Evaluating Outcomes of Spinopelvic Fixation for Patients Undergoing Long Segment Thoracolumbar Fusion with a Prior Total Hip Arthroplasty.

Abstract

Understanding the relationship between spinal fusion and its effects on relative spinopelvic alignment in patients with prior total hip arthroplasty (THA) is critical. However, limited data exist on the effects of long spinal fusions on hip alignment in patients with a prior THA. Our objective was to compare clinical outcomes and changes in hip alignment between patients undergoing long fusion to the sacrum versus to the pelvis in the setting of prior THA. Patients with a prior THA who underwent elective thoracolumbar spinal fusion starting at L2 or above were retrospectively identified. Patients were placed into one of two groups: fusion to the sacrum or pelvis. Preoperative, six-month postoperative, one-year postoperative, and delta spinopelvic and acetabular measurements were measured from standing lumbar radiographs. A total of 112 patients (55 sacral fusions, 57 pelvic fusions) were included. Patients who underwent fusion to the pelvis experienced longer length of stay (LOS) (8.31 vs. 4.21, P < 0.001) and less frequent home discharges (30.8% vs. 61.9%, P = 0.010), but fewer spinal revisions (12.3% vs. 30.9%, P = 0.030). No difference was observed in hip dislocation rates (3.51% vs. 1.82%, P = 1.000) or hip revisions (5.26% vs. 3.64%, P = 1.000) based on fusion construct. Fusion to the sacrum alone was an independent predictor of an increased spine revision rate (odds ratio: 3.56, P = 0.023). Patients in the pelvic fusion group had lower baseline lumbar lordosis (LL) (29.2 vs. 42.9, P < 0.001), six-month postoperative LL (38.7 vs. 47.3, P = 0.038), and greater 1-year ∆ pelvic incidence-lumbar lordosis (-7.98 vs. 0.21, P = 0.032). Patients with prior THA undergoing long fusion to the pelvis experienced longer LOS, more surgical complications, and lower rate of spinal revisions. Patients with instrumentation to the pelvis had lower LL preoperatively with greater changes in LL and pelvic incidence-lumbar lordosis postoperatively. No differences were observed in acetabular positioning, hip dislocations, or THA revision rates between groups.