Percutaneous fixation and balloon kyphoplasty for the treatment of A3 thoracolumbar fractures

Abstract

Background

Despite a long history of descriptive and clinical series, there is still no consensus in the treatment of traumatic thoracolumbar fractures. It is now widely accepted that percutaneous surgery in thoracolumbar spine trauma management can achieve the same results as conventional treatment but less morbidity but it is still not clear which are the best indications for these minimal invasive procedures.

Methods

Thirty-two adult patients with single type A3 thoracolumbar burst fractures without neurologic deficits were included in this retrospective review of clinical and radiological outcomes after surgical management. All patients underwent combined percutaneous kyphoplasty and short fixation with screws in the vertebral pedicles above and below the fracture. Radiographic evaluation of segmental kyphosis and local kyphotic corrections were made preoperatively, 3 days postoperatively, 12 months post-operatively and at the last follow-up (the mean last follow-up was 41 months post-operatively). Clinical outcomes were determined by SF-36^® Health Survey and Oswestry Disability Index scores at 3-month and 12-month follow-ups.

Results

Clinical assessments suggested good outcomes as early as the third postoperative month. The clinical outcomes were sustained at one year follow-up. At the last follow-up the segmental kyphosis correction and local kyphotic correction were maintained.

Conclusions

Our analysis demonstrates that minimally invasive kyphoplasty and percutaneous short fixation applied to thoracolumbar A3 burst fractures without neurological deficit may achieve results comparable to nonsurgical or open surgical treatment, but with less morbidity and complication, and should be considered as a valid treatment option.

1. Introduction

Despite their frequency, the management of thoracolumbar fractures varies widely and remains controversial.¹ Specifically, there is a lack of a reproducible classification system, predicting the outcome of the treatment as well as the natural history of an injury pattern. This tends to limit the method of management of the thoracolumbar fractures to the experience of each practioner.²

Surgical management of thoracolumbar fractures is the consensus for patients with progressive neurological loss, unstable fractures, or polytrauma who require fixation for earlier and easier rehabilitation. Nevertheless, a universally accepted algorithm to decide the appropriate surgical technique remains lacking.

For neurologically intact patients, surgical management of thoracolumbar fracture remains controversial. Several authors have reported good clinical results after non-operative management without fracture reduction.³ Proponents of surgery believe that decompression, fracture reduction and stabilization are essential for stabilizing the spine and reducing the pain. However, for surgical management of thoracolumbar fractures in neurologically intact patients, there is no consensus regarding the timing of surgery, the approach and the points of stabilization (e.g., posterior only, anterior only or combined dorsal-ventral stabilization).

Internal fixation of unstable thoracolumbar spine fractures requires correction of the missing anterior column support. This usually entails the insertion of a vertebral body replacement strut through an anterior approach, or a long posterior construct spanning at least two vertebrae above and two vertebrae below the fracture. Each approach requires extensive exposure of the spine, with prolonged operative times and intraoperative blood loss.

Recently, minimally-invasive percutaneous posterior fixation has been used for trauma and tumor-related conditions. This procedure was initially developed to improve the functional results of surgery for degenerative spine disease. Indirect reduction of a fractured endplate, which can be achieved with a posterior instrumentation by traction on the annular fibers, will easily restore the cortical periphery, giving the impression of a good restoration of the vertebral body on antero-posterior and lateral radiographs.⁴ However, the central endplate portion cannot be reduced this way, leading to distorted disc space morphology. The nucleus can subsequently herniate through the fractured endplate and cause anterior column insufficiency. Consequently, the progressive collapse of the discal space characterizes the failure of such surgical treatment.⁵

The application of kyphoplasty in the fractured vertebra allows a transpedicular reduction and augmentation of the central endplate, preventing some complications of posterior short segment constructs and obviating the need of the high cost and morbidity associated with anterior surgery.^6,7 Recent findings suggest that intervertebral discs adjacent to traumatic burst fractures treated with pedicle screw instrumentation and direct end-plate restoration do not seem to routinely progress to severe degeneration at 12–18 months post-injury.⁸

Despite many studies suggesting the safety and efficacy of percutaneous fixation for thoracolumbar fracture, the full range of indications for these minimal invasive procedures has to be better documented and investigated. The objective of this study was to report the results of management of type A3 fractures of the OTA/AO classification system,⁹ through combined kyphoplasty and percutaneous fixation.

