Introduction and Aims: Speculation exists with regard to the exact mechanism of remodelling of thoracolumbar burst fractures treated non-operatively. We prospectively evaluated spinal canal remodelling in 30 patients with burst fractures by measuring the epidural pressure following ethical approval.

Method: Thirty-four patients (average age 37 years) were recruited into the study. The injury followed a fall from a height in 18 patients; and 12 resulted from a road traffic accident. All patients were neurologically intact. Plain X-rays and CT scans were obtained to evaluate the injuries. The patients were treated non-operatively with orthoses. At two weeks and 12 months post-injury epidural pressures were measured in theatre. The measurements were performed in the lateral decubitis position, and a fluoroscopically guided radio-opaque catheter was positioned below the fracture site to record the epidural pressure.

Results: Thirty patients were included in the study and four were excluded due to inadequate follow-up. The average progress in the Cobb’s angle at one-year follow-up was 2.690 (range 10–60). At the time of injury the mean canal compromise measured on CT scans was 43% (range 32%–83%) and at follow-up had improved to 28% (range 44%–100%). The CT volumetric measurements showed an average improvement of 10% in volume at follow-up (range 7%–16%).

The average epidural pressure recorded at the time of injury was 16.65mmHg (range 2.5–30.85mmHg) and at follow-up was –5.85mmHg (range 0 to –10.17mmHg). There were no complications following epidural pressure monitoring.

Conclusion: Scapinelli and Candiotti hypothesised that burst fractures remodelled secondary to respiratory oscillations transmitted to the epidural space. The raised epidural pressure (p< 0.001), by virtue of its mechanical effect, may be one of the factors responsible for the remodelling of burst fractures.

The exact mechanism of remodelling of burst fractures is uncertain. We studied the relationship between epidural pressures and remodelling.

In a prospective, ethically-approved study in 34 patients with burst fractures at the levels T12 to L4, epidural pressure was measured. Four patients were lost to follow-up. In 18 patients the fractures were due to a fall and in 12 to motor vehicle accidents. The mean age was 37 years. All patients were neurologically intact and treated non-operatively. Plain radiographs and CT scans measuring the sagittal plane deformity and mid-sagittal diameter respectively were obtained. Using a fluoroscopically-guided radio-opaque catheter placed at the normal interspace below the burst fracture, epidural pressure was measured at 2 weeks and at 12 months after the injury. The mean canal compromise shown on CT scan at the time of injury and at follow-up was 43% (32% to 83%) and 28% (44% to 100%) respectively. CT volumetric measurements showed a mean improvement of 10% at follow-up (7% to 16%). The epidural pressures recorded at the time of injury and at follow-up were 16.65 mmHg (2.5 to 30.85) and 5.85mmHg (0 to 10.17) respectively. At 1-year follow-up, the Cobbs angle had progressed by a mean of 2.69° (10° to 60°).

The retropulsed burst fracture fragments cause localised constriction of the spinal cord (Venturi effect). Epidural pressure, raised to maintain a constant flow rate across this constriction, has a mechanical effect on the retropulsed fragments, thus promoting remodelling.

Several studies have reported that remodelling of the spinal canal occurs in lumbar burst fractures following non-operative treatment. Various theories have been proposed for spinal canal remodelling, including the possible effect of the oscillatory pulsations of the subdural space, but no studies have been done to evaluate this effect.

In a prospective study between September 1999 and April 2002, we evaluated 17 men and seven women, with a mean age of 35.25 years (19 to 59), who had sustained a burst fracture in the upper lumbar region. The fractures were at the L1 and L2 regions in 14 and 10 patients respectively. The epidural pressure and radiological appearances were initially evaluated approximately two weeks after injury, and again 12 months after injury. All patients were neurologically intact and treated non-operatively.

CT evaluation of the initial injury showed a mean initial canal compromise of 49.81% (22.3% to 80%) as measured by mid-sagittal diameter and 13.9% (8.2% to 16.9%) as measured by volumetric assessment, with a mean epidural pressure of 14.56mmHg (2.5 to 30.38). At follow-up 12 months later, the mean epidural pressure was -4.67mmHg (−1.1 to −8.9) and the mean canal compromise as measured by the mid-sagittal diameter and volumetric measurements on CT scan were 24.56% and 8.9% respectively.

Our data show that the epidural pressure was raised in acute burst fractures and reverts to normal with remodelling. We can conclude that the raised epidural pressure may be one of the mechanisms that contribute to the remodelling process.