T9 versus T10 as the upper instrumented vertebra for correction of adult deformity-rationale and recommendations.

BACKGROUND CONTEXT: Adult spinal deformity correction sometimes involves long posterior pedicle screw constructs extending from the lumbosacral spine to the thoracic vertebra. As fusion obliterates motion and places supraphysiological stress on adjacent spinal segments, it is crucial to ascertain the ideal upper instrumented vertebra (UIV) to minimize risk of proximal junctional failure (PJF). The T10 vertebra is often chosen to allow bridging of the thoracolumbar junction into the immobile thoracic vertebrae on the basis that it is the lowest immobile thoracic vertebra strut by the rib cage.

PURPOSE: This study aimed to characterize the range of motion (ROM) of each vertebral segment from T7 to S1 to determine if T10 is truly the lowest immobile thoracic vertebra.

STUDY DESIGN/SETTING: This is a prospective, comparative study.

PATIENT SAMPLE: Seventy-nine adults (mean age of 45.4 years) presenting with low back pain or lower limb radiculopathy or both, without previous spinal intervention, metastases, fractures, infection, or congenital deformities of the spine, were included in the study.

OUTCOME MEASURES: A ROM >5° across two vertebral segments as determined by the Cobb method from radiographs.

METHODS: Lumbar flexion-extension and neutral erect radiographs were obtained in randomized order using a slot scanner. Segmental ROM was measured from T7-T8 to L5-S1 and analyzed for significant differences using t tests. Age, gender, radiographical indices such as standard spinopelvic parameters, sagittal vertical axis (SVA), C7-T12 SVA, T1 slope, thoracic kyphosis (TK), and lumbar lordosis (LL) were studied via multivariate analysis to identify predictive factors for >5° change in ROM at the various segmental levels. There were no sources of funding and no conflicts of interest associated with this study.

RESULTS: In the thoracolumbar spine, significant decreases in ROM when compared with the adjacent caudad segment occurs up to T9-T10, with mean total ROM of 1.98±1.47° (p<.001) seen in T9-T10, 2.19±1.67° (p<.001) in T10-T11, and 3.92±3.21°(p<.001) in T11-T12. The total ROM of T8-T9 (2.53±1.79°) was not significantly different from that of T9-T10 (p=.261). At the thoracolumbar junction, absence of scoliosis (OR 11.37, p=.020), high pelvic incidence (OR 1.14, p=.046), and low T1 slope (OR 1.45, p=.030) were predictive of ROM >5°.

CONCLUSIONS: Lumbar spine flexion-extension ROM decreases as it approaches the thoracolumbar junction. T10 is indeed the lowest immobile thoracic vertebra strut by the rib cage, and the last significant decrease in ROM is observed at T9-T10, in relation to T10-T11. However, because this also implies that a UIV of T10 would mean there is only one level of fixation above the relatively mobile segment, while respecting other factors that influence UIV selection, we propose the T9 vertebra as a more ideal UIV to fulfill the biomechanical concept of bridge fixation. However, this decision should still be taken on a case-by-case basis.