Accuracy of minimally invasive navigated acetabular and iliosacral fracture stabilization using a targeting and noninvasive registration device

To assess the feasibility and accuracy of guide pin (GP) placement using a combined noninvasive patient immobilization and stereotactic targeting system in computer-assisted percutaneous pelvic fracture stabilization.

A total of 12 patients with negligible dislocated unstable pelvic fractures were enrolled in this study, performed between February 2002 and October 2005. Our original plans included 13 GP placements in the iliosacral area (SF) and 8 in the acetabular (AF) area. Patients were bedded on a noninvasive dual-vacuum immobilization device. Interventions were planned on a navigation system using intraoperatively acquired CT data. Radiodense markers glued to the skin and the immobilization device provided synchronization between virtual data set and real anatomical situation. A stereotactic targeting device was used for stabilization of GP tracking. GP positions were verified intraoperatively by CT, followed by fracture stabilization with cannulated screws.

Mean GP placement accuracy according to plan: (1) SF-cohort: 2.8 mm (SD 2.0 mm, range 0.5–9.0 mm) at the bony entry point and 3.8 mm (SD 2.3 mm, range 0.6–9.5 mm) at the target point. (2) AF-cohort: 3.0 mm (SD 0.9 mm, range 1.6–4.9 mm) at the bony entry point and 3.9 mm (SD 1.9 mm, range 1.6–7.5 mm) at the target point. GP placement succeeded optimally in 11 out of 13 cases in the SF-cohort, and 6 out of 8 cases in the AF-cohort. The individual average dose–length product (DLP) per successful finished procedure was 1,576 mGy × cm (SD 812 mGy × cm, range 561–2,739 mGy × cm).

Our findings substantiate application of the noninvasive patient immobilization and stereotactic targeting system as effective in computer-assited percutaneous stabilization of sacral bone fractures/SI joint disruptions and coronally oriented acetabular dome fractures. We recommend according to the ALARA (as low as reasonable achievable) principle: first, the kV and mAs values have to be reduced. Second, the scanned volume has to be strictly limited to the area of interest. Third, the number of control CTs have to be minimized. Also, the IsoC might be a better choice for implant tracking below 12 cm to reduce the radiation dose to the minimum. We believe that for all high-precise GP placements in the acetabular column area, further improvements in GP guidance (inhibiting pin tip slipping and detecting intraosseous GP deflection) are necessary.