BACKGROUND:Loss of axial rotation and lateral bending after atlantoaxial fusion reduces a patient's quality of life. Therefore, effective, nonfusion fixation alternatives are needed for atlantoaxial instability.OBJECTIVE:To evaluate the initial stability and function of posterior atlantoaxial restricted nonfusion fixation (PAARNF), a new protocol, using cadaveric cervical spines compared with the intact state, destabilization, and posterior C1-C2 rod fixation.METHODS:Cervical areas C0 through C3 were used from 6 cadaveric spines to test flexion-extension, lateral bending, and axial rotation range of motion (ROM). With the use of a machine, 1.5-Nm torque at a rate of 0.1 Nm/s was used and held for 10 seconds. The specimens were loaded 3 times, and data were collected in the third cycle and tested in the following sequence: (1) intact, (2) destabilization (using a type II odontoid fracture model), (3) destabilization with PAARNF (PAARNF group), and (4) rod implantation (rod group). The order of tests for the PAARNF and rod groups was randomly assigned.RESULTS:The average flexion-extension ROM in the PAARNF group was 7.44 2.05 degrees, which was significantly less than in the intact (P = .00) and destabilization (P = .00) groups but not significantly different from that of the rod group (P = .07). The average lateral bending ROM (10.59 2.33 degrees; P = .00) and axial rotation ROM (38.79 13.41 degrees; P = .00) of the PAARNF group were significantly greater than in the rod group. However, the values of the PAARNF group showed no significant differences compared with those of the intact group.CONCLUSION:PAARNF restricted atlantoaxial flexion-extension but preserved axial rotation and lateral bending at the atlantoaxial joint in a type II odontoid fracture model. However, it should not be used clinically until further studies have been performed to test the long-term effects of this procedure.ABBREVIATIONS:PAARNF, posterior atlantoaxial restricted nonfusion fixationROM, range of motion