2. Methods

2.1. Patient selection

We retrospectively reviewed the data of 51 patients with single type A3 thoracolumbar burst fractures without neurologic deficits. These patients were admitted to our institution between the years 2011 and 2013 and evaluated preoperatively by means of X-Rays, CT scan and MRI.

Nineteen of these patients, with a greater than 20% spinal canal compromise, were excluded for the purpose of this study, as were treated with open decompression and posterior stabilization.

The remaining 32 patients (18 men and 14 women), with a lesser than 20% spinal canal compromise, were treated with a percutaneous minimally invasive technique and included in our study.

2.2. Data collection

These fractures have been classified according to OTA/AO classification⁹: 16 type A3.1, 3 type A3.2, and 13 A3.3. The injuries were located in T11 in 5 patients, T12 in 13 patients, L1 in 11 patients and L2 in 3 patients. Mean age of the patients was 45.5 (range, 24–72) years.

All patients underwent bilateral balloon kyphoplasty (Kyphon^®, Medtronic, Memphis, TN) and segmental percutaneous pedicle screw instrumentation (ILLICO^® MIS – SE system, Alphatec Spine, Inc., Carlsbad, CA) to reduce segmental kyphosis and restore vertebral body height. The assembly was short, with screws in the vertebral pedicles above and below the fracture.

For the vertebral body augmentation, we used polymethylmethacrylate in patients older than 60 years of age, polymethylmethacrylate and hydroxyapatite cement (Kyphon^® ActivOs™, Medtronic, Memphis, TN) in patients between their 30's and 50's, and calcium magnesium hydroxyapatite cement (KyphOs FS™, Medtronic, Medtronic, Memphis, TN) in patients under 30.

All patients were treated postoperatively with early mobilization (in the first or second postoperative day) and were ordered to wear a Jewett brace for 4 weeks when standing.

Radiographs taken in the supine position were evaluated preoperatively, 3 days postoperatively, 12 months postoperatively and at the last follow-up (the mean follow-up was 41 months post-operatively, range 17–60 months). For the segmental kyphosis correction, we used the Cobb angle determined by the angle formed by the perpendiculars of the extension lines of the upper endplate of the upper fractured vertebra and the lower endplate of the lower fractured vertebra. For the local kyphotic correction, we used the Cobb angle determined by the angle formed by the perpendiculars of the extension lines between the upper and the lower endplate of the fractured vertebra. Changes of the preoperative and postoperative parameters were analyzed by independent sample t-test. All tests were two-sided. P < 0.05 was considered to indicate a statistically significant difference.

Quality of life Italian SF-36^® Health Survey Scoring¹⁰ and Oswestry Disability Index (ODI) were used to evaluate the functional outcome at the third and twelfth postoperative month.

Ethics: This study was approved by the ethics committee of Ferrara (AOU Ferrara Italy 26/3/2015). Written informed consent to participate and to share data anonymously were obtained from all participants in accordance with the 1964 Helsinki declaration and its later amendments.

3. Results

Mean hospital stay was 7 days (minimum 6, maximum 10). All patients completed the follow-up. The only intraoperative complication observed was cement leakage, detected in 7 patients (an intradiscal leak in 3 cases and a paravertebral leak in 4 cases), without clinical consequences. No major complications (e.g., infection, bleeding, pulmonary embolism, neurologic deficit) were observed.

At the 1-year radiographic follow-up, we observed one case of pedicle screw breakage in the distal vertebra. We decided to remove the implant, but we failed to remove the portion of the broken screw from the vertebral body. The patient (male, 28 years old) did not show any clinical symptoms.

In four cases, we removed the implants twelve months or more after the fixation. Implant removal was easily performed using the cutaneous scar at each level, allowing screw and rod removal without extended approach.

We analyzed radiographs preoperatively and postoperatively (Fig. 1) in supine position. Contrast analysis was performed on Cobb angle changes preoperatively and at follow-up (Table 1). Three days after surgery, segmental kyphosis improved from an average preoperative kyphosis of 12.4°–7.5° (P < 0.05), and local kyphotic correction improved from an average preoperative kyphosis of 11.6°–7.8° (P < 0.05). In the 1-year follow-up, there was a slight mean loss of 0.7° for segmental kyphosis and 0.6° for local kyphotic correction (P > 0.05) compared to 3 days postoperatively. At the last follow-up the local kyphotic correction was 8.1° and the segmental kyphosis was 8.2°. The final correction maintained a statistically significant difference compared to the preoperative (P < 0.05) with no difference in terms of cement used.

