Adjacent-segment effects of lumbar cortical screw–rod fixation versus pedicle screw–rod fixation with and without interbody support

Abstract

OBJECTIVE Cortical screw–rod (CSR) fixation has emerged as an alternative to the traditional pedicle screw–rod (PSR) fixation for posterior lumbar fixation. Previous studies have concluded that CSR provides the same stability in cadaveric specimens as PSR and is comparable in clinical outcomes. However, recent clinical studies reported a lower incidence of radiographic and symptomatic adjacent-segment degeneration with CSR. No biomechanical study to date has focused on how the adjacent-segment mobility of these two constructs compares. This study aimed to investigate adjacent-segment mobility of CSR and PSR fixation, with and without interbody support (lateral lumbar interbody fusion [LLIF] or transforaminal lumbar interbody fusion [TLIF]). METHODS A retroactive analysis was done using normalized range of motion (ROM) data at levels adjacent to single-level (L3–4) bilateral screw–rod fixation using pedicle or cortical screws, with and without LLIF or TLIF. Intact and instrumented specimens (n = 28, all L2–5) were tested using pure moment loads (7.5 Nm) in flexion, extension, lateral bending, and axial rotation. Adjacent-segment ROM data were normalized to intact ROM data. Statistical comparisons of adjacent-segment normalized ROM between two of the groups (PSR followed by PSR+TLIF [n = 7] and CSR followed by CSR+TLIF [n = 7]) were performed using 2-way ANOVA with replication. Statistical comparisons among four of the groups (PSR+TLIF [n = 7], PSR+LLIF [n = 7], CSR+TLIF [n = 7], and CSR+LLIF [n = 7]) were made using 2-way ANOVA without replication. Statistical significance was set at p < 0.05. RESULTS Proximal adjacent-segment normalized ROM was significantly larger with PSR than CSR during flexion-extension regardless of TLIF (p = 0.02), or with either TLIF or LLIF (p = 0.04). During lateral bending with TLIF, the distal adjacent-segment normalized ROM was significantly larger with PSR than CSR (p < 0.001). Moreover, regardless of the types of screw-rod fixations (CSR or PSR), TLIF had a significantly larger normalized ROM than LLIF in all directions at both proximal and distal adjacent segments (p ≤ 0.04). CONCLUSIONS The use of PSR versus CSR during single-level lumbar fusion can significantly affect mobility at the adjacent segment, regardless of the presence of TLIF or with either TLIF or LLIF. Moreover, the type of interbody support also had a significant effect on adjacent-segment mobility.