Biomechanical evaluation of Anterior spinal instrumentation systems for scoliosis: In vitro fatigue simulation

A biomechanical study was designed to assess the bone-screw interface fixation strength among five anterior spinal instrumentation systems for scoliosis before and after a fatigue simulation. The objectives of the current study were twofold: 1) evaluate the static (initial) strength at the bone-screw interface and 2) evaluate dynamic (post fatigue) strength of the bone-screw interface after a fatigue simulation to investigate a possible mechanism for postoperative loss of correction. Although the recent advancement of anterior instrumentation for scoliosis has permitted shorter fusion segments and improved surgical correction, the loss of correction over the instrumented segments still has been reported in one-rod systems. Little is known about the mechanism for loss of correction. Twenty-five fresh-frozen calf spines (T6-L6) were used. A total of five instrumentation systems included the following: Anterior ISOLA (ISOLA), Bad Wildungen Metz (BWM), Texas Scottish Rite Hospital system (TSRH), Cotrel-Dubousset Hoph (CDH), and Kaneda Anterior Scoliosis System (KASS). Screw pullout and rotational tests in the sagittal plane using a single vertebra were performed to investigate bone-screw interface fixation strength before and after a fatigue simulation. To simulate cyclic loading that the spine could undergo in vivo, a fatigue simulation using compressive-flexion loading up to 24,000 cycles was carried out. Mean maximum tensile pullout force decreased in the following order: KASS > CDH > BWM > TSRH > ISOLA (F = 29.91, P < 0.0001). KASS blunt tip screw was 26% stronger in pullout force than KASS sharp tip screw (P < 0.05). The one-rod system demonstrated a positive correlation between pullout force and both bone mineral density and screw insertional torque. For fatigue analysis the rotational strength at the most cephalad and caudal segments significantly decreased after a fatigue simulation in the one-rod system (P < 0.05). The two-rod system showed no significant decrease after a fatigue simulation. Simulating the cyclic loading to the construct, screw loosening at the bone-screw interface was produced in the one-rod system. This screw loosening may elucidate one mechanism for loss of correction in the one-rod system. The two-rod system may have the potential to minimize the risk of loss of correction.