Introduction Vertebroplasty (VP), where vertebral bodies are injected with polymethylmethacrylate (PMMA) cement, is used to treat various spinal lesions. More recently VP has been used for augmenting osteoporotic vertebral bodies that have fractured or are at risk of fracture. Although the complication rate for VP is low, thermal damage caused by the exothermic curing of PMMA has been implicated.

The aim of this series of experiments was to measure the temperatures reached during VP using a sheep model. The cement volume effect and inter cement differences were assessed. Spinal cord monitoring was undertaken to monitor spinal cord function during this procedure to validate this for clinical use.

Methods In the in vivo experiment each of the lumbar vertebral bodies of 10 sheep were injected with one of two cements (Simplex & Vertebroplastic) and one of two volumes (3.0ml or 6.0ml). This was undertaken through an open approach in the lumbar vertebrae. While performing the in vivo experimental studies 6 of the sheep were concurrently monitored using epidural Motor Evoked Potentials (MEP’s).

Results There was a significant increase in the temperature at the bone cement interface. The mean peak temperature at the bone-cement interface was 49.5 C (3.0ml Simplex); 61.47 C (6.0ml Simplex); 42.1 C (DePuy 3ml) and 47.2 (DePuy 6ml).

Spinal cord monitoring showed that when PMMA was injected into the correct location within the vertebral body there was no change in amplitude of the evoked potentials. When significant leakage of PMMA occurred, there was a decrease in amplitude of MEP’s.

Discussion In this sheep model, using cement volumes similar to those used in human clinical practice, we were able to monitor temperature changes within the vertebral body at the bone cement interface. The temperature of the bone cement interface reached temperatures that are known to cause tissue necrosis.

Using epidural monitoring we were able to show that when PMMA is injected into the correct location within the vertebral body there is no change in amplitude of MEP’s.

Percutaneous vertebroplasty (PVP), where vertebral bodies are injected with polymethylmethacrylate (PMMA) cement, is used to treat various spinal lesions. Although the complication rate for PVP is low, thermal damage caused by the exothermic curing of PMMA has been implicated.

This study was to measure the temperatures reached during PVP as PMMA cures as well as assessing the cement volume effect and inter cement differences. Validating spinal cord monitoring during PVP was also undertaken.

In the in vivo experiment each of the lumbar vertebral bodies of 10 sheep were injected with one of two cements and one of two volumes. Thermocouple monitoring was undertaken at the bone cement interface. While undertaking the in vivo experimental studies 6 sheep underwent epidural monitoring using Motor Evoked Potentials (MEPs).

The mean peak temperature at the bone-cement interface was 49.5 C (3.0ml Simplex); 61.47 C (6.0ml Simplex); 42.1 C (DePuy 3ml) and 47.2 (DePuy 6ml). Spinal cord monitoring showed that when cement was injected into the correct location within the trabeculae of the vertebral body no change in amplitude monitoring was noted. When leakage occurred, deliberate or unintended, amplitude changes were noted.

Using cement volumes similar to those used in human clinical practice in a sheep model we were able to monitor temperature changes. The temperature of the bone cement interface reached temperatures that are known to cause tissue necrosis. Using epidural monitoring we were able to detect leakage of cement during injection.

The aim was to compare the initial pull out strengths of various interference screw devices used for tibial fixation of hamstring grafts and the effect of concentric or eccentric screw position.

Quadrupled tendon grafts were harvested from freshly killed sheep. The grafts were then prepared and fixed in the distal femur using various devices (Intrafix, RCI screw, Wedge screw +/− transfix pin, screw and post) in both concentric and eccentric positions. A single load to failure test was then performed.

The highest pull out strength was with the Intrafix device inserted concentrically (mean 941N). This was significantly higher than the wedge screw inserted concentrically (737N) (p=0.015). This in turn had significantly greater initial pull out strength than the wedge eccentric with post or pin (p=0.03) and the RCI screw (464N) (p=0.00036).

In this sheep model the Intrafix device inserted concentrically had a significantly greater initial pull-out strength than the other interference screws tested. Concentric positioning of an interference screw gave significantly greater initial pullout strength of a quadruple hamstring graft than eccentric positioning. Addition of a cross pin or post made no difference to initial pullout strength.