding vertebral fracture management decisions and prevent patients from having unnecessary surgery and the risks associated with it.
In this study, the effectiveness of upright radiographs as a tool to guide operative versus non-operative decisions for traumatic vertebral fracture management was evaluated. Standing radiographs led to a change of plan from presumed non-operative to operative treatment in 6 (8.6%) of 70 patients. Four of the six operative patients converted to surgery during the index admission after showing an increase in kyphosis ranging from 12.3 to 22.7 degrees and an increase an anterior body height loss ranging from 8 to 53% when comparing initial CT scan to first upright radiograph. These patients had acute fractures and the time from initial CT scan to first X-ray ranged from 0 to 8 days. This suggests that factors such as age or undiagnosed osteopenia did not play a role in the fracture subsidence. Instead, findings on first upright radiographs such as increased kyphosis compared to CT scan and average anterior body height loss guided management decisions. There was a significant difference between non-operative (38%) and operative (53%) anterior body height loss in this study. This amount of anterior body height loss in the operative group is similar to literature suggesting > 50% loss of height indicates mechanical instability and the need for surgery [8]. Within the operative group, the increase in kyphosis on first X-ray (13 degrees) compared to initial CT scan (2 degrees) was found to be statistically significant. Patients in this study had a conversion to surgery at a ~ 11 degree increase in kyphosis from CT scan to first X-ray and an average 13 degrees of kyphosis on first upright radiograph. This degree of kyphosis is lower than the > 30 degrees that has been reported as a level of kyphosis indicating instability and likely need for surgical management [8]. Our results illustrate how physicians in clinical practice may decide to operate at a lower increase in kyphotic threshold than is suggested. However, there is no universally agreed upon angle of kyphosis that warrants surgical management and there is conflicting evidence about whether or not increased kyphosis is predictive of conservative management failure [21].
Alimohammadi et al. [25] suggested that there is a possible relationship between increased kyphosis and failure of conservative therapy for vertebral fractures. However, this study focused on only burst fractures and compared kyphosis on all available imaging (CT, X-ray, MRI) instead of specifically upright radiographs. We assessed whether the degree of kyphotic angulation through the fractured vertebra on the first upright radiograph had a correlation with the decision to proceed to surgical stabilization. There was a significant difference in the degree of kyphotic collapse on the first standing radiograph in operative patients (~ 12 degrees) versus non-operative patients (~ 5 degrees). Additionally, in the first upright radiograph, there was a significant difference in anterior body height loss of 38% and 53% between the non-operative and operative groups, respectively. This supports that an increase in kyphosis from CT scan to the first upright radiograph could be relevant in fracture management decision making processes, but larger studies are needed to confirm this.
There is no general consensus on the use of a TLSO for TL fractures. Research has shown that there is no significant difference in functional outcomes and kyphosis for TL fracture patients treated with or without a TLSO [24]. This could explain why there was variable use of a TLSO in radiographs for the non-operative patients in this study. However, 50% of non-operative patients had their first upright radiograph in a TLSO. In the operative group, five of six patients who converted to surgical management had kyphotic collapse in their first upright radiograph when compared to initial CT scan despite the use of a TLSO. Additionally, the two patients who underwent surgical stabilization after outpatient follow up did so only after their first upright radiograph without a TLSO, which showed a further progression of kyphosis. TLSO use both allowed and hid kyphosis progression on upright radiographs for the operative patients in this study, therefore, it is yet to be determined if and/or how TLSO use affects markers of stability on upright radiographic imaging.
There were limitations in this study. First, this was a retrospective study, which is prone to confounding. Data was collected for all eligible patients in an attempt to mitigate selection bias, though factors that could influence measurements such as BMI and smoking were unable to be abstracted from all patient charts. However, four of the six patients who converted to surgery during their index admission did so shortly after an upright radiograph, which suggests that factors influencing the conversion to surgery were more acute and related to the increase in kyphosis seen on first X-ray. Additionally, patients with neurological injury were eliminated from this study because acute change in neurologic function could be an indication for surgery regardless of radiological findings. Eliminating patients with things that could influence management substantially (pathological fracture, neurological injury) further supports the association between imaging measurements and surgical decisions. Second, upright radiographs are vulnerable to slight differences in patient position and image quality (imaging blurs), which could affect measurement values. Measurements were taken at specific points for each image and steps were taken including adjusting image contrast to best visual measurement points and having measurements checked for consistency. The slight differences in measurements that could occur would be identical to that seen in clinical practice, which supports the application of this study. Third, the sample size in this study is small. Larger multicenter studies with the inclusion of relevant confounding factors that could affect traumatic TL fracture measurements and management decisions are necessary.