Understanding the role of pelvic obliquity and leg length discrepancy in adult spinal deformity patients with coronal malalignment: unlocking the black box

OBJECTIVE This study had 3 objectives: 1) to describe pelvic obliquity (PO) and leg-length discrepancy (LLD) and their relationship with coronal malalignment (CM); 2) to report rates of isolated PO and PO secondary to LLD; and 3) to assess the importance of preoperative PO and LLD in postoperative complications, readmission, reoperation, and patient-reported outcomes. METHODS Patients undergoing surgery (z 6-level fusions) for adult spinal deformity at a single institution were reviewed. Variables evaluated were as follows: 1) PO, angle between the horizontal plane and a line touching bilateral iliac crests; and 2) LLD, distance from the head to the tibial plafond. Coronal vertical axis (CVA) and sagittal vertical axis measurements were collected, both from C7. The cutoff for CM was CVA > 3 cm. The Oswestry Disability Index (ODI) was collected preoperatively and at 2 years. RESULTS Of 242 patients undergoing surgery for adult spinal deformity, 90 (37.0%) had preoperative CM. Patients with preoperative CM had a higher PO (2.8 degrees +/- 3.2 degrees vs 2.0 degrees +/- 1.7 degrees, p = 0.013), a higher percentage of patients with PO > 3 degrees (35.6% vs 23.5%, p = 0.044), and higher a percentage of patients with LLD > 1 cm (21.1% vs 9.8%, p = 0.014). Whereas preoperative PO was significantly positively correlated with CVA (r = 0.26, p < 0.001) and maximum Cobb angle (r = 0.30, p < 0.001), preoperative LLD was only significantly correlated with CVA (r = 0.14, p = 0.035). A total of 12.2% of patients with CM had significant PO and LLD, defined as follows: PO z 3 degrees; LLD z 1 cm. Postoperatively, preoperative PO was significantly associated with both postoperative CM (OR 1.22, 95% CI 1.05-1.40, p = 0.008) and postoperative CVA (beta = 0.14, 95% CI 0.06-0.22, p < 0.001). A higher preoperative PO was independently associated with postoperative complications after multivariate logistic regression (OR 1.24, 95% CI 1.05-1.45, p = 0.010); however, 2-year ODI scores were not. Preoperative LLD had no significant relationship with postoperative CM, CVA, ODI, or complications. CONCLUSIONS A PO z 3 degrees or LLD z 1 cm was seen in 44.1% of patients with preoperative CM and in 23.5% of patients with normal coronal alignment. Preoperative PO was significantly associated with preoperative CVA and maximum Cobb angle, whereas preoperative LLD was only associated with preoperative CVA. The direction of PO and LLD showed no consistent pattern with CVA. Preoperative PO was independently associated with complications but not with 2-year ODI scores.