Fig. 1.

Preoperative and postoperative radiographs of one patient with type A3.2 fracture. The patient with L1 type A3.2 fracture was treated with kyphoplasty at L1 vertebra and pedicle screw fixation in adjacent vertebrae. A-C, Preoperative transverse views of the L1 vertebra. D and E, Preoperative radiographs. F and G, Postoperative radiographs.

Table 1.

Results of radiographic evaluation of the segmental kyphosis and local kyphotic correction.

	Preoperatively	3 Days Postoperatively	12 Months Postoperatively	Last Follow-up (mean 41 months)
Segmental Kyphosis	12.4°	7.5°*	8.2°*	8.1°*
Local Kyphotic Correction	11.6°	7.8°*	8.4°*	8.2°*

*P < 0.05 in comparison to preoperative.

Clinical outcome assessments showed good results at the third postoperative month as confirmed by norm-based¹⁰ comparison of the eight SF-36^® Health Survey scales (Fig. 2) and by the low degree of disability as assessed by Oswestry Disability Index (ODI) scoring (Table 2). SF-36^® Health Survey and ODI scores at the 12-month follow-up demonstrated a sustained improvement of clinical outcomes. By the one-year follow-up, 100% of the patients resumed work activities held before the injury.

Fig. 2.

Follow-up SF-36^® Survey Scores. The mean values for SF-36^® Survey Scores were derived from the 32 patients with type A3 burst fractures and plotted. Black bars and grey bars represent SF-36^® Survey Scores at 3-month and 12-month follow-up, respectively. The pink lines indicate the normal scores for the general Italian population. Error bars indicate standard deviation (SD).

Table 2.

Postoperative Oswestry disability index (ODI) scores.

	3 Months Postoperatively	12 Months Postoperatively
ODI Score (±SD)	18.3 ± 3.6	14.7 ± 2.5

4. Discussion

A number of techniques have been described in the management of thoracolumbar spinal fractures¹¹ For thoracolumbar fractures with progressive neurological loss, unstable fractures, or polytrauma, surgical treatment is becoming the consensus. In terms of AO-Magerl Classification, type B and type C fractures are treated surgically. However, for type A fractures, often presented without neurological deficit, considerable controversy remains over their treatment options.

Prospective studies directly comparing operative and nonoperative treatment of thoracolumbar burst fracture without neurological deficit are equivocal in clearly demonstrating the superiority of either techniques for various outcomes.^12,13

Siebenga et al. reported better clinical outcome in the operative group.¹⁴ Wood et al. on the other hand, reported similar clinical outcomes in both groups, although complications were more frequent in the operative group.¹⁵ It is noteworthy, that the operative techniques in these studies involved standard open posterior pedicle fixation with autologous bone harvesting, both of which are substantially invasive procedures.

Successful non-surgical treatments of burst fractures without neurological deficits have been well documented. However, many of these studies do not adequately address the fact that conservative treatments, which can vary widely, place substantial burden on patients for a long duration. For instance, in a well cited report advocating conservative treatment for burst fractures without neurological deficits, most of the patients were in nonambulatory body cast for 2–4.5 months.¹⁶ Other examples include 3 months in Jewett hyperextension orthosis,¹⁴ and 2–3 months of molded plaster cast followed further by 1–2 months of orthosis, that were not tolerated by all patients.¹⁵

With an open posterior approach to stabilize the spine surgically, detaching muscles and continuously aggressively retracting them has been implicated in muscle denervation and devascularisation, leading to muscle atrophy.¹⁷ These complications can result in postoperative pain and muscle strength loss, which delays functional recovery and can result in chronic pain18.

Moreover, isolated posterior stabilization is often associated with a loss of correction and implant failure.¹⁹ The anterior approach has been proposed to correct the kyphosis and achieve bone graft fusion, in order to establish the fused segment.²⁰ However, this surgery requires a high level of skills and is not without complication.

Augmentation of posterior instrumentation with kyphoplasty has been proposed as a safe and less invasive alternative to anterior approach for the treatment of burst fractures without neurologic compromise.²¹ Verlaan et al. reported significant reduction of the Cobb angle and restoration of the vertebral body height. No serious complications were observed.

