Biomechanics of open versus minimally invasive deformity correction:​ comparison of stability and rod strain between pedicle subtraction osteotomy and anterior column realignment.

ABSTRACT

OBJECTIVE: Anterior column realignment (ACR) is a new minimally invasive approach for deformity correction that achieves a degree of lordosis similar to that obtained with pedicle subtraction osteotomy (PSO). This study compared the biomechanical profiles of ACR with PSO using range of motion (ROM) and posterior rod strain (RS) to gain insight into the ACR technique and the necessary surgical strategies to optimize longevity and stability. METHODS: An in vitro biomechanical study using standard flexibility testing (7.5 Nm) was performed on 14 human cadaveric specimens, separated into 2 groups similar in age, sex, bone mineral density, and intact ROM. For group 1 (n = 7, instrumented L1-S1), a 30° ACR was performed at L3-4. For group 2 (n = 7, instrumented T12-S1), a 30° L3 PSO was performed. Specimens were subjected to nondestructive loads in flexion, extension, axial rotation, lateral bending, and compression. Conditions tested were 1) intact, 2) pedicle screw with 2 rods (PSR), 3) ACR or PSO with 2 rods (+2R), and 4) ACR or PSO with 4 rods (+4R). Primary outcome measures of interest were ROM stability and posterior RS at L3-4. RESULTS: No difference was observed between groups in lumbar lordosis (p = 0.83) or focal angular lordosis at L3-4 (p = 0.75). No differences in stability were observed between ACR+2R and PSO+2R (p ≥ 0.06);​ however, ACR+2R was significantly less stable than PSR in flexion and extension (p ≤ 0.02), whereas PSO+2R was less stable than PSR only in extension (p = 0.04). ACR+4R was more stable than ACR+2R in flexion, extension, left axial rotation, and compression (p ≤ 0.02). PSO+4R was more stable than PSO+2R only in extension (p = 0.04). Both ACR+2R and PSO+2R resulted in significant increases in RS in flexion and extension compared with PSR (p ≤ 0.032). RS in flexion and extension decreased significantly for ACR+4R versus ACR+2R and for PSO+4R versus PSO+2R (p ≤ 0.047). PSO+2R yielded lower RS than ACR+2R in compression (p = 0.03). No differences existed in RS between ACR+4R and PSO+4R (p ≥ 0.05). CONCLUSIONS: Although ACR appeared to be slightly more destabilizing than PSO using traditional 2R fixation, both techniques resulted in significant increases in posterior RS. The 4R technique increased stability in ACR and decreased RS in both ACR and PSO but may be more beneficial in ACR. Longer-term clinical studies are needed to appropriately identify the durability of the ACR technique in deformity correction.