Should pelvic incidence influence realignment strategy? A detailed analysis in adult spinal deformity

BACKGROUND CONTEXT: Pelvic incidence (PI) serves as the cornerstone for realignment schema to create a more individualized realignment target. Yet, it is not known if outcomes of realignment schema are dependent on the amount of pelvic incidence. The purpose of this study is to assess how varying realignment strategies affect mechanical failure and clinical outcomes in PI-stratified cohorts following ASD surgery. PURPOSE: The purpose of this study is to assess how varying realignment strategies affect mechanical failure and clinical outcomes in PI-stratified cohorts following ASD surgery. STUDY DESIGN/SETTING: Retrospective cohort study; Single academic center. PATIENT SAMPLE: A total of 445 adult patients met radiographic criteria for adult spinal deformity. OUTCOME MEASURES: Mechanical failure (either a major hardware failure requiring intervention or proximal junctional failure [PJF]); Clinical Improvement at two years: [meeting either: (1) Substantial Clinical Benefit for Oswestry Disability Index (change >18.8), or (2) Oswestry Disability Index <15 and Scoliosis Research Society Total>4.5]; Good Outcome involved meeting Clinical Improvement criteria with absence of mechanical failure by two years. METHODS: Conditional Inference Tree (CIT) analysis was utilized to define subsets within pelvic incidence generating significantly different rates of mechanical failure. These subsets of pelvic incidence were further analyzed as sub-cohorts for the outcomes and effects of realignment within each. Multivariate logistic regression analysis controlling for baseline frailty and lumbar lordosis (L1-S1) analyzed the association of age-adjusted realignment (Sagittal Age-Adjusted Score [SAAS]; Lafage et al) and Global and Alignment Proportionment (GAP; Yilgor et al) strategies with meeting Good Outcome within PI-stratified groups. RESULTS: Using CIT analysis, a parabolic relationship between PI and mechanical failure was seen, whereas patients with either less than 51° (n=174; 39.1% of cohort) or greater than 63° (n=114; 25.2% of cohort) of pelvic incidence generated higher rates of mechanical failure (18.0% and 20.0%, respectively) and lower rates of Good Outcome (80.3% and 77.6%, respectively) than those with moderate (51-63°) pelvic incidence (n=145, 32.6% of cohort; 8.9% mechanical failure, 92.2% Good Outcome). Patients with Lower PI (<51°) more often met Good Outcome when undercorrected in age-adjusted PI-LL and SAAS overall (12.3% vs 0.0%; p=.004). Patients not meeting Good Outcome in this group were more likely to deteriorate in GAP Relative Lordosis from first to final follow-up (OR: 13.4, 95% CI: 1.3-39.2), leading to a higher likelihood of mechanical failure (OR: 3.2, 95% CI: 1.34-7.52; p=.009). In those with moderate pelvic incidence (51-63°), patients were more likely to meet Good Outcome when aligned in GAP Lumbar Distribution Index (OR: 1.7, 95% CI: 1.1-3.3; p=.029), and those not meeting Good Outcome criteria were more likely to deteriorate in Lumbar Distribution Index from first to final follow-up (OR: 5.8, 95% CI: 1.7-19.8; p=.005). While these patients had noticeably higher rates of failure (20.0% vs 8.9%; p=.123), patients losing alignment in LDI had significantly lower rates of reaching Clinical Improvement criteria (30.0% vs 66.7%, p=.004). Patients with higher pelvic incidence (>63°) meeting Good Outcome were more likely to be overcorrected in SVA (OR: 2.4, 95% CI: 1.1-5.2; p=.033) at first follow-up. and were less likely to be undercorrected in T1PA (OR: 0.4, [0.17-0.86]; p=.020) by final follow-up. When assessing GAP alignment, patients were more likely to meet Good Outcome when aligned in GAP Lumbar Distribution Index (OR: 3.5, 95% CI: 1.4-8.9; p=.007). CONCLUSIONS: There is a parabolic relationship between pelvic incidence and both mechanical failure and clinical improvement following correction of adult spinal deformity. Patients with lower pelvic incidence may fare better with undercorrection in age-adjusted alignment, while those with higher pelvic may necessitate proper distribution of lordosis within the lumbar spine. In addition, loss of in-construct alignment led to higher rates of mechanical failure within low pelvic incidence and less clinical improvement among those with a higher grade. Understanding of the associations this fixed parameter has with poor outcomes can aid the surgeon in strategical planning when seeking to realign adult spinal deformity. FDA Device/Drug Status: This abstract does not discuss or include any applicable devices or drugs.

Duke Scholars

Author Peter Passias Orthopaedic Surgery