Introduction and Aims: Acute intramedullary stabilisation of femoral shaft fractures in multiply injured patients remains controversial. Intravasation of medullary fat during nailing has been suspected to trigger ARDS. This study investigates the effect of a filter placed into the ipsilateral common iliac vein during medullary canal pressurisation in a canine experiment.

Method: Using an established model, 12 mongrel dogs were randomised into two groups. Under general anesthesia, cannulations were performed to measure left and right-sided pressures. Transoesophageal echocardiography was performed in all dogs. A special ‘TRAP ‘filter was inserted percutaneously into the left common iliac vein in six dogs, whereas the other six served as controls. In all dogs, the left femora and tibiae were then pressurised by injection of bone cement and insertion of intramedullary rods. Hemodynamic measurements and echocardiography images were recorded continuously. After one hour, the animals were sacrificed and the lungs were harvested for histomorphommetric analysis.

Results: The mean pulmonary artery pressure at three minutes of pressurisation was 12mm of Hg in the filter group and 28mm of Hg in the control group. There was increase in the peak systolic pulmonary artery pressure and the right ventricular pressure after canal pressurisation in the control group, whereas no such changes were observed in the filter groups. The pulmonary vascular resistance as denoted by the difference between the mean pulmonary artery pressure and the end diastolic left ventricular pressure increased significantly (p< 0.05) at three, five, 10,15 and 30 minutes after pressurisation in the control group when compared to the baseline value. In the filter group, the pulmonary vascular resistance increased only slightly after pressurisation. Transesophageal echocardiography images were analysed by a blinded echocardiologist. There was evidence of moderate to severe embolisation in the control group with detection of large echogenic particles. In comparison, there was mild grade of embolisation in the filter group. Histological analysis showed statistically significant difference between the two groups, when comparison of the percentage of area of lung tissue occupied by fat, the percentage of pulmonary vasculature occupied by fat and the maximum size of the embolus were made (p< 0.05).

Conclusion: This study has conclusively demonstrated that mechanical blockade by venous filters prior to medullary canal pressurisation, significantly reduces the embolic load and its effect on the lungs. A retrievable filter with a system to remove the accumulated marrow content is being developed for use in high-risk patients.