Korovessis et al. reported a prospective study of 23 patients with A3 thoracolumbar burst fractures with and without neurologic deficit treated with posterior reduction and fixation and balloon kyphoplasty with calcium phosphate cement.²² The authors reported a significant improvement of sagittal alignment with a correction of kyphosis from 16° preoperatively to 1° postoperatively.

Several techniques have been used to augment the stability of posterior implants, including spinal fusion. However, evidence of the need for posterior spinal fusion remains inconclusive. Several studies recommend posteriori fusion,²³ but others have claimed that it affords no benefit.^24,25

With the availability of percutaneous surgery, it is now possible to minimize the disadvantages of traditional surgery, to spare the paravertebral musculature, reducing bleeding, infection rates and postoperative pain. Clinical outcomes of percutaneous fixation and kyphoplasty for burst fractures without neurological deficit have been positive. As assessed by VAS, patients displayed improvement as early as time of discharge (3–7 days).^26,27 The rapid improvement in VAS was sustained through the last follow-up, averaging around 2 years. Another study with a 15-month mean follow-up VAS also found significant improvement in patients who underwent combined percutaneous fixation and kyphoplasty for the treatment of thoracolumbar fractures without neurological deficit.²⁸

In our study, we report additional clinical measures, ODI and SF-36^® Health Survey, which show satisfactory clinical outcomes as early as 3 months and persist into the 1-year follow-up. Together, these studies cover various clinical outcome measures in a range of follow-up periods to strongly support the safety and effectiveness of combined kyphoplasty and percutaneous fixation in the treatment of thoracolumbar burst fractures in the absence of neurological deficit.

Along with the obvious advantages, disadvantages of combined kyphoplasty and percutaneous fixation do exist and must be considered. Proper percutaneous screw placement requires a substantial technical experience and quality fluoroscopy.

One must also be able to perform a kyphoplsty without creating neurological complications, especially in the presence of variable spinal canal compromise. The risk of cement leakage into the spinal canal in A3 type fractures exist, and there appears to be no clear consensus when to perform this procedure safely. Based on our experience, we believe kyphoplasty can be performed without neurological complications when the spinal canal compromise is less than 20%.

Additional consideration for treatment with combined kyphoplasty and fixation is the decision regarding implant removal. Removal of the fixation system aims to free up the immobilized thoracolumbar level. Although no consensus exists on whether this is necessary,^29,30 it seems reasonable to remove the instrumentation at 8–12 months, when the fracture is healed. This might be recommended in particular for younger patients to preserve mobility at adjacent levels.

We removed the implants in 4 (18%) of our patients 1 year postoperatively at their discretion. This was performed percutaneously without any complications. Wild et al. removed the instrumentation in all of their patients after fixation and observed partial loss of intervertebral space correction during the year after the removal, although this loss did stabilize over time.³¹

In conclusion our study and the literature demonstrate that treatment of thoracolumbar burst fractures without neurological loss with percutaneous fixation in combination with balloon kyphoplasty provides satisfactory, long-lasting fracture reduction with excellent clinical results. Importantly, the results of our study demonstrate that good clinical results as measured by ODI and SF-36^® Health Survey can be achieved as early as 3 months. The rapidity with which patients with thoracolumbar burst fractures without neurological deficit achieve good clinical outcomes after combined percutaneous fixation and kyphoplasty has not been demonstrated for conservative treatment. Although more studies are needed, there is some evidence to consider the possibility that the conservative care generates indirect costs (e.g. delayed return to work) which may far exceed the direct cost of surgery for patients.³²

Our study adds to the growing evidence that the combined kyphoplasty and percutaneous fixation offers patients a safe and effective alternative to conservative treatment or open surgical treatment of thoracolumbar burst fracture without neurological deficit. Using this technique, we found that patients achieved a faster clinical improvement not expected with conservative management, without the co-morbidities associated with open surgery or conservative method.

Obviously the main limitation of this paper is being a retrospective analysis of the population of patient without a control group and therefore can't be conclusive.

Therefore prospective and randomized studies with a larger number of patients will be important to rigorously assess various potential advantages of combined percutaneous fixation and kyphoplasty over many other methods of treatment of thoracolumbar burst fractures without neurological deficit.

Conflict of interest statement

The authors declare that they have no competing interests.