Pedicle screw shift without loosening following instrumented posterior fusion: limitations of pedicle screw fixation

Abstract

The disc angle at the fused segment is extended in operative prone position, but eventually returns to preoperative neutral position within 6 months to 1 year. This study aimed to assess pedicle screw (PS) shift without loosening to identify the mechanism of the change in disc angle after posterior fusion for degenerative lumbar spondylolisthesis (DLS). Sixty-three consecutive patients who underwent facet fusion for L4 single-level DLS were retrospectively reviewed using computed tomography (CT) immediately after surgery and 6 months postoperatively. Twenty-two patients (88 PSs) in whom the disc angle had decreased by more than 4° at 6-month postoperative radiographic follow-up were selected to more readily identify and quantify PS shift. Six patients with PS loosening and/or nonunion were excluded. We reconstructed a CT plane, vertical to the cranial endplate of the vertebrae and passing through the cannula used for percutaneous PSs. Angle α, which is formed by the cranial endplate and the cannula on this plane, was measured. A change in angle α of more than 2° between the immediate postoperative period and the 6-month follow-up was defined as a PS shift. Angle α did not change by more than 2° in any of the 44 PSs in the upper vertebrae of the fused segment. In the lower vertebrae, angle α changed by more than 2° in 22 of 44 PSs. The change in angle α in the lower vertebrae (average, 2.3°) was significantly greater than that in the upper vertebrae (average, 0.2°) (P < 0.0001). The change in the disc angle was not relevant to clinical outcomes assessed by the Japanese Orthopaedic Association Back Pain Evaluation Questionnaire, the Roland-Morris Disability Questionnaire, and the visual analogue scale. The disc angle at the fused segment returned to preoperative neutral position due to PS shift without loosening, mainly in the lower vertebrae. PS shift is caused by bone remodeling in response to biomechanical load, similar to that in orthodontic tooth movement. As PS has limited ability to maintain a lordotic disc angle, even with the insertion of a cage, lumbar alignment will return to preoperative neutral position owing to cage subsidence or adjacent segment disease. These findings might indicate that it is not necessary to correct the spinal alignment for DLS.